{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.727803672712978,
  "eval_steps": 3000,
  "global_step": 51000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.25639886856079,
      "epoch": 0.0005615769079575448,
      "grad_norm": 3.71875,
      "learning_rate": 2e-06,
      "loss": 11.7409,
      "mean_token_accuracy": 0.0026910859567578884,
      "num_tokens": 11659.0,
      "step": 5
    },
    {
      "entropy": 10.252698040008545,
      "epoch": 0.0011231538159150896,
      "grad_norm": 4.25,
      "learning_rate": 4.5e-06,
      "loss": 11.7094,
      "mean_token_accuracy": 0.0012989068985916675,
      "num_tokens": 21794.0,
      "step": 10
    },
    {
      "entropy": 10.256062889099121,
      "epoch": 0.0016847307238726343,
      "grad_norm": 4.09375,
      "learning_rate": 7e-06,
      "loss": 11.6893,
      "mean_token_accuracy": 0.002295289159519598,
      "num_tokens": 33206.0,
      "step": 15
    },
    {
      "entropy": 10.28339958190918,
      "epoch": 0.0022463076318301792,
      "grad_norm": 3.96875,
      "learning_rate": 9.5e-06,
      "loss": 11.5471,
      "mean_token_accuracy": 0.003405291523085907,
      "num_tokens": 44487.0,
      "step": 20
    },
    {
      "entropy": 10.283167362213135,
      "epoch": 0.002807884539787724,
      "grad_norm": 3.625,
      "learning_rate": 1.2e-05,
      "loss": 11.3861,
      "mean_token_accuracy": 0.006172553007490933,
      "num_tokens": 56062.0,
      "step": 25
    },
    {
      "entropy": 10.233351135253907,
      "epoch": 0.0033694614477452686,
      "grad_norm": 3.59375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 11.2207,
      "mean_token_accuracy": 0.022563868667930365,
      "num_tokens": 67453.0,
      "step": 30
    },
    {
      "entropy": 10.17716064453125,
      "epoch": 0.003931038355702813,
      "grad_norm": 3.25,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.9781,
      "mean_token_accuracy": 0.02911745272576809,
      "num_tokens": 78527.0,
      "step": 35
    },
    {
      "entropy": 10.325019550323486,
      "epoch": 0.0044926152636603585,
      "grad_norm": 2.640625,
      "learning_rate": 1.95e-05,
      "loss": 10.6651,
      "mean_token_accuracy": 0.030208175629377367,
      "num_tokens": 89267.0,
      "step": 40
    },
    {
      "entropy": 10.383963871002198,
      "epoch": 0.005054192171617903,
      "grad_norm": 2.59375,
      "learning_rate": 2.2e-05,
      "loss": 10.5824,
      "mean_token_accuracy": 0.02886309754103422,
      "num_tokens": 102035.0,
      "step": 45
    },
    {
      "entropy": 10.449808597564697,
      "epoch": 0.005615769079575448,
      "grad_norm": 2.546875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.4605,
      "mean_token_accuracy": 0.038186770118772986,
      "num_tokens": 112734.0,
      "step": 50
    },
    {
      "entropy": 10.461854267120362,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.4375,
      "learning_rate": 2.7e-05,
      "loss": 10.3849,
      "mean_token_accuracy": 0.040889688022434714,
      "num_tokens": 123854.0,
      "step": 55
    },
    {
      "entropy": 10.472981071472168,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.390625,
      "learning_rate": 2.95e-05,
      "loss": 10.295,
      "mean_token_accuracy": 0.037749376334249976,
      "num_tokens": 133444.0,
      "step": 60
    },
    {
      "entropy": 10.491286468505859,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.515625,
      "learning_rate": 3.2e-05,
      "loss": 10.2382,
      "mean_token_accuracy": 0.03989502489566803,
      "num_tokens": 144678.0,
      "step": 65
    },
    {
      "entropy": 10.506498050689697,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.546875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 10.1488,
      "mean_token_accuracy": 0.03764319252222777,
      "num_tokens": 155866.0,
      "step": 70
    },
    {
      "entropy": 10.518252658843995,
      "epoch": 0.008423653619363172,
      "grad_norm": 2.4375,
      "learning_rate": 3.7e-05,
      "loss": 10.0531,
      "mean_token_accuracy": 0.038073062710464,
      "num_tokens": 166965.0,
      "step": 75
    },
    {
      "entropy": 10.528139972686768,
      "epoch": 0.008985230527320717,
      "grad_norm": 2.34375,
      "learning_rate": 3.95e-05,
      "loss": 9.9445,
      "mean_token_accuracy": 0.039117959514260295,
      "num_tokens": 177423.0,
      "step": 80
    },
    {
      "entropy": 10.524243068695068,
      "epoch": 0.00954680743527826,
      "grad_norm": 2.375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.8407,
      "mean_token_accuracy": 0.044676617905497554,
      "num_tokens": 188687.0,
      "step": 85
    },
    {
      "entropy": 10.510469150543212,
      "epoch": 0.010108384343235806,
      "grad_norm": 2.328125,
      "learning_rate": 4.45e-05,
      "loss": 9.7592,
      "mean_token_accuracy": 0.038761510699987414,
      "num_tokens": 200173.0,
      "step": 90
    },
    {
      "entropy": 10.509880542755127,
      "epoch": 0.010669961251193351,
      "grad_norm": 2.453125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.6288,
      "mean_token_accuracy": 0.042973919026553634,
      "num_tokens": 210877.0,
      "step": 95
    },
    {
      "entropy": 10.515023994445801,
      "epoch": 0.011231538159150897,
      "grad_norm": 2.40625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.5036,
      "mean_token_accuracy": 0.039228996820747854,
      "num_tokens": 221770.0,
      "step": 100
    },
    {
      "entropy": 10.511959838867188,
      "epoch": 0.01179311506710844,
      "grad_norm": 2.359375,
      "learning_rate": 5.2e-05,
      "loss": 9.3655,
      "mean_token_accuracy": 0.041880004480481146,
      "num_tokens": 232140.0,
      "step": 105
    },
    {
      "entropy": 10.486880874633789,
      "epoch": 0.012354691975065986,
      "grad_norm": 2.234375,
      "learning_rate": 5.45e-05,
      "loss": 9.2756,
      "mean_token_accuracy": 0.03760720025748014,
      "num_tokens": 244186.0,
      "step": 110
    },
    {
      "entropy": 10.462724304199218,
      "epoch": 0.012916268883023531,
      "grad_norm": 2.25,
      "learning_rate": 5.7e-05,
      "loss": 9.084,
      "mean_token_accuracy": 0.038566050492227076,
      "num_tokens": 254805.0,
      "step": 115
    },
    {
      "entropy": 10.42501459121704,
      "epoch": 0.013477845790981074,
      "grad_norm": 2.34375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8662,
      "mean_token_accuracy": 0.041522559337317945,
      "num_tokens": 265055.0,
      "step": 120
    },
    {
      "entropy": 10.37605791091919,
      "epoch": 0.01403942269893862,
      "grad_norm": 2.21875,
      "learning_rate": 6.2e-05,
      "loss": 8.7259,
      "mean_token_accuracy": 0.035992196202278136,
      "num_tokens": 275850.0,
      "step": 125
    },
    {
      "entropy": 10.301450824737548,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.9609375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5828,
      "mean_token_accuracy": 0.03876302167773247,
      "num_tokens": 286802.0,
      "step": 130
    },
    {
      "entropy": 10.22424955368042,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.984375,
      "learning_rate": 6.7e-05,
      "loss": 8.4842,
      "mean_token_accuracy": 0.03725951910018921,
      "num_tokens": 297070.0,
      "step": 135
    },
    {
      "entropy": 10.137630462646484,
      "epoch": 0.015724153422811252,
      "grad_norm": 1.890625,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.3182,
      "mean_token_accuracy": 0.04086887463927269,
      "num_tokens": 309617.0,
      "step": 140
    },
    {
      "entropy": 9.952799987792968,
      "epoch": 0.0162857303307688,
      "grad_norm": 1.734375,
      "learning_rate": 7.2e-05,
      "loss": 8.1687,
      "mean_token_accuracy": 0.04170650318264961,
      "num_tokens": 321310.0,
      "step": 145
    },
    {
      "entropy": 9.707593727111817,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.484375,
      "learning_rate": 7.45e-05,
      "loss": 7.9867,
      "mean_token_accuracy": 0.04252769723534584,
      "num_tokens": 332243.0,
      "step": 150
    },
    {
      "entropy": 9.44668788909912,
      "epoch": 0.017408884146683887,
      "grad_norm": 1.15625,
      "learning_rate": 7.7e-05,
      "loss": 7.8717,
      "mean_token_accuracy": 0.03686592467129231,
      "num_tokens": 342847.0,
      "step": 155
    },
    {
      "entropy": 9.195470142364503,
      "epoch": 0.017970461054641434,
      "grad_norm": 1.140625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.7169,
      "mean_token_accuracy": 0.0400078721344471,
      "num_tokens": 353570.0,
      "step": 160
    },
    {
      "entropy": 8.776480579376221,
      "epoch": 0.018532037962598977,
      "grad_norm": 0.9453125,
      "learning_rate": 8.2e-05,
      "loss": 7.6303,
      "mean_token_accuracy": 0.04083114080131054,
      "num_tokens": 364330.0,
      "step": 165
    },
    {
      "entropy": 8.479571533203124,
      "epoch": 0.01909361487055652,
      "grad_norm": 0.98828125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.6035,
      "mean_token_accuracy": 0.039075618609786034,
      "num_tokens": 374930.0,
      "step": 170
    },
    {
      "entropy": 8.282562637329102,
      "epoch": 0.019655191778514068,
      "grad_norm": 0.9453125,
      "learning_rate": 8.7e-05,
      "loss": 7.5491,
      "mean_token_accuracy": 0.03750791251659393,
      "num_tokens": 384690.0,
      "step": 175
    },
    {
      "entropy": 8.006285572052002,
      "epoch": 0.020216768686471612,
      "grad_norm": 0.99609375,
      "learning_rate": 8.95e-05,
      "loss": 7.5439,
      "mean_token_accuracy": 0.038402245566248897,
      "num_tokens": 395753.0,
      "step": 180
    },
    {
      "entropy": 7.9631373405456545,
      "epoch": 0.020778345594429155,
      "grad_norm": 0.98046875,
      "learning_rate": 9.2e-05,
      "loss": 7.493,
      "mean_token_accuracy": 0.0437284167855978,
      "num_tokens": 407049.0,
      "step": 185
    },
    {
      "entropy": 7.781239318847656,
      "epoch": 0.021339922502386702,
      "grad_norm": 0.88671875,
      "learning_rate": 9.45e-05,
      "loss": 7.4957,
      "mean_token_accuracy": 0.04678881764411926,
      "num_tokens": 417942.0,
      "step": 190
    },
    {
      "entropy": 7.785053539276123,
      "epoch": 0.021901499410344246,
      "grad_norm": 0.94921875,
      "learning_rate": 9.7e-05,
      "loss": 7.5136,
      "mean_token_accuracy": 0.04727039746940136,
      "num_tokens": 428151.0,
      "step": 195
    },
    {
      "entropy": 7.740858125686645,
      "epoch": 0.022463076318301793,
      "grad_norm": 0.828125,
      "learning_rate": 9.95e-05,
      "loss": 7.5696,
      "mean_token_accuracy": 0.04193144179880619,
      "num_tokens": 439837.0,
      "step": 200
    },
    {
      "entropy": 7.758722639083862,
      "epoch": 0.023024653226259337,
      "grad_norm": 0.7578125,
      "learning_rate": 0.000102,
      "loss": 7.4302,
      "mean_token_accuracy": 0.045322371646761894,
      "num_tokens": 450357.0,
      "step": 205
    },
    {
      "entropy": 7.715395164489746,
      "epoch": 0.02358623013421688,
      "grad_norm": 0.859375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4423,
      "mean_token_accuracy": 0.058253410086035726,
      "num_tokens": 461174.0,
      "step": 210
    },
    {
      "entropy": 7.59510555267334,
      "epoch": 0.024147807042174427,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000107,
      "loss": 7.5059,
      "mean_token_accuracy": 0.058244950324296954,
      "num_tokens": 472785.0,
      "step": 215
    },
    {
      "entropy": 7.6779868602752686,
      "epoch": 0.02470938395013197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001095,
      "loss": 7.458,
      "mean_token_accuracy": 0.06584881767630577,
      "num_tokens": 483248.0,
      "step": 220
    },
    {
      "entropy": 7.6262555599212645,
      "epoch": 0.025270960858089515,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000112,
      "loss": 7.3474,
      "mean_token_accuracy": 0.06667502745985984,
      "num_tokens": 494690.0,
      "step": 225
    },
    {
      "entropy": 7.683849191665649,
      "epoch": 0.025832537766047062,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001145,
      "loss": 7.5293,
      "mean_token_accuracy": 0.0678394366055727,
      "num_tokens": 506192.0,
      "step": 230
    },
    {
      "entropy": 7.675125026702881,
      "epoch": 0.026394114674004605,
      "grad_norm": 0.875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.3906,
      "mean_token_accuracy": 0.07262160144746303,
      "num_tokens": 517152.0,
      "step": 235
    },
    {
      "entropy": 7.534460210800171,
      "epoch": 0.02695569158196215,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.3269,
      "mean_token_accuracy": 0.0788414567708969,
      "num_tokens": 527836.0,
      "step": 240
    },
    {
      "entropy": 7.556740760803223,
      "epoch": 0.027517268489919696,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000122,
      "loss": 7.4559,
      "mean_token_accuracy": 0.07313848771154881,
      "num_tokens": 539491.0,
      "step": 245
    },
    {
      "entropy": 7.651741409301758,
      "epoch": 0.02807884539787724,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0001245,
      "loss": 7.4231,
      "mean_token_accuracy": 0.06984101124107837,
      "num_tokens": 550816.0,
      "step": 250
    },
    {
      "entropy": 7.559566640853882,
      "epoch": 0.028640422305834783,
      "grad_norm": 0.765625,
      "learning_rate": 0.000127,
      "loss": 7.4417,
      "mean_token_accuracy": 0.07528160624206066,
      "num_tokens": 561566.0,
      "step": 255
    },
    {
      "entropy": 7.60768313407898,
      "epoch": 0.02920199921379233,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0001295,
      "loss": 7.3847,
      "mean_token_accuracy": 0.07447496987879276,
      "num_tokens": 573032.0,
      "step": 260
    },
    {
      "entropy": 7.602000761032104,
      "epoch": 0.029763576121749874,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000132,
      "loss": 7.3083,
      "mean_token_accuracy": 0.07689321786165237,
      "num_tokens": 584140.0,
      "step": 265
    },
    {
      "entropy": 7.546748113632202,
      "epoch": 0.030325153029707418,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.3863,
      "mean_token_accuracy": 0.0756129652261734,
      "num_tokens": 595097.0,
      "step": 270
    },
    {
      "entropy": 7.626759958267212,
      "epoch": 0.030886729937664965,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3776,
      "mean_token_accuracy": 0.07868644669651985,
      "num_tokens": 606797.0,
      "step": 275
    },
    {
      "entropy": 7.603859090805054,
      "epoch": 0.031448306845622505,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001395,
      "loss": 7.3381,
      "mean_token_accuracy": 0.0748847633600235,
      "num_tokens": 618187.0,
      "step": 280
    },
    {
      "entropy": 7.569405174255371,
      "epoch": 0.032009883753580055,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.2915,
      "mean_token_accuracy": 0.08101793453097343,
      "num_tokens": 629829.0,
      "step": 285
    },
    {
      "entropy": 7.554937267303467,
      "epoch": 0.0325714606615376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0001445,
      "loss": 7.2766,
      "mean_token_accuracy": 0.08518879413604737,
      "num_tokens": 640914.0,
      "step": 290
    },
    {
      "entropy": 7.543597173690796,
      "epoch": 0.03313303756949514,
      "grad_norm": 0.9375,
      "learning_rate": 0.000147,
      "loss": 7.2827,
      "mean_token_accuracy": 0.0838129960000515,
      "num_tokens": 653057.0,
      "step": 295
    },
    {
      "entropy": 7.536286020278931,
      "epoch": 0.033694614477452686,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0001495,
      "loss": 7.2208,
      "mean_token_accuracy": 0.08279290944337844,
      "num_tokens": 663437.0,
      "step": 300
    },
    {
      "entropy": 7.500700044631958,
      "epoch": 0.03425619138541023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000152,
      "loss": 7.2719,
      "mean_token_accuracy": 0.08277729079127312,
      "num_tokens": 675416.0,
      "step": 305
    },
    {
      "entropy": 7.403319644927978,
      "epoch": 0.03481776829336777,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0849,
      "mean_token_accuracy": 0.08483515977859497,
      "num_tokens": 686633.0,
      "step": 310
    },
    {
      "entropy": 7.422772455215454,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.015625,
      "learning_rate": 0.000157,
      "loss": 7.173,
      "mean_token_accuracy": 0.08649655506014824,
      "num_tokens": 698065.0,
      "step": 315
    },
    {
      "entropy": 7.5332183837890625,
      "epoch": 0.03594092210928287,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0001595,
      "loss": 7.1756,
      "mean_token_accuracy": 0.08641982153058052,
      "num_tokens": 708652.0,
      "step": 320
    },
    {
      "entropy": 7.479076910018921,
      "epoch": 0.03650249901724041,
      "grad_norm": 0.9375,
      "learning_rate": 0.000162,
      "loss": 7.2977,
      "mean_token_accuracy": 0.08264525830745698,
      "num_tokens": 720199.0,
      "step": 325
    },
    {
      "entropy": 7.496834087371826,
      "epoch": 0.037064075925197955,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.2999,
      "mean_token_accuracy": 0.08251902386546135,
      "num_tokens": 732964.0,
      "step": 330
    },
    {
      "entropy": 7.395641613006592,
      "epoch": 0.0376256528331555,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0579,
      "mean_token_accuracy": 0.08676571622490883,
      "num_tokens": 743528.0,
      "step": 335
    },
    {
      "entropy": 7.39327449798584,
      "epoch": 0.03818722974111304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.1187,
      "mean_token_accuracy": 0.0880549892783165,
      "num_tokens": 754365.0,
      "step": 340
    },
    {
      "entropy": 7.318950414657593,
      "epoch": 0.03874880664907059,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0332,
      "mean_token_accuracy": 0.08635925874114037,
      "num_tokens": 766094.0,
      "step": 345
    },
    {
      "entropy": 7.353262042999267,
      "epoch": 0.039310383557028136,
      "grad_norm": 0.984375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0952,
      "mean_token_accuracy": 0.09182592332363129,
      "num_tokens": 777275.0,
      "step": 350
    },
    {
      "entropy": 7.3257616519927975,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000177,
      "loss": 7.136,
      "mean_token_accuracy": 0.08920007646083832,
      "num_tokens": 789436.0,
      "step": 355
    },
    {
      "entropy": 7.399164628982544,
      "epoch": 0.040433537372943223,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001795,
      "loss": 7.1632,
      "mean_token_accuracy": 0.08764473721385002,
      "num_tokens": 800164.0,
      "step": 360
    },
    {
      "entropy": 7.224786233901978,
      "epoch": 0.04099511428090077,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000182,
      "loss": 7.0352,
      "mean_token_accuracy": 0.09125346541404725,
      "num_tokens": 811797.0,
      "step": 365
    },
    {
      "entropy": 7.286064577102661,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001845,
      "loss": 7.0431,
      "mean_token_accuracy": 0.09055957570672035,
      "num_tokens": 824158.0,
      "step": 370
    },
    {
      "entropy": 7.305629539489746,
      "epoch": 0.04211826809681586,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000187,
      "loss": 7.0457,
      "mean_token_accuracy": 0.09202495589852333,
      "num_tokens": 835803.0,
      "step": 375
    },
    {
      "entropy": 7.371310424804688,
      "epoch": 0.042679845004773405,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0001895,
      "loss": 7.1098,
      "mean_token_accuracy": 0.08957164734601974,
      "num_tokens": 848106.0,
      "step": 380
    },
    {
      "entropy": 7.336284589767456,
      "epoch": 0.04324142191273095,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000192,
      "loss": 7.063,
      "mean_token_accuracy": 0.09114922881126404,
      "num_tokens": 859101.0,
      "step": 385
    },
    {
      "entropy": 7.226832294464112,
      "epoch": 0.04380299882068849,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001945,
      "loss": 6.9601,
      "mean_token_accuracy": 0.0964093454182148,
      "num_tokens": 870309.0,
      "step": 390
    },
    {
      "entropy": 7.222394752502441,
      "epoch": 0.044364575728646036,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0239,
      "mean_token_accuracy": 0.10131923705339432,
      "num_tokens": 881515.0,
      "step": 395
    },
    {
      "entropy": 7.244077253341675,
      "epoch": 0.044926152636603586,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9904,
      "mean_token_accuracy": 0.09794511124491692,
      "num_tokens": 892129.0,
      "step": 400
    },
    {
      "entropy": 7.264095592498779,
      "epoch": 0.04548772954456113,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000202,
      "loss": 7.0037,
      "mean_token_accuracy": 0.09216101467609406,
      "num_tokens": 903186.0,
      "step": 405
    },
    {
      "entropy": 7.212151622772216,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9559,
      "mean_token_accuracy": 0.10168264731764794,
      "num_tokens": 913471.0,
      "step": 410
    },
    {
      "entropy": 7.21423306465149,
      "epoch": 0.04661088336047622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000207,
      "loss": 6.9665,
      "mean_token_accuracy": 0.09732877537608146,
      "num_tokens": 924283.0,
      "step": 415
    },
    {
      "entropy": 7.138846921920776,
      "epoch": 0.04717246026843376,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0002095,
      "loss": 6.9948,
      "mean_token_accuracy": 0.09952060505747795,
      "num_tokens": 934975.0,
      "step": 420
    },
    {
      "entropy": 7.175388956069947,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000212,
      "loss": 6.8805,
      "mean_token_accuracy": 0.10059212520718575,
      "num_tokens": 945587.0,
      "step": 425
    },
    {
      "entropy": 7.089182806015015,
      "epoch": 0.048295614084348855,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0002145,
      "loss": 6.9372,
      "mean_token_accuracy": 0.09849796593189239,
      "num_tokens": 955935.0,
      "step": 430
    },
    {
      "entropy": 7.184536552429199,
      "epoch": 0.0488571909923064,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.889,
      "mean_token_accuracy": 0.10151714682579041,
      "num_tokens": 967960.0,
      "step": 435
    },
    {
      "entropy": 7.127185440063476,
      "epoch": 0.04941876790026394,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002195,
      "loss": 6.9301,
      "mean_token_accuracy": 0.10072232410311699,
      "num_tokens": 979680.0,
      "step": 440
    },
    {
      "entropy": 7.1458930492401125,
      "epoch": 0.049980344808221486,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000222,
      "loss": 7.0148,
      "mean_token_accuracy": 0.1004769742488861,
      "num_tokens": 990600.0,
      "step": 445
    },
    {
      "entropy": 7.197234964370727,
      "epoch": 0.05054192171617903,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002245,
      "loss": 7.0244,
      "mean_token_accuracy": 0.10357746407389641,
      "num_tokens": 1002331.0,
      "step": 450
    },
    {
      "entropy": 6.973642158508301,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8315,
      "mean_token_accuracy": 0.1050386182963848,
      "num_tokens": 1013473.0,
      "step": 455
    },
    {
      "entropy": 7.1875935077667235,
      "epoch": 0.051665075532094124,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.9257,
      "mean_token_accuracy": 0.09776141494512558,
      "num_tokens": 1024434.0,
      "step": 460
    },
    {
      "entropy": 7.147069692611694,
      "epoch": 0.05222665244005167,
      "grad_norm": 0.84375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.9539,
      "mean_token_accuracy": 0.09493176713585853,
      "num_tokens": 1035452.0,
      "step": 465
    },
    {
      "entropy": 7.040191555023194,
      "epoch": 0.05278822934800921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8166,
      "mean_token_accuracy": 0.10042703226208687,
      "num_tokens": 1046213.0,
      "step": 470
    },
    {
      "entropy": 7.027799129486084,
      "epoch": 0.053349806255966754,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000237,
      "loss": 6.8431,
      "mean_token_accuracy": 0.1014677569270134,
      "num_tokens": 1057469.0,
      "step": 475
    },
    {
      "entropy": 7.141836309432984,
      "epoch": 0.0539113831639243,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002395,
      "loss": 6.9903,
      "mean_token_accuracy": 0.10095212161540985,
      "num_tokens": 1069543.0,
      "step": 480
    },
    {
      "entropy": 6.945070791244507,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.015625,
      "learning_rate": 0.000242,
      "loss": 6.8419,
      "mean_token_accuracy": 0.10394882559776306,
      "num_tokens": 1080476.0,
      "step": 485
    },
    {
      "entropy": 7.250262975692749,
      "epoch": 0.05503453697983939,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002445,
      "loss": 6.9611,
      "mean_token_accuracy": 0.10339747741818428,
      "num_tokens": 1091700.0,
      "step": 490
    },
    {
      "entropy": 6.982399654388428,
      "epoch": 0.055596113887796936,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000247,
      "loss": 6.8471,
      "mean_token_accuracy": 0.0988765001296997,
      "num_tokens": 1103056.0,
      "step": 495
    },
    {
      "entropy": 7.01871132850647,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002495,
      "loss": 6.9121,
      "mean_token_accuracy": 0.1074085734784603,
      "num_tokens": 1114836.0,
      "step": 500
    },
    {
      "entropy": 7.023473834991455,
      "epoch": 0.05671926770371202,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000252,
      "loss": 6.8057,
      "mean_token_accuracy": 0.10938784778118134,
      "num_tokens": 1125993.0,
      "step": 505
    },
    {
      "entropy": 7.023718929290771,
      "epoch": 0.05728084461166957,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002545,
      "loss": 6.8839,
      "mean_token_accuracy": 0.11009863540530204,
      "num_tokens": 1136719.0,
      "step": 510
    },
    {
      "entropy": 6.948973846435547,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000257,
      "loss": 6.7549,
      "mean_token_accuracy": 0.10685163512825965,
      "num_tokens": 1146947.0,
      "step": 515
    },
    {
      "entropy": 7.0727332592010494,
      "epoch": 0.05840399842758466,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002595,
      "loss": 6.8908,
      "mean_token_accuracy": 0.10683172717690467,
      "num_tokens": 1157974.0,
      "step": 520
    },
    {
      "entropy": 6.95055775642395,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.046875,
      "learning_rate": 0.000262,
      "loss": 6.8077,
      "mean_token_accuracy": 0.10525572448968887,
      "num_tokens": 1168573.0,
      "step": 525
    },
    {
      "entropy": 7.000631856918335,
      "epoch": 0.05952715224349975,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7251,
      "mean_token_accuracy": 0.10736186131834984,
      "num_tokens": 1178446.0,
      "step": 530
    },
    {
      "entropy": 6.959245443344116,
      "epoch": 0.06008872915145729,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7478,
      "mean_token_accuracy": 0.10534990206360817,
      "num_tokens": 1190558.0,
      "step": 535
    },
    {
      "entropy": 6.902206563949585,
      "epoch": 0.060650306059414835,
      "grad_norm": 0.875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.9124,
      "mean_token_accuracy": 0.10460811406373978,
      "num_tokens": 1201929.0,
      "step": 540
    },
    {
      "entropy": 6.9541998386383055,
      "epoch": 0.06121188296737238,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7425,
      "mean_token_accuracy": 0.11181432232260705,
      "num_tokens": 1213155.0,
      "step": 545
    },
    {
      "entropy": 7.080395078659057,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002745,
      "loss": 6.8042,
      "mean_token_accuracy": 0.10288820937275886,
      "num_tokens": 1223961.0,
      "step": 550
    },
    {
      "entropy": 6.755504512786866,
      "epoch": 0.06233503678328747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000277,
      "loss": 6.7008,
      "mean_token_accuracy": 0.11182279288768768,
      "num_tokens": 1233888.0,
      "step": 555
    },
    {
      "entropy": 6.958348941802979,
      "epoch": 0.06289661369124501,
      "grad_norm": 0.84375,
      "learning_rate": 0.0002795,
      "loss": 6.8077,
      "mean_token_accuracy": 0.1030250370502472,
      "num_tokens": 1244405.0,
      "step": 560
    },
    {
      "entropy": 7.0104207515716555,
      "epoch": 0.06345819059920256,
      "grad_norm": 0.796875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7471,
      "mean_token_accuracy": 0.1139664702117443,
      "num_tokens": 1257099.0,
      "step": 565
    },
    {
      "entropy": 6.860793161392212,
      "epoch": 0.06401976750716011,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002845,
      "loss": 6.8584,
      "mean_token_accuracy": 0.11190064251422882,
      "num_tokens": 1267965.0,
      "step": 570
    },
    {
      "entropy": 6.909835290908814,
      "epoch": 0.06458134441511765,
      "grad_norm": 0.953125,
      "learning_rate": 0.000287,
      "loss": 6.747,
      "mean_token_accuracy": 0.11047227904200554,
      "num_tokens": 1280075.0,
      "step": 575
    },
    {
      "entropy": 6.996085596084595,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002895,
      "loss": 6.7767,
      "mean_token_accuracy": 0.11513174846768379,
      "num_tokens": 1290574.0,
      "step": 580
    },
    {
      "entropy": 6.839344120025634,
      "epoch": 0.06570449823103273,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000292,
      "loss": 6.7449,
      "mean_token_accuracy": 0.11083959862589836,
      "num_tokens": 1302680.0,
      "step": 585
    },
    {
      "entropy": 6.863824939727783,
      "epoch": 0.06626607513899029,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002945,
      "loss": 6.727,
      "mean_token_accuracy": 0.11484655365347862,
      "num_tokens": 1314577.0,
      "step": 590
    },
    {
      "entropy": 6.9280845642089846,
      "epoch": 0.06682765204694784,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000297,
      "loss": 6.8363,
      "mean_token_accuracy": 0.11602571532130242,
      "num_tokens": 1326560.0,
      "step": 595
    },
    {
      "entropy": 7.001092004776001,
      "epoch": 0.06738922895490537,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002995,
      "loss": 6.9057,
      "mean_token_accuracy": 0.10357883349061012,
      "num_tokens": 1339105.0,
      "step": 600
    },
    {
      "entropy": 6.722957897186279,
      "epoch": 0.06795080586286292,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000302,
      "loss": 6.5765,
      "mean_token_accuracy": 0.11965182796120644,
      "num_tokens": 1349914.0,
      "step": 605
    },
    {
      "entropy": 6.931371068954467,
      "epoch": 0.06851238277082046,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003045,
      "loss": 6.7924,
      "mean_token_accuracy": 0.11453562378883361,
      "num_tokens": 1361700.0,
      "step": 610
    },
    {
      "entropy": 6.858430242538452,
      "epoch": 0.06907395967877801,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000307,
      "loss": 6.8079,
      "mean_token_accuracy": 0.1133706197142601,
      "num_tokens": 1373103.0,
      "step": 615
    },
    {
      "entropy": 6.822653102874756,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003095,
      "loss": 6.6143,
      "mean_token_accuracy": 0.11548932567238808,
      "num_tokens": 1383742.0,
      "step": 620
    },
    {
      "entropy": 6.799988889694214,
      "epoch": 0.0701971134946931,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000312,
      "loss": 6.7658,
      "mean_token_accuracy": 0.11335846111178398,
      "num_tokens": 1396656.0,
      "step": 625
    },
    {
      "entropy": 6.776160717010498,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.0,
      "learning_rate": 0.0003145,
      "loss": 6.6291,
      "mean_token_accuracy": 0.12655726373195647,
      "num_tokens": 1406709.0,
      "step": 630
    },
    {
      "entropy": 6.81230092048645,
      "epoch": 0.07132026731060818,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000317,
      "loss": 6.677,
      "mean_token_accuracy": 0.120645322650671,
      "num_tokens": 1417793.0,
      "step": 635
    },
    {
      "entropy": 6.884833669662475,
      "epoch": 0.07188184421856574,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003195,
      "loss": 6.7582,
      "mean_token_accuracy": 0.11574307680130005,
      "num_tokens": 1429093.0,
      "step": 640
    },
    {
      "entropy": 6.699341487884522,
      "epoch": 0.07244342112652327,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000322,
      "loss": 6.6679,
      "mean_token_accuracy": 0.12265671268105507,
      "num_tokens": 1440907.0,
      "step": 645
    },
    {
      "entropy": 6.890548658370972,
      "epoch": 0.07300499803448082,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.7498,
      "mean_token_accuracy": 0.1140442967414856,
      "num_tokens": 1452089.0,
      "step": 650
    },
    {
      "entropy": 6.8591334342956545,
      "epoch": 0.07356657494243837,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.7635,
      "mean_token_accuracy": 0.11165911853313445,
      "num_tokens": 1464378.0,
      "step": 655
    },
    {
      "entropy": 6.80465030670166,
      "epoch": 0.07412815185039591,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.687,
      "mean_token_accuracy": 0.11683566421270371,
      "num_tokens": 1476678.0,
      "step": 660
    },
    {
      "entropy": 6.709757089614868,
      "epoch": 0.07468972875835346,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5363,
      "mean_token_accuracy": 0.121712426841259,
      "num_tokens": 1486808.0,
      "step": 665
    },
    {
      "entropy": 6.838316631317139,
      "epoch": 0.075251305666311,
      "grad_norm": 0.859375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.8144,
      "mean_token_accuracy": 0.1120360791683197,
      "num_tokens": 1499281.0,
      "step": 670
    },
    {
      "entropy": 6.873385381698609,
      "epoch": 0.07581288257426855,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000337,
      "loss": 6.7205,
      "mean_token_accuracy": 0.11793086603283882,
      "num_tokens": 1510202.0,
      "step": 675
    },
    {
      "entropy": 6.792718982696533,
      "epoch": 0.07637445948222608,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0003395,
      "loss": 6.5831,
      "mean_token_accuracy": 0.12022969275712966,
      "num_tokens": 1521492.0,
      "step": 680
    },
    {
      "entropy": 6.644273567199707,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.046875,
      "learning_rate": 0.000342,
      "loss": 6.6175,
      "mean_token_accuracy": 0.12695625126361848,
      "num_tokens": 1532856.0,
      "step": 685
    },
    {
      "entropy": 6.711900901794434,
      "epoch": 0.07749761329814119,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6209,
      "mean_token_accuracy": 0.1197846882045269,
      "num_tokens": 1543408.0,
      "step": 690
    },
    {
      "entropy": 6.7214684009552,
      "epoch": 0.07805919020609872,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000347,
      "loss": 6.6278,
      "mean_token_accuracy": 0.11783769056200981,
      "num_tokens": 1555474.0,
      "step": 695
    },
    {
      "entropy": 6.796803903579712,
      "epoch": 0.07862076711405627,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003495,
      "loss": 6.7268,
      "mean_token_accuracy": 0.1210656888782978,
      "num_tokens": 1567484.0,
      "step": 700
    },
    {
      "entropy": 6.696943473815918,
      "epoch": 0.07918234402201381,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000352,
      "loss": 6.5718,
      "mean_token_accuracy": 0.1262438789010048,
      "num_tokens": 1578204.0,
      "step": 705
    },
    {
      "entropy": 6.750957107543945,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003545,
      "loss": 6.5652,
      "mean_token_accuracy": 0.12253414019942284,
      "num_tokens": 1589012.0,
      "step": 710
    },
    {
      "entropy": 6.546931219100952,
      "epoch": 0.08030549783792891,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000357,
      "loss": 6.5747,
      "mean_token_accuracy": 0.12381779551506042,
      "num_tokens": 1601099.0,
      "step": 715
    },
    {
      "entropy": 6.817351818084717,
      "epoch": 0.08086707474588645,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003595,
      "loss": 6.7402,
      "mean_token_accuracy": 0.11719322726130485,
      "num_tokens": 1612814.0,
      "step": 720
    },
    {
      "entropy": 6.713132858276367,
      "epoch": 0.081428651653844,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000362,
      "loss": 6.6325,
      "mean_token_accuracy": 0.11970261111855507,
      "num_tokens": 1624953.0,
      "step": 725
    },
    {
      "entropy": 6.730032444000244,
      "epoch": 0.08199022856180153,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0003645,
      "loss": 6.6488,
      "mean_token_accuracy": 0.12532211914658548,
      "num_tokens": 1635673.0,
      "step": 730
    },
    {
      "entropy": 6.764851093292236,
      "epoch": 0.08255180546975908,
      "grad_norm": 0.875,
      "learning_rate": 0.000367,
      "loss": 6.5684,
      "mean_token_accuracy": 0.1264908105134964,
      "num_tokens": 1647282.0,
      "step": 735
    },
    {
      "entropy": 6.644136762619018,
      "epoch": 0.08311338237771662,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003695,
      "loss": 6.576,
      "mean_token_accuracy": 0.12845728248357774,
      "num_tokens": 1658426.0,
      "step": 740
    },
    {
      "entropy": 6.66550931930542,
      "epoch": 0.08367495928567417,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000372,
      "loss": 6.6157,
      "mean_token_accuracy": 0.1237633615732193,
      "num_tokens": 1669958.0,
      "step": 745
    },
    {
      "entropy": 6.739824438095093,
      "epoch": 0.08423653619363172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003745,
      "loss": 6.6554,
      "mean_token_accuracy": 0.12233619689941407,
      "num_tokens": 1682409.0,
      "step": 750
    },
    {
      "entropy": 6.677445888519287,
      "epoch": 0.08479811310158926,
      "grad_norm": 0.96875,
      "learning_rate": 0.000377,
      "loss": 6.5781,
      "mean_token_accuracy": 0.13024282306432725,
      "num_tokens": 1692409.0,
      "step": 755
    },
    {
      "entropy": 6.733970499038696,
      "epoch": 0.08535969000954681,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003795,
      "loss": 6.6821,
      "mean_token_accuracy": 0.12274667993187904,
      "num_tokens": 1705018.0,
      "step": 760
    },
    {
      "entropy": 6.672050523757934,
      "epoch": 0.08592126691750435,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000382,
      "loss": 6.6111,
      "mean_token_accuracy": 0.12385557517409325,
      "num_tokens": 1715625.0,
      "step": 765
    },
    {
      "entropy": 6.624249505996704,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003845,
      "loss": 6.6708,
      "mean_token_accuracy": 0.11638602316379547,
      "num_tokens": 1727045.0,
      "step": 770
    },
    {
      "entropy": 6.845441389083862,
      "epoch": 0.08704442073341945,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.7396,
      "mean_token_accuracy": 0.12483799159526825,
      "num_tokens": 1737374.0,
      "step": 775
    },
    {
      "entropy": 6.763451385498047,
      "epoch": 0.08760599764137698,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.7368,
      "mean_token_accuracy": 0.1217820793390274,
      "num_tokens": 1747788.0,
      "step": 780
    },
    {
      "entropy": 6.595237827301025,
      "epoch": 0.08816757454933453,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4325,
      "mean_token_accuracy": 0.13330992460250854,
      "num_tokens": 1757769.0,
      "step": 785
    },
    {
      "entropy": 6.58812689781189,
      "epoch": 0.08872915145729207,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.5469,
      "mean_token_accuracy": 0.12191837057471275,
      "num_tokens": 1769246.0,
      "step": 790
    },
    {
      "entropy": 6.780652809143066,
      "epoch": 0.08929072836524962,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.6634,
      "mean_token_accuracy": 0.123942082375288,
      "num_tokens": 1781306.0,
      "step": 795
    },
    {
      "entropy": 6.602289724349975,
      "epoch": 0.08985230527320717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003995,
      "loss": 6.5703,
      "mean_token_accuracy": 0.12607999593019487,
      "num_tokens": 1792888.0,
      "step": 800
    },
    {
      "entropy": 6.695853853225708,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000402,
      "loss": 6.5673,
      "mean_token_accuracy": 0.12844647020101546,
      "num_tokens": 1802938.0,
      "step": 805
    },
    {
      "entropy": 6.667530870437622,
      "epoch": 0.09097545908912226,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004045,
      "loss": 6.6049,
      "mean_token_accuracy": 0.12256239578127862,
      "num_tokens": 1814738.0,
      "step": 810
    },
    {
      "entropy": 6.709681272506714,
      "epoch": 0.0915370359970798,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.6675,
      "mean_token_accuracy": 0.12430930361151696,
      "num_tokens": 1825100.0,
      "step": 815
    },
    {
      "entropy": 6.590344858169556,
      "epoch": 0.09209861290503735,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004095,
      "loss": 6.5098,
      "mean_token_accuracy": 0.12951224520802498,
      "num_tokens": 1835879.0,
      "step": 820
    },
    {
      "entropy": 6.738299226760864,
      "epoch": 0.09266018981299488,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000412,
      "loss": 6.6605,
      "mean_token_accuracy": 0.11758637428283691,
      "num_tokens": 1848352.0,
      "step": 825
    },
    {
      "entropy": 6.652737951278686,
      "epoch": 0.09322176672095243,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004145,
      "loss": 6.6038,
      "mean_token_accuracy": 0.12358284816145897,
      "num_tokens": 1859687.0,
      "step": 830
    },
    {
      "entropy": 6.627905178070068,
      "epoch": 0.09378334362890998,
      "grad_norm": 0.921875,
      "learning_rate": 0.000417,
      "loss": 6.598,
      "mean_token_accuracy": 0.12505071982741356,
      "num_tokens": 1871438.0,
      "step": 835
    },
    {
      "entropy": 6.549618434906006,
      "epoch": 0.09434492053686752,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004195,
      "loss": 6.4806,
      "mean_token_accuracy": 0.13302377462387086,
      "num_tokens": 1881519.0,
      "step": 840
    },
    {
      "entropy": 6.733680200576782,
      "epoch": 0.09490649744482507,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000422,
      "loss": 6.6641,
      "mean_token_accuracy": 0.12029701620340347,
      "num_tokens": 1892428.0,
      "step": 845
    },
    {
      "entropy": 6.674782085418701,
      "epoch": 0.09546807435278261,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004245,
      "loss": 6.5505,
      "mean_token_accuracy": 0.13130439519882203,
      "num_tokens": 1903760.0,
      "step": 850
    },
    {
      "entropy": 6.632183504104614,
      "epoch": 0.09602965126074016,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000427,
      "loss": 6.6123,
      "mean_token_accuracy": 0.11986933276057243,
      "num_tokens": 1915430.0,
      "step": 855
    },
    {
      "entropy": 6.6506672382354735,
      "epoch": 0.09659122816869771,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004295,
      "loss": 6.566,
      "mean_token_accuracy": 0.128663220256567,
      "num_tokens": 1925709.0,
      "step": 860
    },
    {
      "entropy": 6.659067535400391,
      "epoch": 0.09715280507665525,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000432,
      "loss": 6.6133,
      "mean_token_accuracy": 0.12502505853772164,
      "num_tokens": 1938148.0,
      "step": 865
    },
    {
      "entropy": 6.522532606124878,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004345,
      "loss": 6.465,
      "mean_token_accuracy": 0.1305969774723053,
      "num_tokens": 1949028.0,
      "step": 870
    },
    {
      "entropy": 6.521027231216431,
      "epoch": 0.09827595889257033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000437,
      "loss": 6.5332,
      "mean_token_accuracy": 0.12862171232700348,
      "num_tokens": 1960353.0,
      "step": 875
    },
    {
      "entropy": 6.733491516113281,
      "epoch": 0.09883753580052788,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004395,
      "loss": 6.6259,
      "mean_token_accuracy": 0.12553723603487016,
      "num_tokens": 1972335.0,
      "step": 880
    },
    {
      "entropy": 6.552764797210694,
      "epoch": 0.09939911270848542,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000442,
      "loss": 6.4797,
      "mean_token_accuracy": 0.12476090043783188,
      "num_tokens": 1982851.0,
      "step": 885
    },
    {
      "entropy": 6.5761137962341305,
      "epoch": 0.09996068961644297,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004445,
      "loss": 6.4659,
      "mean_token_accuracy": 0.13477788046002387,
      "num_tokens": 1993931.0,
      "step": 890
    },
    {
      "entropy": 6.574522161483765,
      "epoch": 0.10052226652440052,
      "grad_norm": 0.828125,
      "learning_rate": 0.000447,
      "loss": 6.5948,
      "mean_token_accuracy": 0.12630576267838478,
      "num_tokens": 2004838.0,
      "step": 895
    },
    {
      "entropy": 6.519236660003662,
      "epoch": 0.10108384343235806,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3958,
      "mean_token_accuracy": 0.13307079225778579,
      "num_tokens": 2015620.0,
      "step": 900
    },
    {
      "entropy": 6.607462596893311,
      "epoch": 0.10164542034031561,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5494,
      "mean_token_accuracy": 0.13406692892313005,
      "num_tokens": 2027489.0,
      "step": 905
    },
    {
      "entropy": 6.4195150375366214,
      "epoch": 0.10220699724827315,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3915,
      "mean_token_accuracy": 0.13798633739352226,
      "num_tokens": 2038475.0,
      "step": 910
    },
    {
      "entropy": 6.49401707649231,
      "epoch": 0.1027685741562307,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.411,
      "mean_token_accuracy": 0.13152770772576333,
      "num_tokens": 2048938.0,
      "step": 915
    },
    {
      "entropy": 6.484944677352905,
      "epoch": 0.10333015106418825,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4981,
      "mean_token_accuracy": 0.12505212873220445,
      "num_tokens": 2060372.0,
      "step": 920
    },
    {
      "entropy": 6.572954750061035,
      "epoch": 0.10389172797214578,
      "grad_norm": 0.9375,
      "learning_rate": 0.000462,
      "loss": 6.5125,
      "mean_token_accuracy": 0.12695388346910477,
      "num_tokens": 2071558.0,
      "step": 925
    },
    {
      "entropy": 6.553452968597412,
      "epoch": 0.10445330488010333,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004645,
      "loss": 6.5288,
      "mean_token_accuracy": 0.12948471456766128,
      "num_tokens": 2082350.0,
      "step": 930
    },
    {
      "entropy": 6.6269059658050535,
      "epoch": 0.10501488178806087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000467,
      "loss": 6.5882,
      "mean_token_accuracy": 0.12736724987626075,
      "num_tokens": 2093448.0,
      "step": 935
    },
    {
      "entropy": 6.718510961532592,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004695,
      "loss": 6.6336,
      "mean_token_accuracy": 0.12538549676537514,
      "num_tokens": 2104171.0,
      "step": 940
    },
    {
      "entropy": 6.5074928283691404,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472,
      "loss": 6.5012,
      "mean_token_accuracy": 0.13048772215843202,
      "num_tokens": 2114496.0,
      "step": 945
    },
    {
      "entropy": 6.615377426147461,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745,
      "loss": 6.506,
      "mean_token_accuracy": 0.1323799431324005,
      "num_tokens": 2125930.0,
      "step": 950
    },
    {
      "entropy": 6.48550271987915,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477,
      "loss": 6.4768,
      "mean_token_accuracy": 0.13284032344818114,
      "num_tokens": 2136452.0,
      "step": 955
    },
    {
      "entropy": 6.5667078495025635,
      "epoch": 0.1078227663278486,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004795,
      "loss": 6.5651,
      "mean_token_accuracy": 0.12846669927239418,
      "num_tokens": 2148233.0,
      "step": 960
    },
    {
      "entropy": 6.492109537124634,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 6.3508,
      "mean_token_accuracy": 0.13522292152047158,
      "num_tokens": 2159475.0,
      "step": 965
    },
    {
      "entropy": 6.475024604797364,
      "epoch": 0.10894592014376368,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004845,
      "loss": 6.4622,
      "mean_token_accuracy": 0.12998763769865035,
      "num_tokens": 2170042.0,
      "step": 970
    },
    {
      "entropy": 6.565961837768555,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000487,
      "loss": 6.4639,
      "mean_token_accuracy": 0.12638997063040733,
      "num_tokens": 2182234.0,
      "step": 975
    },
    {
      "entropy": 6.398307752609253,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895,
      "loss": 6.4561,
      "mean_token_accuracy": 0.13463474065065384,
      "num_tokens": 2192477.0,
      "step": 980
    },
    {
      "entropy": 6.4558258056640625,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.15625,
      "learning_rate": 0.000492,
      "loss": 6.4511,
      "mean_token_accuracy": 0.12646445855498314,
      "num_tokens": 2203082.0,
      "step": 985
    },
    {
      "entropy": 6.489946889877319,
      "epoch": 0.11119222777559387,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945,
      "loss": 6.4149,
      "mean_token_accuracy": 0.13445554673671722,
      "num_tokens": 2213684.0,
      "step": 990
    },
    {
      "entropy": 6.453236436843872,
      "epoch": 0.11175380468355141,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497,
      "loss": 6.4207,
      "mean_token_accuracy": 0.13301730901002884,
      "num_tokens": 2225524.0,
      "step": 995
    },
    {
      "entropy": 6.585108804702759,
      "epoch": 0.11231538159150896,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995,
      "loss": 6.3966,
      "mean_token_accuracy": 0.1316801317036152,
      "num_tokens": 2236004.0,
      "step": 1000
    },
    {
      "entropy": 6.456496858596802,
      "epoch": 0.1128769584994665,
      "grad_norm": 0.875,
      "learning_rate": 0.000499999997707492,
      "loss": 6.5247,
      "mean_token_accuracy": 0.12453527525067329,
      "num_tokens": 2247730.0,
      "step": 1005
    },
    {
      "entropy": 6.560186100006104,
      "epoch": 0.11343853540742405,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.491,
      "mean_token_accuracy": 0.1362232081592083,
      "num_tokens": 2260132.0,
      "step": 1010
    },
    {
      "entropy": 6.490068674087524,
      "epoch": 0.1140001123153816,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.464,
      "mean_token_accuracy": 0.1341004975140095,
      "num_tokens": 2270060.0,
      "step": 1015
    },
    {
      "entropy": 6.507891464233398,
      "epoch": 0.11456168922333913,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.4733,
      "mean_token_accuracy": 0.12934465631842612,
      "num_tokens": 2281568.0,
      "step": 1020
    },
    {
      "entropy": 6.524950933456421,
      "epoch": 0.11512326613129668,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.4998,
      "mean_token_accuracy": 0.127571739256382,
      "num_tokens": 2293325.0,
      "step": 1025
    },
    {
      "entropy": 6.621273231506348,
      "epoch": 0.11568484303925422,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.5086,
      "mean_token_accuracy": 0.1368153616786003,
      "num_tokens": 2304780.0,
      "step": 1030
    },
    {
      "entropy": 6.402975273132324,
      "epoch": 0.11624641994721177,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.4343,
      "mean_token_accuracy": 0.1310793451964855,
      "num_tokens": 2315999.0,
      "step": 1035
    },
    {
      "entropy": 6.539117860794067,
      "epoch": 0.11680799685516932,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999997820685,
      "loss": 6.5788,
      "mean_token_accuracy": 0.131030685454607,
      "num_tokens": 2326695.0,
      "step": 1040
    },
    {
      "entropy": 6.555915832519531,
      "epoch": 0.11736957376312686,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.4067,
      "mean_token_accuracy": 0.14168997928500177,
      "num_tokens": 2336925.0,
      "step": 1045
    },
    {
      "entropy": 6.3779439449310305,
      "epoch": 0.11793115067108441,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.3148,
      "mean_token_accuracy": 0.13549636751413346,
      "num_tokens": 2347431.0,
      "step": 1050
    },
    {
      "entropy": 6.440852689743042,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.4194,
      "mean_token_accuracy": 0.13899283483624458,
      "num_tokens": 2359894.0,
      "step": 1055
    },
    {
      "entropy": 6.419764804840088,
      "epoch": 0.1190543044869995,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.4644,
      "mean_token_accuracy": 0.1340063691139221,
      "num_tokens": 2371181.0,
      "step": 1060
    },
    {
      "entropy": 6.476871919631958,
      "epoch": 0.11961588139495703,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.3616,
      "mean_token_accuracy": 0.1311580441892147,
      "num_tokens": 2381531.0,
      "step": 1065
    },
    {
      "entropy": 6.585131025314331,
      "epoch": 0.12017745830291458,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.4729,
      "mean_token_accuracy": 0.13767909109592438,
      "num_tokens": 2391573.0,
      "step": 1070
    },
    {
      "entropy": 6.413916444778442,
      "epoch": 0.12073903521087213,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.446,
      "mean_token_accuracy": 0.13483774214982985,
      "num_tokens": 2402822.0,
      "step": 1075
    },
    {
      "entropy": 6.425882148742676,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.3556,
      "mean_token_accuracy": 0.13548125475645065,
      "num_tokens": 2414322.0,
      "step": 1080
    },
    {
      "entropy": 6.4980425357818605,
      "epoch": 0.12186218902678722,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.5464,
      "mean_token_accuracy": 0.1351483039557934,
      "num_tokens": 2426560.0,
      "step": 1085
    },
    {
      "entropy": 6.573987579345703,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.4631,
      "mean_token_accuracy": 0.13326464518904685,
      "num_tokens": 2437723.0,
      "step": 1090
    },
    {
      "entropy": 6.402809953689575,
      "epoch": 0.12298534284270231,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.4139,
      "mean_token_accuracy": 0.13775320574641228,
      "num_tokens": 2448497.0,
      "step": 1095
    },
    {
      "entropy": 6.311347532272339,
      "epoch": 0.12354691975065986,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.2693,
      "mean_token_accuracy": 0.14299433752894403,
      "num_tokens": 2459089.0,
      "step": 1100
    },
    {
      "entropy": 6.469817876815796,
      "epoch": 0.1241084966586174,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.3627,
      "mean_token_accuracy": 0.14109985455870627,
      "num_tokens": 2469882.0,
      "step": 1105
    },
    {
      "entropy": 6.417138671875,
      "epoch": 0.12467007356657495,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.415,
      "mean_token_accuracy": 0.13988902568817138,
      "num_tokens": 2482120.0,
      "step": 1110
    },
    {
      "entropy": 6.481253623962402,
      "epoch": 0.1252316504745325,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.4102,
      "mean_token_accuracy": 0.13518037721514703,
      "num_tokens": 2492626.0,
      "step": 1115
    },
    {
      "entropy": 6.59227557182312,
      "epoch": 0.12579322738249002,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499997970990249,
      "loss": 6.4646,
      "mean_token_accuracy": 0.1327376738190651,
      "num_tokens": 2506347.0,
      "step": 1120
    },
    {
      "entropy": 6.266043615341187,
      "epoch": 0.12635480429044757,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.3464,
      "mean_token_accuracy": 0.13951586112380027,
      "num_tokens": 2516584.0,
      "step": 1125
    },
    {
      "entropy": 6.528832149505615,
      "epoch": 0.12691638119840512,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.3984,
      "mean_token_accuracy": 0.13773541748523713,
      "num_tokens": 2526731.0,
      "step": 1130
    },
    {
      "entropy": 6.33425726890564,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.2973,
      "mean_token_accuracy": 0.1425212301313877,
      "num_tokens": 2537180.0,
      "step": 1135
    },
    {
      "entropy": 6.518994569778442,
      "epoch": 0.12803953501432022,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.3819,
      "mean_token_accuracy": 0.13330503478646277,
      "num_tokens": 2547617.0,
      "step": 1140
    },
    {
      "entropy": 6.3555159091949465,
      "epoch": 0.12860111192227774,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.2686,
      "mean_token_accuracy": 0.1448584720492363,
      "num_tokens": 2557743.0,
      "step": 1145
    },
    {
      "entropy": 6.3979815483093265,
      "epoch": 0.1291626888302353,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.3601,
      "mean_token_accuracy": 0.13572428226470948,
      "num_tokens": 2568351.0,
      "step": 1150
    },
    {
      "entropy": 6.344897937774658,
      "epoch": 0.12972426573819285,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.4169,
      "mean_token_accuracy": 0.14001458808779715,
      "num_tokens": 2579703.0,
      "step": 1155
    },
    {
      "entropy": 6.554988098144531,
      "epoch": 0.1302858426461504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.3852,
      "mean_token_accuracy": 0.13615019991993904,
      "num_tokens": 2591197.0,
      "step": 1160
    },
    {
      "entropy": 6.3376793384552,
      "epoch": 0.13084741955410795,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499996146305169,
      "loss": 6.3748,
      "mean_token_accuracy": 0.13186048045754434,
      "num_tokens": 2601836.0,
      "step": 1165
    },
    {
      "entropy": 6.344990921020508,
      "epoch": 0.13140899646206547,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.3009,
      "mean_token_accuracy": 0.1399838626384735,
      "num_tokens": 2613513.0,
      "step": 1170
    },
    {
      "entropy": 6.420182085037231,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.3476,
      "mean_token_accuracy": 0.13971240893006326,
      "num_tokens": 2625246.0,
      "step": 1175
    },
    {
      "entropy": 6.348382759094238,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.3225,
      "mean_token_accuracy": 0.13963563665747641,
      "num_tokens": 2636375.0,
      "step": 1180
    },
    {
      "entropy": 6.378366088867187,
      "epoch": 0.13309372718593812,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.4077,
      "mean_token_accuracy": 0.13064486235380174,
      "num_tokens": 2648586.0,
      "step": 1185
    },
    {
      "entropy": 6.471061944961548,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.4102,
      "mean_token_accuracy": 0.142090106010437,
      "num_tokens": 2660847.0,
      "step": 1190
    },
    {
      "entropy": 6.4110862731933596,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.828125,
      "learning_rate": 0.00049999460746976,
      "loss": 6.3302,
      "mean_token_accuracy": 0.14113106280565263,
      "num_tokens": 2671160.0,
      "step": 1195
    },
    {
      "entropy": 6.469895267486573,
      "epoch": 0.13477845790981074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.4551,
      "mean_token_accuracy": 0.1292402721941471,
      "num_tokens": 2681622.0,
      "step": 1200
    },
    {
      "entropy": 6.354708957672119,
      "epoch": 0.1353400348177683,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.2578,
      "mean_token_accuracy": 0.14201848655939103,
      "num_tokens": 2691503.0,
      "step": 1205
    },
    {
      "entropy": 6.4176633834838865,
      "epoch": 0.13590161172572585,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.3486,
      "mean_token_accuracy": 0.13264842703938484,
      "num_tokens": 2702207.0,
      "step": 1210
    },
    {
      "entropy": 6.377297592163086,
      "epoch": 0.1364631886336834,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.3505,
      "mean_token_accuracy": 0.13756700977683067,
      "num_tokens": 2713178.0,
      "step": 1215
    },
    {
      "entropy": 6.4666331768035885,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.3452,
      "mean_token_accuracy": 0.14598208293318748,
      "num_tokens": 2723717.0,
      "step": 1220
    },
    {
      "entropy": 6.428584241867066,
      "epoch": 0.13758634244959847,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.4134,
      "mean_token_accuracy": 0.1348503641784191,
      "num_tokens": 2735669.0,
      "step": 1225
    },
    {
      "entropy": 6.266319847106933,
      "epoch": 0.13814791935755602,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.3068,
      "mean_token_accuracy": 0.1396687351167202,
      "num_tokens": 2746087.0,
      "step": 1230
    },
    {
      "entropy": 6.460708808898926,
      "epoch": 0.13870949626551357,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.3242,
      "mean_token_accuracy": 0.13702112659811974,
      "num_tokens": 2757143.0,
      "step": 1235
    },
    {
      "entropy": 6.414711141586304,
      "epoch": 0.1392710731734711,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.4365,
      "mean_token_accuracy": 0.13701202794909478,
      "num_tokens": 2769340.0,
      "step": 1240
    },
    {
      "entropy": 6.432314443588257,
      "epoch": 0.13983265008142864,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.3401,
      "mean_token_accuracy": 0.13881635367870332,
      "num_tokens": 2779514.0,
      "step": 1245
    },
    {
      "entropy": 6.358312511444092,
      "epoch": 0.1403942269893862,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.2974,
      "mean_token_accuracy": 0.14529771953821183,
      "num_tokens": 2790801.0,
      "step": 1250
    },
    {
      "entropy": 6.419577169418335,
      "epoch": 0.14095580389734375,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.2782,
      "mean_token_accuracy": 0.14040456488728523,
      "num_tokens": 2801082.0,
      "step": 1255
    },
    {
      "entropy": 6.296146583557129,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.2855,
      "mean_token_accuracy": 0.14076389670372008,
      "num_tokens": 2812719.0,
      "step": 1260
    },
    {
      "entropy": 6.348859882354736,
      "epoch": 0.14207895771325882,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499990013909322,
      "loss": 6.2849,
      "mean_token_accuracy": 0.1368976838886738,
      "num_tokens": 2823766.0,
      "step": 1265
    },
    {
      "entropy": 6.348705673217774,
      "epoch": 0.14264053462121637,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499989632069227,
      "loss": 6.3652,
      "mean_token_accuracy": 0.13619414418935777,
      "num_tokens": 2834801.0,
      "step": 1270
    },
    {
      "entropy": 6.347746229171753,
      "epoch": 0.14320211152917392,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.2028,
      "mean_token_accuracy": 0.14546942338347435,
      "num_tokens": 2845769.0,
      "step": 1275
    },
    {
      "entropy": 6.289702272415161,
      "epoch": 0.14376368843713147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.2401,
      "mean_token_accuracy": 0.14375891238451005,
      "num_tokens": 2856001.0,
      "step": 1280
    },
    {
      "entropy": 6.35533881187439,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.2663,
      "mean_token_accuracy": 0.14633105844259262,
      "num_tokens": 2867703.0,
      "step": 1285
    },
    {
      "entropy": 6.233186101913452,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999880330714,
      "loss": 6.234,
      "mean_token_accuracy": 0.14503639936447144,
      "num_tokens": 2878671.0,
      "step": 1290
    },
    {
      "entropy": 6.362540769577026,
      "epoch": 0.1454484191610041,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.3403,
      "mean_token_accuracy": 0.13664028346538543,
      "num_tokens": 2890137.0,
      "step": 1295
    },
    {
      "entropy": 6.306020212173462,
      "epoch": 0.14600999606896164,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.2357,
      "mean_token_accuracy": 0.13921404853463173,
      "num_tokens": 2900344.0,
      "step": 1300
    },
    {
      "entropy": 6.284024572372436,
      "epoch": 0.1465715729769192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.2557,
      "mean_token_accuracy": 0.14088367968797683,
      "num_tokens": 2912270.0,
      "step": 1305
    },
    {
      "entropy": 6.376524925231934,
      "epoch": 0.14713314988487675,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.3348,
      "mean_token_accuracy": 0.13199515640735626,
      "num_tokens": 2924677.0,
      "step": 1310
    },
    {
      "entropy": 6.2676990032196045,
      "epoch": 0.14769472679283427,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.2504,
      "mean_token_accuracy": 0.1430131196975708,
      "num_tokens": 2935794.0,
      "step": 1315
    },
    {
      "entropy": 6.3416272640228275,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.3627,
      "mean_token_accuracy": 0.13728150352835655,
      "num_tokens": 2945958.0,
      "step": 1320
    },
    {
      "entropy": 6.376453304290772,
      "epoch": 0.14881788060874937,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.3296,
      "mean_token_accuracy": 0.13831673040986062,
      "num_tokens": 2957536.0,
      "step": 1325
    },
    {
      "entropy": 6.417525577545166,
      "epoch": 0.14937945751670692,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.313,
      "mean_token_accuracy": 0.13861921206116676,
      "num_tokens": 2967787.0,
      "step": 1330
    },
    {
      "entropy": 6.407558107376099,
      "epoch": 0.14994103442466447,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499984016250827,
      "loss": 6.3224,
      "mean_token_accuracy": 0.13716086149215698,
      "num_tokens": 2980522.0,
      "step": 1335
    },
    {
      "entropy": 6.188078784942627,
      "epoch": 0.150502611332622,
      "grad_norm": 1.0,
      "learning_rate": 0.000499983534119342,
      "loss": 6.2513,
      "mean_token_accuracy": 0.13945341408252715,
      "num_tokens": 2990497.0,
      "step": 1340
    },
    {
      "entropy": 6.482673597335816,
      "epoch": 0.15106418824057954,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.353,
      "mean_token_accuracy": 0.14002300277352334,
      "num_tokens": 3002931.0,
      "step": 1345
    },
    {
      "entropy": 6.254575872421265,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.2748,
      "mean_token_accuracy": 0.14595276713371277,
      "num_tokens": 3014137.0,
      "step": 1350
    },
    {
      "entropy": 6.553881549835205,
      "epoch": 0.15218734205649465,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.4367,
      "mean_token_accuracy": 0.13388766795396806,
      "num_tokens": 3026145.0,
      "step": 1355
    },
    {
      "entropy": 6.27535490989685,
      "epoch": 0.15274891896445217,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.2853,
      "mean_token_accuracy": 0.13878872245550156,
      "num_tokens": 3037068.0,
      "step": 1360
    },
    {
      "entropy": 6.293557262420654,
      "epoch": 0.15331049587240972,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.2339,
      "mean_token_accuracy": 0.14105164185166358,
      "num_tokens": 3047836.0,
      "step": 1365
    },
    {
      "entropy": 6.372267818450927,
      "epoch": 0.15387207278036727,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.3584,
      "mean_token_accuracy": 0.1421847365796566,
      "num_tokens": 3060309.0,
      "step": 1370
    },
    {
      "entropy": 6.353737211227417,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.3003,
      "mean_token_accuracy": 0.1419496037065983,
      "num_tokens": 3071789.0,
      "step": 1375
    },
    {
      "entropy": 6.360533714294434,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499979419180532,
      "loss": 6.2236,
      "mean_token_accuracy": 0.1449585184454918,
      "num_tokens": 3081756.0,
      "step": 1380
    },
    {
      "entropy": 6.235655450820923,
      "epoch": 0.1555568035042399,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.3434,
      "mean_token_accuracy": 0.13456685617566108,
      "num_tokens": 3093456.0,
      "step": 1385
    },
    {
      "entropy": 6.440259408950806,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.265,
      "mean_token_accuracy": 0.13899147212505342,
      "num_tokens": 3104565.0,
      "step": 1390
    },
    {
      "entropy": 6.34439787864685,
      "epoch": 0.156679957320155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999777578813904,
      "loss": 6.2861,
      "mean_token_accuracy": 0.14071865901350974,
      "num_tokens": 3115577.0,
      "step": 1395
    },
    {
      "entropy": 6.224101924896241,
      "epoch": 0.15724153422811254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499977189788228,
      "loss": 6.1836,
      "mean_token_accuracy": 0.14295614287257194,
      "num_tokens": 3126019.0,
      "step": 1400
    },
    {
      "entropy": 6.388071966171265,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.3394,
      "mean_token_accuracy": 0.13431284055113793,
      "num_tokens": 3136191.0,
      "step": 1405
    },
    {
      "entropy": 6.397411298751831,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.2113,
      "mean_token_accuracy": 0.143301859498024,
      "num_tokens": 3147737.0,
      "step": 1410
    },
    {
      "entropy": 6.158065271377564,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.2684,
      "mean_token_accuracy": 0.13974310755729674,
      "num_tokens": 3158494.0,
      "step": 1415
    },
    {
      "entropy": 6.432788991928101,
      "epoch": 0.15948784185994272,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.315,
      "mean_token_accuracy": 0.14211769849061967,
      "num_tokens": 3170550.0,
      "step": 1420
    },
    {
      "entropy": 6.21512804031372,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499974241872371,
      "loss": 6.2877,
      "mean_token_accuracy": 0.14602836072444916,
      "num_tokens": 3182860.0,
      "step": 1425
    },
    {
      "entropy": 6.338991689682007,
      "epoch": 0.16061099567585782,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.2843,
      "mean_token_accuracy": 0.13677069991827012,
      "num_tokens": 3193807.0,
      "step": 1430
    },
    {
      "entropy": 6.18875823020935,
      "epoch": 0.16117257258381534,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.1569,
      "mean_token_accuracy": 0.13699881434440614,
      "num_tokens": 3203959.0,
      "step": 1435
    },
    {
      "entropy": 6.351223182678223,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.1745,
      "mean_token_accuracy": 0.14297635555267335,
      "num_tokens": 3215416.0,
      "step": 1440
    },
    {
      "entropy": 6.344785833358765,
      "epoch": 0.16229572639973044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.3179,
      "mean_token_accuracy": 0.13555386140942574,
      "num_tokens": 3226938.0,
      "step": 1445
    },
    {
      "entropy": 6.229445600509644,
      "epoch": 0.162857303307688,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999711148748476,
      "loss": 6.1367,
      "mean_token_accuracy": 0.14848765432834626,
      "num_tokens": 3238406.0,
      "step": 1450
    },
    {
      "entropy": 6.131477737426758,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.1427,
      "mean_token_accuracy": 0.14189836829900743,
      "num_tokens": 3249695.0,
      "step": 1455
    },
    {
      "entropy": 6.446213865280152,
      "epoch": 0.16398045712360307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.3058,
      "mean_token_accuracy": 0.1473018154501915,
      "num_tokens": 3259656.0,
      "step": 1460
    },
    {
      "entropy": 6.181229066848755,
      "epoch": 0.16454203403156062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.1311,
      "mean_token_accuracy": 0.15136088579893112,
      "num_tokens": 3269712.0,
      "step": 1465
    },
    {
      "entropy": 6.132031583786011,
      "epoch": 0.16510361093951817,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999684843396968,
      "loss": 6.1254,
      "mean_token_accuracy": 0.1496811881661415,
      "num_tokens": 3280343.0,
      "step": 1470
    },
    {
      "entropy": 6.244436836242675,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.1549,
      "mean_token_accuracy": 0.14772350192070008,
      "num_tokens": 3291649.0,
      "step": 1475
    },
    {
      "entropy": 6.332507038116455,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.3432,
      "mean_token_accuracy": 0.14269206747412683,
      "num_tokens": 3304190.0,
      "step": 1480
    },
    {
      "entropy": 6.224244928359985,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999664362258814,
      "loss": 6.0663,
      "mean_token_accuracy": 0.1542764872312546,
      "num_tokens": 3313966.0,
      "step": 1485
    },
    {
      "entropy": 6.275435161590576,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.3114,
      "mean_token_accuracy": 0.14152566939592362,
      "num_tokens": 3326017.0,
      "step": 1490
    },
    {
      "entropy": 6.349188041687012,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499965035001542,
      "loss": 6.2104,
      "mean_token_accuracy": 0.13980446606874466,
      "num_tokens": 3338264.0,
      "step": 1495
    },
    {
      "entropy": 6.26988787651062,
      "epoch": 0.16847307238726345,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.2569,
      "mean_token_accuracy": 0.15189649015665055,
      "num_tokens": 3348901.0,
      "step": 1500
    },
    {
      "entropy": 6.249851369857788,
      "epoch": 0.16903464929522097,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.1784,
      "mean_token_accuracy": 0.14386393129825592,
      "num_tokens": 3358441.0,
      "step": 1505
    },
    {
      "entropy": 6.336975526809693,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.2653,
      "mean_token_accuracy": 0.1504651814699173,
      "num_tokens": 3369334.0,
      "step": 1510
    },
    {
      "entropy": 6.2502185821533205,
      "epoch": 0.17015780311113607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.2748,
      "mean_token_accuracy": 0.1409212589263916,
      "num_tokens": 3380698.0,
      "step": 1515
    },
    {
      "entropy": 6.462999057769776,
      "epoch": 0.17071938001909362,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.288,
      "mean_token_accuracy": 0.1381421558558941,
      "num_tokens": 3391061.0,
      "step": 1520
    },
    {
      "entropy": 6.246134901046753,
      "epoch": 0.17128095692705117,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.3371,
      "mean_token_accuracy": 0.13986739218235017,
      "num_tokens": 3403983.0,
      "step": 1525
    },
    {
      "entropy": 6.3834983825683596,
      "epoch": 0.1718425338350087,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049995990508388,
      "loss": 6.2555,
      "mean_token_accuracy": 0.1448330372571945,
      "num_tokens": 3413838.0,
      "step": 1530
    },
    {
      "entropy": 6.145442771911621,
      "epoch": 0.17240411074296624,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999591435870674,
      "loss": 6.0926,
      "mean_token_accuracy": 0.14555919021368027,
      "num_tokens": 3423978.0,
      "step": 1535
    },
    {
      "entropy": 6.297583770751953,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.2691,
      "mean_token_accuracy": 0.1472916215658188,
      "num_tokens": 3433416.0,
      "step": 1540
    },
    {
      "entropy": 6.213202953338623,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999575991051074,
      "loss": 6.1183,
      "mean_token_accuracy": 0.1482724741101265,
      "num_tokens": 3444983.0,
      "step": 1545
    },
    {
      "entropy": 6.228994274139405,
      "epoch": 0.1740888414668389,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999568161200093,
      "loss": 6.0923,
      "mean_token_accuracy": 0.1479077994823456,
      "num_tokens": 3455808.0,
      "step": 1550
    },
    {
      "entropy": 6.121140384674073,
      "epoch": 0.17465041837479642,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.2406,
      "mean_token_accuracy": 0.14061250612139703,
      "num_tokens": 3467503.0,
      "step": 1555
    },
    {
      "entropy": 6.5244382381439205,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.3998,
      "mean_token_accuracy": 0.14157005473971368,
      "num_tokens": 3479941.0,
      "step": 1560
    },
    {
      "entropy": 6.189446020126343,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999544241885421,
      "loss": 6.1028,
      "mean_token_accuracy": 0.15490597486495972,
      "num_tokens": 3491069.0,
      "step": 1565
    },
    {
      "entropy": 6.284439182281494,
      "epoch": 0.17633514909866907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.3132,
      "mean_token_accuracy": 0.13803954273462296,
      "num_tokens": 3502065.0,
      "step": 1570
    },
    {
      "entropy": 6.291994380950928,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499952793754341,
      "loss": 6.2551,
      "mean_token_accuracy": 0.14772795587778093,
      "num_tokens": 3514228.0,
      "step": 1575
    },
    {
      "entropy": 6.30311017036438,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.1707,
      "mean_token_accuracy": 0.14605900645256042,
      "num_tokens": 3525801.0,
      "step": 1580
    },
    {
      "entropy": 6.187708377838135,
      "epoch": 0.1780198798225417,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999511346698035,
      "loss": 6.1216,
      "mean_token_accuracy": 0.1448603853583336,
      "num_tokens": 3536431.0,
      "step": 1585
    },
    {
      "entropy": 6.21040096282959,
      "epoch": 0.17858145673049924,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.2162,
      "mean_token_accuracy": 0.14277410134673119,
      "num_tokens": 3549098.0,
      "step": 1590
    },
    {
      "entropy": 6.308737564086914,
      "epoch": 0.1791430336384568,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.187,
      "mean_token_accuracy": 0.1415993995964527,
      "num_tokens": 3560532.0,
      "step": 1595
    },
    {
      "entropy": 6.151534795761108,
      "epoch": 0.17970461054641435,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.1807,
      "mean_token_accuracy": 0.14444077014923096,
      "num_tokens": 3572175.0,
      "step": 1600
    },
    {
      "entropy": 6.327757406234741,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999477305505675,
      "loss": 6.1126,
      "mean_token_accuracy": 0.15372487977147103,
      "num_tokens": 3582862.0,
      "step": 1605
    },
    {
      "entropy": 6.3059488296508786,
      "epoch": 0.18082776436232942,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.2841,
      "mean_token_accuracy": 0.14689953327178956,
      "num_tokens": 3594685.0,
      "step": 1610
    },
    {
      "entropy": 6.120879793167115,
      "epoch": 0.18138934127028697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999459855163026,
      "loss": 6.0994,
      "mean_token_accuracy": 0.152867791056633,
      "num_tokens": 3604847.0,
      "step": 1615
    },
    {
      "entropy": 6.267522573471069,
      "epoch": 0.18195091817824452,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499945102255605,
      "loss": 6.1771,
      "mean_token_accuracy": 0.14639166221022606,
      "num_tokens": 3615549.0,
      "step": 1620
    },
    {
      "entropy": 6.195144271850586,
      "epoch": 0.18251249508620204,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999442118325682,
      "loss": 6.0792,
      "mean_token_accuracy": 0.1583167091012001,
      "num_tokens": 3626300.0,
      "step": 1625
    },
    {
      "entropy": 6.165605783462524,
      "epoch": 0.1830740719941596,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999433142472205,
      "loss": 6.1142,
      "mean_token_accuracy": 0.1476339131593704,
      "num_tokens": 3637832.0,
      "step": 1630
    },
    {
      "entropy": 6.1696093559265135,
      "epoch": 0.18363564890211714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.1569,
      "mean_token_accuracy": 0.1420426405966282,
      "num_tokens": 3649179.0,
      "step": 1635
    },
    {
      "entropy": 6.245587396621704,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.198,
      "mean_token_accuracy": 0.1402585081756115,
      "num_tokens": 3660011.0,
      "step": 1640
    },
    {
      "entropy": 6.302711725234985,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.2802,
      "mean_token_accuracy": 0.14278458282351494,
      "num_tokens": 3672791.0,
      "step": 1645
    },
    {
      "entropy": 6.166055631637573,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999396522832949,
      "loss": 6.0875,
      "mean_token_accuracy": 0.1390815943479538,
      "num_tokens": 3684709.0,
      "step": 1650
    },
    {
      "entropy": 6.152737283706665,
      "epoch": 0.18588195653394732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999387188868256,
      "loss": 6.0627,
      "mean_token_accuracy": 0.15113601088523865,
      "num_tokens": 3694675.0,
      "step": 1655
    },
    {
      "entropy": 6.079391670227051,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999377783282203,
      "loss": 6.0739,
      "mean_token_accuracy": 0.14676920920610428,
      "num_tokens": 3706384.0,
      "step": 1660
    },
    {
      "entropy": 6.2547712326049805,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.1527,
      "mean_token_accuracy": 0.14625280648469924,
      "num_tokens": 3717419.0,
      "step": 1665
    },
    {
      "entropy": 6.255711317062378,
      "epoch": 0.18756668725781997,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.1702,
      "mean_token_accuracy": 0.1443650782108307,
      "num_tokens": 3727306.0,
      "step": 1670
    },
    {
      "entropy": 6.278718280792236,
      "epoch": 0.1881282641657775,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.2849,
      "mean_token_accuracy": 0.14049319103360175,
      "num_tokens": 3739476.0,
      "step": 1675
    },
    {
      "entropy": 6.170718765258789,
      "epoch": 0.18868984107373504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999339444730409,
      "loss": 6.0598,
      "mean_token_accuracy": 0.15319431126117705,
      "num_tokens": 3751035.0,
      "step": 1680
    },
    {
      "entropy": 6.1924333572387695,
      "epoch": 0.1892514179816926,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999329681042092,
      "loss": 6.1581,
      "mean_token_accuracy": 0.1471312403678894,
      "num_tokens": 3762881.0,
      "step": 1685
    },
    {
      "entropy": 6.224253702163696,
      "epoch": 0.18981299488965014,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.1835,
      "mean_token_accuracy": 0.14361522495746612,
      "num_tokens": 3774180.0,
      "step": 1690
    },
    {
      "entropy": 6.230266857147217,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.1492,
      "mean_token_accuracy": 0.1505300983786583,
      "num_tokens": 3784864.0,
      "step": 1695
    },
    {
      "entropy": 6.259658145904541,
      "epoch": 0.19093614870556522,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.1322,
      "mean_token_accuracy": 0.14401278346776963,
      "num_tokens": 3797693.0,
      "step": 1700
    },
    {
      "entropy": 6.167710828781128,
      "epoch": 0.19149772561352277,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.1574,
      "mean_token_accuracy": 0.14654720053076745,
      "num_tokens": 3809379.0,
      "step": 1705
    },
    {
      "entropy": 6.081889629364014,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999279788313858,
      "loss": 5.999,
      "mean_token_accuracy": 0.15570705085992814,
      "num_tokens": 3819473.0,
      "step": 1710
    },
    {
      "entropy": 6.077426338195801,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999269594913107,
      "loss": 5.9932,
      "mean_token_accuracy": 0.15049063116312028,
      "num_tokens": 3831037.0,
      "step": 1715
    },
    {
      "entropy": 6.112188291549683,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.1417,
      "mean_token_accuracy": 0.148158148676157,
      "num_tokens": 3841767.0,
      "step": 1720
    },
    {
      "entropy": 6.089222717285156,
      "epoch": 0.19374403324535294,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999248993259081,
      "loss": 5.9764,
      "mean_token_accuracy": 0.1548941805958748,
      "num_tokens": 3851629.0,
      "step": 1725
    },
    {
      "entropy": 6.0848329067230225,
      "epoch": 0.1943056101533105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999238585006463,
      "loss": 6.0616,
      "mean_token_accuracy": 0.14957866221666336,
      "num_tokens": 3863247.0,
      "step": 1730
    },
    {
      "entropy": 6.147818183898925,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999228105137218,
      "loss": 6.0335,
      "mean_token_accuracy": 0.1490088775753975,
      "num_tokens": 3873949.0,
      "step": 1735
    },
    {
      "entropy": 6.188397979736328,
      "epoch": 0.1954287639692256,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999217553651676,
      "loss": 6.1385,
      "mean_token_accuracy": 0.15051164627075195,
      "num_tokens": 3884992.0,
      "step": 1740
    },
    {
      "entropy": 6.246903657913208,
      "epoch": 0.19599034087718312,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.2617,
      "mean_token_accuracy": 0.15150685235857964,
      "num_tokens": 3897029.0,
      "step": 1745
    },
    {
      "entropy": 6.131383800506592,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999196235833054,
      "loss": 6.105,
      "mean_token_accuracy": 0.14395989701151848,
      "num_tokens": 3909510.0,
      "step": 1750
    },
    {
      "entropy": 6.214809513092041,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999185469500651,
      "loss": 6.0185,
      "mean_token_accuracy": 0.1482173793017864,
      "num_tokens": 3920593.0,
      "step": 1755
    },
    {
      "entropy": 6.142779541015625,
      "epoch": 0.19767507160105577,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499917463155331,
      "loss": 6.1469,
      "mean_token_accuracy": 0.14746149256825447,
      "num_tokens": 3932215.0,
      "step": 1760
    },
    {
      "entropy": 6.216650819778442,
      "epoch": 0.19823664850901332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999163721991378,
      "loss": 6.1024,
      "mean_token_accuracy": 0.14692942351102828,
      "num_tokens": 3943876.0,
      "step": 1765
    },
    {
      "entropy": 6.062715196609497,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049991527408152,
      "loss": 6.0895,
      "mean_token_accuracy": 0.15319886654615403,
      "num_tokens": 3954533.0,
      "step": 1770
    },
    {
      "entropy": 6.210793972015381,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.148,
      "mean_token_accuracy": 0.15144960433244706,
      "num_tokens": 3965722.0,
      "step": 1775
    },
    {
      "entropy": 6.174288940429688,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499913056362151,
      "loss": 6.088,
      "mean_token_accuracy": 0.14908579289913176,
      "num_tokens": 3976737.0,
      "step": 1780
    },
    {
      "entropy": 6.225276279449463,
      "epoch": 0.2004829561408435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999119367604703,
      "loss": 6.1585,
      "mean_token_accuracy": 0.1512651652097702,
      "num_tokens": 3986102.0,
      "step": 1785
    },
    {
      "entropy": 6.183860206604004,
      "epoch": 0.20104453304880104,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999108099975064,
      "loss": 6.1194,
      "mean_token_accuracy": 0.15166645050048827,
      "num_tokens": 3996534.0,
      "step": 1790
    },
    {
      "entropy": 6.145701837539673,
      "epoch": 0.20160610995675857,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499909676073295,
      "loss": 6.0466,
      "mean_token_accuracy": 0.14939896166324615,
      "num_tokens": 4006947.0,
      "step": 1795
    },
    {
      "entropy": 6.029658460617066,
      "epoch": 0.20216768686471612,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999085349878723,
      "loss": 6.0612,
      "mean_token_accuracy": 0.1588248908519745,
      "num_tokens": 4017723.0,
      "step": 1800
    },
    {
      "entropy": 6.135398435592651,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999073867412746,
      "loss": 6.1535,
      "mean_token_accuracy": 0.14331067353487015,
      "num_tokens": 4028347.0,
      "step": 1805
    },
    {
      "entropy": 6.15322699546814,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999062313335385,
      "loss": 6.0206,
      "mean_token_accuracy": 0.1573499321937561,
      "num_tokens": 4039648.0,
      "step": 1810
    },
    {
      "entropy": 6.208154487609863,
      "epoch": 0.20385241758858877,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999050687647008,
      "loss": 6.071,
      "mean_token_accuracy": 0.1487251937389374,
      "num_tokens": 4050786.0,
      "step": 1815
    },
    {
      "entropy": 6.128590059280396,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999038990347984,
      "loss": 6.1181,
      "mean_token_accuracy": 0.1585391640663147,
      "num_tokens": 4061184.0,
      "step": 1820
    },
    {
      "entropy": 6.332269668579102,
      "epoch": 0.20497557140450384,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.2477,
      "mean_token_accuracy": 0.1459768995642662,
      "num_tokens": 4073607.0,
      "step": 1825
    },
    {
      "entropy": 6.105072927474976,
      "epoch": 0.2055371483124614,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499901538091949,
      "loss": 6.063,
      "mean_token_accuracy": 0.14765892326831817,
      "num_tokens": 4083924.0,
      "step": 1830
    },
    {
      "entropy": 6.109910249710083,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999003468790773,
      "loss": 6.0556,
      "mean_token_accuracy": 0.15023878514766692,
      "num_tokens": 4094083.0,
      "step": 1835
    },
    {
      "entropy": 6.175615119934082,
      "epoch": 0.2066603021283765,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998991485052911,
      "loss": 6.0894,
      "mean_token_accuracy": 0.157446925342083,
      "num_tokens": 4105124.0,
      "step": 1840
    },
    {
      "entropy": 6.236642742156983,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499897942970629,
      "loss": 6.2515,
      "mean_token_accuracy": 0.14365296810865402,
      "num_tokens": 4116792.0,
      "step": 1845
    },
    {
      "entropy": 6.346523237228394,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499896730275129,
      "loss": 6.2425,
      "mean_token_accuracy": 0.14365033358335494,
      "num_tokens": 4127441.0,
      "step": 1850
    },
    {
      "entropy": 6.223315095901489,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049989551041883,
      "loss": 6.216,
      "mean_token_accuracy": 0.14539413154125214,
      "num_tokens": 4138696.0,
      "step": 1855
    },
    {
      "entropy": 6.098178768157959,
      "epoch": 0.20890660976020667,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998942834017707,
      "loss": 6.0751,
      "mean_token_accuracy": 0.1485121563076973,
      "num_tokens": 4151254.0,
      "step": 1860
    },
    {
      "entropy": 6.124366617202758,
      "epoch": 0.2094681866681642,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998930492239901,
      "loss": 6.0368,
      "mean_token_accuracy": 0.1491617165505886,
      "num_tokens": 4162046.0,
      "step": 1865
    },
    {
      "entropy": 6.119881010055542,
      "epoch": 0.21002976357612174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998918078855277,
      "loss": 6.0023,
      "mean_token_accuracy": 0.15643363893032075,
      "num_tokens": 4172409.0,
      "step": 1870
    },
    {
      "entropy": 6.118890285491943,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998905593864227,
      "loss": 6.1515,
      "mean_token_accuracy": 0.1560173213481903,
      "num_tokens": 4182949.0,
      "step": 1875
    },
    {
      "entropy": 6.123640012741089,
      "epoch": 0.21115291739203684,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998893037267153,
      "loss": 6.1166,
      "mean_token_accuracy": 0.15241918265819548,
      "num_tokens": 4194164.0,
      "step": 1880
    },
    {
      "entropy": 6.308550214767456,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998880409064452,
      "loss": 6.0822,
      "mean_token_accuracy": 0.15168551802635194,
      "num_tokens": 4206056.0,
      "step": 1885
    },
    {
      "entropy": 6.042774629592896,
      "epoch": 0.21227607120795192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998867709256527,
      "loss": 6.1398,
      "mean_token_accuracy": 0.14869935661554337,
      "num_tokens": 4216660.0,
      "step": 1890
    },
    {
      "entropy": 6.086363315582275,
      "epoch": 0.21283764811590947,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000499885493784378,
      "loss": 6.0304,
      "mean_token_accuracy": 0.15354848504066468,
      "num_tokens": 4228495.0,
      "step": 1895
    },
    {
      "entropy": 6.137011957168579,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998842094826621,
      "loss": 6.0173,
      "mean_token_accuracy": 0.15438924431800843,
      "num_tokens": 4238560.0,
      "step": 1900
    },
    {
      "entropy": 6.263278245925903,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.2303,
      "mean_token_accuracy": 0.14329903498291968,
      "num_tokens": 4250553.0,
      "step": 1905
    },
    {
      "entropy": 6.198366451263428,
      "epoch": 0.21452237883978212,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998816193980701,
      "loss": 6.161,
      "mean_token_accuracy": 0.14747311770915986,
      "num_tokens": 4262050.0,
      "step": 1910
    },
    {
      "entropy": 6.277743291854859,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998803136152764,
      "loss": 6.1105,
      "mean_token_accuracy": 0.14608076959848404,
      "num_tokens": 4273019.0,
      "step": 1915
    },
    {
      "entropy": 6.069996547698975,
      "epoch": 0.2156455326556972,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998790006722063,
      "loss": 6.0012,
      "mean_token_accuracy": 0.15598759353160857,
      "num_tokens": 4283585.0,
      "step": 1920
    },
    {
      "entropy": 6.026457405090332,
      "epoch": 0.21620710956365474,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998776805689016,
      "loss": 6.0773,
      "mean_token_accuracy": 0.15107496082782745,
      "num_tokens": 4294905.0,
      "step": 1925
    },
    {
      "entropy": 6.290648698806763,
      "epoch": 0.2167686864716123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998763533054045,
      "loss": 6.1873,
      "mean_token_accuracy": 0.15128218084573747,
      "num_tokens": 4305926.0,
      "step": 1930
    },
    {
      "entropy": 6.10584077835083,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998750188817568,
      "loss": 6.1502,
      "mean_token_accuracy": 0.14796533063054085,
      "num_tokens": 4318297.0,
      "step": 1935
    },
    {
      "entropy": 6.090063381195068,
      "epoch": 0.21789184028752737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.979,
      "mean_token_accuracy": 0.15715189576148986,
      "num_tokens": 4328795.0,
      "step": 1940
    },
    {
      "entropy": 6.042044544219971,
      "epoch": 0.21845341719548492,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499872328554181,
      "loss": 5.9812,
      "mean_token_accuracy": 0.15147724151611328,
      "num_tokens": 4340279.0,
      "step": 1945
    },
    {
      "entropy": 6.2164843559265135,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998709726503383,
      "loss": 6.1623,
      "mean_token_accuracy": 0.1436479665338993,
      "num_tokens": 4352157.0,
      "step": 1950
    },
    {
      "entropy": 6.175703525543213,
      "epoch": 0.21957657101140002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998696095865168,
      "loss": 6.0064,
      "mean_token_accuracy": 0.1556989625096321,
      "num_tokens": 4361909.0,
      "step": 1955
    },
    {
      "entropy": 5.869294500350952,
      "epoch": 0.22013814791935757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.9129,
      "mean_token_accuracy": 0.15484603345394135,
      "num_tokens": 4372177.0,
      "step": 1960
    },
    {
      "entropy": 6.039911985397339,
      "epoch": 0.2206997248273151,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.982,
      "mean_token_accuracy": 0.1563266322016716,
      "num_tokens": 4382872.0,
      "step": 1965
    },
    {
      "entropy": 6.159225559234619,
      "epoch": 0.22126130173527264,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998654774356131,
      "loss": 6.0804,
      "mean_token_accuracy": 0.1488205224275589,
      "num_tokens": 4393623.0,
      "step": 1970
    },
    {
      "entropy": 6.015854930877685,
      "epoch": 0.2218228786432302,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998640857323118,
      "loss": 6.0076,
      "mean_token_accuracy": 0.14695556685328484,
      "num_tokens": 4404215.0,
      "step": 1975
    },
    {
      "entropy": 6.244098138809204,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998626868692509,
      "loss": 6.2583,
      "mean_token_accuracy": 0.14658438861370088,
      "num_tokens": 4415978.0,
      "step": 1980
    },
    {
      "entropy": 6.135475778579712,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998612808464748,
      "loss": 6.0107,
      "mean_token_accuracy": 0.15329247564077378,
      "num_tokens": 4427060.0,
      "step": 1985
    },
    {
      "entropy": 6.063277769088745,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998598676640283,
      "loss": 6.0722,
      "mean_token_accuracy": 0.14928158596158028,
      "num_tokens": 4438145.0,
      "step": 1990
    },
    {
      "entropy": 6.266687870025635,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.1493,
      "mean_token_accuracy": 0.15549051016569138,
      "num_tokens": 4449324.0,
      "step": 1995
    },
    {
      "entropy": 6.027319240570068,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.8902,
      "mean_token_accuracy": 0.15551431477069855,
      "num_tokens": 4459728.0,
      "step": 2000
    },
    {
      "entropy": 6.019188499450683,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998555851591175,
      "loss": 6.1022,
      "mean_token_accuracy": 0.15041673630476,
      "num_tokens": 4471164.0,
      "step": 2005
    },
    {
      "entropy": 6.231358623504638,
      "epoch": 0.225753916998933,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.1488,
      "mean_token_accuracy": 0.15111713707447053,
      "num_tokens": 4483349.0,
      "step": 2010
    },
    {
      "entropy": 6.196571016311646,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998526943583226,
      "loss": 6.1081,
      "mean_token_accuracy": 0.1539139837026596,
      "num_tokens": 4495311.0,
      "step": 2015
    },
    {
      "entropy": 6.084535074234009,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998512382188066,
      "loss": 6.0404,
      "mean_token_accuracy": 0.15315846651792525,
      "num_tokens": 4507017.0,
      "step": 2020
    },
    {
      "entropy": 6.096588039398194,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998497749199401,
      "loss": 6.0761,
      "mean_token_accuracy": 0.15145348012447357,
      "num_tokens": 4517746.0,
      "step": 2025
    },
    {
      "entropy": 6.213549852371216,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998483044617694,
      "loss": 6.1664,
      "mean_token_accuracy": 0.145511694252491,
      "num_tokens": 4529149.0,
      "step": 2030
    },
    {
      "entropy": 6.154675054550171,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998468268443416,
      "loss": 6.051,
      "mean_token_accuracy": 0.15400808304548264,
      "num_tokens": 4540630.0,
      "step": 2035
    },
    {
      "entropy": 6.01413254737854,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.9758,
      "mean_token_accuracy": 0.1529553934931755,
      "num_tokens": 4552877.0,
      "step": 2040
    },
    {
      "entropy": 6.154169607162475,
      "epoch": 0.22968495535463582,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998438501319025,
      "loss": 6.0693,
      "mean_token_accuracy": 0.15515930503606795,
      "num_tokens": 4563634.0,
      "step": 2045
    },
    {
      "entropy": 6.079659175872803,
      "epoch": 0.23024653226259337,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998423510369862,
      "loss": 6.0185,
      "mean_token_accuracy": 0.16131191849708557,
      "num_tokens": 4574007.0,
      "step": 2050
    },
    {
      "entropy": 6.063276243209839,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998408447830022,
      "loss": 6.0141,
      "mean_token_accuracy": 0.1526772379875183,
      "num_tokens": 4584146.0,
      "step": 2055
    },
    {
      "entropy": 6.107676076889038,
      "epoch": 0.23136968607850844,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998393313699985,
      "loss": 6.0371,
      "mean_token_accuracy": 0.1490740567445755,
      "num_tokens": 4594839.0,
      "step": 2060
    },
    {
      "entropy": 6.250979852676392,
      "epoch": 0.231931262986466,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998378107980233,
      "loss": 6.1898,
      "mean_token_accuracy": 0.14502785950899125,
      "num_tokens": 4608229.0,
      "step": 2065
    },
    {
      "entropy": 6.109724712371826,
      "epoch": 0.23249283989442354,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998362830671249,
      "loss": 5.9936,
      "mean_token_accuracy": 0.15285840034484863,
      "num_tokens": 4619206.0,
      "step": 2070
    },
    {
      "entropy": 6.003591156005859,
      "epoch": 0.2330544168023811,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998347481773522,
      "loss": 6.0237,
      "mean_token_accuracy": 0.15680260062217713,
      "num_tokens": 4628966.0,
      "step": 2075
    },
    {
      "entropy": 6.083307552337646,
      "epoch": 0.23361599371033864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998332061287538,
      "loss": 5.9268,
      "mean_token_accuracy": 0.1513105571269989,
      "num_tokens": 4639935.0,
      "step": 2080
    },
    {
      "entropy": 6.155953454971313,
      "epoch": 0.23417757061829617,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499831656921379,
      "loss": 6.1565,
      "mean_token_accuracy": 0.14754048883914947,
      "num_tokens": 4651998.0,
      "step": 2085
    },
    {
      "entropy": 6.174464988708496,
      "epoch": 0.23473914752625372,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499830100555277,
      "loss": 6.1424,
      "mean_token_accuracy": 0.1461272232234478,
      "num_tokens": 4664002.0,
      "step": 2090
    },
    {
      "entropy": 6.1856427669525145,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998285370304974,
      "loss": 6.0264,
      "mean_token_accuracy": 0.15473159700632094,
      "num_tokens": 4675343.0,
      "step": 2095
    },
    {
      "entropy": 6.077418518066406,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00049982696634709,
      "loss": 6.0756,
      "mean_token_accuracy": 0.15717330425977707,
      "num_tokens": 4686410.0,
      "step": 2100
    },
    {
      "entropy": 6.082688808441162,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998253885051048,
      "loss": 6.0996,
      "mean_token_accuracy": 0.15195055305957794,
      "num_tokens": 4697768.0,
      "step": 2105
    },
    {
      "entropy": 6.22024130821228,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499823803504592,
      "loss": 6.1022,
      "mean_token_accuracy": 0.1504859909415245,
      "num_tokens": 4709933.0,
      "step": 2110
    },
    {
      "entropy": 6.151685571670532,
      "epoch": 0.23754703206604144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998222113456022,
      "loss": 6.1375,
      "mean_token_accuracy": 0.14684750735759736,
      "num_tokens": 4721385.0,
      "step": 2115
    },
    {
      "entropy": 6.061194324493409,
      "epoch": 0.238108608973999,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998206120281858,
      "loss": 5.956,
      "mean_token_accuracy": 0.1638572856783867,
      "num_tokens": 4732562.0,
      "step": 2120
    },
    {
      "entropy": 6.022126531600952,
      "epoch": 0.23867018588195654,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499819005552394,
      "loss": 5.9696,
      "mean_token_accuracy": 0.15314062237739562,
      "num_tokens": 4742794.0,
      "step": 2125
    },
    {
      "entropy": 5.996016979217529,
      "epoch": 0.23923176278991407,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.934,
      "mean_token_accuracy": 0.1606994539499283,
      "num_tokens": 4754903.0,
      "step": 2130
    },
    {
      "entropy": 6.060186147689819,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499815771125889,
      "loss": 5.9783,
      "mean_token_accuracy": 0.15231975466012954,
      "num_tokens": 4766165.0,
      "step": 2135
    },
    {
      "entropy": 6.0076862335205075,
      "epoch": 0.24035491660582917,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998141431752785,
      "loss": 6.0022,
      "mean_token_accuracy": 0.1594335064291954,
      "num_tokens": 4776551.0,
      "step": 2140
    },
    {
      "entropy": 6.193838834762573,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998125080664987,
      "loss": 6.1794,
      "mean_token_accuracy": 0.14647017419338226,
      "num_tokens": 4788344.0,
      "step": 2145
    },
    {
      "entropy": 6.1166502952575685,
      "epoch": 0.24147807042174427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.9985,
      "mean_token_accuracy": 0.1459207035601139,
      "num_tokens": 4799314.0,
      "step": 2150
    },
    {
      "entropy": 6.083969259262085,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998092163746387,
      "loss": 6.0212,
      "mean_token_accuracy": 0.15580957233905793,
      "num_tokens": 4809588.0,
      "step": 2155
    },
    {
      "entropy": 5.976114320755005,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.8474,
      "mean_token_accuracy": 0.16323217898607253,
      "num_tokens": 4820076.0,
      "step": 2160
    },
    {
      "entropy": 6.061397123336792,
      "epoch": 0.2431628011456169,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998058960507286,
      "loss": 6.1208,
      "mean_token_accuracy": 0.14964161813259125,
      "num_tokens": 4831504.0,
      "step": 2165
    },
    {
      "entropy": 6.107447004318237,
      "epoch": 0.24372437805357444,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.9668,
      "mean_token_accuracy": 0.15928057283163072,
      "num_tokens": 4842826.0,
      "step": 2170
    },
    {
      "entropy": 5.872031927108765,
      "epoch": 0.244285954961532,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998025470951908,
      "loss": 5.9572,
      "mean_token_accuracy": 0.15852829664945603,
      "num_tokens": 4853279.0,
      "step": 2175
    },
    {
      "entropy": 6.157621288299561,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998008618806949,
      "loss": 6.0044,
      "mean_token_accuracy": 0.16119896918535231,
      "num_tokens": 4863332.0,
      "step": 2180
    },
    {
      "entropy": 6.080954504013062,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997991695084523,
      "loss": 6.1308,
      "mean_token_accuracy": 0.14659972339868546,
      "num_tokens": 4874217.0,
      "step": 2185
    },
    {
      "entropy": 6.237659978866577,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997974699785169,
      "loss": 6.0082,
      "mean_token_accuracy": 0.16113914996385575,
      "num_tokens": 4885322.0,
      "step": 2190
    },
    {
      "entropy": 5.919554424285889,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.9381,
      "mean_token_accuracy": 0.15738066136837006,
      "num_tokens": 4895546.0,
      "step": 2195
    },
    {
      "entropy": 6.218607234954834,
      "epoch": 0.24709383950131972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997940494457846,
      "loss": 6.154,
      "mean_token_accuracy": 0.14981306791305543,
      "num_tokens": 4906957.0,
      "step": 2200
    },
    {
      "entropy": 6.19651517868042,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997923284430967,
      "loss": 6.0821,
      "mean_token_accuracy": 0.14576388597488404,
      "num_tokens": 4917965.0,
      "step": 2205
    },
    {
      "entropy": 6.0613480567932125,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.9757,
      "mean_token_accuracy": 0.15590714514255524,
      "num_tokens": 4928585.0,
      "step": 2210
    },
    {
      "entropy": 5.968810176849365,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000499788864965351,
      "loss": 5.9858,
      "mean_token_accuracy": 0.1578755110502243,
      "num_tokens": 4939818.0,
      "step": 2215
    },
    {
      "entropy": 6.131110334396363,
      "epoch": 0.2493401471331499,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997871224904036,
      "loss": 6.0622,
      "mean_token_accuracy": 0.14955551028251649,
      "num_tokens": 4951752.0,
      "step": 2220
    },
    {
      "entropy": 5.9506998538970945,
      "epoch": 0.24990172404110744,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.8901,
      "mean_token_accuracy": 0.16224861592054368,
      "num_tokens": 4961846.0,
      "step": 2225
    },
    {
      "entropy": 6.0852518558502195,
      "epoch": 0.250463300949065,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997836160686368,
      "loss": 6.0521,
      "mean_token_accuracy": 0.15131033658981324,
      "num_tokens": 4973693.0,
      "step": 2230
    },
    {
      "entropy": 6.060042428970337,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.9604,
      "mean_token_accuracy": 0.1576952561736107,
      "num_tokens": 4986156.0,
      "step": 2235
    },
    {
      "entropy": 6.013088703155518,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997800810180803,
      "loss": 6.0447,
      "mean_token_accuracy": 0.15376421064138412,
      "num_tokens": 4997615.0,
      "step": 2240
    },
    {
      "entropy": 6.1519105434417725,
      "epoch": 0.2521480316729376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.9679,
      "mean_token_accuracy": 0.15947780162096023,
      "num_tokens": 5007655.0,
      "step": 2245
    },
    {
      "entropy": 6.026669692993164,
      "epoch": 0.25270960858089514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997765173391841,
      "loss": 6.0482,
      "mean_token_accuracy": 0.15224288105964662,
      "num_tokens": 5018933.0,
      "step": 2250
    },
    {
      "entropy": 6.166677522659302,
      "epoch": 0.2532711854888527,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997747247642505,
      "loss": 6.0525,
      "mean_token_accuracy": 0.15358851104974747,
      "num_tokens": 5030495.0,
      "step": 2255
    },
    {
      "entropy": 6.102503728866577,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997729250324023,
      "loss": 6.0817,
      "mean_token_accuracy": 0.1501933827996254,
      "num_tokens": 5041553.0,
      "step": 2260
    },
    {
      "entropy": 6.090739488601685,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997711181436971,
      "loss": 6.007,
      "mean_token_accuracy": 0.15912422835826873,
      "num_tokens": 5052312.0,
      "step": 2265
    },
    {
      "entropy": 6.151770973205567,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997693040981922,
      "loss": 6.1953,
      "mean_token_accuracy": 0.14825651198625564,
      "num_tokens": 5063445.0,
      "step": 2270
    },
    {
      "entropy": 6.138964891433716,
      "epoch": 0.2555174931206829,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.9824,
      "mean_token_accuracy": 0.15490673780441283,
      "num_tokens": 5073052.0,
      "step": 2275
    },
    {
      "entropy": 6.038366365432739,
      "epoch": 0.25607907002864044,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997656545370151,
      "loss": 6.0267,
      "mean_token_accuracy": 0.15771733224391937,
      "num_tokens": 5084174.0,
      "step": 2280
    },
    {
      "entropy": 6.103801107406616,
      "epoch": 0.256640646936598,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499763819021459,
      "loss": 6.0128,
      "mean_token_accuracy": 0.15576983839273453,
      "num_tokens": 5094106.0,
      "step": 2285
    },
    {
      "entropy": 6.129735326766967,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997619763493357,
      "loss": 6.0902,
      "mean_token_accuracy": 0.15617480129003525,
      "num_tokens": 5106963.0,
      "step": 2290
    },
    {
      "entropy": 5.987654972076416,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499760126520704,
      "loss": 5.8927,
      "mean_token_accuracy": 0.15116089582443237,
      "num_tokens": 5119476.0,
      "step": 2295
    },
    {
      "entropy": 6.157898807525635,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997582695356225,
      "loss": 6.1371,
      "mean_token_accuracy": 0.15195467919111252,
      "num_tokens": 5132169.0,
      "step": 2300
    },
    {
      "entropy": 6.094862699508667,
      "epoch": 0.25888695456842814,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.9806,
      "mean_token_accuracy": 0.1550524652004242,
      "num_tokens": 5143354.0,
      "step": 2305
    },
    {
      "entropy": 5.958035707473755,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997545340963478,
      "loss": 6.0541,
      "mean_token_accuracy": 0.1468196243047714,
      "num_tokens": 5154624.0,
      "step": 2310
    },
    {
      "entropy": 6.126406049728393,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997526556422733,
      "loss": 5.9392,
      "mean_token_accuracy": 0.15484741032123567,
      "num_tokens": 5165349.0,
      "step": 2315
    },
    {
      "entropy": 6.043291521072388,
      "epoch": 0.2605716852923008,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997507700319872,
      "loss": 6.0235,
      "mean_token_accuracy": 0.15324235707521439,
      "num_tokens": 5175874.0,
      "step": 2320
    },
    {
      "entropy": 6.055290842056275,
      "epoch": 0.26113326220025834,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997488772655492,
      "loss": 5.9108,
      "mean_token_accuracy": 0.16078570783138274,
      "num_tokens": 5186591.0,
      "step": 2325
    },
    {
      "entropy": 5.993036842346191,
      "epoch": 0.2616948391082159,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997469773430199,
      "loss": 6.0215,
      "mean_token_accuracy": 0.1626984179019928,
      "num_tokens": 5197307.0,
      "step": 2330
    },
    {
      "entropy": 6.009312629699707,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.9781,
      "mean_token_accuracy": 0.15396428406238555,
      "num_tokens": 5209868.0,
      "step": 2335
    },
    {
      "entropy": 6.020443344116211,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.9995,
      "mean_token_accuracy": 0.15726363211870192,
      "num_tokens": 5220365.0,
      "step": 2340
    },
    {
      "entropy": 6.102507829666138,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.9229,
      "mean_token_accuracy": 0.15023885518312455,
      "num_tokens": 5231575.0,
      "step": 2345
    },
    {
      "entropy": 6.035394668579102,
      "epoch": 0.26394114674004604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.9938,
      "mean_token_accuracy": 0.15539337694644928,
      "num_tokens": 5241852.0,
      "step": 2350
    },
    {
      "entropy": 6.038948059082031,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997373703911273,
      "loss": 6.0288,
      "mean_token_accuracy": 0.15632582753896712,
      "num_tokens": 5254536.0,
      "step": 2355
    },
    {
      "entropy": 6.083648633956909,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.9606,
      "mean_token_accuracy": 0.15450441390275954,
      "num_tokens": 5264946.0,
      "step": 2360
    },
    {
      "entropy": 5.8433245658874515,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499733477519865,
      "loss": 5.784,
      "mean_token_accuracy": 0.15305204540491105,
      "num_tokens": 5275527.0,
      "step": 2365
    },
    {
      "entropy": 6.025034093856812,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997315203508013,
      "loss": 5.9569,
      "mean_token_accuracy": 0.15324413478374482,
      "num_tokens": 5285471.0,
      "step": 2370
    },
    {
      "entropy": 5.916028690338135,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.89,
      "mean_token_accuracy": 0.15359498262405397,
      "num_tokens": 5295743.0,
      "step": 2375
    },
    {
      "entropy": 6.049281930923462,
      "epoch": 0.26731060818779134,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049972758454612,
      "loss": 6.0335,
      "mean_token_accuracy": 0.1482474312186241,
      "num_tokens": 5307963.0,
      "step": 2380
    },
    {
      "entropy": 6.043624353408814,
      "epoch": 0.26787218509574884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.9176,
      "mean_token_accuracy": 0.15914819538593292,
      "num_tokens": 5318394.0,
      "step": 2385
    },
    {
      "entropy": 6.079010963439941,
      "epoch": 0.2684337620037064,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997236201197853,
      "loss": 6.144,
      "mean_token_accuracy": 0.1458563432097435,
      "num_tokens": 5331728.0,
      "step": 2390
    },
    {
      "entropy": 6.132135200500488,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997216271736554,
      "loss": 6.0232,
      "mean_token_accuracy": 0.16378798931837082,
      "num_tokens": 5342825.0,
      "step": 2395
    },
    {
      "entropy": 6.050425243377686,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.9357,
      "mean_token_accuracy": 0.15482408478856086,
      "num_tokens": 5354232.0,
      "step": 2400
    },
    {
      "entropy": 6.0575597286224365,
      "epoch": 0.27011849272757904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997176198157881,
      "loss": 6.0253,
      "mean_token_accuracy": 0.1507862962782383,
      "num_tokens": 5364807.0,
      "step": 2405
    },
    {
      "entropy": 6.040148830413818,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.9844,
      "mean_token_accuracy": 0.15217866897583007,
      "num_tokens": 5375593.0,
      "step": 2410
    },
    {
      "entropy": 6.045368623733521,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.9655,
      "mean_token_accuracy": 0.15985994040966034,
      "num_tokens": 5386740.0,
      "step": 2415
    },
    {
      "entropy": 5.910307359695435,
      "epoch": 0.2718032234514517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.9274,
      "mean_token_accuracy": 0.15814046263694764,
      "num_tokens": 5397732.0,
      "step": 2420
    },
    {
      "entropy": 6.159975671768189,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499709519239414,
      "loss": 5.9925,
      "mean_token_accuracy": 0.14930515587329865,
      "num_tokens": 5409392.0,
      "step": 2425
    },
    {
      "entropy": 5.926601505279541,
      "epoch": 0.2729263772673668,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.9607,
      "mean_token_accuracy": 0.16299881041049957,
      "num_tokens": 5420903.0,
      "step": 2430
    },
    {
      "entropy": 5.892569589614868,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499705426021939,
      "loss": 5.8882,
      "mean_token_accuracy": 0.15828869342803956,
      "num_tokens": 5432631.0,
      "step": 2435
    },
    {
      "entropy": 6.01947603225708,
      "epoch": 0.27404953108328184,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499703368681107,
      "loss": 5.8652,
      "mean_token_accuracy": 0.16353946775197983,
      "num_tokens": 5443463.0,
      "step": 2440
    },
    {
      "entropy": 6.1192179203033445,
      "epoch": 0.2746111079912394,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997013041856326,
      "loss": 6.1688,
      "mean_token_accuracy": 0.15179817527532577,
      "num_tokens": 5455903.0,
      "step": 2445
    },
    {
      "entropy": 6.088233995437622,
      "epoch": 0.27517268489919694,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.9814,
      "mean_token_accuracy": 0.15265644788742067,
      "num_tokens": 5467201.0,
      "step": 2450
    },
    {
      "entropy": 5.960464429855347,
      "epoch": 0.2757342618071545,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.8577,
      "mean_token_accuracy": 0.15811959505081177,
      "num_tokens": 5478109.0,
      "step": 2455
    },
    {
      "entropy": 6.074854564666748,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.9994,
      "mean_token_accuracy": 0.15298570096492767,
      "num_tokens": 5489621.0,
      "step": 2460
    },
    {
      "entropy": 5.961808776855468,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499692974658629,
      "loss": 5.9402,
      "mean_token_accuracy": 0.15553341507911683,
      "num_tokens": 5500606.0,
      "step": 2465
    },
    {
      "entropy": 5.985595941543579,
      "epoch": 0.27741899253102714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.8963,
      "mean_token_accuracy": 0.16250037103891374,
      "num_tokens": 5510768.0,
      "step": 2470
    },
    {
      "entropy": 5.943635702133179,
      "epoch": 0.2779805694389847,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.9121,
      "mean_token_accuracy": 0.15700624138116837,
      "num_tokens": 5522210.0,
      "step": 2475
    },
    {
      "entropy": 6.043688678741455,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996866523928747,
      "loss": 6.0396,
      "mean_token_accuracy": 0.1516438513994217,
      "num_tokens": 5533055.0,
      "step": 2480
    },
    {
      "entropy": 6.1347261428833,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996845306626467,
      "loss": 6.0314,
      "mean_token_accuracy": 0.15147426277399062,
      "num_tokens": 5544345.0,
      "step": 2485
    },
    {
      "entropy": 6.014614534378052,
      "epoch": 0.2796653001628573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.9502,
      "mean_token_accuracy": 0.15230564698576926,
      "num_tokens": 5556734.0,
      "step": 2490
    },
    {
      "entropy": 6.008048486709595,
      "epoch": 0.28022687707081484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.9865,
      "mean_token_accuracy": 0.15721939951181413,
      "num_tokens": 5566874.0,
      "step": 2495
    },
    {
      "entropy": 6.101916837692261,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.9791,
      "mean_token_accuracy": 0.15063416808843613,
      "num_tokens": 5578812.0,
      "step": 2500
    },
    {
      "entropy": 6.013814878463745,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996759722019872,
      "loss": 6.0271,
      "mean_token_accuracy": 0.15371579229831694,
      "num_tokens": 5590021.0,
      "step": 2505
    },
    {
      "entropy": 6.024594354629516,
      "epoch": 0.2819116077946875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996738147022262,
      "loss": 6.0567,
      "mean_token_accuracy": 0.14784704446792601,
      "num_tokens": 5600921.0,
      "step": 2510
    },
    {
      "entropy": 6.051389265060425,
      "epoch": 0.28247318470264504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.9445,
      "mean_token_accuracy": 0.15591129958629607,
      "num_tokens": 5610615.0,
      "step": 2515
    },
    {
      "entropy": 5.990168857574463,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499669478241669,
      "loss": 5.8949,
      "mean_token_accuracy": 0.15703389495611192,
      "num_tokens": 5621596.0,
      "step": 2520
    },
    {
      "entropy": 6.092303705215454,
      "epoch": 0.28359633851856014,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.9413,
      "mean_token_accuracy": 0.15879166424274443,
      "num_tokens": 5633047.0,
      "step": 2525
    },
    {
      "entropy": 5.900774335861206,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.9041,
      "mean_token_accuracy": 0.1561056926846504,
      "num_tokens": 5644970.0,
      "step": 2530
    },
    {
      "entropy": 5.969459009170532,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996629198992823,
      "loss": 6.0379,
      "mean_token_accuracy": 0.15508987456560136,
      "num_tokens": 5656631.0,
      "step": 2535
    },
    {
      "entropy": 6.097274971008301,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1543091729283333,
      "num_tokens": 5668808.0,
      "step": 2540
    },
    {
      "entropy": 5.94429612159729,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.9251,
      "mean_token_accuracy": 0.15240908712148665,
      "num_tokens": 5679493.0,
      "step": 2545
    },
    {
      "entropy": 6.034660053253174,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.9655,
      "mean_token_accuracy": 0.15858010351657867,
      "num_tokens": 5689879.0,
      "step": 2550
    },
    {
      "entropy": 6.085075855255127,
      "epoch": 0.2869657999663054,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499654075296133,
      "loss": 6.0574,
      "mean_token_accuracy": 0.15524982064962387,
      "num_tokens": 5699934.0,
      "step": 2555
    },
    {
      "entropy": 6.04909462928772,
      "epoch": 0.28752737687426294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499651846262487,
      "loss": 5.9431,
      "mean_token_accuracy": 0.1601421758532524,
      "num_tokens": 5710171.0,
      "step": 2560
    },
    {
      "entropy": 5.977365207672119,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.9509,
      "mean_token_accuracy": 0.15049331039190292,
      "num_tokens": 5721190.0,
      "step": 2565
    },
    {
      "entropy": 6.137782859802246,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996473667362604,
      "loss": 6.0564,
      "mean_token_accuracy": 0.15206752270460128,
      "num_tokens": 5733706.0,
      "step": 2570
    },
    {
      "entropy": 5.952969312667847,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996451162438224,
      "loss": 6.0291,
      "mean_token_accuracy": 0.1582716077566147,
      "num_tokens": 5745712.0,
      "step": 2575
    },
    {
      "entropy": 5.9377197265625,
      "epoch": 0.2897736845060931,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.8341,
      "mean_token_accuracy": 0.167465503513813,
      "num_tokens": 5757915.0,
      "step": 2580
    },
    {
      "entropy": 6.0136088848114015,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.983,
      "mean_token_accuracy": 0.1571892872452736,
      "num_tokens": 5768788.0,
      "step": 2585
    },
    {
      "entropy": 6.062278985977173,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.9163,
      "mean_token_accuracy": 0.15861130356788636,
      "num_tokens": 5779806.0,
      "step": 2590
    },
    {
      "entropy": 5.848001956939697,
      "epoch": 0.29145841522996574,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499636042746914,
      "loss": 5.8739,
      "mean_token_accuracy": 0.1530233860015869,
      "num_tokens": 5790177.0,
      "step": 2595
    },
    {
      "entropy": 6.030651712417603,
      "epoch": 0.2920199921379233,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499633756491259,
      "loss": 5.9058,
      "mean_token_accuracy": 0.16201242804527283,
      "num_tokens": 5800251.0,
      "step": 2600
    },
    {
      "entropy": 6.013105869293213,
      "epoch": 0.29258156904588084,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996314630831781,
      "loss": 6.0029,
      "mean_token_accuracy": 0.1577301412820816,
      "num_tokens": 5812210.0,
      "step": 2605
    },
    {
      "entropy": 6.03218035697937,
      "epoch": 0.2931431459538384,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.972,
      "mean_token_accuracy": 0.15995176434516906,
      "num_tokens": 5823280.0,
      "step": 2610
    },
    {
      "entropy": 6.098472261428833,
      "epoch": 0.29370472286179594,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996268548100307,
      "loss": 6.0516,
      "mean_token_accuracy": 0.14447705671191216,
      "num_tokens": 5835076.0,
      "step": 2615
    },
    {
      "entropy": 5.97599425315857,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499624539945111,
      "loss": 5.8353,
      "mean_token_accuracy": 0.15990931689739227,
      "num_tokens": 5845376.0,
      "step": 2620
    },
    {
      "entropy": 5.9088949203491214,
      "epoch": 0.294827876677711,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.8904,
      "mean_token_accuracy": 0.15515819191932678,
      "num_tokens": 5857124.0,
      "step": 2625
    },
    {
      "entropy": 6.012001180648804,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.9792,
      "mean_token_accuracy": 0.14992934688925744,
      "num_tokens": 5868681.0,
      "step": 2630
    },
    {
      "entropy": 6.072331762313842,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.9727,
      "mean_token_accuracy": 0.1609386384487152,
      "num_tokens": 5880532.0,
      "step": 2635
    },
    {
      "entropy": 5.996478271484375,
      "epoch": 0.29651260740158364,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499615208964848,
      "loss": 5.9347,
      "mean_token_accuracy": 0.16165445894002914,
      "num_tokens": 5891582.0,
      "step": 2640
    },
    {
      "entropy": 5.978861856460571,
      "epoch": 0.2970741843095412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.932,
      "mean_token_accuracy": 0.150858586281538,
      "num_tokens": 5903543.0,
      "step": 2645
    },
    {
      "entropy": 5.955687618255615,
      "epoch": 0.29763576121749874,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.9003,
      "mean_token_accuracy": 0.1639521986246109,
      "num_tokens": 5913140.0,
      "step": 2650
    },
    {
      "entropy": 5.951510906219482,
      "epoch": 0.2981973381254563,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.87,
      "mean_token_accuracy": 0.15955302566289903,
      "num_tokens": 5924010.0,
      "step": 2655
    },
    {
      "entropy": 6.063926649093628,
      "epoch": 0.29875891503341384,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996057635552237,
      "loss": 6.0587,
      "mean_token_accuracy": 0.15360975861549378,
      "num_tokens": 5936246.0,
      "step": 2660
    },
    {
      "entropy": 6.02715482711792,
      "epoch": 0.2993204919413714,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.9615,
      "mean_token_accuracy": 0.1463753491640091,
      "num_tokens": 5947594.0,
      "step": 2665
    },
    {
      "entropy": 6.025208616256714,
      "epoch": 0.29988206884932894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996009979409025,
      "loss": 5.8914,
      "mean_token_accuracy": 0.1480000451207161,
      "num_tokens": 5957660.0,
      "step": 2670
    },
    {
      "entropy": 5.800934886932373,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.7274,
      "mean_token_accuracy": 0.16546161770820617,
      "num_tokens": 5968881.0,
      "step": 2675
    },
    {
      "entropy": 5.921378183364868,
      "epoch": 0.301005222665244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049959620372105,
      "loss": 5.9065,
      "mean_token_accuracy": 0.1579037681221962,
      "num_tokens": 5980032.0,
      "step": 2680
    },
    {
      "entropy": 5.919540977478027,
      "epoch": 0.30156679957320154,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.8937,
      "mean_token_accuracy": 0.15591801702976227,
      "num_tokens": 5991781.0,
      "step": 2685
    },
    {
      "entropy": 5.953354835510254,
      "epoch": 0.3021283764811591,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.948,
      "mean_token_accuracy": 0.15558021366596222,
      "num_tokens": 6002303.0,
      "step": 2690
    },
    {
      "entropy": 6.10141167640686,
      "epoch": 0.30268995338911664,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995889587572366,
      "loss": 6.0053,
      "mean_token_accuracy": 0.1546380952000618,
      "num_tokens": 6011523.0,
      "step": 2695
    },
    {
      "entropy": 5.984244871139526,
      "epoch": 0.3032515302970742,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499586529467197,
      "loss": 5.8236,
      "mean_token_accuracy": 0.15790919214487076,
      "num_tokens": 6022169.0,
      "step": 2700
    },
    {
      "entropy": 5.809388446807861,
      "epoch": 0.30381310720503174,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.8721,
      "mean_token_accuracy": 0.1557815633714199,
      "num_tokens": 6033393.0,
      "step": 2705
    },
    {
      "entropy": 6.070313596725464,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.9486,
      "mean_token_accuracy": 0.15179309844970704,
      "num_tokens": 6045365.0,
      "step": 2710
    },
    {
      "entropy": 5.969070339202881,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.9637,
      "mean_token_accuracy": 0.15632378160953522,
      "num_tokens": 6057482.0,
      "step": 2715
    },
    {
      "entropy": 5.966453695297242,
      "epoch": 0.30549783792890434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.8717,
      "mean_token_accuracy": 0.16056556701660157,
      "num_tokens": 6068194.0,
      "step": 2720
    },
    {
      "entropy": 5.998284339904785,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.9356,
      "mean_token_accuracy": 0.1577038884162903,
      "num_tokens": 6080540.0,
      "step": 2725
    },
    {
      "entropy": 5.920602607727051,
      "epoch": 0.30662099174481944,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499571803560316,
      "loss": 5.9674,
      "mean_token_accuracy": 0.15675391405820846,
      "num_tokens": 6092464.0,
      "step": 2730
    },
    {
      "entropy": 6.051038122177124,
      "epoch": 0.307182568652777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.9221,
      "mean_token_accuracy": 0.1568444013595581,
      "num_tokens": 6103242.0,
      "step": 2735
    },
    {
      "entropy": 5.945160531997681,
      "epoch": 0.30774414556073454,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.9509,
      "mean_token_accuracy": 0.15690646022558213,
      "num_tokens": 6116383.0,
      "step": 2740
    },
    {
      "entropy": 5.832088613510132,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.7578,
      "mean_token_accuracy": 0.1716798350214958,
      "num_tokens": 6126883.0,
      "step": 2745
    },
    {
      "entropy": 6.002315855026245,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.9508,
      "mean_token_accuracy": 0.15082325488328935,
      "num_tokens": 6139613.0,
      "step": 2750
    },
    {
      "entropy": 6.019475936889648,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995593353330753,
      "loss": 6.0145,
      "mean_token_accuracy": 0.14947013407945633,
      "num_tokens": 6151950.0,
      "step": 2755
    },
    {
      "entropy": 5.9253801822662355,
      "epoch": 0.30999045319256474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.798,
      "mean_token_accuracy": 0.1644776478409767,
      "num_tokens": 6161563.0,
      "step": 2760
    },
    {
      "entropy": 5.855195665359497,
      "epoch": 0.3105520301005223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.8162,
      "mean_token_accuracy": 0.16404712200164795,
      "num_tokens": 6170966.0,
      "step": 2765
    },
    {
      "entropy": 5.913083553314209,
      "epoch": 0.3111136070084798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.888,
      "mean_token_accuracy": 0.15659398585557938,
      "num_tokens": 6182563.0,
      "step": 2770
    },
    {
      "entropy": 6.013282346725464,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.9971,
      "mean_token_accuracy": 0.15299334228038788,
      "num_tokens": 6194495.0,
      "step": 2775
    },
    {
      "entropy": 6.030122756958008,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.895,
      "mean_token_accuracy": 0.1562565177679062,
      "num_tokens": 6206594.0,
      "step": 2780
    },
    {
      "entropy": 5.946122312545777,
      "epoch": 0.31279833773235244,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.9247,
      "mean_token_accuracy": 0.15526296347379684,
      "num_tokens": 6218371.0,
      "step": 2785
    },
    {
      "entropy": 6.001246643066406,
      "epoch": 0.31335991464031,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.8648,
      "mean_token_accuracy": 0.15969813764095306,
      "num_tokens": 6228723.0,
      "step": 2790
    },
    {
      "entropy": 5.9897809505462645,
      "epoch": 0.31392149154826754,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995390143707665,
      "loss": 6.0495,
      "mean_token_accuracy": 0.15685173273086547,
      "num_tokens": 6240535.0,
      "step": 2795
    },
    {
      "entropy": 5.88944444656372,
      "epoch": 0.3144830684562251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.7423,
      "mean_token_accuracy": 0.16546466201543808,
      "num_tokens": 6250643.0,
      "step": 2800
    },
    {
      "entropy": 5.891298627853393,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.8873,
      "mean_token_accuracy": 0.16037205457687378,
      "num_tokens": 6259569.0,
      "step": 2805
    },
    {
      "entropy": 6.019705200195313,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499531276042413,
      "loss": 6.0154,
      "mean_token_accuracy": 0.16337641030550004,
      "num_tokens": 6271108.0,
      "step": 2810
    },
    {
      "entropy": 6.061544942855835,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.8578,
      "mean_token_accuracy": 0.15389910191297532,
      "num_tokens": 6282352.0,
      "step": 2815
    },
    {
      "entropy": 5.868814754486084,
      "epoch": 0.31672937608805524,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.9603,
      "mean_token_accuracy": 0.157982175052166,
      "num_tokens": 6294946.0,
      "step": 2820
    },
    {
      "entropy": 5.909223985671997,
      "epoch": 0.3172909529960128,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.7985,
      "mean_token_accuracy": 0.1622636839747429,
      "num_tokens": 6305552.0,
      "step": 2825
    },
    {
      "entropy": 5.955548286437988,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.8637,
      "mean_token_accuracy": 0.1647718846797943,
      "num_tokens": 6315815.0,
      "step": 2830
    },
    {
      "entropy": 5.898680114746094,
      "epoch": 0.3184141068119279,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.8973,
      "mean_token_accuracy": 0.1575847841799259,
      "num_tokens": 6327384.0,
      "step": 2835
    },
    {
      "entropy": 5.998651695251465,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.8787,
      "mean_token_accuracy": 0.15362707823514937,
      "num_tokens": 6338083.0,
      "step": 2840
    },
    {
      "entropy": 6.001203966140747,
      "epoch": 0.319537260627843,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.8944,
      "mean_token_accuracy": 0.16161568313837052,
      "num_tokens": 6349058.0,
      "step": 2845
    },
    {
      "entropy": 5.875576114654541,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.9006,
      "mean_token_accuracy": 0.16204904466867448,
      "num_tokens": 6359655.0,
      "step": 2850
    },
    {
      "entropy": 6.017832040786743,
      "epoch": 0.3206604144437581,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995076750115258,
      "loss": 6.027,
      "mean_token_accuracy": 0.15353279262781144,
      "num_tokens": 6372088.0,
      "step": 2855
    },
    {
      "entropy": 5.99152102470398,
      "epoch": 0.32122199135171564,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.9362,
      "mean_token_accuracy": 0.15284417122602462,
      "num_tokens": 6382946.0,
      "step": 2860
    },
    {
      "entropy": 6.00714955329895,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.9206,
      "mean_token_accuracy": 0.15483998730778695,
      "num_tokens": 6392591.0,
      "step": 2865
    },
    {
      "entropy": 5.8733357906341555,
      "epoch": 0.3223451451676307,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.8586,
      "mean_token_accuracy": 0.16367374062538148,
      "num_tokens": 6403159.0,
      "step": 2870
    },
    {
      "entropy": 5.974071979522705,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.88,
      "mean_token_accuracy": 0.15994496941566466,
      "num_tokens": 6414232.0,
      "step": 2875
    },
    {
      "entropy": 5.964216232299805,
      "epoch": 0.3234682989835458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.8534,
      "mean_token_accuracy": 0.16927893757820128,
      "num_tokens": 6425762.0,
      "step": 2880
    },
    {
      "entropy": 5.935434436798095,
      "epoch": 0.32402987589150334,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.9197,
      "mean_token_accuracy": 0.1512233465909958,
      "num_tokens": 6439077.0,
      "step": 2885
    },
    {
      "entropy": 6.02995548248291,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.9628,
      "mean_token_accuracy": 0.1587122306227684,
      "num_tokens": 6450425.0,
      "step": 2890
    },
    {
      "entropy": 5.9784949779510494,
      "epoch": 0.32515302970741844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.9359,
      "mean_token_accuracy": 0.15752865448594094,
      "num_tokens": 6461347.0,
      "step": 2895
    },
    {
      "entropy": 6.020418548583985,
      "epoch": 0.325714606615376,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.924,
      "mean_token_accuracy": 0.1578624352812767,
      "num_tokens": 6472428.0,
      "step": 2900
    },
    {
      "entropy": 5.879127836227417,
      "epoch": 0.32627618352333354,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.8904,
      "mean_token_accuracy": 0.16676969081163406,
      "num_tokens": 6483393.0,
      "step": 2905
    },
    {
      "entropy": 5.935239315032959,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.8648,
      "mean_token_accuracy": 0.16207789927721022,
      "num_tokens": 6493520.0,
      "step": 2910
    },
    {
      "entropy": 5.898339509963989,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.7977,
      "mean_token_accuracy": 0.16786165684461593,
      "num_tokens": 6505282.0,
      "step": 2915
    },
    {
      "entropy": 5.955675268173218,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.971,
      "mean_token_accuracy": 0.16191720515489577,
      "num_tokens": 6516808.0,
      "step": 2920
    },
    {
      "entropy": 5.991767835617066,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.9314,
      "mean_token_accuracy": 0.1532451778650284,
      "num_tokens": 6529184.0,
      "step": 2925
    },
    {
      "entropy": 5.945623350143433,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.7807,
      "mean_token_accuracy": 0.1646365851163864,
      "num_tokens": 6539909.0,
      "step": 2930
    },
    {
      "entropy": 5.8691150665283205,
      "epoch": 0.3296456449710788,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.8382,
      "mean_token_accuracy": 0.16271614283323288,
      "num_tokens": 6550870.0,
      "step": 2935
    },
    {
      "entropy": 5.8914587020874025,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499461515538303,
      "loss": 5.8833,
      "mean_token_accuracy": 0.1637248530983925,
      "num_tokens": 6561784.0,
      "step": 2940
    },
    {
      "entropy": 5.8711829662323,
      "epoch": 0.3307687987869939,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.8381,
      "mean_token_accuracy": 0.15968684256076812,
      "num_tokens": 6573422.0,
      "step": 2945
    },
    {
      "entropy": 6.02776026725769,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499455949211845,
      "loss": 5.9274,
      "mean_token_accuracy": 0.16189566254615784,
      "num_tokens": 6584565.0,
      "step": 2950
    },
    {
      "entropy": 5.9975940704345705,
      "epoch": 0.331891952602909,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.823,
      "mean_token_accuracy": 0.16222766041755676,
      "num_tokens": 6594238.0,
      "step": 2955
    },
    {
      "entropy": 5.858269691467285,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.8909,
      "mean_token_accuracy": 0.15825014412403107,
      "num_tokens": 6606309.0,
      "step": 2960
    },
    {
      "entropy": 6.004867458343506,
      "epoch": 0.33301510641882404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.987,
      "mean_token_accuracy": 0.15539744198322297,
      "num_tokens": 6617916.0,
      "step": 2965
    },
    {
      "entropy": 6.02716212272644,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.8425,
      "mean_token_accuracy": 0.160085965692997,
      "num_tokens": 6627951.0,
      "step": 2970
    },
    {
      "entropy": 5.868495655059815,
      "epoch": 0.33413826023473914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.9095,
      "mean_token_accuracy": 0.1606067419052124,
      "num_tokens": 6639285.0,
      "step": 2975
    },
    {
      "entropy": 5.955172061920166,
      "epoch": 0.3346998371426967,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.7658,
      "mean_token_accuracy": 0.17116498202085495,
      "num_tokens": 6649973.0,
      "step": 2980
    },
    {
      "entropy": 5.75142731666565,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.8012,
      "mean_token_accuracy": 0.16292239129543304,
      "num_tokens": 6660625.0,
      "step": 2985
    },
    {
      "entropy": 5.8972403049469,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.875,
      "learning_rate": 0.00049943339804258,
      "loss": 5.8733,
      "mean_token_accuracy": 0.1556589461863041,
      "num_tokens": 6673191.0,
      "step": 2990
    },
    {
      "entropy": 5.964262771606445,
      "epoch": 0.33638456786656934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.8682,
      "mean_token_accuracy": 0.1654741033911705,
      "num_tokens": 6683993.0,
      "step": 2995
    },
    {
      "entropy": 5.850085067749023,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.8926,
      "mean_token_accuracy": 0.16476777344942092,
      "num_tokens": 6694881.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.856326704514774,
      "eval_loss": 5.906490325927734,
      "eval_mean_token_accuracy": 0.1660596932999661,
      "eval_num_tokens": 6694881.0,
      "eval_runtime": 60.6974,
      "eval_samples_per_second": 510.944,
      "eval_steps_per_second": 63.874,
      "step": 3000
    },
    {
      "entropy": 6.009792947769165,
      "epoch": 0.33750772168248444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.8779,
      "mean_token_accuracy": 0.16287034898996353,
      "num_tokens": 6706648.0,
      "step": 3005
    },
    {
      "entropy": 5.946700239181519,
      "epoch": 0.33806929859044194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499421950949953,
      "loss": 5.8992,
      "mean_token_accuracy": 0.16301757097244263,
      "num_tokens": 6716858.0,
      "step": 3010
    },
    {
      "entropy": 5.883094263076782,
      "epoch": 0.3386308754983995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499419071312363,
      "loss": 5.8546,
      "mean_token_accuracy": 0.1635548561811447,
      "num_tokens": 6726558.0,
      "step": 3015
    },
    {
      "entropy": 5.962727642059326,
      "epoch": 0.33919245240635704,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.855,
      "mean_token_accuracy": 0.15484548956155778,
      "num_tokens": 6737819.0,
      "step": 3020
    },
    {
      "entropy": 5.8366251468658445,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.8073,
      "mean_token_accuracy": 0.16233350038528443,
      "num_tokens": 6748593.0,
      "step": 3025
    },
    {
      "entropy": 5.829312515258789,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.9023,
      "mean_token_accuracy": 0.15177217721939087,
      "num_tokens": 6760631.0,
      "step": 3030
    },
    {
      "entropy": 6.040865039825439,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.8109,
      "mean_token_accuracy": 0.1586655080318451,
      "num_tokens": 6772318.0,
      "step": 3035
    },
    {
      "entropy": 5.788278579711914,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.8562,
      "mean_token_accuracy": 0.16249461621046066,
      "num_tokens": 6782684.0,
      "step": 3040
    },
    {
      "entropy": 5.948060512542725,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.7292,
      "mean_token_accuracy": 0.16573673039674758,
      "num_tokens": 6793895.0,
      "step": 3045
    },
    {
      "entropy": 5.8175944805145265,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.7933,
      "mean_token_accuracy": 0.16275266855955123,
      "num_tokens": 6804794.0,
      "step": 3050
    },
    {
      "entropy": 5.914697074890137,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499395776978143,
      "loss": 5.9175,
      "mean_token_accuracy": 0.15746153593063356,
      "num_tokens": 6815976.0,
      "step": 3055
    },
    {
      "entropy": 5.967100524902344,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.8987,
      "mean_token_accuracy": 0.15376825481653214,
      "num_tokens": 6828027.0,
      "step": 3060
    },
    {
      "entropy": 5.991644477844238,
      "epoch": 0.34424664457797494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.8724,
      "mean_token_accuracy": 0.15755766332149507,
      "num_tokens": 6839195.0,
      "step": 3065
    },
    {
      "entropy": 5.703225994110108,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.7587,
      "mean_token_accuracy": 0.1565806671977043,
      "num_tokens": 6851731.0,
      "step": 3070
    },
    {
      "entropy": 5.916368103027343,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.7981,
      "mean_token_accuracy": 0.16409124583005905,
      "num_tokens": 6862916.0,
      "step": 3075
    },
    {
      "entropy": 6.025053215026856,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.9562,
      "mean_token_accuracy": 0.15552304387092591,
      "num_tokens": 6876003.0,
      "step": 3080
    },
    {
      "entropy": 5.906529664993286,
      "epoch": 0.34649295220980514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.946,
      "mean_token_accuracy": 0.15784678012132644,
      "num_tokens": 6886040.0,
      "step": 3085
    },
    {
      "entropy": 5.927360343933105,
      "epoch": 0.3470545291177627,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.841,
      "mean_token_accuracy": 0.16224057227373123,
      "num_tokens": 6896758.0,
      "step": 3090
    },
    {
      "entropy": 5.953919267654419,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.9075,
      "mean_token_accuracy": 0.1569637894630432,
      "num_tokens": 6907287.0,
      "step": 3095
    },
    {
      "entropy": 5.9171771049499515,
      "epoch": 0.3481776829336778,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.9252,
      "mean_token_accuracy": 0.1543675921857357,
      "num_tokens": 6918992.0,
      "step": 3100
    },
    {
      "entropy": 6.060032367706299,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.9307,
      "mean_token_accuracy": 0.15680571645498276,
      "num_tokens": 6931493.0,
      "step": 3105
    },
    {
      "entropy": 5.858033657073975,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1642995670437813,
      "num_tokens": 6941807.0,
      "step": 3110
    },
    {
      "entropy": 5.926855850219726,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.9636,
      "mean_token_accuracy": 0.15738050788640975,
      "num_tokens": 6953436.0,
      "step": 3115
    },
    {
      "entropy": 6.154252576828003,
      "epoch": 0.35042399056550794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993569484348359,
      "loss": 6.0503,
      "mean_token_accuracy": 0.15361388698220252,
      "num_tokens": 6965247.0,
      "step": 3120
    },
    {
      "entropy": 5.882022333145142,
      "epoch": 0.3509855674734655,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.8606,
      "mean_token_accuracy": 0.1635250687599182,
      "num_tokens": 6976605.0,
      "step": 3125
    },
    {
      "entropy": 5.892963695526123,
      "epoch": 0.35154714438142304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1708419293165207,
      "num_tokens": 6986205.0,
      "step": 3130
    },
    {
      "entropy": 5.826607275009155,
      "epoch": 0.3521087212893806,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.7705,
      "mean_token_accuracy": 0.1646653413772583,
      "num_tokens": 6997699.0,
      "step": 3135
    },
    {
      "entropy": 5.784408044815064,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.7973,
      "mean_token_accuracy": 0.15687063187360764,
      "num_tokens": 7008968.0,
      "step": 3140
    },
    {
      "entropy": 6.000455713272094,
      "epoch": 0.3532318751052957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499341692905686,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1572865605354309,
      "num_tokens": 7020633.0,
      "step": 3145
    },
    {
      "entropy": 5.800573825836182,
      "epoch": 0.35379345201325324,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.7457,
      "mean_token_accuracy": 0.1712852880358696,
      "num_tokens": 7031187.0,
      "step": 3150
    },
    {
      "entropy": 5.866760349273681,
      "epoch": 0.35435502892121074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.9662,
      "mean_token_accuracy": 0.15580730140209198,
      "num_tokens": 7042662.0,
      "step": 3155
    },
    {
      "entropy": 6.075199604034424,
      "epoch": 0.3549166058291683,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.9474,
      "mean_token_accuracy": 0.16323336064815522,
      "num_tokens": 7054297.0,
      "step": 3160
    },
    {
      "entropy": 5.99321403503418,
      "epoch": 0.35547818273712584,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.8053,
      "mean_token_accuracy": 0.1664621964097023,
      "num_tokens": 7064961.0,
      "step": 3165
    },
    {
      "entropy": 5.883630561828613,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.9654,
      "mean_token_accuracy": 0.15905895978212356,
      "num_tokens": 7076857.0,
      "step": 3170
    },
    {
      "entropy": 5.92643666267395,
      "epoch": 0.35660133655304094,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499323150548625,
      "loss": 5.8288,
      "mean_token_accuracy": 0.1658812627196312,
      "num_tokens": 7088926.0,
      "step": 3175
    },
    {
      "entropy": 5.85989089012146,
      "epoch": 0.3571629134609985,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.7505,
      "mean_token_accuracy": 0.16490727439522743,
      "num_tokens": 7099478.0,
      "step": 3180
    },
    {
      "entropy": 5.790369987487793,
      "epoch": 0.35772449036895604,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.7955,
      "mean_token_accuracy": 0.16942433565855025,
      "num_tokens": 7111776.0,
      "step": 3185
    },
    {
      "entropy": 6.009975481033325,
      "epoch": 0.3582860672769136,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.8808,
      "mean_token_accuracy": 0.16010644435882568,
      "num_tokens": 7123084.0,
      "step": 3190
    },
    {
      "entropy": 5.8817391872406,
      "epoch": 0.35884764418487114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.8452,
      "mean_token_accuracy": 0.16542601436376572,
      "num_tokens": 7134973.0,
      "step": 3195
    },
    {
      "entropy": 5.892574882507324,
      "epoch": 0.3594092210928287,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.8804,
      "mean_token_accuracy": 0.15410309433937072,
      "num_tokens": 7147227.0,
      "step": 3200
    },
    {
      "entropy": 5.97821683883667,
      "epoch": 0.3599707980007862,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.864,
      "mean_token_accuracy": 0.16147272139787675,
      "num_tokens": 7158201.0,
      "step": 3205
    },
    {
      "entropy": 5.785186910629273,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499301192814866,
      "loss": 5.7684,
      "mean_token_accuracy": 0.1673009842634201,
      "num_tokens": 7168909.0,
      "step": 3210
    },
    {
      "entropy": 5.995763444900513,
      "epoch": 0.3610939518167013,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.9293,
      "mean_token_accuracy": 0.1569584757089615,
      "num_tokens": 7179877.0,
      "step": 3215
    },
    {
      "entropy": 5.913835334777832,
      "epoch": 0.36165552872465884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.8438,
      "mean_token_accuracy": 0.16550108194351196,
      "num_tokens": 7190823.0,
      "step": 3220
    },
    {
      "entropy": 5.784553384780883,
      "epoch": 0.3622171056326164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.7717,
      "mean_token_accuracy": 0.16480892449617385,
      "num_tokens": 7201700.0,
      "step": 3225
    },
    {
      "entropy": 5.910831880569458,
      "epoch": 0.36277868254057394,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499288488415819,
      "loss": 5.8078,
      "mean_token_accuracy": 0.1658710226416588,
      "num_tokens": 7213618.0,
      "step": 3230
    },
    {
      "entropy": 6.051048278808594,
      "epoch": 0.3633402594485315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.9842,
      "mean_token_accuracy": 0.1599672853946686,
      "num_tokens": 7226207.0,
      "step": 3235
    },
    {
      "entropy": 5.871135187149048,
      "epoch": 0.36390183635648904,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.8453,
      "mean_token_accuracy": 0.16316219717264174,
      "num_tokens": 7237545.0,
      "step": 3240
    },
    {
      "entropy": 5.8658225536346436,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.8152,
      "mean_token_accuracy": 0.16614115238189697,
      "num_tokens": 7247799.0,
      "step": 3245
    },
    {
      "entropy": 5.896589946746826,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.7922,
      "mean_token_accuracy": 0.1614817500114441,
      "num_tokens": 7260229.0,
      "step": 3250
    },
    {
      "entropy": 5.829941320419311,
      "epoch": 0.36558656708036164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.9007,
      "mean_token_accuracy": 0.1580626793205738,
      "num_tokens": 7271457.0,
      "step": 3255
    },
    {
      "entropy": 5.933812761306763,
      "epoch": 0.3661481439883192,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.8052,
      "mean_token_accuracy": 0.16287915259599686,
      "num_tokens": 7282124.0,
      "step": 3260
    },
    {
      "entropy": 5.927642107009888,
      "epoch": 0.36670972089627674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.9091,
      "mean_token_accuracy": 0.16330223083496093,
      "num_tokens": 7292981.0,
      "step": 3265
    },
    {
      "entropy": 6.152715539932251,
      "epoch": 0.3672712978042343,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992627368355112,
      "loss": 6.0666,
      "mean_token_accuracy": 0.15554819852113724,
      "num_tokens": 7304591.0,
      "step": 3270
    },
    {
      "entropy": 5.956924867630005,
      "epoch": 0.36783287471219184,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.797,
      "mean_token_accuracy": 0.1661202237010002,
      "num_tokens": 7315918.0,
      "step": 3275
    },
    {
      "entropy": 5.8168786525726315,
      "epoch": 0.3683944516201494,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.9015,
      "mean_token_accuracy": 0.16148915886878967,
      "num_tokens": 7327725.0,
      "step": 3280
    },
    {
      "entropy": 6.069124317169189,
      "epoch": 0.36895602852810694,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.9152,
      "mean_token_accuracy": 0.15593092143535614,
      "num_tokens": 7339787.0,
      "step": 3285
    },
    {
      "entropy": 5.9365582942962645,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.8939,
      "mean_token_accuracy": 0.16256406754255295,
      "num_tokens": 7351284.0,
      "step": 3290
    },
    {
      "entropy": 5.742411994934082,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.7132,
      "mean_token_accuracy": 0.1695790097117424,
      "num_tokens": 7362105.0,
      "step": 3295
    },
    {
      "entropy": 5.893797731399536,
      "epoch": 0.37064075925197953,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.816,
      "mean_token_accuracy": 0.16402172893285752,
      "num_tokens": 7375204.0,
      "step": 3300
    },
    {
      "entropy": 6.095653533935547,
      "epoch": 0.3712023361599371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.98,
      "mean_token_accuracy": 0.1561136357486248,
      "num_tokens": 7385853.0,
      "step": 3305
    },
    {
      "entropy": 5.926312065124511,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.8333,
      "mean_token_accuracy": 0.1528450667858124,
      "num_tokens": 7397459.0,
      "step": 3310
    },
    {
      "entropy": 5.815288496017456,
      "epoch": 0.3723254899758522,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.8029,
      "mean_token_accuracy": 0.15823303014039994,
      "num_tokens": 7408996.0,
      "step": 3315
    },
    {
      "entropy": 5.959876823425293,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.9063,
      "mean_token_accuracy": 0.15332845151424407,
      "num_tokens": 7419938.0,
      "step": 3320
    },
    {
      "entropy": 5.944911193847656,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.912,
      "mean_token_accuracy": 0.1548202320933342,
      "num_tokens": 7430181.0,
      "step": 3325
    },
    {
      "entropy": 5.948005962371826,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.8674,
      "mean_token_accuracy": 0.15860404074192047,
      "num_tokens": 7440774.0,
      "step": 3330
    },
    {
      "entropy": 5.933913946151733,
      "epoch": 0.3745717976076824,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499219915847757,
      "loss": 5.7852,
      "mean_token_accuracy": 0.16459606140851973,
      "num_tokens": 7451744.0,
      "step": 3335
    },
    {
      "entropy": 5.72104172706604,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.8251,
      "mean_token_accuracy": 0.16036779880523683,
      "num_tokens": 7463731.0,
      "step": 3340
    },
    {
      "entropy": 5.943351984024048,
      "epoch": 0.37569495142359743,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.8207,
      "mean_token_accuracy": 0.16885581910610198,
      "num_tokens": 7474392.0,
      "step": 3345
    },
    {
      "entropy": 5.8482115268707275,
      "epoch": 0.376256528331555,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.7597,
      "mean_token_accuracy": 0.17417206317186357,
      "num_tokens": 7484860.0,
      "step": 3350
    },
    {
      "entropy": 5.800925064086914,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499206497413985,
      "loss": 5.8007,
      "mean_token_accuracy": 0.16721806675195694,
      "num_tokens": 7495877.0,
      "step": 3355
    },
    {
      "entropy": 5.954871892929077,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.8626,
      "mean_token_accuracy": 0.1659214660525322,
      "num_tokens": 7506879.0,
      "step": 3360
    },
    {
      "entropy": 5.829053115844727,
      "epoch": 0.37794125905542764,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.7166,
      "mean_token_accuracy": 0.16946493536233903,
      "num_tokens": 7517405.0,
      "step": 3365
    },
    {
      "entropy": 5.846356964111328,
      "epoch": 0.3785028359633852,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.7839,
      "mean_token_accuracy": 0.16438688188791276,
      "num_tokens": 7528832.0,
      "step": 3370
    },
    {
      "entropy": 5.828200101852417,
      "epoch": 0.37906441287134274,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.7811,
      "mean_token_accuracy": 0.15837691128253936,
      "num_tokens": 7541048.0,
      "step": 3375
    },
    {
      "entropy": 5.882894372940063,
      "epoch": 0.3796259897793003,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.7892,
      "mean_token_accuracy": 0.16829562336206436,
      "num_tokens": 7551849.0,
      "step": 3380
    },
    {
      "entropy": 5.801930713653564,
      "epoch": 0.38018756668725784,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.8084,
      "mean_token_accuracy": 0.16716636419296266,
      "num_tokens": 7563148.0,
      "step": 3385
    },
    {
      "entropy": 6.005253744125366,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.9857,
      "mean_token_accuracy": 0.15619939416646958,
      "num_tokens": 7575119.0,
      "step": 3390
    },
    {
      "entropy": 5.920757913589478,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.8477,
      "mean_token_accuracy": 0.16746386140584946,
      "num_tokens": 7585899.0,
      "step": 3395
    },
    {
      "entropy": 5.927013492584228,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.7408,
      "mean_token_accuracy": 0.17119719982147216,
      "num_tokens": 7598938.0,
      "step": 3400
    },
    {
      "entropy": 5.718728494644165,
      "epoch": 0.382433874319088,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.8176,
      "mean_token_accuracy": 0.15823707431554795,
      "num_tokens": 7610087.0,
      "step": 3405
    },
    {
      "entropy": 6.031653833389282,
      "epoch": 0.38299545122704554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.9304,
      "mean_token_accuracy": 0.15817293301224708,
      "num_tokens": 7622186.0,
      "step": 3410
    },
    {
      "entropy": 5.95524001121521,
      "epoch": 0.3835570281350031,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.8724,
      "mean_token_accuracy": 0.15889675095677375,
      "num_tokens": 7633501.0,
      "step": 3415
    },
    {
      "entropy": 5.811369943618774,
      "epoch": 0.38411860504296064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.8081,
      "mean_token_accuracy": 0.1626387983560562,
      "num_tokens": 7643656.0,
      "step": 3420
    },
    {
      "entropy": 5.975064659118653,
      "epoch": 0.3846801819509182,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.8101,
      "mean_token_accuracy": 0.16192732900381088,
      "num_tokens": 7654565.0,
      "step": 3425
    },
    {
      "entropy": 5.829861116409302,
      "epoch": 0.38524175885887574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.7368,
      "mean_token_accuracy": 0.17527267187833787,
      "num_tokens": 7664725.0,
      "step": 3430
    },
    {
      "entropy": 5.78968071937561,
      "epoch": 0.3858033357668333,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.8136,
      "mean_token_accuracy": 0.1644575446844101,
      "num_tokens": 7674772.0,
      "step": 3435
    },
    {
      "entropy": 5.963439416885376,
      "epoch": 0.38636491267479084,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499148194874525,
      "loss": 5.8914,
      "mean_token_accuracy": 0.15519467890262603,
      "num_tokens": 7686488.0,
      "step": 3440
    },
    {
      "entropy": 5.9735828876495365,
      "epoch": 0.38692648958274833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.8796,
      "mean_token_accuracy": 0.16299588978290558,
      "num_tokens": 7698842.0,
      "step": 3445
    },
    {
      "entropy": 5.723750066757202,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.6652,
      "mean_token_accuracy": 0.16502486169338226,
      "num_tokens": 7709255.0,
      "step": 3450
    },
    {
      "entropy": 5.712912368774414,
      "epoch": 0.38804964339866344,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1652544140815735,
      "num_tokens": 7720182.0,
      "step": 3455
    },
    {
      "entropy": 5.820716857910156,
      "epoch": 0.388611220306621,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.7476,
      "mean_token_accuracy": 0.17566618472337722,
      "num_tokens": 7730944.0,
      "step": 3460
    },
    {
      "entropy": 5.849359560012817,
      "epoch": 0.38917279721457854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049913065451149,
      "loss": 5.8219,
      "mean_token_accuracy": 0.16147851198911667,
      "num_tokens": 7741686.0,
      "step": 3465
    },
    {
      "entropy": 5.922025728225708,
      "epoch": 0.3897343741225361,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.807,
      "mean_token_accuracy": 0.16342652589082718,
      "num_tokens": 7752245.0,
      "step": 3470
    },
    {
      "entropy": 5.757428932189941,
      "epoch": 0.39029595103049364,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.7788,
      "mean_token_accuracy": 0.159398677945137,
      "num_tokens": 7762622.0,
      "step": 3475
    },
    {
      "entropy": 5.963356447219849,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.9069,
      "mean_token_accuracy": 0.16293678879737855,
      "num_tokens": 7773550.0,
      "step": 3480
    },
    {
      "entropy": 5.894080686569214,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.789,
      "mean_token_accuracy": 0.16059745252132415,
      "num_tokens": 7784988.0,
      "step": 3485
    },
    {
      "entropy": 5.940509033203125,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.8767,
      "mean_token_accuracy": 0.16270290464162826,
      "num_tokens": 7797706.0,
      "step": 3490
    },
    {
      "entropy": 5.870716524124146,
      "epoch": 0.3925422586623238,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.7541,
      "mean_token_accuracy": 0.1708513543009758,
      "num_tokens": 7808960.0,
      "step": 3495
    },
    {
      "entropy": 5.8290143489837645,
      "epoch": 0.39310383557028133,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.7225,
      "mean_token_accuracy": 0.17448564767837524,
      "num_tokens": 7819271.0,
      "step": 3500
    },
    {
      "entropy": 5.827299976348877,
      "epoch": 0.3936654124782389,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.8587,
      "mean_token_accuracy": 0.16258958131074905,
      "num_tokens": 7832257.0,
      "step": 3505
    },
    {
      "entropy": 5.969977903366089,
      "epoch": 0.39422698938619644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049909863227389,
      "loss": 5.8743,
      "mean_token_accuracy": 0.16018556654453278,
      "num_tokens": 7842684.0,
      "step": 3510
    },
    {
      "entropy": 5.831171321868896,
      "epoch": 0.394788566294154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.878,
      "mean_token_accuracy": 0.1610353022813797,
      "num_tokens": 7855031.0,
      "step": 3515
    },
    {
      "entropy": 5.840944480895996,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.6858,
      "mean_token_accuracy": 0.16333473473787308,
      "num_tokens": 7865291.0,
      "step": 3520
    },
    {
      "entropy": 5.744883012771607,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.6451,
      "mean_token_accuracy": 0.16607558876276016,
      "num_tokens": 7876401.0,
      "step": 3525
    },
    {
      "entropy": 5.718347787857056,
      "epoch": 0.39647329701802664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.6944,
      "mean_token_accuracy": 0.16728755831718445,
      "num_tokens": 7887787.0,
      "step": 3530
    },
    {
      "entropy": 5.763777160644532,
      "epoch": 0.3970348739259842,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.6914,
      "mean_token_accuracy": 0.16545438915491104,
      "num_tokens": 7898550.0,
      "step": 3535
    },
    {
      "entropy": 5.8369975090026855,
      "epoch": 0.3975964508339417,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.8377,
      "mean_token_accuracy": 0.16232438683509826,
      "num_tokens": 7909254.0,
      "step": 3540
    },
    {
      "entropy": 6.022816181182861,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990733265032,
      "loss": 5.8797,
      "mean_token_accuracy": 0.16756701320409775,
      "num_tokens": 7920164.0,
      "step": 3545
    },
    {
      "entropy": 5.7947046756744385,
      "epoch": 0.3987196046498568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.7971,
      "mean_token_accuracy": 0.16098637133836746,
      "num_tokens": 7931390.0,
      "step": 3550
    },
    {
      "entropy": 5.927447414398193,
      "epoch": 0.39928118155781434,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499066032070036,
      "loss": 5.8199,
      "mean_token_accuracy": 0.16026633083820344,
      "num_tokens": 7943765.0,
      "step": 3555
    },
    {
      "entropy": 5.817334890365601,
      "epoch": 0.3998427584657719,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.6543,
      "mean_token_accuracy": 0.16805292963981627,
      "num_tokens": 7954621.0,
      "step": 3560
    },
    {
      "entropy": 5.7308307647705075,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499058709099475,
      "loss": 5.8617,
      "mean_token_accuracy": 0.16110717505216599,
      "num_tokens": 7966063.0,
      "step": 3565
    },
    {
      "entropy": 5.984078693389892,
      "epoch": 0.400965912281687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.7822,
      "mean_token_accuracy": 0.16557336151599883,
      "num_tokens": 7977391.0,
      "step": 3570
    },
    {
      "entropy": 5.949883031845093,
      "epoch": 0.40152748918964454,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.9017,
      "mean_token_accuracy": 0.1615771010518074,
      "num_tokens": 7988786.0,
      "step": 3575
    },
    {
      "entropy": 5.823601341247558,
      "epoch": 0.4020890660976021,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.7066,
      "mean_token_accuracy": 0.16304796189069748,
      "num_tokens": 8000848.0,
      "step": 3580
    },
    {
      "entropy": 5.806692171096802,
      "epoch": 0.40265064300555964,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.8641,
      "mean_token_accuracy": 0.16117497086524962,
      "num_tokens": 8012621.0,
      "step": 3585
    },
    {
      "entropy": 6.056632232666016,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499040276828089,
      "loss": 5.9032,
      "mean_token_accuracy": 0.16068349182605743,
      "num_tokens": 8023599.0,
      "step": 3590
    },
    {
      "entropy": 5.734525108337403,
      "epoch": 0.4037737968214747,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.7073,
      "mean_token_accuracy": 0.16968922764062883,
      "num_tokens": 8037151.0,
      "step": 3595
    },
    {
      "entropy": 5.866984462738037,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.9032,
      "mean_token_accuracy": 0.15894825309514998,
      "num_tokens": 8048606.0,
      "step": 3600
    },
    {
      "entropy": 5.814764881134034,
      "epoch": 0.4048969506373898,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.6844,
      "mean_token_accuracy": 0.1796335518360138,
      "num_tokens": 8058430.0,
      "step": 3605
    },
    {
      "entropy": 5.724921464920044,
      "epoch": 0.40545852754534734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.7002,
      "mean_token_accuracy": 0.16601582020521163,
      "num_tokens": 8069554.0,
      "step": 3610
    },
    {
      "entropy": 5.946437072753906,
      "epoch": 0.4060201044533049,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.8534,
      "mean_token_accuracy": 0.15956933349370955,
      "num_tokens": 8082131.0,
      "step": 3615
    },
    {
      "entropy": 5.800406789779663,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.6988,
      "mean_token_accuracy": 0.16990892142057418,
      "num_tokens": 8093454.0,
      "step": 3620
    },
    {
      "entropy": 5.98328971862793,
      "epoch": 0.40714325826922,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.9571,
      "mean_token_accuracy": 0.1504925012588501,
      "num_tokens": 8105502.0,
      "step": 3625
    },
    {
      "entropy": 5.859578943252563,
      "epoch": 0.40770483517717754,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.8145,
      "mean_token_accuracy": 0.1626613110303879,
      "num_tokens": 8116524.0,
      "step": 3630
    },
    {
      "entropy": 5.894063568115234,
      "epoch": 0.40826641208513503,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499006649337227,
      "loss": 5.7959,
      "mean_token_accuracy": 0.17154594510793686,
      "num_tokens": 8126949.0,
      "step": 3635
    },
    {
      "entropy": 5.830201005935669,
      "epoch": 0.4088279889930926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.763,
      "mean_token_accuracy": 0.16976310908794404,
      "num_tokens": 8137346.0,
      "step": 3640
    },
    {
      "entropy": 5.788245964050293,
      "epoch": 0.40938956590105013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.7978,
      "mean_token_accuracy": 0.16071729511022567,
      "num_tokens": 8148499.0,
      "step": 3645
    },
    {
      "entropy": 5.900784349441528,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498995311785053,
      "loss": 5.8276,
      "mean_token_accuracy": 0.16623455286026,
      "num_tokens": 8159645.0,
      "step": 3650
    },
    {
      "entropy": 5.830116033554077,
      "epoch": 0.41051271971696524,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.6977,
      "mean_token_accuracy": 0.16828417778015137,
      "num_tokens": 8171033.0,
      "step": 3655
    },
    {
      "entropy": 5.790131378173828,
      "epoch": 0.4110742966249228,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.792,
      "mean_token_accuracy": 0.16119424402713775,
      "num_tokens": 8181582.0,
      "step": 3660
    },
    {
      "entropy": 5.773313522338867,
      "epoch": 0.41163587353288034,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498983910044002,
      "loss": 5.6858,
      "mean_token_accuracy": 0.16789519041776657,
      "num_tokens": 8192432.0,
      "step": 3665
    },
    {
      "entropy": 5.975875806808472,
      "epoch": 0.4121974504408379,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.867,
      "mean_token_accuracy": 0.16418623179197311,
      "num_tokens": 8203995.0,
      "step": 3670
    },
    {
      "entropy": 5.781466293334961,
      "epoch": 0.41275902734879544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.7005,
      "mean_token_accuracy": 0.16910858303308487,
      "num_tokens": 8214323.0,
      "step": 3675
    },
    {
      "entropy": 5.71599268913269,
      "epoch": 0.413320604256753,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.7628,
      "mean_token_accuracy": 0.15963411182165146,
      "num_tokens": 8226285.0,
      "step": 3680
    },
    {
      "entropy": 6.009037590026855,
      "epoch": 0.4138821811647105,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.9395,
      "mean_token_accuracy": 0.15572211444377898,
      "num_tokens": 8237961.0,
      "step": 3685
    },
    {
      "entropy": 5.771542167663574,
      "epoch": 0.41444375807266803,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.6755,
      "mean_token_accuracy": 0.17040250152349473,
      "num_tokens": 8248463.0,
      "step": 3690
    },
    {
      "entropy": 5.865232706069946,
      "epoch": 0.4150053349806256,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.815,
      "mean_token_accuracy": 0.16708862632513047,
      "num_tokens": 8260342.0,
      "step": 3695
    },
    {
      "entropy": 5.873272800445557,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.9115,
      "mean_token_accuracy": 0.16665390282869338,
      "num_tokens": 8272055.0,
      "step": 3700
    },
    {
      "entropy": 6.020225572586059,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.876,
      "mean_token_accuracy": 0.16337255090475084,
      "num_tokens": 8282129.0,
      "step": 3705
    },
    {
      "entropy": 5.9397468090057375,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.8663,
      "mean_token_accuracy": 0.16446712464094163,
      "num_tokens": 8294663.0,
      "step": 3710
    },
    {
      "entropy": 5.813158178329468,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498945440696253,
      "loss": 5.6755,
      "mean_token_accuracy": 0.17049953639507293,
      "num_tokens": 8305882.0,
      "step": 3715
    },
    {
      "entropy": 5.772831439971924,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16710352897644043,
      "num_tokens": 8316604.0,
      "step": 3720
    },
    {
      "entropy": 5.902248382568359,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.8747,
      "mean_token_accuracy": 0.16121989041566848,
      "num_tokens": 8328126.0,
      "step": 3725
    },
    {
      "entropy": 5.865182256698608,
      "epoch": 0.4189363733363284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.799,
      "mean_token_accuracy": 0.1652168944478035,
      "num_tokens": 8339050.0,
      "step": 3730
    },
    {
      "entropy": 5.732286357879639,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17028117328882217,
      "num_tokens": 8350216.0,
      "step": 3735
    },
    {
      "entropy": 5.8672754764556885,
      "epoch": 0.4200595271522435,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.7858,
      "mean_token_accuracy": 0.16723376661539077,
      "num_tokens": 8362311.0,
      "step": 3740
    },
    {
      "entropy": 5.879049682617188,
      "epoch": 0.42062110406020103,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.9504,
      "mean_token_accuracy": 0.16739826500415803,
      "num_tokens": 8373751.0,
      "step": 3745
    },
    {
      "entropy": 5.871591663360595,
      "epoch": 0.4211826809681586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.7839,
      "mean_token_accuracy": 0.1667262703180313,
      "num_tokens": 8384932.0,
      "step": 3750
    },
    {
      "entropy": 5.870432186126709,
      "epoch": 0.42174425787611614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.7206,
      "mean_token_accuracy": 0.172009739279747,
      "num_tokens": 8394973.0,
      "step": 3755
    },
    {
      "entropy": 5.825662612915039,
      "epoch": 0.4223058347840737,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.8181,
      "mean_token_accuracy": 0.1665082648396492,
      "num_tokens": 8406590.0,
      "step": 3760
    },
    {
      "entropy": 5.801623344421387,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.7343,
      "mean_token_accuracy": 0.1675838053226471,
      "num_tokens": 8417127.0,
      "step": 3765
    },
    {
      "entropy": 5.918402051925659,
      "epoch": 0.4234289885999888,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.8453,
      "mean_token_accuracy": 0.1659783571958542,
      "num_tokens": 8429442.0,
      "step": 3770
    },
    {
      "entropy": 5.811355018615723,
      "epoch": 0.42399056550794634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.7607,
      "mean_token_accuracy": 0.16928921788930892,
      "num_tokens": 8439375.0,
      "step": 3775
    },
    {
      "entropy": 5.832507801055908,
      "epoch": 0.42455214241590383,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.7695,
      "mean_token_accuracy": 0.1644688591361046,
      "num_tokens": 8449817.0,
      "step": 3780
    },
    {
      "entropy": 5.78357048034668,
      "epoch": 0.4251137193238614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.653,
      "mean_token_accuracy": 0.17269241362810134,
      "num_tokens": 8460096.0,
      "step": 3785
    },
    {
      "entropy": 5.856718921661377,
      "epoch": 0.42567529623181893,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.783,
      "mean_token_accuracy": 0.1635545700788498,
      "num_tokens": 8470241.0,
      "step": 3790
    },
    {
      "entropy": 5.866709804534912,
      "epoch": 0.4262368731397765,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.7959,
      "mean_token_accuracy": 0.15800238102674485,
      "num_tokens": 8481438.0,
      "step": 3795
    },
    {
      "entropy": 5.85970401763916,
      "epoch": 0.42679845004773403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498878406420253,
      "loss": 5.8495,
      "mean_token_accuracy": 0.16143330335617065,
      "num_tokens": 8492588.0,
      "step": 3800
    },
    {
      "entropy": 5.779029798507691,
      "epoch": 0.4273600269556916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.7801,
      "mean_token_accuracy": 0.1666340261697769,
      "num_tokens": 8502298.0,
      "step": 3805
    },
    {
      "entropy": 5.804246425628662,
      "epoch": 0.42792160386364914,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.6317,
      "mean_token_accuracy": 0.1682993531227112,
      "num_tokens": 8513507.0,
      "step": 3810
    },
    {
      "entropy": 5.814224147796631,
      "epoch": 0.4284831807716067,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.7407,
      "mean_token_accuracy": 0.17205402106046677,
      "num_tokens": 8523888.0,
      "step": 3815
    },
    {
      "entropy": 5.748360204696655,
      "epoch": 0.42904475767956424,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498862334234653,
      "loss": 5.6829,
      "mean_token_accuracy": 0.17145948857069016,
      "num_tokens": 8534032.0,
      "step": 3820
    },
    {
      "entropy": 5.761185073852539,
      "epoch": 0.4296063345875218,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.819,
      "mean_token_accuracy": 0.16719216257333755,
      "num_tokens": 8545544.0,
      "step": 3825
    },
    {
      "entropy": 5.888425254821778,
      "epoch": 0.4301679114954793,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.7701,
      "mean_token_accuracy": 0.167030468583107,
      "num_tokens": 8555428.0,
      "step": 3830
    },
    {
      "entropy": 5.7430023670196535,
      "epoch": 0.43072948840343683,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.6721,
      "mean_token_accuracy": 0.16957473009824753,
      "num_tokens": 8566456.0,
      "step": 3835
    },
    {
      "entropy": 5.826628875732422,
      "epoch": 0.4312910653113944,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.6896,
      "mean_token_accuracy": 0.1679348900914192,
      "num_tokens": 8578125.0,
      "step": 3840
    },
    {
      "entropy": 5.783590936660767,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.7499,
      "mean_token_accuracy": 0.1610526993870735,
      "num_tokens": 8589530.0,
      "step": 3845
    },
    {
      "entropy": 5.828071880340576,
      "epoch": 0.4324142191273095,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.8504,
      "mean_token_accuracy": 0.16466037034988404,
      "num_tokens": 8601655.0,
      "step": 3850
    },
    {
      "entropy": 5.8106616020202635,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.763,
      "mean_token_accuracy": 0.1670367658138275,
      "num_tokens": 8613355.0,
      "step": 3855
    },
    {
      "entropy": 5.823573637008667,
      "epoch": 0.4335373729432246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.7118,
      "mean_token_accuracy": 0.16885844767093658,
      "num_tokens": 8624154.0,
      "step": 3860
    },
    {
      "entropy": 5.815149307250977,
      "epoch": 0.43409894985118214,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049882575484962,
      "loss": 5.7858,
      "mean_token_accuracy": 0.16578275412321092,
      "num_tokens": 8633888.0,
      "step": 3865
    },
    {
      "entropy": 5.878628587722778,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.7605,
      "mean_token_accuracy": 0.16726396083831788,
      "num_tokens": 8644965.0,
      "step": 3870
    },
    {
      "entropy": 5.801642656326294,
      "epoch": 0.4352221036670972,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.7145,
      "mean_token_accuracy": 0.17267416566610336,
      "num_tokens": 8655483.0,
      "step": 3875
    },
    {
      "entropy": 5.81993408203125,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498813433436022,
      "loss": 5.6916,
      "mean_token_accuracy": 0.16909216344356537,
      "num_tokens": 8666443.0,
      "step": 3880
    },
    {
      "entropy": 5.816588592529297,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.7545,
      "mean_token_accuracy": 0.17053562998771668,
      "num_tokens": 8677198.0,
      "step": 3885
    },
    {
      "entropy": 5.756594896316528,
      "epoch": 0.43690683439096983,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.7977,
      "mean_token_accuracy": 0.16542441099882127,
      "num_tokens": 8688303.0,
      "step": 3890
    },
    {
      "entropy": 5.910586500167847,
      "epoch": 0.4374684112989274,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.7561,
      "mean_token_accuracy": 0.16605307906866074,
      "num_tokens": 8700326.0,
      "step": 3895
    },
    {
      "entropy": 5.762407350540161,
      "epoch": 0.43802998820688493,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.721,
      "mean_token_accuracy": 0.1687159538269043,
      "num_tokens": 8711101.0,
      "step": 3900
    },
    {
      "entropy": 5.769402265548706,
      "epoch": 0.4385915651148425,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.8423,
      "mean_token_accuracy": 0.16267140656709672,
      "num_tokens": 8721751.0,
      "step": 3905
    },
    {
      "entropy": 5.954661417007446,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.9352,
      "mean_token_accuracy": 0.15480559915304185,
      "num_tokens": 8733550.0,
      "step": 3910
    },
    {
      "entropy": 5.869749975204468,
      "epoch": 0.4397147189307576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.7293,
      "mean_token_accuracy": 0.16600938737392426,
      "num_tokens": 8744218.0,
      "step": 3915
    },
    {
      "entropy": 5.753007507324218,
      "epoch": 0.44027629583871514,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.7763,
      "mean_token_accuracy": 0.16374737173318862,
      "num_tokens": 8756531.0,
      "step": 3920
    },
    {
      "entropy": 6.002922201156617,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.8171,
      "mean_token_accuracy": 0.16575318574905396,
      "num_tokens": 8768114.0,
      "step": 3925
    },
    {
      "entropy": 5.745939636230469,
      "epoch": 0.4413994496546302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.805,
      "mean_token_accuracy": 0.1609507381916046,
      "num_tokens": 8780010.0,
      "step": 3930
    },
    {
      "entropy": 5.886521244049073,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.7653,
      "mean_token_accuracy": 0.16114584356546402,
      "num_tokens": 8792183.0,
      "step": 3935
    },
    {
      "entropy": 5.888902807235718,
      "epoch": 0.4425226034705453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.8566,
      "mean_token_accuracy": 0.16036604791879655,
      "num_tokens": 8804513.0,
      "step": 3940
    },
    {
      "entropy": 5.777865171432495,
      "epoch": 0.44308418037850283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.7553,
      "mean_token_accuracy": 0.17000913769006729,
      "num_tokens": 8814918.0,
      "step": 3945
    },
    {
      "entropy": 5.8164514064788815,
      "epoch": 0.4436457572864604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.704,
      "mean_token_accuracy": 0.16863586157560348,
      "num_tokens": 8826176.0,
      "step": 3950
    },
    {
      "entropy": 5.743463325500488,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.7513,
      "mean_token_accuracy": 0.16684129387140273,
      "num_tokens": 8837242.0,
      "step": 3955
    },
    {
      "entropy": 5.9041355609893795,
      "epoch": 0.4447689111023755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.7139,
      "mean_token_accuracy": 0.1682710364460945,
      "num_tokens": 8848972.0,
      "step": 3960
    },
    {
      "entropy": 5.703074598312378,
      "epoch": 0.44533048801033304,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.674,
      "mean_token_accuracy": 0.1637897565960884,
      "num_tokens": 8861108.0,
      "step": 3965
    },
    {
      "entropy": 5.782131719589233,
      "epoch": 0.4458920649182906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.782,
      "mean_token_accuracy": 0.16714437305927277,
      "num_tokens": 8872161.0,
      "step": 3970
    },
    {
      "entropy": 6.058572435379029,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.8653,
      "mean_token_accuracy": 0.16646484285593033,
      "num_tokens": 8882497.0,
      "step": 3975
    },
    {
      "entropy": 5.645896625518799,
      "epoch": 0.44701521873420563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.5833,
      "mean_token_accuracy": 0.17363847494125367,
      "num_tokens": 8893155.0,
      "step": 3980
    },
    {
      "entropy": 5.686550903320312,
      "epoch": 0.4475767956421632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498725387911674,
      "loss": 5.6662,
      "mean_token_accuracy": 0.1726548358798027,
      "num_tokens": 8904478.0,
      "step": 3985
    },
    {
      "entropy": 5.747406625747681,
      "epoch": 0.44813837255012073,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.7828,
      "mean_token_accuracy": 0.15897889584302902,
      "num_tokens": 8915470.0,
      "step": 3990
    },
    {
      "entropy": 5.84643874168396,
      "epoch": 0.4486999494580783,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498716838763451,
      "loss": 5.718,
      "mean_token_accuracy": 0.1680638760328293,
      "num_tokens": 8925644.0,
      "step": 3995
    },
    {
      "entropy": 5.809060144424438,
      "epoch": 0.44926152636603583,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498712553504425,
      "loss": 5.7545,
      "mean_token_accuracy": 0.16400753408670427,
      "num_tokens": 8937000.0,
      "step": 4000
    },
    {
      "entropy": 5.759871149063111,
      "epoch": 0.4498231032739934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.7301,
      "mean_token_accuracy": 0.16376389265060426,
      "num_tokens": 8948258.0,
      "step": 4005
    },
    {
      "entropy": 5.8029169082641605,
      "epoch": 0.45038468018195094,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.746,
      "mean_token_accuracy": 0.16891608238220215,
      "num_tokens": 8960107.0,
      "step": 4010
    },
    {
      "entropy": 5.846726274490356,
      "epoch": 0.4509462570899085,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.7237,
      "mean_token_accuracy": 0.16962783485651017,
      "num_tokens": 8969612.0,
      "step": 4015
    },
    {
      "entropy": 5.89162974357605,
      "epoch": 0.451507833997866,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16363618075847625,
      "num_tokens": 8980970.0,
      "step": 4020
    },
    {
      "entropy": 5.861903858184815,
      "epoch": 0.45206941090582353,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.771,
      "mean_token_accuracy": 0.16609935760498046,
      "num_tokens": 8992218.0,
      "step": 4025
    },
    {
      "entropy": 5.727895593643188,
      "epoch": 0.4526309878137811,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.6925,
      "mean_token_accuracy": 0.17172871232032777,
      "num_tokens": 9003737.0,
      "step": 4030
    },
    {
      "entropy": 5.871471118927002,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.7609,
      "mean_token_accuracy": 0.16709625869989395,
      "num_tokens": 9015427.0,
      "step": 4035
    },
    {
      "entropy": 5.7899058818817135,
      "epoch": 0.4537541416296962,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.7803,
      "mean_token_accuracy": 0.16394250839948654,
      "num_tokens": 9025003.0,
      "step": 4040
    },
    {
      "entropy": 5.775137424468994,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.726,
      "mean_token_accuracy": 0.16647830456495286,
      "num_tokens": 9036074.0,
      "step": 4045
    },
    {
      "entropy": 5.936799144744873,
      "epoch": 0.4548772954456113,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.738,
      "mean_token_accuracy": 0.16546613425016404,
      "num_tokens": 9046878.0,
      "step": 4050
    },
    {
      "entropy": 5.783922004699707,
      "epoch": 0.45543887235356884,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.8007,
      "mean_token_accuracy": 0.16725005060434342,
      "num_tokens": 9058366.0,
      "step": 4055
    },
    {
      "entropy": 5.899766874313355,
      "epoch": 0.4560004492615264,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498660574833995,
      "loss": 5.8188,
      "mean_token_accuracy": 0.16507331877946854,
      "num_tokens": 9069542.0,
      "step": 4060
    },
    {
      "entropy": 5.914427661895752,
      "epoch": 0.45656202616948394,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.9796,
      "mean_token_accuracy": 0.15912373960018159,
      "num_tokens": 9081219.0,
      "step": 4065
    },
    {
      "entropy": 5.8226522445678714,
      "epoch": 0.45712360307744143,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17707761228084565,
      "num_tokens": 9092101.0,
      "step": 4070
    },
    {
      "entropy": 5.781389141082764,
      "epoch": 0.457685179985399,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498647419924648,
      "loss": 5.7367,
      "mean_token_accuracy": 0.16802326887845992,
      "num_tokens": 9101900.0,
      "step": 4075
    },
    {
      "entropy": 5.965834522247315,
      "epoch": 0.45824675689335653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498643020712638,
      "loss": 5.867,
      "mean_token_accuracy": 0.1598948821425438,
      "num_tokens": 9114369.0,
      "step": 4080
    },
    {
      "entropy": 5.796750783920288,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.6732,
      "mean_token_accuracy": 0.16585982143878936,
      "num_tokens": 9126548.0,
      "step": 4085
    },
    {
      "entropy": 5.825146055221557,
      "epoch": 0.45936991070927163,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.8567,
      "mean_token_accuracy": 0.16830722838640214,
      "num_tokens": 9138070.0,
      "step": 4090
    },
    {
      "entropy": 5.863235330581665,
      "epoch": 0.4599314876172292,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.759,
      "mean_token_accuracy": 0.16653492152690888,
      "num_tokens": 9149489.0,
      "step": 4095
    },
    {
      "entropy": 5.8261312484741214,
      "epoch": 0.46049306452518673,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.7561,
      "mean_token_accuracy": 0.1725111186504364,
      "num_tokens": 9160363.0,
      "step": 4100
    },
    {
      "entropy": 5.8416753768920895,
      "epoch": 0.4610546414331443,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.8144,
      "mean_token_accuracy": 0.16463642865419387,
      "num_tokens": 9171619.0,
      "step": 4105
    },
    {
      "entropy": 5.7307634353637695,
      "epoch": 0.46161621834110184,
      "grad_norm": 1.0,
      "learning_rate": 0.000498616475907006,
      "loss": 5.6512,
      "mean_token_accuracy": 0.17227994948625563,
      "num_tokens": 9182828.0,
      "step": 4110
    },
    {
      "entropy": 5.877857303619384,
      "epoch": 0.46217779524905933,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.8204,
      "mean_token_accuracy": 0.17060580402612685,
      "num_tokens": 9195249.0,
      "step": 4115
    },
    {
      "entropy": 5.769808769226074,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.6353,
      "mean_token_accuracy": 0.1722049430012703,
      "num_tokens": 9205770.0,
      "step": 4120
    },
    {
      "entropy": 5.683967733383179,
      "epoch": 0.46330094906497443,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1706199139356613,
      "num_tokens": 9216367.0,
      "step": 4125
    },
    {
      "entropy": 5.781439018249512,
      "epoch": 0.463862525972932,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.7368,
      "mean_token_accuracy": 0.1613730251789093,
      "num_tokens": 9228524.0,
      "step": 4130
    },
    {
      "entropy": 5.818389558792115,
      "epoch": 0.46442410288088953,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.7121,
      "mean_token_accuracy": 0.17004109174013138,
      "num_tokens": 9238907.0,
      "step": 4135
    },
    {
      "entropy": 5.794554662704468,
      "epoch": 0.4649856797888471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498589674780126,
      "loss": 5.8502,
      "mean_token_accuracy": 0.16455847769975662,
      "num_tokens": 9250491.0,
      "step": 4140
    },
    {
      "entropy": 5.864607095718384,
      "epoch": 0.46554725669680463,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.6501,
      "mean_token_accuracy": 0.17331554889678955,
      "num_tokens": 9260984.0,
      "step": 4145
    },
    {
      "entropy": 5.840168809890747,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.808,
      "mean_token_accuracy": 0.1647871918976307,
      "num_tokens": 9272416.0,
      "step": 4150
    },
    {
      "entropy": 5.794215488433838,
      "epoch": 0.46667041051271974,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498576178105807,
      "loss": 5.7851,
      "mean_token_accuracy": 0.1661459356546402,
      "num_tokens": 9284513.0,
      "step": 4155
    },
    {
      "entropy": 5.86452431678772,
      "epoch": 0.4672319874206773,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.6809,
      "mean_token_accuracy": 0.17553289085626603,
      "num_tokens": 9293422.0,
      "step": 4160
    },
    {
      "entropy": 5.720726060867309,
      "epoch": 0.4677935643286348,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.6794,
      "mean_token_accuracy": 0.16760085970163346,
      "num_tokens": 9304735.0,
      "step": 4165
    },
    {
      "entropy": 5.788778066635132,
      "epoch": 0.46835514123659233,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.7873,
      "mean_token_accuracy": 0.16755305379629135,
      "num_tokens": 9315659.0,
      "step": 4170
    },
    {
      "entropy": 5.810567426681518,
      "epoch": 0.4689167181445499,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.8001,
      "mean_token_accuracy": 0.1698588401079178,
      "num_tokens": 9327106.0,
      "step": 4175
    },
    {
      "entropy": 5.9411545753479,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.8433,
      "mean_token_accuracy": 0.16779095083475112,
      "num_tokens": 9338762.0,
      "step": 4180
    },
    {
      "entropy": 5.756154012680054,
      "epoch": 0.470039871960465,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498548992554749,
      "loss": 5.7029,
      "mean_token_accuracy": 0.17006189972162247,
      "num_tokens": 9349495.0,
      "step": 4185
    },
    {
      "entropy": 5.910188484191894,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.8079,
      "mean_token_accuracy": 0.17130580097436904,
      "num_tokens": 9360830.0,
      "step": 4190
    },
    {
      "entropy": 5.705734586715698,
      "epoch": 0.4711630257763801,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.6845,
      "mean_token_accuracy": 0.16564561575651168,
      "num_tokens": 9371668.0,
      "step": 4195
    },
    {
      "entropy": 5.817333698272705,
      "epoch": 0.47172460268433763,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16629635244607927,
      "num_tokens": 9383437.0,
      "step": 4200
    },
    {
      "entropy": 5.768382501602173,
      "epoch": 0.4722861795922952,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.6339,
      "mean_token_accuracy": 0.17444221079349517,
      "num_tokens": 9394597.0,
      "step": 4205
    },
    {
      "entropy": 5.6198773860931395,
      "epoch": 0.47284775650025274,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.577,
      "mean_token_accuracy": 0.17500848025083543,
      "num_tokens": 9404322.0,
      "step": 4210
    },
    {
      "entropy": 5.8928163051605225,
      "epoch": 0.47340933340821023,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.9324,
      "mean_token_accuracy": 0.15648610070347785,
      "num_tokens": 9416740.0,
      "step": 4215
    },
    {
      "entropy": 5.868355560302734,
      "epoch": 0.4739709103161678,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.7415,
      "mean_token_accuracy": 0.17201738208532333,
      "num_tokens": 9428617.0,
      "step": 4220
    },
    {
      "entropy": 5.704996824264526,
      "epoch": 0.47453248722412533,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.6938,
      "mean_token_accuracy": 0.1724829375743866,
      "num_tokens": 9439686.0,
      "step": 4225
    },
    {
      "entropy": 5.8679687023162845,
      "epoch": 0.4750940641320829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17257601916790008,
      "num_tokens": 9450378.0,
      "step": 4230
    },
    {
      "entropy": 5.819633293151855,
      "epoch": 0.47565564104004043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.7126,
      "mean_token_accuracy": 0.1686561957001686,
      "num_tokens": 9461659.0,
      "step": 4235
    },
    {
      "entropy": 5.743727445602417,
      "epoch": 0.476217217947998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.7547,
      "mean_token_accuracy": 0.17106645703315734,
      "num_tokens": 9472528.0,
      "step": 4240
    },
    {
      "entropy": 5.846922206878662,
      "epoch": 0.47677879485595553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.6776,
      "mean_token_accuracy": 0.17797594964504243,
      "num_tokens": 9483388.0,
      "step": 4245
    },
    {
      "entropy": 5.644798994064331,
      "epoch": 0.4773403717639131,
      "grad_norm": 1.0,
      "learning_rate": 0.000498489211510521,
      "loss": 5.5383,
      "mean_token_accuracy": 0.1787099227309227,
      "num_tokens": 9495178.0,
      "step": 4250
    },
    {
      "entropy": 5.699044275283813,
      "epoch": 0.47790194867187064,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.7797,
      "mean_token_accuracy": 0.16378788501024247,
      "num_tokens": 9506226.0,
      "step": 4255
    },
    {
      "entropy": 5.863605213165283,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.6729,
      "mean_token_accuracy": 0.16951856166124343,
      "num_tokens": 9517671.0,
      "step": 4260
    },
    {
      "entropy": 5.764629888534546,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.7478,
      "mean_token_accuracy": 0.1722684383392334,
      "num_tokens": 9527637.0,
      "step": 4265
    },
    {
      "entropy": 5.832716703414917,
      "epoch": 0.47958667939574323,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498470575384638,
      "loss": 5.8458,
      "mean_token_accuracy": 0.16255225837230683,
      "num_tokens": 9539686.0,
      "step": 4270
    },
    {
      "entropy": 5.893230867385864,
      "epoch": 0.4801482563037008,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.8008,
      "mean_token_accuracy": 0.17014207541942597,
      "num_tokens": 9550775.0,
      "step": 4275
    },
    {
      "entropy": 5.86106767654419,
      "epoch": 0.48070983321165833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.7186,
      "mean_token_accuracy": 0.1644670069217682,
      "num_tokens": 9561760.0,
      "step": 4280
    },
    {
      "entropy": 5.694846820831299,
      "epoch": 0.4812714101196159,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.7358,
      "mean_token_accuracy": 0.1661587029695511,
      "num_tokens": 9573899.0,
      "step": 4285
    },
    {
      "entropy": 5.78567066192627,
      "epoch": 0.48183298702757343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498451825412523,
      "loss": 5.7017,
      "mean_token_accuracy": 0.16853241473436356,
      "num_tokens": 9584336.0,
      "step": 4290
    },
    {
      "entropy": 5.927900266647339,
      "epoch": 0.482394563935531,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.88,
      "mean_token_accuracy": 0.162874473631382,
      "num_tokens": 9596614.0,
      "step": 4295
    },
    {
      "entropy": 5.835678482055664,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.7325,
      "mean_token_accuracy": 0.17014361917972565,
      "num_tokens": 9606796.0,
      "step": 4300
    },
    {
      "entropy": 5.778892517089844,
      "epoch": 0.4835177177514461,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.7093,
      "mean_token_accuracy": 0.16853998601436615,
      "num_tokens": 9617705.0,
      "step": 4305
    },
    {
      "entropy": 5.704960536956787,
      "epoch": 0.4840792946594036,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.7265,
      "mean_token_accuracy": 0.17152929455041885,
      "num_tokens": 9628098.0,
      "step": 4310
    },
    {
      "entropy": 5.815963077545166,
      "epoch": 0.48464087156736113,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17872087508440018,
      "num_tokens": 9637697.0,
      "step": 4315
    },
    {
      "entropy": 5.674087190628052,
      "epoch": 0.4852024484753187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.5793,
      "mean_token_accuracy": 0.17939777821302413,
      "num_tokens": 9648221.0,
      "step": 4320
    },
    {
      "entropy": 5.805170249938965,
      "epoch": 0.48576402538327623,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.7692,
      "mean_token_accuracy": 0.1637537583708763,
      "num_tokens": 9660182.0,
      "step": 4325
    },
    {
      "entropy": 5.825180864334106,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.777,
      "mean_token_accuracy": 0.16494843363761902,
      "num_tokens": 9672121.0,
      "step": 4330
    },
    {
      "entropy": 5.823313140869141,
      "epoch": 0.48688717919919133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.6781,
      "mean_token_accuracy": 0.17020544558763503,
      "num_tokens": 9682651.0,
      "step": 4335
    },
    {
      "entropy": 5.806669759750366,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.8241,
      "mean_token_accuracy": 0.16280035227537154,
      "num_tokens": 9693989.0,
      "step": 4340
    },
    {
      "entropy": 5.678175973892212,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17800934761762618,
      "num_tokens": 9703802.0,
      "step": 4345
    },
    {
      "entropy": 5.73551115989685,
      "epoch": 0.488571909923064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498394892514596,
      "loss": 5.6309,
      "mean_token_accuracy": 0.17766664773225785,
      "num_tokens": 9714287.0,
      "step": 4350
    },
    {
      "entropy": 5.779881143569947,
      "epoch": 0.48913348683102154,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.655,
      "mean_token_accuracy": 0.17147285342216492,
      "num_tokens": 9724819.0,
      "step": 4355
    },
    {
      "entropy": 5.801750993728637,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.8086,
      "mean_token_accuracy": 0.16176895052194595,
      "num_tokens": 9736326.0,
      "step": 4360
    },
    {
      "entropy": 5.879187440872192,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.8104,
      "mean_token_accuracy": 0.16969723701477052,
      "num_tokens": 9747883.0,
      "step": 4365
    },
    {
      "entropy": 5.767481184005737,
      "epoch": 0.49081821755489413,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.6812,
      "mean_token_accuracy": 0.17253028899431228,
      "num_tokens": 9758191.0,
      "step": 4370
    },
    {
      "entropy": 5.8729065418243405,
      "epoch": 0.4913797944628517,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.7807,
      "mean_token_accuracy": 0.16696447283029556,
      "num_tokens": 9768852.0,
      "step": 4375
    },
    {
      "entropy": 5.849040746688843,
      "epoch": 0.49194137137080923,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.7301,
      "mean_token_accuracy": 0.16603904068470002,
      "num_tokens": 9779033.0,
      "step": 4380
    },
    {
      "entropy": 5.716362190246582,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.6373,
      "mean_token_accuracy": 0.17526944130659103,
      "num_tokens": 9790777.0,
      "step": 4385
    },
    {
      "entropy": 5.915458679199219,
      "epoch": 0.49306452518672433,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.909,
      "mean_token_accuracy": 0.15452502816915512,
      "num_tokens": 9802305.0,
      "step": 4390
    },
    {
      "entropy": 5.837260770797729,
      "epoch": 0.4936261020946819,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.7088,
      "mean_token_accuracy": 0.16492092460393906,
      "num_tokens": 9813590.0,
      "step": 4395
    },
    {
      "entropy": 5.762714004516601,
      "epoch": 0.49418767900263944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.78,
      "mean_token_accuracy": 0.1626877874135971,
      "num_tokens": 9824080.0,
      "step": 4400
    },
    {
      "entropy": 5.770403957366943,
      "epoch": 0.49474925591059693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.6743,
      "mean_token_accuracy": 0.16747995167970658,
      "num_tokens": 9835441.0,
      "step": 4405
    },
    {
      "entropy": 5.74335503578186,
      "epoch": 0.4953108328185545,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.659,
      "mean_token_accuracy": 0.1735517680644989,
      "num_tokens": 9846750.0,
      "step": 4410
    },
    {
      "entropy": 5.842858934402466,
      "epoch": 0.49587240972651203,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.81,
      "mean_token_accuracy": 0.16727619916200637,
      "num_tokens": 9857987.0,
      "step": 4415
    },
    {
      "entropy": 5.785632085800171,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.7654,
      "mean_token_accuracy": 0.16815820336341858,
      "num_tokens": 9869736.0,
      "step": 4420
    },
    {
      "entropy": 5.8420435905456545,
      "epoch": 0.49699556354242713,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498322286045317,
      "loss": 5.7281,
      "mean_token_accuracy": 0.16822068840265275,
      "num_tokens": 9881610.0,
      "step": 4425
    },
    {
      "entropy": 5.783283519744873,
      "epoch": 0.4975571404503847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.7289,
      "mean_token_accuracy": 0.16912795081734658,
      "num_tokens": 9893203.0,
      "step": 4430
    },
    {
      "entropy": 5.775019788742066,
      "epoch": 0.49811871735834223,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.6411,
      "mean_token_accuracy": 0.16674574017524718,
      "num_tokens": 9904392.0,
      "step": 4435
    },
    {
      "entropy": 5.601576042175293,
      "epoch": 0.4986802942662998,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.6408,
      "mean_token_accuracy": 0.17083830386400223,
      "num_tokens": 9916184.0,
      "step": 4440
    },
    {
      "entropy": 5.817759895324707,
      "epoch": 0.49924187117425733,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.7422,
      "mean_token_accuracy": 0.17348867505788804,
      "num_tokens": 9927516.0,
      "step": 4445
    },
    {
      "entropy": 5.817559862136841,
      "epoch": 0.4998034480822149,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.7749,
      "mean_token_accuracy": 0.16315798461437225,
      "num_tokens": 9939182.0,
      "step": 4450
    },
    {
      "entropy": 5.855092334747314,
      "epoch": 0.5003650249901724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498292795456261,
      "loss": 5.7898,
      "mean_token_accuracy": 0.16669366359710694,
      "num_tokens": 9950382.0,
      "step": 4455
    },
    {
      "entropy": 5.796814346313477,
      "epoch": 0.50092660189813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.7932,
      "mean_token_accuracy": 0.17010375410318374,
      "num_tokens": 9961053.0,
      "step": 4460
    },
    {
      "entropy": 5.765320682525635,
      "epoch": 0.5014881788060875,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.6947,
      "mean_token_accuracy": 0.17564453780651093,
      "num_tokens": 9971313.0,
      "step": 4465
    },
    {
      "entropy": 5.784207010269165,
      "epoch": 0.5020497557140451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.7036,
      "mean_token_accuracy": 0.17723664939403533,
      "num_tokens": 9981257.0,
      "step": 4470
    },
    {
      "entropy": 5.7435914993286135,
      "epoch": 0.5026113326220026,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.6139,
      "mean_token_accuracy": 0.17061744183301925,
      "num_tokens": 9991860.0,
      "step": 4475
    },
    {
      "entropy": 5.721191549301148,
      "epoch": 0.5031729095299601,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.7319,
      "mean_token_accuracy": 0.170281982421875,
      "num_tokens": 10002320.0,
      "step": 4480
    },
    {
      "entropy": 5.869790172576904,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.8147,
      "mean_token_accuracy": 0.16252802163362504,
      "num_tokens": 10013187.0,
      "step": 4485
    },
    {
      "entropy": 5.748865604400635,
      "epoch": 0.5042960633458752,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.551,
      "mean_token_accuracy": 0.18117449134588243,
      "num_tokens": 10024268.0,
      "step": 4490
    },
    {
      "entropy": 5.773206615447998,
      "epoch": 0.5048576402538327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.8156,
      "mean_token_accuracy": 0.16586428880691528,
      "num_tokens": 10034738.0,
      "step": 4495
    },
    {
      "entropy": 5.867972803115845,
      "epoch": 0.5054192171617903,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.7633,
      "mean_token_accuracy": 0.16974856257438659,
      "num_tokens": 10045666.0,
      "step": 4500
    },
    {
      "entropy": 5.755939483642578,
      "epoch": 0.5059807940697478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17190644890069962,
      "num_tokens": 10057084.0,
      "step": 4505
    },
    {
      "entropy": 5.667339277267456,
      "epoch": 0.5065423709777054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.6199,
      "mean_token_accuracy": 0.17011208087205887,
      "num_tokens": 10066754.0,
      "step": 4510
    },
    {
      "entropy": 5.678653812408447,
      "epoch": 0.5071039478856629,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.6282,
      "mean_token_accuracy": 0.17506971210241318,
      "num_tokens": 10078139.0,
      "step": 4515
    },
    {
      "entropy": 5.80544900894165,
      "epoch": 0.5076655247936205,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.6884,
      "mean_token_accuracy": 0.16866966784000398,
      "num_tokens": 10089652.0,
      "step": 4520
    },
    {
      "entropy": 5.6051665306091305,
      "epoch": 0.508227101701578,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.5358,
      "mean_token_accuracy": 0.17593766152858734,
      "num_tokens": 10100642.0,
      "step": 4525
    },
    {
      "entropy": 5.782035064697266,
      "epoch": 0.5087886786095356,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.7448,
      "mean_token_accuracy": 0.16809683442115783,
      "num_tokens": 10112036.0,
      "step": 4530
    },
    {
      "entropy": 5.784066915512085,
      "epoch": 0.5093502555174931,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.6198,
      "mean_token_accuracy": 0.17079970091581345,
      "num_tokens": 10121653.0,
      "step": 4535
    },
    {
      "entropy": 5.767913055419922,
      "epoch": 0.5099118324254507,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1688558965921402,
      "num_tokens": 10132271.0,
      "step": 4540
    },
    {
      "entropy": 5.658374261856079,
      "epoch": 0.5104734093334082,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.606,
      "mean_token_accuracy": 0.17197940349578858,
      "num_tokens": 10143411.0,
      "step": 4545
    },
    {
      "entropy": 5.847708702087402,
      "epoch": 0.5110349862413658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1733257830142975,
      "num_tokens": 10154338.0,
      "step": 4550
    },
    {
      "entropy": 5.802937316894531,
      "epoch": 0.5115965631493233,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.7516,
      "mean_token_accuracy": 0.17324941903352736,
      "num_tokens": 10164810.0,
      "step": 4555
    },
    {
      "entropy": 5.7534888744354244,
      "epoch": 0.5121581400572809,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.6731,
      "mean_token_accuracy": 0.17414593547582627,
      "num_tokens": 10176323.0,
      "step": 4560
    },
    {
      "entropy": 5.736021661758423,
      "epoch": 0.5127197169652384,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.7073,
      "mean_token_accuracy": 0.17113523185253143,
      "num_tokens": 10187110.0,
      "step": 4565
    },
    {
      "entropy": 5.761826086044311,
      "epoch": 0.513281293873196,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.6717,
      "mean_token_accuracy": 0.16766324490308762,
      "num_tokens": 10199099.0,
      "step": 4570
    },
    {
      "entropy": 5.785593271255493,
      "epoch": 0.5138428707811534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.6747,
      "mean_token_accuracy": 0.1726672261953354,
      "num_tokens": 10209726.0,
      "step": 4575
    },
    {
      "entropy": 5.755669784545899,
      "epoch": 0.514404447689111,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049816716337126,
      "loss": 5.7604,
      "mean_token_accuracy": 0.16705486178398132,
      "num_tokens": 10220858.0,
      "step": 4580
    },
    {
      "entropy": 5.644749641418457,
      "epoch": 0.5149660245970685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.577,
      "mean_token_accuracy": 0.1802625373005867,
      "num_tokens": 10232250.0,
      "step": 4585
    },
    {
      "entropy": 5.893209981918335,
      "epoch": 0.5155276015050261,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1636040985584259,
      "num_tokens": 10244584.0,
      "step": 4590
    },
    {
      "entropy": 5.823698377609253,
      "epoch": 0.5160891784129836,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.6569,
      "mean_token_accuracy": 0.16793317794799806,
      "num_tokens": 10255926.0,
      "step": 4595
    },
    {
      "entropy": 5.751343584060669,
      "epoch": 0.5166507553209412,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498146650039855,
      "loss": 5.7326,
      "mean_token_accuracy": 0.17038354724645616,
      "num_tokens": 10266691.0,
      "step": 4600
    },
    {
      "entropy": 5.818839836120605,
      "epoch": 0.5172123322288987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.7986,
      "mean_token_accuracy": 0.1667577788233757,
      "num_tokens": 10277744.0,
      "step": 4605
    },
    {
      "entropy": 5.718520307540894,
      "epoch": 0.5177739091368563,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.6473,
      "mean_token_accuracy": 0.17276567965745926,
      "num_tokens": 10288792.0,
      "step": 4610
    },
    {
      "entropy": 5.818396329879761,
      "epoch": 0.5183354860448138,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.6795,
      "mean_token_accuracy": 0.16973294615745543,
      "num_tokens": 10299569.0,
      "step": 4615
    },
    {
      "entropy": 5.722053909301758,
      "epoch": 0.5188970629527714,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.6642,
      "mean_token_accuracy": 0.17573922574520112,
      "num_tokens": 10309742.0,
      "step": 4620
    },
    {
      "entropy": 5.8336011409759525,
      "epoch": 0.5194586398607289,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.7048,
      "mean_token_accuracy": 0.1665973499417305,
      "num_tokens": 10320603.0,
      "step": 4625
    },
    {
      "entropy": 5.736460113525391,
      "epoch": 0.5200202167686865,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17377568930387496,
      "num_tokens": 10331498.0,
      "step": 4630
    },
    {
      "entropy": 5.715865516662598,
      "epoch": 0.520581793676644,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.6988,
      "mean_token_accuracy": 0.17312800139188766,
      "num_tokens": 10342002.0,
      "step": 4635
    },
    {
      "entropy": 5.749444532394409,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.6599,
      "mean_token_accuracy": 0.16870316565036775,
      "num_tokens": 10351797.0,
      "step": 4640
    },
    {
      "entropy": 5.719133377075195,
      "epoch": 0.5217049474925591,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.6073,
      "mean_token_accuracy": 0.17236316353082656,
      "num_tokens": 10362538.0,
      "step": 4645
    },
    {
      "entropy": 5.737236261367798,
      "epoch": 0.5222665244005167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.6399,
      "mean_token_accuracy": 0.17452512383461,
      "num_tokens": 10373365.0,
      "step": 4650
    },
    {
      "entropy": 5.641112422943115,
      "epoch": 0.5228281013084742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.6009,
      "mean_token_accuracy": 0.17220466434955597,
      "num_tokens": 10385087.0,
      "step": 4655
    },
    {
      "entropy": 5.7947471141815186,
      "epoch": 0.5233896782164318,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.7315,
      "mean_token_accuracy": 0.1702256292104721,
      "num_tokens": 10396014.0,
      "step": 4660
    },
    {
      "entropy": 5.765877151489258,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.6451,
      "mean_token_accuracy": 0.1674633339047432,
      "num_tokens": 10407844.0,
      "step": 4665
    },
    {
      "entropy": 5.6722806930542,
      "epoch": 0.5245128320323468,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.7118,
      "mean_token_accuracy": 0.16796951591968537,
      "num_tokens": 10419892.0,
      "step": 4670
    },
    {
      "entropy": 5.8044257164001465,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.6947,
      "mean_token_accuracy": 0.16804990619421006,
      "num_tokens": 10430686.0,
      "step": 4675
    },
    {
      "entropy": 5.7502185821533205,
      "epoch": 0.5256359858482619,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.7058,
      "mean_token_accuracy": 0.16406390219926834,
      "num_tokens": 10443785.0,
      "step": 4680
    },
    {
      "entropy": 5.69890947341919,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.6032,
      "mean_token_accuracy": 0.17852845937013626,
      "num_tokens": 10454889.0,
      "step": 4685
    },
    {
      "entropy": 5.756927967071533,
      "epoch": 0.526759139664177,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.7589,
      "mean_token_accuracy": 0.1650112196803093,
      "num_tokens": 10466525.0,
      "step": 4690
    },
    {
      "entropy": 5.854841518402099,
      "epoch": 0.5273207165721345,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.7132,
      "mean_token_accuracy": 0.16974902898073196,
      "num_tokens": 10476329.0,
      "step": 4695
    },
    {
      "entropy": 5.827983236312866,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.7684,
      "mean_token_accuracy": 0.16705012321472168,
      "num_tokens": 10487178.0,
      "step": 4700
    },
    {
      "entropy": 5.820889663696289,
      "epoch": 0.5284438703880496,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.7131,
      "mean_token_accuracy": 0.1603904828429222,
      "num_tokens": 10499293.0,
      "step": 4705
    },
    {
      "entropy": 5.8819056987762455,
      "epoch": 0.5290054472960072,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049803179494862,
      "loss": 5.8989,
      "mean_token_accuracy": 0.1597862184047699,
      "num_tokens": 10511762.0,
      "step": 4710
    },
    {
      "entropy": 5.775479793548584,
      "epoch": 0.5295670242039647,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.6241,
      "mean_token_accuracy": 0.17840833216905594,
      "num_tokens": 10521725.0,
      "step": 4715
    },
    {
      "entropy": 5.639372968673706,
      "epoch": 0.5301286011119223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17198170274496077,
      "num_tokens": 10531764.0,
      "step": 4720
    },
    {
      "entropy": 5.737982130050659,
      "epoch": 0.5306901780198798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.6481,
      "mean_token_accuracy": 0.17444745302200318,
      "num_tokens": 10542127.0,
      "step": 4725
    },
    {
      "entropy": 5.851061010360718,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.7351,
      "mean_token_accuracy": 0.16558413803577424,
      "num_tokens": 10552580.0,
      "step": 4730
    },
    {
      "entropy": 5.777568435668945,
      "epoch": 0.5318133318357949,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.6917,
      "mean_token_accuracy": 0.1689352035522461,
      "num_tokens": 10562961.0,
      "step": 4735
    },
    {
      "entropy": 5.761757183074951,
      "epoch": 0.5323749087437525,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.6857,
      "mean_token_accuracy": 0.174189792573452,
      "num_tokens": 10573210.0,
      "step": 4740
    },
    {
      "entropy": 5.648348140716553,
      "epoch": 0.53293648565171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.6351,
      "mean_token_accuracy": 0.16878298968076705,
      "num_tokens": 10583878.0,
      "step": 4745
    },
    {
      "entropy": 5.798595285415649,
      "epoch": 0.5334980625596676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.7171,
      "mean_token_accuracy": 0.16758286952972412,
      "num_tokens": 10593420.0,
      "step": 4750
    },
    {
      "entropy": 5.799276733398438,
      "epoch": 0.5340596394676251,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497983818025976,
      "loss": 5.7368,
      "mean_token_accuracy": 0.16662465855479242,
      "num_tokens": 10604213.0,
      "step": 4755
    },
    {
      "entropy": 5.797427368164063,
      "epoch": 0.5346212163755827,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.7054,
      "mean_token_accuracy": 0.1650031179189682,
      "num_tokens": 10615502.0,
      "step": 4760
    },
    {
      "entropy": 5.812700986862183,
      "epoch": 0.5351827932835401,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.7668,
      "mean_token_accuracy": 0.1636280044913292,
      "num_tokens": 10627589.0,
      "step": 4765
    },
    {
      "entropy": 5.721961879730225,
      "epoch": 0.5357443701914977,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.6092,
      "mean_token_accuracy": 0.17420833557844162,
      "num_tokens": 10638365.0,
      "step": 4770
    },
    {
      "entropy": 5.775761985778809,
      "epoch": 0.5363059470994552,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497962310344703,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1731460452079773,
      "num_tokens": 10650664.0,
      "step": 4775
    },
    {
      "entropy": 5.7953284740447994,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049795691567594,
      "loss": 5.7564,
      "mean_token_accuracy": 0.17478667497634887,
      "num_tokens": 10662418.0,
      "step": 4780
    },
    {
      "entropy": 5.774968767166138,
      "epoch": 0.5374291009153703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.6265,
      "mean_token_accuracy": 0.1747649058699608,
      "num_tokens": 10672793.0,
      "step": 4785
    },
    {
      "entropy": 5.7399228572845455,
      "epoch": 0.5379906778233279,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.6728,
      "mean_token_accuracy": 0.1787702426314354,
      "num_tokens": 10684082.0,
      "step": 4790
    },
    {
      "entropy": 5.687348508834839,
      "epoch": 0.5385522547312854,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.6123,
      "mean_token_accuracy": 0.17388249784708024,
      "num_tokens": 10694490.0,
      "step": 4795
    },
    {
      "entropy": 5.682422208786011,
      "epoch": 0.539113831639243,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.6102,
      "mean_token_accuracy": 0.1730710655450821,
      "num_tokens": 10704731.0,
      "step": 4800
    },
    {
      "entropy": 5.7284330368042,
      "epoch": 0.5396754085472005,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.6458,
      "mean_token_accuracy": 0.17584740072488786,
      "num_tokens": 10715735.0,
      "step": 4805
    },
    {
      "entropy": 5.70305609703064,
      "epoch": 0.5402369854551581,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.6098,
      "mean_token_accuracy": 0.1710106059908867,
      "num_tokens": 10726951.0,
      "step": 4810
    },
    {
      "entropy": 5.645564556121826,
      "epoch": 0.5407985623631156,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.6851,
      "mean_token_accuracy": 0.16763894110918046,
      "num_tokens": 10737609.0,
      "step": 4815
    },
    {
      "entropy": 5.723374080657959,
      "epoch": 0.5413601392710732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.5759,
      "mean_token_accuracy": 0.17617718130350113,
      "num_tokens": 10748957.0,
      "step": 4820
    },
    {
      "entropy": 5.7252906322479244,
      "epoch": 0.5419217161790307,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.7624,
      "mean_token_accuracy": 0.16397164314985274,
      "num_tokens": 10759398.0,
      "step": 4825
    },
    {
      "entropy": 5.8732177734375,
      "epoch": 0.5424832930869883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.7446,
      "mean_token_accuracy": 0.16531347930431367,
      "num_tokens": 10770436.0,
      "step": 4830
    },
    {
      "entropy": 5.633301305770874,
      "epoch": 0.5430448699949458,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.5752,
      "mean_token_accuracy": 0.1801727831363678,
      "num_tokens": 10781160.0,
      "step": 4835
    },
    {
      "entropy": 5.753065013885498,
      "epoch": 0.5436064469029034,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.7015,
      "mean_token_accuracy": 0.1770429030060768,
      "num_tokens": 10793170.0,
      "step": 4840
    },
    {
      "entropy": 5.803515291213989,
      "epoch": 0.5441680238108609,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.6435,
      "mean_token_accuracy": 0.1693033382296562,
      "num_tokens": 10803515.0,
      "step": 4845
    },
    {
      "entropy": 5.7509010314941404,
      "epoch": 0.5447296007188185,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.6651,
      "mean_token_accuracy": 0.1700579084455967,
      "num_tokens": 10814508.0,
      "step": 4850
    },
    {
      "entropy": 5.723728084564209,
      "epoch": 0.545291177626776,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.7722,
      "mean_token_accuracy": 0.16889193505048752,
      "num_tokens": 10825924.0,
      "step": 4855
    },
    {
      "entropy": 5.863556575775147,
      "epoch": 0.5458527545347336,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.7463,
      "mean_token_accuracy": 0.16716790348291397,
      "num_tokens": 10837854.0,
      "step": 4860
    },
    {
      "entropy": 5.762776231765747,
      "epoch": 0.546414331442691,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.6418,
      "mean_token_accuracy": 0.1727205768227577,
      "num_tokens": 10848279.0,
      "step": 4865
    },
    {
      "entropy": 5.611614179611206,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.654,
      "mean_token_accuracy": 0.1656600594520569,
      "num_tokens": 10859615.0,
      "step": 4870
    },
    {
      "entropy": 5.882994222640991,
      "epoch": 0.5475374852586061,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.6263,
      "mean_token_accuracy": 0.17195343524217604,
      "num_tokens": 10871388.0,
      "step": 4875
    },
    {
      "entropy": 5.682742977142334,
      "epoch": 0.5480990621665637,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1701396718621254,
      "num_tokens": 10882411.0,
      "step": 4880
    },
    {
      "entropy": 5.731553316116333,
      "epoch": 0.5486606390745212,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.7924,
      "mean_token_accuracy": 0.16876837164163588,
      "num_tokens": 10894116.0,
      "step": 4885
    },
    {
      "entropy": 5.752131938934326,
      "epoch": 0.5492222159824788,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.5682,
      "mean_token_accuracy": 0.17538228034973144,
      "num_tokens": 10904514.0,
      "step": 4890
    },
    {
      "entropy": 5.689201021194458,
      "epoch": 0.5497837928904363,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.6134,
      "mean_token_accuracy": 0.17261481434106826,
      "num_tokens": 10915602.0,
      "step": 4895
    },
    {
      "entropy": 5.682745933532715,
      "epoch": 0.5503453697983939,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.6033,
      "mean_token_accuracy": 0.17805052399635315,
      "num_tokens": 10926172.0,
      "step": 4900
    },
    {
      "entropy": 5.778084802627563,
      "epoch": 0.5509069467063514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.7268,
      "mean_token_accuracy": 0.16874644458293914,
      "num_tokens": 10938435.0,
      "step": 4905
    },
    {
      "entropy": 5.671151447296142,
      "epoch": 0.551468523614309,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.6205,
      "mean_token_accuracy": 0.17252751886844636,
      "num_tokens": 10949118.0,
      "step": 4910
    },
    {
      "entropy": 5.777276611328125,
      "epoch": 0.5520301005222665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.657,
      "mean_token_accuracy": 0.17057308554649353,
      "num_tokens": 10960120.0,
      "step": 4915
    },
    {
      "entropy": 5.742342805862426,
      "epoch": 0.5525916774302241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497802983376339,
      "loss": 5.6397,
      "mean_token_accuracy": 0.17520381212234498,
      "num_tokens": 10971210.0,
      "step": 4920
    },
    {
      "entropy": 5.655767631530762,
      "epoch": 0.5531532543381816,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.6121,
      "mean_token_accuracy": 0.1807950958609581,
      "num_tokens": 10983446.0,
      "step": 4925
    },
    {
      "entropy": 5.819978809356689,
      "epoch": 0.5537148312461392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.7389,
      "mean_token_accuracy": 0.16750801503658294,
      "num_tokens": 10994776.0,
      "step": 4930
    },
    {
      "entropy": 5.725039482116699,
      "epoch": 0.5542764081540967,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497786160684314,
      "loss": 5.6497,
      "mean_token_accuracy": 0.16738180369138717,
      "num_tokens": 11005280.0,
      "step": 4935
    },
    {
      "entropy": 5.634877109527588,
      "epoch": 0.5548379850620543,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.6471,
      "mean_token_accuracy": 0.1671821042895317,
      "num_tokens": 11017509.0,
      "step": 4940
    },
    {
      "entropy": 5.773461437225341,
      "epoch": 0.5553995619700118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497774910088011,
      "loss": 5.6786,
      "mean_token_accuracy": 0.17146782875061034,
      "num_tokens": 11029087.0,
      "step": 4945
    },
    {
      "entropy": 5.748759794235229,
      "epoch": 0.5559611388779694,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.7066,
      "mean_token_accuracy": 0.16947160810232162,
      "num_tokens": 11041144.0,
      "step": 4950
    },
    {
      "entropy": 5.878700208663941,
      "epoch": 0.5565227157859269,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049776363111875,
      "loss": 5.7887,
      "mean_token_accuracy": 0.16643262207508086,
      "num_tokens": 11054050.0,
      "step": 4955
    },
    {
      "entropy": 5.735003042221069,
      "epoch": 0.5570842926938844,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.6076,
      "mean_token_accuracy": 0.17405114471912383,
      "num_tokens": 11064257.0,
      "step": 4960
    },
    {
      "entropy": 5.790982389450074,
      "epoch": 0.5576458696018419,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.7207,
      "mean_token_accuracy": 0.17392538487911224,
      "num_tokens": 11075568.0,
      "step": 4965
    },
    {
      "entropy": 5.759704113006592,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.6607,
      "mean_token_accuracy": 0.17160382568836213,
      "num_tokens": 11087078.0,
      "step": 4970
    },
    {
      "entropy": 5.8532062530517575,
      "epoch": 0.558769023417757,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.8178,
      "mean_token_accuracy": 0.16156835705041886,
      "num_tokens": 11100655.0,
      "step": 4975
    },
    {
      "entropy": 5.802464723587036,
      "epoch": 0.5593306003257146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.7037,
      "mean_token_accuracy": 0.170867520570755,
      "num_tokens": 11110675.0,
      "step": 4980
    },
    {
      "entropy": 5.717938423156738,
      "epoch": 0.5598921772336721,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.712,
      "mean_token_accuracy": 0.16717755943536758,
      "num_tokens": 11122337.0,
      "step": 4985
    },
    {
      "entropy": 5.822988271713257,
      "epoch": 0.5604537541416297,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497723931310065,
      "loss": 5.6137,
      "mean_token_accuracy": 0.17525150924921035,
      "num_tokens": 11132672.0,
      "step": 4990
    },
    {
      "entropy": 5.674412488937378,
      "epoch": 0.5610153310495872,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.6591,
      "mean_token_accuracy": 0.1736418917775154,
      "num_tokens": 11143740.0,
      "step": 4995
    },
    {
      "entropy": 5.701447057723999,
      "epoch": 0.5615769079575448,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.6127,
      "mean_token_accuracy": 0.17458404526114463,
      "num_tokens": 11153722.0,
      "step": 5000
    },
    {
      "entropy": 5.794809436798095,
      "epoch": 0.5621384848655023,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.7267,
      "mean_token_accuracy": 0.16622802317142488,
      "num_tokens": 11165042.0,
      "step": 5005
    },
    {
      "entropy": 5.705378532409668,
      "epoch": 0.5627000617734599,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.5937,
      "mean_token_accuracy": 0.1812054619193077,
      "num_tokens": 11176834.0,
      "step": 5010
    },
    {
      "entropy": 5.714614629745483,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.6806,
      "mean_token_accuracy": 0.17363234609365463,
      "num_tokens": 11188442.0,
      "step": 5015
    },
    {
      "entropy": 5.6922242641448975,
      "epoch": 0.563823215589375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497689626327256,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17609855234622956,
      "num_tokens": 11197855.0,
      "step": 5020
    },
    {
      "entropy": 5.740423107147217,
      "epoch": 0.5643847924973325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497683884012223,
      "loss": 5.6819,
      "mean_token_accuracy": 0.16573509573936462,
      "num_tokens": 11209300.0,
      "step": 5025
    },
    {
      "entropy": 5.729487133026123,
      "epoch": 0.5649463694052901,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.6998,
      "mean_token_accuracy": 0.1707627922296524,
      "num_tokens": 11222175.0,
      "step": 5030
    },
    {
      "entropy": 5.830688619613648,
      "epoch": 0.5655079463132476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.6957,
      "mean_token_accuracy": 0.16623380333185195,
      "num_tokens": 11233277.0,
      "step": 5035
    },
    {
      "entropy": 5.7766283512115475,
      "epoch": 0.5660695232212052,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.7199,
      "mean_token_accuracy": 0.1678628459572792,
      "num_tokens": 11245241.0,
      "step": 5040
    },
    {
      "entropy": 5.7631550312042235,
      "epoch": 0.5666311001291627,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.7385,
      "mean_token_accuracy": 0.17289155423641206,
      "num_tokens": 11257359.0,
      "step": 5045
    },
    {
      "entropy": 5.713629865646363,
      "epoch": 0.5671926770371203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.6593,
      "mean_token_accuracy": 0.17672290354967118,
      "num_tokens": 11270651.0,
      "step": 5050
    },
    {
      "entropy": 5.727716875076294,
      "epoch": 0.5677542539450777,
      "grad_norm": 1.0,
      "learning_rate": 0.000497649281231274,
      "loss": 5.63,
      "mean_token_accuracy": 0.1713034063577652,
      "num_tokens": 11280164.0,
      "step": 5055
    },
    {
      "entropy": 5.755387592315674,
      "epoch": 0.5683158308530353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.586,
      "mean_token_accuracy": 0.17397502511739732,
      "num_tokens": 11290156.0,
      "step": 5060
    },
    {
      "entropy": 5.699924278259277,
      "epoch": 0.5688774077609928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.6365,
      "mean_token_accuracy": 0.1798318862915039,
      "num_tokens": 11301929.0,
      "step": 5065
    },
    {
      "entropy": 5.655545091629028,
      "epoch": 0.5694389846689504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1807105928659439,
      "num_tokens": 11312805.0,
      "step": 5070
    },
    {
      "entropy": 5.56551160812378,
      "epoch": 0.5700005615769079,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.4619,
      "mean_token_accuracy": 0.18117846846580504,
      "num_tokens": 11322316.0,
      "step": 5075
    },
    {
      "entropy": 5.671806049346924,
      "epoch": 0.5705621384848655,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497620250624724,
      "loss": 5.6488,
      "mean_token_accuracy": 0.1746806189417839,
      "num_tokens": 11332994.0,
      "step": 5080
    },
    {
      "entropy": 5.757491254806519,
      "epoch": 0.571123715392823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497614423237728,
      "loss": 5.6294,
      "mean_token_accuracy": 0.17021082043647767,
      "num_tokens": 11344355.0,
      "step": 5085
    },
    {
      "entropy": 5.715160655975342,
      "epoch": 0.5716852923007806,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.6708,
      "mean_token_accuracy": 0.16981630623340607,
      "num_tokens": 11356275.0,
      "step": 5090
    },
    {
      "entropy": 5.7371093273162845,
      "epoch": 0.5722468692087381,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.6543,
      "mean_token_accuracy": 0.1713169738650322,
      "num_tokens": 11368624.0,
      "step": 5095
    },
    {
      "entropy": 5.846412563323975,
      "epoch": 0.5728084461166957,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497596898548706,
      "loss": 5.7707,
      "mean_token_accuracy": 0.17011373490095139,
      "num_tokens": 11380520.0,
      "step": 5100
    },
    {
      "entropy": 5.677037048339844,
      "epoch": 0.5733700230246532,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.6728,
      "mean_token_accuracy": 0.17191081047058104,
      "num_tokens": 11392293.0,
      "step": 5105
    },
    {
      "entropy": 5.776775360107422,
      "epoch": 0.5739315999326108,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.7284,
      "mean_token_accuracy": 0.16902240514755248,
      "num_tokens": 11403866.0,
      "step": 5110
    },
    {
      "entropy": 5.731493425369263,
      "epoch": 0.5744931768405683,
      "grad_norm": 0.875,
      "learning_rate": 0.000497579310071541,
      "loss": 5.5573,
      "mean_token_accuracy": 0.18312533646821977,
      "num_tokens": 11415334.0,
      "step": 5115
    },
    {
      "entropy": 5.651573848724365,
      "epoch": 0.5750547537485259,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.6967,
      "mean_token_accuracy": 0.1741857036948204,
      "num_tokens": 11426840.0,
      "step": 5120
    },
    {
      "entropy": 5.840533256530762,
      "epoch": 0.5756163306564834,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.8478,
      "mean_token_accuracy": 0.16357955038547517,
      "num_tokens": 11438497.0,
      "step": 5125
    },
    {
      "entropy": 5.727730226516724,
      "epoch": 0.576177907564441,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.5642,
      "mean_token_accuracy": 0.177400603890419,
      "num_tokens": 11448705.0,
      "step": 5130
    },
    {
      "entropy": 5.817451763153076,
      "epoch": 0.5767394844723985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.8326,
      "mean_token_accuracy": 0.1604374922811985,
      "num_tokens": 11460610.0,
      "step": 5135
    },
    {
      "entropy": 5.747437524795532,
      "epoch": 0.5773010613803561,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.5884,
      "mean_token_accuracy": 0.17517827302217484,
      "num_tokens": 11471170.0,
      "step": 5140
    },
    {
      "entropy": 5.702098035812378,
      "epoch": 0.5778626382883136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.578,
      "mean_token_accuracy": 0.18181243389844895,
      "num_tokens": 11482090.0,
      "step": 5145
    },
    {
      "entropy": 5.636244106292724,
      "epoch": 0.5784242151962711,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.6729,
      "mean_token_accuracy": 0.1727846696972847,
      "num_tokens": 11493650.0,
      "step": 5150
    },
    {
      "entropy": 5.775764989852905,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.6583,
      "mean_token_accuracy": 0.16539878398180008,
      "num_tokens": 11505903.0,
      "step": 5155
    },
    {
      "entropy": 5.8385509014129635,
      "epoch": 0.5795473690121862,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.7001,
      "mean_token_accuracy": 0.16813577711582184,
      "num_tokens": 11517194.0,
      "step": 5160
    },
    {
      "entropy": 5.7446431636810305,
      "epoch": 0.5801089459201437,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.7074,
      "mean_token_accuracy": 0.17194776237010956,
      "num_tokens": 11528683.0,
      "step": 5165
    },
    {
      "entropy": 5.736820793151855,
      "epoch": 0.5806705228281013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.7343,
      "mean_token_accuracy": 0.17425447255373,
      "num_tokens": 11541940.0,
      "step": 5170
    },
    {
      "entropy": 5.717590761184693,
      "epoch": 0.5812320997360588,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.6288,
      "mean_token_accuracy": 0.17126648128032684,
      "num_tokens": 11553045.0,
      "step": 5175
    },
    {
      "entropy": 5.62059440612793,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.5967,
      "mean_token_accuracy": 0.17394222617149352,
      "num_tokens": 11564401.0,
      "step": 5180
    },
    {
      "entropy": 5.729012060165405,
      "epoch": 0.5823552535519739,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.6396,
      "mean_token_accuracy": 0.17092307806015014,
      "num_tokens": 11576424.0,
      "step": 5185
    },
    {
      "entropy": 5.755575132369995,
      "epoch": 0.5829168304599315,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049749041104062,
      "loss": 5.6283,
      "mean_token_accuracy": 0.17128757536411285,
      "num_tokens": 11587853.0,
      "step": 5190
    },
    {
      "entropy": 5.693592929840088,
      "epoch": 0.583478407367889,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.6439,
      "mean_token_accuracy": 0.17354430258274078,
      "num_tokens": 11597938.0,
      "step": 5195
    },
    {
      "entropy": 5.937219858169556,
      "epoch": 0.5840399842758466,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.796,
      "mean_token_accuracy": 0.17016382068395614,
      "num_tokens": 11610077.0,
      "step": 5200
    },
    {
      "entropy": 5.715516757965088,
      "epoch": 0.5846015611838041,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.6748,
      "mean_token_accuracy": 0.1713270962238312,
      "num_tokens": 11622079.0,
      "step": 5205
    },
    {
      "entropy": 5.604576063156128,
      "epoch": 0.5851631380917617,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17965344488620758,
      "num_tokens": 11633047.0,
      "step": 5210
    },
    {
      "entropy": 5.649144887924194,
      "epoch": 0.5857247149997192,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.5514,
      "mean_token_accuracy": 0.17715423107147216,
      "num_tokens": 11643858.0,
      "step": 5215
    },
    {
      "entropy": 5.6603387832641605,
      "epoch": 0.5862862919076768,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.5828,
      "mean_token_accuracy": 0.1777600809931755,
      "num_tokens": 11654550.0,
      "step": 5220
    },
    {
      "entropy": 5.6631739139556885,
      "epoch": 0.5868478688156343,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497448379305403,
      "loss": 5.6081,
      "mean_token_accuracy": 0.17178307473659515,
      "num_tokens": 11665787.0,
      "step": 5225
    },
    {
      "entropy": 5.768941593170166,
      "epoch": 0.5874094457235919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.7174,
      "mean_token_accuracy": 0.16885657459497452,
      "num_tokens": 11677717.0,
      "step": 5230
    },
    {
      "entropy": 5.720490026473999,
      "epoch": 0.5879710226315494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497436306489741,
      "loss": 5.6263,
      "mean_token_accuracy": 0.17350273728370666,
      "num_tokens": 11687714.0,
      "step": 5235
    },
    {
      "entropy": 5.643569993972778,
      "epoch": 0.588532599539507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18226350098848343,
      "num_tokens": 11698777.0,
      "step": 5240
    },
    {
      "entropy": 5.668469762802124,
      "epoch": 0.5890941764474645,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.607,
      "mean_token_accuracy": 0.1782697170972824,
      "num_tokens": 11710922.0,
      "step": 5245
    },
    {
      "entropy": 5.705522632598877,
      "epoch": 0.589655753355422,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.6433,
      "mean_token_accuracy": 0.17290348261594773,
      "num_tokens": 11721450.0,
      "step": 5250
    },
    {
      "entropy": 5.651781797409058,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.5491,
      "mean_token_accuracy": 0.1756184220314026,
      "num_tokens": 11732458.0,
      "step": 5255
    },
    {
      "entropy": 5.618138551712036,
      "epoch": 0.5907789071713371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.5286,
      "mean_token_accuracy": 0.18139473646879195,
      "num_tokens": 11742649.0,
      "step": 5260
    },
    {
      "entropy": 5.63271131515503,
      "epoch": 0.5913404840792946,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497399918069926,
      "loss": 5.5522,
      "mean_token_accuracy": 0.18027863204479216,
      "num_tokens": 11753811.0,
      "step": 5265
    },
    {
      "entropy": 5.733545780181885,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.6592,
      "mean_token_accuracy": 0.16973895877599715,
      "num_tokens": 11763965.0,
      "step": 5270
    },
    {
      "entropy": 5.8242546081542965,
      "epoch": 0.5924636378952097,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.6572,
      "mean_token_accuracy": 0.18044903129339218,
      "num_tokens": 11774323.0,
      "step": 5275
    },
    {
      "entropy": 5.680837869644165,
      "epoch": 0.5930252148031673,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.6447,
      "mean_token_accuracy": 0.1720179185271263,
      "num_tokens": 11786157.0,
      "step": 5280
    },
    {
      "entropy": 5.722670650482177,
      "epoch": 0.5935867917111248,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.6275,
      "mean_token_accuracy": 0.17616336196660995,
      "num_tokens": 11797344.0,
      "step": 5285
    },
    {
      "entropy": 5.539758396148682,
      "epoch": 0.5941483686190824,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.5538,
      "mean_token_accuracy": 0.17561053335666657,
      "num_tokens": 11808149.0,
      "step": 5290
    },
    {
      "entropy": 5.658942747116089,
      "epoch": 0.5947099455270399,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.5717,
      "mean_token_accuracy": 0.17923365384340287,
      "num_tokens": 11820203.0,
      "step": 5295
    },
    {
      "entropy": 5.755408239364624,
      "epoch": 0.5952715224349975,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17345385253429413,
      "num_tokens": 11830172.0,
      "step": 5300
    },
    {
      "entropy": 5.731786251068115,
      "epoch": 0.595833099342955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.6034,
      "mean_token_accuracy": 0.18058293759822847,
      "num_tokens": 11840203.0,
      "step": 5305
    },
    {
      "entropy": 5.706607627868652,
      "epoch": 0.5963946762509126,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.5648,
      "mean_token_accuracy": 0.17356404513120652,
      "num_tokens": 11851002.0,
      "step": 5310
    },
    {
      "entropy": 5.624784660339356,
      "epoch": 0.5969562531588701,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.5772,
      "mean_token_accuracy": 0.17031607925891876,
      "num_tokens": 11861559.0,
      "step": 5315
    },
    {
      "entropy": 5.6799314498901365,
      "epoch": 0.5975178300668277,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.6395,
      "mean_token_accuracy": 0.17312609702348708,
      "num_tokens": 11872535.0,
      "step": 5320
    },
    {
      "entropy": 5.7497875690460205,
      "epoch": 0.5980794069747852,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.659,
      "mean_token_accuracy": 0.1717180833220482,
      "num_tokens": 11884308.0,
      "step": 5325
    },
    {
      "entropy": 5.81054277420044,
      "epoch": 0.5986409838827428,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.7253,
      "mean_token_accuracy": 0.17106300741434097,
      "num_tokens": 11896136.0,
      "step": 5330
    },
    {
      "entropy": 5.704445934295654,
      "epoch": 0.5992025607907003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497314020447805,
      "loss": 5.555,
      "mean_token_accuracy": 0.18488236963748933,
      "num_tokens": 11906497.0,
      "step": 5335
    },
    {
      "entropy": 5.71569709777832,
      "epoch": 0.5997641376986579,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.6882,
      "mean_token_accuracy": 0.16954067945480347,
      "num_tokens": 11917872.0,
      "step": 5340
    },
    {
      "entropy": 5.770382022857666,
      "epoch": 0.6003257146066153,
      "grad_norm": 1.0,
      "learning_rate": 0.000497301636089344,
      "loss": 5.7133,
      "mean_token_accuracy": 0.16962868273258208,
      "num_tokens": 11929727.0,
      "step": 5345
    },
    {
      "entropy": 5.726130390167237,
      "epoch": 0.6008872915145729,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17450084090232848,
      "num_tokens": 11939996.0,
      "step": 5350
    },
    {
      "entropy": 5.598533058166504,
      "epoch": 0.6014488684225304,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.5753,
      "mean_token_accuracy": 0.18218182176351547,
      "num_tokens": 11950683.0,
      "step": 5355
    },
    {
      "entropy": 5.614047431945801,
      "epoch": 0.602010445330488,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.5811,
      "mean_token_accuracy": 0.17111939787864686,
      "num_tokens": 11962447.0,
      "step": 5360
    },
    {
      "entropy": 5.748089361190796,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.6663,
      "mean_token_accuracy": 0.17336540073156356,
      "num_tokens": 11972740.0,
      "step": 5365
    },
    {
      "entropy": 5.6867390155792235,
      "epoch": 0.6031335991464031,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.5378,
      "mean_token_accuracy": 0.17462631613016127,
      "num_tokens": 11983292.0,
      "step": 5370
    },
    {
      "entropy": 5.741811800003052,
      "epoch": 0.6036951760543606,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.7533,
      "mean_token_accuracy": 0.16546933352947235,
      "num_tokens": 11996384.0,
      "step": 5375
    },
    {
      "entropy": 5.790655422210693,
      "epoch": 0.6042567529623182,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1804477110505104,
      "num_tokens": 12006938.0,
      "step": 5380
    },
    {
      "entropy": 5.586878156661987,
      "epoch": 0.6048183298702757,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.4956,
      "mean_token_accuracy": 0.17651916891336442,
      "num_tokens": 12018729.0,
      "step": 5385
    },
    {
      "entropy": 5.913886404037475,
      "epoch": 0.6053799067782333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.8358,
      "mean_token_accuracy": 0.16637750566005707,
      "num_tokens": 12029625.0,
      "step": 5390
    },
    {
      "entropy": 5.72364821434021,
      "epoch": 0.6059414836861908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.5423,
      "mean_token_accuracy": 0.1724603295326233,
      "num_tokens": 12040243.0,
      "step": 5395
    },
    {
      "entropy": 5.774748659133911,
      "epoch": 0.6065030605941484,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.8269,
      "mean_token_accuracy": 0.16337401419878006,
      "num_tokens": 12051961.0,
      "step": 5400
    },
    {
      "entropy": 5.790932512283325,
      "epoch": 0.6070646375021059,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.6903,
      "mean_token_accuracy": 0.1754745662212372,
      "num_tokens": 12063969.0,
      "step": 5405
    },
    {
      "entropy": 5.711603879928589,
      "epoch": 0.6076262144100635,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.5856,
      "mean_token_accuracy": 0.17822082340717316,
      "num_tokens": 12075682.0,
      "step": 5410
    },
    {
      "entropy": 5.753958749771118,
      "epoch": 0.608187791318021,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.7246,
      "mean_token_accuracy": 0.168320532143116,
      "num_tokens": 12086595.0,
      "step": 5415
    },
    {
      "entropy": 5.753818988800049,
      "epoch": 0.6087493682259786,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.5994,
      "mean_token_accuracy": 0.17213478088378906,
      "num_tokens": 12099421.0,
      "step": 5420
    },
    {
      "entropy": 5.753985404968262,
      "epoch": 0.6093109451339361,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.706,
      "mean_token_accuracy": 0.16944072097539903,
      "num_tokens": 12109680.0,
      "step": 5425
    },
    {
      "entropy": 5.648378658294678,
      "epoch": 0.6098725220418937,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.5771,
      "mean_token_accuracy": 0.18387772440910338,
      "num_tokens": 12121196.0,
      "step": 5430
    },
    {
      "entropy": 5.64174075126648,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497188902983013,
      "loss": 5.5387,
      "mean_token_accuracy": 0.16641906052827835,
      "num_tokens": 12132262.0,
      "step": 5435
    },
    {
      "entropy": 5.77960057258606,
      "epoch": 0.6109956758578087,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.7221,
      "mean_token_accuracy": 0.1732988551259041,
      "num_tokens": 12143747.0,
      "step": 5440
    },
    {
      "entropy": 5.90015869140625,
      "epoch": 0.6115572527657662,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.7968,
      "mean_token_accuracy": 0.16648507714271546,
      "num_tokens": 12155887.0,
      "step": 5445
    },
    {
      "entropy": 5.683777236938477,
      "epoch": 0.6121188296737238,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.5861,
      "mean_token_accuracy": 0.17328004390001298,
      "num_tokens": 12166718.0,
      "step": 5450
    },
    {
      "entropy": 5.721788024902343,
      "epoch": 0.6126804065816813,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.7248,
      "mean_token_accuracy": 0.17312407493591309,
      "num_tokens": 12178540.0,
      "step": 5455
    },
    {
      "entropy": 5.88222484588623,
      "epoch": 0.6132419834896389,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.7589,
      "mean_token_accuracy": 0.16493648439645767,
      "num_tokens": 12189820.0,
      "step": 5460
    },
    {
      "entropy": 5.733379125595093,
      "epoch": 0.6138035603975964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.683,
      "mean_token_accuracy": 0.1734658569097519,
      "num_tokens": 12201679.0,
      "step": 5465
    },
    {
      "entropy": 5.737128114700317,
      "epoch": 0.614365137305554,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.6765,
      "mean_token_accuracy": 0.16735841035842897,
      "num_tokens": 12213793.0,
      "step": 5470
    },
    {
      "entropy": 5.807603883743286,
      "epoch": 0.6149267142135115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.7514,
      "mean_token_accuracy": 0.16945342123508453,
      "num_tokens": 12224519.0,
      "step": 5475
    },
    {
      "entropy": 5.8681245803833,
      "epoch": 0.6154882911214691,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.7608,
      "mean_token_accuracy": 0.1658295512199402,
      "num_tokens": 12236401.0,
      "step": 5480
    },
    {
      "entropy": 5.783669519424438,
      "epoch": 0.6160498680294266,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.7865,
      "mean_token_accuracy": 0.16470268964767457,
      "num_tokens": 12247670.0,
      "step": 5485
    },
    {
      "entropy": 5.696588134765625,
      "epoch": 0.6166114449373842,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.6372,
      "mean_token_accuracy": 0.17362091988325118,
      "num_tokens": 12258877.0,
      "step": 5490
    },
    {
      "entropy": 5.721828269958496,
      "epoch": 0.6171730218453417,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.6085,
      "mean_token_accuracy": 0.17151312828063964,
      "num_tokens": 12268507.0,
      "step": 5495
    },
    {
      "entropy": 5.705202722549439,
      "epoch": 0.6177345987532993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497106059033725,
      "loss": 5.606,
      "mean_token_accuracy": 0.17059515416622162,
      "num_tokens": 12280420.0,
      "step": 5500
    },
    {
      "entropy": 5.597872352600097,
      "epoch": 0.6182961756612568,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.5981,
      "mean_token_accuracy": 0.17261921167373656,
      "num_tokens": 12290876.0,
      "step": 5505
    },
    {
      "entropy": 5.676345777511597,
      "epoch": 0.6188577525692144,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.5022,
      "mean_token_accuracy": 0.18050920814275742,
      "num_tokens": 12301831.0,
      "step": 5510
    },
    {
      "entropy": 5.628951120376587,
      "epoch": 0.6194193294771719,
      "grad_norm": 1.0625,
      "learning_rate": 0.000497086771456353,
      "loss": 5.5275,
      "mean_token_accuracy": 0.17723134756088257,
      "num_tokens": 12312678.0,
      "step": 5515
    },
    {
      "entropy": 5.774626207351685,
      "epoch": 0.6199809063851295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.7205,
      "mean_token_accuracy": 0.16646523922681808,
      "num_tokens": 12322905.0,
      "step": 5520
    },
    {
      "entropy": 5.714275932312011,
      "epoch": 0.620542483293087,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.5894,
      "mean_token_accuracy": 0.1710187703371048,
      "num_tokens": 12333091.0,
      "step": 5525
    },
    {
      "entropy": 5.697361612319947,
      "epoch": 0.6211040602010446,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17744589745998382,
      "num_tokens": 12346310.0,
      "step": 5530
    },
    {
      "entropy": 5.6302886486053465,
      "epoch": 0.621665637109002,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.5137,
      "mean_token_accuracy": 0.17561311572790145,
      "num_tokens": 12357138.0,
      "step": 5535
    },
    {
      "entropy": 5.731592130661011,
      "epoch": 0.6222272140169596,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.6047,
      "mean_token_accuracy": 0.1761130392551422,
      "num_tokens": 12367382.0,
      "step": 5540
    },
    {
      "entropy": 5.6389649391174315,
      "epoch": 0.6227887909249171,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.5956,
      "mean_token_accuracy": 0.17054942697286607,
      "num_tokens": 12380357.0,
      "step": 5545
    },
    {
      "entropy": 5.804252243041992,
      "epoch": 0.6233503678328747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.7713,
      "mean_token_accuracy": 0.16149107813835145,
      "num_tokens": 12392187.0,
      "step": 5550
    },
    {
      "entropy": 5.722889518737793,
      "epoch": 0.6239119447408322,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.633,
      "mean_token_accuracy": 0.17469703853130342,
      "num_tokens": 12402924.0,
      "step": 5555
    },
    {
      "entropy": 5.734255170822143,
      "epoch": 0.6244735216487898,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497028526894682,
      "loss": 5.7521,
      "mean_token_accuracy": 0.16999632269144058,
      "num_tokens": 12413554.0,
      "step": 5560
    },
    {
      "entropy": 5.798130083084106,
      "epoch": 0.6250350985567473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.6273,
      "mean_token_accuracy": 0.1768548682332039,
      "num_tokens": 12424525.0,
      "step": 5565
    },
    {
      "entropy": 5.680708885192871,
      "epoch": 0.6255966754647049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.6322,
      "mean_token_accuracy": 0.17046258449554444,
      "num_tokens": 12436405.0,
      "step": 5570
    },
    {
      "entropy": 5.718565797805786,
      "epoch": 0.6261582523726624,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.7037,
      "mean_token_accuracy": 0.17012005001306535,
      "num_tokens": 12446909.0,
      "step": 5575
    },
    {
      "entropy": 5.737675189971924,
      "epoch": 0.62671982928062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.6804,
      "mean_token_accuracy": 0.1764724850654602,
      "num_tokens": 12458739.0,
      "step": 5580
    },
    {
      "entropy": 5.707290410995483,
      "epoch": 0.6272814061885775,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.6001,
      "mean_token_accuracy": 0.17520488798618317,
      "num_tokens": 12469529.0,
      "step": 5585
    },
    {
      "entropy": 5.761801242828369,
      "epoch": 0.6278429830965351,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.6483,
      "mean_token_accuracy": 0.16846525967121123,
      "num_tokens": 12480067.0,
      "step": 5590
    },
    {
      "entropy": 5.673601484298706,
      "epoch": 0.6284045600044926,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.5278,
      "mean_token_accuracy": 0.1837966412305832,
      "num_tokens": 12491043.0,
      "step": 5595
    },
    {
      "entropy": 5.561583518981934,
      "epoch": 0.6289661369124502,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.4627,
      "mean_token_accuracy": 0.1810373306274414,
      "num_tokens": 12501852.0,
      "step": 5600
    },
    {
      "entropy": 5.743855667114258,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.7447,
      "mean_token_accuracy": 0.1711232617497444,
      "num_tokens": 12512696.0,
      "step": 5605
    },
    {
      "entropy": 5.74007658958435,
      "epoch": 0.6300892907283653,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.6186,
      "mean_token_accuracy": 0.1756168007850647,
      "num_tokens": 12523643.0,
      "step": 5610
    },
    {
      "entropy": 5.71359224319458,
      "epoch": 0.6306508676363228,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.5883,
      "mean_token_accuracy": 0.1746621236205101,
      "num_tokens": 12534827.0,
      "step": 5615
    },
    {
      "entropy": 5.563238573074341,
      "epoch": 0.6312124445442804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.4719,
      "mean_token_accuracy": 0.17915562987327577,
      "num_tokens": 12545977.0,
      "step": 5620
    },
    {
      "entropy": 5.544950914382935,
      "epoch": 0.6317740214522379,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.5384,
      "mean_token_accuracy": 0.1715235486626625,
      "num_tokens": 12555890.0,
      "step": 5625
    },
    {
      "entropy": 5.698396301269531,
      "epoch": 0.6323355983601955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.6064,
      "mean_token_accuracy": 0.17803364992141724,
      "num_tokens": 12566412.0,
      "step": 5630
    },
    {
      "entropy": 5.7760552883148195,
      "epoch": 0.6328971752681529,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17108214646577835,
      "num_tokens": 12577175.0,
      "step": 5635
    },
    {
      "entropy": 5.595721292495727,
      "epoch": 0.6334587521761105,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496923567209739,
      "loss": 5.5375,
      "mean_token_accuracy": 0.17472511678934097,
      "num_tokens": 12588760.0,
      "step": 5640
    },
    {
      "entropy": 5.609034824371338,
      "epoch": 0.634020329084068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.6121,
      "mean_token_accuracy": 0.1745341196656227,
      "num_tokens": 12599899.0,
      "step": 5645
    },
    {
      "entropy": 5.744676113128662,
      "epoch": 0.6345819059920256,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.6017,
      "mean_token_accuracy": 0.1837252125144005,
      "num_tokens": 12610836.0,
      "step": 5650
    },
    {
      "entropy": 5.61579442024231,
      "epoch": 0.6351434828999831,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.5485,
      "mean_token_accuracy": 0.18136483132839204,
      "num_tokens": 12621855.0,
      "step": 5655
    },
    {
      "entropy": 5.69185848236084,
      "epoch": 0.6357050598079407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.6491,
      "mean_token_accuracy": 0.17270753532648087,
      "num_tokens": 12632964.0,
      "step": 5660
    },
    {
      "entropy": 5.634363222122192,
      "epoch": 0.6362666367158982,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496890396295883,
      "loss": 5.4726,
      "mean_token_accuracy": 0.1809499219059944,
      "num_tokens": 12643717.0,
      "step": 5665
    },
    {
      "entropy": 5.70470118522644,
      "epoch": 0.6368282136238558,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.6115,
      "mean_token_accuracy": 0.16996596604585648,
      "num_tokens": 12654491.0,
      "step": 5670
    },
    {
      "entropy": 5.6480659484863285,
      "epoch": 0.6373897905318133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.645,
      "mean_token_accuracy": 0.17329516261816025,
      "num_tokens": 12666128.0,
      "step": 5675
    },
    {
      "entropy": 5.748110628128051,
      "epoch": 0.6379513674397709,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.7332,
      "mean_token_accuracy": 0.16888405978679658,
      "num_tokens": 12677957.0,
      "step": 5680
    },
    {
      "entropy": 5.756485652923584,
      "epoch": 0.6385129443477284,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.5693,
      "mean_token_accuracy": 0.17527114748954772,
      "num_tokens": 12687964.0,
      "step": 5685
    },
    {
      "entropy": 5.590578317642212,
      "epoch": 0.639074521255686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.574,
      "mean_token_accuracy": 0.17056965082883835,
      "num_tokens": 12698805.0,
      "step": 5690
    },
    {
      "entropy": 5.7417315483093265,
      "epoch": 0.6396360981636435,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.584,
      "mean_token_accuracy": 0.17162402719259262,
      "num_tokens": 12709942.0,
      "step": 5695
    },
    {
      "entropy": 5.778721570968628,
      "epoch": 0.6401976750716011,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.6731,
      "mean_token_accuracy": 0.1761099860072136,
      "num_tokens": 12721856.0,
      "step": 5700
    },
    {
      "entropy": 5.556879568099975,
      "epoch": 0.6407592519795586,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.4929,
      "mean_token_accuracy": 0.17659999579191207,
      "num_tokens": 12732619.0,
      "step": 5705
    },
    {
      "entropy": 5.642232465744018,
      "epoch": 0.6413208288875162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.578,
      "mean_token_accuracy": 0.1764561727643013,
      "num_tokens": 12743478.0,
      "step": 5710
    },
    {
      "entropy": 5.622849702835083,
      "epoch": 0.6418824057954737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.5327,
      "mean_token_accuracy": 0.16981768161058425,
      "num_tokens": 12754761.0,
      "step": 5715
    },
    {
      "entropy": 5.762554883956909,
      "epoch": 0.6424439827034313,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17977579683065414,
      "num_tokens": 12765803.0,
      "step": 5720
    },
    {
      "entropy": 5.767837190628052,
      "epoch": 0.6430055596113888,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.6577,
      "mean_token_accuracy": 0.17164772003889084,
      "num_tokens": 12777882.0,
      "step": 5725
    },
    {
      "entropy": 5.702736282348633,
      "epoch": 0.6435671365193463,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.6754,
      "mean_token_accuracy": 0.17198235988616944,
      "num_tokens": 12788268.0,
      "step": 5730
    },
    {
      "entropy": 5.629304265975952,
      "epoch": 0.6441287134273038,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.6715,
      "mean_token_accuracy": 0.17235511243343354,
      "num_tokens": 12800506.0,
      "step": 5735
    },
    {
      "entropy": 5.7705975532531735,
      "epoch": 0.6446902903352614,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17663808166980743,
      "num_tokens": 12813040.0,
      "step": 5740
    },
    {
      "entropy": 5.780026578903199,
      "epoch": 0.6452518672432189,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.6714,
      "mean_token_accuracy": 0.1773954689502716,
      "num_tokens": 12825534.0,
      "step": 5745
    },
    {
      "entropy": 5.6190712451934814,
      "epoch": 0.6458134441511765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.5549,
      "mean_token_accuracy": 0.17475787103176116,
      "num_tokens": 12835882.0,
      "step": 5750
    },
    {
      "entropy": 5.759916305541992,
      "epoch": 0.646375021059134,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.7408,
      "mean_token_accuracy": 0.16839111149311065,
      "num_tokens": 12847663.0,
      "step": 5755
    },
    {
      "entropy": 5.961766767501831,
      "epoch": 0.6469365979670916,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.8788,
      "mean_token_accuracy": 0.16633762419223785,
      "num_tokens": 12860789.0,
      "step": 5760
    },
    {
      "entropy": 5.7413105964660645,
      "epoch": 0.6474981748750491,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.6103,
      "mean_token_accuracy": 0.17463897466659545,
      "num_tokens": 12872509.0,
      "step": 5765
    },
    {
      "entropy": 5.628509473800659,
      "epoch": 0.6480597517830067,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.6059,
      "mean_token_accuracy": 0.17851067781448365,
      "num_tokens": 12884321.0,
      "step": 5770
    },
    {
      "entropy": 5.554707098007202,
      "epoch": 0.6486213286909642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.5081,
      "mean_token_accuracy": 0.17908329516649246,
      "num_tokens": 12894559.0,
      "step": 5775
    },
    {
      "entropy": 5.679509878158569,
      "epoch": 0.6491829055989218,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.6027,
      "mean_token_accuracy": 0.17926470041275025,
      "num_tokens": 12904934.0,
      "step": 5780
    },
    {
      "entropy": 5.823178243637085,
      "epoch": 0.6497444825068793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496728717732762,
      "loss": 5.7257,
      "mean_token_accuracy": 0.1665002003312111,
      "num_tokens": 12916551.0,
      "step": 5785
    },
    {
      "entropy": 5.698605918884278,
      "epoch": 0.6503060594148369,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.6256,
      "mean_token_accuracy": 0.1717161476612091,
      "num_tokens": 12928097.0,
      "step": 5790
    },
    {
      "entropy": 5.642832279205322,
      "epoch": 0.6508676363227944,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.4937,
      "mean_token_accuracy": 0.18256949335336686,
      "num_tokens": 12939161.0,
      "step": 5795
    },
    {
      "entropy": 5.546928739547729,
      "epoch": 0.651429213230752,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.5389,
      "mean_token_accuracy": 0.1791675016283989,
      "num_tokens": 12950862.0,
      "step": 5800
    },
    {
      "entropy": 5.664335155487061,
      "epoch": 0.6519907901387095,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.5791,
      "mean_token_accuracy": 0.17750437557697296,
      "num_tokens": 12962333.0,
      "step": 5805
    },
    {
      "entropy": 5.673435211181641,
      "epoch": 0.6525523670466671,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.5719,
      "mean_token_accuracy": 0.17666879445314407,
      "num_tokens": 12973136.0,
      "step": 5810
    },
    {
      "entropy": 5.6725842475891115,
      "epoch": 0.6531139439546246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.5988,
      "mean_token_accuracy": 0.17476778477430344,
      "num_tokens": 12983602.0,
      "step": 5815
    },
    {
      "entropy": 5.700492525100708,
      "epoch": 0.6536755208625822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.5205,
      "mean_token_accuracy": 0.18146025538444518,
      "num_tokens": 12994129.0,
      "step": 5820
    },
    {
      "entropy": 5.636948108673096,
      "epoch": 0.6542370977705396,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17263583242893218,
      "num_tokens": 13005071.0,
      "step": 5825
    },
    {
      "entropy": 5.795610237121582,
      "epoch": 0.6547986746784972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.7477,
      "mean_token_accuracy": 0.172401587665081,
      "num_tokens": 13016202.0,
      "step": 5830
    },
    {
      "entropy": 5.8175863265991214,
      "epoch": 0.6553602515864547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.7552,
      "mean_token_accuracy": 0.16232562959194183,
      "num_tokens": 13028468.0,
      "step": 5835
    },
    {
      "entropy": 5.709358787536621,
      "epoch": 0.6559218284944123,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496653255842169,
      "loss": 5.6296,
      "mean_token_accuracy": 0.1753317266702652,
      "num_tokens": 13039667.0,
      "step": 5840
    },
    {
      "entropy": 5.678438663482666,
      "epoch": 0.6564834054023698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.6781,
      "mean_token_accuracy": 0.16587844491004944,
      "num_tokens": 13051583.0,
      "step": 5845
    },
    {
      "entropy": 5.732926893234253,
      "epoch": 0.6570449823103274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.6319,
      "mean_token_accuracy": 0.17361444979906082,
      "num_tokens": 13062401.0,
      "step": 5850
    },
    {
      "entropy": 5.604018449783325,
      "epoch": 0.6576065592182849,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.524,
      "mean_token_accuracy": 0.1782347083091736,
      "num_tokens": 13074219.0,
      "step": 5855
    },
    {
      "entropy": 5.75310869216919,
      "epoch": 0.6581681361262425,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.7095,
      "mean_token_accuracy": 0.17001335322856903,
      "num_tokens": 13086073.0,
      "step": 5860
    },
    {
      "entropy": 5.7198251247406,
      "epoch": 0.6587297130342,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.6264,
      "mean_token_accuracy": 0.1701858565211296,
      "num_tokens": 13096340.0,
      "step": 5865
    },
    {
      "entropy": 5.67186975479126,
      "epoch": 0.6592912899421576,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.5863,
      "mean_token_accuracy": 0.17610248178243637,
      "num_tokens": 13107273.0,
      "step": 5870
    },
    {
      "entropy": 5.664875936508179,
      "epoch": 0.6598528668501151,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.5649,
      "mean_token_accuracy": 0.1737078085541725,
      "num_tokens": 13117388.0,
      "step": 5875
    },
    {
      "entropy": 5.7048064231872555,
      "epoch": 0.6604144437580727,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496597838078686,
      "loss": 5.6793,
      "mean_token_accuracy": 0.1677414134144783,
      "num_tokens": 13128720.0,
      "step": 5880
    },
    {
      "entropy": 5.75440559387207,
      "epoch": 0.6609760206660302,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496590879105005,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17483343333005905,
      "num_tokens": 13139405.0,
      "step": 5885
    },
    {
      "entropy": 5.713056468963623,
      "epoch": 0.6615375975739878,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.6805,
      "mean_token_accuracy": 0.1753469303250313,
      "num_tokens": 13151408.0,
      "step": 5890
    },
    {
      "entropy": 5.69801516532898,
      "epoch": 0.6620991744819453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.6001,
      "mean_token_accuracy": 0.1750464230775833,
      "num_tokens": 13162592.0,
      "step": 5895
    },
    {
      "entropy": 5.758023262023926,
      "epoch": 0.6626607513899029,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.6202,
      "mean_token_accuracy": 0.1782073572278023,
      "num_tokens": 13173040.0,
      "step": 5900
    },
    {
      "entropy": 5.64671893119812,
      "epoch": 0.6632223282978604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.69,
      "mean_token_accuracy": 0.1743127301335335,
      "num_tokens": 13185475.0,
      "step": 5905
    },
    {
      "entropy": 5.667902898788452,
      "epoch": 0.663783905205818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.5685,
      "mean_token_accuracy": 0.17888267785310746,
      "num_tokens": 13196481.0,
      "step": 5910
    },
    {
      "entropy": 5.656558847427368,
      "epoch": 0.6643454821137755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.4688,
      "mean_token_accuracy": 0.18201548606157303,
      "num_tokens": 13207947.0,
      "step": 5915
    },
    {
      "entropy": 5.50324239730835,
      "epoch": 0.664907059021733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.5078,
      "mean_token_accuracy": 0.18712942004203797,
      "num_tokens": 13219392.0,
      "step": 5920
    },
    {
      "entropy": 5.573181676864624,
      "epoch": 0.6654686359296905,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.5411,
      "mean_token_accuracy": 0.18074103444814682,
      "num_tokens": 13231556.0,
      "step": 5925
    },
    {
      "entropy": 5.654805421829224,
      "epoch": 0.6660302128376481,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.4946,
      "mean_token_accuracy": 0.18450129479169847,
      "num_tokens": 13241287.0,
      "step": 5930
    },
    {
      "entropy": 5.568438386917114,
      "epoch": 0.6665917897456056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.5469,
      "mean_token_accuracy": 0.16648050397634506,
      "num_tokens": 13254120.0,
      "step": 5935
    },
    {
      "entropy": 5.712764930725098,
      "epoch": 0.6671533666535632,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.6151,
      "mean_token_accuracy": 0.17329633831977845,
      "num_tokens": 13265745.0,
      "step": 5940
    },
    {
      "entropy": 5.793698453903199,
      "epoch": 0.6677149435615207,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496506821152345,
      "loss": 5.6867,
      "mean_token_accuracy": 0.17168214321136474,
      "num_tokens": 13277791.0,
      "step": 5945
    },
    {
      "entropy": 5.608066082000732,
      "epoch": 0.6682765204694783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.5327,
      "mean_token_accuracy": 0.17143328189849855,
      "num_tokens": 13288078.0,
      "step": 5950
    },
    {
      "entropy": 5.670023727416992,
      "epoch": 0.6688380973774358,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496492712743046,
      "loss": 5.5634,
      "mean_token_accuracy": 0.18261950612068176,
      "num_tokens": 13300547.0,
      "step": 5955
    },
    {
      "entropy": 5.675726366043091,
      "epoch": 0.6693996742853934,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.6515,
      "mean_token_accuracy": 0.17231980115175247,
      "num_tokens": 13310339.0,
      "step": 5960
    },
    {
      "entropy": 5.7539581775665285,
      "epoch": 0.6699612511933509,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.6133,
      "mean_token_accuracy": 0.1755753353238106,
      "num_tokens": 13321294.0,
      "step": 5965
    },
    {
      "entropy": 5.729451990127563,
      "epoch": 0.6705228281013085,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.646,
      "mean_token_accuracy": 0.174198018014431,
      "num_tokens": 13332608.0,
      "step": 5970
    },
    {
      "entropy": 5.684158277511597,
      "epoch": 0.671084405009266,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.585,
      "mean_token_accuracy": 0.18588267415761947,
      "num_tokens": 13344169.0,
      "step": 5975
    },
    {
      "entropy": 5.595675802230835,
      "epoch": 0.6716459819172236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.5457,
      "mean_token_accuracy": 0.17660977244377135,
      "num_tokens": 13355713.0,
      "step": 5980
    },
    {
      "entropy": 5.7029187202453615,
      "epoch": 0.6722075588251811,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.6341,
      "mean_token_accuracy": 0.17223501354455947,
      "num_tokens": 13366644.0,
      "step": 5985
    },
    {
      "entropy": 5.672948265075684,
      "epoch": 0.6727691357331387,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.6001,
      "mean_token_accuracy": 0.17075798511505128,
      "num_tokens": 13378125.0,
      "step": 5990
    },
    {
      "entropy": 5.659946870803833,
      "epoch": 0.6733307126410962,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.6462,
      "mean_token_accuracy": 0.17172100096940995,
      "num_tokens": 13389681.0,
      "step": 5995
    },
    {
      "entropy": 5.716019535064698,
      "epoch": 0.6738922895490538,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.6374,
      "mean_token_accuracy": 0.1769365757703781,
      "num_tokens": 13399814.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.609915955582629,
      "eval_loss": 5.632114887237549,
      "eval_mean_token_accuracy": 0.18043994106619451,
      "eval_num_tokens": 13399814.0,
      "eval_runtime": 59.1053,
      "eval_samples_per_second": 524.707,
      "eval_steps_per_second": 65.595,
      "step": 6000
    },
    {
      "entropy": 5.779499101638794,
      "epoch": 0.6744538664570113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.6522,
      "mean_token_accuracy": 0.17724189162254333,
      "num_tokens": 13410614.0,
      "step": 6005
    },
    {
      "entropy": 5.697342586517334,
      "epoch": 0.6750154433649689,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496414612292672,
      "loss": 5.6599,
      "mean_token_accuracy": 0.1712150201201439,
      "num_tokens": 13421392.0,
      "step": 6010
    },
    {
      "entropy": 5.715792560577393,
      "epoch": 0.6755770202729264,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.6966,
      "mean_token_accuracy": 0.17103470414876937,
      "num_tokens": 13432004.0,
      "step": 6015
    },
    {
      "entropy": 5.720300006866455,
      "epoch": 0.6761385971808839,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.5345,
      "mean_token_accuracy": 0.18039939254522325,
      "num_tokens": 13442450.0,
      "step": 6020
    },
    {
      "entropy": 5.674740743637085,
      "epoch": 0.6767001740888414,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.6041,
      "mean_token_accuracy": 0.17143985629081726,
      "num_tokens": 13453848.0,
      "step": 6025
    },
    {
      "entropy": 5.611522245407104,
      "epoch": 0.677261750996799,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.5994,
      "mean_token_accuracy": 0.17509289979934692,
      "num_tokens": 13465119.0,
      "step": 6030
    },
    {
      "entropy": 5.611239814758301,
      "epoch": 0.6778233279047565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.582,
      "mean_token_accuracy": 0.17205574214458466,
      "num_tokens": 13475937.0,
      "step": 6035
    },
    {
      "entropy": 5.642318058013916,
      "epoch": 0.6783849048127141,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.5671,
      "mean_token_accuracy": 0.17940536588430406,
      "num_tokens": 13487430.0,
      "step": 6040
    },
    {
      "entropy": 5.672739744186401,
      "epoch": 0.6789464817206716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.5517,
      "mean_token_accuracy": 0.17838966995477676,
      "num_tokens": 13499631.0,
      "step": 6045
    },
    {
      "entropy": 5.693057012557984,
      "epoch": 0.6795080586286292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496357276153527,
      "loss": 5.629,
      "mean_token_accuracy": 0.1752404034137726,
      "num_tokens": 13511753.0,
      "step": 6050
    },
    {
      "entropy": 5.735835170745849,
      "epoch": 0.6800696355365867,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.6887,
      "mean_token_accuracy": 0.16904593259096146,
      "num_tokens": 13524228.0,
      "step": 6055
    },
    {
      "entropy": 5.823309230804443,
      "epoch": 0.6806312124445443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496342871633153,
      "loss": 5.7107,
      "mean_token_accuracy": 0.1734006255865097,
      "num_tokens": 13535700.0,
      "step": 6060
    },
    {
      "entropy": 5.719172239303589,
      "epoch": 0.6811927893525018,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.6879,
      "mean_token_accuracy": 0.1680354133248329,
      "num_tokens": 13547337.0,
      "step": 6065
    },
    {
      "entropy": 5.625253963470459,
      "epoch": 0.6817543662604594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.5363,
      "mean_token_accuracy": 0.17368890196084977,
      "num_tokens": 13558350.0,
      "step": 6070
    },
    {
      "entropy": 5.730729627609253,
      "epoch": 0.6823159431684169,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.5879,
      "mean_token_accuracy": 0.1806545540690422,
      "num_tokens": 13569230.0,
      "step": 6075
    },
    {
      "entropy": 5.726511764526367,
      "epoch": 0.6828775200763745,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.6743,
      "mean_token_accuracy": 0.17548759877681733,
      "num_tokens": 13580502.0,
      "step": 6080
    },
    {
      "entropy": 5.736097192764282,
      "epoch": 0.683439096984332,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.5845,
      "mean_token_accuracy": 0.17914303243160248,
      "num_tokens": 13591524.0,
      "step": 6085
    },
    {
      "entropy": 5.7245134830474855,
      "epoch": 0.6840006738922896,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.5997,
      "mean_token_accuracy": 0.1826063796877861,
      "num_tokens": 13602432.0,
      "step": 6090
    },
    {
      "entropy": 5.560599994659424,
      "epoch": 0.6845622508002471,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.5253,
      "mean_token_accuracy": 0.18265436589717865,
      "num_tokens": 13613052.0,
      "step": 6095
    },
    {
      "entropy": 5.642383337020874,
      "epoch": 0.6851238277082047,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.5741,
      "mean_token_accuracy": 0.17706613391637802,
      "num_tokens": 13623961.0,
      "step": 6100
    },
    {
      "entropy": 5.657310056686401,
      "epoch": 0.6856854046161622,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.556,
      "mean_token_accuracy": 0.1735036551952362,
      "num_tokens": 13635285.0,
      "step": 6105
    },
    {
      "entropy": 5.779331064224243,
      "epoch": 0.6862469815241198,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.7065,
      "mean_token_accuracy": 0.16852724254131318,
      "num_tokens": 13647720.0,
      "step": 6110
    },
    {
      "entropy": 5.626996564865112,
      "epoch": 0.6868085584320772,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.5526,
      "mean_token_accuracy": 0.17468138933181762,
      "num_tokens": 13659726.0,
      "step": 6115
    },
    {
      "entropy": 5.763138675689698,
      "epoch": 0.6873701353400348,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496255852573374,
      "loss": 5.7382,
      "mean_token_accuracy": 0.16682715564966202,
      "num_tokens": 13670332.0,
      "step": 6120
    },
    {
      "entropy": 5.783021879196167,
      "epoch": 0.6879317122479923,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.676,
      "mean_token_accuracy": 0.16434723436832427,
      "num_tokens": 13682419.0,
      "step": 6125
    },
    {
      "entropy": 5.578999662399292,
      "epoch": 0.6884932891559499,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.434,
      "mean_token_accuracy": 0.18486667722463607,
      "num_tokens": 13692698.0,
      "step": 6130
    },
    {
      "entropy": 5.5511950016021725,
      "epoch": 0.6890548660639074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.6314,
      "mean_token_accuracy": 0.17299454361200334,
      "num_tokens": 13704791.0,
      "step": 6135
    },
    {
      "entropy": 5.669870662689209,
      "epoch": 0.689616442971865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496226620764454,
      "loss": 5.5595,
      "mean_token_accuracy": 0.18140525370836258,
      "num_tokens": 13715860.0,
      "step": 6140
    },
    {
      "entropy": 5.706946134567261,
      "epoch": 0.6901780198798225,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496219295201788,
      "loss": 5.5431,
      "mean_token_accuracy": 0.18075013756752015,
      "num_tokens": 13727029.0,
      "step": 6145
    },
    {
      "entropy": 5.627535486221314,
      "epoch": 0.6907395967877801,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.5307,
      "mean_token_accuracy": 0.182185660302639,
      "num_tokens": 13737168.0,
      "step": 6150
    },
    {
      "entropy": 5.73567795753479,
      "epoch": 0.6913011736957376,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.6696,
      "mean_token_accuracy": 0.16650990694761275,
      "num_tokens": 13749828.0,
      "step": 6155
    },
    {
      "entropy": 5.654178047180176,
      "epoch": 0.6918627506036952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.4991,
      "mean_token_accuracy": 0.17278295755386353,
      "num_tokens": 13761617.0,
      "step": 6160
    },
    {
      "entropy": 5.6681290626525875,
      "epoch": 0.6924243275116527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496189922516379,
      "loss": 5.5524,
      "mean_token_accuracy": 0.18426286280155182,
      "num_tokens": 13772721.0,
      "step": 6165
    },
    {
      "entropy": 5.575658178329467,
      "epoch": 0.6929859044196103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.5412,
      "mean_token_accuracy": 0.1766442507505417,
      "num_tokens": 13783215.0,
      "step": 6170
    },
    {
      "entropy": 5.564092636108398,
      "epoch": 0.6935474813275678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.4812,
      "mean_token_accuracy": 0.18373349010944368,
      "num_tokens": 13793496.0,
      "step": 6175
    },
    {
      "entropy": 5.509256601333618,
      "epoch": 0.6941090582355254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496167819052389,
      "loss": 5.4176,
      "mean_token_accuracy": 0.17860936522483825,
      "num_tokens": 13802818.0,
      "step": 6180
    },
    {
      "entropy": 5.770023536682129,
      "epoch": 0.6946706351434829,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.7072,
      "mean_token_accuracy": 0.1726786971092224,
      "num_tokens": 13814449.0,
      "step": 6185
    },
    {
      "entropy": 5.710700941085816,
      "epoch": 0.6952322120514405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.6321,
      "mean_token_accuracy": 0.17531704306602477,
      "num_tokens": 13825369.0,
      "step": 6190
    },
    {
      "entropy": 5.586572647094727,
      "epoch": 0.695793788959398,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.4985,
      "mean_token_accuracy": 0.17635693699121474,
      "num_tokens": 13837992.0,
      "step": 6195
    },
    {
      "entropy": 5.714752912521362,
      "epoch": 0.6963553658673556,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.6089,
      "mean_token_accuracy": 0.17184996008872985,
      "num_tokens": 13849771.0,
      "step": 6200
    },
    {
      "entropy": 5.652769136428833,
      "epoch": 0.6969169427753131,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.587,
      "mean_token_accuracy": 0.17513276189565657,
      "num_tokens": 13861997.0,
      "step": 6205
    },
    {
      "entropy": 5.711466979980469,
      "epoch": 0.6974785196832706,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.6714,
      "mean_token_accuracy": 0.17424398809671401,
      "num_tokens": 13872972.0,
      "step": 6210
    },
    {
      "entropy": 5.5919266700744625,
      "epoch": 0.6980400965912281,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.4473,
      "mean_token_accuracy": 0.18001363873481752,
      "num_tokens": 13883139.0,
      "step": 6215
    },
    {
      "entropy": 5.670935344696045,
      "epoch": 0.6986016734991857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.6201,
      "mean_token_accuracy": 0.1752718135714531,
      "num_tokens": 13894326.0,
      "step": 6220
    },
    {
      "entropy": 5.690594577789307,
      "epoch": 0.6991632504071432,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.6457,
      "mean_token_accuracy": 0.1730250060558319,
      "num_tokens": 13905525.0,
      "step": 6225
    },
    {
      "entropy": 5.682399272918701,
      "epoch": 0.6997248273151008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.5583,
      "mean_token_accuracy": 0.17501617968082428,
      "num_tokens": 13915973.0,
      "step": 6230
    },
    {
      "entropy": 5.584595727920532,
      "epoch": 0.7002864042230583,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.3906,
      "mean_token_accuracy": 0.1863795131444931,
      "num_tokens": 13926380.0,
      "step": 6235
    },
    {
      "entropy": 5.582453870773316,
      "epoch": 0.7008479811310159,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.5793,
      "mean_token_accuracy": 0.17046487629413604,
      "num_tokens": 13938523.0,
      "step": 6240
    },
    {
      "entropy": 5.564077138900757,
      "epoch": 0.7014095580389734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.5593,
      "mean_token_accuracy": 0.1796951949596405,
      "num_tokens": 13949343.0,
      "step": 6245
    },
    {
      "entropy": 5.766108608245849,
      "epoch": 0.701971134946931,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.6802,
      "mean_token_accuracy": 0.17632725834846497,
      "num_tokens": 13960056.0,
      "step": 6250
    },
    {
      "entropy": 5.794866371154785,
      "epoch": 0.7025327118548885,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.6507,
      "mean_token_accuracy": 0.1752863645553589,
      "num_tokens": 13971932.0,
      "step": 6255
    },
    {
      "entropy": 5.695323705673218,
      "epoch": 0.7030942887628461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.648,
      "mean_token_accuracy": 0.17090354561805726,
      "num_tokens": 13981992.0,
      "step": 6260
    },
    {
      "entropy": 5.699549341201783,
      "epoch": 0.7036558656708036,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.5927,
      "mean_token_accuracy": 0.18272750079631805,
      "num_tokens": 13992215.0,
      "step": 6265
    },
    {
      "entropy": 5.66727876663208,
      "epoch": 0.7042174425787612,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.4959,
      "mean_token_accuracy": 0.18945090025663375,
      "num_tokens": 14002689.0,
      "step": 6270
    },
    {
      "entropy": 5.585068273544311,
      "epoch": 0.7047790194867187,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.615,
      "mean_token_accuracy": 0.17799130380153655,
      "num_tokens": 14014951.0,
      "step": 6275
    },
    {
      "entropy": 5.652612018585205,
      "epoch": 0.7053405963946763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.5438,
      "mean_token_accuracy": 0.18065126091241837,
      "num_tokens": 14025502.0,
      "step": 6280
    },
    {
      "entropy": 5.7588221549987795,
      "epoch": 0.7059021733026338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.5959,
      "mean_token_accuracy": 0.18291736096143724,
      "num_tokens": 14036576.0,
      "step": 6285
    },
    {
      "entropy": 5.642173147201538,
      "epoch": 0.7064637502105914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.6226,
      "mean_token_accuracy": 0.17990546971559523,
      "num_tokens": 14046493.0,
      "step": 6290
    },
    {
      "entropy": 5.751773023605347,
      "epoch": 0.7070253271185489,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.6775,
      "mean_token_accuracy": 0.18058259636163712,
      "num_tokens": 14057297.0,
      "step": 6295
    },
    {
      "entropy": 5.614803981781006,
      "epoch": 0.7075869040265065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.4618,
      "mean_token_accuracy": 0.18240975141525267,
      "num_tokens": 14068947.0,
      "step": 6300
    },
    {
      "entropy": 5.579032039642334,
      "epoch": 0.7081484809344639,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1755152612924576,
      "num_tokens": 14078979.0,
      "step": 6305
    },
    {
      "entropy": 5.620485639572143,
      "epoch": 0.7087100578424215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.4855,
      "mean_token_accuracy": 0.18234243988990784,
      "num_tokens": 14089785.0,
      "step": 6310
    },
    {
      "entropy": 5.627900648117065,
      "epoch": 0.709271634750379,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.5878,
      "mean_token_accuracy": 0.17533970773220062,
      "num_tokens": 14099921.0,
      "step": 6315
    },
    {
      "entropy": 5.62828893661499,
      "epoch": 0.7098332116583366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.4869,
      "mean_token_accuracy": 0.17739137262105942,
      "num_tokens": 14110965.0,
      "step": 6320
    },
    {
      "entropy": 5.638380813598633,
      "epoch": 0.7103947885662941,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.63,
      "mean_token_accuracy": 0.17418512403964997,
      "num_tokens": 14122409.0,
      "step": 6325
    },
    {
      "entropy": 5.67718539237976,
      "epoch": 0.7109563654742517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.5665,
      "mean_token_accuracy": 0.1799087792634964,
      "num_tokens": 14134017.0,
      "step": 6330
    },
    {
      "entropy": 5.709336662292481,
      "epoch": 0.7115179423822092,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.6862,
      "mean_token_accuracy": 0.1742307111620903,
      "num_tokens": 14144843.0,
      "step": 6335
    },
    {
      "entropy": 5.6433673858642575,
      "epoch": 0.7120795192901668,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.5399,
      "mean_token_accuracy": 0.1767047733068466,
      "num_tokens": 14155944.0,
      "step": 6340
    },
    {
      "entropy": 5.630907773971558,
      "epoch": 0.7126410961981243,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.5973,
      "mean_token_accuracy": 0.17689424902200698,
      "num_tokens": 14167064.0,
      "step": 6345
    },
    {
      "entropy": 5.731440496444702,
      "epoch": 0.7132026731060819,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.6192,
      "mean_token_accuracy": 0.16884241998195648,
      "num_tokens": 14178583.0,
      "step": 6350
    },
    {
      "entropy": 5.6261039733886715,
      "epoch": 0.7137642500140394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.5129,
      "mean_token_accuracy": 0.17680273950099945,
      "num_tokens": 14188752.0,
      "step": 6355
    },
    {
      "entropy": 5.608508205413818,
      "epoch": 0.714325826921997,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495897635781793,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18105533123016357,
      "num_tokens": 14199962.0,
      "step": 6360
    },
    {
      "entropy": 5.588458156585693,
      "epoch": 0.7148874038299545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.5275,
      "mean_token_accuracy": 0.17510385662317277,
      "num_tokens": 14209877.0,
      "step": 6365
    },
    {
      "entropy": 5.5747175216674805,
      "epoch": 0.7154489807379121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.5736,
      "mean_token_accuracy": 0.17937180995941163,
      "num_tokens": 14220672.0,
      "step": 6370
    },
    {
      "entropy": 5.715974426269531,
      "epoch": 0.7160105576458696,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.5765,
      "mean_token_accuracy": 0.18228082805871965,
      "num_tokens": 14232348.0,
      "step": 6375
    },
    {
      "entropy": 5.65572338104248,
      "epoch": 0.7165721345538272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.5655,
      "mean_token_accuracy": 0.17139022350311278,
      "num_tokens": 14243985.0,
      "step": 6380
    },
    {
      "entropy": 5.591464376449585,
      "epoch": 0.7171337114617847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495859389142044,
      "loss": 5.4465,
      "mean_token_accuracy": 0.18223742246627808,
      "num_tokens": 14253548.0,
      "step": 6385
    },
    {
      "entropy": 5.677018022537231,
      "epoch": 0.7176952883697423,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495851718716373,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17394395023584366,
      "num_tokens": 14264757.0,
      "step": 6390
    },
    {
      "entropy": 5.55725588798523,
      "epoch": 0.7182568652776998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.5014,
      "mean_token_accuracy": 0.19020429253578186,
      "num_tokens": 14275946.0,
      "step": 6395
    },
    {
      "entropy": 5.633136177062989,
      "epoch": 0.7188184421856574,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.5917,
      "mean_token_accuracy": 0.1715858042240143,
      "num_tokens": 14287006.0,
      "step": 6400
    },
    {
      "entropy": 5.769589424133301,
      "epoch": 0.7193800190936148,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.6154,
      "mean_token_accuracy": 0.17177885621786118,
      "num_tokens": 14299328.0,
      "step": 6405
    },
    {
      "entropy": 5.683426475524902,
      "epoch": 0.7199415960015724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.5995,
      "mean_token_accuracy": 0.17586500495672225,
      "num_tokens": 14310048.0,
      "step": 6410
    },
    {
      "entropy": 5.578950309753418,
      "epoch": 0.7205031729095299,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.5329,
      "mean_token_accuracy": 0.18161720037460327,
      "num_tokens": 14321107.0,
      "step": 6415
    },
    {
      "entropy": 5.588766717910767,
      "epoch": 0.7210647498174875,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.4388,
      "mean_token_accuracy": 0.17925141602754593,
      "num_tokens": 14332561.0,
      "step": 6420
    },
    {
      "entropy": 5.719823837280273,
      "epoch": 0.721626326725445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.6527,
      "mean_token_accuracy": 0.17188818156719207,
      "num_tokens": 14343238.0,
      "step": 6425
    },
    {
      "entropy": 5.61526780128479,
      "epoch": 0.7221879036334026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.5451,
      "mean_token_accuracy": 0.18082940131425856,
      "num_tokens": 14356067.0,
      "step": 6430
    },
    {
      "entropy": 5.63355016708374,
      "epoch": 0.7227494805413601,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17994758635759353,
      "num_tokens": 14367302.0,
      "step": 6435
    },
    {
      "entropy": 5.613122844696045,
      "epoch": 0.7233110574493177,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.5053,
      "mean_token_accuracy": 0.18515312373638154,
      "num_tokens": 14377535.0,
      "step": 6440
    },
    {
      "entropy": 5.712108516693116,
      "epoch": 0.7238726343572752,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.6411,
      "mean_token_accuracy": 0.1715245634317398,
      "num_tokens": 14389493.0,
      "step": 6445
    },
    {
      "entropy": 5.64812970161438,
      "epoch": 0.7244342112652328,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.5475,
      "mean_token_accuracy": 0.17638586610555648,
      "num_tokens": 14401564.0,
      "step": 6450
    },
    {
      "entropy": 5.692028951644898,
      "epoch": 0.7249957881731903,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.6146,
      "mean_token_accuracy": 0.17962251603603363,
      "num_tokens": 14412123.0,
      "step": 6455
    },
    {
      "entropy": 5.654779052734375,
      "epoch": 0.7255573650811479,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.5839,
      "mean_token_accuracy": 0.17979903519153595,
      "num_tokens": 14423614.0,
      "step": 6460
    },
    {
      "entropy": 5.698367547988892,
      "epoch": 0.7261189419891054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.5821,
      "mean_token_accuracy": 0.1727018788456917,
      "num_tokens": 14434830.0,
      "step": 6465
    },
    {
      "entropy": 5.673839664459228,
      "epoch": 0.726680518897063,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.6245,
      "mean_token_accuracy": 0.17713413536548614,
      "num_tokens": 14445015.0,
      "step": 6470
    },
    {
      "entropy": 5.682993030548095,
      "epoch": 0.7272420958050205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.697,
      "mean_token_accuracy": 0.17095409333705902,
      "num_tokens": 14455809.0,
      "step": 6475
    },
    {
      "entropy": 5.685393333435059,
      "epoch": 0.7278036727129781,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.4851,
      "mean_token_accuracy": 0.184173022210598,
      "num_tokens": 14466565.0,
      "step": 6480
    },
    {
      "entropy": 5.679110431671143,
      "epoch": 0.7283652496209356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.5926,
      "mean_token_accuracy": 0.18195963203907012,
      "num_tokens": 14477383.0,
      "step": 6485
    },
    {
      "entropy": 5.625385284423828,
      "epoch": 0.7289268265288932,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.5654,
      "mean_token_accuracy": 0.1758306458592415,
      "num_tokens": 14489968.0,
      "step": 6490
    },
    {
      "entropy": 5.6930827617645265,
      "epoch": 0.7294884034368507,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.6123,
      "mean_token_accuracy": 0.17990701347589494,
      "num_tokens": 14500808.0,
      "step": 6495
    },
    {
      "entropy": 5.624130773544311,
      "epoch": 0.7300499803448082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.6243,
      "mean_token_accuracy": 0.17792190462350846,
      "num_tokens": 14511444.0,
      "step": 6500
    },
    {
      "entropy": 5.620875692367553,
      "epoch": 0.7306115572527657,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1812374547123909,
      "num_tokens": 14522562.0,
      "step": 6505
    },
    {
      "entropy": 5.64646201133728,
      "epoch": 0.7311731341607233,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.5665,
      "mean_token_accuracy": 0.17758426070213318,
      "num_tokens": 14533497.0,
      "step": 6510
    },
    {
      "entropy": 5.692796039581299,
      "epoch": 0.7317347110686808,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004956576733121,
      "loss": 5.598,
      "mean_token_accuracy": 0.16916249245405196,
      "num_tokens": 14545637.0,
      "step": 6515
    },
    {
      "entropy": 5.753426265716553,
      "epoch": 0.7322962879766384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.5956,
      "mean_token_accuracy": 0.17224169969558717,
      "num_tokens": 14556826.0,
      "step": 6520
    },
    {
      "entropy": 5.550675916671753,
      "epoch": 0.7328578648845959,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000495641959931013,
      "loss": 5.5334,
      "mean_token_accuracy": 0.17258528470993043,
      "num_tokens": 14567315.0,
      "step": 6525
    },
    {
      "entropy": 5.674978303909302,
      "epoch": 0.7334194417925535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.5959,
      "mean_token_accuracy": 0.18003043234348298,
      "num_tokens": 14577685.0,
      "step": 6530
    },
    {
      "entropy": 5.717158794403076,
      "epoch": 0.733981018700511,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495626218448624,
      "loss": 5.5573,
      "mean_token_accuracy": 0.18203723579645156,
      "num_tokens": 14588885.0,
      "step": 6535
    },
    {
      "entropy": 5.646932458877563,
      "epoch": 0.7345425956084686,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.5742,
      "mean_token_accuracy": 0.18059505373239518,
      "num_tokens": 14599774.0,
      "step": 6540
    },
    {
      "entropy": 5.73839750289917,
      "epoch": 0.7351041725164261,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.6226,
      "mean_token_accuracy": 0.1707258015871048,
      "num_tokens": 14612101.0,
      "step": 6545
    },
    {
      "entropy": 5.768576145172119,
      "epoch": 0.7356657494243837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.6829,
      "mean_token_accuracy": 0.16432776749134065,
      "num_tokens": 14622151.0,
      "step": 6550
    },
    {
      "entropy": 5.67973108291626,
      "epoch": 0.7362273263323412,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.5817,
      "mean_token_accuracy": 0.17310305088758468,
      "num_tokens": 14634631.0,
      "step": 6555
    },
    {
      "entropy": 5.577903175354004,
      "epoch": 0.7367889032402988,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.5306,
      "mean_token_accuracy": 0.1801718667149544,
      "num_tokens": 14645429.0,
      "step": 6560
    },
    {
      "entropy": 5.575721406936646,
      "epoch": 0.7373504801482563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.5515,
      "mean_token_accuracy": 0.18089888244867325,
      "num_tokens": 14657309.0,
      "step": 6565
    },
    {
      "entropy": 5.642233753204346,
      "epoch": 0.7379120570562139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.4547,
      "mean_token_accuracy": 0.17947259545326233,
      "num_tokens": 14668578.0,
      "step": 6570
    },
    {
      "entropy": 5.633334541320801,
      "epoch": 0.7384736339641714,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.6982,
      "mean_token_accuracy": 0.17283702045679092,
      "num_tokens": 14679120.0,
      "step": 6575
    },
    {
      "entropy": 5.71691861152649,
      "epoch": 0.739035210872129,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.637,
      "mean_token_accuracy": 0.17072526067495347,
      "num_tokens": 14690849.0,
      "step": 6580
    },
    {
      "entropy": 5.725338459014893,
      "epoch": 0.7395967877800865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.5853,
      "mean_token_accuracy": 0.17834530174732208,
      "num_tokens": 14702431.0,
      "step": 6585
    },
    {
      "entropy": 5.563197231292724,
      "epoch": 0.7401583646880441,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.4968,
      "mean_token_accuracy": 0.18283657729625702,
      "num_tokens": 14714271.0,
      "step": 6590
    },
    {
      "entropy": 5.62459077835083,
      "epoch": 0.7407199415960015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.6289,
      "mean_token_accuracy": 0.17922379374504088,
      "num_tokens": 14724796.0,
      "step": 6595
    },
    {
      "entropy": 5.7125084400177,
      "epoch": 0.7412815185039591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495523213982657,
      "loss": 5.5526,
      "mean_token_accuracy": 0.1789660319685936,
      "num_tokens": 14735331.0,
      "step": 6600
    },
    {
      "entropy": 5.759558486938476,
      "epoch": 0.7418430954119166,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.6177,
      "mean_token_accuracy": 0.17553897202014923,
      "num_tokens": 14748595.0,
      "step": 6605
    },
    {
      "entropy": 5.696309328079224,
      "epoch": 0.7424046723198742,
      "grad_norm": 1.0,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1770661875605583,
      "num_tokens": 14759501.0,
      "step": 6610
    },
    {
      "entropy": 5.581427764892578,
      "epoch": 0.7429662492278317,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.5808,
      "mean_token_accuracy": 0.17926053404808046,
      "num_tokens": 14769752.0,
      "step": 6615
    },
    {
      "entropy": 5.649709987640381,
      "epoch": 0.7435278261357893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.5369,
      "mean_token_accuracy": 0.1772114470601082,
      "num_tokens": 14781166.0,
      "step": 6620
    },
    {
      "entropy": 5.623387956619263,
      "epoch": 0.7440894030437468,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495483280880511,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18195586204528807,
      "num_tokens": 14790536.0,
      "step": 6625
    },
    {
      "entropy": 5.708026552200318,
      "epoch": 0.7446509799517044,
      "grad_norm": 1.0625,
      "learning_rate": 0.000495475273198244,
      "loss": 5.633,
      "mean_token_accuracy": 0.17637256681919097,
      "num_tokens": 14801673.0,
      "step": 6630
    },
    {
      "entropy": 5.685859251022339,
      "epoch": 0.7452125568596619,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.534,
      "mean_token_accuracy": 0.17445341795682906,
      "num_tokens": 14812345.0,
      "step": 6635
    },
    {
      "entropy": 5.727948570251465,
      "epoch": 0.7457741337676195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.71,
      "mean_token_accuracy": 0.17131739258766174,
      "num_tokens": 14823620.0,
      "step": 6640
    },
    {
      "entropy": 5.659303665161133,
      "epoch": 0.746335710675577,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1815517619252205,
      "num_tokens": 14833487.0,
      "step": 6645
    },
    {
      "entropy": 5.528368377685547,
      "epoch": 0.7468972875835346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.3856,
      "mean_token_accuracy": 0.18170241862535477,
      "num_tokens": 14844421.0,
      "step": 6650
    },
    {
      "entropy": 5.58285117149353,
      "epoch": 0.7474588644914921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.5516,
      "mean_token_accuracy": 0.18573013246059417,
      "num_tokens": 14855525.0,
      "step": 6655
    },
    {
      "entropy": 5.746717357635498,
      "epoch": 0.7480204413994497,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.6344,
      "mean_token_accuracy": 0.17335387468338012,
      "num_tokens": 14866717.0,
      "step": 6660
    },
    {
      "entropy": 5.617247200012207,
      "epoch": 0.7485820183074072,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.524,
      "mean_token_accuracy": 0.18003176599740983,
      "num_tokens": 14877372.0,
      "step": 6665
    },
    {
      "entropy": 5.609625339508057,
      "epoch": 0.7491435952153648,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.5619,
      "mean_token_accuracy": 0.18102132230997087,
      "num_tokens": 14888618.0,
      "step": 6670
    },
    {
      "entropy": 5.618066883087158,
      "epoch": 0.7497051721233223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.582,
      "mean_token_accuracy": 0.1763724699616432,
      "num_tokens": 14900421.0,
      "step": 6675
    },
    {
      "entropy": 5.688990259170533,
      "epoch": 0.7502667490312799,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.5426,
      "mean_token_accuracy": 0.17627472281455994,
      "num_tokens": 14910634.0,
      "step": 6680
    },
    {
      "entropy": 5.718801259994507,
      "epoch": 0.7508283259392374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.6102,
      "mean_token_accuracy": 0.16975564509630203,
      "num_tokens": 14921408.0,
      "step": 6685
    },
    {
      "entropy": 5.664679384231567,
      "epoch": 0.7513899028471949,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1800124704837799,
      "num_tokens": 14933435.0,
      "step": 6690
    },
    {
      "entropy": 5.587392330169678,
      "epoch": 0.7519514797551524,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.6067,
      "mean_token_accuracy": 0.17205231189727782,
      "num_tokens": 14946488.0,
      "step": 6695
    },
    {
      "entropy": 5.611816883087158,
      "epoch": 0.75251305666311,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.4678,
      "mean_token_accuracy": 0.1828654393553734,
      "num_tokens": 14958053.0,
      "step": 6700
    },
    {
      "entropy": 5.687585306167603,
      "epoch": 0.7530746335710675,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17251309305429458,
      "num_tokens": 14968922.0,
      "step": 6705
    },
    {
      "entropy": 5.63105788230896,
      "epoch": 0.7536362104790251,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.5565,
      "mean_token_accuracy": 0.18053768426179886,
      "num_tokens": 14979541.0,
      "step": 6710
    },
    {
      "entropy": 5.682077217102051,
      "epoch": 0.7541977873869826,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.592,
      "mean_token_accuracy": 0.17432711124420167,
      "num_tokens": 14990009.0,
      "step": 6715
    },
    {
      "entropy": 5.6482017040252686,
      "epoch": 0.7547593642949402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.5922,
      "mean_token_accuracy": 0.1729342371225357,
      "num_tokens": 15001712.0,
      "step": 6720
    },
    {
      "entropy": 5.6821846008300785,
      "epoch": 0.7553209412028977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.5338,
      "mean_token_accuracy": 0.18202275931835174,
      "num_tokens": 15012574.0,
      "step": 6725
    },
    {
      "entropy": 5.576630592346191,
      "epoch": 0.7558825181108553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.536,
      "mean_token_accuracy": 0.18243104964494705,
      "num_tokens": 15023034.0,
      "step": 6730
    },
    {
      "entropy": 5.7357114315032955,
      "epoch": 0.7564440950188128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.583,
      "mean_token_accuracy": 0.17332218885421752,
      "num_tokens": 15033925.0,
      "step": 6735
    },
    {
      "entropy": 5.629609394073486,
      "epoch": 0.7570056719267704,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.5648,
      "mean_token_accuracy": 0.18474339842796325,
      "num_tokens": 15044947.0,
      "step": 6740
    },
    {
      "entropy": 5.641076469421387,
      "epoch": 0.7575672488347279,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.4805,
      "mean_token_accuracy": 0.18399559259414672,
      "num_tokens": 15054784.0,
      "step": 6745
    },
    {
      "entropy": 5.573967218399048,
      "epoch": 0.7581288257426855,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.557,
      "mean_token_accuracy": 0.17349208146333694,
      "num_tokens": 15066272.0,
      "step": 6750
    },
    {
      "entropy": 5.703629016876221,
      "epoch": 0.758690402650643,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.6326,
      "mean_token_accuracy": 0.17263609766960145,
      "num_tokens": 15079504.0,
      "step": 6755
    },
    {
      "entropy": 5.73172698020935,
      "epoch": 0.7592519795586006,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.5724,
      "mean_token_accuracy": 0.18210249692201613,
      "num_tokens": 15089623.0,
      "step": 6760
    },
    {
      "entropy": 5.607945919036865,
      "epoch": 0.7598135564665581,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.4771,
      "mean_token_accuracy": 0.17305199503898622,
      "num_tokens": 15100816.0,
      "step": 6765
    },
    {
      "entropy": 5.619532680511474,
      "epoch": 0.7603751333745157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.5681,
      "mean_token_accuracy": 0.18450591266155242,
      "num_tokens": 15111697.0,
      "step": 6770
    },
    {
      "entropy": 5.6809234619140625,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.6583,
      "mean_token_accuracy": 0.17277388274669647,
      "num_tokens": 15122869.0,
      "step": 6775
    },
    {
      "entropy": 5.585271644592285,
      "epoch": 0.7614982871904308,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.4476,
      "mean_token_accuracy": 0.1824674978852272,
      "num_tokens": 15134709.0,
      "step": 6780
    },
    {
      "entropy": 5.618035459518433,
      "epoch": 0.7620598640983883,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.4912,
      "mean_token_accuracy": 0.1805910736322403,
      "num_tokens": 15145319.0,
      "step": 6785
    },
    {
      "entropy": 5.5981072902679445,
      "epoch": 0.7626214410063458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495215322197552,
      "loss": 5.5212,
      "mean_token_accuracy": 0.1760521337389946,
      "num_tokens": 15156463.0,
      "step": 6790
    },
    {
      "entropy": 5.692944288253784,
      "epoch": 0.7631830179143033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.6445,
      "mean_token_accuracy": 0.16605111286044122,
      "num_tokens": 15167618.0,
      "step": 6795
    },
    {
      "entropy": 5.672136211395264,
      "epoch": 0.7637445948222609,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.5881,
      "mean_token_accuracy": 0.17731750905513763,
      "num_tokens": 15179312.0,
      "step": 6800
    },
    {
      "entropy": 5.575615406036377,
      "epoch": 0.7643061717302184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.531,
      "mean_token_accuracy": 0.18109908252954482,
      "num_tokens": 15191370.0,
      "step": 6805
    },
    {
      "entropy": 5.690320682525635,
      "epoch": 0.764867748638176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.4933,
      "mean_token_accuracy": 0.1843721464276314,
      "num_tokens": 15201060.0,
      "step": 6810
    },
    {
      "entropy": 5.567106008529663,
      "epoch": 0.7654293255461335,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18041964024305343,
      "num_tokens": 15212186.0,
      "step": 6815
    },
    {
      "entropy": 5.544131898880005,
      "epoch": 0.7659909024540911,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.4382,
      "mean_token_accuracy": 0.18858288377523422,
      "num_tokens": 15223301.0,
      "step": 6820
    },
    {
      "entropy": 5.615245676040649,
      "epoch": 0.7665524793620486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.5977,
      "mean_token_accuracy": 0.17640408724546433,
      "num_tokens": 15234475.0,
      "step": 6825
    },
    {
      "entropy": 5.6895167350769045,
      "epoch": 0.7671140562700062,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.5712,
      "mean_token_accuracy": 0.17795952558517455,
      "num_tokens": 15245877.0,
      "step": 6830
    },
    {
      "entropy": 5.657568693161011,
      "epoch": 0.7676756331779637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.5566,
      "mean_token_accuracy": 0.18239046037197112,
      "num_tokens": 15257375.0,
      "step": 6835
    },
    {
      "entropy": 5.628437662124634,
      "epoch": 0.7682372100859213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.5809,
      "mean_token_accuracy": 0.17615204602479934,
      "num_tokens": 15268079.0,
      "step": 6840
    },
    {
      "entropy": 5.630699539184571,
      "epoch": 0.7687987869938788,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.5634,
      "mean_token_accuracy": 0.17514654099941254,
      "num_tokens": 15278678.0,
      "step": 6845
    },
    {
      "entropy": 5.731458759307861,
      "epoch": 0.7693603639018364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.6736,
      "mean_token_accuracy": 0.17262617200613023,
      "num_tokens": 15290080.0,
      "step": 6850
    },
    {
      "entropy": 5.666080188751221,
      "epoch": 0.7699219408097939,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.596,
      "mean_token_accuracy": 0.17108689546585082,
      "num_tokens": 15300602.0,
      "step": 6855
    },
    {
      "entropy": 5.679002046585083,
      "epoch": 0.7704835177177515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.5994,
      "mean_token_accuracy": 0.18032609522342682,
      "num_tokens": 15312823.0,
      "step": 6860
    },
    {
      "entropy": 5.705432748794555,
      "epoch": 0.771045094625709,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495090998008418,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17205487936735153,
      "num_tokens": 15323870.0,
      "step": 6865
    },
    {
      "entropy": 5.601526308059692,
      "epoch": 0.7716066715336666,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495082653656997,
      "loss": 5.4999,
      "mean_token_accuracy": 0.18249612003564836,
      "num_tokens": 15334887.0,
      "step": 6870
    },
    {
      "entropy": 5.6566027164459225,
      "epoch": 0.7721682484416241,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.5961,
      "mean_token_accuracy": 0.17155962884426118,
      "num_tokens": 15345871.0,
      "step": 6875
    },
    {
      "entropy": 5.596802568435669,
      "epoch": 0.7727298253495817,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.4575,
      "mean_token_accuracy": 0.1836513876914978,
      "num_tokens": 15356530.0,
      "step": 6880
    },
    {
      "entropy": 5.604869699478149,
      "epoch": 0.7732914022575391,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.5235,
      "mean_token_accuracy": 0.1817237541079521,
      "num_tokens": 15368756.0,
      "step": 6885
    },
    {
      "entropy": 5.684064817428589,
      "epoch": 0.7738529791654967,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.5977,
      "mean_token_accuracy": 0.18085209876298905,
      "num_tokens": 15379450.0,
      "step": 6890
    },
    {
      "entropy": 5.624396705627442,
      "epoch": 0.7744145560734542,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.5198,
      "mean_token_accuracy": 0.1743107855319977,
      "num_tokens": 15389501.0,
      "step": 6895
    },
    {
      "entropy": 5.5190822124481205,
      "epoch": 0.7749761329814118,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.5206,
      "mean_token_accuracy": 0.18011448979377748,
      "num_tokens": 15400608.0,
      "step": 6900
    },
    {
      "entropy": 5.710084533691406,
      "epoch": 0.7755377098893693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.603,
      "mean_token_accuracy": 0.17645572125911713,
      "num_tokens": 15411709.0,
      "step": 6905
    },
    {
      "entropy": 5.582209825515747,
      "epoch": 0.7760992867973269,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18789805173873902,
      "num_tokens": 15422292.0,
      "step": 6910
    },
    {
      "entropy": 5.6498126029968265,
      "epoch": 0.7766608637052844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495007239187896,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17521766871213912,
      "num_tokens": 15434439.0,
      "step": 6915
    },
    {
      "entropy": 5.655129098892212,
      "epoch": 0.777222440613242,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17417643517255782,
      "num_tokens": 15445480.0,
      "step": 6920
    },
    {
      "entropy": 5.625287294387817,
      "epoch": 0.7777840175211995,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.5337,
      "mean_token_accuracy": 0.1788464143872261,
      "num_tokens": 15456960.0,
      "step": 6925
    },
    {
      "entropy": 5.583571100234986,
      "epoch": 0.7783455944291571,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494981974929824,
      "loss": 5.544,
      "mean_token_accuracy": 0.17995237857103347,
      "num_tokens": 15468169.0,
      "step": 6930
    },
    {
      "entropy": 5.6113402366638185,
      "epoch": 0.7789071713371146,
      "grad_norm": 1.09375,
      "learning_rate": 0.000494973539501486,
      "loss": 5.5293,
      "mean_token_accuracy": 0.17978586107492447,
      "num_tokens": 15478740.0,
      "step": 6935
    },
    {
      "entropy": 5.64364652633667,
      "epoch": 0.7794687482450722,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1706491157412529,
      "num_tokens": 15489705.0,
      "step": 6940
    },
    {
      "entropy": 5.578392839431762,
      "epoch": 0.7800303251530297,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.4462,
      "mean_token_accuracy": 0.18012455254793167,
      "num_tokens": 15500549.0,
      "step": 6945
    },
    {
      "entropy": 5.718480587005615,
      "epoch": 0.7805919020609873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.6087,
      "mean_token_accuracy": 0.17435625344514846,
      "num_tokens": 15511124.0,
      "step": 6950
    },
    {
      "entropy": 5.609439659118652,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.5027,
      "mean_token_accuracy": 0.18258694410324097,
      "num_tokens": 15521810.0,
      "step": 6955
    },
    {
      "entropy": 5.66283917427063,
      "epoch": 0.7817150558769024,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17872778922319413,
      "num_tokens": 15533108.0,
      "step": 6960
    },
    {
      "entropy": 5.627768659591675,
      "epoch": 0.7822766327848599,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.5627,
      "mean_token_accuracy": 0.17870569974184036,
      "num_tokens": 15544841.0,
      "step": 6965
    },
    {
      "entropy": 5.61465892791748,
      "epoch": 0.7828382096928175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.5608,
      "mean_token_accuracy": 0.17548853754997254,
      "num_tokens": 15556297.0,
      "step": 6970
    },
    {
      "entropy": 5.61036696434021,
      "epoch": 0.783399786600775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.4589,
      "mean_token_accuracy": 0.18257362693548201,
      "num_tokens": 15566607.0,
      "step": 6975
    },
    {
      "entropy": 5.70405011177063,
      "epoch": 0.7839613635087325,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.6147,
      "mean_token_accuracy": 0.17051844596862792,
      "num_tokens": 15577435.0,
      "step": 6980
    },
    {
      "entropy": 5.639044189453125,
      "epoch": 0.78452294041669,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.5254,
      "mean_token_accuracy": 0.18027030378580094,
      "num_tokens": 15587844.0,
      "step": 6985
    },
    {
      "entropy": 5.669044351577758,
      "epoch": 0.7850845173246476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.5886,
      "mean_token_accuracy": 0.17704415023326875,
      "num_tokens": 15599586.0,
      "step": 6990
    },
    {
      "entropy": 5.573870754241943,
      "epoch": 0.7856460942326051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.4768,
      "mean_token_accuracy": 0.18209463059902192,
      "num_tokens": 15610838.0,
      "step": 6995
    },
    {
      "entropy": 5.6168153285980225,
      "epoch": 0.7862076711405627,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.5594,
      "mean_token_accuracy": 0.17618024200201035,
      "num_tokens": 15622652.0,
      "step": 7000
    },
    {
      "entropy": 5.716824579238891,
      "epoch": 0.7867692480485202,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.654,
      "mean_token_accuracy": 0.17388251274824143,
      "num_tokens": 15634280.0,
      "step": 7005
    },
    {
      "entropy": 5.572071170806884,
      "epoch": 0.7873308249564778,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.3767,
      "mean_token_accuracy": 0.18992173969745635,
      "num_tokens": 15644088.0,
      "step": 7010
    },
    {
      "entropy": 5.62109866142273,
      "epoch": 0.7878924018644353,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.6222,
      "mean_token_accuracy": 0.17247122526168823,
      "num_tokens": 15654968.0,
      "step": 7015
    },
    {
      "entropy": 5.619490337371826,
      "epoch": 0.7884539787723929,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.5262,
      "mean_token_accuracy": 0.18326697051525115,
      "num_tokens": 15666758.0,
      "step": 7020
    },
    {
      "entropy": 5.530732345581055,
      "epoch": 0.7890155556803504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19031909704208375,
      "num_tokens": 15677295.0,
      "step": 7025
    },
    {
      "entropy": 5.647847747802734,
      "epoch": 0.789577132588308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.5898,
      "mean_token_accuracy": 0.17960943132638932,
      "num_tokens": 15689508.0,
      "step": 7030
    },
    {
      "entropy": 5.71213321685791,
      "epoch": 0.7901387094962655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.5478,
      "mean_token_accuracy": 0.17901325672864915,
      "num_tokens": 15701289.0,
      "step": 7035
    },
    {
      "entropy": 5.562152433395386,
      "epoch": 0.7907002864042231,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.4906,
      "mean_token_accuracy": 0.1727774813771248,
      "num_tokens": 15711579.0,
      "step": 7040
    },
    {
      "entropy": 5.611237621307373,
      "epoch": 0.7912618633121806,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.5803,
      "mean_token_accuracy": 0.18268261551856996,
      "num_tokens": 15722629.0,
      "step": 7045
    },
    {
      "entropy": 5.645194244384766,
      "epoch": 0.7918234402201382,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.5212,
      "mean_token_accuracy": 0.17456017136573793,
      "num_tokens": 15735818.0,
      "step": 7050
    },
    {
      "entropy": 5.5891368865966795,
      "epoch": 0.7923850171280957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.5819,
      "mean_token_accuracy": 0.17798920273780822,
      "num_tokens": 15746914.0,
      "step": 7055
    },
    {
      "entropy": 5.586161708831787,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.15625,
      "learning_rate": 0.000494760378181364,
      "loss": 5.4843,
      "mean_token_accuracy": 0.17352930903434755,
      "num_tokens": 15757597.0,
      "step": 7060
    },
    {
      "entropy": 5.605369901657104,
      "epoch": 0.7935081709440108,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.4104,
      "mean_token_accuracy": 0.1861075982451439,
      "num_tokens": 15768860.0,
      "step": 7065
    },
    {
      "entropy": 5.6157104015350345,
      "epoch": 0.7940697478519684,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.5444,
      "mean_token_accuracy": 0.17149641513824462,
      "num_tokens": 15779476.0,
      "step": 7070
    },
    {
      "entropy": 5.578100347518921,
      "epoch": 0.7946313247599258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.5868,
      "mean_token_accuracy": 0.17395118325948716,
      "num_tokens": 15792142.0,
      "step": 7075
    },
    {
      "entropy": 5.61973295211792,
      "epoch": 0.7951929016678834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.5594,
      "mean_token_accuracy": 0.17755297869443892,
      "num_tokens": 15802307.0,
      "step": 7080
    },
    {
      "entropy": 5.578988790512085,
      "epoch": 0.7957544785758409,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.5129,
      "mean_token_accuracy": 0.17733030319213866,
      "num_tokens": 15812986.0,
      "step": 7085
    },
    {
      "entropy": 5.567875957489013,
      "epoch": 0.7963160554837985,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.463,
      "mean_token_accuracy": 0.18132586628198624,
      "num_tokens": 15823938.0,
      "step": 7090
    },
    {
      "entropy": 5.565356779098511,
      "epoch": 0.796877632391756,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.4428,
      "mean_token_accuracy": 0.18291499763727187,
      "num_tokens": 15835813.0,
      "step": 7095
    },
    {
      "entropy": 5.620179748535156,
      "epoch": 0.7974392092997136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494691242717217,
      "loss": 5.5135,
      "mean_token_accuracy": 0.1817249044775963,
      "num_tokens": 15846203.0,
      "step": 7100
    },
    {
      "entropy": 5.617189073562622,
      "epoch": 0.7980007862076711,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.491,
      "mean_token_accuracy": 0.18329188376665115,
      "num_tokens": 15856919.0,
      "step": 7105
    },
    {
      "entropy": 5.640201139450073,
      "epoch": 0.7985623631156287,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494673888895125,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17678974568843842,
      "num_tokens": 15867149.0,
      "step": 7110
    },
    {
      "entropy": 5.534529066085815,
      "epoch": 0.7991239400235862,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.4821,
      "mean_token_accuracy": 0.17967871576547623,
      "num_tokens": 15877229.0,
      "step": 7115
    },
    {
      "entropy": 5.612982368469238,
      "epoch": 0.7996855169315438,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.4931,
      "mean_token_accuracy": 0.17960063070058824,
      "num_tokens": 15888799.0,
      "step": 7120
    },
    {
      "entropy": 5.567969942092896,
      "epoch": 0.8002470938395013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.4686,
      "mean_token_accuracy": 0.19139161407947541,
      "num_tokens": 15900144.0,
      "step": 7125
    },
    {
      "entropy": 5.665867805480957,
      "epoch": 0.8008086707474589,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.6336,
      "mean_token_accuracy": 0.17360662966966628,
      "num_tokens": 15912030.0,
      "step": 7130
    },
    {
      "entropy": 5.627162742614746,
      "epoch": 0.8013702476554164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.5136,
      "mean_token_accuracy": 0.1804039403796196,
      "num_tokens": 15923606.0,
      "step": 7135
    },
    {
      "entropy": 5.525466537475586,
      "epoch": 0.801931824563374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.5001,
      "mean_token_accuracy": 0.1816282406449318,
      "num_tokens": 15935077.0,
      "step": 7140
    },
    {
      "entropy": 5.528133392333984,
      "epoch": 0.8024934014713315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.4395,
      "mean_token_accuracy": 0.18267178535461426,
      "num_tokens": 15945331.0,
      "step": 7145
    },
    {
      "entropy": 5.625937223434448,
      "epoch": 0.8030549783792891,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.5317,
      "mean_token_accuracy": 0.177494777739048,
      "num_tokens": 15957031.0,
      "step": 7150
    },
    {
      "entropy": 5.671985340118408,
      "epoch": 0.8036165552872466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.5756,
      "mean_token_accuracy": 0.17943753749132157,
      "num_tokens": 15967545.0,
      "step": 7155
    },
    {
      "entropy": 5.710970115661621,
      "epoch": 0.8041781321952042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.614,
      "mean_token_accuracy": 0.1680646523833275,
      "num_tokens": 15978700.0,
      "step": 7160
    },
    {
      "entropy": 5.549652862548828,
      "epoch": 0.8047397091031617,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.5004,
      "mean_token_accuracy": 0.18439516574144363,
      "num_tokens": 15989712.0,
      "step": 7165
    },
    {
      "entropy": 5.742206001281739,
      "epoch": 0.8053012860111193,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494569178502031,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1759636864066124,
      "num_tokens": 16002380.0,
      "step": 7170
    },
    {
      "entropy": 5.711532402038574,
      "epoch": 0.8058628629190767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.4894,
      "mean_token_accuracy": 0.1773687332868576,
      "num_tokens": 16013749.0,
      "step": 7175
    },
    {
      "entropy": 5.645091247558594,
      "epoch": 0.8064244398270343,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.6528,
      "mean_token_accuracy": 0.1744043931365013,
      "num_tokens": 16025131.0,
      "step": 7180
    },
    {
      "entropy": 5.638998174667359,
      "epoch": 0.8069860167349918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.5341,
      "mean_token_accuracy": 0.17833992689847947,
      "num_tokens": 16037269.0,
      "step": 7185
    },
    {
      "entropy": 5.7754237174987795,
      "epoch": 0.8075475936429494,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.7265,
      "mean_token_accuracy": 0.17038890421390535,
      "num_tokens": 16050992.0,
      "step": 7190
    },
    {
      "entropy": 5.6251636981964115,
      "epoch": 0.8081091705509069,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.5598,
      "mean_token_accuracy": 0.1733875721693039,
      "num_tokens": 16062309.0,
      "step": 7195
    },
    {
      "entropy": 5.756367492675781,
      "epoch": 0.8086707474588645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.581,
      "mean_token_accuracy": 0.17815395891666413,
      "num_tokens": 16073079.0,
      "step": 7200
    },
    {
      "entropy": 5.636918449401856,
      "epoch": 0.809232324366822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.4812,
      "mean_token_accuracy": 0.18240586668252945,
      "num_tokens": 16082885.0,
      "step": 7205
    },
    {
      "entropy": 5.44354248046875,
      "epoch": 0.8097939012747796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1906597673892975,
      "num_tokens": 16094173.0,
      "step": 7210
    },
    {
      "entropy": 5.610472202301025,
      "epoch": 0.8103554781827371,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494489985015675,
      "loss": 5.4802,
      "mean_token_accuracy": 0.17765818387269974,
      "num_tokens": 16106351.0,
      "step": 7215
    },
    {
      "entropy": 5.573151445388794,
      "epoch": 0.8109170550906947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.4836,
      "mean_token_accuracy": 0.17927531152963638,
      "num_tokens": 16116892.0,
      "step": 7220
    },
    {
      "entropy": 5.669474697113037,
      "epoch": 0.8114786319986522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.6198,
      "mean_token_accuracy": 0.1687384530901909,
      "num_tokens": 16128316.0,
      "step": 7225
    },
    {
      "entropy": 5.626483821868897,
      "epoch": 0.8120402089066098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.4535,
      "mean_token_accuracy": 0.17954584062099457,
      "num_tokens": 16139245.0,
      "step": 7230
    },
    {
      "entropy": 5.552922773361206,
      "epoch": 0.8126017858145673,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049445460619367,
      "loss": 5.4844,
      "mean_token_accuracy": 0.1813029646873474,
      "num_tokens": 16150893.0,
      "step": 7235
    },
    {
      "entropy": 5.611425065994263,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.6503,
      "mean_token_accuracy": 0.16410815566778184,
      "num_tokens": 16163430.0,
      "step": 7240
    },
    {
      "entropy": 5.757813405990601,
      "epoch": 0.8137249396304824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.53,
      "mean_token_accuracy": 0.18086985647678375,
      "num_tokens": 16173954.0,
      "step": 7245
    },
    {
      "entropy": 5.586259508132935,
      "epoch": 0.81428651653844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.5687,
      "mean_token_accuracy": 0.17602573484182357,
      "num_tokens": 16185491.0,
      "step": 7250
    },
    {
      "entropy": 5.497488117218017,
      "epoch": 0.8148480934463975,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19166387915611266,
      "num_tokens": 16195301.0,
      "step": 7255
    },
    {
      "entropy": 5.659170913696289,
      "epoch": 0.8154096703543551,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.5571,
      "mean_token_accuracy": 0.17331268936395644,
      "num_tokens": 16206620.0,
      "step": 7260
    },
    {
      "entropy": 5.648644590377808,
      "epoch": 0.8159712472623126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.6375,
      "mean_token_accuracy": 0.17144708782434465,
      "num_tokens": 16218693.0,
      "step": 7265
    },
    {
      "entropy": 5.7061583518981935,
      "epoch": 0.8165328241702701,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.5237,
      "mean_token_accuracy": 0.18040265440940856,
      "num_tokens": 16229124.0,
      "step": 7270
    },
    {
      "entropy": 5.559056901931763,
      "epoch": 0.8170944010782276,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.5089,
      "mean_token_accuracy": 0.17560482025146484,
      "num_tokens": 16239999.0,
      "step": 7275
    },
    {
      "entropy": 5.628651857376099,
      "epoch": 0.8176559779861852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.571,
      "mean_token_accuracy": 0.18164185136556626,
      "num_tokens": 16251027.0,
      "step": 7280
    },
    {
      "entropy": 5.667185211181641,
      "epoch": 0.8182175548941427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.5314,
      "mean_token_accuracy": 0.1802248865365982,
      "num_tokens": 16262924.0,
      "step": 7285
    },
    {
      "entropy": 5.697255945205688,
      "epoch": 0.8187791318021003,
      "grad_norm": 1.125,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.6383,
      "mean_token_accuracy": 0.17880614697933198,
      "num_tokens": 16274481.0,
      "step": 7290
    },
    {
      "entropy": 5.637316036224365,
      "epoch": 0.8193407087100578,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.5605,
      "mean_token_accuracy": 0.1813304215669632,
      "num_tokens": 16285214.0,
      "step": 7295
    },
    {
      "entropy": 5.672893047332764,
      "epoch": 0.8199022856180154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.6307,
      "mean_token_accuracy": 0.1746322527527809,
      "num_tokens": 16296077.0,
      "step": 7300
    },
    {
      "entropy": 5.700228357315064,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.5368,
      "mean_token_accuracy": 0.17813003361225127,
      "num_tokens": 16307514.0,
      "step": 7305
    },
    {
      "entropy": 5.680727100372314,
      "epoch": 0.8210254394339305,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.5332,
      "mean_token_accuracy": 0.17984469681978227,
      "num_tokens": 16318734.0,
      "step": 7310
    },
    {
      "entropy": 5.579897356033325,
      "epoch": 0.821587016341888,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494311973083621,
      "loss": 5.5628,
      "mean_token_accuracy": 0.17839521467685698,
      "num_tokens": 16329057.0,
      "step": 7315
    },
    {
      "entropy": 5.579474544525146,
      "epoch": 0.8221485932498456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.5024,
      "mean_token_accuracy": 0.18085712641477586,
      "num_tokens": 16339646.0,
      "step": 7320
    },
    {
      "entropy": 5.652965307235718,
      "epoch": 0.8227101701578031,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.5888,
      "mean_token_accuracy": 0.17340163737535477,
      "num_tokens": 16349783.0,
      "step": 7325
    },
    {
      "entropy": 5.668771743774414,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18269888162612916,
      "num_tokens": 16361078.0,
      "step": 7330
    },
    {
      "entropy": 5.528364753723144,
      "epoch": 0.8238333239737182,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18162541687488556,
      "num_tokens": 16371308.0,
      "step": 7335
    },
    {
      "entropy": 5.628272008895874,
      "epoch": 0.8243949008816758,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.5518,
      "mean_token_accuracy": 0.18339990228414535,
      "num_tokens": 16382682.0,
      "step": 7340
    },
    {
      "entropy": 5.7226831912994385,
      "epoch": 0.8249564777896333,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.5871,
      "mean_token_accuracy": 0.17570515275001525,
      "num_tokens": 16393715.0,
      "step": 7345
    },
    {
      "entropy": 5.574453592300415,
      "epoch": 0.8255180546975909,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.5441,
      "mean_token_accuracy": 0.1787221685051918,
      "num_tokens": 16406332.0,
      "step": 7350
    },
    {
      "entropy": 5.5816456317901615,
      "epoch": 0.8260796316055484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.4876,
      "mean_token_accuracy": 0.18046286553144456,
      "num_tokens": 16417702.0,
      "step": 7355
    },
    {
      "entropy": 5.6356000900268555,
      "epoch": 0.826641208513506,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.5239,
      "mean_token_accuracy": 0.17496747821569442,
      "num_tokens": 16428497.0,
      "step": 7360
    },
    {
      "entropy": 5.63239164352417,
      "epoch": 0.8272027854214634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17938542664051055,
      "num_tokens": 16440637.0,
      "step": 7365
    },
    {
      "entropy": 5.597147846221924,
      "epoch": 0.827764362329421,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.547,
      "mean_token_accuracy": 0.1762922652065754,
      "num_tokens": 16453213.0,
      "step": 7370
    },
    {
      "entropy": 5.601246643066406,
      "epoch": 0.8283259392373785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.5616,
      "mean_token_accuracy": 0.1792689397931099,
      "num_tokens": 16464345.0,
      "step": 7375
    },
    {
      "entropy": 5.570774888992309,
      "epoch": 0.8288875161453361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.4458,
      "mean_token_accuracy": 0.1834470361471176,
      "num_tokens": 16475399.0,
      "step": 7380
    },
    {
      "entropy": 5.552033090591431,
      "epoch": 0.8294490930532936,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.4295,
      "mean_token_accuracy": 0.18250147253274918,
      "num_tokens": 16485439.0,
      "step": 7385
    },
    {
      "entropy": 5.557843732833862,
      "epoch": 0.8300106699612512,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.4211,
      "mean_token_accuracy": 0.18478182405233384,
      "num_tokens": 16496765.0,
      "step": 7390
    },
    {
      "entropy": 5.50337781906128,
      "epoch": 0.8305722468692087,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.4742,
      "mean_token_accuracy": 0.18376317471265793,
      "num_tokens": 16509199.0,
      "step": 7395
    },
    {
      "entropy": 5.668723487854004,
      "epoch": 0.8311338237771663,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.6168,
      "mean_token_accuracy": 0.17944542020559312,
      "num_tokens": 16520073.0,
      "step": 7400
    },
    {
      "entropy": 5.6101233005523685,
      "epoch": 0.8316954006851238,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.4928,
      "mean_token_accuracy": 0.17857892513275148,
      "num_tokens": 16530919.0,
      "step": 7405
    },
    {
      "entropy": 5.585693073272705,
      "epoch": 0.8322569775930814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18006205558776855,
      "num_tokens": 16542591.0,
      "step": 7410
    },
    {
      "entropy": 5.639138460159302,
      "epoch": 0.8328185545010389,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.5166,
      "mean_token_accuracy": 0.17835039049386978,
      "num_tokens": 16554984.0,
      "step": 7415
    },
    {
      "entropy": 5.569074535369873,
      "epoch": 0.8333801314089965,
      "grad_norm": 1.0,
      "learning_rate": 0.000494122054427151,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18038525581359863,
      "num_tokens": 16566598.0,
      "step": 7420
    },
    {
      "entropy": 5.645244073867798,
      "epoch": 0.833941708316954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.5727,
      "mean_token_accuracy": 0.17438869774341584,
      "num_tokens": 16577868.0,
      "step": 7425
    },
    {
      "entropy": 5.605511569976807,
      "epoch": 0.8345032852249116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.5199,
      "mean_token_accuracy": 0.17894256263971328,
      "num_tokens": 16587891.0,
      "step": 7430
    },
    {
      "entropy": 5.581606817245484,
      "epoch": 0.8350648621328691,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494094671958544,
      "loss": 5.4789,
      "mean_token_accuracy": 0.185255566239357,
      "num_tokens": 16599924.0,
      "step": 7435
    },
    {
      "entropy": 5.564290761947632,
      "epoch": 0.8356264390408267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.4221,
      "mean_token_accuracy": 0.1857829600572586,
      "num_tokens": 16609958.0,
      "step": 7440
    },
    {
      "entropy": 5.592352771759034,
      "epoch": 0.8361880159487842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.5451,
      "mean_token_accuracy": 0.17597225308418274,
      "num_tokens": 16621159.0,
      "step": 7445
    },
    {
      "entropy": 5.665910339355468,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.6495,
      "mean_token_accuracy": 0.17905145585536958,
      "num_tokens": 16632263.0,
      "step": 7450
    },
    {
      "entropy": 5.610781002044678,
      "epoch": 0.8373111697646993,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.5127,
      "mean_token_accuracy": 0.1742522358894348,
      "num_tokens": 16644117.0,
      "step": 7455
    },
    {
      "entropy": 5.504357767105103,
      "epoch": 0.8378727466726568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1833467125892639,
      "num_tokens": 16654529.0,
      "step": 7460
    },
    {
      "entropy": 5.552529144287109,
      "epoch": 0.8384343235806143,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.4581,
      "mean_token_accuracy": 0.18409384936094284,
      "num_tokens": 16666191.0,
      "step": 7465
    },
    {
      "entropy": 5.602557229995727,
      "epoch": 0.8389959004885719,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.4532,
      "mean_token_accuracy": 0.18632374703884125,
      "num_tokens": 16677252.0,
      "step": 7470
    },
    {
      "entropy": 5.514086151123047,
      "epoch": 0.8395574773965294,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494021345117024,
      "loss": 5.5045,
      "mean_token_accuracy": 0.1806218221783638,
      "num_tokens": 16689487.0,
      "step": 7475
    },
    {
      "entropy": 5.5145933628082275,
      "epoch": 0.840119054304487,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18738940060138704,
      "num_tokens": 16699955.0,
      "step": 7480
    },
    {
      "entropy": 5.490093421936035,
      "epoch": 0.8406806312124445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.4548,
      "mean_token_accuracy": 0.18099683672189712,
      "num_tokens": 16711071.0,
      "step": 7485
    },
    {
      "entropy": 5.5467767238616945,
      "epoch": 0.8412422081204021,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.4595,
      "mean_token_accuracy": 0.18532306998968123,
      "num_tokens": 16722115.0,
      "step": 7490
    },
    {
      "entropy": 5.573823499679565,
      "epoch": 0.8418037850283596,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.4847,
      "mean_token_accuracy": 0.1772324413061142,
      "num_tokens": 16733146.0,
      "step": 7495
    },
    {
      "entropy": 5.528927516937256,
      "epoch": 0.8423653619363172,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.4651,
      "mean_token_accuracy": 0.17675635814666749,
      "num_tokens": 16745543.0,
      "step": 7500
    },
    {
      "entropy": 5.652834272384643,
      "epoch": 0.8429269388442747,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.5381,
      "mean_token_accuracy": 0.18086207807064056,
      "num_tokens": 16755782.0,
      "step": 7505
    },
    {
      "entropy": 5.671106624603271,
      "epoch": 0.8434885157522323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.5523,
      "mean_token_accuracy": 0.17602942287921905,
      "num_tokens": 16767509.0,
      "step": 7510
    },
    {
      "entropy": 5.536266279220581,
      "epoch": 0.8440500926601898,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18344242572784425,
      "num_tokens": 16779257.0,
      "step": 7515
    },
    {
      "entropy": 5.494624519348145,
      "epoch": 0.8446116695681474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1824428215622902,
      "num_tokens": 16790769.0,
      "step": 7520
    },
    {
      "entropy": 5.61299409866333,
      "epoch": 0.8451732464761049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.4786,
      "mean_token_accuracy": 0.1779421776533127,
      "num_tokens": 16800054.0,
      "step": 7525
    },
    {
      "entropy": 5.661533546447754,
      "epoch": 0.8457348233840625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.5634,
      "mean_token_accuracy": 0.17657241970300674,
      "num_tokens": 16811485.0,
      "step": 7530
    },
    {
      "entropy": 5.6560791492462155,
      "epoch": 0.84629640029202,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.6189,
      "mean_token_accuracy": 0.18123217225074767,
      "num_tokens": 16822683.0,
      "step": 7535
    },
    {
      "entropy": 5.659206247329712,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.4949,
      "mean_token_accuracy": 0.18079070150852203,
      "num_tokens": 16834200.0,
      "step": 7540
    },
    {
      "entropy": 5.665460824966431,
      "epoch": 0.8474195541079351,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.552,
      "mean_token_accuracy": 0.1727047935128212,
      "num_tokens": 16846399.0,
      "step": 7545
    },
    {
      "entropy": 5.670234203338623,
      "epoch": 0.8479811310158927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.6857,
      "mean_token_accuracy": 0.17400158941745758,
      "num_tokens": 16858322.0,
      "step": 7550
    },
    {
      "entropy": 5.603891086578369,
      "epoch": 0.8485427079238502,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.5161,
      "mean_token_accuracy": 0.18002558946609498,
      "num_tokens": 16869537.0,
      "step": 7555
    },
    {
      "entropy": 5.7228978157043455,
      "epoch": 0.8491042848318077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.5528,
      "mean_token_accuracy": 0.1807059243321419,
      "num_tokens": 16881566.0,
      "step": 7560
    },
    {
      "entropy": 5.505270767211914,
      "epoch": 0.8496658617397652,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18655762821435928,
      "num_tokens": 16892959.0,
      "step": 7565
    },
    {
      "entropy": 5.627756643295288,
      "epoch": 0.8502274386477228,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.5327,
      "mean_token_accuracy": 0.17650023996829986,
      "num_tokens": 16904809.0,
      "step": 7570
    },
    {
      "entropy": 5.6635175228118895,
      "epoch": 0.8507890155556803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.5555,
      "mean_token_accuracy": 0.1722218871116638,
      "num_tokens": 16916969.0,
      "step": 7575
    },
    {
      "entropy": 5.547809886932373,
      "epoch": 0.8513505924636379,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.4493,
      "mean_token_accuracy": 0.18617565035820008,
      "num_tokens": 16928253.0,
      "step": 7580
    },
    {
      "entropy": 5.5388415336608885,
      "epoch": 0.8519121693715954,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.4925,
      "mean_token_accuracy": 0.1745935708284378,
      "num_tokens": 16939875.0,
      "step": 7585
    },
    {
      "entropy": 5.716016340255737,
      "epoch": 0.852473746279553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17623613476753236,
      "num_tokens": 16950173.0,
      "step": 7590
    },
    {
      "entropy": 5.595033264160156,
      "epoch": 0.8530353231875105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.5257,
      "mean_token_accuracy": 0.1806575506925583,
      "num_tokens": 16960933.0,
      "step": 7595
    },
    {
      "entropy": 5.53103814125061,
      "epoch": 0.8535969000954681,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.4185,
      "mean_token_accuracy": 0.18351842612028121,
      "num_tokens": 16972284.0,
      "step": 7600
    },
    {
      "entropy": 5.60010724067688,
      "epoch": 0.8541584770034256,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.4401,
      "mean_token_accuracy": 0.18844986706972122,
      "num_tokens": 16983541.0,
      "step": 7605
    },
    {
      "entropy": 5.6820151805877686,
      "epoch": 0.8547200539113832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049377057282307,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17229961454868317,
      "num_tokens": 16993773.0,
      "step": 7610
    },
    {
      "entropy": 5.676660823822021,
      "epoch": 0.8552816308193407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493761187403909,
      "loss": 5.5623,
      "mean_token_accuracy": 0.17587521821260452,
      "num_tokens": 17005287.0,
      "step": 7615
    },
    {
      "entropy": 5.610708808898925,
      "epoch": 0.8558432077272983,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493751795019307,
      "loss": 5.4996,
      "mean_token_accuracy": 0.18012612909078599,
      "num_tokens": 17015664.0,
      "step": 7620
    },
    {
      "entropy": 5.535806179046631,
      "epoch": 0.8564047846352558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.4242,
      "mean_token_accuracy": 0.18311096727848053,
      "num_tokens": 17026876.0,
      "step": 7625
    },
    {
      "entropy": 5.506599283218383,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.4716,
      "mean_token_accuracy": 0.19143964052200318,
      "num_tokens": 17037321.0,
      "step": 7630
    },
    {
      "entropy": 5.6227470397949215,
      "epoch": 0.8575279384511709,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.5503,
      "mean_token_accuracy": 0.1803719401359558,
      "num_tokens": 17047506.0,
      "step": 7635
    },
    {
      "entropy": 5.7091453075408936,
      "epoch": 0.8580895153591285,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.5766,
      "mean_token_accuracy": 0.1779593601822853,
      "num_tokens": 17059710.0,
      "step": 7640
    },
    {
      "entropy": 5.681270933151245,
      "epoch": 0.858651092267086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.6052,
      "mean_token_accuracy": 0.1721954107284546,
      "num_tokens": 17070956.0,
      "step": 7645
    },
    {
      "entropy": 5.67075514793396,
      "epoch": 0.8592126691750436,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.569,
      "mean_token_accuracy": 0.17900737822055818,
      "num_tokens": 17081404.0,
      "step": 7650
    },
    {
      "entropy": 5.679018259048462,
      "epoch": 0.859774246083001,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.5798,
      "mean_token_accuracy": 0.17749113738536834,
      "num_tokens": 17092692.0,
      "step": 7655
    },
    {
      "entropy": 5.587037515640259,
      "epoch": 0.8603358229909586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000493676405222546,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18479957133531572,
      "num_tokens": 17103730.0,
      "step": 7660
    },
    {
      "entropy": 5.5954996109008786,
      "epoch": 0.8608973998989161,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.4967,
      "mean_token_accuracy": 0.17741685062646867,
      "num_tokens": 17115086.0,
      "step": 7665
    },
    {
      "entropy": 5.562705373764038,
      "epoch": 0.8614589768068737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.5177,
      "mean_token_accuracy": 0.17818861007690429,
      "num_tokens": 17125928.0,
      "step": 7670
    },
    {
      "entropy": 5.633713483810425,
      "epoch": 0.8620205537148312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.5344,
      "mean_token_accuracy": 0.17698321789503096,
      "num_tokens": 17137328.0,
      "step": 7675
    },
    {
      "entropy": 5.56837158203125,
      "epoch": 0.8625821306227888,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.5116,
      "mean_token_accuracy": 0.1792793244123459,
      "num_tokens": 17148274.0,
      "step": 7680
    },
    {
      "entropy": 5.567792510986328,
      "epoch": 0.8631437075307463,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.5524,
      "mean_token_accuracy": 0.1794579342007637,
      "num_tokens": 17158834.0,
      "step": 7685
    },
    {
      "entropy": 5.5691451072692875,
      "epoch": 0.8637052844387039,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.3966,
      "mean_token_accuracy": 0.18155379891395568,
      "num_tokens": 17169276.0,
      "step": 7690
    },
    {
      "entropy": 5.575228214263916,
      "epoch": 0.8642668613466614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.4476,
      "mean_token_accuracy": 0.18865410089492798,
      "num_tokens": 17180642.0,
      "step": 7695
    },
    {
      "entropy": 5.5229919910430905,
      "epoch": 0.864828438254619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004936005698104,
      "loss": 5.5157,
      "mean_token_accuracy": 0.1827980622649193,
      "num_tokens": 17192477.0,
      "step": 7700
    },
    {
      "entropy": 5.683761119842529,
      "epoch": 0.8653900151625765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.648,
      "mean_token_accuracy": 0.17949193716049194,
      "num_tokens": 17202653.0,
      "step": 7705
    },
    {
      "entropy": 5.522699785232544,
      "epoch": 0.8659515920705341,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.3877,
      "mean_token_accuracy": 0.18448692113161086,
      "num_tokens": 17212607.0,
      "step": 7710
    },
    {
      "entropy": 5.577323579788208,
      "epoch": 0.8665131689784916,
      "grad_norm": 0.984375,
      "learning_rate": 0.000493572016677141,
      "loss": 5.5312,
      "mean_token_accuracy": 0.18407864719629288,
      "num_tokens": 17223701.0,
      "step": 7715
    },
    {
      "entropy": 5.618215799331665,
      "epoch": 0.8670747458864492,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.4317,
      "mean_token_accuracy": 0.1864711031317711,
      "num_tokens": 17234878.0,
      "step": 7720
    },
    {
      "entropy": 5.507683944702149,
      "epoch": 0.8676363227944067,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.5017,
      "mean_token_accuracy": 0.1830043986439705,
      "num_tokens": 17245813.0,
      "step": 7725
    },
    {
      "entropy": 5.595293378829956,
      "epoch": 0.8681978997023643,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.5815,
      "mean_token_accuracy": 0.18132366389036178,
      "num_tokens": 17257378.0,
      "step": 7730
    },
    {
      "entropy": 5.655421590805053,
      "epoch": 0.8687594766103218,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.5168,
      "mean_token_accuracy": 0.1777476266026497,
      "num_tokens": 17269108.0,
      "step": 7735
    },
    {
      "entropy": 5.585360717773438,
      "epoch": 0.8693210535182794,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.4943,
      "mean_token_accuracy": 0.17545924484729766,
      "num_tokens": 17280643.0,
      "step": 7740
    },
    {
      "entropy": 5.690800285339355,
      "epoch": 0.8698826304262369,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.5603,
      "mean_token_accuracy": 0.17998234629631044,
      "num_tokens": 17291236.0,
      "step": 7745
    },
    {
      "entropy": 5.7303039073944095,
      "epoch": 0.8704442073341944,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.6305,
      "mean_token_accuracy": 0.1695631831884384,
      "num_tokens": 17303547.0,
      "step": 7750
    },
    {
      "entropy": 5.726223516464233,
      "epoch": 0.8710057842421519,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.6027,
      "mean_token_accuracy": 0.17243569046258928,
      "num_tokens": 17314803.0,
      "step": 7755
    },
    {
      "entropy": 5.568598127365112,
      "epoch": 0.8715673611501095,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.4792,
      "mean_token_accuracy": 0.1801397055387497,
      "num_tokens": 17326009.0,
      "step": 7760
    },
    {
      "entropy": 5.635626983642578,
      "epoch": 0.872128938058067,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.4904,
      "mean_token_accuracy": 0.18378062695264816,
      "num_tokens": 17337078.0,
      "step": 7765
    },
    {
      "entropy": 5.6486553192138675,
      "epoch": 0.8726905149660246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.6029,
      "mean_token_accuracy": 0.17538124173879624,
      "num_tokens": 17349266.0,
      "step": 7770
    },
    {
      "entropy": 5.464595460891724,
      "epoch": 0.8732520918739821,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.4034,
      "mean_token_accuracy": 0.18144616186618806,
      "num_tokens": 17360327.0,
      "step": 7775
    },
    {
      "entropy": 5.679508781433105,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17058178782463074,
      "num_tokens": 17372727.0,
      "step": 7780
    },
    {
      "entropy": 5.712011480331421,
      "epoch": 0.8743752456898972,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.6117,
      "mean_token_accuracy": 0.17371562570333482,
      "num_tokens": 17384160.0,
      "step": 7785
    },
    {
      "entropy": 5.605370426177979,
      "epoch": 0.8749368225978548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.5269,
      "mean_token_accuracy": 0.17663144171237946,
      "num_tokens": 17396009.0,
      "step": 7790
    },
    {
      "entropy": 5.570238256454468,
      "epoch": 0.8754983995058123,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.5287,
      "mean_token_accuracy": 0.18028004467487335,
      "num_tokens": 17406903.0,
      "step": 7795
    },
    {
      "entropy": 5.586734771728516,
      "epoch": 0.8760599764137699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.3941,
      "mean_token_accuracy": 0.18741592317819594,
      "num_tokens": 17417694.0,
      "step": 7800
    },
    {
      "entropy": 5.5621052265167235,
      "epoch": 0.8766215533217274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.6368,
      "mean_token_accuracy": 0.1808529257774353,
      "num_tokens": 17428731.0,
      "step": 7805
    },
    {
      "entropy": 5.559157037734986,
      "epoch": 0.877183130229685,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.3837,
      "mean_token_accuracy": 0.18207520693540574,
      "num_tokens": 17439242.0,
      "step": 7810
    },
    {
      "entropy": 5.6173442840576175,
      "epoch": 0.8777447071376425,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.4863,
      "mean_token_accuracy": 0.17750116288661957,
      "num_tokens": 17449980.0,
      "step": 7815
    },
    {
      "entropy": 5.623306751251221,
      "epoch": 0.8783062840456001,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.592,
      "mean_token_accuracy": 0.16987940967082976,
      "num_tokens": 17464140.0,
      "step": 7820
    },
    {
      "entropy": 5.725029325485229,
      "epoch": 0.8788678609535576,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.5997,
      "mean_token_accuracy": 0.17623118460178375,
      "num_tokens": 17476246.0,
      "step": 7825
    },
    {
      "entropy": 5.660158109664917,
      "epoch": 0.8794294378615152,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.6103,
      "mean_token_accuracy": 0.16733368188142778,
      "num_tokens": 17488169.0,
      "step": 7830
    },
    {
      "entropy": 5.5224262237548825,
      "epoch": 0.8799910147694727,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.4298,
      "mean_token_accuracy": 0.18274897038936616,
      "num_tokens": 17498465.0,
      "step": 7835
    },
    {
      "entropy": 5.593181943893432,
      "epoch": 0.8805525916774303,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.4271,
      "mean_token_accuracy": 0.18860915303230286,
      "num_tokens": 17509217.0,
      "step": 7840
    },
    {
      "entropy": 5.572647142410278,
      "epoch": 0.8811141685853877,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.4612,
      "mean_token_accuracy": 0.18687058687210084,
      "num_tokens": 17519851.0,
      "step": 7845
    },
    {
      "entropy": 5.569455528259278,
      "epoch": 0.8816757454933453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.5286,
      "mean_token_accuracy": 0.1822614923119545,
      "num_tokens": 17532144.0,
      "step": 7850
    },
    {
      "entropy": 5.6123206615448,
      "epoch": 0.8822373224013028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18794535398483275,
      "num_tokens": 17542929.0,
      "step": 7855
    },
    {
      "entropy": 5.6424397945404055,
      "epoch": 0.8827988993092604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.6132,
      "mean_token_accuracy": 0.1730242982506752,
      "num_tokens": 17554931.0,
      "step": 7860
    },
    {
      "entropy": 5.524155998229981,
      "epoch": 0.8833604762172179,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18392786830663682,
      "num_tokens": 17566026.0,
      "step": 7865
    },
    {
      "entropy": 5.602767896652222,
      "epoch": 0.8839220531251755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049327330152104,
      "loss": 5.525,
      "mean_token_accuracy": 0.18074765652418137,
      "num_tokens": 17578396.0,
      "step": 7870
    },
    {
      "entropy": 5.66438627243042,
      "epoch": 0.884483630033133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.5199,
      "mean_token_accuracy": 0.182549124956131,
      "num_tokens": 17590559.0,
      "step": 7875
    },
    {
      "entropy": 5.553819227218628,
      "epoch": 0.8850452069410906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.4882,
      "mean_token_accuracy": 0.17640931904315948,
      "num_tokens": 17600833.0,
      "step": 7880
    },
    {
      "entropy": 5.5474964618682865,
      "epoch": 0.8856067838490481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.4701,
      "mean_token_accuracy": 0.18419776856899261,
      "num_tokens": 17612166.0,
      "step": 7885
    },
    {
      "entropy": 5.653920030593872,
      "epoch": 0.8861683607570057,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.5832,
      "mean_token_accuracy": 0.17716454714536667,
      "num_tokens": 17623070.0,
      "step": 7890
    },
    {
      "entropy": 5.673707628250122,
      "epoch": 0.8867299376649632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493224495924948,
      "loss": 5.6412,
      "mean_token_accuracy": 0.17596878111362457,
      "num_tokens": 17633319.0,
      "step": 7895
    },
    {
      "entropy": 5.601524114608765,
      "epoch": 0.8872915145729208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.4527,
      "mean_token_accuracy": 0.18103813529014587,
      "num_tokens": 17643818.0,
      "step": 7900
    },
    {
      "entropy": 5.580865526199341,
      "epoch": 0.8878530914808783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.4478,
      "mean_token_accuracy": 0.18300647139549256,
      "num_tokens": 17655665.0,
      "step": 7905
    },
    {
      "entropy": 5.56022891998291,
      "epoch": 0.8884146683888359,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.3737,
      "mean_token_accuracy": 0.1908242642879486,
      "num_tokens": 17666200.0,
      "step": 7910
    },
    {
      "entropy": 5.550500535964966,
      "epoch": 0.8889762452967934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.5856,
      "mean_token_accuracy": 0.17779973596334459,
      "num_tokens": 17677299.0,
      "step": 7915
    },
    {
      "entropy": 5.598557758331299,
      "epoch": 0.889537822204751,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.4302,
      "mean_token_accuracy": 0.1908480182290077,
      "num_tokens": 17687870.0,
      "step": 7920
    },
    {
      "entropy": 5.579096794128418,
      "epoch": 0.8900993991127085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.4567,
      "mean_token_accuracy": 0.18092368841171264,
      "num_tokens": 17698853.0,
      "step": 7925
    },
    {
      "entropy": 5.567067193984985,
      "epoch": 0.8906609760206661,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.5095,
      "mean_token_accuracy": 0.18371810168027877,
      "num_tokens": 17709897.0,
      "step": 7930
    },
    {
      "entropy": 5.595690393447876,
      "epoch": 0.8912225529286236,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1784860759973526,
      "num_tokens": 17720974.0,
      "step": 7935
    },
    {
      "entropy": 5.736320924758911,
      "epoch": 0.8917841298365812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.6477,
      "mean_token_accuracy": 0.17302757054567336,
      "num_tokens": 17734449.0,
      "step": 7940
    },
    {
      "entropy": 5.485985565185547,
      "epoch": 0.8923457067445386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.3903,
      "mean_token_accuracy": 0.18473621755838393,
      "num_tokens": 17744735.0,
      "step": 7945
    },
    {
      "entropy": 5.533071231842041,
      "epoch": 0.8929072836524962,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18630819767713547,
      "num_tokens": 17755641.0,
      "step": 7950
    },
    {
      "entropy": 5.641396427154541,
      "epoch": 0.8934688605604537,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.5324,
      "mean_token_accuracy": 0.17234542667865754,
      "num_tokens": 17767621.0,
      "step": 7955
    },
    {
      "entropy": 5.680049991607666,
      "epoch": 0.8940304374684113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.6277,
      "mean_token_accuracy": 0.1737061321735382,
      "num_tokens": 17779507.0,
      "step": 7960
    },
    {
      "entropy": 5.543487548828125,
      "epoch": 0.8945920143763688,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.4693,
      "mean_token_accuracy": 0.17781761288642883,
      "num_tokens": 17790371.0,
      "step": 7965
    },
    {
      "entropy": 5.603549337387085,
      "epoch": 0.8951535912843264,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.4923,
      "mean_token_accuracy": 0.1885839194059372,
      "num_tokens": 17800903.0,
      "step": 7970
    },
    {
      "entropy": 5.712726497650147,
      "epoch": 0.8957151681922839,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.6715,
      "mean_token_accuracy": 0.16963825598359109,
      "num_tokens": 17812053.0,
      "step": 7975
    },
    {
      "entropy": 5.577488851547241,
      "epoch": 0.8962767451002415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.4038,
      "mean_token_accuracy": 0.18471201211214067,
      "num_tokens": 17822113.0,
      "step": 7980
    },
    {
      "entropy": 5.536539506912232,
      "epoch": 0.896838322008199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.4442,
      "mean_token_accuracy": 0.18067950308322905,
      "num_tokens": 17832741.0,
      "step": 7985
    },
    {
      "entropy": 5.445788621902466,
      "epoch": 0.8973998989161566,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.3738,
      "mean_token_accuracy": 0.17739315032958985,
      "num_tokens": 17842968.0,
      "step": 7990
    },
    {
      "entropy": 5.68609652519226,
      "epoch": 0.8979614758241141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.6441,
      "mean_token_accuracy": 0.167500863969326,
      "num_tokens": 17854295.0,
      "step": 7995
    },
    {
      "entropy": 5.724550437927246,
      "epoch": 0.8985230527320717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.565,
      "mean_token_accuracy": 0.18175942301750184,
      "num_tokens": 17865597.0,
      "step": 8000
    },
    {
      "entropy": 5.452319383621216,
      "epoch": 0.8990846296400292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.342,
      "mean_token_accuracy": 0.18693463057279586,
      "num_tokens": 17877405.0,
      "step": 8005
    },
    {
      "entropy": 5.69283709526062,
      "epoch": 0.8996462065479868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17947353273630143,
      "num_tokens": 17889273.0,
      "step": 8010
    },
    {
      "entropy": 5.5469183921813965,
      "epoch": 0.9002077834559443,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.4435,
      "mean_token_accuracy": 0.1848416194319725,
      "num_tokens": 17899761.0,
      "step": 8015
    },
    {
      "entropy": 5.553060245513916,
      "epoch": 0.9007693603639019,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.4433,
      "mean_token_accuracy": 0.18374273478984832,
      "num_tokens": 17910607.0,
      "step": 8020
    },
    {
      "entropy": 5.6265302181243895,
      "epoch": 0.9013309372718594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.5055,
      "mean_token_accuracy": 0.17801229804754257,
      "num_tokens": 17922538.0,
      "step": 8025
    },
    {
      "entropy": 5.638562345504761,
      "epoch": 0.901892514179817,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.5479,
      "mean_token_accuracy": 0.1817954033613205,
      "num_tokens": 17933313.0,
      "step": 8030
    },
    {
      "entropy": 5.673544836044312,
      "epoch": 0.9024540910877745,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.632,
      "mean_token_accuracy": 0.1749018445611,
      "num_tokens": 17944893.0,
      "step": 8035
    },
    {
      "entropy": 5.606023740768433,
      "epoch": 0.903015667995732,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.5149,
      "mean_token_accuracy": 0.1786926954984665,
      "num_tokens": 17955019.0,
      "step": 8040
    },
    {
      "entropy": 5.6634604930877686,
      "epoch": 0.9035772449036895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.5422,
      "mean_token_accuracy": 0.18044948875904082,
      "num_tokens": 17965793.0,
      "step": 8045
    },
    {
      "entropy": 5.551074838638305,
      "epoch": 0.9041388218116471,
      "grad_norm": 1.125,
      "learning_rate": 0.000492918025562663,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18530188649892806,
      "num_tokens": 17975709.0,
      "step": 8050
    },
    {
      "entropy": 5.63584451675415,
      "epoch": 0.9047003987196046,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.5954,
      "mean_token_accuracy": 0.17713356465101243,
      "num_tokens": 17985405.0,
      "step": 8055
    },
    {
      "entropy": 5.69279899597168,
      "epoch": 0.9052619756275622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.6199,
      "mean_token_accuracy": 0.17835464477539062,
      "num_tokens": 17997105.0,
      "step": 8060
    },
    {
      "entropy": 5.630202341079712,
      "epoch": 0.9058235525355197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.5167,
      "mean_token_accuracy": 0.18688323348760605,
      "num_tokens": 18007914.0,
      "step": 8065
    },
    {
      "entropy": 5.534109878540039,
      "epoch": 0.9063851294434773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.5166,
      "mean_token_accuracy": 0.18196988403797149,
      "num_tokens": 18018328.0,
      "step": 8070
    },
    {
      "entropy": 5.592997598648071,
      "epoch": 0.9069467063514348,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492867970138744,
      "loss": 5.4732,
      "mean_token_accuracy": 0.18148116916418075,
      "num_tokens": 18028568.0,
      "step": 8075
    },
    {
      "entropy": 5.627342462539673,
      "epoch": 0.9075082832593924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.4917,
      "mean_token_accuracy": 0.18100592792034148,
      "num_tokens": 18038655.0,
      "step": 8080
    },
    {
      "entropy": 5.5573585510253904,
      "epoch": 0.9080698601673499,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492847899407939,
      "loss": 5.4214,
      "mean_token_accuracy": 0.18531959056854247,
      "num_tokens": 18049716.0,
      "step": 8085
    },
    {
      "entropy": 5.418399143218994,
      "epoch": 0.9086314370753075,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18372095227241517,
      "num_tokens": 18062530.0,
      "step": 8090
    },
    {
      "entropy": 5.559502267837525,
      "epoch": 0.909193013983265,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.508,
      "mean_token_accuracy": 0.17610028833150865,
      "num_tokens": 18073076.0,
      "step": 8095
    },
    {
      "entropy": 5.695038843154907,
      "epoch": 0.9097545908912226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.4742,
      "mean_token_accuracy": 0.17959958761930467,
      "num_tokens": 18083759.0,
      "step": 8100
    },
    {
      "entropy": 5.470184278488159,
      "epoch": 0.9103161677991801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.4145,
      "mean_token_accuracy": 0.1759122133255005,
      "num_tokens": 18095980.0,
      "step": 8105
    },
    {
      "entropy": 5.582666730880737,
      "epoch": 0.9108777447071377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.5138,
      "mean_token_accuracy": 0.18386195302009584,
      "num_tokens": 18106672.0,
      "step": 8110
    },
    {
      "entropy": 5.564055871963501,
      "epoch": 0.9114393216150952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1788438469171524,
      "num_tokens": 18117481.0,
      "step": 8115
    },
    {
      "entropy": 5.546481370925903,
      "epoch": 0.9120008985230528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18264017403125762,
      "num_tokens": 18127602.0,
      "step": 8120
    },
    {
      "entropy": 5.665448760986328,
      "epoch": 0.9125624754310103,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.5225,
      "mean_token_accuracy": 0.17809127867221833,
      "num_tokens": 18140109.0,
      "step": 8125
    },
    {
      "entropy": 5.551574611663819,
      "epoch": 0.9131240523389679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.4111,
      "mean_token_accuracy": 0.18579337745904922,
      "num_tokens": 18150271.0,
      "step": 8130
    },
    {
      "entropy": 5.551731109619141,
      "epoch": 0.9136856292469253,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.5347,
      "mean_token_accuracy": 0.18156878054141998,
      "num_tokens": 18160746.0,
      "step": 8135
    },
    {
      "entropy": 5.681353235244751,
      "epoch": 0.9142472061548829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.5083,
      "mean_token_accuracy": 0.1790511950850487,
      "num_tokens": 18172271.0,
      "step": 8140
    },
    {
      "entropy": 5.579694557189941,
      "epoch": 0.9148087830628404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.4419,
      "mean_token_accuracy": 0.18102690130472182,
      "num_tokens": 18183116.0,
      "step": 8145
    },
    {
      "entropy": 5.51436710357666,
      "epoch": 0.915370359970798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.4092,
      "mean_token_accuracy": 0.19384038895368577,
      "num_tokens": 18195147.0,
      "step": 8150
    },
    {
      "entropy": 5.570596933364868,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.21875,
      "learning_rate": 0.000492706627563404,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1778377816081047,
      "num_tokens": 18206217.0,
      "step": 8155
    },
    {
      "entropy": 5.550006151199341,
      "epoch": 0.9164935137867131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.4855,
      "mean_token_accuracy": 0.1773753508925438,
      "num_tokens": 18217434.0,
      "step": 8160
    },
    {
      "entropy": 5.608560085296631,
      "epoch": 0.9170550906946706,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.4922,
      "mean_token_accuracy": 0.17910895496606827,
      "num_tokens": 18229277.0,
      "step": 8165
    },
    {
      "entropy": 5.526487493515015,
      "epoch": 0.9176166676026282,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18623987585306168,
      "num_tokens": 18240320.0,
      "step": 8170
    },
    {
      "entropy": 5.463681316375732,
      "epoch": 0.9181782445105857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.3871,
      "mean_token_accuracy": 0.18620867878198624,
      "num_tokens": 18250380.0,
      "step": 8175
    },
    {
      "entropy": 5.583861589431763,
      "epoch": 0.9187398214185433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.5694,
      "mean_token_accuracy": 0.18073032200336456,
      "num_tokens": 18261703.0,
      "step": 8180
    },
    {
      "entropy": 5.621598863601685,
      "epoch": 0.9193013983265008,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.496,
      "mean_token_accuracy": 0.1756924122571945,
      "num_tokens": 18272472.0,
      "step": 8185
    },
    {
      "entropy": 5.674906826019287,
      "epoch": 0.9198629752344584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.5723,
      "mean_token_accuracy": 0.1792101725935936,
      "num_tokens": 18282168.0,
      "step": 8190
    },
    {
      "entropy": 5.687933826446534,
      "epoch": 0.9204245521424159,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.5731,
      "mean_token_accuracy": 0.17163886725902558,
      "num_tokens": 18293750.0,
      "step": 8195
    },
    {
      "entropy": 5.5886390686035154,
      "epoch": 0.9209861290503735,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.5,
      "mean_token_accuracy": 0.18049391061067582,
      "num_tokens": 18304568.0,
      "step": 8200
    },
    {
      "entropy": 5.562101364135742,
      "epoch": 0.921547705958331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.4241,
      "mean_token_accuracy": 0.1870491161942482,
      "num_tokens": 18315129.0,
      "step": 8205
    },
    {
      "entropy": 5.434761047363281,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492594675081777,
      "loss": 5.3787,
      "mean_token_accuracy": 0.18584675639867781,
      "num_tokens": 18324628.0,
      "step": 8210
    },
    {
      "entropy": 5.466102457046508,
      "epoch": 0.9226708597742461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492584456007161,
      "loss": 5.4009,
      "mean_token_accuracy": 0.18903011679649354,
      "num_tokens": 18334899.0,
      "step": 8215
    },
    {
      "entropy": 5.617522048950195,
      "epoch": 0.9232324366822037,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.4078,
      "mean_token_accuracy": 0.1870238482952118,
      "num_tokens": 18345719.0,
      "step": 8220
    },
    {
      "entropy": 5.561037397384643,
      "epoch": 0.9237940135901612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.4906,
      "mean_token_accuracy": 0.18776837289333342,
      "num_tokens": 18355814.0,
      "step": 8225
    },
    {
      "entropy": 5.588109254837036,
      "epoch": 0.9243555904981187,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.5545,
      "mean_token_accuracy": 0.17607142478227616,
      "num_tokens": 18366746.0,
      "step": 8230
    },
    {
      "entropy": 5.51508412361145,
      "epoch": 0.9249171674060762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.3827,
      "mean_token_accuracy": 0.1891380324959755,
      "num_tokens": 18375846.0,
      "step": 8235
    },
    {
      "entropy": 5.514292907714844,
      "epoch": 0.9254787443140338,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000492533256720994,
      "loss": 5.5051,
      "mean_token_accuracy": 0.1810300201177597,
      "num_tokens": 18386850.0,
      "step": 8240
    },
    {
      "entropy": 5.538437843322754,
      "epoch": 0.9260403212219913,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18939542025327682,
      "num_tokens": 18396347.0,
      "step": 8245
    },
    {
      "entropy": 5.567156553268433,
      "epoch": 0.9266018981299489,
      "grad_norm": 1.125,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.4436,
      "mean_token_accuracy": 0.18140185326337815,
      "num_tokens": 18406583.0,
      "step": 8250
    },
    {
      "entropy": 5.679202651977539,
      "epoch": 0.9271634750379064,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492502454030566,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17679764479398727,
      "num_tokens": 18418586.0,
      "step": 8255
    },
    {
      "entropy": 5.5931939601898195,
      "epoch": 0.927725051945864,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.4015,
      "mean_token_accuracy": 0.18384822010993956,
      "num_tokens": 18429752.0,
      "step": 8260
    },
    {
      "entropy": 5.575476551055909,
      "epoch": 0.9282866288538215,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.4969,
      "mean_token_accuracy": 0.1765847384929657,
      "num_tokens": 18440663.0,
      "step": 8265
    },
    {
      "entropy": 5.456879949569702,
      "epoch": 0.9288482057617791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.4371,
      "mean_token_accuracy": 0.18375977724790574,
      "num_tokens": 18451497.0,
      "step": 8270
    },
    {
      "entropy": 5.543888425827026,
      "epoch": 0.9294097826697366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.431,
      "mean_token_accuracy": 0.17825741618871688,
      "num_tokens": 18462794.0,
      "step": 8275
    },
    {
      "entropy": 5.6037522315979,
      "epoch": 0.9299713595776942,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.4277,
      "mean_token_accuracy": 0.18658117055892945,
      "num_tokens": 18472973.0,
      "step": 8280
    },
    {
      "entropy": 5.635124921798706,
      "epoch": 0.9305329364856517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.5903,
      "mean_token_accuracy": 0.1796143800020218,
      "num_tokens": 18484451.0,
      "step": 8285
    },
    {
      "entropy": 5.611140489578247,
      "epoch": 0.9310945133936093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.4427,
      "mean_token_accuracy": 0.1901862293481827,
      "num_tokens": 18496412.0,
      "step": 8290
    },
    {
      "entropy": 5.556115007400512,
      "epoch": 0.9316560903015668,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.4706,
      "mean_token_accuracy": 0.1861564189195633,
      "num_tokens": 18508547.0,
      "step": 8295
    },
    {
      "entropy": 5.514981889724732,
      "epoch": 0.9322176672095244,
      "grad_norm": 1.09375,
      "learning_rate": 0.000492409672025112,
      "loss": 5.513,
      "mean_token_accuracy": 0.175574454665184,
      "num_tokens": 18518954.0,
      "step": 8300
    },
    {
      "entropy": 5.528355646133423,
      "epoch": 0.9327792441174819,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18600136190652847,
      "num_tokens": 18530444.0,
      "step": 8305
    },
    {
      "entropy": 5.545026731491089,
      "epoch": 0.9333408210254395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.4264,
      "mean_token_accuracy": 0.18126846700906754,
      "num_tokens": 18541612.0,
      "step": 8310
    },
    {
      "entropy": 5.616086959838867,
      "epoch": 0.933902397933397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.5977,
      "mean_token_accuracy": 0.17694150507450104,
      "num_tokens": 18553316.0,
      "step": 8315
    },
    {
      "entropy": 5.60566201210022,
      "epoch": 0.9344639748413546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049236825558139,
      "loss": 5.4627,
      "mean_token_accuracy": 0.18421150743961334,
      "num_tokens": 18564468.0,
      "step": 8320
    },
    {
      "entropy": 5.501161336898804,
      "epoch": 0.9350255517493121,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.3976,
      "mean_token_accuracy": 0.18672065287828446,
      "num_tokens": 18575536.0,
      "step": 8325
    },
    {
      "entropy": 5.5153326988220215,
      "epoch": 0.9355871286572696,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.5212,
      "mean_token_accuracy": 0.1734934478998184,
      "num_tokens": 18586906.0,
      "step": 8330
    },
    {
      "entropy": 5.623011255264283,
      "epoch": 0.9361487055652271,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.503,
      "mean_token_accuracy": 0.18519700616598128,
      "num_tokens": 18599064.0,
      "step": 8335
    },
    {
      "entropy": 5.60233063697815,
      "epoch": 0.9367102824731847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.5367,
      "mean_token_accuracy": 0.18010803312063217,
      "num_tokens": 18610057.0,
      "step": 8340
    },
    {
      "entropy": 5.434579849243164,
      "epoch": 0.9372718593811422,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.3641,
      "mean_token_accuracy": 0.18760113418102264,
      "num_tokens": 18620008.0,
      "step": 8345
    },
    {
      "entropy": 5.479099082946777,
      "epoch": 0.9378334362890998,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18557711094617843,
      "num_tokens": 18630599.0,
      "step": 8350
    },
    {
      "entropy": 5.573970603942871,
      "epoch": 0.9383950131970573,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.5379,
      "mean_token_accuracy": 0.1812148243188858,
      "num_tokens": 18643898.0,
      "step": 8355
    },
    {
      "entropy": 5.660813236236573,
      "epoch": 0.9389565901050149,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000492285090502816,
      "loss": 5.5152,
      "mean_token_accuracy": 0.17586288154125213,
      "num_tokens": 18654760.0,
      "step": 8360
    },
    {
      "entropy": 5.6697930812835695,
      "epoch": 0.9395181670129724,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.49,
      "mean_token_accuracy": 0.18287123441696168,
      "num_tokens": 18666036.0,
      "step": 8365
    },
    {
      "entropy": 5.622818899154663,
      "epoch": 0.94007974392093,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.5341,
      "mean_token_accuracy": 0.18295766413211823,
      "num_tokens": 18677806.0,
      "step": 8370
    },
    {
      "entropy": 5.612994861602783,
      "epoch": 0.9406413208288875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492253789434953,
      "loss": 5.6516,
      "mean_token_accuracy": 0.1783110097050667,
      "num_tokens": 18688971.0,
      "step": 8375
    },
    {
      "entropy": 5.589858484268189,
      "epoch": 0.9412028977368451,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.5247,
      "mean_token_accuracy": 0.17627423405647277,
      "num_tokens": 18700153.0,
      "step": 8380
    },
    {
      "entropy": 5.715108442306518,
      "epoch": 0.9417644746448026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.5649,
      "mean_token_accuracy": 0.179406775534153,
      "num_tokens": 18711935.0,
      "step": 8385
    },
    {
      "entropy": 5.487036943435669,
      "epoch": 0.9423260515527602,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18586247116327287,
      "num_tokens": 18722694.0,
      "step": 8390
    },
    {
      "entropy": 5.586714458465576,
      "epoch": 0.9428876284607177,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.5388,
      "mean_token_accuracy": 0.18508632779121398,
      "num_tokens": 18733443.0,
      "step": 8395
    },
    {
      "entropy": 5.6142387866973875,
      "epoch": 0.9434492053686753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.4596,
      "mean_token_accuracy": 0.18226936012506484,
      "num_tokens": 18743965.0,
      "step": 8400
    },
    {
      "entropy": 5.560277986526489,
      "epoch": 0.9440107822766328,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.5514,
      "mean_token_accuracy": 0.17946840226650237,
      "num_tokens": 18755712.0,
      "step": 8405
    },
    {
      "entropy": 5.579993200302124,
      "epoch": 0.9445723591845904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18157842755317688,
      "num_tokens": 18766756.0,
      "step": 8410
    },
    {
      "entropy": 5.581419563293457,
      "epoch": 0.9451339360925479,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17493135780096053,
      "num_tokens": 18776932.0,
      "step": 8415
    },
    {
      "entropy": 5.607511425018311,
      "epoch": 0.9456955130005055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.4681,
      "mean_token_accuracy": 0.18051346987485886,
      "num_tokens": 18787197.0,
      "step": 8420
    },
    {
      "entropy": 5.512584781646728,
      "epoch": 0.9462570899084629,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.4303,
      "mean_token_accuracy": 0.18762568533420562,
      "num_tokens": 18797861.0,
      "step": 8425
    },
    {
      "entropy": 5.535125875473023,
      "epoch": 0.9468186668164205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.4368,
      "mean_token_accuracy": 0.18415761291980742,
      "num_tokens": 18809111.0,
      "step": 8430
    },
    {
      "entropy": 5.6051353931427,
      "epoch": 0.947380243724378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.4542,
      "mean_token_accuracy": 0.1931390032172203,
      "num_tokens": 18820236.0,
      "step": 8435
    },
    {
      "entropy": 5.523641204833984,
      "epoch": 0.9479418206323356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.4247,
      "mean_token_accuracy": 0.18735523670911788,
      "num_tokens": 18831942.0,
      "step": 8440
    },
    {
      "entropy": 5.559072494506836,
      "epoch": 0.9485033975402931,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18535945266485215,
      "num_tokens": 18843271.0,
      "step": 8445
    },
    {
      "entropy": 5.581586647033691,
      "epoch": 0.9490649744482507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.5181,
      "mean_token_accuracy": 0.18354747593402862,
      "num_tokens": 18856285.0,
      "step": 8450
    },
    {
      "entropy": 5.667445230484009,
      "epoch": 0.9496265513562082,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00049208579917436,
      "loss": 5.5354,
      "mean_token_accuracy": 0.17863369584083558,
      "num_tokens": 18867203.0,
      "step": 8455
    },
    {
      "entropy": 5.6186872005462645,
      "epoch": 0.9501881282641658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17248342782258988,
      "num_tokens": 18878598.0,
      "step": 8460
    },
    {
      "entropy": 5.656963253021241,
      "epoch": 0.9507497051721233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.486,
      "mean_token_accuracy": 0.18291398584842683,
      "num_tokens": 18890233.0,
      "step": 8465
    },
    {
      "entropy": 5.614354228973388,
      "epoch": 0.9513112820800809,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.5555,
      "mean_token_accuracy": 0.17716809660196303,
      "num_tokens": 18900362.0,
      "step": 8470
    },
    {
      "entropy": 5.601590394973755,
      "epoch": 0.9518728589880384,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.5011,
      "mean_token_accuracy": 0.18710409849882126,
      "num_tokens": 18911924.0,
      "step": 8475
    },
    {
      "entropy": 5.6120593547821045,
      "epoch": 0.952434435895996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.5114,
      "mean_token_accuracy": 0.1798287019133568,
      "num_tokens": 18923931.0,
      "step": 8480
    },
    {
      "entropy": 5.569701528549194,
      "epoch": 0.9529960128039535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.4259,
      "mean_token_accuracy": 0.18338057398796082,
      "num_tokens": 18935859.0,
      "step": 8485
    },
    {
      "entropy": 5.604145479202271,
      "epoch": 0.9535575897119111,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.5531,
      "mean_token_accuracy": 0.17262073755264282,
      "num_tokens": 18946939.0,
      "step": 8490
    },
    {
      "entropy": 5.617562627792358,
      "epoch": 0.9541191666198686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.5297,
      "mean_token_accuracy": 0.17581794559955596,
      "num_tokens": 18958067.0,
      "step": 8495
    },
    {
      "entropy": 5.567305421829223,
      "epoch": 0.9546807435278262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.4349,
      "mean_token_accuracy": 0.18979771435260773,
      "num_tokens": 18967920.0,
      "step": 8500
    },
    {
      "entropy": 5.5682543277740475,
      "epoch": 0.9552423204357837,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491979906602092,
      "loss": 5.4654,
      "mean_token_accuracy": 0.18660126626491547,
      "num_tokens": 18978244.0,
      "step": 8505
    },
    {
      "entropy": 5.584858179092407,
      "epoch": 0.9558038973437413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18310647010803222,
      "num_tokens": 18988464.0,
      "step": 8510
    },
    {
      "entropy": 5.487370204925537,
      "epoch": 0.9563654742516988,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.3872,
      "mean_token_accuracy": 0.1830912634730339,
      "num_tokens": 18998318.0,
      "step": 8515
    },
    {
      "entropy": 5.42850890159607,
      "epoch": 0.9569270511596563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.3693,
      "mean_token_accuracy": 0.187215156853199,
      "num_tokens": 19009802.0,
      "step": 8520
    },
    {
      "entropy": 5.691130685806274,
      "epoch": 0.9574886280676138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.5971,
      "mean_token_accuracy": 0.1797230139374733,
      "num_tokens": 19020174.0,
      "step": 8525
    },
    {
      "entropy": 5.621046876907348,
      "epoch": 0.9580502049755714,
      "grad_norm": 1.0,
      "learning_rate": 0.000491926701217703,
      "loss": 5.5008,
      "mean_token_accuracy": 0.17844899594783784,
      "num_tokens": 19031505.0,
      "step": 8530
    },
    {
      "entropy": 5.537479114532471,
      "epoch": 0.9586117818835289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.4162,
      "mean_token_accuracy": 0.18335415422916412,
      "num_tokens": 19042052.0,
      "step": 8535
    },
    {
      "entropy": 5.73462405204773,
      "epoch": 0.9591733587914865,
      "grad_norm": 1.046875,
      "learning_rate": 0.000491905370712362,
      "loss": 5.6042,
      "mean_token_accuracy": 0.17487475723028184,
      "num_tokens": 19052050.0,
      "step": 8540
    },
    {
      "entropy": 5.642431116104126,
      "epoch": 0.959734935699444,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.5188,
      "mean_token_accuracy": 0.17947326004505157,
      "num_tokens": 19062053.0,
      "step": 8545
    },
    {
      "entropy": 5.530859279632568,
      "epoch": 0.9602965126074016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.4788,
      "mean_token_accuracy": 0.18358104079961776,
      "num_tokens": 19073095.0,
      "step": 8550
    },
    {
      "entropy": 5.615685939788818,
      "epoch": 0.9608580895153591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.5361,
      "mean_token_accuracy": 0.18429580926895142,
      "num_tokens": 19084521.0,
      "step": 8555
    },
    {
      "entropy": 5.64449553489685,
      "epoch": 0.9614196664233167,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.5147,
      "mean_token_accuracy": 0.18146532624959946,
      "num_tokens": 19096372.0,
      "step": 8560
    },
    {
      "entropy": 5.6696178913116455,
      "epoch": 0.9619812433312742,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.6342,
      "mean_token_accuracy": 0.1733379617333412,
      "num_tokens": 19108535.0,
      "step": 8565
    },
    {
      "entropy": 5.70990104675293,
      "epoch": 0.9625428202392318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.4846,
      "mean_token_accuracy": 0.18403034955263137,
      "num_tokens": 19119480.0,
      "step": 8570
    },
    {
      "entropy": 5.588913059234619,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.49,
      "mean_token_accuracy": 0.18181291073560715,
      "num_tokens": 19131168.0,
      "step": 8575
    },
    {
      "entropy": 5.604631042480468,
      "epoch": 0.9636659740551469,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.621,
      "mean_token_accuracy": 0.17726342529058456,
      "num_tokens": 19142454.0,
      "step": 8580
    },
    {
      "entropy": 5.597691631317138,
      "epoch": 0.9642275509631044,
      "grad_norm": 0.921875,
      "learning_rate": 0.000491809041613254,
      "loss": 5.4787,
      "mean_token_accuracy": 0.18039288967847825,
      "num_tokens": 19154682.0,
      "step": 8585
    },
    {
      "entropy": 5.572999572753906,
      "epoch": 0.964789127871062,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.4083,
      "mean_token_accuracy": 0.17833466827869415,
      "num_tokens": 19165190.0,
      "step": 8590
    },
    {
      "entropy": 5.609723138809204,
      "epoch": 0.9653507047790195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.5429,
      "mean_token_accuracy": 0.18797778934240342,
      "num_tokens": 19176075.0,
      "step": 8595
    },
    {
      "entropy": 5.6036262035369875,
      "epoch": 0.9659122816869771,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.5542,
      "mean_token_accuracy": 0.17775871157646178,
      "num_tokens": 19188450.0,
      "step": 8600
    },
    {
      "entropy": 5.532081127166748,
      "epoch": 0.9664738585949346,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.4163,
      "mean_token_accuracy": 0.19002363234758377,
      "num_tokens": 19201254.0,
      "step": 8605
    },
    {
      "entropy": 5.505360412597656,
      "epoch": 0.9670354355028922,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.4888,
      "mean_token_accuracy": 0.18102688789367677,
      "num_tokens": 19214062.0,
      "step": 8610
    },
    {
      "entropy": 5.557801866531372,
      "epoch": 0.9675970124108496,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.4048,
      "mean_token_accuracy": 0.18907466828823088,
      "num_tokens": 19224844.0,
      "step": 8615
    },
    {
      "entropy": 5.545368671417236,
      "epoch": 0.9681585893188072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.5717,
      "mean_token_accuracy": 0.17443536072969437,
      "num_tokens": 19237347.0,
      "step": 8620
    },
    {
      "entropy": 5.599031209945679,
      "epoch": 0.9687201662267647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.4525,
      "mean_token_accuracy": 0.18267473876476287,
      "num_tokens": 19248743.0,
      "step": 8625
    },
    {
      "entropy": 5.597088861465454,
      "epoch": 0.9692817431347223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.4472,
      "mean_token_accuracy": 0.18550329953432082,
      "num_tokens": 19259904.0,
      "step": 8630
    },
    {
      "entropy": 5.488122892379761,
      "epoch": 0.9698433200426798,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.3845,
      "mean_token_accuracy": 0.1810299724340439,
      "num_tokens": 19271388.0,
      "step": 8635
    },
    {
      "entropy": 5.54411039352417,
      "epoch": 0.9704048969506374,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.5203,
      "mean_token_accuracy": 0.18260214924812318,
      "num_tokens": 19282790.0,
      "step": 8640
    },
    {
      "entropy": 5.609619045257569,
      "epoch": 0.9709664738585949,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.4125,
      "mean_token_accuracy": 0.1887241557240486,
      "num_tokens": 19293389.0,
      "step": 8645
    },
    {
      "entropy": 5.559426498413086,
      "epoch": 0.9715280507665525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.4924,
      "mean_token_accuracy": 0.18119360953569413,
      "num_tokens": 19304593.0,
      "step": 8650
    },
    {
      "entropy": 5.577017593383789,
      "epoch": 0.97208962767451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.4134,
      "mean_token_accuracy": 0.18593296855688096,
      "num_tokens": 19317388.0,
      "step": 8655
    },
    {
      "entropy": 5.610171699523926,
      "epoch": 0.9726512045824676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.4759,
      "mean_token_accuracy": 0.184097957611084,
      "num_tokens": 19329228.0,
      "step": 8660
    },
    {
      "entropy": 5.570147609710693,
      "epoch": 0.9732127814904251,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.439,
      "mean_token_accuracy": 0.18606530874967575,
      "num_tokens": 19339421.0,
      "step": 8665
    },
    {
      "entropy": 5.468735218048096,
      "epoch": 0.9737743583983827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.3461,
      "mean_token_accuracy": 0.18966307491064072,
      "num_tokens": 19349618.0,
      "step": 8670
    },
    {
      "entropy": 5.42205400466919,
      "epoch": 0.9743359353063402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.446,
      "mean_token_accuracy": 0.18687217086553573,
      "num_tokens": 19361324.0,
      "step": 8675
    },
    {
      "entropy": 5.5919225215911865,
      "epoch": 0.9748975122142978,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.4594,
      "mean_token_accuracy": 0.18415236473083496,
      "num_tokens": 19373631.0,
      "step": 8680
    },
    {
      "entropy": 5.509151649475098,
      "epoch": 0.9754590891222553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.3967,
      "mean_token_accuracy": 0.1740434393286705,
      "num_tokens": 19384682.0,
      "step": 8685
    },
    {
      "entropy": 5.621597242355347,
      "epoch": 0.9760206660302129,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.5324,
      "mean_token_accuracy": 0.17876951843500138,
      "num_tokens": 19396344.0,
      "step": 8690
    },
    {
      "entropy": 5.5938441276550295,
      "epoch": 0.9765822429381704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.3791,
      "mean_token_accuracy": 0.18256505727767944,
      "num_tokens": 19406921.0,
      "step": 8695
    },
    {
      "entropy": 5.543767309188842,
      "epoch": 0.977143819846128,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491560327430032,
      "loss": 5.5064,
      "mean_token_accuracy": 0.1808873161673546,
      "num_tokens": 19416851.0,
      "step": 8700
    },
    {
      "entropy": 5.577412939071655,
      "epoch": 0.9777053967540855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.4647,
      "mean_token_accuracy": 0.1804933190345764,
      "num_tokens": 19428180.0,
      "step": 8705
    },
    {
      "entropy": 5.691901683807373,
      "epoch": 0.9782669736620431,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.5244,
      "mean_token_accuracy": 0.1787811204791069,
      "num_tokens": 19439575.0,
      "step": 8710
    },
    {
      "entropy": 5.61657485961914,
      "epoch": 0.9788285505700005,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.4937,
      "mean_token_accuracy": 0.18056508004665375,
      "num_tokens": 19452074.0,
      "step": 8715
    },
    {
      "entropy": 5.4843040943145756,
      "epoch": 0.9793901274779581,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.423,
      "mean_token_accuracy": 0.18719545304775237,
      "num_tokens": 19463085.0,
      "step": 8720
    },
    {
      "entropy": 5.699585342407227,
      "epoch": 0.9799517043859156,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.5769,
      "mean_token_accuracy": 0.17127641886472703,
      "num_tokens": 19472726.0,
      "step": 8725
    },
    {
      "entropy": 5.559404087066651,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049149484540214,
      "loss": 5.4047,
      "mean_token_accuracy": 0.18202415704727173,
      "num_tokens": 19482275.0,
      "step": 8730
    },
    {
      "entropy": 5.5444048881530765,
      "epoch": 0.9810748582018307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.5032,
      "mean_token_accuracy": 0.18604240268468858,
      "num_tokens": 19492782.0,
      "step": 8735
    },
    {
      "entropy": 5.496974086761474,
      "epoch": 0.9816364351097883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.3922,
      "mean_token_accuracy": 0.18256796002388,
      "num_tokens": 19505709.0,
      "step": 8740
    },
    {
      "entropy": 5.586014890670777,
      "epoch": 0.9821980120177458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.4607,
      "mean_token_accuracy": 0.1845390111207962,
      "num_tokens": 19516269.0,
      "step": 8745
    },
    {
      "entropy": 5.560197639465332,
      "epoch": 0.9827595889257034,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.4564,
      "mean_token_accuracy": 0.18129220306873323,
      "num_tokens": 19526786.0,
      "step": 8750
    },
    {
      "entropy": 5.616420602798462,
      "epoch": 0.9833211658336609,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.4868,
      "mean_token_accuracy": 0.17830210477113723,
      "num_tokens": 19537865.0,
      "step": 8755
    },
    {
      "entropy": 5.50406551361084,
      "epoch": 0.9838827427416185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19029248505830765,
      "num_tokens": 19548320.0,
      "step": 8760
    },
    {
      "entropy": 5.565492105484009,
      "epoch": 0.984444319649576,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491418136064015,
      "loss": 5.4142,
      "mean_token_accuracy": 0.18464195132255554,
      "num_tokens": 19558856.0,
      "step": 8765
    },
    {
      "entropy": 5.56738600730896,
      "epoch": 0.9850058965575336,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491407150020996,
      "loss": 5.5146,
      "mean_token_accuracy": 0.1884256422519684,
      "num_tokens": 19570792.0,
      "step": 8770
    },
    {
      "entropy": 5.599941253662109,
      "epoch": 0.9855674734654911,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18806307315826415,
      "num_tokens": 19581982.0,
      "step": 8775
    },
    {
      "entropy": 5.512841844558716,
      "epoch": 0.9861290503734487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.3796,
      "mean_token_accuracy": 0.1856641262769699,
      "num_tokens": 19592882.0,
      "step": 8780
    },
    {
      "entropy": 5.651014518737793,
      "epoch": 0.9866906272814062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.5993,
      "mean_token_accuracy": 0.17879325300455093,
      "num_tokens": 19605967.0,
      "step": 8785
    },
    {
      "entropy": 5.645204353332519,
      "epoch": 0.9872522041893638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.4488,
      "mean_token_accuracy": 0.18693710565567018,
      "num_tokens": 19616494.0,
      "step": 8790
    },
    {
      "entropy": 5.582544946670533,
      "epoch": 0.9878137810973213,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.5433,
      "mean_token_accuracy": 0.1812344253063202,
      "num_tokens": 19627013.0,
      "step": 8795
    },
    {
      "entropy": 5.5240236759185795,
      "epoch": 0.9883753580052789,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491341089074901,
      "loss": 5.4555,
      "mean_token_accuracy": 0.18922722935676575,
      "num_tokens": 19639099.0,
      "step": 8800
    },
    {
      "entropy": 5.535636138916016,
      "epoch": 0.9889369349132364,
      "grad_norm": 1.046875,
      "learning_rate": 0.000491330054805827,
      "loss": 5.4266,
      "mean_token_accuracy": 0.18756327778100967,
      "num_tokens": 19650650.0,
      "step": 8805
    },
    {
      "entropy": 5.602877569198609,
      "epoch": 0.9894985118211939,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.4585,
      "mean_token_accuracy": 0.18385438323020936,
      "num_tokens": 19662869.0,
      "step": 8810
    },
    {
      "entropy": 5.521768283843994,
      "epoch": 0.9900600887291514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.4634,
      "mean_token_accuracy": 0.18303659409284592,
      "num_tokens": 19673232.0,
      "step": 8815
    },
    {
      "entropy": 5.587317323684692,
      "epoch": 0.990621665637109,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.4813,
      "mean_token_accuracy": 0.1822674974799156,
      "num_tokens": 19685069.0,
      "step": 8820
    },
    {
      "entropy": 5.6474627494812015,
      "epoch": 0.9911832425450665,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491285848852719,
      "loss": 5.5704,
      "mean_token_accuracy": 0.17589719742536544,
      "num_tokens": 19697018.0,
      "step": 8825
    },
    {
      "entropy": 5.591678476333618,
      "epoch": 0.9917448194530241,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.4193,
      "mean_token_accuracy": 0.1865503191947937,
      "num_tokens": 19707922.0,
      "step": 8830
    },
    {
      "entropy": 5.410496044158935,
      "epoch": 0.9923063963609816,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491263704555005,
      "loss": 5.37,
      "mean_token_accuracy": 0.18731970489025115,
      "num_tokens": 19718693.0,
      "step": 8835
    },
    {
      "entropy": 5.526531171798706,
      "epoch": 0.9928679732689392,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.4524,
      "mean_token_accuracy": 0.17944142073392869,
      "num_tokens": 19730846.0,
      "step": 8840
    },
    {
      "entropy": 5.548200798034668,
      "epoch": 0.9934295501768967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.4637,
      "mean_token_accuracy": 0.18399159759283065,
      "num_tokens": 19742099.0,
      "step": 8845
    },
    {
      "entropy": 5.576262474060059,
      "epoch": 0.9939911270848543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000491230436464915,
      "loss": 5.4773,
      "mean_token_accuracy": 0.18230542540550232,
      "num_tokens": 19753216.0,
      "step": 8850
    },
    {
      "entropy": 5.6016761302948,
      "epoch": 0.9945527039928118,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.4088,
      "mean_token_accuracy": 0.18425570875406266,
      "num_tokens": 19765316.0,
      "step": 8855
    },
    {
      "entropy": 5.562940406799316,
      "epoch": 0.9951142809007694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.4337,
      "mean_token_accuracy": 0.18214668035507203,
      "num_tokens": 19776208.0,
      "step": 8860
    },
    {
      "entropy": 5.593229818344116,
      "epoch": 0.9956758578087269,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.4778,
      "mean_token_accuracy": 0.1849060982465744,
      "num_tokens": 19788065.0,
      "step": 8865
    },
    {
      "entropy": 5.582309246063232,
      "epoch": 0.9962374347166845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18344223350286484,
      "num_tokens": 19799388.0,
      "step": 8870
    },
    {
      "entropy": 5.531808280944825,
      "epoch": 0.996799011624642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.491,
      "mean_token_accuracy": 0.17774271667003633,
      "num_tokens": 19809933.0,
      "step": 8875
    },
    {
      "entropy": 5.685315370559692,
      "epoch": 0.9973605885325996,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.5734,
      "mean_token_accuracy": 0.1801734372973442,
      "num_tokens": 19821566.0,
      "step": 8880
    },
    {
      "entropy": 5.6796534061431885,
      "epoch": 0.9979221654405571,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.525,
      "mean_token_accuracy": 0.18233054876327515,
      "num_tokens": 19833727.0,
      "step": 8885
    },
    {
      "entropy": 5.610952854156494,
      "epoch": 0.9984837423485147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.5418,
      "mean_token_accuracy": 0.1787643238902092,
      "num_tokens": 19845176.0,
      "step": 8890
    },
    {
      "entropy": 5.476277828216553,
      "epoch": 0.9990453192564722,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.4456,
      "mean_token_accuracy": 0.17728385776281358,
      "num_tokens": 19857528.0,
      "step": 8895
    },
    {
      "entropy": 5.590266132354737,
      "epoch": 0.9996068961644298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.373,
      "mean_token_accuracy": 0.18531470745801926,
      "num_tokens": 19868232.0,
      "step": 8900
    },
    {
      "entropy": 5.526914808485243,
      "epoch": 1.0001123153815914,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.3123,
      "mean_token_accuracy": 0.18969544933901894,
      "num_tokens": 19877881.0,
      "step": 8905
    },
    {
      "entropy": 5.416088581085205,
      "epoch": 1.000673892289549,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.4079,
      "mean_token_accuracy": 0.19263606518507004,
      "num_tokens": 19888398.0,
      "step": 8910
    },
    {
      "entropy": 5.603803396224976,
      "epoch": 1.0012354691975065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.3823,
      "mean_token_accuracy": 0.18999042063951493,
      "num_tokens": 19899661.0,
      "step": 8915
    },
    {
      "entropy": 5.565470886230469,
      "epoch": 1.0017970461054642,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.4358,
      "mean_token_accuracy": 0.19043354392051698,
      "num_tokens": 19911073.0,
      "step": 8920
    },
    {
      "entropy": 5.462346839904785,
      "epoch": 1.0023586230134216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18880361914634705,
      "num_tokens": 19922567.0,
      "step": 8925
    },
    {
      "entropy": 5.537394285202026,
      "epoch": 1.0029201999213793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.4645,
      "mean_token_accuracy": 0.18664260804653168,
      "num_tokens": 19934547.0,
      "step": 8930
    },
    {
      "entropy": 5.555563068389892,
      "epoch": 1.0034817768293367,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.4688,
      "mean_token_accuracy": 0.18146257624030113,
      "num_tokens": 19947659.0,
      "step": 8935
    },
    {
      "entropy": 5.588209009170532,
      "epoch": 1.0040433537372944,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000491029527020594,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18237140774726868,
      "num_tokens": 19958150.0,
      "step": 8940
    },
    {
      "entropy": 5.49904727935791,
      "epoch": 1.0046049306452518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.361,
      "mean_token_accuracy": 0.18561754673719405,
      "num_tokens": 19968173.0,
      "step": 8945
    },
    {
      "entropy": 5.523112201690674,
      "epoch": 1.0051665075532095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1836772307753563,
      "num_tokens": 19978543.0,
      "step": 8950
    },
    {
      "entropy": 5.5357386589050295,
      "epoch": 1.005728084461167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.358,
      "mean_token_accuracy": 0.18340061455965043,
      "num_tokens": 19989586.0,
      "step": 8955
    },
    {
      "entropy": 5.521802759170532,
      "epoch": 1.0062896613691246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.3736,
      "mean_token_accuracy": 0.1864837720990181,
      "num_tokens": 20000524.0,
      "step": 8960
    },
    {
      "entropy": 5.5667784214019775,
      "epoch": 1.006851238277082,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.494,
      "mean_token_accuracy": 0.18308472037315368,
      "num_tokens": 20012935.0,
      "step": 8965
    },
    {
      "entropy": 5.56749906539917,
      "epoch": 1.0074128151850397,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.46,
      "mean_token_accuracy": 0.18224538564682008,
      "num_tokens": 20024435.0,
      "step": 8970
    },
    {
      "entropy": 5.587008428573609,
      "epoch": 1.007974392092997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.4596,
      "mean_token_accuracy": 0.17887090891599655,
      "num_tokens": 20036017.0,
      "step": 8975
    },
    {
      "entropy": 5.5421875,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.441,
      "mean_token_accuracy": 0.1802113637328148,
      "num_tokens": 20047216.0,
      "step": 8980
    },
    {
      "entropy": 5.53623366355896,
      "epoch": 1.0090975459089122,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.3309,
      "mean_token_accuracy": 0.185406456887722,
      "num_tokens": 20057439.0,
      "step": 8985
    },
    {
      "entropy": 5.550770998001099,
      "epoch": 1.0096591228168699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.4458,
      "mean_token_accuracy": 0.18372596949338912,
      "num_tokens": 20069602.0,
      "step": 8990
    },
    {
      "entropy": 5.505699682235718,
      "epoch": 1.0102206997248273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.396,
      "mean_token_accuracy": 0.18840807676315308,
      "num_tokens": 20081366.0,
      "step": 8995
    },
    {
      "entropy": 5.554901027679444,
      "epoch": 1.010782276632785,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.3988,
      "mean_token_accuracy": 0.18910853266716005,
      "num_tokens": 20091384.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.457474174898003,
      "eval_loss": 5.499125957489014,
      "eval_mean_token_accuracy": 0.18832171713534107,
      "eval_num_tokens": 20091384.0,
      "eval_runtime": 58.5199,
      "eval_samples_per_second": 529.956,
      "eval_steps_per_second": 66.251,
      "step": 9000
    },
    {
      "entropy": 5.616485166549682,
      "epoch": 1.0113438535407424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.4308,
      "mean_token_accuracy": 0.18055644035339355,
      "num_tokens": 20102953.0,
      "step": 9005
    },
    {
      "entropy": 5.5084339618682865,
      "epoch": 1.0119054304486998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.3845,
      "mean_token_accuracy": 0.1884632647037506,
      "num_tokens": 20113672.0,
      "step": 9010
    },
    {
      "entropy": 5.578459787368774,
      "epoch": 1.0124670073566575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.4048,
      "mean_token_accuracy": 0.186214779317379,
      "num_tokens": 20125143.0,
      "step": 9015
    },
    {
      "entropy": 5.5441014766693115,
      "epoch": 1.013028584264615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.4129,
      "mean_token_accuracy": 0.18566339910030366,
      "num_tokens": 20135489.0,
      "step": 9020
    },
    {
      "entropy": 5.513772201538086,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18757753372192382,
      "num_tokens": 20146768.0,
      "step": 9025
    },
    {
      "entropy": 5.329649972915649,
      "epoch": 1.01415173808053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.2137,
      "mean_token_accuracy": 0.18804068714380265,
      "num_tokens": 20156894.0,
      "step": 9030
    },
    {
      "entropy": 5.542341899871826,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.4357,
      "mean_token_accuracy": 0.1877053663134575,
      "num_tokens": 20168174.0,
      "step": 9035
    },
    {
      "entropy": 5.622216033935547,
      "epoch": 1.0152748918964452,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490803680581439,
      "loss": 5.5145,
      "mean_token_accuracy": 0.17867926359176636,
      "num_tokens": 20179962.0,
      "step": 9040
    },
    {
      "entropy": 5.598105525970459,
      "epoch": 1.0158364688044028,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490792316087133,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1849098801612854,
      "num_tokens": 20189842.0,
      "step": 9045
    },
    {
      "entropy": 5.5179485321044925,
      "epoch": 1.0163980457123603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.3081,
      "mean_token_accuracy": 0.1819055676460266,
      "num_tokens": 20200471.0,
      "step": 9050
    },
    {
      "entropy": 5.528376197814941,
      "epoch": 1.016959622620318,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.384,
      "mean_token_accuracy": 0.1852393552660942,
      "num_tokens": 20212145.0,
      "step": 9055
    },
    {
      "entropy": 5.502276039123535,
      "epoch": 1.0175211995282754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.3749,
      "mean_token_accuracy": 0.18804652988910675,
      "num_tokens": 20222454.0,
      "step": 9060
    },
    {
      "entropy": 5.597353744506836,
      "epoch": 1.018082776436233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1800994336605072,
      "num_tokens": 20233402.0,
      "step": 9065
    },
    {
      "entropy": 5.487265300750733,
      "epoch": 1.0186443533441905,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19445917904376983,
      "num_tokens": 20244102.0,
      "step": 9070
    },
    {
      "entropy": 5.594339370727539,
      "epoch": 1.0192059302521481,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490723984844845,
      "loss": 5.5201,
      "mean_token_accuracy": 0.18200081139802932,
      "num_tokens": 20255003.0,
      "step": 9075
    },
    {
      "entropy": 5.551009750366211,
      "epoch": 1.0197675071601056,
      "grad_norm": 1.125,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18559956699609756,
      "num_tokens": 20265211.0,
      "step": 9080
    },
    {
      "entropy": 5.475327205657959,
      "epoch": 1.0203290840680632,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.3241,
      "mean_token_accuracy": 0.18909415304660798,
      "num_tokens": 20276552.0,
      "step": 9085
    },
    {
      "entropy": 5.559994316101074,
      "epoch": 1.0208906609760207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.4538,
      "mean_token_accuracy": 0.18224422484636307,
      "num_tokens": 20286805.0,
      "step": 9090
    },
    {
      "entropy": 5.557529258728027,
      "epoch": 1.0214522378839783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.3922,
      "mean_token_accuracy": 0.18283914625644684,
      "num_tokens": 20298536.0,
      "step": 9095
    },
    {
      "entropy": 5.530365753173828,
      "epoch": 1.0220138147919358,
      "grad_norm": 1.125,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.4019,
      "mean_token_accuracy": 0.18648336976766586,
      "num_tokens": 20309842.0,
      "step": 9100
    },
    {
      "entropy": 5.517857694625855,
      "epoch": 1.0225753916998932,
      "grad_norm": 1.125,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.3217,
      "mean_token_accuracy": 0.1901137039065361,
      "num_tokens": 20320413.0,
      "step": 9105
    },
    {
      "entropy": 5.549293136596679,
      "epoch": 1.0231369686078509,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.3748,
      "mean_token_accuracy": 0.1853329733014107,
      "num_tokens": 20331812.0,
      "step": 9110
    },
    {
      "entropy": 5.475205707550049,
      "epoch": 1.0236985455158083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1900770142674446,
      "num_tokens": 20342148.0,
      "step": 9115
    },
    {
      "entropy": 5.516740608215332,
      "epoch": 1.024260122423766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18869406282901763,
      "num_tokens": 20353843.0,
      "step": 9120
    },
    {
      "entropy": 5.548943662643433,
      "epoch": 1.0248216993317234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.4208,
      "mean_token_accuracy": 0.18616133481264113,
      "num_tokens": 20364846.0,
      "step": 9125
    },
    {
      "entropy": 5.565707874298096,
      "epoch": 1.025383276239681,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.3691,
      "mean_token_accuracy": 0.1888698384165764,
      "num_tokens": 20376874.0,
      "step": 9130
    },
    {
      "entropy": 5.49728045463562,
      "epoch": 1.0259448531476385,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.3856,
      "mean_token_accuracy": 0.18706257939338683,
      "num_tokens": 20388189.0,
      "step": 9135
    },
    {
      "entropy": 5.536900901794434,
      "epoch": 1.0265064300555962,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18550562411546706,
      "num_tokens": 20399155.0,
      "step": 9140
    },
    {
      "entropy": 5.513206052780151,
      "epoch": 1.0270680069635536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1886839136481285,
      "num_tokens": 20410569.0,
      "step": 9145
    },
    {
      "entropy": 5.426887464523316,
      "epoch": 1.0276295838715113,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.3469,
      "mean_token_accuracy": 0.18439428955316545,
      "num_tokens": 20420438.0,
      "step": 9150
    },
    {
      "entropy": 5.56753511428833,
      "epoch": 1.0281911607794687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1855535164475441,
      "num_tokens": 20431874.0,
      "step": 9155
    },
    {
      "entropy": 5.522236251831055,
      "epoch": 1.0287527376874264,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.3824,
      "mean_token_accuracy": 0.18540708273649215,
      "num_tokens": 20442792.0,
      "step": 9160
    },
    {
      "entropy": 5.567186307907105,
      "epoch": 1.0293143145953838,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490517507786001,
      "loss": 5.3997,
      "mean_token_accuracy": 0.18406621366739273,
      "num_tokens": 20454410.0,
      "step": 9165
    },
    {
      "entropy": 5.655465030670166,
      "epoch": 1.0298758915033415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.5354,
      "mean_token_accuracy": 0.1752452775835991,
      "num_tokens": 20467072.0,
      "step": 9170
    },
    {
      "entropy": 5.57279372215271,
      "epoch": 1.030437468411299,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.4558,
      "mean_token_accuracy": 0.18608310520648957,
      "num_tokens": 20477697.0,
      "step": 9175
    },
    {
      "entropy": 5.52084174156189,
      "epoch": 1.0309990453192566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.2931,
      "mean_token_accuracy": 0.1903368964791298,
      "num_tokens": 20489024.0,
      "step": 9180
    },
    {
      "entropy": 5.500179243087769,
      "epoch": 1.031560622227214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18683810234069825,
      "num_tokens": 20499928.0,
      "step": 9185
    },
    {
      "entropy": 5.44192795753479,
      "epoch": 1.0321221991351717,
      "grad_norm": 1.125,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.4284,
      "mean_token_accuracy": 0.1826045975089073,
      "num_tokens": 20511085.0,
      "step": 9190
    },
    {
      "entropy": 5.55776572227478,
      "epoch": 1.032683776043129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18520693182945253,
      "num_tokens": 20522164.0,
      "step": 9195
    },
    {
      "entropy": 5.586649608612061,
      "epoch": 1.0332453529510865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.4627,
      "mean_token_accuracy": 0.186662457883358,
      "num_tokens": 20534187.0,
      "step": 9200
    },
    {
      "entropy": 5.59186577796936,
      "epoch": 1.0338069298590442,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.4379,
      "mean_token_accuracy": 0.1879410594701767,
      "num_tokens": 20545819.0,
      "step": 9205
    },
    {
      "entropy": 5.610886096954346,
      "epoch": 1.0343685067670016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.449,
      "mean_token_accuracy": 0.17868003100156785,
      "num_tokens": 20556803.0,
      "step": 9210
    },
    {
      "entropy": 5.562782859802246,
      "epoch": 1.0349300836749593,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.4392,
      "mean_token_accuracy": 0.17891971468925477,
      "num_tokens": 20567366.0,
      "step": 9215
    },
    {
      "entropy": 5.5469132423400875,
      "epoch": 1.0354916605829168,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1822926864027977,
      "num_tokens": 20579182.0,
      "step": 9220
    },
    {
      "entropy": 5.515037584304809,
      "epoch": 1.0360532374908744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.3174,
      "mean_token_accuracy": 0.18531483113765718,
      "num_tokens": 20590346.0,
      "step": 9225
    },
    {
      "entropy": 5.52108678817749,
      "epoch": 1.0366148143988319,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.4186,
      "mean_token_accuracy": 0.18376436084508896,
      "num_tokens": 20601279.0,
      "step": 9230
    },
    {
      "entropy": 5.487068128585816,
      "epoch": 1.0371763913067895,
      "grad_norm": 1.21875,
      "learning_rate": 0.000490355377285356,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18688741326332092,
      "num_tokens": 20612163.0,
      "step": 9235
    },
    {
      "entropy": 5.550949001312256,
      "epoch": 1.037737968214747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.3069,
      "mean_token_accuracy": 0.1903759852051735,
      "num_tokens": 20622587.0,
      "step": 9240
    },
    {
      "entropy": 5.528546476364136,
      "epoch": 1.0382995451227046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.4755,
      "mean_token_accuracy": 0.17715801149606705,
      "num_tokens": 20633341.0,
      "step": 9245
    },
    {
      "entropy": 5.494615077972412,
      "epoch": 1.038861122030662,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.3492,
      "mean_token_accuracy": 0.18869257122278213,
      "num_tokens": 20644582.0,
      "step": 9250
    },
    {
      "entropy": 5.527494478225708,
      "epoch": 1.0394226989386197,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1975536048412323,
      "num_tokens": 20654936.0,
      "step": 9255
    },
    {
      "entropy": 5.434938859939575,
      "epoch": 1.0399842758465772,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.3381,
      "mean_token_accuracy": 0.1842801809310913,
      "num_tokens": 20666348.0,
      "step": 9260
    },
    {
      "entropy": 5.5995471477508545,
      "epoch": 1.0405458527545348,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.5141,
      "mean_token_accuracy": 0.18195110112428664,
      "num_tokens": 20677482.0,
      "step": 9265
    },
    {
      "entropy": 5.579679870605469,
      "epoch": 1.0411074296624923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.425,
      "mean_token_accuracy": 0.19345418363809586,
      "num_tokens": 20688369.0,
      "step": 9270
    },
    {
      "entropy": 5.517041492462158,
      "epoch": 1.04166900657045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.437,
      "mean_token_accuracy": 0.18158860951662065,
      "num_tokens": 20698933.0,
      "step": 9275
    },
    {
      "entropy": 5.515263271331787,
      "epoch": 1.0422305834784074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.3057,
      "mean_token_accuracy": 0.1855906665325165,
      "num_tokens": 20710192.0,
      "step": 9280
    },
    {
      "entropy": 5.498098850250244,
      "epoch": 1.042792160386365,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.362,
      "mean_token_accuracy": 0.1852239578962326,
      "num_tokens": 20721408.0,
      "step": 9285
    },
    {
      "entropy": 5.419748306274414,
      "epoch": 1.0433537372943225,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.272,
      "mean_token_accuracy": 0.19246668368577957,
      "num_tokens": 20732155.0,
      "step": 9290
    },
    {
      "entropy": 5.4803431034088135,
      "epoch": 1.04391531420228,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490215338554978,
      "loss": 5.2751,
      "mean_token_accuracy": 0.19139215499162673,
      "num_tokens": 20741819.0,
      "step": 9295
    },
    {
      "entropy": 5.537647771835327,
      "epoch": 1.0444768911102376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.4062,
      "mean_token_accuracy": 0.18556405305862428,
      "num_tokens": 20754931.0,
      "step": 9300
    },
    {
      "entropy": 5.5981498718261715,
      "epoch": 1.045038468018195,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.4801,
      "mean_token_accuracy": 0.1823945239186287,
      "num_tokens": 20766533.0,
      "step": 9305
    },
    {
      "entropy": 5.502779340744018,
      "epoch": 1.0456000449261527,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.346,
      "mean_token_accuracy": 0.19028619229793547,
      "num_tokens": 20777055.0,
      "step": 9310
    },
    {
      "entropy": 5.418141555786133,
      "epoch": 1.04616162183411,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.2939,
      "mean_token_accuracy": 0.18705033957958223,
      "num_tokens": 20786871.0,
      "step": 9315
    },
    {
      "entropy": 5.565149259567261,
      "epoch": 1.0467231987420678,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.4456,
      "mean_token_accuracy": 0.182110396027565,
      "num_tokens": 20798546.0,
      "step": 9320
    },
    {
      "entropy": 5.530739164352417,
      "epoch": 1.0472847756500252,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.38,
      "mean_token_accuracy": 0.18562474697828293,
      "num_tokens": 20809657.0,
      "step": 9325
    },
    {
      "entropy": 5.46672682762146,
      "epoch": 1.0478463525579829,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18574664890766143,
      "num_tokens": 20819662.0,
      "step": 9330
    },
    {
      "entropy": 5.450799417495728,
      "epoch": 1.0484079294659403,
      "grad_norm": 1.046875,
      "learning_rate": 0.000490121431158592,
      "loss": 5.2663,
      "mean_token_accuracy": 0.19446823745965958,
      "num_tokens": 20830554.0,
      "step": 9335
    },
    {
      "entropy": 5.571157646179199,
      "epoch": 1.048969506373898,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.4337,
      "mean_token_accuracy": 0.1810958907008171,
      "num_tokens": 20842003.0,
      "step": 9340
    },
    {
      "entropy": 5.559390020370484,
      "epoch": 1.0495310832818554,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.4178,
      "mean_token_accuracy": 0.1890198290348053,
      "num_tokens": 20853453.0,
      "step": 9345
    },
    {
      "entropy": 5.6493274688720705,
      "epoch": 1.050092660189813,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.4767,
      "mean_token_accuracy": 0.1816459745168686,
      "num_tokens": 20867296.0,
      "step": 9350
    },
    {
      "entropy": 5.473407840728759,
      "epoch": 1.0506542370977705,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.3723,
      "mean_token_accuracy": 0.18959630727767945,
      "num_tokens": 20878951.0,
      "step": 9355
    },
    {
      "entropy": 5.556913042068482,
      "epoch": 1.0512158140057282,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.4364,
      "mean_token_accuracy": 0.18974503874778748,
      "num_tokens": 20891103.0,
      "step": 9360
    },
    {
      "entropy": 5.632567405700684,
      "epoch": 1.0517773909136856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.5273,
      "mean_token_accuracy": 0.1795767903327942,
      "num_tokens": 20903570.0,
      "step": 9365
    },
    {
      "entropy": 5.604644870758056,
      "epoch": 1.0523389678216433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.4774,
      "mean_token_accuracy": 0.182725690305233,
      "num_tokens": 20914156.0,
      "step": 9370
    },
    {
      "entropy": 5.50830864906311,
      "epoch": 1.0529005447296007,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490027085391092,
      "loss": 5.3634,
      "mean_token_accuracy": 0.18633141964673997,
      "num_tokens": 20925192.0,
      "step": 9375
    },
    {
      "entropy": 5.559418725967407,
      "epoch": 1.0534621216375584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490015261356751,
      "loss": 5.2579,
      "mean_token_accuracy": 0.20030733942985535,
      "num_tokens": 20935281.0,
      "step": 9380
    },
    {
      "entropy": 5.603970336914062,
      "epoch": 1.0540236985455158,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.4817,
      "mean_token_accuracy": 0.18674329668283463,
      "num_tokens": 20947397.0,
      "step": 9385
    },
    {
      "entropy": 5.462509298324585,
      "epoch": 1.0545852754534732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489991592749938,
      "loss": 5.3527,
      "mean_token_accuracy": 0.18962117433547973,
      "num_tokens": 20958473.0,
      "step": 9390
    },
    {
      "entropy": 5.624472379684448,
      "epoch": 1.055146852361431,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.4511,
      "mean_token_accuracy": 0.17956508100032806,
      "num_tokens": 20970112.0,
      "step": 9395
    },
    {
      "entropy": 5.5020238876342775,
      "epoch": 1.0557084292693883,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.3606,
      "mean_token_accuracy": 0.18722193837165832,
      "num_tokens": 20980762.0,
      "step": 9400
    },
    {
      "entropy": 5.46553521156311,
      "epoch": 1.056270006177346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.2688,
      "mean_token_accuracy": 0.19204796999692916,
      "num_tokens": 20991663.0,
      "step": 9405
    },
    {
      "entropy": 5.472022581100464,
      "epoch": 1.0568315830853034,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489944173394342,
      "loss": 5.2958,
      "mean_token_accuracy": 0.19311801642179488,
      "num_tokens": 21002121.0,
      "step": 9410
    },
    {
      "entropy": 5.4545652866363525,
      "epoch": 1.0573931599932611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.3564,
      "mean_token_accuracy": 0.1815158247947693,
      "num_tokens": 21013791.0,
      "step": 9415
    },
    {
      "entropy": 5.448367071151734,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18644743114709855,
      "num_tokens": 21022977.0,
      "step": 9420
    },
    {
      "entropy": 5.521354913711548,
      "epoch": 1.0585163138091762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.4551,
      "mean_token_accuracy": 0.18517175167798997,
      "num_tokens": 21036372.0,
      "step": 9425
    },
    {
      "entropy": 5.680944919586182,
      "epoch": 1.0590778907171337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.5038,
      "mean_token_accuracy": 0.17759744226932525,
      "num_tokens": 21047984.0,
      "step": 9430
    },
    {
      "entropy": 5.606042003631591,
      "epoch": 1.0596394676250913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.3977,
      "mean_token_accuracy": 0.18267227858304977,
      "num_tokens": 21060127.0,
      "step": 9435
    },
    {
      "entropy": 5.482920551300049,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18611850142478942,
      "num_tokens": 21070659.0,
      "step": 9440
    },
    {
      "entropy": 5.472131824493408,
      "epoch": 1.0607626214410064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.3223,
      "mean_token_accuracy": 0.19197576344013215,
      "num_tokens": 21081667.0,
      "step": 9445
    },
    {
      "entropy": 5.533467817306518,
      "epoch": 1.0613241983489639,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.3432,
      "mean_token_accuracy": 0.18749047070741653,
      "num_tokens": 21094275.0,
      "step": 9450
    },
    {
      "entropy": 5.549678325653076,
      "epoch": 1.0618857752569215,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.4108,
      "mean_token_accuracy": 0.18219694793224334,
      "num_tokens": 21106360.0,
      "step": 9455
    },
    {
      "entropy": 5.505332660675049,
      "epoch": 1.062447352164879,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1860309511423111,
      "num_tokens": 21117647.0,
      "step": 9460
    },
    {
      "entropy": 5.3988227367401125,
      "epoch": 1.0630089290728366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19441557824611663,
      "num_tokens": 21129599.0,
      "step": 9465
    },
    {
      "entropy": 5.485315370559692,
      "epoch": 1.063570505980794,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.3569,
      "mean_token_accuracy": 0.1858839511871338,
      "num_tokens": 21140573.0,
      "step": 9470
    },
    {
      "entropy": 5.580947637557983,
      "epoch": 1.0641320828887517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.4311,
      "mean_token_accuracy": 0.1874650239944458,
      "num_tokens": 21151899.0,
      "step": 9475
    },
    {
      "entropy": 5.602825069427491,
      "epoch": 1.0646936597967092,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.4881,
      "mean_token_accuracy": 0.17654771059751512,
      "num_tokens": 21162748.0,
      "step": 9480
    },
    {
      "entropy": 5.550855731964111,
      "epoch": 1.0652552367046666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000489765375752192,
      "loss": 5.4042,
      "mean_token_accuracy": 0.183906626701355,
      "num_tokens": 21172962.0,
      "step": 9485
    },
    {
      "entropy": 5.54920072555542,
      "epoch": 1.0658168136126243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.3891,
      "mean_token_accuracy": 0.18949666619300842,
      "num_tokens": 21183805.0,
      "step": 9490
    },
    {
      "entropy": 5.438908433914184,
      "epoch": 1.0663783905205817,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19069612473249437,
      "num_tokens": 21195129.0,
      "step": 9495
    },
    {
      "entropy": 5.4912261962890625,
      "epoch": 1.0669399674285394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000489729431550941,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1807676836848259,
      "num_tokens": 21206300.0,
      "step": 9500
    },
    {
      "entropy": 5.565548753738403,
      "epoch": 1.0675015443364968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18868244737386702,
      "num_tokens": 21217137.0,
      "step": 9505
    },
    {
      "entropy": 5.481102037429809,
      "epoch": 1.0680631212444545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.3322,
      "mean_token_accuracy": 0.19163161218166352,
      "num_tokens": 21228399.0,
      "step": 9510
    },
    {
      "entropy": 5.528740978240966,
      "epoch": 1.068624698152412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18235985487699508,
      "num_tokens": 21239396.0,
      "step": 9515
    },
    {
      "entropy": 5.504907941818237,
      "epoch": 1.0691862750603696,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1862892672419548,
      "num_tokens": 21249966.0,
      "step": 9520
    },
    {
      "entropy": 5.5240458965301515,
      "epoch": 1.069747851968327,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18622422069311143,
      "num_tokens": 21260157.0,
      "step": 9525
    },
    {
      "entropy": 5.523571395874024,
      "epoch": 1.0703094288762847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.4258,
      "mean_token_accuracy": 0.18874689936637878,
      "num_tokens": 21270738.0,
      "step": 9530
    },
    {
      "entropy": 5.4886860847473145,
      "epoch": 1.070871005784242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18944532722234725,
      "num_tokens": 21282469.0,
      "step": 9535
    },
    {
      "entropy": 5.496334505081177,
      "epoch": 1.0714325826921998,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1876911997795105,
      "num_tokens": 21293474.0,
      "step": 9540
    },
    {
      "entropy": 5.541016721725464,
      "epoch": 1.0719941596001572,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1798918917775154,
      "num_tokens": 21304904.0,
      "step": 9545
    },
    {
      "entropy": 5.574141883850098,
      "epoch": 1.0725557365081149,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18713437765836716,
      "num_tokens": 21315407.0,
      "step": 9550
    },
    {
      "entropy": 5.469040870666504,
      "epoch": 1.0731173134160723,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.3148,
      "mean_token_accuracy": 0.19208430349826813,
      "num_tokens": 21326496.0,
      "step": 9555
    },
    {
      "entropy": 5.515277433395386,
      "epoch": 1.07367889032403,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.4026,
      "mean_token_accuracy": 0.1864038661122322,
      "num_tokens": 21339449.0,
      "step": 9560
    },
    {
      "entropy": 5.527117300033569,
      "epoch": 1.0742404672319874,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.4633,
      "mean_token_accuracy": 0.18114811927080154,
      "num_tokens": 21350431.0,
      "step": 9565
    },
    {
      "entropy": 5.497921133041382,
      "epoch": 1.074802044139945,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18641359955072404,
      "num_tokens": 21361168.0,
      "step": 9570
    },
    {
      "entropy": 5.535551834106445,
      "epoch": 1.0753636210479025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.3458,
      "mean_token_accuracy": 0.1902617409825325,
      "num_tokens": 21372635.0,
      "step": 9575
    },
    {
      "entropy": 5.446909427642822,
      "epoch": 1.07592519795586,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.3517,
      "mean_token_accuracy": 0.18696731477975845,
      "num_tokens": 21384051.0,
      "step": 9580
    },
    {
      "entropy": 5.511615562438965,
      "epoch": 1.0764867748638176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.3767,
      "mean_token_accuracy": 0.17953940927982331,
      "num_tokens": 21397496.0,
      "step": 9585
    },
    {
      "entropy": 5.510258388519287,
      "epoch": 1.077048351771775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.4354,
      "mean_token_accuracy": 0.1822570264339447,
      "num_tokens": 21409349.0,
      "step": 9590
    },
    {
      "entropy": 5.527426385879517,
      "epoch": 1.0776099286797327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.3954,
      "mean_token_accuracy": 0.1880100414156914,
      "num_tokens": 21421843.0,
      "step": 9595
    },
    {
      "entropy": 5.507824325561524,
      "epoch": 1.0781715055876901,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.275,
      "mean_token_accuracy": 0.18905281722545625,
      "num_tokens": 21433230.0,
      "step": 9600
    },
    {
      "entropy": 5.541763591766357,
      "epoch": 1.0787330824956478,
      "grad_norm": 1.109375,
      "learning_rate": 0.000489476099780026,
      "loss": 5.412,
      "mean_token_accuracy": 0.1857568547129631,
      "num_tokens": 21444295.0,
      "step": 9605
    },
    {
      "entropy": 5.568008661270142,
      "epoch": 1.0792946594036052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.3545,
      "mean_token_accuracy": 0.1823578417301178,
      "num_tokens": 21455721.0,
      "step": 9610
    },
    {
      "entropy": 5.496934843063355,
      "epoch": 1.079856236311563,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.3864,
      "mean_token_accuracy": 0.18858225643634796,
      "num_tokens": 21466514.0,
      "step": 9615
    },
    {
      "entropy": 5.519130754470825,
      "epoch": 1.0804178132195204,
      "grad_norm": 1.21875,
      "learning_rate": 0.000489439663599751,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18560597747564317,
      "num_tokens": 21478782.0,
      "step": 9620
    },
    {
      "entropy": 5.6073750972747805,
      "epoch": 1.080979390127478,
      "grad_norm": 1.09375,
      "learning_rate": 0.000489427504550126,
      "loss": 5.3838,
      "mean_token_accuracy": 0.19413492530584336,
      "num_tokens": 21490428.0,
      "step": 9625
    },
    {
      "entropy": 5.39088807106018,
      "epoch": 1.0815409670354355,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18894489258527755,
      "num_tokens": 21501530.0,
      "step": 9630
    },
    {
      "entropy": 5.502291059494018,
      "epoch": 1.0821025439433931,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.3327,
      "mean_token_accuracy": 0.19191098362207412,
      "num_tokens": 21512122.0,
      "step": 9635
    },
    {
      "entropy": 5.477610349655151,
      "epoch": 1.0826641208513506,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.2542,
      "mean_token_accuracy": 0.18706072121858597,
      "num_tokens": 21521896.0,
      "step": 9640
    },
    {
      "entropy": 5.502262163162231,
      "epoch": 1.0832256977593082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000489378800080952,
      "loss": 5.4232,
      "mean_token_accuracy": 0.18383677750825883,
      "num_tokens": 21533584.0,
      "step": 9645
    },
    {
      "entropy": 5.418279790878296,
      "epoch": 1.0837872746672657,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.3868,
      "mean_token_accuracy": 0.19056437462568282,
      "num_tokens": 21545432.0,
      "step": 9650
    },
    {
      "entropy": 5.534392881393432,
      "epoch": 1.0843488515752233,
      "grad_norm": 1.125,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.3864,
      "mean_token_accuracy": 0.1842215746641159,
      "num_tokens": 21556547.0,
      "step": 9655
    },
    {
      "entropy": 5.5567206859588625,
      "epoch": 1.0849104284831808,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1804027333855629,
      "num_tokens": 21567354.0,
      "step": 9660
    },
    {
      "entropy": 5.567187309265137,
      "epoch": 1.0854720053911384,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.3219,
      "mean_token_accuracy": 0.18749481737613677,
      "num_tokens": 21578694.0,
      "step": 9665
    },
    {
      "entropy": 5.474405145645141,
      "epoch": 1.0860335822990959,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.274,
      "mean_token_accuracy": 0.19735342413187026,
      "num_tokens": 21589365.0,
      "step": 9670
    },
    {
      "entropy": 5.520022010803222,
      "epoch": 1.0865951592070533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18308279663324356,
      "num_tokens": 21600766.0,
      "step": 9675
    },
    {
      "entropy": 5.533956050872803,
      "epoch": 1.087156736115011,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.534,
      "mean_token_accuracy": 0.178544320166111,
      "num_tokens": 21613969.0,
      "step": 9680
    },
    {
      "entropy": 5.494122648239136,
      "epoch": 1.0877183130229684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.3484,
      "mean_token_accuracy": 0.18665636330842972,
      "num_tokens": 21626302.0,
      "step": 9685
    },
    {
      "entropy": 5.582063674926758,
      "epoch": 1.088279889930926,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.4242,
      "mean_token_accuracy": 0.18517893850803374,
      "num_tokens": 21637058.0,
      "step": 9690
    },
    {
      "entropy": 5.57568507194519,
      "epoch": 1.0888414668388835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.4384,
      "mean_token_accuracy": 0.17996353805065154,
      "num_tokens": 21648729.0,
      "step": 9695
    },
    {
      "entropy": 5.421488428115845,
      "epoch": 1.0894030437468412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19667865931987763,
      "num_tokens": 21659961.0,
      "step": 9700
    },
    {
      "entropy": 5.4760987758636475,
      "epoch": 1.0899646206547986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.3471,
      "mean_token_accuracy": 0.1910601496696472,
      "num_tokens": 21670577.0,
      "step": 9705
    },
    {
      "entropy": 5.474126720428467,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.3096,
      "mean_token_accuracy": 0.195725616812706,
      "num_tokens": 21679882.0,
      "step": 9710
    },
    {
      "entropy": 5.569431114196777,
      "epoch": 1.0910877744707137,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.4347,
      "mean_token_accuracy": 0.19174170345067978,
      "num_tokens": 21690595.0,
      "step": 9715
    },
    {
      "entropy": 5.459299182891845,
      "epoch": 1.0916493513786714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19190464466810225,
      "num_tokens": 21701251.0,
      "step": 9720
    },
    {
      "entropy": 5.563612699508667,
      "epoch": 1.0922109282866288,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18425981253385543,
      "num_tokens": 21713286.0,
      "step": 9725
    },
    {
      "entropy": 5.58175835609436,
      "epoch": 1.0927725051945865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.5244,
      "mean_token_accuracy": 0.1812610000371933,
      "num_tokens": 21725508.0,
      "step": 9730
    },
    {
      "entropy": 5.508434200286866,
      "epoch": 1.093334082102544,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18853978514671327,
      "num_tokens": 21736467.0,
      "step": 9735
    },
    {
      "entropy": 5.599121046066284,
      "epoch": 1.0938956590105016,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.4172,
      "mean_token_accuracy": 0.1899057596921921,
      "num_tokens": 21746114.0,
      "step": 9740
    },
    {
      "entropy": 5.482849264144898,
      "epoch": 1.094457235918459,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000489133640016346,
      "loss": 5.4166,
      "mean_token_accuracy": 0.1853234961628914,
      "num_tokens": 21757145.0,
      "step": 9745
    },
    {
      "entropy": 5.4897778034210205,
      "epoch": 1.0950188128264167,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.3906,
      "mean_token_accuracy": 0.18188587725162506,
      "num_tokens": 21768294.0,
      "step": 9750
    },
    {
      "entropy": 5.565903949737549,
      "epoch": 1.095580389734374,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.4016,
      "mean_token_accuracy": 0.18574694842100142,
      "num_tokens": 21779667.0,
      "step": 9755
    },
    {
      "entropy": 5.454914569854736,
      "epoch": 1.0961419666423318,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.2588,
      "mean_token_accuracy": 0.19705150574445723,
      "num_tokens": 21789635.0,
      "step": 9760
    },
    {
      "entropy": 5.410786819458008,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000489084280634532,
      "loss": 5.2576,
      "mean_token_accuracy": 0.19602492451667786,
      "num_tokens": 21799462.0,
      "step": 9765
    },
    {
      "entropy": 5.416881847381592,
      "epoch": 1.0972651204582466,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1895948529243469,
      "num_tokens": 21811520.0,
      "step": 9770
    },
    {
      "entropy": 5.622190618515015,
      "epoch": 1.0978266973662043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.3953,
      "mean_token_accuracy": 0.18550430834293366,
      "num_tokens": 21821997.0,
      "step": 9775
    },
    {
      "entropy": 5.57149920463562,
      "epoch": 1.0983882742741617,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.4493,
      "mean_token_accuracy": 0.18469329476356505,
      "num_tokens": 21833353.0,
      "step": 9780
    },
    {
      "entropy": 5.491009521484375,
      "epoch": 1.0989498511821194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1915326952934265,
      "num_tokens": 21843691.0,
      "step": 9785
    },
    {
      "entropy": 5.468305063247681,
      "epoch": 1.0995114280900768,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.3459,
      "mean_token_accuracy": 0.19110767245292665,
      "num_tokens": 21853771.0,
      "step": 9790
    },
    {
      "entropy": 5.413413190841675,
      "epoch": 1.1000730049980345,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19955267161130905,
      "num_tokens": 21863424.0,
      "step": 9795
    },
    {
      "entropy": 5.507965183258056,
      "epoch": 1.100634581905992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.3446,
      "mean_token_accuracy": 0.18554313480854034,
      "num_tokens": 21875407.0,
      "step": 9800
    },
    {
      "entropy": 5.513894748687744,
      "epoch": 1.1011961588139496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.4146,
      "mean_token_accuracy": 0.18251552283763886,
      "num_tokens": 21887081.0,
      "step": 9805
    },
    {
      "entropy": 5.485643911361694,
      "epoch": 1.101757735721907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.3215,
      "mean_token_accuracy": 0.18611925095319748,
      "num_tokens": 21897858.0,
      "step": 9810
    },
    {
      "entropy": 5.50360255241394,
      "epoch": 1.1023193126298647,
      "grad_norm": 1.125,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.367,
      "mean_token_accuracy": 0.1846587523818016,
      "num_tokens": 21907985.0,
      "step": 9815
    },
    {
      "entropy": 5.49332365989685,
      "epoch": 1.1028808895378222,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048894798004696,
      "loss": 5.4128,
      "mean_token_accuracy": 0.18538483530282973,
      "num_tokens": 21918726.0,
      "step": 9820
    },
    {
      "entropy": 5.463434791564941,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18947667479515076,
      "num_tokens": 21929027.0,
      "step": 9825
    },
    {
      "entropy": 5.456523656845093,
      "epoch": 1.1040040433537373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.257,
      "mean_token_accuracy": 0.19990183860063554,
      "num_tokens": 21938356.0,
      "step": 9830
    },
    {
      "entropy": 5.46082034111023,
      "epoch": 1.104565620261695,
      "grad_norm": 1.125,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.2924,
      "mean_token_accuracy": 0.19054461866617203,
      "num_tokens": 21947889.0,
      "step": 9835
    },
    {
      "entropy": 5.50875072479248,
      "epoch": 1.1051271971696524,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.3827,
      "mean_token_accuracy": 0.18880287557840347,
      "num_tokens": 21958409.0,
      "step": 9840
    },
    {
      "entropy": 5.438988304138183,
      "epoch": 1.10568877407761,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.3542,
      "mean_token_accuracy": 0.18572020828723906,
      "num_tokens": 21970120.0,
      "step": 9845
    },
    {
      "entropy": 5.491817474365234,
      "epoch": 1.1062503509855675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.36,
      "mean_token_accuracy": 0.1902606576681137,
      "num_tokens": 21981499.0,
      "step": 9850
    },
    {
      "entropy": 5.42215747833252,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.2482,
      "mean_token_accuracy": 0.20036101937294007,
      "num_tokens": 21991263.0,
      "step": 9855
    },
    {
      "entropy": 5.460143899917602,
      "epoch": 1.1073735048014826,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.2468,
      "mean_token_accuracy": 0.1842061311006546,
      "num_tokens": 22001582.0,
      "step": 9860
    },
    {
      "entropy": 5.446298122406006,
      "epoch": 1.10793508170944,
      "grad_norm": 1.140625,
      "learning_rate": 0.000488835848264024,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19687162339687347,
      "num_tokens": 22012447.0,
      "step": 9865
    },
    {
      "entropy": 5.496721887588501,
      "epoch": 1.1084966586173977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18907029330730438,
      "num_tokens": 22024865.0,
      "step": 9870
    },
    {
      "entropy": 5.469445371627808,
      "epoch": 1.109058235525355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.3311,
      "mean_token_accuracy": 0.190682515501976,
      "num_tokens": 22036298.0,
      "step": 9875
    },
    {
      "entropy": 5.547363233566284,
      "epoch": 1.1096198124333128,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1882574364542961,
      "num_tokens": 22047499.0,
      "step": 9880
    },
    {
      "entropy": 5.517907047271729,
      "epoch": 1.1101813893412702,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.3848,
      "mean_token_accuracy": 0.18664682954549788,
      "num_tokens": 22057612.0,
      "step": 9885
    },
    {
      "entropy": 5.4974652290344235,
      "epoch": 1.1107429662492279,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000488773314508967,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18530358076095582,
      "num_tokens": 22068566.0,
      "step": 9890
    },
    {
      "entropy": 5.473715686798096,
      "epoch": 1.1113045431571853,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.2856,
      "mean_token_accuracy": 0.1880566105246544,
      "num_tokens": 22079821.0,
      "step": 9895
    },
    {
      "entropy": 5.489977693557739,
      "epoch": 1.111866120065143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.335,
      "mean_token_accuracy": 0.1866757184267044,
      "num_tokens": 22090143.0,
      "step": 9900
    },
    {
      "entropy": 5.445848846435547,
      "epoch": 1.1124276969731004,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.395,
      "mean_token_accuracy": 0.18721785098314286,
      "num_tokens": 22101601.0,
      "step": 9905
    },
    {
      "entropy": 5.50714430809021,
      "epoch": 1.112989273881058,
      "grad_norm": 1.109375,
      "learning_rate": 0.000488723164983281,
      "loss": 5.3121,
      "mean_token_accuracy": 0.1923036679625511,
      "num_tokens": 22113102.0,
      "step": 9910
    },
    {
      "entropy": 5.5688591480255125,
      "epoch": 1.1135508507890155,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18365613073110582,
      "num_tokens": 22124650.0,
      "step": 9915
    },
    {
      "entropy": 5.545090818405152,
      "epoch": 1.1141124276969732,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.388,
      "mean_token_accuracy": 0.1909029722213745,
      "num_tokens": 22136637.0,
      "step": 9920
    },
    {
      "entropy": 5.465398502349854,
      "epoch": 1.1146740046049306,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.3099,
      "mean_token_accuracy": 0.19008917212486268,
      "num_tokens": 22149318.0,
      "step": 9925
    },
    {
      "entropy": 5.384241342544556,
      "epoch": 1.1152355815128883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.1965,
      "mean_token_accuracy": 0.19351300746202468,
      "num_tokens": 22159410.0,
      "step": 9930
    },
    {
      "entropy": 5.425002002716065,
      "epoch": 1.1157971584208457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.2794,
      "mean_token_accuracy": 0.18795328289270402,
      "num_tokens": 22170017.0,
      "step": 9935
    },
    {
      "entropy": 5.449740934371948,
      "epoch": 1.1163587353288034,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.2828,
      "mean_token_accuracy": 0.1909773141145706,
      "num_tokens": 22179961.0,
      "step": 9940
    },
    {
      "entropy": 5.555003643035889,
      "epoch": 1.1169203122367608,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1794225588440895,
      "num_tokens": 22191415.0,
      "step": 9945
    },
    {
      "entropy": 5.509974718093872,
      "epoch": 1.1174818891447185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.3652,
      "mean_token_accuracy": 0.19871425330638887,
      "num_tokens": 22205354.0,
      "step": 9950
    },
    {
      "entropy": 5.595667791366577,
      "epoch": 1.118043466052676,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.4172,
      "mean_token_accuracy": 0.1836776241660118,
      "num_tokens": 22216989.0,
      "step": 9955
    },
    {
      "entropy": 5.58510389328003,
      "epoch": 1.1186050429606333,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.4163,
      "mean_token_accuracy": 0.18037967681884765,
      "num_tokens": 22228421.0,
      "step": 9960
    },
    {
      "entropy": 5.463361406326294,
      "epoch": 1.119166619868591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.2873,
      "mean_token_accuracy": 0.18683721423149108,
      "num_tokens": 22240621.0,
      "step": 9965
    },
    {
      "entropy": 5.410719871520996,
      "epoch": 1.1197281967765484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.2622,
      "mean_token_accuracy": 0.19615741223096847,
      "num_tokens": 22250936.0,
      "step": 9970
    },
    {
      "entropy": 5.475217628479004,
      "epoch": 1.120289773684506,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.3882,
      "mean_token_accuracy": 0.18656658232212067,
      "num_tokens": 22261766.0,
      "step": 9975
    },
    {
      "entropy": 5.475364446640015,
      "epoch": 1.1208513505924635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048854678437799,
      "loss": 5.319,
      "mean_token_accuracy": 0.1898321345448494,
      "num_tokens": 22271831.0,
      "step": 9980
    },
    {
      "entropy": 5.427309131622314,
      "epoch": 1.1214129275004212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.3458,
      "mean_token_accuracy": 0.18728182464838028,
      "num_tokens": 22283756.0,
      "step": 9985
    },
    {
      "entropy": 5.4209051609039305,
      "epoch": 1.1219745044083786,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.2965,
      "mean_token_accuracy": 0.18717077374458313,
      "num_tokens": 22294576.0,
      "step": 9990
    },
    {
      "entropy": 5.495046854019165,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18578314632177353,
      "num_tokens": 22305003.0,
      "step": 9995
    },
    {
      "entropy": 5.46302809715271,
      "epoch": 1.1230976582242937,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.3157,
      "mean_token_accuracy": 0.19266755282878875,
      "num_tokens": 22316505.0,
      "step": 10000
    },
    {
      "entropy": 5.515137434005737,
      "epoch": 1.1236592351322514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.3772,
      "mean_token_accuracy": 0.19046587198972703,
      "num_tokens": 22327853.0,
      "step": 10005
    },
    {
      "entropy": 5.558890104293823,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.2952,
      "mean_token_accuracy": 0.188982817530632,
      "num_tokens": 22337386.0,
      "step": 10010
    },
    {
      "entropy": 5.539405822753906,
      "epoch": 1.1247823889481665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.4759,
      "mean_token_accuracy": 0.18247807174921035,
      "num_tokens": 22348742.0,
      "step": 10015
    },
    {
      "entropy": 5.583350753784179,
      "epoch": 1.125343965856124,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1852103129029274,
      "num_tokens": 22360083.0,
      "step": 10020
    },
    {
      "entropy": 5.500770807266235,
      "epoch": 1.1259055427640816,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000488432693037651,
      "loss": 5.3609,
      "mean_token_accuracy": 0.1815495163202286,
      "num_tokens": 22370982.0,
      "step": 10025
    },
    {
      "entropy": 5.451694679260254,
      "epoch": 1.126467119672039,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000488419982237771,
      "loss": 5.3063,
      "mean_token_accuracy": 0.20067557394504548,
      "num_tokens": 22382383.0,
      "step": 10030
    },
    {
      "entropy": 5.519569683074951,
      "epoch": 1.1270286965799967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.3186,
      "mean_token_accuracy": 0.18760548233985902,
      "num_tokens": 22393994.0,
      "step": 10035
    },
    {
      "entropy": 5.419602727890014,
      "epoch": 1.1275902734879542,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19836754649877547,
      "num_tokens": 22405354.0,
      "step": 10040
    },
    {
      "entropy": 5.387629604339599,
      "epoch": 1.1281518503959118,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.2984,
      "mean_token_accuracy": 0.18958158791065216,
      "num_tokens": 22416708.0,
      "step": 10045
    },
    {
      "entropy": 5.441644954681396,
      "epoch": 1.1287134273038693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19262771159410477,
      "num_tokens": 22427724.0,
      "step": 10050
    },
    {
      "entropy": 5.5489989757537845,
      "epoch": 1.1292750042118267,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.3878,
      "mean_token_accuracy": 0.18374199718236922,
      "num_tokens": 22437823.0,
      "step": 10055
    },
    {
      "entropy": 5.523835563659668,
      "epoch": 1.1298365811197844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1875823125243187,
      "num_tokens": 22450155.0,
      "step": 10060
    },
    {
      "entropy": 5.53595986366272,
      "epoch": 1.130398158027742,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19261740893125534,
      "num_tokens": 22460484.0,
      "step": 10065
    },
    {
      "entropy": 5.434231376647949,
      "epoch": 1.1309597349356995,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18646218478679658,
      "num_tokens": 22472063.0,
      "step": 10070
    },
    {
      "entropy": 5.421738958358764,
      "epoch": 1.131521311843657,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19664047062397003,
      "num_tokens": 22483669.0,
      "step": 10075
    },
    {
      "entropy": 5.559923267364502,
      "epoch": 1.1320828887516146,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18587087243795394,
      "num_tokens": 22495505.0,
      "step": 10080
    },
    {
      "entropy": 5.499471282958984,
      "epoch": 1.132644465659572,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.3628,
      "mean_token_accuracy": 0.18663066774606704,
      "num_tokens": 22505260.0,
      "step": 10085
    },
    {
      "entropy": 5.515253400802612,
      "epoch": 1.1332060425675297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.4206,
      "mean_token_accuracy": 0.1849902629852295,
      "num_tokens": 22516362.0,
      "step": 10090
    },
    {
      "entropy": 5.570123624801636,
      "epoch": 1.133767619475487,
      "grad_norm": 1.125,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.3758,
      "mean_token_accuracy": 0.18518655598163605,
      "num_tokens": 22527760.0,
      "step": 10095
    },
    {
      "entropy": 5.689021730422974,
      "epoch": 1.1343291963834448,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.4669,
      "mean_token_accuracy": 0.17877719551324844,
      "num_tokens": 22540601.0,
      "step": 10100
    },
    {
      "entropy": 5.5218689918518065,
      "epoch": 1.1348907732914022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.3798,
      "mean_token_accuracy": 0.18207437992095948,
      "num_tokens": 22552469.0,
      "step": 10105
    },
    {
      "entropy": 5.379859828948975,
      "epoch": 1.1354523501993599,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.3306,
      "mean_token_accuracy": 0.1915702611207962,
      "num_tokens": 22563552.0,
      "step": 10110
    },
    {
      "entropy": 5.4801393985748295,
      "epoch": 1.1360139271073173,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.306,
      "mean_token_accuracy": 0.19141027182340623,
      "num_tokens": 22575619.0,
      "step": 10115
    },
    {
      "entropy": 5.467701959609985,
      "epoch": 1.136575504015275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1910892367362976,
      "num_tokens": 22585854.0,
      "step": 10120
    },
    {
      "entropy": 5.524414014816284,
      "epoch": 1.1371370809232324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.3095,
      "mean_token_accuracy": 0.19435176998376846,
      "num_tokens": 22595899.0,
      "step": 10125
    },
    {
      "entropy": 5.407589101791382,
      "epoch": 1.13769865783119,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19016751348972322,
      "num_tokens": 22607517.0,
      "step": 10130
    },
    {
      "entropy": 5.501319694519043,
      "epoch": 1.1382602347391475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.4031,
      "mean_token_accuracy": 0.18679691404104232,
      "num_tokens": 22618531.0,
      "step": 10135
    },
    {
      "entropy": 5.388666439056396,
      "epoch": 1.1388218116471052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.2495,
      "mean_token_accuracy": 0.1937074512243271,
      "num_tokens": 22629193.0,
      "step": 10140
    },
    {
      "entropy": 5.523422384262085,
      "epoch": 1.1393833885550626,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.3675,
      "mean_token_accuracy": 0.19131201058626174,
      "num_tokens": 22639794.0,
      "step": 10145
    },
    {
      "entropy": 5.546735048294067,
      "epoch": 1.13994496546302,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.3187,
      "mean_token_accuracy": 0.1938394457101822,
      "num_tokens": 22650061.0,
      "step": 10150
    },
    {
      "entropy": 5.480262565612793,
      "epoch": 1.1405065423709777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048810000479995,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18121499419212342,
      "num_tokens": 22661399.0,
      "step": 10155
    },
    {
      "entropy": 5.635375690460205,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.4886,
      "mean_token_accuracy": 0.18087699860334397,
      "num_tokens": 22672445.0,
      "step": 10160
    },
    {
      "entropy": 5.555114841461181,
      "epoch": 1.1416296961868928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.2962,
      "mean_token_accuracy": 0.19056684970855714,
      "num_tokens": 22682545.0,
      "step": 10165
    },
    {
      "entropy": 5.468265962600708,
      "epoch": 1.1421912730948502,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.3508,
      "mean_token_accuracy": 0.1824083372950554,
      "num_tokens": 22693885.0,
      "step": 10170
    },
    {
      "entropy": 5.4522919178009035,
      "epoch": 1.142752850002808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.2971,
      "mean_token_accuracy": 0.18848683089017867,
      "num_tokens": 22703841.0,
      "step": 10175
    },
    {
      "entropy": 5.590392589569092,
      "epoch": 1.1433144269107653,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.4413,
      "mean_token_accuracy": 0.1882365733385086,
      "num_tokens": 22714142.0,
      "step": 10180
    },
    {
      "entropy": 5.53915376663208,
      "epoch": 1.143876003818723,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.3155,
      "mean_token_accuracy": 0.18552613705396653,
      "num_tokens": 22725056.0,
      "step": 10185
    },
    {
      "entropy": 5.4065001010894775,
      "epoch": 1.1444375807266804,
      "grad_norm": 1.3125,
      "learning_rate": 0.000488009650901768,
      "loss": 5.3452,
      "mean_token_accuracy": 0.1924957364797592,
      "num_tokens": 22735664.0,
      "step": 10190
    },
    {
      "entropy": 5.450916433334351,
      "epoch": 1.144999157634638,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19113609790802003,
      "num_tokens": 22748071.0,
      "step": 10195
    },
    {
      "entropy": 5.533325099945069,
      "epoch": 1.1455607345425955,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.3782,
      "mean_token_accuracy": 0.1861809179186821,
      "num_tokens": 22759477.0,
      "step": 10200
    },
    {
      "entropy": 5.493514156341552,
      "epoch": 1.1461223114505532,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.2539,
      "mean_token_accuracy": 0.1958856463432312,
      "num_tokens": 22769955.0,
      "step": 10205
    },
    {
      "entropy": 5.419086170196533,
      "epoch": 1.1466838883585107,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.3503,
      "mean_token_accuracy": 0.18614689409732818,
      "num_tokens": 22781216.0,
      "step": 10210
    },
    {
      "entropy": 5.436713027954101,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.2921,
      "mean_token_accuracy": 0.18933203518390657,
      "num_tokens": 22792581.0,
      "step": 10215
    },
    {
      "entropy": 5.422752380371094,
      "epoch": 1.1478070421744258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.2549,
      "mean_token_accuracy": 0.19116601198911667,
      "num_tokens": 22803305.0,
      "step": 10220
    },
    {
      "entropy": 5.560664892196655,
      "epoch": 1.1483686190823834,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.3888,
      "mean_token_accuracy": 0.18957023620605468,
      "num_tokens": 22815558.0,
      "step": 10225
    },
    {
      "entropy": 5.477578258514404,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19268295168876648,
      "num_tokens": 22825428.0,
      "step": 10230
    },
    {
      "entropy": 5.495377683639527,
      "epoch": 1.1494917728982985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19178205877542495,
      "num_tokens": 22836294.0,
      "step": 10235
    },
    {
      "entropy": 5.4914785861969,
      "epoch": 1.150053349806256,
      "grad_norm": 1.25,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.4317,
      "mean_token_accuracy": 0.19050806164741516,
      "num_tokens": 22847835.0,
      "step": 10240
    },
    {
      "entropy": 5.513508224487305,
      "epoch": 1.1506149267142134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.3659,
      "mean_token_accuracy": 0.18653237372636794,
      "num_tokens": 22859726.0,
      "step": 10245
    },
    {
      "entropy": 5.433402252197266,
      "epoch": 1.151176503622171,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.3195,
      "mean_token_accuracy": 0.18754699677228928,
      "num_tokens": 22871250.0,
      "step": 10250
    },
    {
      "entropy": 5.541570472717285,
      "epoch": 1.1517380805301287,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.3975,
      "mean_token_accuracy": 0.1847960025072098,
      "num_tokens": 22882034.0,
      "step": 10255
    },
    {
      "entropy": 5.631517171859741,
      "epoch": 1.1522996574380862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.4759,
      "mean_token_accuracy": 0.18365521728992462,
      "num_tokens": 22894606.0,
      "step": 10260
    },
    {
      "entropy": 5.440877389907837,
      "epoch": 1.1528612343460436,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.3115,
      "mean_token_accuracy": 0.19253068268299103,
      "num_tokens": 22905246.0,
      "step": 10265
    },
    {
      "entropy": 5.446545505523682,
      "epoch": 1.1534228112540013,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.2604,
      "mean_token_accuracy": 0.18477636873722075,
      "num_tokens": 22916202.0,
      "step": 10270
    },
    {
      "entropy": 5.470248508453369,
      "epoch": 1.1539843881619587,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.2578,
      "mean_token_accuracy": 0.18614826798439027,
      "num_tokens": 22926451.0,
      "step": 10275
    },
    {
      "entropy": 5.520329093933105,
      "epoch": 1.1545459650699164,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.348,
      "mean_token_accuracy": 0.19069368839263917,
      "num_tokens": 22936618.0,
      "step": 10280
    },
    {
      "entropy": 5.560961484909058,
      "epoch": 1.1551075419778738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.381,
      "mean_token_accuracy": 0.17827016711235047,
      "num_tokens": 22948278.0,
      "step": 10285
    },
    {
      "entropy": 5.44008960723877,
      "epoch": 1.1556691188858315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.2742,
      "mean_token_accuracy": 0.19207292199134826,
      "num_tokens": 22959928.0,
      "step": 10290
    },
    {
      "entropy": 5.384089946746826,
      "epoch": 1.156230695793789,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000487736595774831,
      "loss": 5.3298,
      "mean_token_accuracy": 0.1838379830121994,
      "num_tokens": 22971050.0,
      "step": 10295
    },
    {
      "entropy": 5.41632490158081,
      "epoch": 1.1567922727017466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1910719946026802,
      "num_tokens": 22981471.0,
      "step": 10300
    },
    {
      "entropy": 5.601700258255005,
      "epoch": 1.157353849609704,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.4813,
      "mean_token_accuracy": 0.18961743414402008,
      "num_tokens": 22992526.0,
      "step": 10305
    },
    {
      "entropy": 5.472655868530273,
      "epoch": 1.1579154265176617,
      "grad_norm": 1.015625,
      "learning_rate": 0.000487697343959723,
      "loss": 5.3413,
      "mean_token_accuracy": 0.18970987647771836,
      "num_tokens": 23003408.0,
      "step": 10310
    },
    {
      "entropy": 5.460650539398193,
      "epoch": 1.158477003425619,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1915055900812149,
      "num_tokens": 23013406.0,
      "step": 10315
    },
    {
      "entropy": 5.437350940704346,
      "epoch": 1.1590385803335768,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.3264,
      "mean_token_accuracy": 0.19100383818149566,
      "num_tokens": 23024960.0,
      "step": 10320
    },
    {
      "entropy": 5.54367356300354,
      "epoch": 1.1596001572415342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487658031193323,
      "loss": 5.3505,
      "mean_token_accuracy": 0.1825086921453476,
      "num_tokens": 23037514.0,
      "step": 10325
    },
    {
      "entropy": 5.5518410205841064,
      "epoch": 1.1601617341494919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.4485,
      "mean_token_accuracy": 0.17935476303100586,
      "num_tokens": 23048460.0,
      "step": 10330
    },
    {
      "entropy": 5.54437141418457,
      "epoch": 1.1607233110574493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.3051,
      "mean_token_accuracy": 0.1868131637573242,
      "num_tokens": 23058866.0,
      "step": 10335
    },
    {
      "entropy": 5.425964832305908,
      "epoch": 1.1612848879654067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.34,
      "mean_token_accuracy": 0.1852980837225914,
      "num_tokens": 23069097.0,
      "step": 10340
    },
    {
      "entropy": 5.55105037689209,
      "epoch": 1.1618464648733644,
      "grad_norm": 1.125,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.3217,
      "mean_token_accuracy": 0.18968528807163237,
      "num_tokens": 23079553.0,
      "step": 10345
    },
    {
      "entropy": 5.473508358001709,
      "epoch": 1.162408041781322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.3361,
      "mean_token_accuracy": 0.18588747680187226,
      "num_tokens": 23090554.0,
      "step": 10350
    },
    {
      "entropy": 5.41131443977356,
      "epoch": 1.1629696186892795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.2312,
      "mean_token_accuracy": 0.190905624628067,
      "num_tokens": 23101740.0,
      "step": 10355
    },
    {
      "entropy": 5.473917245864868,
      "epoch": 1.163531195597237,
      "grad_norm": 1.5,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18988496959209442,
      "num_tokens": 23112842.0,
      "step": 10360
    },
    {
      "entropy": 5.434205722808838,
      "epoch": 1.1640927725051946,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.3221,
      "mean_token_accuracy": 0.1870110347867012,
      "num_tokens": 23123822.0,
      "step": 10365
    },
    {
      "entropy": 5.526471328735352,
      "epoch": 1.164654349413152,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.3419,
      "mean_token_accuracy": 0.18134610950946808,
      "num_tokens": 23134045.0,
      "step": 10370
    },
    {
      "entropy": 5.447522306442261,
      "epoch": 1.1652159263211097,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487526548579656,
      "loss": 5.3734,
      "mean_token_accuracy": 0.18452945798635484,
      "num_tokens": 23144200.0,
      "step": 10375
    },
    {
      "entropy": 5.455108690261841,
      "epoch": 1.1657775032290671,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.307,
      "mean_token_accuracy": 0.192894583940506,
      "num_tokens": 23155671.0,
      "step": 10380
    },
    {
      "entropy": 5.567395973205566,
      "epoch": 1.1663390801370248,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18633103370666504,
      "num_tokens": 23166644.0,
      "step": 10385
    },
    {
      "entropy": 5.525482654571533,
      "epoch": 1.1669006570449822,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1920440375804901,
      "num_tokens": 23177074.0,
      "step": 10390
    },
    {
      "entropy": 5.534948301315308,
      "epoch": 1.16746223395294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.4461,
      "mean_token_accuracy": 0.18229351192712784,
      "num_tokens": 23188724.0,
      "step": 10395
    },
    {
      "entropy": 5.466331672668457,
      "epoch": 1.1680238108608973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.3522,
      "mean_token_accuracy": 0.1878226652741432,
      "num_tokens": 23198459.0,
      "step": 10400
    },
    {
      "entropy": 5.553414821624756,
      "epoch": 1.168585387768855,
      "grad_norm": 1.078125,
      "learning_rate": 0.000487447334172127,
      "loss": 5.361,
      "mean_token_accuracy": 0.18774810135364534,
      "num_tokens": 23208991.0,
      "step": 10405
    },
    {
      "entropy": 5.479293727874756,
      "epoch": 1.1691469646768125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.2926,
      "mean_token_accuracy": 0.18909573256969453,
      "num_tokens": 23219309.0,
      "step": 10410
    },
    {
      "entropy": 5.473764657974243,
      "epoch": 1.1697085415847701,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.3993,
      "mean_token_accuracy": 0.18177039921283722,
      "num_tokens": 23230551.0,
      "step": 10415
    },
    {
      "entropy": 5.559957695007324,
      "epoch": 1.1702701184927276,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.461,
      "mean_token_accuracy": 0.18303379118442537,
      "num_tokens": 23242799.0,
      "step": 10420
    },
    {
      "entropy": 5.5983470439910885,
      "epoch": 1.1708316954006852,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.4458,
      "mean_token_accuracy": 0.179739548265934,
      "num_tokens": 23255112.0,
      "step": 10425
    },
    {
      "entropy": 5.47545485496521,
      "epoch": 1.1713932723086427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.3497,
      "mean_token_accuracy": 0.19257825911045073,
      "num_tokens": 23264800.0,
      "step": 10430
    },
    {
      "entropy": 5.449072074890137,
      "epoch": 1.1719548492166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000487367876246022,
      "loss": 5.2913,
      "mean_token_accuracy": 0.1865327313542366,
      "num_tokens": 23276672.0,
      "step": 10435
    },
    {
      "entropy": 5.482340717315674,
      "epoch": 1.1725164261245578,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18801960945129395,
      "num_tokens": 23288282.0,
      "step": 10440
    },
    {
      "entropy": 5.491970252990723,
      "epoch": 1.1730780030325154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.3468,
      "mean_token_accuracy": 0.18586714565753937,
      "num_tokens": 23299078.0,
      "step": 10445
    },
    {
      "entropy": 5.497157001495362,
      "epoch": 1.1736395799404729,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.3358,
      "mean_token_accuracy": 0.19111865609884263,
      "num_tokens": 23310036.0,
      "step": 10450
    },
    {
      "entropy": 5.4897033214569095,
      "epoch": 1.1742011568484303,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.268,
      "mean_token_accuracy": 0.19594541937112808,
      "num_tokens": 23321029.0,
      "step": 10455
    },
    {
      "entropy": 5.50123085975647,
      "epoch": 1.174762733756388,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.3796,
      "mean_token_accuracy": 0.1866510823369026,
      "num_tokens": 23331584.0,
      "step": 10460
    },
    {
      "entropy": 5.635983562469482,
      "epoch": 1.1753243106643454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.4467,
      "mean_token_accuracy": 0.17778573036193848,
      "num_tokens": 23343171.0,
      "step": 10465
    },
    {
      "entropy": 5.559485721588135,
      "epoch": 1.175885887572303,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18856250047683715,
      "num_tokens": 23353047.0,
      "step": 10470
    },
    {
      "entropy": 5.4335010051727295,
      "epoch": 1.1764474644802605,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18665113747119905,
      "num_tokens": 23364447.0,
      "step": 10475
    },
    {
      "entropy": 5.5190877437591555,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487248232958846,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18602234572172166,
      "num_tokens": 23375616.0,
      "step": 10480
    },
    {
      "entropy": 5.6089424133300785,
      "epoch": 1.1775706182961756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.4896,
      "mean_token_accuracy": 0.18161313533782958,
      "num_tokens": 23388176.0,
      "step": 10485
    },
    {
      "entropy": 5.583784341812134,
      "epoch": 1.1781321952041333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.4174,
      "mean_token_accuracy": 0.18961915075778962,
      "num_tokens": 23400563.0,
      "step": 10490
    },
    {
      "entropy": 5.504327201843262,
      "epoch": 1.1786937721120907,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.3573,
      "mean_token_accuracy": 0.18537813574075698,
      "num_tokens": 23410740.0,
      "step": 10495
    },
    {
      "entropy": 5.484878921508789,
      "epoch": 1.1792553490200484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.4446,
      "mean_token_accuracy": 0.18258217871189117,
      "num_tokens": 23423086.0,
      "step": 10500
    },
    {
      "entropy": 5.565518712997436,
      "epoch": 1.1798169259280058,
      "grad_norm": 1.171875,
      "learning_rate": 0.000487181527913898,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18355826139450074,
      "num_tokens": 23435869.0,
      "step": 10505
    },
    {
      "entropy": 5.514801931381226,
      "epoch": 1.1803785028359635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.2992,
      "mean_token_accuracy": 0.20019642412662506,
      "num_tokens": 23446635.0,
      "step": 10510
    },
    {
      "entropy": 5.476276874542236,
      "epoch": 1.180940079743921,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.3254,
      "mean_token_accuracy": 0.18937005400657653,
      "num_tokens": 23458214.0,
      "step": 10515
    },
    {
      "entropy": 5.51998610496521,
      "epoch": 1.1815016566518786,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.352,
      "mean_token_accuracy": 0.19283366650342942,
      "num_tokens": 23469383.0,
      "step": 10520
    },
    {
      "entropy": 5.449571895599365,
      "epoch": 1.182063233559836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.3195,
      "mean_token_accuracy": 0.19173895865678786,
      "num_tokens": 23480841.0,
      "step": 10525
    },
    {
      "entropy": 5.428569221496582,
      "epoch": 1.1826248104677934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.2743,
      "mean_token_accuracy": 0.1885039508342743,
      "num_tokens": 23490796.0,
      "step": 10530
    },
    {
      "entropy": 5.445909738540649,
      "epoch": 1.183186387375751,
      "grad_norm": 1.25,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.3153,
      "mean_token_accuracy": 0.1842427521944046,
      "num_tokens": 23501759.0,
      "step": 10535
    },
    {
      "entropy": 5.496493244171143,
      "epoch": 1.1837479642837088,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.3457,
      "mean_token_accuracy": 0.18848443180322647,
      "num_tokens": 23512209.0,
      "step": 10540
    },
    {
      "entropy": 5.517921113967896,
      "epoch": 1.1843095411916662,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.4205,
      "mean_token_accuracy": 0.1864883452653885,
      "num_tokens": 23524316.0,
      "step": 10545
    },
    {
      "entropy": 5.5062822818756105,
      "epoch": 1.1848711180996236,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1910739853978157,
      "num_tokens": 23536199.0,
      "step": 10550
    },
    {
      "entropy": 5.4898323059082035,
      "epoch": 1.1854326950075813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.3967,
      "mean_token_accuracy": 0.18637777268886566,
      "num_tokens": 23548034.0,
      "step": 10555
    },
    {
      "entropy": 5.551316785812378,
      "epoch": 1.1859942719155387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19020506888628005,
      "num_tokens": 23559280.0,
      "step": 10560
    },
    {
      "entropy": 5.5113687992095945,
      "epoch": 1.1865558488234964,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.3361,
      "mean_token_accuracy": 0.19059309214353562,
      "num_tokens": 23569163.0,
      "step": 10565
    },
    {
      "entropy": 5.510538244247437,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.396,
      "mean_token_accuracy": 0.18955078572034836,
      "num_tokens": 23580526.0,
      "step": 10570
    },
    {
      "entropy": 5.416756916046142,
      "epoch": 1.1876790026394115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.2844,
      "mean_token_accuracy": 0.18405936062335967,
      "num_tokens": 23590552.0,
      "step": 10575
    },
    {
      "entropy": 5.332808208465576,
      "epoch": 1.188240579547369,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19522219747304917,
      "num_tokens": 23601889.0,
      "step": 10580
    },
    {
      "entropy": 5.490883445739746,
      "epoch": 1.1888021564553266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.4437,
      "mean_token_accuracy": 0.19266334921121597,
      "num_tokens": 23613833.0,
      "step": 10585
    },
    {
      "entropy": 5.590994930267334,
      "epoch": 1.189363733363284,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.3926,
      "mean_token_accuracy": 0.19101812690496445,
      "num_tokens": 23625404.0,
      "step": 10590
    },
    {
      "entropy": 5.470618438720703,
      "epoch": 1.1899253102712417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19313479363918304,
      "num_tokens": 23635670.0,
      "step": 10595
    },
    {
      "entropy": 5.41665506362915,
      "epoch": 1.1904868871791991,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.2622,
      "mean_token_accuracy": 0.1956461876630783,
      "num_tokens": 23646893.0,
      "step": 10600
    },
    {
      "entropy": 5.395202207565307,
      "epoch": 1.1910484640871568,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1856496348977089,
      "num_tokens": 23657733.0,
      "step": 10605
    },
    {
      "entropy": 5.422374296188354,
      "epoch": 1.1916100409951143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048689952297394,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19550398737192154,
      "num_tokens": 23668100.0,
      "step": 10610
    },
    {
      "entropy": 5.540370607376099,
      "epoch": 1.192171617903072,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.3806,
      "mean_token_accuracy": 0.1858476459980011,
      "num_tokens": 23678295.0,
      "step": 10615
    },
    {
      "entropy": 5.485654497146607,
      "epoch": 1.1927331948110294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.2324,
      "mean_token_accuracy": 0.1951396644115448,
      "num_tokens": 23689310.0,
      "step": 10620
    },
    {
      "entropy": 5.436852025985718,
      "epoch": 1.1932947717189868,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.3966,
      "mean_token_accuracy": 0.18616850078105926,
      "num_tokens": 23699475.0,
      "step": 10625
    },
    {
      "entropy": 5.580835056304932,
      "epoch": 1.1938563486269445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.3794,
      "mean_token_accuracy": 0.18866919577121735,
      "num_tokens": 23710186.0,
      "step": 10630
    },
    {
      "entropy": 5.396511554718018,
      "epoch": 1.1944179255349021,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.1921,
      "mean_token_accuracy": 0.1933600828051567,
      "num_tokens": 23720449.0,
      "step": 10635
    },
    {
      "entropy": 5.5301899909973145,
      "epoch": 1.1949795024428596,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.452,
      "mean_token_accuracy": 0.18851781636476517,
      "num_tokens": 23732295.0,
      "step": 10640
    },
    {
      "entropy": 5.452880477905273,
      "epoch": 1.195541079350817,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.3516,
      "mean_token_accuracy": 0.1847606897354126,
      "num_tokens": 23743616.0,
      "step": 10645
    },
    {
      "entropy": 5.56827073097229,
      "epoch": 1.1961026562587747,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1836438223719597,
      "num_tokens": 23755288.0,
      "step": 10650
    },
    {
      "entropy": 5.506766748428345,
      "epoch": 1.196664233166732,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.2596,
      "mean_token_accuracy": 0.19082026481628417,
      "num_tokens": 23765916.0,
      "step": 10655
    },
    {
      "entropy": 5.3702747344970705,
      "epoch": 1.1972258100746898,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19341834783554077,
      "num_tokens": 23778891.0,
      "step": 10660
    },
    {
      "entropy": 5.4559619426727295,
      "epoch": 1.1977873869826472,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1918458342552185,
      "num_tokens": 23790402.0,
      "step": 10665
    },
    {
      "entropy": 5.436944055557251,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.3074,
      "mean_token_accuracy": 0.1992880582809448,
      "num_tokens": 23801928.0,
      "step": 10670
    },
    {
      "entropy": 5.55572361946106,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1797810286283493,
      "num_tokens": 23812302.0,
      "step": 10675
    },
    {
      "entropy": 5.549669694900513,
      "epoch": 1.19947211770652,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.3553,
      "mean_token_accuracy": 0.1992868646979332,
      "num_tokens": 23824033.0,
      "step": 10680
    },
    {
      "entropy": 5.441835021972656,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19039777517318726,
      "num_tokens": 23834907.0,
      "step": 10685
    },
    {
      "entropy": 5.53205418586731,
      "epoch": 1.200595271522435,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.4973,
      "mean_token_accuracy": 0.18706956952810289,
      "num_tokens": 23846418.0,
      "step": 10690
    },
    {
      "entropy": 5.594804286956787,
      "epoch": 1.2011568484303925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.426,
      "mean_token_accuracy": 0.18486818969249724,
      "num_tokens": 23858423.0,
      "step": 10695
    },
    {
      "entropy": 5.525947713851929,
      "epoch": 1.2017184253383502,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19153492003679276,
      "num_tokens": 23868340.0,
      "step": 10700
    },
    {
      "entropy": 5.427148675918579,
      "epoch": 1.2022800022463076,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.2579,
      "mean_token_accuracy": 0.19293180108070374,
      "num_tokens": 23878752.0,
      "step": 10705
    },
    {
      "entropy": 5.4996744155883786,
      "epoch": 1.2028415791542653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.318,
      "mean_token_accuracy": 0.19087314158678054,
      "num_tokens": 23889374.0,
      "step": 10710
    },
    {
      "entropy": 5.530946254730225,
      "epoch": 1.2034031560622227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.374,
      "mean_token_accuracy": 0.18851903080940247,
      "num_tokens": 23901651.0,
      "step": 10715
    },
    {
      "entropy": 5.414171886444092,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000486600897985503,
      "loss": 5.2956,
      "mean_token_accuracy": 0.1927599087357521,
      "num_tokens": 23911818.0,
      "step": 10720
    },
    {
      "entropy": 5.508558034896851,
      "epoch": 1.2045263098781378,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.342,
      "mean_token_accuracy": 0.19158872216939926,
      "num_tokens": 23923840.0,
      "step": 10725
    },
    {
      "entropy": 5.47757625579834,
      "epoch": 1.2050878867860955,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.2345,
      "mean_token_accuracy": 0.1947578877210617,
      "num_tokens": 23935406.0,
      "step": 10730
    },
    {
      "entropy": 5.48161211013794,
      "epoch": 1.205649463694053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.3391,
      "mean_token_accuracy": 0.1922893762588501,
      "num_tokens": 23946350.0,
      "step": 10735
    },
    {
      "entropy": 5.378890943527222,
      "epoch": 1.2062110406020103,
      "grad_norm": 1.125,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.2911,
      "mean_token_accuracy": 0.19203284680843352,
      "num_tokens": 23956315.0,
      "step": 10740
    },
    {
      "entropy": 5.4803853034973145,
      "epoch": 1.206772617509968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.2019,
      "mean_token_accuracy": 0.18931492567062377,
      "num_tokens": 23966649.0,
      "step": 10745
    },
    {
      "entropy": 5.5576153755187985,
      "epoch": 1.2073341944179254,
      "grad_norm": 1.125,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.4741,
      "mean_token_accuracy": 0.17762160748243333,
      "num_tokens": 23977674.0,
      "step": 10750
    },
    {
      "entropy": 5.442694044113159,
      "epoch": 1.207895771325883,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.3609,
      "mean_token_accuracy": 0.18723692148923873,
      "num_tokens": 23987985.0,
      "step": 10755
    },
    {
      "entropy": 5.5725297927856445,
      "epoch": 1.2084573482338405,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.4555,
      "mean_token_accuracy": 0.18374131619930267,
      "num_tokens": 23998730.0,
      "step": 10760
    },
    {
      "entropy": 5.442189359664917,
      "epoch": 1.2090189251417982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.2953,
      "mean_token_accuracy": 0.18207336962223053,
      "num_tokens": 24008743.0,
      "step": 10765
    },
    {
      "entropy": 5.4576087474823,
      "epoch": 1.2095805020497556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.3407,
      "mean_token_accuracy": 0.18913229405879975,
      "num_tokens": 24019805.0,
      "step": 10770
    },
    {
      "entropy": 5.501690101623535,
      "epoch": 1.2101420789577133,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18727027773857116,
      "num_tokens": 24031437.0,
      "step": 10775
    },
    {
      "entropy": 5.5729974746704105,
      "epoch": 1.2107036558656707,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.4505,
      "mean_token_accuracy": 0.1780677929520607,
      "num_tokens": 24043922.0,
      "step": 10780
    },
    {
      "entropy": 5.526761817932129,
      "epoch": 1.2112652327736284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.2709,
      "mean_token_accuracy": 0.1968682363629341,
      "num_tokens": 24054755.0,
      "step": 10785
    },
    {
      "entropy": 5.44307689666748,
      "epoch": 1.2118268096815858,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.2823,
      "mean_token_accuracy": 0.18202337771654128,
      "num_tokens": 24067347.0,
      "step": 10790
    },
    {
      "entropy": 5.503612089157104,
      "epoch": 1.2123883865895435,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.3385,
      "mean_token_accuracy": 0.19037413448095322,
      "num_tokens": 24078938.0,
      "step": 10795
    },
    {
      "entropy": 5.5415795803070065,
      "epoch": 1.212949963497501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.355,
      "mean_token_accuracy": 0.18644218295812606,
      "num_tokens": 24090275.0,
      "step": 10800
    },
    {
      "entropy": 5.482690811157227,
      "epoch": 1.2135115404054586,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.3654,
      "mean_token_accuracy": 0.19203082770109176,
      "num_tokens": 24101518.0,
      "step": 10805
    },
    {
      "entropy": 5.555949640274048,
      "epoch": 1.214073117313416,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18214682191610337,
      "num_tokens": 24113320.0,
      "step": 10810
    },
    {
      "entropy": 5.512860536575317,
      "epoch": 1.2146346942213737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19154454469680787,
      "num_tokens": 24124842.0,
      "step": 10815
    },
    {
      "entropy": 5.527637052536011,
      "epoch": 1.2151962711293312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19004831165075303,
      "num_tokens": 24136391.0,
      "step": 10820
    },
    {
      "entropy": 5.518775367736817,
      "epoch": 1.2157578480372888,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.3768,
      "mean_token_accuracy": 0.18957658410072326,
      "num_tokens": 24147194.0,
      "step": 10825
    },
    {
      "entropy": 5.56171498298645,
      "epoch": 1.2163194249452463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.4485,
      "mean_token_accuracy": 0.18261944651603698,
      "num_tokens": 24158170.0,
      "step": 10830
    },
    {
      "entropy": 5.516436910629272,
      "epoch": 1.2168810018532037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1917682930827141,
      "num_tokens": 24168671.0,
      "step": 10835
    },
    {
      "entropy": 5.451374912261963,
      "epoch": 1.2174425787611614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.3267,
      "mean_token_accuracy": 0.19494537264108658,
      "num_tokens": 24179898.0,
      "step": 10840
    },
    {
      "entropy": 5.390147399902344,
      "epoch": 1.2180041556691188,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19178963899612428,
      "num_tokens": 24190622.0,
      "step": 10845
    },
    {
      "entropy": 5.666590404510498,
      "epoch": 1.2185657325770765,
      "grad_norm": 1.125,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.4901,
      "mean_token_accuracy": 0.18107927292585374,
      "num_tokens": 24202526.0,
      "step": 10850
    },
    {
      "entropy": 5.502965974807739,
      "epoch": 1.219127309485034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.3159,
      "mean_token_accuracy": 0.18561350256204606,
      "num_tokens": 24213186.0,
      "step": 10855
    },
    {
      "entropy": 5.423367118835449,
      "epoch": 1.2196888863929916,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.3253,
      "mean_token_accuracy": 0.1877934679389,
      "num_tokens": 24225490.0,
      "step": 10860
    },
    {
      "entropy": 5.504482269287109,
      "epoch": 1.220250463300949,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.3113,
      "mean_token_accuracy": 0.19293032884597777,
      "num_tokens": 24236653.0,
      "step": 10865
    },
    {
      "entropy": 5.50580325126648,
      "epoch": 1.2208120402089067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1966002181172371,
      "num_tokens": 24247843.0,
      "step": 10870
    },
    {
      "entropy": 5.400511264801025,
      "epoch": 1.221373617116864,
      "grad_norm": 1.125,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.3601,
      "mean_token_accuracy": 0.19113782048225403,
      "num_tokens": 24258332.0,
      "step": 10875
    },
    {
      "entropy": 5.492725467681884,
      "epoch": 1.2219351940248218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.3457,
      "mean_token_accuracy": 0.1893364816904068,
      "num_tokens": 24268685.0,
      "step": 10880
    },
    {
      "entropy": 5.556686973571777,
      "epoch": 1.2224967709327792,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.3703,
      "mean_token_accuracy": 0.18632747530937194,
      "num_tokens": 24280660.0,
      "step": 10885
    },
    {
      "entropy": 5.531473541259766,
      "epoch": 1.2230583478407369,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19230025857686997,
      "num_tokens": 24291649.0,
      "step": 10890
    },
    {
      "entropy": 5.449414205551148,
      "epoch": 1.2236199247486943,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.3174,
      "mean_token_accuracy": 0.19521680921316148,
      "num_tokens": 24303108.0,
      "step": 10895
    },
    {
      "entropy": 5.438139867782593,
      "epoch": 1.224181501656652,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.3431,
      "mean_token_accuracy": 0.183199942111969,
      "num_tokens": 24315939.0,
      "step": 10900
    },
    {
      "entropy": 5.492931270599366,
      "epoch": 1.2247430785646094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.3215,
      "mean_token_accuracy": 0.19137313961982727,
      "num_tokens": 24326916.0,
      "step": 10905
    },
    {
      "entropy": 5.507109022140503,
      "epoch": 1.225304655472567,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19142188876867294,
      "num_tokens": 24337790.0,
      "step": 10910
    },
    {
      "entropy": 5.408120346069336,
      "epoch": 1.2258662323805245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19919081181287765,
      "num_tokens": 24347930.0,
      "step": 10915
    },
    {
      "entropy": 5.454737520217895,
      "epoch": 1.2264278092884822,
      "grad_norm": 1.125,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.3036,
      "mean_token_accuracy": 0.18901157826185228,
      "num_tokens": 24359213.0,
      "step": 10920
    },
    {
      "entropy": 5.445453786849976,
      "epoch": 1.2269893861964396,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1886252284049988,
      "num_tokens": 24371000.0,
      "step": 10925
    },
    {
      "entropy": 5.448771858215332,
      "epoch": 1.227550963104397,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.1989,
      "mean_token_accuracy": 0.19589506089687347,
      "num_tokens": 24381301.0,
      "step": 10930
    },
    {
      "entropy": 5.559473848342895,
      "epoch": 1.2281125400123547,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.4411,
      "mean_token_accuracy": 0.17990793585777282,
      "num_tokens": 24391808.0,
      "step": 10935
    },
    {
      "entropy": 5.6326392650604244,
      "epoch": 1.2286741169203121,
      "grad_norm": 1.0,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.5473,
      "mean_token_accuracy": 0.18333044052124023,
      "num_tokens": 24404220.0,
      "step": 10940
    },
    {
      "entropy": 5.559707450866699,
      "epoch": 1.2292356938282698,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.3616,
      "mean_token_accuracy": 0.1912491202354431,
      "num_tokens": 24414906.0,
      "step": 10945
    },
    {
      "entropy": 5.422472763061523,
      "epoch": 1.2297972707362272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.3387,
      "mean_token_accuracy": 0.18478153496980668,
      "num_tokens": 24426469.0,
      "step": 10950
    },
    {
      "entropy": 5.3907982349395756,
      "epoch": 1.230358847644185,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.298,
      "mean_token_accuracy": 0.19148481339216233,
      "num_tokens": 24437683.0,
      "step": 10955
    },
    {
      "entropy": 5.4061651706695555,
      "epoch": 1.2309204245521423,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19314789026975632,
      "num_tokens": 24449038.0,
      "step": 10960
    },
    {
      "entropy": 5.50235333442688,
      "epoch": 1.2314820014601,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.3554,
      "mean_token_accuracy": 0.18964497298002242,
      "num_tokens": 24460228.0,
      "step": 10965
    },
    {
      "entropy": 5.462199687957764,
      "epoch": 1.2320435783680574,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.3812,
      "mean_token_accuracy": 0.18497704714536667,
      "num_tokens": 24471222.0,
      "step": 10970
    },
    {
      "entropy": 5.413447046279908,
      "epoch": 1.232605155276015,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.3304,
      "mean_token_accuracy": 0.18635518997907638,
      "num_tokens": 24482127.0,
      "step": 10975
    },
    {
      "entropy": 5.511392974853516,
      "epoch": 1.2331667321839725,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1897481694817543,
      "num_tokens": 24492963.0,
      "step": 10980
    },
    {
      "entropy": 5.548450136184693,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19336990565061568,
      "num_tokens": 24502885.0,
      "step": 10985
    },
    {
      "entropy": 5.5679937362670895,
      "epoch": 1.2342898859998876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.438,
      "mean_token_accuracy": 0.18847732096910477,
      "num_tokens": 24514990.0,
      "step": 10990
    },
    {
      "entropy": 5.528681898117066,
      "epoch": 1.2348514629078453,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.3529,
      "mean_token_accuracy": 0.18421241790056228,
      "num_tokens": 24525555.0,
      "step": 10995
    },
    {
      "entropy": 5.485701847076416,
      "epoch": 1.2354130398158027,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18213511556386947,
      "num_tokens": 24536350.0,
      "step": 11000
    },
    {
      "entropy": 5.588993835449219,
      "epoch": 1.2359746167237604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.4034,
      "mean_token_accuracy": 0.18713052719831466,
      "num_tokens": 24547756.0,
      "step": 11005
    },
    {
      "entropy": 5.509750890731811,
      "epoch": 1.2365361936317179,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.2776,
      "mean_token_accuracy": 0.19096165597438813,
      "num_tokens": 24559897.0,
      "step": 11010
    },
    {
      "entropy": 5.467511320114136,
      "epoch": 1.2370977705396755,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.3089,
      "mean_token_accuracy": 0.18433302342891694,
      "num_tokens": 24570168.0,
      "step": 11015
    },
    {
      "entropy": 5.460204935073852,
      "epoch": 1.237659347447633,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19639374762773515,
      "num_tokens": 24580918.0,
      "step": 11020
    },
    {
      "entropy": 5.410836982727051,
      "epoch": 1.2382209243555904,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18541559875011443,
      "num_tokens": 24592378.0,
      "step": 11025
    },
    {
      "entropy": 5.5328937530517575,
      "epoch": 1.238782501263548,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.3732,
      "mean_token_accuracy": 0.18891598731279374,
      "num_tokens": 24603666.0,
      "step": 11030
    },
    {
      "entropy": 5.484682369232178,
      "epoch": 1.2393440781715055,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.3345,
      "mean_token_accuracy": 0.18914095163345337,
      "num_tokens": 24615299.0,
      "step": 11035
    },
    {
      "entropy": 5.496534729003907,
      "epoch": 1.2399056550794632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1921449452638626,
      "num_tokens": 24626138.0,
      "step": 11040
    },
    {
      "entropy": 5.4452039241790775,
      "epoch": 1.2404672319874206,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.2981,
      "mean_token_accuracy": 0.19092509895563126,
      "num_tokens": 24636720.0,
      "step": 11045
    },
    {
      "entropy": 5.506981277465821,
      "epoch": 1.2410288088953783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.3201,
      "mean_token_accuracy": 0.18461062908172607,
      "num_tokens": 24648147.0,
      "step": 11050
    },
    {
      "entropy": 5.461820936203003,
      "epoch": 1.2415903858033357,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.3638,
      "mean_token_accuracy": 0.18919588029384612,
      "num_tokens": 24659825.0,
      "step": 11055
    },
    {
      "entropy": 5.477649545669555,
      "epoch": 1.2421519627112934,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.3236,
      "mean_token_accuracy": 0.1912950709462166,
      "num_tokens": 24671583.0,
      "step": 11060
    },
    {
      "entropy": 5.443061256408692,
      "epoch": 1.2427135396192508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19250702112913132,
      "num_tokens": 24682034.0,
      "step": 11065
    },
    {
      "entropy": 5.463092708587647,
      "epoch": 1.2432751165272085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.3473,
      "mean_token_accuracy": 0.18822134882211686,
      "num_tokens": 24693624.0,
      "step": 11070
    },
    {
      "entropy": 5.557461404800415,
      "epoch": 1.243836693435166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.3631,
      "mean_token_accuracy": 0.18502741307020187,
      "num_tokens": 24705340.0,
      "step": 11075
    },
    {
      "entropy": 5.620899343490601,
      "epoch": 1.2443982703431236,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18347597867250443,
      "num_tokens": 24717115.0,
      "step": 11080
    },
    {
      "entropy": 5.468814563751221,
      "epoch": 1.244959847251081,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.2434,
      "mean_token_accuracy": 0.19057439416646957,
      "num_tokens": 24727768.0,
      "step": 11085
    },
    {
      "entropy": 5.479076480865478,
      "epoch": 1.2455214241590387,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.4109,
      "mean_token_accuracy": 0.1810232073068619,
      "num_tokens": 24740054.0,
      "step": 11090
    },
    {
      "entropy": 5.50800929069519,
      "epoch": 1.246083001066996,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.283,
      "mean_token_accuracy": 0.19207758456468582,
      "num_tokens": 24750658.0,
      "step": 11095
    },
    {
      "entropy": 5.605063104629517,
      "epoch": 1.2466445779749538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000485544223542299,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1885771170258522,
      "num_tokens": 24762730.0,
      "step": 11100
    },
    {
      "entropy": 5.432342052459717,
      "epoch": 1.2472061548829112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.2305,
      "mean_token_accuracy": 0.19524402618408204,
      "num_tokens": 24773865.0,
      "step": 11105
    },
    {
      "entropy": 5.421746683120728,
      "epoch": 1.2477677317908689,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1914224237203598,
      "num_tokens": 24783030.0,
      "step": 11110
    },
    {
      "entropy": 5.414378118515015,
      "epoch": 1.2483293086988263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.2622,
      "mean_token_accuracy": 0.19432681947946548,
      "num_tokens": 24793177.0,
      "step": 11115
    },
    {
      "entropy": 5.558236885070801,
      "epoch": 1.2488908856067837,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.3723,
      "mean_token_accuracy": 0.19053647071123123,
      "num_tokens": 24804796.0,
      "step": 11120
    },
    {
      "entropy": 5.559025812149048,
      "epoch": 1.2494524625147414,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.3767,
      "mean_token_accuracy": 0.18704819679260254,
      "num_tokens": 24816806.0,
      "step": 11125
    },
    {
      "entropy": 5.454526090621949,
      "epoch": 1.250014039422699,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.3259,
      "mean_token_accuracy": 0.1941091910004616,
      "num_tokens": 24826888.0,
      "step": 11130
    },
    {
      "entropy": 5.470983648300171,
      "epoch": 1.2505756163306565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19485928863286972,
      "num_tokens": 24837731.0,
      "step": 11135
    },
    {
      "entropy": 5.458429908752441,
      "epoch": 1.251137193238614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.3003,
      "mean_token_accuracy": 0.1871982917189598,
      "num_tokens": 24847900.0,
      "step": 11140
    },
    {
      "entropy": 5.462789726257324,
      "epoch": 1.2516987701465716,
      "grad_norm": 1.125,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.3874,
      "mean_token_accuracy": 0.1854668065905571,
      "num_tokens": 24858878.0,
      "step": 11145
    },
    {
      "entropy": 5.493943071365356,
      "epoch": 1.252260347054529,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1905045062303543,
      "num_tokens": 24870051.0,
      "step": 11150
    },
    {
      "entropy": 5.4933915615081785,
      "epoch": 1.2528219239624867,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.4166,
      "mean_token_accuracy": 0.18264361917972566,
      "num_tokens": 24881671.0,
      "step": 11155
    },
    {
      "entropy": 5.537044477462769,
      "epoch": 1.2533835008704441,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.2761,
      "mean_token_accuracy": 0.18795724660158158,
      "num_tokens": 24892944.0,
      "step": 11160
    },
    {
      "entropy": 5.504327917098999,
      "epoch": 1.2539450777784018,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.3644,
      "mean_token_accuracy": 0.1891561046242714,
      "num_tokens": 24903346.0,
      "step": 11165
    },
    {
      "entropy": 5.45168251991272,
      "epoch": 1.2545066546863592,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.338,
      "mean_token_accuracy": 0.18961320370435714,
      "num_tokens": 24913904.0,
      "step": 11170
    },
    {
      "entropy": 5.454988813400268,
      "epoch": 1.255068231594317,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19362047165632248,
      "num_tokens": 24924635.0,
      "step": 11175
    },
    {
      "entropy": 5.553615617752075,
      "epoch": 1.2556298085022743,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.4802,
      "mean_token_accuracy": 0.18075439780950547,
      "num_tokens": 24935466.0,
      "step": 11180
    },
    {
      "entropy": 5.584972953796386,
      "epoch": 1.256191385410232,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.4355,
      "mean_token_accuracy": 0.18358389437198638,
      "num_tokens": 24948348.0,
      "step": 11185
    },
    {
      "entropy": 5.489812278747559,
      "epoch": 1.2567529623181894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.2809,
      "mean_token_accuracy": 0.19066839814186096,
      "num_tokens": 24959242.0,
      "step": 11190
    },
    {
      "entropy": 5.447453498840332,
      "epoch": 1.2573145392261469,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.2568,
      "mean_token_accuracy": 0.19476295560598372,
      "num_tokens": 24969855.0,
      "step": 11195
    },
    {
      "entropy": 5.429177618026733,
      "epoch": 1.2578761161341045,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.2817,
      "mean_token_accuracy": 0.19262396395206452,
      "num_tokens": 24980378.0,
      "step": 11200
    },
    {
      "entropy": 5.509761619567871,
      "epoch": 1.2584376930420622,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.3507,
      "mean_token_accuracy": 0.18462255895137786,
      "num_tokens": 24990646.0,
      "step": 11205
    },
    {
      "entropy": 5.61767258644104,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.4643,
      "mean_token_accuracy": 0.1814809709787369,
      "num_tokens": 25002644.0,
      "step": 11210
    },
    {
      "entropy": 5.524403715133667,
      "epoch": 1.259560846857977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.3586,
      "mean_token_accuracy": 0.1936364233493805,
      "num_tokens": 25012635.0,
      "step": 11215
    },
    {
      "entropy": 5.49802942276001,
      "epoch": 1.2601224237659348,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000485202483023983,
      "loss": 5.4706,
      "mean_token_accuracy": 0.18091963231563568,
      "num_tokens": 25024289.0,
      "step": 11220
    },
    {
      "entropy": 5.516169881820678,
      "epoch": 1.2606840006738924,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.2383,
      "mean_token_accuracy": 0.18986262381076813,
      "num_tokens": 25035316.0,
      "step": 11225
    },
    {
      "entropy": 5.516177320480347,
      "epoch": 1.2612455775818499,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.3003,
      "mean_token_accuracy": 0.19051390886306763,
      "num_tokens": 25046011.0,
      "step": 11230
    },
    {
      "entropy": 5.484417676925659,
      "epoch": 1.2618071544898073,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.2965,
      "mean_token_accuracy": 0.1923714891076088,
      "num_tokens": 25057402.0,
      "step": 11235
    },
    {
      "entropy": 5.447501087188721,
      "epoch": 1.262368731397765,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.2949,
      "mean_token_accuracy": 0.18824667632579803,
      "num_tokens": 25068287.0,
      "step": 11240
    },
    {
      "entropy": 5.436086130142212,
      "epoch": 1.2629303083057224,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.2727,
      "mean_token_accuracy": 0.19113888293504716,
      "num_tokens": 25079534.0,
      "step": 11245
    },
    {
      "entropy": 5.405421876907349,
      "epoch": 1.26349188521368,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.2436,
      "mean_token_accuracy": 0.19242125302553176,
      "num_tokens": 25090825.0,
      "step": 11250
    },
    {
      "entropy": 5.475717782974243,
      "epoch": 1.2640534621216375,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.3484,
      "mean_token_accuracy": 0.18806565552949905,
      "num_tokens": 25101859.0,
      "step": 11255
    },
    {
      "entropy": 5.460765314102173,
      "epoch": 1.2646150390295952,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19475947916507722,
      "num_tokens": 25111416.0,
      "step": 11260
    },
    {
      "entropy": 5.496658134460449,
      "epoch": 1.2651766159375526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1914309784770012,
      "num_tokens": 25122129.0,
      "step": 11265
    },
    {
      "entropy": 5.514855623245239,
      "epoch": 1.2657381928455103,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.3453,
      "mean_token_accuracy": 0.1911257103085518,
      "num_tokens": 25132429.0,
      "step": 11270
    },
    {
      "entropy": 5.402274036407471,
      "epoch": 1.2662997697534677,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.383,
      "mean_token_accuracy": 0.19443860203027724,
      "num_tokens": 25144242.0,
      "step": 11275
    },
    {
      "entropy": 5.562759017944336,
      "epoch": 1.2668613466614254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.3843,
      "mean_token_accuracy": 0.19072728753089904,
      "num_tokens": 25155332.0,
      "step": 11280
    },
    {
      "entropy": 5.5600745677948,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.3138,
      "mean_token_accuracy": 0.18983711004257203,
      "num_tokens": 25166024.0,
      "step": 11285
    },
    {
      "entropy": 5.466601705551147,
      "epoch": 1.2679845004773402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.254,
      "mean_token_accuracy": 0.19609452188014984,
      "num_tokens": 25176697.0,
      "step": 11290
    },
    {
      "entropy": 5.4619403839111325,
      "epoch": 1.268546077385298,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.3584,
      "mean_token_accuracy": 0.19177097380161284,
      "num_tokens": 25187421.0,
      "step": 11295
    },
    {
      "entropy": 5.467381525039673,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.3589,
      "mean_token_accuracy": 0.1935126945376396,
      "num_tokens": 25198851.0,
      "step": 11300
    },
    {
      "entropy": 5.543435716629029,
      "epoch": 1.269669231201213,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.406,
      "mean_token_accuracy": 0.1843758761882782,
      "num_tokens": 25209984.0,
      "step": 11305
    },
    {
      "entropy": 5.54110951423645,
      "epoch": 1.2702308081091704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.2978,
      "mean_token_accuracy": 0.18914757221937178,
      "num_tokens": 25220672.0,
      "step": 11310
    },
    {
      "entropy": 5.520363998413086,
      "epoch": 1.270792385017128,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18871739655733108,
      "num_tokens": 25231906.0,
      "step": 11315
    },
    {
      "entropy": 5.525923109054565,
      "epoch": 1.2713539619250858,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18501406908035278,
      "num_tokens": 25242902.0,
      "step": 11320
    },
    {
      "entropy": 5.501807594299317,
      "epoch": 1.2719155388330432,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.4056,
      "mean_token_accuracy": 0.18155012279748917,
      "num_tokens": 25254363.0,
      "step": 11325
    },
    {
      "entropy": 5.500497674942016,
      "epoch": 1.2724771157410006,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1882040560245514,
      "num_tokens": 25267108.0,
      "step": 11330
    },
    {
      "entropy": 5.43377833366394,
      "epoch": 1.2730386926489583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19312814027070999,
      "num_tokens": 25278583.0,
      "step": 11335
    },
    {
      "entropy": 5.571519756317139,
      "epoch": 1.2736002695569157,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.3909,
      "mean_token_accuracy": 0.1879279851913452,
      "num_tokens": 25289813.0,
      "step": 11340
    },
    {
      "entropy": 5.465385484695434,
      "epoch": 1.2741618464648734,
      "grad_norm": 1.125,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19203702062368394,
      "num_tokens": 25301003.0,
      "step": 11345
    },
    {
      "entropy": 5.427354907989502,
      "epoch": 1.2747234233728308,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.2768,
      "mean_token_accuracy": 0.19592732936143875,
      "num_tokens": 25312535.0,
      "step": 11350
    },
    {
      "entropy": 5.470835161209107,
      "epoch": 1.2752850002807885,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.3309,
      "mean_token_accuracy": 0.18613635152578353,
      "num_tokens": 25323333.0,
      "step": 11355
    },
    {
      "entropy": 5.600289726257325,
      "epoch": 1.275846577188746,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.4115,
      "mean_token_accuracy": 0.1884475588798523,
      "num_tokens": 25334865.0,
      "step": 11360
    },
    {
      "entropy": 5.5824517726898195,
      "epoch": 1.2764081540967036,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.3916,
      "mean_token_accuracy": 0.18760250210762025,
      "num_tokens": 25346672.0,
      "step": 11365
    },
    {
      "entropy": 5.489925146102905,
      "epoch": 1.276969731004661,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.3809,
      "mean_token_accuracy": 0.1921190619468689,
      "num_tokens": 25358702.0,
      "step": 11370
    },
    {
      "entropy": 5.46632194519043,
      "epoch": 1.2775313079126187,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.3192,
      "mean_token_accuracy": 0.18536871969699859,
      "num_tokens": 25369813.0,
      "step": 11375
    },
    {
      "entropy": 5.527243518829346,
      "epoch": 1.2780928848205761,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.3592,
      "mean_token_accuracy": 0.18779368549585343,
      "num_tokens": 25381042.0,
      "step": 11380
    },
    {
      "entropy": 5.503709030151367,
      "epoch": 1.2786544617285336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.3471,
      "mean_token_accuracy": 0.18134421855211258,
      "num_tokens": 25391286.0,
      "step": 11385
    },
    {
      "entropy": 5.471360301971435,
      "epoch": 1.2792160386364912,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.2919,
      "mean_token_accuracy": 0.19129288643598558,
      "num_tokens": 25402335.0,
      "step": 11390
    },
    {
      "entropy": 5.434795999526978,
      "epoch": 1.279777615544449,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1946100503206253,
      "num_tokens": 25413484.0,
      "step": 11395
    },
    {
      "entropy": 5.5469069480896,
      "epoch": 1.2803391924524063,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.3572,
      "mean_token_accuracy": 0.18792201727628707,
      "num_tokens": 25424276.0,
      "step": 11400
    },
    {
      "entropy": 5.342579078674317,
      "epoch": 1.2809007693603638,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.235,
      "mean_token_accuracy": 0.19518037289381027,
      "num_tokens": 25435178.0,
      "step": 11405
    },
    {
      "entropy": 5.389122056961059,
      "epoch": 1.2814623462683215,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.199,
      "mean_token_accuracy": 0.19640120565891267,
      "num_tokens": 25445792.0,
      "step": 11410
    },
    {
      "entropy": 5.437396717071533,
      "epoch": 1.2820239231762791,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.2889,
      "mean_token_accuracy": 0.19388676136732103,
      "num_tokens": 25456598.0,
      "step": 11415
    },
    {
      "entropy": 5.45359468460083,
      "epoch": 1.2825855000842366,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.3734,
      "mean_token_accuracy": 0.18930955827236176,
      "num_tokens": 25468311.0,
      "step": 11420
    },
    {
      "entropy": 5.552885103225708,
      "epoch": 1.283147076992194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18507194221019746,
      "num_tokens": 25480680.0,
      "step": 11425
    },
    {
      "entropy": 5.604588317871094,
      "epoch": 1.2837086539001517,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19158392101526261,
      "num_tokens": 25492255.0,
      "step": 11430
    },
    {
      "entropy": 5.481392669677734,
      "epoch": 1.284270230808109,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000484580560960249,
      "loss": 5.3929,
      "mean_token_accuracy": 0.18605994880199433,
      "num_tokens": 25504530.0,
      "step": 11435
    },
    {
      "entropy": 5.42602310180664,
      "epoch": 1.2848318077160668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.3281,
      "mean_token_accuracy": 0.19711543321609498,
      "num_tokens": 25515273.0,
      "step": 11440
    },
    {
      "entropy": 5.551933288574219,
      "epoch": 1.2853933846240242,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.4505,
      "mean_token_accuracy": 0.18545989394187928,
      "num_tokens": 25526050.0,
      "step": 11445
    },
    {
      "entropy": 5.565883159637451,
      "epoch": 1.2859549615319819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.3579,
      "mean_token_accuracy": 0.186768040060997,
      "num_tokens": 25538372.0,
      "step": 11450
    },
    {
      "entropy": 5.620044374465943,
      "epoch": 1.2865165384399393,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.442,
      "mean_token_accuracy": 0.1926871195435524,
      "num_tokens": 25549636.0,
      "step": 11455
    },
    {
      "entropy": 5.497498512268066,
      "epoch": 1.287078115347897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.3677,
      "mean_token_accuracy": 0.18753843009471893,
      "num_tokens": 25560850.0,
      "step": 11460
    },
    {
      "entropy": 5.476309728622437,
      "epoch": 1.2876396922558544,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19111284762620925,
      "num_tokens": 25572433.0,
      "step": 11465
    },
    {
      "entropy": 5.607483530044556,
      "epoch": 1.288201269163812,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1838137537240982,
      "num_tokens": 25582372.0,
      "step": 11470
    },
    {
      "entropy": 5.4621470928192135,
      "epoch": 1.2887628460717695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.2084,
      "mean_token_accuracy": 0.1927051931619644,
      "num_tokens": 25592570.0,
      "step": 11475
    },
    {
      "entropy": 5.490028619766235,
      "epoch": 1.289324422979727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.3351,
      "mean_token_accuracy": 0.1845156028866768,
      "num_tokens": 25603573.0,
      "step": 11480
    },
    {
      "entropy": 5.530058574676514,
      "epoch": 1.2898859998876846,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.3641,
      "mean_token_accuracy": 0.18942673951387407,
      "num_tokens": 25614233.0,
      "step": 11485
    },
    {
      "entropy": 5.527749490737915,
      "epoch": 1.2904475767956423,
      "grad_norm": 1.25,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.3323,
      "mean_token_accuracy": 0.1844804048538208,
      "num_tokens": 25625077.0,
      "step": 11490
    },
    {
      "entropy": 5.434228038787841,
      "epoch": 1.2910091537035997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.3518,
      "mean_token_accuracy": 0.18301459401845932,
      "num_tokens": 25636693.0,
      "step": 11495
    },
    {
      "entropy": 5.331040525436402,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.2095,
      "mean_token_accuracy": 0.19238260984420777,
      "num_tokens": 25647644.0,
      "step": 11500
    },
    {
      "entropy": 5.435154771804809,
      "epoch": 1.2921323075195148,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.2446,
      "mean_token_accuracy": 0.19497229605913163,
      "num_tokens": 25659373.0,
      "step": 11505
    },
    {
      "entropy": 5.556627988815308,
      "epoch": 1.2926938844274725,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19794285893440247,
      "num_tokens": 25670479.0,
      "step": 11510
    },
    {
      "entropy": 5.544531679153442,
      "epoch": 1.29325546133543,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.3877,
      "mean_token_accuracy": 0.1912296235561371,
      "num_tokens": 25682748.0,
      "step": 11515
    },
    {
      "entropy": 5.4173712730407715,
      "epoch": 1.2938170382433873,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.2218,
      "mean_token_accuracy": 0.19448208659887314,
      "num_tokens": 25693170.0,
      "step": 11520
    },
    {
      "entropy": 5.427676868438721,
      "epoch": 1.294378615151345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.3497,
      "mean_token_accuracy": 0.1826302319765091,
      "num_tokens": 25706009.0,
      "step": 11525
    },
    {
      "entropy": 5.553053140640259,
      "epoch": 1.2949401920593024,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.3782,
      "mean_token_accuracy": 0.19116197675466537,
      "num_tokens": 25716036.0,
      "step": 11530
    },
    {
      "entropy": 5.505878114700318,
      "epoch": 1.29550176896726,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.1892,
      "mean_token_accuracy": 0.19445182532072067,
      "num_tokens": 25726516.0,
      "step": 11535
    },
    {
      "entropy": 5.4927653789520265,
      "epoch": 1.2960633458752175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.345,
      "mean_token_accuracy": 0.18722363859415053,
      "num_tokens": 25736340.0,
      "step": 11540
    },
    {
      "entropy": 5.356808090209961,
      "epoch": 1.2966249227831752,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.2995,
      "mean_token_accuracy": 0.18733727484941481,
      "num_tokens": 25746439.0,
      "step": 11545
    },
    {
      "entropy": 5.466232585906982,
      "epoch": 1.2971864996911326,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.2113,
      "mean_token_accuracy": 0.193405120074749,
      "num_tokens": 25757065.0,
      "step": 11550
    },
    {
      "entropy": 5.494842433929444,
      "epoch": 1.2977480765990903,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.2876,
      "mean_token_accuracy": 0.18820154219865798,
      "num_tokens": 25768021.0,
      "step": 11555
    },
    {
      "entropy": 5.496604585647583,
      "epoch": 1.2983096535070477,
      "grad_norm": 1.125,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.39,
      "mean_token_accuracy": 0.19395477026700975,
      "num_tokens": 25780197.0,
      "step": 11560
    },
    {
      "entropy": 5.477793979644775,
      "epoch": 1.2988712304150054,
      "grad_norm": 1.125,
      "learning_rate": 0.000484198526925203,
      "loss": 5.3126,
      "mean_token_accuracy": 0.18827086836099624,
      "num_tokens": 25790536.0,
      "step": 11565
    },
    {
      "entropy": 5.445866107940674,
      "epoch": 1.2994328073229628,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.276,
      "mean_token_accuracy": 0.19228814989328386,
      "num_tokens": 25802397.0,
      "step": 11570
    },
    {
      "entropy": 5.501689720153808,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.3892,
      "mean_token_accuracy": 0.19443353563547133,
      "num_tokens": 25813961.0,
      "step": 11575
    },
    {
      "entropy": 5.5217243194580075,
      "epoch": 1.300555961138878,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.4093,
      "mean_token_accuracy": 0.18545944094657899,
      "num_tokens": 25825610.0,
      "step": 11580
    },
    {
      "entropy": 5.471556091308594,
      "epoch": 1.3011175380468356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19547210782766342,
      "num_tokens": 25836076.0,
      "step": 11585
    },
    {
      "entropy": 5.4894482612609865,
      "epoch": 1.301679114954793,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.3344,
      "mean_token_accuracy": 0.19530290961265565,
      "num_tokens": 25847341.0,
      "step": 11590
    },
    {
      "entropy": 5.3722395420074465,
      "epoch": 1.3022406918627505,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.2376,
      "mean_token_accuracy": 0.1964733585715294,
      "num_tokens": 25856912.0,
      "step": 11595
    },
    {
      "entropy": 5.431223630905151,
      "epoch": 1.3028022687707081,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.2595,
      "mean_token_accuracy": 0.18715425878763198,
      "num_tokens": 25868370.0,
      "step": 11600
    },
    {
      "entropy": 5.4193699836730955,
      "epoch": 1.3033638456786658,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19211488366127014,
      "num_tokens": 25879104.0,
      "step": 11605
    },
    {
      "entropy": 5.535993194580078,
      "epoch": 1.3039254225866233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1839719831943512,
      "num_tokens": 25889744.0,
      "step": 11610
    },
    {
      "entropy": 5.553609228134155,
      "epoch": 1.3044869994945807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.4663,
      "mean_token_accuracy": 0.1850156307220459,
      "num_tokens": 25901615.0,
      "step": 11615
    },
    {
      "entropy": 5.526863670349121,
      "epoch": 1.3050485764025384,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.3245,
      "mean_token_accuracy": 0.18772584199905396,
      "num_tokens": 25911418.0,
      "step": 11620
    },
    {
      "entropy": 5.549993276596069,
      "epoch": 1.3056101533104958,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.4049,
      "mean_token_accuracy": 0.19036079645156861,
      "num_tokens": 25922141.0,
      "step": 11625
    },
    {
      "entropy": 5.539162349700928,
      "epoch": 1.3061717302184535,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.4201,
      "mean_token_accuracy": 0.18496872782707213,
      "num_tokens": 25933897.0,
      "step": 11630
    },
    {
      "entropy": 5.5644574642181395,
      "epoch": 1.306733307126411,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.3437,
      "mean_token_accuracy": 0.18744596987962722,
      "num_tokens": 25943960.0,
      "step": 11635
    },
    {
      "entropy": 5.605818891525269,
      "epoch": 1.3072948840343686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.3931,
      "mean_token_accuracy": 0.18857479095458984,
      "num_tokens": 25955104.0,
      "step": 11640
    },
    {
      "entropy": 5.493829298019409,
      "epoch": 1.307856460942326,
      "grad_norm": 1.125,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.3311,
      "mean_token_accuracy": 0.193758088350296,
      "num_tokens": 25965180.0,
      "step": 11645
    },
    {
      "entropy": 5.444762754440307,
      "epoch": 1.3084180378502837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19077154099941254,
      "num_tokens": 25976600.0,
      "step": 11650
    },
    {
      "entropy": 5.396129751205445,
      "epoch": 1.308979614758241,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18808194547891616,
      "num_tokens": 25987509.0,
      "step": 11655
    },
    {
      "entropy": 5.474160957336426,
      "epoch": 1.3095411916661988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000483916500180593,
      "loss": 5.242,
      "mean_token_accuracy": 0.19153997898101807,
      "num_tokens": 25997962.0,
      "step": 11660
    },
    {
      "entropy": 5.487403345108032,
      "epoch": 1.3101027685741562,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.3025,
      "mean_token_accuracy": 0.1925146296620369,
      "num_tokens": 26008932.0,
      "step": 11665
    },
    {
      "entropy": 5.488673543930053,
      "epoch": 1.3106643454821139,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1836942106485367,
      "num_tokens": 26020513.0,
      "step": 11670
    },
    {
      "entropy": 5.565633153915405,
      "epoch": 1.3112259223900713,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.373,
      "mean_token_accuracy": 0.18528845757246018,
      "num_tokens": 26032526.0,
      "step": 11675
    },
    {
      "entropy": 5.514460468292237,
      "epoch": 1.311787499298029,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19298763275146485,
      "num_tokens": 26044216.0,
      "step": 11680
    },
    {
      "entropy": 5.417226934432984,
      "epoch": 1.3123490762059864,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.2108,
      "mean_token_accuracy": 0.18900535702705384,
      "num_tokens": 26054012.0,
      "step": 11685
    },
    {
      "entropy": 5.4516205310821535,
      "epoch": 1.3129106531139438,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.3687,
      "mean_token_accuracy": 0.18900944739580156,
      "num_tokens": 26065365.0,
      "step": 11690
    },
    {
      "entropy": 5.529718637466431,
      "epoch": 1.3134722300219015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1843010053038597,
      "num_tokens": 26075964.0,
      "step": 11695
    },
    {
      "entropy": 5.4653685092926025,
      "epoch": 1.3140338069298592,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.3094,
      "mean_token_accuracy": 0.1927257478237152,
      "num_tokens": 26087572.0,
      "step": 11700
    },
    {
      "entropy": 5.428178977966309,
      "epoch": 1.3145953838378166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000483782070216197,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1969181165099144,
      "num_tokens": 26098612.0,
      "step": 11705
    },
    {
      "entropy": 5.512505435943604,
      "epoch": 1.315156960745774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.3931,
      "mean_token_accuracy": 0.19104406088590623,
      "num_tokens": 26109491.0,
      "step": 11710
    },
    {
      "entropy": 5.535184383392334,
      "epoch": 1.3157185376537317,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.2866,
      "mean_token_accuracy": 0.19219576567411423,
      "num_tokens": 26120544.0,
      "step": 11715
    },
    {
      "entropy": 5.515174674987793,
      "epoch": 1.3162801145616891,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.3864,
      "mean_token_accuracy": 0.1831790253520012,
      "num_tokens": 26131549.0,
      "step": 11720
    },
    {
      "entropy": 5.464992904663086,
      "epoch": 1.3168416914696468,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.3755,
      "mean_token_accuracy": 0.19335275739431382,
      "num_tokens": 26142871.0,
      "step": 11725
    },
    {
      "entropy": 5.580825424194336,
      "epoch": 1.3174032683776042,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.4526,
      "mean_token_accuracy": 0.189836722612381,
      "num_tokens": 26155293.0,
      "step": 11730
    },
    {
      "entropy": 5.571844244003296,
      "epoch": 1.317964845285562,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.3931,
      "mean_token_accuracy": 0.19178189486265182,
      "num_tokens": 26167069.0,
      "step": 11735
    },
    {
      "entropy": 5.498572158813476,
      "epoch": 1.3185264221935193,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.3932,
      "mean_token_accuracy": 0.18102272152900695,
      "num_tokens": 26179142.0,
      "step": 11740
    },
    {
      "entropy": 5.5183288097381595,
      "epoch": 1.319087999101477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.3208,
      "mean_token_accuracy": 0.1880206748843193,
      "num_tokens": 26190059.0,
      "step": 11745
    },
    {
      "entropy": 5.520627403259278,
      "epoch": 1.3196495760094344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.3416,
      "mean_token_accuracy": 0.1924460709095001,
      "num_tokens": 26199859.0,
      "step": 11750
    },
    {
      "entropy": 5.497109222412109,
      "epoch": 1.320211152917392,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.3722,
      "mean_token_accuracy": 0.18477653115987777,
      "num_tokens": 26211920.0,
      "step": 11755
    },
    {
      "entropy": 5.493609380722046,
      "epoch": 1.3207727298253495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18585283011198045,
      "num_tokens": 26222875.0,
      "step": 11760
    },
    {
      "entropy": 5.530766820907592,
      "epoch": 1.3213343067333072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18562915027141572,
      "num_tokens": 26233549.0,
      "step": 11765
    },
    {
      "entropy": 5.429447841644287,
      "epoch": 1.3218958836412646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.2386,
      "mean_token_accuracy": 0.20302475094795228,
      "num_tokens": 26243886.0,
      "step": 11770
    },
    {
      "entropy": 5.40765905380249,
      "epoch": 1.3224574605492223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.28,
      "mean_token_accuracy": 0.19504840672016144,
      "num_tokens": 26254475.0,
      "step": 11775
    },
    {
      "entropy": 5.541731977462769,
      "epoch": 1.3230190374571797,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18699074387550355,
      "num_tokens": 26265412.0,
      "step": 11780
    },
    {
      "entropy": 5.616350173950195,
      "epoch": 1.3235806143651372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.3941,
      "mean_token_accuracy": 0.18800213932991028,
      "num_tokens": 26276942.0,
      "step": 11785
    },
    {
      "entropy": 5.531629276275635,
      "epoch": 1.3241421912730948,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.3516,
      "mean_token_accuracy": 0.18680412024259568,
      "num_tokens": 26288009.0,
      "step": 11790
    },
    {
      "entropy": 5.51987886428833,
      "epoch": 1.3247037681810525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.5002,
      "mean_token_accuracy": 0.17894561290740968,
      "num_tokens": 26301583.0,
      "step": 11795
    },
    {
      "entropy": 5.576613235473633,
      "epoch": 1.32526534508901,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.3854,
      "mean_token_accuracy": 0.19318455308675767,
      "num_tokens": 26312922.0,
      "step": 11800
    },
    {
      "entropy": 5.575756168365478,
      "epoch": 1.3258269219969674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19539692103862763,
      "num_tokens": 26324108.0,
      "step": 11805
    },
    {
      "entropy": 5.520961284637451,
      "epoch": 1.326388498904925,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.3991,
      "mean_token_accuracy": 0.18515405729413031,
      "num_tokens": 26335669.0,
      "step": 11810
    },
    {
      "entropy": 5.383474254608155,
      "epoch": 1.3269500758128827,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.2692,
      "mean_token_accuracy": 0.1946129947900772,
      "num_tokens": 26346065.0,
      "step": 11815
    },
    {
      "entropy": 5.486476135253906,
      "epoch": 1.3275116527208402,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1924546480178833,
      "num_tokens": 26357567.0,
      "step": 11820
    },
    {
      "entropy": 5.472987461090088,
      "epoch": 1.3280732296287976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.3009,
      "mean_token_accuracy": 0.18726590871810914,
      "num_tokens": 26367805.0,
      "step": 11825
    },
    {
      "entropy": 5.41779203414917,
      "epoch": 1.3286348065367553,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.2161,
      "mean_token_accuracy": 0.19460426717996598,
      "num_tokens": 26379343.0,
      "step": 11830
    },
    {
      "entropy": 5.520466279983521,
      "epoch": 1.3291963834447127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.4277,
      "mean_token_accuracy": 0.18180939704179763,
      "num_tokens": 26392039.0,
      "step": 11835
    },
    {
      "entropy": 5.365854406356812,
      "epoch": 1.3297579603526704,
      "grad_norm": 1.0,
      "learning_rate": 0.000483375550934239,
      "loss": 5.2016,
      "mean_token_accuracy": 0.20187190175056458,
      "num_tokens": 26403324.0,
      "step": 11840
    },
    {
      "entropy": 5.469034719467163,
      "epoch": 1.3303195372606278,
      "grad_norm": 1.125,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.3181,
      "mean_token_accuracy": 0.1933395341038704,
      "num_tokens": 26413435.0,
      "step": 11845
    },
    {
      "entropy": 5.518855142593384,
      "epoch": 1.3308811141685855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.3947,
      "mean_token_accuracy": 0.18085718899965286,
      "num_tokens": 26424254.0,
      "step": 11850
    },
    {
      "entropy": 5.524479627609253,
      "epoch": 1.331442691076543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19006485342979432,
      "num_tokens": 26435658.0,
      "step": 11855
    },
    {
      "entropy": 5.539576292037964,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.373,
      "mean_token_accuracy": 0.1878064513206482,
      "num_tokens": 26447174.0,
      "step": 11860
    },
    {
      "entropy": 5.55002121925354,
      "epoch": 1.332565844892458,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.4465,
      "mean_token_accuracy": 0.1872912123799324,
      "num_tokens": 26459303.0,
      "step": 11865
    },
    {
      "entropy": 5.508304023742676,
      "epoch": 1.3331274218004157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.323,
      "mean_token_accuracy": 0.18582605868577956,
      "num_tokens": 26469981.0,
      "step": 11870
    },
    {
      "entropy": 5.49251012802124,
      "epoch": 1.333688998708373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19587670117616654,
      "num_tokens": 26481176.0,
      "step": 11875
    },
    {
      "entropy": 5.430894279479981,
      "epoch": 1.3342505756163305,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.2728,
      "mean_token_accuracy": 0.19216970801353456,
      "num_tokens": 26492795.0,
      "step": 11880
    },
    {
      "entropy": 5.539825582504273,
      "epoch": 1.3348121525242882,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.4408,
      "mean_token_accuracy": 0.1842477634549141,
      "num_tokens": 26504731.0,
      "step": 11885
    },
    {
      "entropy": 5.599880743026733,
      "epoch": 1.3353737294322459,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.3162,
      "mean_token_accuracy": 0.19063470065593718,
      "num_tokens": 26516816.0,
      "step": 11890
    },
    {
      "entropy": 5.533083200454712,
      "epoch": 1.3359353063402033,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1902502804994583,
      "num_tokens": 26528861.0,
      "step": 11895
    },
    {
      "entropy": 5.542850351333618,
      "epoch": 1.3364968832481607,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18914323598146437,
      "num_tokens": 26538816.0,
      "step": 11900
    },
    {
      "entropy": 5.593536853790283,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18843986690044404,
      "num_tokens": 26548959.0,
      "step": 11905
    },
    {
      "entropy": 5.429722595214844,
      "epoch": 1.337620037064076,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19304779022932053,
      "num_tokens": 26559189.0,
      "step": 11910
    },
    {
      "entropy": 5.376905107498169,
      "epoch": 1.3381816139720335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19619487822055817,
      "num_tokens": 26570066.0,
      "step": 11915
    },
    {
      "entropy": 5.4953361511230465,
      "epoch": 1.338743190879991,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.4143,
      "mean_token_accuracy": 0.1912911430001259,
      "num_tokens": 26581489.0,
      "step": 11920
    },
    {
      "entropy": 5.55296745300293,
      "epoch": 1.3393047677879486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.3355,
      "mean_token_accuracy": 0.18815976679325103,
      "num_tokens": 26592051.0,
      "step": 11925
    },
    {
      "entropy": 5.568896102905273,
      "epoch": 1.339866344695906,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18929506689310074,
      "num_tokens": 26602296.0,
      "step": 11930
    },
    {
      "entropy": 5.487189292907715,
      "epoch": 1.3404279216038637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.315,
      "mean_token_accuracy": 0.18514619916677474,
      "num_tokens": 26613837.0,
      "step": 11935
    },
    {
      "entropy": 5.446844005584717,
      "epoch": 1.3409894985118211,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.3757,
      "mean_token_accuracy": 0.18660722821950912,
      "num_tokens": 26624898.0,
      "step": 11940
    },
    {
      "entropy": 5.501018333435058,
      "epoch": 1.3415510754197788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.359,
      "mean_token_accuracy": 0.19223895370960237,
      "num_tokens": 26636449.0,
      "step": 11945
    },
    {
      "entropy": 5.602386617660523,
      "epoch": 1.3421126523277362,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.4509,
      "mean_token_accuracy": 0.18780952990055083,
      "num_tokens": 26649192.0,
      "step": 11950
    },
    {
      "entropy": 5.484031391143799,
      "epoch": 1.342674229235694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000483025441381746,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19373395442962646,
      "num_tokens": 26660005.0,
      "step": 11955
    },
    {
      "entropy": 5.59054012298584,
      "epoch": 1.3432358061436513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.4115,
      "mean_token_accuracy": 0.18633730560541154,
      "num_tokens": 26672165.0,
      "step": 11960
    },
    {
      "entropy": 5.466653633117676,
      "epoch": 1.343797383051609,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.359,
      "mean_token_accuracy": 0.1914936989545822,
      "num_tokens": 26683135.0,
      "step": 11965
    },
    {
      "entropy": 5.461401414871216,
      "epoch": 1.3443589599595664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.241,
      "mean_token_accuracy": 0.1981671690940857,
      "num_tokens": 26693724.0,
      "step": 11970
    },
    {
      "entropy": 5.5157993793487545,
      "epoch": 1.3449205368675239,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.4293,
      "mean_token_accuracy": 0.18887880593538284,
      "num_tokens": 26705880.0,
      "step": 11975
    },
    {
      "entropy": 5.547498750686645,
      "epoch": 1.3454821137754815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.419,
      "mean_token_accuracy": 0.19185321778059006,
      "num_tokens": 26716978.0,
      "step": 11980
    },
    {
      "entropy": 5.626371669769287,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.3592,
      "mean_token_accuracy": 0.19536844044923782,
      "num_tokens": 26728380.0,
      "step": 11985
    },
    {
      "entropy": 5.436957788467407,
      "epoch": 1.3466052675913966,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.3076,
      "mean_token_accuracy": 0.1934093102812767,
      "num_tokens": 26739034.0,
      "step": 11990
    },
    {
      "entropy": 5.292926263809204,
      "epoch": 1.347166844499354,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004829028425172487,
      "loss": 5.114,
      "mean_token_accuracy": 0.2030069813132286,
      "num_tokens": 26750123.0,
      "step": 11995
    },
    {
      "entropy": 5.428149938583374,
      "epoch": 1.3477284214073117,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.2874,
      "mean_token_accuracy": 0.20094300508499147,
      "num_tokens": 26760411.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.320394583513419,
      "eval_loss": 5.423473834991455,
      "eval_mean_token_accuracy": 0.1944776664912962,
      "eval_num_tokens": 26760411.0,
      "eval_runtime": 59.5728,
      "eval_samples_per_second": 520.59,
      "eval_steps_per_second": 65.08,
      "step": 12000
    },
    {
      "entropy": 5.478197383880615,
      "epoch": 1.3482899983152694,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.3597,
      "mean_token_accuracy": 0.19468082040548323,
      "num_tokens": 26770821.0,
      "step": 12005
    },
    {
      "entropy": 5.5496132373809814,
      "epoch": 1.3488515752232269,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.3429,
      "mean_token_accuracy": 0.19109699130058289,
      "num_tokens": 26782099.0,
      "step": 12010
    },
    {
      "entropy": 5.573636913299561,
      "epoch": 1.3494131521311843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.4507,
      "mean_token_accuracy": 0.18952461928129197,
      "num_tokens": 26794056.0,
      "step": 12015
    },
    {
      "entropy": 5.463382816314697,
      "epoch": 1.349974729039142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.3013,
      "mean_token_accuracy": 0.18759446889162062,
      "num_tokens": 26804501.0,
      "step": 12020
    },
    {
      "entropy": 5.553075218200684,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.3504,
      "mean_token_accuracy": 0.19351499825716018,
      "num_tokens": 26814996.0,
      "step": 12025
    },
    {
      "entropy": 5.499054431915283,
      "epoch": 1.351097882855057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1943981885910034,
      "num_tokens": 26825838.0,
      "step": 12030
    },
    {
      "entropy": 5.5289122581481935,
      "epoch": 1.3516594597630145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.3142,
      "mean_token_accuracy": 0.19527603238821029,
      "num_tokens": 26836320.0,
      "step": 12035
    },
    {
      "entropy": 5.513550424575806,
      "epoch": 1.3522210366709722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.3089,
      "mean_token_accuracy": 0.19219874292612077,
      "num_tokens": 26846485.0,
      "step": 12040
    },
    {
      "entropy": 5.592586374282837,
      "epoch": 1.3527826135789296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.4285,
      "mean_token_accuracy": 0.1832916557788849,
      "num_tokens": 26858554.0,
      "step": 12045
    },
    {
      "entropy": 5.51733169555664,
      "epoch": 1.3533441904868873,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.3427,
      "mean_token_accuracy": 0.19549855589866638,
      "num_tokens": 26869538.0,
      "step": 12050
    },
    {
      "entropy": 5.557057523727417,
      "epoch": 1.3539057673948447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18891253620386123,
      "num_tokens": 26880284.0,
      "step": 12055
    },
    {
      "entropy": 5.488908147811889,
      "epoch": 1.3544673443028024,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1821691557765007,
      "num_tokens": 26893180.0,
      "step": 12060
    },
    {
      "entropy": 5.480938911437988,
      "epoch": 1.3550289212107598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.251,
      "mean_token_accuracy": 0.194026155769825,
      "num_tokens": 26904213.0,
      "step": 12065
    },
    {
      "entropy": 5.504014730453491,
      "epoch": 1.3555904981187172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.3422,
      "mean_token_accuracy": 0.19684326350688935,
      "num_tokens": 26915566.0,
      "step": 12070
    },
    {
      "entropy": 5.395453691482544,
      "epoch": 1.356152075026675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.2932,
      "mean_token_accuracy": 0.19817951172590256,
      "num_tokens": 26926253.0,
      "step": 12075
    },
    {
      "entropy": 5.429937934875488,
      "epoch": 1.3567136519346326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.292,
      "mean_token_accuracy": 0.19771064221858978,
      "num_tokens": 26936723.0,
      "step": 12080
    },
    {
      "entropy": 5.461929750442505,
      "epoch": 1.35727522884259,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19243800789117813,
      "num_tokens": 26947551.0,
      "step": 12085
    },
    {
      "entropy": 5.534619951248169,
      "epoch": 1.3578368057505474,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.4036,
      "mean_token_accuracy": 0.19200349599123,
      "num_tokens": 26959969.0,
      "step": 12090
    },
    {
      "entropy": 5.553624153137207,
      "epoch": 1.358398382658505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18287914842367173,
      "num_tokens": 26972511.0,
      "step": 12095
    },
    {
      "entropy": 5.565522480010986,
      "epoch": 1.3589599595664628,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.4405,
      "mean_token_accuracy": 0.18577493578195572,
      "num_tokens": 26983827.0,
      "step": 12100
    },
    {
      "entropy": 5.592643737792969,
      "epoch": 1.3595215364744202,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.404,
      "mean_token_accuracy": 0.18419968634843825,
      "num_tokens": 26994668.0,
      "step": 12105
    },
    {
      "entropy": 5.437271165847778,
      "epoch": 1.3600831133823776,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.2605,
      "mean_token_accuracy": 0.1999809518456459,
      "num_tokens": 27006313.0,
      "step": 12110
    },
    {
      "entropy": 5.464850616455078,
      "epoch": 1.3606446902903353,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1898820087313652,
      "num_tokens": 27017628.0,
      "step": 12115
    },
    {
      "entropy": 5.449024724960327,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1939452037215233,
      "num_tokens": 27028303.0,
      "step": 12120
    },
    {
      "entropy": 5.558975076675415,
      "epoch": 1.3617678441062504,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.3899,
      "mean_token_accuracy": 0.18635841608047485,
      "num_tokens": 27039307.0,
      "step": 12125
    },
    {
      "entropy": 5.4659029006958,
      "epoch": 1.3623294210142078,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.2662,
      "mean_token_accuracy": 0.1897073581814766,
      "num_tokens": 27049271.0,
      "step": 12130
    },
    {
      "entropy": 5.365239048004151,
      "epoch": 1.3628909979221655,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.179,
      "mean_token_accuracy": 0.19675081372261047,
      "num_tokens": 27059510.0,
      "step": 12135
    },
    {
      "entropy": 5.606520462036133,
      "epoch": 1.363452574830123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.523,
      "mean_token_accuracy": 0.17754722088575364,
      "num_tokens": 27071685.0,
      "step": 12140
    },
    {
      "entropy": 5.416141366958618,
      "epoch": 1.3640141517380806,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.2462,
      "mean_token_accuracy": 0.1859876334667206,
      "num_tokens": 27081957.0,
      "step": 12145
    },
    {
      "entropy": 5.390149831771851,
      "epoch": 1.364575728646038,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.254,
      "mean_token_accuracy": 0.19522633850574495,
      "num_tokens": 27091609.0,
      "step": 12150
    },
    {
      "entropy": 5.48789267539978,
      "epoch": 1.3651373055539957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.4331,
      "mean_token_accuracy": 0.1839677393436432,
      "num_tokens": 27103889.0,
      "step": 12155
    },
    {
      "entropy": 5.594719934463501,
      "epoch": 1.3656988824619531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000482392646735433,
      "loss": 5.3644,
      "mean_token_accuracy": 0.19539450258016586,
      "num_tokens": 27114186.0,
      "step": 12160
    },
    {
      "entropy": 5.3895775318145756,
      "epoch": 1.3662604593699106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.2358,
      "mean_token_accuracy": 0.1960143566131592,
      "num_tokens": 27124677.0,
      "step": 12165
    },
    {
      "entropy": 5.543695735931396,
      "epoch": 1.3668220362778682,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.4065,
      "mean_token_accuracy": 0.19060794711112977,
      "num_tokens": 27134873.0,
      "step": 12170
    },
    {
      "entropy": 5.506030416488647,
      "epoch": 1.367383613185826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.4083,
      "mean_token_accuracy": 0.1852779746055603,
      "num_tokens": 27146107.0,
      "step": 12175
    },
    {
      "entropy": 5.439063644409179,
      "epoch": 1.3679451900937833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.2579,
      "mean_token_accuracy": 0.19871385544538497,
      "num_tokens": 27158821.0,
      "step": 12180
    },
    {
      "entropy": 5.448226499557495,
      "epoch": 1.3685067670017408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1965023010969162,
      "num_tokens": 27169513.0,
      "step": 12185
    },
    {
      "entropy": 5.390142393112183,
      "epoch": 1.3690683439096984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1985243782401085,
      "num_tokens": 27180059.0,
      "step": 12190
    },
    {
      "entropy": 5.485644292831421,
      "epoch": 1.369629920817656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.3711,
      "mean_token_accuracy": 0.1902673915028572,
      "num_tokens": 27190738.0,
      "step": 12195
    },
    {
      "entropy": 5.505376386642456,
      "epoch": 1.3701914977256135,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.2515,
      "mean_token_accuracy": 0.1893245533108711,
      "num_tokens": 27200664.0,
      "step": 12200
    },
    {
      "entropy": 5.458845520019532,
      "epoch": 1.370753074633571,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.3105,
      "mean_token_accuracy": 0.19332458823919296,
      "num_tokens": 27211011.0,
      "step": 12205
    },
    {
      "entropy": 5.4974650859832765,
      "epoch": 1.3713146515415287,
      "grad_norm": 1.125,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1991906002163887,
      "num_tokens": 27221888.0,
      "step": 12210
    },
    {
      "entropy": 5.569535923004151,
      "epoch": 1.371876228449486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.3575,
      "mean_token_accuracy": 0.18815023750066756,
      "num_tokens": 27234671.0,
      "step": 12215
    },
    {
      "entropy": 5.406680059432984,
      "epoch": 1.3724378053574438,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20078462660312651,
      "num_tokens": 27245283.0,
      "step": 12220
    },
    {
      "entropy": 5.483945226669311,
      "epoch": 1.3729993822654012,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.3197,
      "mean_token_accuracy": 0.19311304688453673,
      "num_tokens": 27255056.0,
      "step": 12225
    },
    {
      "entropy": 5.567039108276367,
      "epoch": 1.3735609591733589,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000482174026472802,
      "loss": 5.3615,
      "mean_token_accuracy": 0.1872144341468811,
      "num_tokens": 27266893.0,
      "step": 12230
    },
    {
      "entropy": 5.468055295944214,
      "epoch": 1.3741225360813163,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.392,
      "mean_token_accuracy": 0.18777517080307007,
      "num_tokens": 27279504.0,
      "step": 12235
    },
    {
      "entropy": 5.524121522903442,
      "epoch": 1.374684112989274,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.4465,
      "mean_token_accuracy": 0.18385884016752244,
      "num_tokens": 27290525.0,
      "step": 12240
    },
    {
      "entropy": 5.534918260574341,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.3009,
      "mean_token_accuracy": 0.1905997797846794,
      "num_tokens": 27301404.0,
      "step": 12245
    },
    {
      "entropy": 5.474128341674804,
      "epoch": 1.375807266805189,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.244,
      "mean_token_accuracy": 0.18948123157024382,
      "num_tokens": 27312635.0,
      "step": 12250
    },
    {
      "entropy": 5.376252698898315,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.2927,
      "mean_token_accuracy": 0.19547967612743378,
      "num_tokens": 27322485.0,
      "step": 12255
    },
    {
      "entropy": 5.478473758697509,
      "epoch": 1.376930420621104,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18763597011566163,
      "num_tokens": 27332568.0,
      "step": 12260
    },
    {
      "entropy": 5.604016065597534,
      "epoch": 1.3774919975290616,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.3856,
      "mean_token_accuracy": 0.18369714319705963,
      "num_tokens": 27343058.0,
      "step": 12265
    },
    {
      "entropy": 5.5251256942749025,
      "epoch": 1.3780535744370193,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.2777,
      "mean_token_accuracy": 0.1988302230834961,
      "num_tokens": 27355316.0,
      "step": 12270
    },
    {
      "entropy": 5.479616785049439,
      "epoch": 1.3786151513449767,
      "grad_norm": 1.25,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1981513559818268,
      "num_tokens": 27366098.0,
      "step": 12275
    },
    {
      "entropy": 5.421827983856201,
      "epoch": 1.3791767282529341,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.2369,
      "mean_token_accuracy": 0.1967311054468155,
      "num_tokens": 27376923.0,
      "step": 12280
    },
    {
      "entropy": 5.4623127460479735,
      "epoch": 1.3797383051608918,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.3739,
      "mean_token_accuracy": 0.18950267285108566,
      "num_tokens": 27386789.0,
      "step": 12285
    },
    {
      "entropy": 5.4712121963500975,
      "epoch": 1.3802998820688495,
      "grad_norm": 1.078125,
      "learning_rate": 0.000481985608570654,
      "loss": 5.2972,
      "mean_token_accuracy": 0.18938076943159105,
      "num_tokens": 27399241.0,
      "step": 12290
    },
    {
      "entropy": 5.493840456008911,
      "epoch": 1.380861458976807,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.1998,
      "mean_token_accuracy": 0.20267933756113052,
      "num_tokens": 27409123.0,
      "step": 12295
    },
    {
      "entropy": 5.411140584945679,
      "epoch": 1.3814230358847643,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.2602,
      "mean_token_accuracy": 0.189808289706707,
      "num_tokens": 27419920.0,
      "step": 12300
    },
    {
      "entropy": 5.349962997436523,
      "epoch": 1.381984612792722,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.1474,
      "mean_token_accuracy": 0.19934788793325425,
      "num_tokens": 27429552.0,
      "step": 12305
    },
    {
      "entropy": 5.522129106521606,
      "epoch": 1.3825461897006794,
      "grad_norm": 1.125,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.3666,
      "mean_token_accuracy": 0.19459759145975114,
      "num_tokens": 27440404.0,
      "step": 12310
    },
    {
      "entropy": 5.489263820648193,
      "epoch": 1.383107766608637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.3942,
      "mean_token_accuracy": 0.1918272614479065,
      "num_tokens": 27452231.0,
      "step": 12315
    },
    {
      "entropy": 5.490846252441406,
      "epoch": 1.3836693435165945,
      "grad_norm": 1.125,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1911655768752098,
      "num_tokens": 27463658.0,
      "step": 12320
    },
    {
      "entropy": 5.5051093101501465,
      "epoch": 1.3842309204245522,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.3354,
      "mean_token_accuracy": 0.18559889048337935,
      "num_tokens": 27475403.0,
      "step": 12325
    },
    {
      "entropy": 5.559214401245117,
      "epoch": 1.3847924973325096,
      "grad_norm": 1.125,
      "learning_rate": 0.000481859469342418,
      "loss": 5.406,
      "mean_token_accuracy": 0.18680911213159562,
      "num_tokens": 27487056.0,
      "step": 12330
    },
    {
      "entropy": 5.427961730957032,
      "epoch": 1.3853540742404673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.269,
      "mean_token_accuracy": 0.1931942269206047,
      "num_tokens": 27498659.0,
      "step": 12335
    },
    {
      "entropy": 5.604191732406616,
      "epoch": 1.3859156511484247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1852207824587822,
      "num_tokens": 27510577.0,
      "step": 12340
    },
    {
      "entropy": 5.513951683044434,
      "epoch": 1.3864772280563824,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.3258,
      "mean_token_accuracy": 0.19546688497066497,
      "num_tokens": 27521441.0,
      "step": 12345
    },
    {
      "entropy": 5.481344890594483,
      "epoch": 1.3870388049643398,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.2698,
      "mean_token_accuracy": 0.1940573751926422,
      "num_tokens": 27533443.0,
      "step": 12350
    },
    {
      "entropy": 5.492029619216919,
      "epoch": 1.3876003818722973,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.3216,
      "mean_token_accuracy": 0.19381196796894073,
      "num_tokens": 27543872.0,
      "step": 12355
    },
    {
      "entropy": 5.63539752960205,
      "epoch": 1.388161958780255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.513,
      "mean_token_accuracy": 0.18065459430217742,
      "num_tokens": 27556556.0,
      "step": 12360
    },
    {
      "entropy": 5.505446863174439,
      "epoch": 1.3887235356882126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.3759,
      "mean_token_accuracy": 0.18792864829301834,
      "num_tokens": 27570739.0,
      "step": 12365
    },
    {
      "entropy": 5.530737257003784,
      "epoch": 1.38928511259617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.4155,
      "mean_token_accuracy": 0.18117647469043732,
      "num_tokens": 27582438.0,
      "step": 12370
    },
    {
      "entropy": 5.536424160003662,
      "epoch": 1.3898466895041275,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.3438,
      "mean_token_accuracy": 0.19049696773290634,
      "num_tokens": 27593854.0,
      "step": 12375
    },
    {
      "entropy": 5.562060594558716,
      "epoch": 1.3904082664120851,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19260254949331285,
      "num_tokens": 27604459.0,
      "step": 12380
    },
    {
      "entropy": 5.590554714202881,
      "epoch": 1.3909698433200428,
      "grad_norm": 1.125,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.4944,
      "mean_token_accuracy": 0.17838713377714158,
      "num_tokens": 27617901.0,
      "step": 12385
    },
    {
      "entropy": 5.562151765823364,
      "epoch": 1.3915314202280002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.4018,
      "mean_token_accuracy": 0.18608833253383636,
      "num_tokens": 27628574.0,
      "step": 12390
    },
    {
      "entropy": 5.520782613754273,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.125,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.271,
      "mean_token_accuracy": 0.19218895584344864,
      "num_tokens": 27639195.0,
      "step": 12395
    },
    {
      "entropy": 5.433777475357056,
      "epoch": 1.3926545740439153,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.203,
      "mean_token_accuracy": 0.2015116885304451,
      "num_tokens": 27649898.0,
      "step": 12400
    },
    {
      "entropy": 5.486884546279907,
      "epoch": 1.3932161509518728,
      "grad_norm": 1.09375,
      "learning_rate": 0.000481621822323602,
      "loss": 5.351,
      "mean_token_accuracy": 0.1842460259795189,
      "num_tokens": 27663617.0,
      "step": 12405
    },
    {
      "entropy": 5.380657386779785,
      "epoch": 1.3937777278598305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.2187,
      "mean_token_accuracy": 0.19785332977771758,
      "num_tokens": 27674823.0,
      "step": 12410
    },
    {
      "entropy": 5.4988696575164795,
      "epoch": 1.394339304767788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19219503402709961,
      "num_tokens": 27686483.0,
      "step": 12415
    },
    {
      "entropy": 5.531269931793213,
      "epoch": 1.3949008816757456,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.2916,
      "mean_token_accuracy": 0.19547995328903198,
      "num_tokens": 27697081.0,
      "step": 12420
    },
    {
      "entropy": 5.444240856170654,
      "epoch": 1.395462458583703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19486741870641708,
      "num_tokens": 27706710.0,
      "step": 12425
    },
    {
      "entropy": 5.48530011177063,
      "epoch": 1.3960240354916607,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.3902,
      "mean_token_accuracy": 0.18007151484489442,
      "num_tokens": 27718213.0,
      "step": 12430
    },
    {
      "entropy": 5.476761341094971,
      "epoch": 1.396585612399618,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1885454311966896,
      "num_tokens": 27729064.0,
      "step": 12435
    },
    {
      "entropy": 5.53508186340332,
      "epoch": 1.3971471893075758,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.3047,
      "mean_token_accuracy": 0.19287462383508683,
      "num_tokens": 27740121.0,
      "step": 12440
    },
    {
      "entropy": 5.447685861587525,
      "epoch": 1.3977087662155332,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.3476,
      "mean_token_accuracy": 0.1958514094352722,
      "num_tokens": 27751597.0,
      "step": 12445
    },
    {
      "entropy": 5.409804916381836,
      "epoch": 1.3982703431234906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19108182787895203,
      "num_tokens": 27761539.0,
      "step": 12450
    },
    {
      "entropy": 5.4863481521606445,
      "epoch": 1.3988319200314483,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.2594,
      "mean_token_accuracy": 0.1919708341360092,
      "num_tokens": 27772904.0,
      "step": 12455
    },
    {
      "entropy": 5.51110053062439,
      "epoch": 1.399393496939406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.266,
      "mean_token_accuracy": 0.19515890181064605,
      "num_tokens": 27783706.0,
      "step": 12460
    },
    {
      "entropy": 5.458293867111206,
      "epoch": 1.3999550738473634,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.2888,
      "mean_token_accuracy": 0.18573704957962037,
      "num_tokens": 27794503.0,
      "step": 12465
    },
    {
      "entropy": 5.452811193466187,
      "epoch": 1.4005166507553208,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.283,
      "mean_token_accuracy": 0.19184025824069978,
      "num_tokens": 27805952.0,
      "step": 12470
    },
    {
      "entropy": 5.502015399932861,
      "epoch": 1.4010782276632785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.2885,
      "mean_token_accuracy": 0.1947245940566063,
      "num_tokens": 27817530.0,
      "step": 12475
    },
    {
      "entropy": 5.519699287414551,
      "epoch": 1.4016398045712362,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.3536,
      "mean_token_accuracy": 0.18735283315181733,
      "num_tokens": 27829932.0,
      "step": 12480
    },
    {
      "entropy": 5.51725401878357,
      "epoch": 1.4022013814791936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.3415,
      "mean_token_accuracy": 0.19383803606033326,
      "num_tokens": 27840993.0,
      "step": 12485
    },
    {
      "entropy": 5.514503240585327,
      "epoch": 1.402762958387151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.2684,
      "mean_token_accuracy": 0.19149025678634643,
      "num_tokens": 27852727.0,
      "step": 12490
    },
    {
      "entropy": 5.466609477996826,
      "epoch": 1.4033245352951087,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.4397,
      "mean_token_accuracy": 0.1922093689441681,
      "num_tokens": 27865183.0,
      "step": 12495
    },
    {
      "entropy": 5.410083341598511,
      "epoch": 1.4038861122030661,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.1917,
      "mean_token_accuracy": 0.20270057320594786,
      "num_tokens": 27877266.0,
      "step": 12500
    },
    {
      "entropy": 5.548344945907592,
      "epoch": 1.4044476891110238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.3368,
      "mean_token_accuracy": 0.18815984725952148,
      "num_tokens": 27888863.0,
      "step": 12505
    },
    {
      "entropy": 5.452338647842407,
      "epoch": 1.4050092660189812,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.274,
      "mean_token_accuracy": 0.19844181090593338,
      "num_tokens": 27900443.0,
      "step": 12510
    },
    {
      "entropy": 5.4826854228973385,
      "epoch": 1.405570842926939,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.2625,
      "mean_token_accuracy": 0.19299466013908387,
      "num_tokens": 27911241.0,
      "step": 12515
    },
    {
      "entropy": 5.488195991516113,
      "epoch": 1.4061324198348963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19225993007421494,
      "num_tokens": 27922171.0,
      "step": 12520
    },
    {
      "entropy": 5.38026533126831,
      "epoch": 1.406693996742854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.2409,
      "mean_token_accuracy": 0.1974858194589615,
      "num_tokens": 27933627.0,
      "step": 12525
    },
    {
      "entropy": 5.318044519424438,
      "epoch": 1.4072555736508114,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.2029,
      "mean_token_accuracy": 0.18923496454954147,
      "num_tokens": 27943585.0,
      "step": 12530
    },
    {
      "entropy": 5.532860851287841,
      "epoch": 1.407817150558769,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.317,
      "mean_token_accuracy": 0.1894823729991913,
      "num_tokens": 27955554.0,
      "step": 12535
    },
    {
      "entropy": 5.52153058052063,
      "epoch": 1.4083787274667265,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.1996,
      "mean_token_accuracy": 0.1972256824374199,
      "num_tokens": 27966150.0,
      "step": 12540
    },
    {
      "entropy": 5.426303052902222,
      "epoch": 1.408940304374684,
      "grad_norm": 1.125,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.2422,
      "mean_token_accuracy": 0.19363798648118974,
      "num_tokens": 27977436.0,
      "step": 12545
    },
    {
      "entropy": 5.5245991230010985,
      "epoch": 1.4095018812826416,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.4071,
      "mean_token_accuracy": 0.18410052061080934,
      "num_tokens": 27988325.0,
      "step": 12550
    },
    {
      "entropy": 5.561803531646729,
      "epoch": 1.4100634581905993,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1947012275457382,
      "num_tokens": 27999760.0,
      "step": 12555
    },
    {
      "entropy": 5.545596790313721,
      "epoch": 1.4106250350985567,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.409,
      "mean_token_accuracy": 0.1847024142742157,
      "num_tokens": 28011333.0,
      "step": 12560
    },
    {
      "entropy": 5.42786660194397,
      "epoch": 1.4111866120065142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.3376,
      "mean_token_accuracy": 0.19361564218997956,
      "num_tokens": 28023464.0,
      "step": 12565
    },
    {
      "entropy": 5.443120670318604,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.171875,
      "learning_rate": 0.000481093790794317,
      "loss": 5.2982,
      "mean_token_accuracy": 0.184749773144722,
      "num_tokens": 28034205.0,
      "step": 12570
    },
    {
      "entropy": 5.538404178619385,
      "epoch": 1.4123097658224295,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.288,
      "mean_token_accuracy": 0.18942586928606034,
      "num_tokens": 28045846.0,
      "step": 12575
    },
    {
      "entropy": 5.497859764099121,
      "epoch": 1.412871342730387,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19742057472467422,
      "num_tokens": 28057365.0,
      "step": 12580
    },
    {
      "entropy": 5.482359743118286,
      "epoch": 1.4134329196383444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.3851,
      "mean_token_accuracy": 0.1854017272591591,
      "num_tokens": 28069098.0,
      "step": 12585
    },
    {
      "entropy": 5.507374238967896,
      "epoch": 1.413994496546302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1919554203748703,
      "num_tokens": 28080008.0,
      "step": 12590
    },
    {
      "entropy": 5.577986812591552,
      "epoch": 1.4145560734542595,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.3612,
      "mean_token_accuracy": 0.18817386478185655,
      "num_tokens": 28091657.0,
      "step": 12595
    },
    {
      "entropy": 5.4135431289672855,
      "epoch": 1.4151176503622171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1927874505519867,
      "num_tokens": 28102586.0,
      "step": 12600
    },
    {
      "entropy": 5.466880178451538,
      "epoch": 1.4156792272701746,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.3369,
      "mean_token_accuracy": 0.1913236677646637,
      "num_tokens": 28113066.0,
      "step": 12605
    },
    {
      "entropy": 5.525369787216187,
      "epoch": 1.4162408041781323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.3299,
      "mean_token_accuracy": 0.18975775837898254,
      "num_tokens": 28124814.0,
      "step": 12610
    },
    {
      "entropy": 5.601166725158691,
      "epoch": 1.4168023810860897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.3664,
      "mean_token_accuracy": 0.18623321056365966,
      "num_tokens": 28138009.0,
      "step": 12615
    },
    {
      "entropy": 5.39503927230835,
      "epoch": 1.4173639579940474,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.2687,
      "mean_token_accuracy": 0.19271760582923889,
      "num_tokens": 28149222.0,
      "step": 12620
    },
    {
      "entropy": 5.469629001617432,
      "epoch": 1.4179255349020048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.33,
      "mean_token_accuracy": 0.19065048396587372,
      "num_tokens": 28161323.0,
      "step": 12625
    },
    {
      "entropy": 5.497536325454712,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19020311385393143,
      "num_tokens": 28173242.0,
      "step": 12630
    },
    {
      "entropy": 5.459263563156128,
      "epoch": 1.41904868871792,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.2768,
      "mean_token_accuracy": 0.19995663762092591,
      "num_tokens": 28183927.0,
      "step": 12635
    },
    {
      "entropy": 5.451407861709595,
      "epoch": 1.4196102656258773,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.2704,
      "mean_token_accuracy": 0.19252085089683532,
      "num_tokens": 28194375.0,
      "step": 12640
    },
    {
      "entropy": 5.494913482666016,
      "epoch": 1.420171842533835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.3253,
      "mean_token_accuracy": 0.19481477290391921,
      "num_tokens": 28206492.0,
      "step": 12645
    },
    {
      "entropy": 5.480983066558838,
      "epoch": 1.4207334194417927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.3536,
      "mean_token_accuracy": 0.18314705938100814,
      "num_tokens": 28217795.0,
      "step": 12650
    },
    {
      "entropy": 5.431267499923706,
      "epoch": 1.42129499634975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.2972,
      "mean_token_accuracy": 0.1943140760064125,
      "num_tokens": 28228796.0,
      "step": 12655
    },
    {
      "entropy": 5.507163000106812,
      "epoch": 1.4218565732577075,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19252949357032775,
      "num_tokens": 28238873.0,
      "step": 12660
    },
    {
      "entropy": 5.46311354637146,
      "epoch": 1.4224181501656652,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048078652978521914,
      "loss": 5.1383,
      "mean_token_accuracy": 0.205023892223835,
      "num_tokens": 28249832.0,
      "step": 12665
    },
    {
      "entropy": 5.472410678863525,
      "epoch": 1.4229797270736229,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.3395,
      "mean_token_accuracy": 0.18793419599533082,
      "num_tokens": 28261532.0,
      "step": 12670
    },
    {
      "entropy": 5.557413625717163,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.3666,
      "mean_token_accuracy": 0.19007611721754075,
      "num_tokens": 28272008.0,
      "step": 12675
    },
    {
      "entropy": 5.501730108261109,
      "epoch": 1.4241028808895377,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.3299,
      "mean_token_accuracy": 0.19588117748498918,
      "num_tokens": 28283239.0,
      "step": 12680
    },
    {
      "entropy": 5.544003295898437,
      "epoch": 1.4246644577974954,
      "grad_norm": 1.0625,
      "learning_rate": 0.000480721541731488,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18821366429328917,
      "num_tokens": 28293597.0,
      "step": 12685
    },
    {
      "entropy": 5.4740571022033695,
      "epoch": 1.4252260347054528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.2833,
      "mean_token_accuracy": 0.1938060313463211,
      "num_tokens": 28305124.0,
      "step": 12690
    },
    {
      "entropy": 5.422293663024902,
      "epoch": 1.4257876116134105,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19523887783288957,
      "num_tokens": 28315875.0,
      "step": 12695
    },
    {
      "entropy": 5.480016803741455,
      "epoch": 1.426349188521368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.2741,
      "mean_token_accuracy": 0.1949415236711502,
      "num_tokens": 28327705.0,
      "step": 12700
    },
    {
      "entropy": 5.4407182216644285,
      "epoch": 1.4269107654293256,
      "grad_norm": 1.0,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.2614,
      "mean_token_accuracy": 0.1960237666964531,
      "num_tokens": 28338300.0,
      "step": 12705
    },
    {
      "entropy": 5.4076940536499025,
      "epoch": 1.427472342337283,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.2238,
      "mean_token_accuracy": 0.1996994972229004,
      "num_tokens": 28348485.0,
      "step": 12710
    },
    {
      "entropy": 5.481090927124024,
      "epoch": 1.4280339192452407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19686545729637145,
      "num_tokens": 28359118.0,
      "step": 12715
    },
    {
      "entropy": 5.447401285171509,
      "epoch": 1.4285954961531981,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.2881,
      "mean_token_accuracy": 0.19529957473278045,
      "num_tokens": 28369608.0,
      "step": 12720
    },
    {
      "entropy": 5.492588376998901,
      "epoch": 1.4291570730611558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.2821,
      "mean_token_accuracy": 0.1956749364733696,
      "num_tokens": 28379709.0,
      "step": 12725
    },
    {
      "entropy": 5.471444416046142,
      "epoch": 1.4297186499691132,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19217366576194764,
      "num_tokens": 28390758.0,
      "step": 12730
    },
    {
      "entropy": 5.665639638900757,
      "epoch": 1.4302802268770707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000480558613151976,
      "loss": 5.5601,
      "mean_token_accuracy": 0.1810941606760025,
      "num_tokens": 28403485.0,
      "step": 12735
    },
    {
      "entropy": 5.611013650894165,
      "epoch": 1.4308418037850283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.3605,
      "mean_token_accuracy": 0.1880544364452362,
      "num_tokens": 28413627.0,
      "step": 12740
    },
    {
      "entropy": 5.499139595031738,
      "epoch": 1.431403380692986,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1897425964474678,
      "num_tokens": 28424202.0,
      "step": 12745
    },
    {
      "entropy": 5.356061124801636,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.125,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.2211,
      "mean_token_accuracy": 0.1956148400902748,
      "num_tokens": 28435136.0,
      "step": 12750
    },
    {
      "entropy": 5.407910776138306,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000480493258429251,
      "loss": 5.235,
      "mean_token_accuracy": 0.19518414586782457,
      "num_tokens": 28444915.0,
      "step": 12755
    },
    {
      "entropy": 5.464826679229736,
      "epoch": 1.4330881114168585,
      "grad_norm": 1.03125,
      "learning_rate": 0.000480476903389806,
      "loss": 5.3008,
      "mean_token_accuracy": 0.1882382333278656,
      "num_tokens": 28458295.0,
      "step": 12760
    },
    {
      "entropy": 5.542707633972168,
      "epoch": 1.4336496883248162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.349,
      "mean_token_accuracy": 0.1964672848582268,
      "num_tokens": 28469587.0,
      "step": 12765
    },
    {
      "entropy": 5.559011840820313,
      "epoch": 1.4342112652327736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.3453,
      "mean_token_accuracy": 0.1865287646651268,
      "num_tokens": 28481498.0,
      "step": 12770
    },
    {
      "entropy": 5.524534511566162,
      "epoch": 1.434772842140731,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18070788830518722,
      "num_tokens": 28492626.0,
      "step": 12775
    },
    {
      "entropy": 5.497053003311157,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.3346,
      "mean_token_accuracy": 0.19229722023010254,
      "num_tokens": 28503882.0,
      "step": 12780
    },
    {
      "entropy": 5.532601833343506,
      "epoch": 1.4358959959566462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.2667,
      "mean_token_accuracy": 0.20189556181430818,
      "num_tokens": 28514786.0,
      "step": 12785
    },
    {
      "entropy": 5.491387796401978,
      "epoch": 1.4364575728646038,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.3003,
      "mean_token_accuracy": 0.19508771896362304,
      "num_tokens": 28526045.0,
      "step": 12790
    },
    {
      "entropy": 5.531581211090088,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1926025003194809,
      "num_tokens": 28536773.0,
      "step": 12795
    },
    {
      "entropy": 5.442726755142212,
      "epoch": 1.437580726680519,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.2426,
      "mean_token_accuracy": 0.19309162199497223,
      "num_tokens": 28548099.0,
      "step": 12800
    },
    {
      "entropy": 5.45169415473938,
      "epoch": 1.4381423035884764,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.3134,
      "mean_token_accuracy": 0.19566312730312346,
      "num_tokens": 28559372.0,
      "step": 12805
    },
    {
      "entropy": 5.4630919933319095,
      "epoch": 1.438703880496434,
      "grad_norm": 1.25,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1951773449778557,
      "num_tokens": 28570106.0,
      "step": 12810
    },
    {
      "entropy": 5.457654142379761,
      "epoch": 1.4392654574043915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000480296566267982,
      "loss": 5.3312,
      "mean_token_accuracy": 0.19525408297777175,
      "num_tokens": 28582417.0,
      "step": 12815
    },
    {
      "entropy": 5.549852800369263,
      "epoch": 1.4398270343123492,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.4421,
      "mean_token_accuracy": 0.19046890884637832,
      "num_tokens": 28594341.0,
      "step": 12820
    },
    {
      "entropy": 5.497651624679565,
      "epoch": 1.4403886112203066,
      "grad_norm": 1.125,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.3253,
      "mean_token_accuracy": 0.19379577934741973,
      "num_tokens": 28605580.0,
      "step": 12825
    },
    {
      "entropy": 5.535317039489746,
      "epoch": 1.440950188128264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.3568,
      "mean_token_accuracy": 0.1853962793946266,
      "num_tokens": 28617779.0,
      "step": 12830
    },
    {
      "entropy": 5.543725109100341,
      "epoch": 1.4415117650362217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18570167124271392,
      "num_tokens": 28629714.0,
      "step": 12835
    },
    {
      "entropy": 5.555856466293335,
      "epoch": 1.4420733419441794,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.3589,
      "mean_token_accuracy": 0.18780214488506317,
      "num_tokens": 28641381.0,
      "step": 12840
    },
    {
      "entropy": 5.518057870864868,
      "epoch": 1.4426349188521368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.3499,
      "mean_token_accuracy": 0.1917356953024864,
      "num_tokens": 28653650.0,
      "step": 12845
    },
    {
      "entropy": 5.477000761032104,
      "epoch": 1.4431964957600942,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.2695,
      "mean_token_accuracy": 0.19412501752376557,
      "num_tokens": 28665252.0,
      "step": 12850
    },
    {
      "entropy": 5.452428817749023,
      "epoch": 1.443758072668052,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.3091,
      "mean_token_accuracy": 0.19653584510087968,
      "num_tokens": 28677072.0,
      "step": 12855
    },
    {
      "entropy": 5.423432111740112,
      "epoch": 1.4443196495760096,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.2378,
      "mean_token_accuracy": 0.19026581794023514,
      "num_tokens": 28688767.0,
      "step": 12860
    },
    {
      "entropy": 5.538924264907837,
      "epoch": 1.444881226483967,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.5004,
      "mean_token_accuracy": 0.18820954412221907,
      "num_tokens": 28699773.0,
      "step": 12865
    },
    {
      "entropy": 5.571140956878662,
      "epoch": 1.4454428033919244,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18863944411277772,
      "num_tokens": 28710971.0,
      "step": 12870
    },
    {
      "entropy": 5.489381122589111,
      "epoch": 1.446004380299882,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.2367,
      "mean_token_accuracy": 0.19461744278669357,
      "num_tokens": 28721544.0,
      "step": 12875
    },
    {
      "entropy": 5.541806602478028,
      "epoch": 1.4465659572078395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18645317256450653,
      "num_tokens": 28733790.0,
      "step": 12880
    },
    {
      "entropy": 5.582160615921021,
      "epoch": 1.4471275341157972,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.3539,
      "mean_token_accuracy": 0.19444509148597716,
      "num_tokens": 28745218.0,
      "step": 12885
    },
    {
      "entropy": 5.6067163944244385,
      "epoch": 1.4476891110237546,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.4006,
      "mean_token_accuracy": 0.18395177125930787,
      "num_tokens": 28758061.0,
      "step": 12890
    },
    {
      "entropy": 5.444657278060913,
      "epoch": 1.4482506879317123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.2122,
      "mean_token_accuracy": 0.20201971381902695,
      "num_tokens": 28770037.0,
      "step": 12895
    },
    {
      "entropy": 5.561487770080566,
      "epoch": 1.4488122648396697,
      "grad_norm": 1.125,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.3851,
      "mean_token_accuracy": 0.18889002948999406,
      "num_tokens": 28781052.0,
      "step": 12900
    },
    {
      "entropy": 5.435133695602417,
      "epoch": 1.4493738417476274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.2454,
      "mean_token_accuracy": 0.19051423966884612,
      "num_tokens": 28791225.0,
      "step": 12905
    },
    {
      "entropy": 5.48399248123169,
      "epoch": 1.4499354186555848,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.2937,
      "mean_token_accuracy": 0.19318269491195678,
      "num_tokens": 28803022.0,
      "step": 12910
    },
    {
      "entropy": 5.42376823425293,
      "epoch": 1.4504969955635425,
      "grad_norm": 1.125,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.2714,
      "mean_token_accuracy": 0.1893945261836052,
      "num_tokens": 28813789.0,
      "step": 12915
    },
    {
      "entropy": 5.540867328643799,
      "epoch": 1.4510585724715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19015136808156968,
      "num_tokens": 28825015.0,
      "step": 12920
    },
    {
      "entropy": 5.6369280338287355,
      "epoch": 1.4516201493794574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18817705661058426,
      "num_tokens": 28835852.0,
      "step": 12925
    },
    {
      "entropy": 5.461359071731567,
      "epoch": 1.452181726287415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.3229,
      "mean_token_accuracy": 0.19205658584833146,
      "num_tokens": 28846684.0,
      "step": 12930
    },
    {
      "entropy": 5.46303915977478,
      "epoch": 1.4527433031953727,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18716940730810167,
      "num_tokens": 28857402.0,
      "step": 12935
    },
    {
      "entropy": 5.437236499786377,
      "epoch": 1.4533048801033301,
      "grad_norm": 1.125,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.1814,
      "mean_token_accuracy": 0.19869955033063888,
      "num_tokens": 28867866.0,
      "step": 12940
    },
    {
      "entropy": 5.452563762664795,
      "epoch": 1.4538664570112876,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19044434130191804,
      "num_tokens": 28878102.0,
      "step": 12945
    },
    {
      "entropy": 5.587057209014892,
      "epoch": 1.4544280339192452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.393,
      "mean_token_accuracy": 0.1933970719575882,
      "num_tokens": 28888981.0,
      "step": 12950
    },
    {
      "entropy": 5.460827445983886,
      "epoch": 1.454989610827203,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.2587,
      "mean_token_accuracy": 0.19057833701372145,
      "num_tokens": 28899733.0,
      "step": 12955
    },
    {
      "entropy": 5.42146863937378,
      "epoch": 1.4555511877351603,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.2516,
      "mean_token_accuracy": 0.1928494781255722,
      "num_tokens": 28910701.0,
      "step": 12960
    },
    {
      "entropy": 5.516459035873413,
      "epoch": 1.4561127646431178,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.3255,
      "mean_token_accuracy": 0.18599175214767455,
      "num_tokens": 28922627.0,
      "step": 12965
    },
    {
      "entropy": 5.539343929290771,
      "epoch": 1.4566743415510754,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.3582,
      "mean_token_accuracy": 0.19016606211662293,
      "num_tokens": 28933292.0,
      "step": 12970
    },
    {
      "entropy": 5.425318622589112,
      "epoch": 1.4572359184590329,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.2292,
      "mean_token_accuracy": 0.1934402659535408,
      "num_tokens": 28943898.0,
      "step": 12975
    },
    {
      "entropy": 5.461353921890259,
      "epoch": 1.4577974953669905,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.3243,
      "mean_token_accuracy": 0.19086166322231293,
      "num_tokens": 28955248.0,
      "step": 12980
    },
    {
      "entropy": 5.464456987380982,
      "epoch": 1.458359072274948,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.2137,
      "mean_token_accuracy": 0.19882305711507797,
      "num_tokens": 28966905.0,
      "step": 12985
    },
    {
      "entropy": 5.506450462341308,
      "epoch": 1.4589206491829056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1928555890917778,
      "num_tokens": 28978515.0,
      "step": 12990
    },
    {
      "entropy": 5.426952695846557,
      "epoch": 1.459482226090863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.3111,
      "mean_token_accuracy": 0.18737949877977372,
      "num_tokens": 28990185.0,
      "step": 12995
    },
    {
      "entropy": 5.498533773422241,
      "epoch": 1.4600438029988208,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.3377,
      "mean_token_accuracy": 0.1894269436597824,
      "num_tokens": 29000779.0,
      "step": 13000
    },
    {
      "entropy": 5.567147445678711,
      "epoch": 1.4606053799067782,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.4434,
      "mean_token_accuracy": 0.1878877818584442,
      "num_tokens": 29012066.0,
      "step": 13005
    },
    {
      "entropy": 5.660548448562622,
      "epoch": 1.4611669568147359,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18517484068870543,
      "num_tokens": 29023416.0,
      "step": 13010
    },
    {
      "entropy": 5.489505100250244,
      "epoch": 1.4617285337226933,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19129559099674226,
      "num_tokens": 29034469.0,
      "step": 13015
    },
    {
      "entropy": 5.468514680862427,
      "epoch": 1.4622901106306507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.2571,
      "mean_token_accuracy": 0.18737759739160537,
      "num_tokens": 29045306.0,
      "step": 13020
    },
    {
      "entropy": 5.424688482284546,
      "epoch": 1.4628516875386084,
      "grad_norm": 1.046875,
      "learning_rate": 0.000479600737020506,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19463146477937698,
      "num_tokens": 29057114.0,
      "step": 13025
    },
    {
      "entropy": 5.459911918640136,
      "epoch": 1.463413264446566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.2317,
      "mean_token_accuracy": 0.18986470401287078,
      "num_tokens": 29067250.0,
      "step": 13030
    },
    {
      "entropy": 5.441561079025268,
      "epoch": 1.4639748413545235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.2798,
      "mean_token_accuracy": 0.19596178978681564,
      "num_tokens": 29078582.0,
      "step": 13035
    },
    {
      "entropy": 5.460959815979004,
      "epoch": 1.464536418262481,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.3071,
      "mean_token_accuracy": 0.19390060752630234,
      "num_tokens": 29090113.0,
      "step": 13040
    },
    {
      "entropy": 5.419605445861817,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.2842,
      "mean_token_accuracy": 0.1960938185453415,
      "num_tokens": 29101117.0,
      "step": 13045
    },
    {
      "entropy": 5.534759855270385,
      "epoch": 1.4656595720783963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.3071,
      "mean_token_accuracy": 0.19129250943660736,
      "num_tokens": 29112569.0,
      "step": 13050
    },
    {
      "entropy": 5.562998580932617,
      "epoch": 1.4662211489863537,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.4059,
      "mean_token_accuracy": 0.19032600075006484,
      "num_tokens": 29124462.0,
      "step": 13055
    },
    {
      "entropy": 5.4725611209869385,
      "epoch": 1.4667827258943111,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.2544,
      "mean_token_accuracy": 0.19727838188409805,
      "num_tokens": 29135401.0,
      "step": 13060
    },
    {
      "entropy": 5.46421971321106,
      "epoch": 1.4673443028022688,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.2914,
      "mean_token_accuracy": 0.18822217732667923,
      "num_tokens": 29146601.0,
      "step": 13065
    },
    {
      "entropy": 5.461003684997559,
      "epoch": 1.4679058797102262,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.2827,
      "mean_token_accuracy": 0.19355401992797852,
      "num_tokens": 29157299.0,
      "step": 13070
    },
    {
      "entropy": 5.557260990142822,
      "epoch": 1.468467456618184,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.3788,
      "mean_token_accuracy": 0.18798348754644395,
      "num_tokens": 29169101.0,
      "step": 13075
    },
    {
      "entropy": 5.402161026000977,
      "epoch": 1.4690290335261413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.1954,
      "mean_token_accuracy": 0.19632158428430557,
      "num_tokens": 29181286.0,
      "step": 13080
    },
    {
      "entropy": 5.496028709411621,
      "epoch": 1.469590610434099,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.3725,
      "mean_token_accuracy": 0.18671411275863647,
      "num_tokens": 29191181.0,
      "step": 13085
    },
    {
      "entropy": 5.458876609802246,
      "epoch": 1.4701521873420564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1940981224179268,
      "num_tokens": 29202426.0,
      "step": 13090
    },
    {
      "entropy": 5.44552435874939,
      "epoch": 1.470713764250014,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19542424231767655,
      "num_tokens": 29212817.0,
      "step": 13095
    },
    {
      "entropy": 5.447294235229492,
      "epoch": 1.4712753411579715,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.2564,
      "mean_token_accuracy": 0.19038020670413972,
      "num_tokens": 29224644.0,
      "step": 13100
    },
    {
      "entropy": 5.465914535522461,
      "epoch": 1.4718369180659292,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.278,
      "mean_token_accuracy": 0.1904280200600624,
      "num_tokens": 29236443.0,
      "step": 13105
    },
    {
      "entropy": 5.559574317932129,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.4181,
      "mean_token_accuracy": 0.18571652621030807,
      "num_tokens": 29247687.0,
      "step": 13110
    },
    {
      "entropy": 5.531558847427368,
      "epoch": 1.4729600718818443,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19203899800777435,
      "num_tokens": 29259739.0,
      "step": 13115
    },
    {
      "entropy": 5.515462970733642,
      "epoch": 1.4735216487898017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19106705933809282,
      "num_tokens": 29270834.0,
      "step": 13120
    },
    {
      "entropy": 5.5195942401885985,
      "epoch": 1.4740832256977594,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.3097,
      "mean_token_accuracy": 0.1902782216668129,
      "num_tokens": 29282291.0,
      "step": 13125
    },
    {
      "entropy": 5.539123392105102,
      "epoch": 1.4746448026057168,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18881076127290725,
      "num_tokens": 29293836.0,
      "step": 13130
    },
    {
      "entropy": 5.499612712860108,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.3694,
      "mean_token_accuracy": 0.18158402442932128,
      "num_tokens": 29305550.0,
      "step": 13135
    },
    {
      "entropy": 5.549182415008545,
      "epoch": 1.475767956421632,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.3601,
      "mean_token_accuracy": 0.18821000307798386,
      "num_tokens": 29316973.0,
      "step": 13140
    },
    {
      "entropy": 5.491867399215698,
      "epoch": 1.4763295333295896,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.2121,
      "mean_token_accuracy": 0.20074503421783446,
      "num_tokens": 29327743.0,
      "step": 13145
    },
    {
      "entropy": 5.446516513824463,
      "epoch": 1.476891110237547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.2028,
      "mean_token_accuracy": 0.18895393311977388,
      "num_tokens": 29338286.0,
      "step": 13150
    },
    {
      "entropy": 5.45173192024231,
      "epoch": 1.4774526871455045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.3388,
      "mean_token_accuracy": 0.1874464675784111,
      "num_tokens": 29348877.0,
      "step": 13155
    },
    {
      "entropy": 5.388352346420288,
      "epoch": 1.4780142640534621,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047914734744424517,
      "loss": 5.1175,
      "mean_token_accuracy": 0.20348843485116958,
      "num_tokens": 29360072.0,
      "step": 13160
    },
    {
      "entropy": 5.493221950531006,
      "epoch": 1.4785758409614196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.3624,
      "mean_token_accuracy": 0.19179106056690215,
      "num_tokens": 29370744.0,
      "step": 13165
    },
    {
      "entropy": 5.541601181030273,
      "epoch": 1.4791374178693772,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.3143,
      "mean_token_accuracy": 0.18578248023986815,
      "num_tokens": 29381463.0,
      "step": 13170
    },
    {
      "entropy": 5.493764734268188,
      "epoch": 1.4796989947773347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.3297,
      "mean_token_accuracy": 0.19343615770339967,
      "num_tokens": 29394227.0,
      "step": 13175
    },
    {
      "entropy": 5.433274221420288,
      "epoch": 1.4802605716852923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.2525,
      "mean_token_accuracy": 0.1978638008236885,
      "num_tokens": 29404754.0,
      "step": 13180
    },
    {
      "entropy": 5.422193717956543,
      "epoch": 1.4808221485932498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19936129003763198,
      "num_tokens": 29416161.0,
      "step": 13185
    },
    {
      "entropy": 5.415967893600464,
      "epoch": 1.4813837255012074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.1809,
      "mean_token_accuracy": 0.20436224192380906,
      "num_tokens": 29426587.0,
      "step": 13190
    },
    {
      "entropy": 5.449272394180298,
      "epoch": 1.4819453024091649,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.3465,
      "mean_token_accuracy": 0.18565037846565247,
      "num_tokens": 29437488.0,
      "step": 13195
    },
    {
      "entropy": 5.500899744033814,
      "epoch": 1.4825068793171226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.322,
      "mean_token_accuracy": 0.19296833872795105,
      "num_tokens": 29448244.0,
      "step": 13200
    },
    {
      "entropy": 5.587065601348877,
      "epoch": 1.48306845622508,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.3906,
      "mean_token_accuracy": 0.18762186765670777,
      "num_tokens": 29458260.0,
      "step": 13205
    },
    {
      "entropy": 5.605184030532837,
      "epoch": 1.4836300331330377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1924204185605049,
      "num_tokens": 29470874.0,
      "step": 13210
    },
    {
      "entropy": 5.439106416702271,
      "epoch": 1.484191610040995,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.251,
      "mean_token_accuracy": 0.19099540412425994,
      "num_tokens": 29481442.0,
      "step": 13215
    },
    {
      "entropy": 5.515398836135864,
      "epoch": 1.4847531869489528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.3977,
      "mean_token_accuracy": 0.1918204203248024,
      "num_tokens": 29492727.0,
      "step": 13220
    },
    {
      "entropy": 5.5656898021698,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.3295,
      "mean_token_accuracy": 0.1868348702788353,
      "num_tokens": 29502729.0,
      "step": 13225
    },
    {
      "entropy": 5.583171081542969,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19567034244537354,
      "num_tokens": 29514315.0,
      "step": 13230
    },
    {
      "entropy": 5.482412528991699,
      "epoch": 1.4864379176728253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19366237223148347,
      "num_tokens": 29526030.0,
      "step": 13235
    },
    {
      "entropy": 5.40715765953064,
      "epoch": 1.486999494580783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.2592,
      "mean_token_accuracy": 0.19387319236993789,
      "num_tokens": 29536972.0,
      "step": 13240
    },
    {
      "entropy": 5.432439184188842,
      "epoch": 1.4875610714887404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.2998,
      "mean_token_accuracy": 0.19370868653059006,
      "num_tokens": 29547325.0,
      "step": 13245
    },
    {
      "entropy": 5.506995725631714,
      "epoch": 1.4881226483966978,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.3174,
      "mean_token_accuracy": 0.19362844675779342,
      "num_tokens": 29557944.0,
      "step": 13250
    },
    {
      "entropy": 5.386491537094116,
      "epoch": 1.4886842253046555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.2346,
      "mean_token_accuracy": 0.20049775540828704,
      "num_tokens": 29568618.0,
      "step": 13255
    },
    {
      "entropy": 5.5095103740692135,
      "epoch": 1.489245802212613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.3153,
      "mean_token_accuracy": 0.1876966953277588,
      "num_tokens": 29580682.0,
      "step": 13260
    },
    {
      "entropy": 5.556739950180054,
      "epoch": 1.4898073791205706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.3558,
      "mean_token_accuracy": 0.188737690448761,
      "num_tokens": 29592302.0,
      "step": 13265
    },
    {
      "entropy": 5.525094366073608,
      "epoch": 1.490368956028528,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.3769,
      "mean_token_accuracy": 0.18692763149738312,
      "num_tokens": 29603591.0,
      "step": 13270
    },
    {
      "entropy": 5.480088710784912,
      "epoch": 1.4909305329364857,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19129645675420762,
      "num_tokens": 29615542.0,
      "step": 13275
    },
    {
      "entropy": 5.581087923049926,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.3518,
      "mean_token_accuracy": 0.19800916761159898,
      "num_tokens": 29626650.0,
      "step": 13280
    },
    {
      "entropy": 5.442483901977539,
      "epoch": 1.4920536867524008,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19341592937707902,
      "num_tokens": 29637135.0,
      "step": 13285
    },
    {
      "entropy": 5.460017681121826,
      "epoch": 1.4926152636603582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18432952463626862,
      "num_tokens": 29648862.0,
      "step": 13290
    },
    {
      "entropy": 5.4734086990356445,
      "epoch": 1.493176840568316,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19206209033727645,
      "num_tokens": 29661015.0,
      "step": 13295
    },
    {
      "entropy": 5.575405168533325,
      "epoch": 1.4937384174762733,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.2945,
      "mean_token_accuracy": 0.1862231731414795,
      "num_tokens": 29673521.0,
      "step": 13300
    },
    {
      "entropy": 5.560371828079224,
      "epoch": 1.494299994384231,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.3667,
      "mean_token_accuracy": 0.19226312190294265,
      "num_tokens": 29685379.0,
      "step": 13305
    },
    {
      "entropy": 5.430146837234497,
      "epoch": 1.4948615712921884,
      "grad_norm": 1.125,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19056206047534943,
      "num_tokens": 29696372.0,
      "step": 13310
    },
    {
      "entropy": 5.481883001327515,
      "epoch": 1.495423148200146,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.2985,
      "mean_token_accuracy": 0.1957177847623825,
      "num_tokens": 29707401.0,
      "step": 13315
    },
    {
      "entropy": 5.429145479202271,
      "epoch": 1.4959847251081035,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.2746,
      "mean_token_accuracy": 0.19296327531337737,
      "num_tokens": 29717524.0,
      "step": 13320
    },
    {
      "entropy": 5.508581018447876,
      "epoch": 1.496546302016061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.3769,
      "mean_token_accuracy": 0.18857701867818832,
      "num_tokens": 29728235.0,
      "step": 13325
    },
    {
      "entropy": 5.580225276947021,
      "epoch": 1.4971078789240186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.3777,
      "mean_token_accuracy": 0.18802305459976196,
      "num_tokens": 29738800.0,
      "step": 13330
    },
    {
      "entropy": 5.5447821617126465,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.3214,
      "mean_token_accuracy": 0.19085756093263626,
      "num_tokens": 29750346.0,
      "step": 13335
    },
    {
      "entropy": 5.464204978942871,
      "epoch": 1.4982310327399337,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.2588,
      "mean_token_accuracy": 0.19634692519903182,
      "num_tokens": 29761931.0,
      "step": 13340
    },
    {
      "entropy": 5.406882429122925,
      "epoch": 1.4987926096478912,
      "grad_norm": 1.0,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19622450917959214,
      "num_tokens": 29773333.0,
      "step": 13345
    },
    {
      "entropy": 5.440138578414917,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19750139564275743,
      "num_tokens": 29785125.0,
      "step": 13350
    },
    {
      "entropy": 5.43392653465271,
      "epoch": 1.4999157634638065,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1961698442697525,
      "num_tokens": 29796288.0,
      "step": 13355
    },
    {
      "entropy": 5.52727780342102,
      "epoch": 1.500477340371764,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.3559,
      "mean_token_accuracy": 0.18841904401779175,
      "num_tokens": 29806348.0,
      "step": 13360
    },
    {
      "entropy": 5.5186844825744625,
      "epoch": 1.5010389172797214,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19519648402929307,
      "num_tokens": 29816342.0,
      "step": 13365
    },
    {
      "entropy": 5.414639472961426,
      "epoch": 1.501600494187679,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.2348,
      "mean_token_accuracy": 0.19205697178840636,
      "num_tokens": 29826789.0,
      "step": 13370
    },
    {
      "entropy": 5.472879409790039,
      "epoch": 1.5021620710956367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19075942933559417,
      "num_tokens": 29837633.0,
      "step": 13375
    },
    {
      "entropy": 5.540190505981445,
      "epoch": 1.502723648003594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1893691673874855,
      "num_tokens": 29848498.0,
      "step": 13380
    },
    {
      "entropy": 5.5391168117523195,
      "epoch": 1.5032852249115516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18855499476194382,
      "num_tokens": 29859777.0,
      "step": 13385
    },
    {
      "entropy": 5.739963531494141,
      "epoch": 1.5038468018195092,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.5347,
      "mean_token_accuracy": 0.1819843664765358,
      "num_tokens": 29873373.0,
      "step": 13390
    },
    {
      "entropy": 5.542487335205078,
      "epoch": 1.5044083787274667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.301,
      "mean_token_accuracy": 0.19327073991298677,
      "num_tokens": 29884749.0,
      "step": 13395
    },
    {
      "entropy": 5.38848180770874,
      "epoch": 1.5049699556354241,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.2235,
      "mean_token_accuracy": 0.19183147847652435,
      "num_tokens": 29895059.0,
      "step": 13400
    },
    {
      "entropy": 5.400285625457764,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.2841,
      "mean_token_accuracy": 0.1841924697160721,
      "num_tokens": 29906611.0,
      "step": 13405
    },
    {
      "entropy": 5.365358114242554,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20012490600347518,
      "num_tokens": 29918179.0,
      "step": 13410
    },
    {
      "entropy": 5.48297004699707,
      "epoch": 1.506654686359297,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18956200480461122,
      "num_tokens": 29929612.0,
      "step": 13415
    },
    {
      "entropy": 5.483995485305786,
      "epoch": 1.5072162632672543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.2149,
      "mean_token_accuracy": 0.20151948630809785,
      "num_tokens": 29939569.0,
      "step": 13420
    },
    {
      "entropy": 5.46502571105957,
      "epoch": 1.507777840175212,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19439188688993453,
      "num_tokens": 29950344.0,
      "step": 13425
    },
    {
      "entropy": 5.446194648742676,
      "epoch": 1.5083394170831697,
      "grad_norm": 1.125,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.2439,
      "mean_token_accuracy": 0.20005802065134048,
      "num_tokens": 29960748.0,
      "step": 13430
    },
    {
      "entropy": 5.604156732559204,
      "epoch": 1.508900993991127,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18356028199195862,
      "num_tokens": 29971678.0,
      "step": 13435
    },
    {
      "entropy": 5.423716306686401,
      "epoch": 1.5094625708990845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19111263155937194,
      "num_tokens": 29982477.0,
      "step": 13440
    },
    {
      "entropy": 5.465841484069824,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.3103,
      "mean_token_accuracy": 0.19847854226827621,
      "num_tokens": 29994539.0,
      "step": 13445
    },
    {
      "entropy": 5.5075523853302,
      "epoch": 1.5105857247149999,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.2557,
      "mean_token_accuracy": 0.19905335009098052,
      "num_tokens": 30004821.0,
      "step": 13450
    },
    {
      "entropy": 5.5712090015411375,
      "epoch": 1.5111473016229573,
      "grad_norm": 1.125,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.3214,
      "mean_token_accuracy": 0.1914004534482956,
      "num_tokens": 30016343.0,
      "step": 13455
    },
    {
      "entropy": 5.456988000869751,
      "epoch": 1.5117088785309147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.2216,
      "mean_token_accuracy": 0.20098562389612198,
      "num_tokens": 30027719.0,
      "step": 13460
    },
    {
      "entropy": 5.44575252532959,
      "epoch": 1.5122704554388724,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19920252710580827,
      "num_tokens": 30037988.0,
      "step": 13465
    },
    {
      "entropy": 5.4822839260101315,
      "epoch": 1.51283203234683,
      "grad_norm": 1.125,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19513847082853317,
      "num_tokens": 30049736.0,
      "step": 13470
    },
    {
      "entropy": 5.507877826690674,
      "epoch": 1.5133936092547875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.3559,
      "mean_token_accuracy": 0.1882481500506401,
      "num_tokens": 30061963.0,
      "step": 13475
    },
    {
      "entropy": 5.536148500442505,
      "epoch": 1.513955186162745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19694520235061647,
      "num_tokens": 30072256.0,
      "step": 13480
    },
    {
      "entropy": 5.430942010879517,
      "epoch": 1.5145167630707026,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.3161,
      "mean_token_accuracy": 0.19644870907068251,
      "num_tokens": 30084804.0,
      "step": 13485
    },
    {
      "entropy": 5.489186716079712,
      "epoch": 1.51507833997866,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.3,
      "mean_token_accuracy": 0.1934249296784401,
      "num_tokens": 30096938.0,
      "step": 13490
    },
    {
      "entropy": 5.5410013675689695,
      "epoch": 1.5156399168866175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.2759,
      "mean_token_accuracy": 0.18755671232938767,
      "num_tokens": 30108233.0,
      "step": 13495
    },
    {
      "entropy": 5.491359090805053,
      "epoch": 1.5162014937945751,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477984505899334,
      "loss": 5.313,
      "mean_token_accuracy": 0.19041243493556975,
      "num_tokens": 30120246.0,
      "step": 13500
    },
    {
      "entropy": 5.469485569000244,
      "epoch": 1.5167630707025328,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047796718188901095,
      "loss": 5.135,
      "mean_token_accuracy": 0.20159560590982437,
      "num_tokens": 30131213.0,
      "step": 13505
    },
    {
      "entropy": 5.372662925720215,
      "epoch": 1.5173246476104902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.162,
      "mean_token_accuracy": 0.20038333684206008,
      "num_tokens": 30143300.0,
      "step": 13510
    },
    {
      "entropy": 5.432745695114136,
      "epoch": 1.5178862245184477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.2379,
      "mean_token_accuracy": 0.19180796593427657,
      "num_tokens": 30153794.0,
      "step": 13515
    },
    {
      "entropy": 5.483161783218383,
      "epoch": 1.5184478014264053,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.3581,
      "mean_token_accuracy": 0.1866893783211708,
      "num_tokens": 30165152.0,
      "step": 13520
    },
    {
      "entropy": 5.492881536483765,
      "epoch": 1.519009378334363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19126631766557695,
      "num_tokens": 30176338.0,
      "step": 13525
    },
    {
      "entropy": 5.481426525115967,
      "epoch": 1.5195709552423204,
      "grad_norm": 1.125,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.3264,
      "mean_token_accuracy": 0.18730773776769638,
      "num_tokens": 30187666.0,
      "step": 13530
    },
    {
      "entropy": 5.484125757217408,
      "epoch": 1.5201325321502779,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18887080997228622,
      "num_tokens": 30198892.0,
      "step": 13535
    },
    {
      "entropy": 5.444531393051148,
      "epoch": 1.5206941090582355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18783968985080718,
      "num_tokens": 30210677.0,
      "step": 13540
    },
    {
      "entropy": 5.513345432281494,
      "epoch": 1.5212556859661932,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.2577,
      "mean_token_accuracy": 0.2003664642572403,
      "num_tokens": 30221217.0,
      "step": 13545
    },
    {
      "entropy": 5.511929988861084,
      "epoch": 1.5218172628741506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.3866,
      "mean_token_accuracy": 0.18828341215848923,
      "num_tokens": 30233576.0,
      "step": 13550
    },
    {
      "entropy": 5.43354434967041,
      "epoch": 1.522378839782108,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.215,
      "mean_token_accuracy": 0.19677537828683853,
      "num_tokens": 30244787.0,
      "step": 13555
    },
    {
      "entropy": 5.497850465774536,
      "epoch": 1.5229404166900657,
      "grad_norm": 1.09375,
      "learning_rate": 0.000477776191368407,
      "loss": 5.3034,
      "mean_token_accuracy": 0.19393382221460342,
      "num_tokens": 30256656.0,
      "step": 13560
    },
    {
      "entropy": 5.48765115737915,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.3694,
      "mean_token_accuracy": 0.19351017475128174,
      "num_tokens": 30267715.0,
      "step": 13565
    },
    {
      "entropy": 5.522074365615845,
      "epoch": 1.5240635705059808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.3456,
      "mean_token_accuracy": 0.18663860112428665,
      "num_tokens": 30278119.0,
      "step": 13570
    },
    {
      "entropy": 5.53045244216919,
      "epoch": 1.5246251474139383,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.2563,
      "mean_token_accuracy": 0.1942450299859047,
      "num_tokens": 30289899.0,
      "step": 13575
    },
    {
      "entropy": 5.475974655151367,
      "epoch": 1.525186724321896,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.2725,
      "mean_token_accuracy": 0.19403212517499924,
      "num_tokens": 30300799.0,
      "step": 13580
    },
    {
      "entropy": 5.604724645614624,
      "epoch": 1.5257483012298534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.4902,
      "mean_token_accuracy": 0.18343800008296968,
      "num_tokens": 30312955.0,
      "step": 13585
    },
    {
      "entropy": 5.53210973739624,
      "epoch": 1.5263098781378108,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.266,
      "mean_token_accuracy": 0.1911618486046791,
      "num_tokens": 30323089.0,
      "step": 13590
    },
    {
      "entropy": 5.426593255996704,
      "epoch": 1.5268714550457685,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477654245141998,
      "loss": 5.2106,
      "mean_token_accuracy": 0.19920119941234588,
      "num_tokens": 30333926.0,
      "step": 13595
    },
    {
      "entropy": 5.377472257614135,
      "epoch": 1.5274330319537262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.3193,
      "mean_token_accuracy": 0.19035524278879165,
      "num_tokens": 30346347.0,
      "step": 13600
    },
    {
      "entropy": 5.471315193176269,
      "epoch": 1.5279946088616836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18975663930177689,
      "num_tokens": 30357757.0,
      "step": 13605
    },
    {
      "entropy": 5.624136877059937,
      "epoch": 1.528556185769641,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.3609,
      "mean_token_accuracy": 0.19210373163223265,
      "num_tokens": 30369323.0,
      "step": 13610
    },
    {
      "entropy": 5.445489358901978,
      "epoch": 1.5291177626775987,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.206,
      "mean_token_accuracy": 0.19743847995996475,
      "num_tokens": 30380335.0,
      "step": 13615
    },
    {
      "entropy": 5.415156698226928,
      "epoch": 1.5296793395855564,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19508120268583298,
      "num_tokens": 30391338.0,
      "step": 13620
    },
    {
      "entropy": 5.4708761215209964,
      "epoch": 1.5302409164935138,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477549468146883,
      "loss": 5.3326,
      "mean_token_accuracy": 0.18398072123527526,
      "num_tokens": 30402917.0,
      "step": 13625
    },
    {
      "entropy": 5.409539222717285,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004775319827387309,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1965908482670784,
      "num_tokens": 30412690.0,
      "step": 13630
    },
    {
      "entropy": 5.469643211364746,
      "epoch": 1.531364070309429,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1923662856221199,
      "num_tokens": 30423020.0,
      "step": 13635
    },
    {
      "entropy": 5.639208745956421,
      "epoch": 1.5319256472173866,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.54,
      "mean_token_accuracy": 0.1781605027616024,
      "num_tokens": 30434039.0,
      "step": 13640
    },
    {
      "entropy": 5.605772399902344,
      "epoch": 1.532487224125344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.3114,
      "mean_token_accuracy": 0.1892308309674263,
      "num_tokens": 30446743.0,
      "step": 13645
    },
    {
      "entropy": 5.502392721176148,
      "epoch": 1.5330488010333014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000477461976624724,
      "loss": 5.2872,
      "mean_token_accuracy": 0.19604190289974213,
      "num_tokens": 30456969.0,
      "step": 13650
    },
    {
      "entropy": 5.387057065963745,
      "epoch": 1.533610377941259,
      "grad_norm": 1.125,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.3414,
      "mean_token_accuracy": 0.18829695284366607,
      "num_tokens": 30467497.0,
      "step": 13655
    },
    {
      "entropy": 5.50472354888916,
      "epoch": 1.5341719548492168,
      "grad_norm": 1.171875,
      "learning_rate": 0.000477426934886684,
      "loss": 5.3023,
      "mean_token_accuracy": 0.1921074241399765,
      "num_tokens": 30479428.0,
      "step": 13660
    },
    {
      "entropy": 5.499163389205933,
      "epoch": 1.5347335317571742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19616036117076874,
      "num_tokens": 30490193.0,
      "step": 13665
    },
    {
      "entropy": 5.537275123596191,
      "epoch": 1.5352951086651316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.3104,
      "mean_token_accuracy": 0.18778176605701447,
      "num_tokens": 30500311.0,
      "step": 13670
    },
    {
      "entropy": 5.450046491622925,
      "epoch": 1.5358566855730893,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.23,
      "mean_token_accuracy": 0.19670809358358382,
      "num_tokens": 30512485.0,
      "step": 13675
    },
    {
      "entropy": 5.453216409683227,
      "epoch": 1.5364182624810467,
      "grad_norm": 1.125,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.214,
      "mean_token_accuracy": 0.194830584526062,
      "num_tokens": 30523495.0,
      "step": 13680
    },
    {
      "entropy": 5.428583288192749,
      "epoch": 1.5369798393890042,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19782847613096238,
      "num_tokens": 30533964.0,
      "step": 13685
    },
    {
      "entropy": 5.384706258773804,
      "epoch": 1.5375414162969618,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19759590327739715,
      "num_tokens": 30544798.0,
      "step": 13690
    },
    {
      "entropy": 5.478674459457397,
      "epoch": 1.5381029932049195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19161437302827836,
      "num_tokens": 30555991.0,
      "step": 13695
    },
    {
      "entropy": 5.47773928642273,
      "epoch": 1.538664570112877,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.1952,
      "mean_token_accuracy": 0.19581881463527678,
      "num_tokens": 30566848.0,
      "step": 13700
    },
    {
      "entropy": 5.470114231109619,
      "epoch": 1.5392261470208344,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.2138,
      "mean_token_accuracy": 0.2048891678452492,
      "num_tokens": 30577361.0,
      "step": 13705
    },
    {
      "entropy": 5.453024816513062,
      "epoch": 1.539787723928792,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.2808,
      "mean_token_accuracy": 0.1933975875377655,
      "num_tokens": 30587685.0,
      "step": 13710
    },
    {
      "entropy": 5.454155158996582,
      "epoch": 1.5403493008367497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.3298,
      "mean_token_accuracy": 0.19296918511390687,
      "num_tokens": 30599864.0,
      "step": 13715
    },
    {
      "entropy": 5.480147457122802,
      "epoch": 1.5409108777447071,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.32,
      "mean_token_accuracy": 0.19130200743675232,
      "num_tokens": 30610885.0,
      "step": 13720
    },
    {
      "entropy": 5.495508050918579,
      "epoch": 1.5414724546526646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.2618,
      "mean_token_accuracy": 0.1925037369132042,
      "num_tokens": 30622537.0,
      "step": 13725
    },
    {
      "entropy": 5.572053241729736,
      "epoch": 1.5420340315606222,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.4054,
      "mean_token_accuracy": 0.18633068799972535,
      "num_tokens": 30635045.0,
      "step": 13730
    },
    {
      "entropy": 5.450768852233887,
      "epoch": 1.54259560846858,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.2344,
      "mean_token_accuracy": 0.18950365036725997,
      "num_tokens": 30645998.0,
      "step": 13735
    },
    {
      "entropy": 5.490896511077881,
      "epoch": 1.5431571853765373,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.2954,
      "mean_token_accuracy": 0.18911807537078856,
      "num_tokens": 30658744.0,
      "step": 13740
    },
    {
      "entropy": 5.440768098831176,
      "epoch": 1.5437187622844948,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20175869762897491,
      "num_tokens": 30669420.0,
      "step": 13745
    },
    {
      "entropy": 5.343130588531494,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1955384060740471,
      "num_tokens": 30680037.0,
      "step": 13750
    },
    {
      "entropy": 5.526803350448608,
      "epoch": 1.54484191610041,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.2926,
      "mean_token_accuracy": 0.1943487599492073,
      "num_tokens": 30691650.0,
      "step": 13755
    },
    {
      "entropy": 5.454730749130249,
      "epoch": 1.5454034930083675,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.2969,
      "mean_token_accuracy": 0.19266333431005478,
      "num_tokens": 30701876.0,
      "step": 13760
    },
    {
      "entropy": 5.634919834136963,
      "epoch": 1.545965069916325,
      "grad_norm": 1.25,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.5084,
      "mean_token_accuracy": 0.18438412100076676,
      "num_tokens": 30714484.0,
      "step": 13765
    },
    {
      "entropy": 5.493253707885742,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477039775179498,
      "loss": 5.2796,
      "mean_token_accuracy": 0.18781240284442902,
      "num_tokens": 30725185.0,
      "step": 13770
    },
    {
      "entropy": 5.450872564315796,
      "epoch": 1.54708822373224,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.2825,
      "mean_token_accuracy": 0.19344881623983384,
      "num_tokens": 30735522.0,
      "step": 13775
    },
    {
      "entropy": 5.437634229660034,
      "epoch": 1.5476498006401975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.2977,
      "mean_token_accuracy": 0.19602542221546174,
      "num_tokens": 30746353.0,
      "step": 13780
    },
    {
      "entropy": 5.497179985046387,
      "epoch": 1.5482113775481552,
      "grad_norm": 1.140625,
      "learning_rate": 0.000476986739302201,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19434531778097153,
      "num_tokens": 30756245.0,
      "step": 13785
    },
    {
      "entropy": 5.527978515625,
      "epoch": 1.5487729544561128,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.3709,
      "mean_token_accuracy": 0.19312793463468553,
      "num_tokens": 30767387.0,
      "step": 13790
    },
    {
      "entropy": 5.480736446380615,
      "epoch": 1.5493345313640703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.3263,
      "mean_token_accuracy": 0.19317321628332138,
      "num_tokens": 30780140.0,
      "step": 13795
    },
    {
      "entropy": 5.465695142745972,
      "epoch": 1.5498961082720277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047693364554288,
      "loss": 5.221,
      "mean_token_accuracy": 0.2014308288693428,
      "num_tokens": 30791137.0,
      "step": 13800
    },
    {
      "entropy": 5.399987125396729,
      "epoch": 1.5504576851799854,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1990022599697113,
      "num_tokens": 30801708.0,
      "step": 13805
    },
    {
      "entropy": 5.493716144561768,
      "epoch": 1.551019262087943,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.3142,
      "mean_token_accuracy": 0.190645994246006,
      "num_tokens": 30813519.0,
      "step": 13810
    },
    {
      "entropy": 5.555122327804566,
      "epoch": 1.5515808389959005,
      "grad_norm": 1.125,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.3754,
      "mean_token_accuracy": 0.19263965040445327,
      "num_tokens": 30823786.0,
      "step": 13815
    },
    {
      "entropy": 5.435871315002442,
      "epoch": 1.552142415903858,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.2836,
      "mean_token_accuracy": 0.1947065681219101,
      "num_tokens": 30834381.0,
      "step": 13820
    },
    {
      "entropy": 5.494338035583496,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.2442,
      "mean_token_accuracy": 0.1950327128171921,
      "num_tokens": 30844865.0,
      "step": 13825
    },
    {
      "entropy": 5.587636661529541,
      "epoch": 1.5532655697197733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.3394,
      "mean_token_accuracy": 0.19499900937080383,
      "num_tokens": 30857088.0,
      "step": 13830
    },
    {
      "entropy": 5.475209426879883,
      "epoch": 1.5538271466277307,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.3698,
      "mean_token_accuracy": 0.18178119659423828,
      "num_tokens": 30868320.0,
      "step": 13835
    },
    {
      "entropy": 5.496590232849121,
      "epoch": 1.5543887235356881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19671133607625962,
      "num_tokens": 30878598.0,
      "step": 13840
    },
    {
      "entropy": 5.479313516616822,
      "epoch": 1.5549503004436458,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19437508583068847,
      "num_tokens": 30889254.0,
      "step": 13845
    },
    {
      "entropy": 5.476692724227905,
      "epoch": 1.5555118773516035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.3502,
      "mean_token_accuracy": 0.1947739079594612,
      "num_tokens": 30901164.0,
      "step": 13850
    },
    {
      "entropy": 5.432692766189575,
      "epoch": 1.556073454259561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19266577064990997,
      "num_tokens": 30911828.0,
      "step": 13855
    },
    {
      "entropy": 5.457504558563232,
      "epoch": 1.5566350311675183,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004767206919899014,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20487585067749023,
      "num_tokens": 30922050.0,
      "step": 13860
    },
    {
      "entropy": 5.384016752243042,
      "epoch": 1.557196608075476,
      "grad_norm": 1.09375,
      "learning_rate": 0.000476702904098382,
      "loss": 5.2087,
      "mean_token_accuracy": 0.19853302389383315,
      "num_tokens": 30932723.0,
      "step": 13865
    },
    {
      "entropy": 5.432240676879883,
      "epoch": 1.5577581849834334,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000476685109784564,
      "loss": 5.2371,
      "mean_token_accuracy": 0.19728236943483352,
      "num_tokens": 30943699.0,
      "step": 13870
    },
    {
      "entropy": 5.456026458740235,
      "epoch": 1.5583197618913909,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.2934,
      "mean_token_accuracy": 0.1911129340529442,
      "num_tokens": 30954315.0,
      "step": 13875
    },
    {
      "entropy": 5.5283287525177,
      "epoch": 1.5588813387993485,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.4669,
      "mean_token_accuracy": 0.18846064060926437,
      "num_tokens": 30965840.0,
      "step": 13880
    },
    {
      "entropy": 5.599387454986572,
      "epoch": 1.5594429157073062,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.3613,
      "mean_token_accuracy": 0.19338658899068834,
      "num_tokens": 30977051.0,
      "step": 13885
    },
    {
      "entropy": 5.436714315414429,
      "epoch": 1.5600044926152636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.2007,
      "mean_token_accuracy": 0.19587831646203996,
      "num_tokens": 30988730.0,
      "step": 13890
    },
    {
      "entropy": 5.416201066970825,
      "epoch": 1.560566069523221,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.2325,
      "mean_token_accuracy": 0.19640145897865297,
      "num_tokens": 31000571.0,
      "step": 13895
    },
    {
      "entropy": 5.421653079986572,
      "epoch": 1.5611276464311787,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.2311,
      "mean_token_accuracy": 0.2024218738079071,
      "num_tokens": 31011003.0,
      "step": 13900
    },
    {
      "entropy": 5.508617877960205,
      "epoch": 1.5616892233391364,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.259,
      "mean_token_accuracy": 0.19444389045238494,
      "num_tokens": 31022510.0,
      "step": 13905
    },
    {
      "entropy": 5.512660360336303,
      "epoch": 1.5622508002470938,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.3107,
      "mean_token_accuracy": 0.1909449055790901,
      "num_tokens": 31033149.0,
      "step": 13910
    },
    {
      "entropy": 5.562669801712036,
      "epoch": 1.5628123771550513,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.3444,
      "mean_token_accuracy": 0.19989275485277175,
      "num_tokens": 31043626.0,
      "step": 13915
    },
    {
      "entropy": 5.5153059482574465,
      "epoch": 1.563373954063009,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.3888,
      "mean_token_accuracy": 0.19250257909297944,
      "num_tokens": 31055083.0,
      "step": 13920
    },
    {
      "entropy": 5.455924987792969,
      "epoch": 1.5639355309709666,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19278991222381592,
      "num_tokens": 31066763.0,
      "step": 13925
    },
    {
      "entropy": 5.484093093872071,
      "epoch": 1.564497107878924,
      "grad_norm": 1.125,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.2875,
      "mean_token_accuracy": 0.1909972682595253,
      "num_tokens": 31078722.0,
      "step": 13930
    },
    {
      "entropy": 5.493951749801636,
      "epoch": 1.5650586847868815,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.3446,
      "mean_token_accuracy": 0.19772130846977234,
      "num_tokens": 31090207.0,
      "step": 13935
    },
    {
      "entropy": 5.5285381317138675,
      "epoch": 1.5656202616948391,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.3663,
      "mean_token_accuracy": 0.19416908472776412,
      "num_tokens": 31101943.0,
      "step": 13940
    },
    {
      "entropy": 5.508974838256836,
      "epoch": 1.5661818386027968,
      "grad_norm": 1.125,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.3699,
      "mean_token_accuracy": 0.18728849440813064,
      "num_tokens": 31115032.0,
      "step": 13945
    },
    {
      "entropy": 5.56921796798706,
      "epoch": 1.5667434155107542,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.2785,
      "mean_token_accuracy": 0.19570743590593337,
      "num_tokens": 31124965.0,
      "step": 13950
    },
    {
      "entropy": 5.532122325897217,
      "epoch": 1.5673049924187117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19396822452545165,
      "num_tokens": 31136692.0,
      "step": 13955
    },
    {
      "entropy": 5.481381416320801,
      "epoch": 1.5678665693266693,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.293,
      "mean_token_accuracy": 0.18791648298501967,
      "num_tokens": 31148352.0,
      "step": 13960
    },
    {
      "entropy": 5.4693190574646,
      "epoch": 1.5684281462346268,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000476345798340041,
      "loss": 5.2338,
      "mean_token_accuracy": 0.19450350254774093,
      "num_tokens": 31159892.0,
      "step": 13965
    },
    {
      "entropy": 5.555439901351929,
      "epoch": 1.5689897231425842,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.4015,
      "mean_token_accuracy": 0.18517508506774902,
      "num_tokens": 31170087.0,
      "step": 13970
    },
    {
      "entropy": 5.374926376342773,
      "epoch": 1.5695513000505419,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19553141742944719,
      "num_tokens": 31180687.0,
      "step": 13975
    },
    {
      "entropy": 5.544255304336548,
      "epoch": 1.5701128769584995,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1890310525894165,
      "num_tokens": 31192513.0,
      "step": 13980
    },
    {
      "entropy": 5.459659147262573,
      "epoch": 1.570674453866457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19225750267505645,
      "num_tokens": 31203432.0,
      "step": 13985
    },
    {
      "entropy": 5.568687295913696,
      "epoch": 1.5712360307744144,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1983028993010521,
      "num_tokens": 31214474.0,
      "step": 13990
    },
    {
      "entropy": 5.50894193649292,
      "epoch": 1.571797607682372,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1880207508802414,
      "num_tokens": 31226489.0,
      "step": 13995
    },
    {
      "entropy": 5.356471872329712,
      "epoch": 1.5723591845903298,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19300511181354524,
      "num_tokens": 31237568.0,
      "step": 14000
    },
    {
      "entropy": 5.421355724334717,
      "epoch": 1.5729207614982872,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.2111,
      "mean_token_accuracy": 0.20428184866905214,
      "num_tokens": 31248587.0,
      "step": 14005
    },
    {
      "entropy": 5.501969671249389,
      "epoch": 1.5734823384062446,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.3303,
      "mean_token_accuracy": 0.18805623501539231,
      "num_tokens": 31259513.0,
      "step": 14010
    },
    {
      "entropy": 5.521801328659057,
      "epoch": 1.5740439153142023,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.3141,
      "mean_token_accuracy": 0.18917859792709352,
      "num_tokens": 31270984.0,
      "step": 14015
    },
    {
      "entropy": 5.44963321685791,
      "epoch": 1.57460549222216,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19760411977767944,
      "num_tokens": 31281907.0,
      "step": 14020
    },
    {
      "entropy": 5.3613365650177,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19272556900978088,
      "num_tokens": 31292903.0,
      "step": 14025
    },
    {
      "entropy": 5.535918760299682,
      "epoch": 1.5757286460380748,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.312,
      "mean_token_accuracy": 0.19466183632612227,
      "num_tokens": 31304185.0,
      "step": 14030
    },
    {
      "entropy": 5.525847482681274,
      "epoch": 1.5762902229460325,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1920395791530609,
      "num_tokens": 31314806.0,
      "step": 14035
    },
    {
      "entropy": 5.479031562805176,
      "epoch": 1.5768517998539902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.3115,
      "mean_token_accuracy": 0.1921336606144905,
      "num_tokens": 31326361.0,
      "step": 14040
    },
    {
      "entropy": 5.527260398864746,
      "epoch": 1.5774133767619476,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1968590497970581,
      "num_tokens": 31336549.0,
      "step": 14045
    },
    {
      "entropy": 5.565581274032593,
      "epoch": 1.577974953669905,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.3934,
      "mean_token_accuracy": 0.1887447103857994,
      "num_tokens": 31346994.0,
      "step": 14050
    },
    {
      "entropy": 5.513060712814331,
      "epoch": 1.5785365305778627,
      "grad_norm": 1.125,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19068933874368668,
      "num_tokens": 31358449.0,
      "step": 14055
    },
    {
      "entropy": 5.538665771484375,
      "epoch": 1.5790981074858201,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1894978925585747,
      "num_tokens": 31368964.0,
      "step": 14060
    },
    {
      "entropy": 5.506728410720825,
      "epoch": 1.5796596843937776,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.269,
      "mean_token_accuracy": 0.19515904784202576,
      "num_tokens": 31379388.0,
      "step": 14065
    },
    {
      "entropy": 5.44234070777893,
      "epoch": 1.5802212613017352,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19756012856960298,
      "num_tokens": 31390757.0,
      "step": 14070
    },
    {
      "entropy": 5.432030057907104,
      "epoch": 1.580782838209693,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1921206921339035,
      "num_tokens": 31402309.0,
      "step": 14075
    },
    {
      "entropy": 5.526041746139526,
      "epoch": 1.5813444151176503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.289,
      "mean_token_accuracy": 0.1956588089466095,
      "num_tokens": 31412559.0,
      "step": 14080
    },
    {
      "entropy": 5.511046600341797,
      "epoch": 1.5819059920256078,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.4156,
      "mean_token_accuracy": 0.18994818180799483,
      "num_tokens": 31424982.0,
      "step": 14085
    },
    {
      "entropy": 5.478686332702637,
      "epoch": 1.5824675689335654,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.1467,
      "mean_token_accuracy": 0.1976684331893921,
      "num_tokens": 31434472.0,
      "step": 14090
    },
    {
      "entropy": 5.516457509994507,
      "epoch": 1.583029145841523,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.3694,
      "mean_token_accuracy": 0.19179917573928834,
      "num_tokens": 31445458.0,
      "step": 14095
    },
    {
      "entropy": 5.471166610717773,
      "epoch": 1.5835907227494805,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.2447,
      "mean_token_accuracy": 0.20258166491985322,
      "num_tokens": 31455851.0,
      "step": 14100
    },
    {
      "entropy": 5.371727085113525,
      "epoch": 1.584152299657438,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.209,
      "mean_token_accuracy": 0.19591001570224761,
      "num_tokens": 31466279.0,
      "step": 14105
    },
    {
      "entropy": 5.527924585342407,
      "epoch": 1.5847138765653956,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.2871,
      "mean_token_accuracy": 0.19851753860712051,
      "num_tokens": 31477933.0,
      "step": 14110
    },
    {
      "entropy": 5.533932399749756,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1907680407166481,
      "num_tokens": 31487640.0,
      "step": 14115
    },
    {
      "entropy": 5.529735708236695,
      "epoch": 1.5858370303813107,
      "grad_norm": 1.25,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19871703535318375,
      "num_tokens": 31499059.0,
      "step": 14120
    },
    {
      "entropy": 5.496701622009278,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19345563650131226,
      "num_tokens": 31510264.0,
      "step": 14125
    },
    {
      "entropy": 5.493211460113526,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.2804,
      "mean_token_accuracy": 0.19585746228694917,
      "num_tokens": 31520873.0,
      "step": 14130
    },
    {
      "entropy": 5.481922245025634,
      "epoch": 1.5875217611051835,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19228129982948303,
      "num_tokens": 31533836.0,
      "step": 14135
    },
    {
      "entropy": 5.43515305519104,
      "epoch": 1.588083338013141,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.301,
      "mean_token_accuracy": 0.19055052399635314,
      "num_tokens": 31545091.0,
      "step": 14140
    },
    {
      "entropy": 5.472206687927246,
      "epoch": 1.5886449149210984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.208,
      "mean_token_accuracy": 0.1969258964061737,
      "num_tokens": 31557139.0,
      "step": 14145
    },
    {
      "entropy": 5.504077005386352,
      "epoch": 1.589206491829056,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.3078,
      "mean_token_accuracy": 0.1930576056241989,
      "num_tokens": 31567928.0,
      "step": 14150
    },
    {
      "entropy": 5.419829034805298,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.3135,
      "mean_token_accuracy": 0.19434235394001007,
      "num_tokens": 31578081.0,
      "step": 14155
    },
    {
      "entropy": 5.4468381881713865,
      "epoch": 1.590329645644971,
      "grad_norm": 1.125,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.226,
      "mean_token_accuracy": 0.19410833418369294,
      "num_tokens": 31589648.0,
      "step": 14160
    },
    {
      "entropy": 5.523640060424805,
      "epoch": 1.5908912225529286,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1986084297299385,
      "num_tokens": 31601085.0,
      "step": 14165
    },
    {
      "entropy": 5.5377849578857425,
      "epoch": 1.5914527994608862,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.3494,
      "mean_token_accuracy": 0.19505300372838974,
      "num_tokens": 31612005.0,
      "step": 14170
    },
    {
      "entropy": 5.507797145843506,
      "epoch": 1.5920143763688437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.319,
      "mean_token_accuracy": 0.19059222042560578,
      "num_tokens": 31623427.0,
      "step": 14175
    },
    {
      "entropy": 5.606068420410156,
      "epoch": 1.5925759532768011,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.4492,
      "mean_token_accuracy": 0.18563432544469832,
      "num_tokens": 31635086.0,
      "step": 14180
    },
    {
      "entropy": 5.5014121532440186,
      "epoch": 1.5931375301847588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.2893,
      "mean_token_accuracy": 0.19231295138597487,
      "num_tokens": 31646707.0,
      "step": 14185
    },
    {
      "entropy": 5.371788024902344,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.1594,
      "mean_token_accuracy": 0.19806861728429795,
      "num_tokens": 31657327.0,
      "step": 14190
    },
    {
      "entropy": 5.447383260726928,
      "epoch": 1.5942606840006739,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.1941,
      "mean_token_accuracy": 0.194957172870636,
      "num_tokens": 31668075.0,
      "step": 14195
    },
    {
      "entropy": 5.505075407028198,
      "epoch": 1.5948222609086313,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.3298,
      "mean_token_accuracy": 0.18746793121099473,
      "num_tokens": 31680270.0,
      "step": 14200
    },
    {
      "entropy": 5.453749942779541,
      "epoch": 1.595383837816589,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20338918566703795,
      "num_tokens": 31691271.0,
      "step": 14205
    },
    {
      "entropy": 5.3994364738464355,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.1951,
      "mean_token_accuracy": 0.19713371396064758,
      "num_tokens": 31702982.0,
      "step": 14210
    },
    {
      "entropy": 5.453500843048095,
      "epoch": 1.596506991632504,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19095061868429183,
      "num_tokens": 31715129.0,
      "step": 14215
    },
    {
      "entropy": 5.560067462921142,
      "epoch": 1.5970685685404615,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.3785,
      "mean_token_accuracy": 0.19212473332881927,
      "num_tokens": 31726392.0,
      "step": 14220
    },
    {
      "entropy": 5.587955665588379,
      "epoch": 1.5976301454484192,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.3388,
      "mean_token_accuracy": 0.19493759274482728,
      "num_tokens": 31738143.0,
      "step": 14225
    },
    {
      "entropy": 5.557267475128174,
      "epoch": 1.5981917223563769,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.3105,
      "mean_token_accuracy": 0.18792663365602494,
      "num_tokens": 31748510.0,
      "step": 14230
    },
    {
      "entropy": 5.346973037719726,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047536881675345,
      "loss": 5.1956,
      "mean_token_accuracy": 0.19649087637662888,
      "num_tokens": 31759378.0,
      "step": 14235
    },
    {
      "entropy": 5.52656397819519,
      "epoch": 1.5993148761722917,
      "grad_norm": 1.125,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18594098836183548,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.472970390319825,
      "epoch": 1.5998764530802494,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.2136,
      "mean_token_accuracy": 0.20129866302013397,
      "num_tokens": 31781990.0,
      "step": 14245
    },
    {
      "entropy": 5.496505451202393,
      "epoch": 1.6004380299882068,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.2773,
      "mean_token_accuracy": 0.19124405235052108,
      "num_tokens": 31792144.0,
      "step": 14250
    },
    {
      "entropy": 5.490231895446778,
      "epoch": 1.6009996068961643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19637819677591323,
      "num_tokens": 31804022.0,
      "step": 14255
    },
    {
      "entropy": 5.553983306884765,
      "epoch": 1.601561183804122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000475277413077462,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1905074432492256,
      "num_tokens": 31815108.0,
      "step": 14260
    },
    {
      "entropy": 5.424648475646973,
      "epoch": 1.6021227607120796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19563350677490235,
      "num_tokens": 31825727.0,
      "step": 14265
    },
    {
      "entropy": 5.419054317474365,
      "epoch": 1.602684337620037,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.2736,
      "mean_token_accuracy": 0.20331913679838182,
      "num_tokens": 31836168.0,
      "step": 14270
    },
    {
      "entropy": 5.577020692825317,
      "epoch": 1.6032459145279945,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.3639,
      "mean_token_accuracy": 0.19066408425569534,
      "num_tokens": 31847383.0,
      "step": 14275
    },
    {
      "entropy": 5.492323684692383,
      "epoch": 1.6038074914359521,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.2019,
      "mean_token_accuracy": 0.2013877347111702,
      "num_tokens": 31858679.0,
      "step": 14280
    },
    {
      "entropy": 5.437363386154175,
      "epoch": 1.6043690683439098,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19410343170166017,
      "num_tokens": 31870195.0,
      "step": 14285
    },
    {
      "entropy": 5.523854541778564,
      "epoch": 1.6049306452518672,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19254668056964874,
      "num_tokens": 31881175.0,
      "step": 14290
    },
    {
      "entropy": 5.435376453399658,
      "epoch": 1.6054922221598247,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.2271,
      "mean_token_accuracy": 0.20184947550296783,
      "num_tokens": 31891199.0,
      "step": 14295
    },
    {
      "entropy": 5.619966793060303,
      "epoch": 1.6060537990677823,
      "grad_norm": 1.125,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.4093,
      "mean_token_accuracy": 0.1852930799126625,
      "num_tokens": 31903379.0,
      "step": 14300
    },
    {
      "entropy": 5.543605899810791,
      "epoch": 1.60661537597574,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.2089,
      "mean_token_accuracy": 0.19692232012748717,
      "num_tokens": 31913958.0,
      "step": 14305
    },
    {
      "entropy": 5.388770532608032,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.1874,
      "mean_token_accuracy": 0.1972719430923462,
      "num_tokens": 31924547.0,
      "step": 14310
    },
    {
      "entropy": 5.561536169052124,
      "epoch": 1.6077385297916549,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1887125328183174,
      "num_tokens": 31936987.0,
      "step": 14315
    },
    {
      "entropy": 5.4675726890563965,
      "epoch": 1.6083001066996125,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19213483482599258,
      "num_tokens": 31948165.0,
      "step": 14320
    },
    {
      "entropy": 5.514142751693726,
      "epoch": 1.6088616836075702,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.3111,
      "mean_token_accuracy": 0.19153206348419188,
      "num_tokens": 31959206.0,
      "step": 14325
    },
    {
      "entropy": 5.5301055908203125,
      "epoch": 1.6094232605155276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.294,
      "mean_token_accuracy": 0.19198443591594697,
      "num_tokens": 31970804.0,
      "step": 14330
    },
    {
      "entropy": 5.441630125045776,
      "epoch": 1.609984837423485,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19461580365896225,
      "num_tokens": 31982961.0,
      "step": 14335
    },
    {
      "entropy": 5.503299951553345,
      "epoch": 1.6105464143314427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1995654433965683,
      "num_tokens": 31993812.0,
      "step": 14340
    },
    {
      "entropy": 5.431442356109619,
      "epoch": 1.6111079912394002,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.3128,
      "mean_token_accuracy": 0.19547715336084365,
      "num_tokens": 32004688.0,
      "step": 14345
    },
    {
      "entropy": 5.433307743072509,
      "epoch": 1.6116695681473576,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19612354934215545,
      "num_tokens": 32014653.0,
      "step": 14350
    },
    {
      "entropy": 5.449719381332398,
      "epoch": 1.6122311450553153,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.2579,
      "mean_token_accuracy": 0.19029328227043152,
      "num_tokens": 32025638.0,
      "step": 14355
    },
    {
      "entropy": 5.34260835647583,
      "epoch": 1.612792721963273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.1635,
      "mean_token_accuracy": 0.19991593956947326,
      "num_tokens": 32036316.0,
      "step": 14360
    },
    {
      "entropy": 5.438609409332275,
      "epoch": 1.6133542988712304,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.1995,
      "mean_token_accuracy": 0.19836467802524566,
      "num_tokens": 32047355.0,
      "step": 14365
    },
    {
      "entropy": 5.4777916431427,
      "epoch": 1.6139158757791878,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.3065,
      "mean_token_accuracy": 0.18453036099672318,
      "num_tokens": 32058897.0,
      "step": 14370
    },
    {
      "entropy": 5.487823629379273,
      "epoch": 1.6144774526871455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.3024,
      "mean_token_accuracy": 0.19205862283706665,
      "num_tokens": 32069932.0,
      "step": 14375
    },
    {
      "entropy": 5.49405083656311,
      "epoch": 1.6150390295951031,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1968003213405609,
      "num_tokens": 32081359.0,
      "step": 14380
    },
    {
      "entropy": 5.389449834823608,
      "epoch": 1.6156006065030606,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.1819,
      "mean_token_accuracy": 0.20150613486766816,
      "num_tokens": 32093050.0,
      "step": 14385
    },
    {
      "entropy": 5.4959498882293705,
      "epoch": 1.616162183411018,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.3132,
      "mean_token_accuracy": 0.19254093915224074,
      "num_tokens": 32104623.0,
      "step": 14390
    },
    {
      "entropy": 5.464283275604248,
      "epoch": 1.6167237603189757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.283,
      "mean_token_accuracy": 0.18424515426158905,
      "num_tokens": 32115845.0,
      "step": 14395
    },
    {
      "entropy": 5.477752304077148,
      "epoch": 1.6172853372269334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19243161231279374,
      "num_tokens": 32127066.0,
      "step": 14400
    },
    {
      "entropy": 5.46799669265747,
      "epoch": 1.6178469141348908,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.3624,
      "mean_token_accuracy": 0.1913568213582039,
      "num_tokens": 32138344.0,
      "step": 14405
    },
    {
      "entropy": 5.5497472286224365,
      "epoch": 1.6184084910428482,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.3297,
      "mean_token_accuracy": 0.18632324635982514,
      "num_tokens": 32149054.0,
      "step": 14410
    },
    {
      "entropy": 5.494712781906128,
      "epoch": 1.618970067950806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.2114,
      "mean_token_accuracy": 0.18590016961097716,
      "num_tokens": 32160385.0,
      "step": 14415
    },
    {
      "entropy": 5.524468421936035,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.3714,
      "mean_token_accuracy": 0.18828795105218887,
      "num_tokens": 32171856.0,
      "step": 14420
    },
    {
      "entropy": 5.449940872192383,
      "epoch": 1.620093221766721,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19384409189224244,
      "num_tokens": 32182690.0,
      "step": 14425
    },
    {
      "entropy": 5.452857494354248,
      "epoch": 1.6206547986746784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.2593,
      "mean_token_accuracy": 0.1913243591785431,
      "num_tokens": 32194045.0,
      "step": 14430
    },
    {
      "entropy": 5.511925172805786,
      "epoch": 1.621216375582636,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18562723100185394,
      "num_tokens": 32205699.0,
      "step": 14435
    },
    {
      "entropy": 5.581693935394287,
      "epoch": 1.6217779524905935,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.3264,
      "mean_token_accuracy": 0.18597105294466018,
      "num_tokens": 32217974.0,
      "step": 14440
    },
    {
      "entropy": 5.365362930297851,
      "epoch": 1.622339529398551,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.213,
      "mean_token_accuracy": 0.19434062540531158,
      "num_tokens": 32228723.0,
      "step": 14445
    },
    {
      "entropy": 5.451690435409546,
      "epoch": 1.6229011063065086,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1877187356352806,
      "num_tokens": 32241048.0,
      "step": 14450
    },
    {
      "entropy": 5.4257649898529055,
      "epoch": 1.6234626832144663,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.2322,
      "mean_token_accuracy": 0.19688658267259598,
      "num_tokens": 32251255.0,
      "step": 14455
    },
    {
      "entropy": 5.471079158782959,
      "epoch": 1.6240242601224237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.276,
      "mean_token_accuracy": 0.18840143382549285,
      "num_tokens": 32262954.0,
      "step": 14460
    },
    {
      "entropy": 5.49017071723938,
      "epoch": 1.6245858370303812,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.2787,
      "mean_token_accuracy": 0.1973302334547043,
      "num_tokens": 32275445.0,
      "step": 14465
    },
    {
      "entropy": 5.48036036491394,
      "epoch": 1.6251474139383388,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.2261,
      "mean_token_accuracy": 0.20021368861198424,
      "num_tokens": 32285976.0,
      "step": 14470
    },
    {
      "entropy": 5.419240140914917,
      "epoch": 1.6257089908462965,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.2253,
      "mean_token_accuracy": 0.19624136984348298,
      "num_tokens": 32296304.0,
      "step": 14475
    },
    {
      "entropy": 5.466180610656738,
      "epoch": 1.626270567754254,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.2573,
      "mean_token_accuracy": 0.202895824611187,
      "num_tokens": 32307085.0,
      "step": 14480
    },
    {
      "entropy": 5.437974071502685,
      "epoch": 1.6268321446622114,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.1813,
      "mean_token_accuracy": 0.2056938111782074,
      "num_tokens": 32316814.0,
      "step": 14485
    },
    {
      "entropy": 5.413638639450073,
      "epoch": 1.627393721570169,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.2688,
      "mean_token_accuracy": 0.19101787954568863,
      "num_tokens": 32327820.0,
      "step": 14490
    },
    {
      "entropy": 5.552432203292847,
      "epoch": 1.6279552984781267,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.2797,
      "mean_token_accuracy": 0.19531487971544265,
      "num_tokens": 32338185.0,
      "step": 14495
    },
    {
      "entropy": 5.474474716186523,
      "epoch": 1.6285168753860841,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.1826,
      "mean_token_accuracy": 0.19765564799308777,
      "num_tokens": 32348537.0,
      "step": 14500
    },
    {
      "entropy": 5.459687805175781,
      "epoch": 1.6290784522940416,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.312,
      "mean_token_accuracy": 0.19006200581789018,
      "num_tokens": 32359094.0,
      "step": 14505
    },
    {
      "entropy": 5.324436140060425,
      "epoch": 1.6296400292019992,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047435462013406554,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20569638013839722,
      "num_tokens": 32370811.0,
      "step": 14510
    },
    {
      "entropy": 5.520537900924682,
      "epoch": 1.630201606109957,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.343,
      "mean_token_accuracy": 0.19003819078207015,
      "num_tokens": 32381522.0,
      "step": 14515
    },
    {
      "entropy": 5.472776699066162,
      "epoch": 1.6307631830179143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.305,
      "mean_token_accuracy": 0.19053868055343628,
      "num_tokens": 32393221.0,
      "step": 14520
    },
    {
      "entropy": 5.555664110183716,
      "epoch": 1.6313247599258718,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.2553,
      "mean_token_accuracy": 0.186443330347538,
      "num_tokens": 32404007.0,
      "step": 14525
    },
    {
      "entropy": 5.52168836593628,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.3659,
      "mean_token_accuracy": 0.18848371654748916,
      "num_tokens": 32414318.0,
      "step": 14530
    },
    {
      "entropy": 5.453628826141357,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.2959,
      "mean_token_accuracy": 0.1955850303173065,
      "num_tokens": 32425493.0,
      "step": 14535
    },
    {
      "entropy": 5.4751489639282225,
      "epoch": 1.6330094906497443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.2545,
      "mean_token_accuracy": 0.19708335846662522,
      "num_tokens": 32435765.0,
      "step": 14540
    },
    {
      "entropy": 5.570137643814087,
      "epoch": 1.633571067557702,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.322,
      "mean_token_accuracy": 0.19859974682331086,
      "num_tokens": 32446478.0,
      "step": 14545
    },
    {
      "entropy": 5.433111381530762,
      "epoch": 1.6341326444656596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.3233,
      "mean_token_accuracy": 0.19823819696903228,
      "num_tokens": 32456866.0,
      "step": 14550
    },
    {
      "entropy": 5.507346677780151,
      "epoch": 1.634694221373617,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474186829994371,
      "loss": 5.2098,
      "mean_token_accuracy": 0.20223504304885864,
      "num_tokens": 32466973.0,
      "step": 14555
    },
    {
      "entropy": 5.591839790344238,
      "epoch": 1.6352557982815745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18848806172609328,
      "num_tokens": 32478118.0,
      "step": 14560
    },
    {
      "entropy": 5.4185503959655765,
      "epoch": 1.6358173751895322,
      "grad_norm": 1.125,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.191,
      "mean_token_accuracy": 0.2056002840399742,
      "num_tokens": 32488255.0,
      "step": 14565
    },
    {
      "entropy": 5.507327938079834,
      "epoch": 1.6363789520974898,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.2908,
      "mean_token_accuracy": 0.19692969173192978,
      "num_tokens": 32498685.0,
      "step": 14570
    },
    {
      "entropy": 5.497486019134522,
      "epoch": 1.6369405290054473,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.3004,
      "mean_token_accuracy": 0.19256396889686583,
      "num_tokens": 32510457.0,
      "step": 14575
    },
    {
      "entropy": 5.448562812805176,
      "epoch": 1.6375021059134047,
      "grad_norm": 1.125,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.2572,
      "mean_token_accuracy": 0.19564854800701142,
      "num_tokens": 32522495.0,
      "step": 14580
    },
    {
      "entropy": 5.607509517669678,
      "epoch": 1.6380636828213624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.3621,
      "mean_token_accuracy": 0.19127316027879715,
      "num_tokens": 32534891.0,
      "step": 14585
    },
    {
      "entropy": 5.461187124252319,
      "epoch": 1.63862525972932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.2229,
      "mean_token_accuracy": 0.19717327356338502,
      "num_tokens": 32545555.0,
      "step": 14590
    },
    {
      "entropy": 5.40444712638855,
      "epoch": 1.6391868366372775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.2846,
      "mean_token_accuracy": 0.19460931420326233,
      "num_tokens": 32556851.0,
      "step": 14595
    },
    {
      "entropy": 5.604703187942505,
      "epoch": 1.639748413545235,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.4212,
      "mean_token_accuracy": 0.1845278963446617,
      "num_tokens": 32569028.0,
      "step": 14600
    },
    {
      "entropy": 5.670565891265869,
      "epoch": 1.6403099904531926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19195715039968492,
      "num_tokens": 32581510.0,
      "step": 14605
    },
    {
      "entropy": 5.455938816070557,
      "epoch": 1.6408715673611503,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.2115,
      "mean_token_accuracy": 0.1932632029056549,
      "num_tokens": 32592914.0,
      "step": 14610
    },
    {
      "entropy": 5.373195362091065,
      "epoch": 1.6414331442691077,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.1598,
      "mean_token_accuracy": 0.20752947479486467,
      "num_tokens": 32603931.0,
      "step": 14615
    },
    {
      "entropy": 5.543744945526123,
      "epoch": 1.6419947211770651,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.3104,
      "mean_token_accuracy": 0.1974663883447647,
      "num_tokens": 32614508.0,
      "step": 14620
    },
    {
      "entropy": 5.410279273986816,
      "epoch": 1.6425562980850228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.164,
      "mean_token_accuracy": 0.20792541056871414,
      "num_tokens": 32625393.0,
      "step": 14625
    },
    {
      "entropy": 5.3330240726470945,
      "epoch": 1.6431178749929805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20206725895404815,
      "num_tokens": 32636645.0,
      "step": 14630
    },
    {
      "entropy": 5.584393167495728,
      "epoch": 1.6436794519009377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.4626,
      "mean_token_accuracy": 0.18707671165466308,
      "num_tokens": 32648792.0,
      "step": 14635
    },
    {
      "entropy": 5.577219247817993,
      "epoch": 1.6442410288088953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.26,
      "mean_token_accuracy": 0.1996319115161896,
      "num_tokens": 32658825.0,
      "step": 14640
    },
    {
      "entropy": 5.532627058029175,
      "epoch": 1.644802605716853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.2861,
      "mean_token_accuracy": 0.19765844494104384,
      "num_tokens": 32670141.0,
      "step": 14645
    },
    {
      "entropy": 5.28882155418396,
      "epoch": 1.6453641826248104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004738309198730219,
      "loss": 5.1366,
      "mean_token_accuracy": 0.20056047588586806,
      "num_tokens": 32681804.0,
      "step": 14650
    },
    {
      "entropy": 5.587597799301148,
      "epoch": 1.6459257595327679,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.3872,
      "mean_token_accuracy": 0.1927493318915367,
      "num_tokens": 32692550.0,
      "step": 14655
    },
    {
      "entropy": 5.517039251327515,
      "epoch": 1.6464873364407255,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.26,
      "mean_token_accuracy": 0.19721782952547073,
      "num_tokens": 32703833.0,
      "step": 14660
    },
    {
      "entropy": 5.41059627532959,
      "epoch": 1.6470489133486832,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.3046,
      "mean_token_accuracy": 0.1915615662932396,
      "num_tokens": 32713928.0,
      "step": 14665
    },
    {
      "entropy": 5.507371091842652,
      "epoch": 1.6476104902566406,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.313,
      "mean_token_accuracy": 0.19521367847919463,
      "num_tokens": 32725393.0,
      "step": 14670
    },
    {
      "entropy": 5.570739507675171,
      "epoch": 1.648172067164598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.2991,
      "mean_token_accuracy": 0.19338775128126146,
      "num_tokens": 32736047.0,
      "step": 14675
    },
    {
      "entropy": 5.498094844818115,
      "epoch": 1.6487336440725557,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.3334,
      "mean_token_accuracy": 0.19092479348182678,
      "num_tokens": 32747960.0,
      "step": 14680
    },
    {
      "entropy": 5.436005353927612,
      "epoch": 1.6492952209805134,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1973534882068634,
      "num_tokens": 32757851.0,
      "step": 14685
    },
    {
      "entropy": 5.494208478927613,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.2718,
      "mean_token_accuracy": 0.19450787007808684,
      "num_tokens": 32767317.0,
      "step": 14690
    },
    {
      "entropy": 5.565016841888427,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.2928,
      "mean_token_accuracy": 0.19438280314207076,
      "num_tokens": 32777729.0,
      "step": 14695
    },
    {
      "entropy": 5.480408334732056,
      "epoch": 1.650979951704386,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.35,
      "mean_token_accuracy": 0.1826064944267273,
      "num_tokens": 32789547.0,
      "step": 14700
    },
    {
      "entropy": 5.497604322433472,
      "epoch": 1.6515415286123436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19785625338554383,
      "num_tokens": 32801487.0,
      "step": 14705
    },
    {
      "entropy": 5.544666910171509,
      "epoch": 1.652103105520301,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.315,
      "mean_token_accuracy": 0.19704079180955886,
      "num_tokens": 32813042.0,
      "step": 14710
    },
    {
      "entropy": 5.505541372299194,
      "epoch": 1.6526646824282585,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19589462578296662,
      "num_tokens": 32825467.0,
      "step": 14715
    },
    {
      "entropy": 5.481626319885254,
      "epoch": 1.6532262593362161,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.2688,
      "mean_token_accuracy": 0.19094224274158478,
      "num_tokens": 32837113.0,
      "step": 14720
    },
    {
      "entropy": 5.55432825088501,
      "epoch": 1.6537878362441738,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.3703,
      "mean_token_accuracy": 0.18702881634235383,
      "num_tokens": 32849779.0,
      "step": 14725
    },
    {
      "entropy": 5.445461559295654,
      "epoch": 1.654349413152131,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19307452589273452,
      "num_tokens": 32861626.0,
      "step": 14730
    },
    {
      "entropy": 5.520581960678101,
      "epoch": 1.6549109900600887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.324,
      "mean_token_accuracy": 0.1982700139284134,
      "num_tokens": 32872875.0,
      "step": 14735
    },
    {
      "entropy": 5.46758804321289,
      "epoch": 1.6554725669680463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.1491,
      "mean_token_accuracy": 0.19871128648519515,
      "num_tokens": 32883899.0,
      "step": 14740
    },
    {
      "entropy": 5.366373872756958,
      "epoch": 1.6560341438760038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.156,
      "mean_token_accuracy": 0.202083320915699,
      "num_tokens": 32894255.0,
      "step": 14745
    },
    {
      "entropy": 5.431836938858032,
      "epoch": 1.6565957207839612,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.2549,
      "mean_token_accuracy": 0.1974031314253807,
      "num_tokens": 32903822.0,
      "step": 14750
    },
    {
      "entropy": 5.462358474731445,
      "epoch": 1.6571572976919189,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.2419,
      "mean_token_accuracy": 0.19894727021455766,
      "num_tokens": 32915879.0,
      "step": 14755
    },
    {
      "entropy": 5.36098313331604,
      "epoch": 1.6577188745998765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.272,
      "mean_token_accuracy": 0.19022632986307145,
      "num_tokens": 32928161.0,
      "step": 14760
    },
    {
      "entropy": 5.436191606521606,
      "epoch": 1.658280451507834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19999379813671112,
      "num_tokens": 32938399.0,
      "step": 14765
    },
    {
      "entropy": 5.447626829147339,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.1865,
      "mean_token_accuracy": 0.20350140631198882,
      "num_tokens": 32948221.0,
      "step": 14770
    },
    {
      "entropy": 5.4412473201751705,
      "epoch": 1.659403605323749,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.2567,
      "mean_token_accuracy": 0.1995250850915909,
      "num_tokens": 32959802.0,
      "step": 14775
    },
    {
      "entropy": 5.488705015182495,
      "epoch": 1.6599651822317067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.2878,
      "mean_token_accuracy": 0.19510946422815323,
      "num_tokens": 32970222.0,
      "step": 14780
    },
    {
      "entropy": 5.461812829971313,
      "epoch": 1.6605267591396642,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.2491,
      "mean_token_accuracy": 0.20073420256376268,
      "num_tokens": 32982241.0,
      "step": 14785
    },
    {
      "entropy": 5.466669034957886,
      "epoch": 1.6610883360476216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20182580798864364,
      "num_tokens": 32992584.0,
      "step": 14790
    },
    {
      "entropy": 5.477460670471191,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.4,
      "mean_token_accuracy": 0.1908192053437233,
      "num_tokens": 33004080.0,
      "step": 14795
    },
    {
      "entropy": 5.446459341049194,
      "epoch": 1.662211489863537,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.2129,
      "mean_token_accuracy": 0.197823166847229,
      "num_tokens": 33015290.0,
      "step": 14800
    },
    {
      "entropy": 5.467676401138306,
      "epoch": 1.6627730667714944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.1932,
      "mean_token_accuracy": 0.19475805908441543,
      "num_tokens": 33026470.0,
      "step": 14805
    },
    {
      "entropy": 5.406273221969604,
      "epoch": 1.6633346436794518,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.1874,
      "mean_token_accuracy": 0.20261722505092622,
      "num_tokens": 33036766.0,
      "step": 14810
    },
    {
      "entropy": 5.3666017055511475,
      "epoch": 1.6638962205874095,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1886460840702057,
      "num_tokens": 33047675.0,
      "step": 14815
    },
    {
      "entropy": 5.49187159538269,
      "epoch": 1.6644577974953672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000473188326686571,
      "loss": 5.2492,
      "mean_token_accuracy": 0.19694626778364183,
      "num_tokens": 33057613.0,
      "step": 14820
    },
    {
      "entropy": 5.5723340034484865,
      "epoch": 1.6650193744033244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.3639,
      "mean_token_accuracy": 0.18635020703077315,
      "num_tokens": 33070498.0,
      "step": 14825
    },
    {
      "entropy": 5.456169319152832,
      "epoch": 1.665580951311282,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.2259,
      "mean_token_accuracy": 0.1980530396103859,
      "num_tokens": 33082026.0,
      "step": 14830
    },
    {
      "entropy": 5.432246971130371,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19229164123535156,
      "num_tokens": 33093841.0,
      "step": 14835
    },
    {
      "entropy": 5.505250883102417,
      "epoch": 1.6667041051271971,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.2696,
      "mean_token_accuracy": 0.19567066580057144,
      "num_tokens": 33103817.0,
      "step": 14840
    },
    {
      "entropy": 5.55364875793457,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047309321185648,
      "loss": 5.3255,
      "mean_token_accuracy": 0.19429995566606523,
      "num_tokens": 33116006.0,
      "step": 14845
    },
    {
      "entropy": 5.445116281509399,
      "epoch": 1.6678272589431122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.318,
      "mean_token_accuracy": 0.18888264745473862,
      "num_tokens": 33127649.0,
      "step": 14850
    },
    {
      "entropy": 5.393762254714966,
      "epoch": 1.66838883585107,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004730551217686495,
      "loss": 5.1534,
      "mean_token_accuracy": 0.202876478433609,
      "num_tokens": 33138748.0,
      "step": 14855
    },
    {
      "entropy": 5.522401428222656,
      "epoch": 1.6689504127590273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1927672177553177,
      "num_tokens": 33150504.0,
      "step": 14860
    },
    {
      "entropy": 5.452508163452149,
      "epoch": 1.6695119896669848,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.3353,
      "mean_token_accuracy": 0.18897348940372466,
      "num_tokens": 33160625.0,
      "step": 14865
    },
    {
      "entropy": 5.478543996810913,
      "epoch": 1.6700735665749424,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19954254180192948,
      "num_tokens": 33172229.0,
      "step": 14870
    },
    {
      "entropy": 5.496869611740112,
      "epoch": 1.6706351434829,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.272,
      "mean_token_accuracy": 0.19200197011232376,
      "num_tokens": 33183234.0,
      "step": 14875
    },
    {
      "entropy": 5.501498699188232,
      "epoch": 1.6711967203908575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.312,
      "mean_token_accuracy": 0.19559772163629532,
      "num_tokens": 33193976.0,
      "step": 14880
    },
    {
      "entropy": 5.437897109985352,
      "epoch": 1.671758297298815,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.272,
      "mean_token_accuracy": 0.20058979392051696,
      "num_tokens": 33205959.0,
      "step": 14885
    },
    {
      "entropy": 5.434696531295776,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20296574831008912,
      "num_tokens": 33218405.0,
      "step": 14890
    },
    {
      "entropy": 5.426598930358887,
      "epoch": 1.6728814511147303,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.249,
      "mean_token_accuracy": 0.194725601375103,
      "num_tokens": 33230778.0,
      "step": 14895
    },
    {
      "entropy": 5.415722751617432,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.25,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.2082,
      "mean_token_accuracy": 0.19918926805257797,
      "num_tokens": 33242069.0,
      "step": 14900
    },
    {
      "entropy": 5.482007837295532,
      "epoch": 1.6740046049306452,
      "grad_norm": 1.125,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.2826,
      "mean_token_accuracy": 0.19658081084489823,
      "num_tokens": 33252942.0,
      "step": 14905
    },
    {
      "entropy": 5.578243350982666,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.3255,
      "mean_token_accuracy": 0.1905709609389305,
      "num_tokens": 33263892.0,
      "step": 14910
    },
    {
      "entropy": 5.404222297668457,
      "epoch": 1.6751277587465605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.2043,
      "mean_token_accuracy": 0.1931039199233055,
      "num_tokens": 33274383.0,
      "step": 14915
    },
    {
      "entropy": 5.502907037734985,
      "epoch": 1.6756893356545177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.2198,
      "mean_token_accuracy": 0.19201160222291946,
      "num_tokens": 33285897.0,
      "step": 14920
    },
    {
      "entropy": 5.535489702224732,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.3199,
      "mean_token_accuracy": 0.19218389838933944,
      "num_tokens": 33297192.0,
      "step": 14925
    },
    {
      "entropy": 5.4477410316467285,
      "epoch": 1.676812489470433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.2113,
      "mean_token_accuracy": 0.19755875170230866,
      "num_tokens": 33308250.0,
      "step": 14930
    },
    {
      "entropy": 5.432844352722168,
      "epoch": 1.6773740663783905,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.2215,
      "mean_token_accuracy": 0.1962057575583458,
      "num_tokens": 33318515.0,
      "step": 14935
    },
    {
      "entropy": 5.542074251174927,
      "epoch": 1.677935643286348,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.359,
      "mean_token_accuracy": 0.19027905613183976,
      "num_tokens": 33329762.0,
      "step": 14940
    },
    {
      "entropy": 5.5367666721344,
      "epoch": 1.6784972201943056,
      "grad_norm": 1.125,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1910214051604271,
      "num_tokens": 33341054.0,
      "step": 14945
    },
    {
      "entropy": 5.457042837142945,
      "epoch": 1.6790587971022632,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19975604116916656,
      "num_tokens": 33351524.0,
      "step": 14950
    },
    {
      "entropy": 5.460369443893432,
      "epoch": 1.6796203740102207,
      "grad_norm": 1.125,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.301,
      "mean_token_accuracy": 0.18565541356801987,
      "num_tokens": 33362953.0,
      "step": 14955
    },
    {
      "entropy": 5.369621610641479,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.1699,
      "mean_token_accuracy": 0.204598967730999,
      "num_tokens": 33373209.0,
      "step": 14960
    },
    {
      "entropy": 5.377118253707886,
      "epoch": 1.6807435278261358,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.1781,
      "mean_token_accuracy": 0.19663191884756087,
      "num_tokens": 33384775.0,
      "step": 14965
    },
    {
      "entropy": 5.392876672744751,
      "epoch": 1.6813051047340934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.23,
      "mean_token_accuracy": 0.19281969368457794,
      "num_tokens": 33395457.0,
      "step": 14970
    },
    {
      "entropy": 5.58880844116211,
      "epoch": 1.6818666816420509,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.3618,
      "mean_token_accuracy": 0.1950633004307747,
      "num_tokens": 33407506.0,
      "step": 14975
    },
    {
      "entropy": 5.549135446548462,
      "epoch": 1.6824282585500083,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.2532,
      "mean_token_accuracy": 0.19640884548425674,
      "num_tokens": 33418154.0,
      "step": 14980
    },
    {
      "entropy": 5.431192016601562,
      "epoch": 1.682989835457966,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.2759,
      "mean_token_accuracy": 0.19720015823841094,
      "num_tokens": 33428319.0,
      "step": 14985
    },
    {
      "entropy": 5.382882356643677,
      "epoch": 1.6835514123659236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.183,
      "mean_token_accuracy": 0.20026712864637375,
      "num_tokens": 33440429.0,
      "step": 14990
    },
    {
      "entropy": 5.638523530960083,
      "epoch": 1.684112989273881,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.4046,
      "mean_token_accuracy": 0.18178529441356658,
      "num_tokens": 33451736.0,
      "step": 14995
    },
    {
      "entropy": 5.484934949874878,
      "epoch": 1.6846745661818385,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19435687810182573,
      "num_tokens": 33461991.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.31126108629543,
      "eval_loss": 5.359016418457031,
      "eval_mean_token_accuracy": 0.1983783744200622,
      "eval_num_tokens": 33461991.0,
      "eval_runtime": 60.7711,
      "eval_samples_per_second": 510.325,
      "eval_steps_per_second": 63.797,
      "step": 15000
    },
    {
      "entropy": 5.467107629776001,
      "epoch": 1.6852361430897962,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.3171,
      "mean_token_accuracy": 0.18784952014684678,
      "num_tokens": 33473646.0,
      "step": 15005
    },
    {
      "entropy": 5.4660145282745365,
      "epoch": 1.6857977199977539,
      "grad_norm": 1.109375,
      "learning_rate": 0.000472461502817644,
      "loss": 5.2711,
      "mean_token_accuracy": 0.19591651260852813,
      "num_tokens": 33485444.0,
      "step": 15010
    },
    {
      "entropy": 5.607774829864502,
      "epoch": 1.6863592969057113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.4451,
      "mean_token_accuracy": 0.1862923875451088,
      "num_tokens": 33497199.0,
      "step": 15015
    },
    {
      "entropy": 5.530071163177491,
      "epoch": 1.6869208738136687,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.3623,
      "mean_token_accuracy": 0.18910333067178725,
      "num_tokens": 33509045.0,
      "step": 15020
    },
    {
      "entropy": 5.4576112747192385,
      "epoch": 1.6874824507216264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20573700964450836,
      "num_tokens": 33519730.0,
      "step": 15025
    },
    {
      "entropy": 5.459954690933228,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.28125,
      "learning_rate": 0.000472384466331328,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18774333000183105,
      "num_tokens": 33530393.0,
      "step": 15030
    },
    {
      "entropy": 5.452465200424195,
      "epoch": 1.6886056045375413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.3035,
      "mean_token_accuracy": 0.19598446488380433,
      "num_tokens": 33543592.0,
      "step": 15035
    },
    {
      "entropy": 5.499931716918946,
      "epoch": 1.689167181445499,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19345757067203523,
      "num_tokens": 33554736.0,
      "step": 15040
    },
    {
      "entropy": 5.4852948665618895,
      "epoch": 1.6897287583534566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20276740491390227,
      "num_tokens": 33565428.0,
      "step": 15045
    },
    {
      "entropy": 5.535889482498169,
      "epoch": 1.690290335261414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.4902,
      "mean_token_accuracy": 0.18611413836479188,
      "num_tokens": 33577102.0,
      "step": 15050
    },
    {
      "entropy": 5.5720357418060305,
      "epoch": 1.6908519121693715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.3081,
      "mean_token_accuracy": 0.19705154448747636,
      "num_tokens": 33587325.0,
      "step": 15055
    },
    {
      "entropy": 5.584289836883545,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.125,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.3471,
      "mean_token_accuracy": 0.18958540111780167,
      "num_tokens": 33599294.0,
      "step": 15060
    },
    {
      "entropy": 5.531365585327149,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.3953,
      "mean_token_accuracy": 0.18766674101352693,
      "num_tokens": 33609025.0,
      "step": 15065
    },
    {
      "entropy": 5.545634460449219,
      "epoch": 1.6925366428932442,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.3226,
      "mean_token_accuracy": 0.1924744427204132,
      "num_tokens": 33620881.0,
      "step": 15070
    },
    {
      "entropy": 5.524413299560547,
      "epoch": 1.6930982198012017,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.2707,
      "mean_token_accuracy": 0.19736731201410293,
      "num_tokens": 33631505.0,
      "step": 15075
    },
    {
      "entropy": 5.534477806091308,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18785444647073746,
      "num_tokens": 33642560.0,
      "step": 15080
    },
    {
      "entropy": 5.542061471939087,
      "epoch": 1.694221373617117,
      "grad_norm": 1.0625,
      "learning_rate": 0.000472172097616375,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19611501097679138,
      "num_tokens": 33654706.0,
      "step": 15085
    },
    {
      "entropy": 5.549911642074585,
      "epoch": 1.6947829505250744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.3324,
      "mean_token_accuracy": 0.18690803200006484,
      "num_tokens": 33665392.0,
      "step": 15090
    },
    {
      "entropy": 5.366191291809082,
      "epoch": 1.6953445274330319,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.1547,
      "mean_token_accuracy": 0.20463827401399612,
      "num_tokens": 33675504.0,
      "step": 15095
    },
    {
      "entropy": 5.481257772445678,
      "epoch": 1.6959061043409895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19443414509296417,
      "num_tokens": 33686505.0,
      "step": 15100
    },
    {
      "entropy": 5.546137285232544,
      "epoch": 1.6964676812489472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.3132,
      "mean_token_accuracy": 0.19328574240207672,
      "num_tokens": 33698752.0,
      "step": 15105
    },
    {
      "entropy": 5.582325458526611,
      "epoch": 1.6970292581569046,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.3589,
      "mean_token_accuracy": 0.1860020011663437,
      "num_tokens": 33710323.0,
      "step": 15110
    },
    {
      "entropy": 5.550843000411987,
      "epoch": 1.697590835064862,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.4268,
      "mean_token_accuracy": 0.17912996262311937,
      "num_tokens": 33721875.0,
      "step": 15115
    },
    {
      "entropy": 5.423304080963135,
      "epoch": 1.6981524119728197,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004720365583207737,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20194465965032576,
      "num_tokens": 33731520.0,
      "step": 15120
    },
    {
      "entropy": 5.565987682342529,
      "epoch": 1.6987139888807772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.3699,
      "mean_token_accuracy": 0.19037993550300597,
      "num_tokens": 33743351.0,
      "step": 15125
    },
    {
      "entropy": 5.545731115341186,
      "epoch": 1.6992755657887346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.307,
      "mean_token_accuracy": 0.19767358154058456,
      "num_tokens": 33754879.0,
      "step": 15130
    },
    {
      "entropy": 5.4189637184143065,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.226,
      "mean_token_accuracy": 0.19588263034820558,
      "num_tokens": 33766671.0,
      "step": 15135
    },
    {
      "entropy": 5.50096173286438,
      "epoch": 1.70039871960465,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.2811,
      "mean_token_accuracy": 0.1901957631111145,
      "num_tokens": 33778236.0,
      "step": 15140
    },
    {
      "entropy": 5.460691261291504,
      "epoch": 1.7009602965126074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.1802,
      "mean_token_accuracy": 0.19543107151985167,
      "num_tokens": 33788898.0,
      "step": 15145
    },
    {
      "entropy": 5.487517404556274,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1876659758388996,
      "num_tokens": 33800641.0,
      "step": 15150
    },
    {
      "entropy": 5.540626287460327,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19562939256429673,
      "num_tokens": 33812581.0,
      "step": 15155
    },
    {
      "entropy": 5.463412809371948,
      "epoch": 1.7026450272364801,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.2045,
      "mean_token_accuracy": 0.20464757829904556,
      "num_tokens": 33823378.0,
      "step": 15160
    },
    {
      "entropy": 5.461296606063843,
      "epoch": 1.7032066041444376,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.3165,
      "mean_token_accuracy": 0.1963017150759697,
      "num_tokens": 33834804.0,
      "step": 15165
    },
    {
      "entropy": 5.43390645980835,
      "epoch": 1.703768181052395,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.2746,
      "mean_token_accuracy": 0.19094635248184205,
      "num_tokens": 33847044.0,
      "step": 15170
    },
    {
      "entropy": 5.508099842071533,
      "epoch": 1.7043297579603527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.2122,
      "mean_token_accuracy": 0.19231095761060715,
      "num_tokens": 33858077.0,
      "step": 15175
    },
    {
      "entropy": 5.490667390823364,
      "epoch": 1.7048913348683103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.2713,
      "mean_token_accuracy": 0.19268581122159958,
      "num_tokens": 33868765.0,
      "step": 15180
    },
    {
      "entropy": 5.427499103546142,
      "epoch": 1.7054529117762678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19691985994577407,
      "num_tokens": 33880415.0,
      "step": 15185
    },
    {
      "entropy": 5.442188405990601,
      "epoch": 1.7060144886842252,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.2758,
      "mean_token_accuracy": 0.19425488114356995,
      "num_tokens": 33892081.0,
      "step": 15190
    },
    {
      "entropy": 5.5596880435943605,
      "epoch": 1.706576065592183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.3918,
      "mean_token_accuracy": 0.19346689730882644,
      "num_tokens": 33903433.0,
      "step": 15195
    },
    {
      "entropy": 5.5551369190216064,
      "epoch": 1.7071376425001406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19400326013565064,
      "num_tokens": 33914706.0,
      "step": 15200
    },
    {
      "entropy": 5.449495697021485,
      "epoch": 1.707699219408098,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.3999,
      "mean_token_accuracy": 0.1895191490650177,
      "num_tokens": 33927195.0,
      "step": 15205
    },
    {
      "entropy": 5.500268507003784,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.2631,
      "mean_token_accuracy": 0.2011074662208557,
      "num_tokens": 33938422.0,
      "step": 15210
    },
    {
      "entropy": 5.5860333919525145,
      "epoch": 1.708822373224013,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.307,
      "mean_token_accuracy": 0.1957274407148361,
      "num_tokens": 33949775.0,
      "step": 15215
    },
    {
      "entropy": 5.467221641540528,
      "epoch": 1.7093839501319705,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.2383,
      "mean_token_accuracy": 0.1993105500936508,
      "num_tokens": 33960371.0,
      "step": 15220
    },
    {
      "entropy": 5.385531806945801,
      "epoch": 1.709945527039928,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19383062571287155,
      "num_tokens": 33971326.0,
      "step": 15225
    },
    {
      "entropy": 5.504047966003418,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.3208,
      "mean_token_accuracy": 0.18932489156723023,
      "num_tokens": 33982744.0,
      "step": 15230
    },
    {
      "entropy": 5.557673597335816,
      "epoch": 1.7110686808558433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1907482475042343,
      "num_tokens": 33993647.0,
      "step": 15235
    },
    {
      "entropy": 5.43685359954834,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.218,
      "mean_token_accuracy": 0.19216079264879227,
      "num_tokens": 34004026.0,
      "step": 15240
    },
    {
      "entropy": 5.456605243682861,
      "epoch": 1.7121918346717582,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.2908,
      "mean_token_accuracy": 0.19980739653110505,
      "num_tokens": 34014882.0,
      "step": 15245
    },
    {
      "entropy": 5.462295484542847,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.3148,
      "mean_token_accuracy": 0.1958867073059082,
      "num_tokens": 34026548.0,
      "step": 15250
    },
    {
      "entropy": 5.478177642822265,
      "epoch": 1.7133149884876735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000471510886150377,
      "loss": 5.1942,
      "mean_token_accuracy": 0.19848761707544327,
      "num_tokens": 34038393.0,
      "step": 15255
    },
    {
      "entropy": 5.488607406616211,
      "epoch": 1.713876565395631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18940568268299102,
      "num_tokens": 34050257.0,
      "step": 15260
    },
    {
      "entropy": 5.4760009765625,
      "epoch": 1.7144381423035884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.2623,
      "mean_token_accuracy": 0.19758107215166093,
      "num_tokens": 34061860.0,
      "step": 15265
    },
    {
      "entropy": 5.56711974143982,
      "epoch": 1.714999719211546,
      "grad_norm": 1.125,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.3305,
      "mean_token_accuracy": 0.1937226578593254,
      "num_tokens": 34075247.0,
      "step": 15270
    },
    {
      "entropy": 5.47876615524292,
      "epoch": 1.7155612961195037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.1915,
      "mean_token_accuracy": 0.19317598789930343,
      "num_tokens": 34085801.0,
      "step": 15275
    },
    {
      "entropy": 5.505070304870605,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1918347880244255,
      "num_tokens": 34095686.0,
      "step": 15280
    },
    {
      "entropy": 5.481535577774048,
      "epoch": 1.7166844499354186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20252699553966522,
      "num_tokens": 34106518.0,
      "step": 15285
    },
    {
      "entropy": 5.483408355712891,
      "epoch": 1.7172460268433762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.2838,
      "mean_token_accuracy": 0.20250847339630126,
      "num_tokens": 34117793.0,
      "step": 15290
    },
    {
      "entropy": 5.519010591506958,
      "epoch": 1.717807603751334,
      "grad_norm": 1.203125,
      "learning_rate": 0.000471354254904924,
      "loss": 5.3781,
      "mean_token_accuracy": 0.19711277782917022,
      "num_tokens": 34128748.0,
      "step": 15295
    },
    {
      "entropy": 5.456810760498047,
      "epoch": 1.7183691806592913,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1997857466340065,
      "num_tokens": 34140597.0,
      "step": 15300
    },
    {
      "entropy": 5.529809427261353,
      "epoch": 1.7189307575672488,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000471315034561129,
      "loss": 5.2623,
      "mean_token_accuracy": 0.18573045134544372,
      "num_tokens": 34151823.0,
      "step": 15305
    },
    {
      "entropy": 5.455437755584716,
      "epoch": 1.7194923344752064,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.2144,
      "mean_token_accuracy": 0.20137128233909607,
      "num_tokens": 34163697.0,
      "step": 15310
    },
    {
      "entropy": 5.5017376899719235,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.2105,
      "mean_token_accuracy": 0.2000478282570839,
      "num_tokens": 34175058.0,
      "step": 15315
    },
    {
      "entropy": 5.513131189346313,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.125,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.3766,
      "mean_token_accuracy": 0.19343230426311492,
      "num_tokens": 34186883.0,
      "step": 15320
    },
    {
      "entropy": 5.455159044265747,
      "epoch": 1.721177065199079,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.2357,
      "mean_token_accuracy": 0.2001147523522377,
      "num_tokens": 34198103.0,
      "step": 15325
    },
    {
      "entropy": 5.435485792160034,
      "epoch": 1.7217386421070366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.2352,
      "mean_token_accuracy": 0.19808426797389983,
      "num_tokens": 34208747.0,
      "step": 15330
    },
    {
      "entropy": 5.4742655754089355,
      "epoch": 1.722300219014994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.344,
      "mean_token_accuracy": 0.19638944715261458,
      "num_tokens": 34220775.0,
      "step": 15335
    },
    {
      "entropy": 5.511271238327026,
      "epoch": 1.7228617959229515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.1973,
      "mean_token_accuracy": 0.20363293141126632,
      "num_tokens": 34231031.0,
      "step": 15340
    },
    {
      "entropy": 5.517343616485595,
      "epoch": 1.7234233728309092,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.306,
      "mean_token_accuracy": 0.19267010390758516,
      "num_tokens": 34241850.0,
      "step": 15345
    },
    {
      "entropy": 5.462200450897217,
      "epoch": 1.7239849497388668,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.2123,
      "mean_token_accuracy": 0.20199716836214066,
      "num_tokens": 34252732.0,
      "step": 15350
    },
    {
      "entropy": 5.44701771736145,
      "epoch": 1.7245465266468243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.2183,
      "mean_token_accuracy": 0.19148160070180892,
      "num_tokens": 34264066.0,
      "step": 15355
    },
    {
      "entropy": 5.579038953781128,
      "epoch": 1.7251081035547817,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.3422,
      "mean_token_accuracy": 0.18472560942173005,
      "num_tokens": 34275239.0,
      "step": 15360
    },
    {
      "entropy": 5.4547121047973635,
      "epoch": 1.7256696804627394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.1846,
      "mean_token_accuracy": 0.19288887530565263,
      "num_tokens": 34285723.0,
      "step": 15365
    },
    {
      "entropy": 5.448443651199341,
      "epoch": 1.726231257370697,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.2297,
      "mean_token_accuracy": 0.20415070056915283,
      "num_tokens": 34296292.0,
      "step": 15370
    },
    {
      "entropy": 5.4620811462402346,
      "epoch": 1.7267928342786545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.271,
      "mean_token_accuracy": 0.20204923003911973,
      "num_tokens": 34307232.0,
      "step": 15375
    },
    {
      "entropy": 5.456731128692627,
      "epoch": 1.727354411186612,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.2846,
      "mean_token_accuracy": 0.20479981154203414,
      "num_tokens": 34318933.0,
      "step": 15380
    },
    {
      "entropy": 5.467982053756714,
      "epoch": 1.7279159880945696,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.219,
      "mean_token_accuracy": 0.20281111299991608,
      "num_tokens": 34330431.0,
      "step": 15385
    },
    {
      "entropy": 5.538415193557739,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19124918133020402,
      "num_tokens": 34342006.0,
      "step": 15390
    },
    {
      "entropy": 5.49035267829895,
      "epoch": 1.7290391419104847,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19346456080675126,
      "num_tokens": 34353390.0,
      "step": 15395
    },
    {
      "entropy": 5.392697095870972,
      "epoch": 1.7296007188184421,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.2252,
      "mean_token_accuracy": 0.19382397532463075,
      "num_tokens": 34364211.0,
      "step": 15400
    },
    {
      "entropy": 5.492289972305298,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.2501,
      "mean_token_accuracy": 0.19321486055850984,
      "num_tokens": 34375362.0,
      "step": 15405
    },
    {
      "entropy": 5.424299049377441,
      "epoch": 1.7307238726343572,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004709017123528462,
      "loss": 5.1217,
      "mean_token_accuracy": 0.2064367413520813,
      "num_tokens": 34385511.0,
      "step": 15410
    },
    {
      "entropy": 5.4561442852020265,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.3609,
      "mean_token_accuracy": 0.1889231473207474,
      "num_tokens": 34397060.0,
      "step": 15415
    },
    {
      "entropy": 5.437074089050293,
      "epoch": 1.7318470264502723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.26,
      "mean_token_accuracy": 0.1962616741657257,
      "num_tokens": 34407703.0,
      "step": 15420
    },
    {
      "entropy": 5.5071210861206055,
      "epoch": 1.73240860335823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.2773,
      "mean_token_accuracy": 0.198504975438118,
      "num_tokens": 34418929.0,
      "step": 15425
    },
    {
      "entropy": 5.429887056350708,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20248452574014664,
      "num_tokens": 34430693.0,
      "step": 15430
    },
    {
      "entropy": 5.418445587158203,
      "epoch": 1.7335317571741449,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19408808797597885,
      "num_tokens": 34442299.0,
      "step": 15435
    },
    {
      "entropy": 5.463304042816162,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.2,
      "mean_token_accuracy": 0.1958053559064865,
      "num_tokens": 34453426.0,
      "step": 15440
    },
    {
      "entropy": 5.412197923660278,
      "epoch": 1.7346549109900602,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19891296476125717,
      "num_tokens": 34463601.0,
      "step": 15445
    },
    {
      "entropy": 5.470721340179443,
      "epoch": 1.7352164878980176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19638391137123107,
      "num_tokens": 34474938.0,
      "step": 15450
    },
    {
      "entropy": 5.44405951499939,
      "epoch": 1.735778064805975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.1614,
      "mean_token_accuracy": 0.20331384241580963,
      "num_tokens": 34485244.0,
      "step": 15455
    },
    {
      "entropy": 5.372856950759887,
      "epoch": 1.7363396417139327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.1628,
      "mean_token_accuracy": 0.2047601506114006,
      "num_tokens": 34496912.0,
      "step": 15460
    },
    {
      "entropy": 5.367724132537842,
      "epoch": 1.7369012186218904,
      "grad_norm": 1.0,
      "learning_rate": 0.000470684112060605,
      "loss": 5.2424,
      "mean_token_accuracy": 0.20117949694395065,
      "num_tokens": 34508122.0,
      "step": 15465
    },
    {
      "entropy": 5.451254892349243,
      "epoch": 1.7374627955298478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.2435,
      "mean_token_accuracy": 0.19873499423265456,
      "num_tokens": 34519522.0,
      "step": 15470
    },
    {
      "entropy": 5.483699464797974,
      "epoch": 1.7380243724378053,
      "grad_norm": 1.03125,
      "learning_rate": 0.000470644467347967,
      "loss": 5.236,
      "mean_token_accuracy": 0.1985923245549202,
      "num_tokens": 34532832.0,
      "step": 15475
    },
    {
      "entropy": 5.485170793533325,
      "epoch": 1.738585949345763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19652123749256134,
      "num_tokens": 34544817.0,
      "step": 15480
    },
    {
      "entropy": 5.406769561767578,
      "epoch": 1.7391475262537206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.1687,
      "mean_token_accuracy": 0.19541578143835067,
      "num_tokens": 34555391.0,
      "step": 15485
    },
    {
      "entropy": 5.411785364151001,
      "epoch": 1.739709103161678,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19859668761491775,
      "num_tokens": 34567531.0,
      "step": 15490
    },
    {
      "entropy": 5.455117177963257,
      "epoch": 1.7402706800696355,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19563011229038238,
      "num_tokens": 34578212.0,
      "step": 15495
    },
    {
      "entropy": 5.396497392654419,
      "epoch": 1.7408322569775931,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.1585,
      "mean_token_accuracy": 0.2035893142223358,
      "num_tokens": 34589331.0,
      "step": 15500
    },
    {
      "entropy": 5.37835464477539,
      "epoch": 1.7413938338855506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.2407,
      "mean_token_accuracy": 0.20102326571941376,
      "num_tokens": 34601467.0,
      "step": 15505
    },
    {
      "entropy": 5.386589241027832,
      "epoch": 1.741955410793508,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.2042,
      "mean_token_accuracy": 0.20691325664520263,
      "num_tokens": 34611608.0,
      "step": 15510
    },
    {
      "entropy": 5.485888051986694,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.125,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20545875430107116,
      "num_tokens": 34622495.0,
      "step": 15515
    },
    {
      "entropy": 5.585084772109985,
      "epoch": 1.7430785646094233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.3814,
      "mean_token_accuracy": 0.19156555086374283,
      "num_tokens": 34633279.0,
      "step": 15520
    },
    {
      "entropy": 5.439335203170776,
      "epoch": 1.7436401415173808,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.2352,
      "mean_token_accuracy": 0.19934522956609727,
      "num_tokens": 34644030.0,
      "step": 15525
    },
    {
      "entropy": 5.572230529785156,
      "epoch": 1.7442017184253382,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.3478,
      "mean_token_accuracy": 0.18335260599851608,
      "num_tokens": 34655540.0,
      "step": 15530
    },
    {
      "entropy": 5.507998752593994,
      "epoch": 1.7447632953332959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.2581,
      "mean_token_accuracy": 0.19626591503620147,
      "num_tokens": 34666752.0,
      "step": 15535
    },
    {
      "entropy": 5.519788789749145,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.2682,
      "mean_token_accuracy": 0.18972883820533754,
      "num_tokens": 34678670.0,
      "step": 15540
    },
    {
      "entropy": 5.502021837234497,
      "epoch": 1.745886449149211,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19230482280254363,
      "num_tokens": 34690848.0,
      "step": 15545
    },
    {
      "entropy": 5.496710634231567,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.307,
      "mean_token_accuracy": 0.1888074278831482,
      "num_tokens": 34703135.0,
      "step": 15550
    },
    {
      "entropy": 5.449506950378418,
      "epoch": 1.747009602965126,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.2274,
      "mean_token_accuracy": 0.1977105662226677,
      "num_tokens": 34714329.0,
      "step": 15555
    },
    {
      "entropy": 5.447259521484375,
      "epoch": 1.7475711798730837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.2857,
      "mean_token_accuracy": 0.18809677958488463,
      "num_tokens": 34727032.0,
      "step": 15560
    },
    {
      "entropy": 5.492578363418579,
      "epoch": 1.7481327567810412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19404125064611435,
      "num_tokens": 34738600.0,
      "step": 15565
    },
    {
      "entropy": 5.538525152206421,
      "epoch": 1.7486943336889986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1916448712348938,
      "num_tokens": 34749810.0,
      "step": 15570
    },
    {
      "entropy": 5.664850044250488,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.4812,
      "mean_token_accuracy": 0.18054275512695311,
      "num_tokens": 34762029.0,
      "step": 15575
    },
    {
      "entropy": 5.4572672843933105,
      "epoch": 1.749817487504914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.3264,
      "mean_token_accuracy": 0.19237176775932313,
      "num_tokens": 34773160.0,
      "step": 15580
    },
    {
      "entropy": 5.430907821655273,
      "epoch": 1.7503790644128714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.3018,
      "mean_token_accuracy": 0.1918123260140419,
      "num_tokens": 34784659.0,
      "step": 15585
    },
    {
      "entropy": 5.607436609268189,
      "epoch": 1.7509406413208288,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000470186763474261,
      "loss": 5.2669,
      "mean_token_accuracy": 0.19227909594774245,
      "num_tokens": 34796976.0,
      "step": 15590
    },
    {
      "entropy": 5.456044387817383,
      "epoch": 1.7515022182287865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.2123,
      "mean_token_accuracy": 0.19207327514886857,
      "num_tokens": 34809271.0,
      "step": 15595
    },
    {
      "entropy": 5.47929048538208,
      "epoch": 1.752063795136744,
      "grad_norm": 1.125,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.2911,
      "mean_token_accuracy": 0.1998542532324791,
      "num_tokens": 34820291.0,
      "step": 15600
    },
    {
      "entropy": 5.388984823226929,
      "epoch": 1.7526253720447014,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.2063,
      "mean_token_accuracy": 0.20087198317050933,
      "num_tokens": 34831224.0,
      "step": 15605
    },
    {
      "entropy": 5.501365089416504,
      "epoch": 1.753186948952659,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.2765,
      "mean_token_accuracy": 0.19637789130210875,
      "num_tokens": 34842947.0,
      "step": 15610
    },
    {
      "entropy": 5.398352193832397,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.2275,
      "mean_token_accuracy": 0.2049104556441307,
      "num_tokens": 34853422.0,
      "step": 15615
    },
    {
      "entropy": 5.430916976928711,
      "epoch": 1.7543101027685741,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.232,
      "mean_token_accuracy": 0.19215001314878463,
      "num_tokens": 34864082.0,
      "step": 15620
    },
    {
      "entropy": 5.470525598526001,
      "epoch": 1.7548716796765316,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.2221,
      "mean_token_accuracy": 0.20426664054393767,
      "num_tokens": 34874094.0,
      "step": 15625
    },
    {
      "entropy": 5.4323986053466795,
      "epoch": 1.7554332565844892,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.2472,
      "mean_token_accuracy": 0.20054281800985335,
      "num_tokens": 34884597.0,
      "step": 15630
    },
    {
      "entropy": 5.378446292877197,
      "epoch": 1.755994833492447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047000676659768477,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20713614374399186,
      "num_tokens": 34894746.0,
      "step": 15635
    },
    {
      "entropy": 5.4310376167297365,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.1406,
      "mean_token_accuracy": 0.2001330718398094,
      "num_tokens": 34906250.0,
      "step": 15640
    },
    {
      "entropy": 5.482785892486572,
      "epoch": 1.7571179873083618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.4081,
      "mean_token_accuracy": 0.1867247298359871,
      "num_tokens": 34917772.0,
      "step": 15645
    },
    {
      "entropy": 5.500496768951416,
      "epoch": 1.7576795642163194,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.2443,
      "mean_token_accuracy": 0.2001166522502899,
      "num_tokens": 34930243.0,
      "step": 15650
    },
    {
      "entropy": 5.471747255325317,
      "epoch": 1.758241141124277,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19575004726648332,
      "num_tokens": 34941357.0,
      "step": 15655
    },
    {
      "entropy": 5.59677939414978,
      "epoch": 1.7588027180322345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.3896,
      "mean_token_accuracy": 0.18695174157619476,
      "num_tokens": 34952116.0,
      "step": 15660
    },
    {
      "entropy": 5.5312329769134525,
      "epoch": 1.759364294940192,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.2817,
      "mean_token_accuracy": 0.19602534025907517,
      "num_tokens": 34963707.0,
      "step": 15665
    },
    {
      "entropy": 5.5054035663604735,
      "epoch": 1.7599258718481496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.3786,
      "mean_token_accuracy": 0.19636891037225723,
      "num_tokens": 34974376.0,
      "step": 15670
    },
    {
      "entropy": 5.486743783950805,
      "epoch": 1.7604874487561073,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.2709,
      "mean_token_accuracy": 0.19557513743638993,
      "num_tokens": 34986183.0,
      "step": 15675
    },
    {
      "entropy": 5.59564528465271,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.125,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.2901,
      "mean_token_accuracy": 0.19822811633348464,
      "num_tokens": 34997137.0,
      "step": 15680
    },
    {
      "entropy": 5.615262269973755,
      "epoch": 1.7616106025720222,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.4418,
      "mean_token_accuracy": 0.18901116251945496,
      "num_tokens": 35008944.0,
      "step": 15685
    },
    {
      "entropy": 5.52647852897644,
      "epoch": 1.7621721794799798,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.3271,
      "mean_token_accuracy": 0.18539503663778306,
      "num_tokens": 35019972.0,
      "step": 15690
    },
    {
      "entropy": 5.438716888427734,
      "epoch": 1.7627337563879373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.2,
      "mean_token_accuracy": 0.19455612301826478,
      "num_tokens": 35031285.0,
      "step": 15695
    },
    {
      "entropy": 5.41926679611206,
      "epoch": 1.7632953332958947,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.1745,
      "mean_token_accuracy": 0.1998610109090805,
      "num_tokens": 35042553.0,
      "step": 15700
    },
    {
      "entropy": 5.589274215698242,
      "epoch": 1.7638569102038524,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.4069,
      "mean_token_accuracy": 0.1867491528391838,
      "num_tokens": 35053637.0,
      "step": 15705
    },
    {
      "entropy": 5.545748233795166,
      "epoch": 1.76441848711181,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.3255,
      "mean_token_accuracy": 0.19714967012405396,
      "num_tokens": 35064754.0,
      "step": 15710
    },
    {
      "entropy": 5.516838312149048,
      "epoch": 1.7649800640197675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.2512,
      "mean_token_accuracy": 0.19179097414016724,
      "num_tokens": 35075816.0,
      "step": 15715
    },
    {
      "entropy": 5.488126039505005,
      "epoch": 1.765541640927725,
      "grad_norm": 1.25,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.3452,
      "mean_token_accuracy": 0.1930240049958229,
      "num_tokens": 35088647.0,
      "step": 15720
    },
    {
      "entropy": 5.480515480041504,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.2294,
      "mean_token_accuracy": 0.19629358798265456,
      "num_tokens": 35099372.0,
      "step": 15725
    },
    {
      "entropy": 5.499585437774658,
      "epoch": 1.7666647947436402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046962512213995,
      "loss": 5.3761,
      "mean_token_accuracy": 0.18661774545907975,
      "num_tokens": 35110663.0,
      "step": 15730
    },
    {
      "entropy": 5.648336124420166,
      "epoch": 1.7672263716515977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.4534,
      "mean_token_accuracy": 0.18458335548639299,
      "num_tokens": 35122406.0,
      "step": 15735
    },
    {
      "entropy": 5.513351345062256,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19630270451307297,
      "num_tokens": 35133444.0,
      "step": 15740
    },
    {
      "entropy": 5.438093185424805,
      "epoch": 1.7683495254675128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.2382,
      "mean_token_accuracy": 0.19741030931472778,
      "num_tokens": 35143840.0,
      "step": 15745
    },
    {
      "entropy": 5.566193914413452,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.4177,
      "mean_token_accuracy": 0.1819252073764801,
      "num_tokens": 35155050.0,
      "step": 15750
    },
    {
      "entropy": 5.610992860794068,
      "epoch": 1.7694726792834279,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19293665736913682,
      "num_tokens": 35167649.0,
      "step": 15755
    },
    {
      "entropy": 5.490816593170166,
      "epoch": 1.7700342561913853,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.2591,
      "mean_token_accuracy": 0.19580239206552505,
      "num_tokens": 35179238.0,
      "step": 15760
    },
    {
      "entropy": 5.520217037200927,
      "epoch": 1.770595833099343,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.309,
      "mean_token_accuracy": 0.19635985642671586,
      "num_tokens": 35189510.0,
      "step": 15765
    },
    {
      "entropy": 5.525603818893432,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19275762587785722,
      "num_tokens": 35199727.0,
      "step": 15770
    },
    {
      "entropy": 5.445923900604248,
      "epoch": 1.771718986915258,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.2391,
      "mean_token_accuracy": 0.2015764221549034,
      "num_tokens": 35210311.0,
      "step": 15775
    },
    {
      "entropy": 5.498425102233886,
      "epoch": 1.7722805638232155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.2613,
      "mean_token_accuracy": 0.19993872940540314,
      "num_tokens": 35222755.0,
      "step": 15780
    },
    {
      "entropy": 5.504217624664307,
      "epoch": 1.7728421407311732,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.2922,
      "mean_token_accuracy": 0.19344490617513657,
      "num_tokens": 35234337.0,
      "step": 15785
    },
    {
      "entropy": 5.456560945510864,
      "epoch": 1.7734037176391306,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.2553,
      "mean_token_accuracy": 0.19621694087982178,
      "num_tokens": 35245736.0,
      "step": 15790
    },
    {
      "entropy": 5.473315620422364,
      "epoch": 1.773965294547088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.2117,
      "mean_token_accuracy": 0.20478270202875137,
      "num_tokens": 35256368.0,
      "step": 15795
    },
    {
      "entropy": 5.440543603897095,
      "epoch": 1.7745268714550457,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19716603606939315,
      "num_tokens": 35267238.0,
      "step": 15800
    },
    {
      "entropy": 5.477315092086792,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.2303,
      "mean_token_accuracy": 0.1946992829442024,
      "num_tokens": 35277113.0,
      "step": 15805
    },
    {
      "entropy": 5.5034326076507565,
      "epoch": 1.7756500252709608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1909641832113266,
      "num_tokens": 35288504.0,
      "step": 15810
    },
    {
      "entropy": 5.443775177001953,
      "epoch": 1.7762116021789183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.2202,
      "mean_token_accuracy": 0.2007802963256836,
      "num_tokens": 35301215.0,
      "step": 15815
    },
    {
      "entropy": 5.40267972946167,
      "epoch": 1.776773179086876,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000469261500582484,
      "loss": 5.2132,
      "mean_token_accuracy": 0.19696372598409653,
      "num_tokens": 35311997.0,
      "step": 15820
    },
    {
      "entropy": 5.481546401977539,
      "epoch": 1.7773347559948336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.2891,
      "mean_token_accuracy": 0.1924848809838295,
      "num_tokens": 35323217.0,
      "step": 15825
    },
    {
      "entropy": 5.484462308883667,
      "epoch": 1.777896332902791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.3161,
      "mean_token_accuracy": 0.19927048087120056,
      "num_tokens": 35333102.0,
      "step": 15830
    },
    {
      "entropy": 5.587183046340942,
      "epoch": 1.7784579098107485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.3372,
      "mean_token_accuracy": 0.19091015607118605,
      "num_tokens": 35346394.0,
      "step": 15835
    },
    {
      "entropy": 5.3949127197265625,
      "epoch": 1.7790194867187061,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004691804235037845,
      "loss": 5.1393,
      "mean_token_accuracy": 0.203298519551754,
      "num_tokens": 35357827.0,
      "step": 15840
    },
    {
      "entropy": 5.410745096206665,
      "epoch": 1.7795810636266638,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1943773955106735,
      "num_tokens": 35370517.0,
      "step": 15845
    },
    {
      "entropy": 5.503803491592407,
      "epoch": 1.7801426405346212,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.2924,
      "mean_token_accuracy": 0.1901368245482445,
      "num_tokens": 35381830.0,
      "step": 15850
    },
    {
      "entropy": 5.606777811050415,
      "epoch": 1.7807042174425787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.3757,
      "mean_token_accuracy": 0.18714867383241654,
      "num_tokens": 35393161.0,
      "step": 15855
    },
    {
      "entropy": 5.52755184173584,
      "epoch": 1.7812657943505363,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.2141,
      "mean_token_accuracy": 0.20724063068628312,
      "num_tokens": 35403338.0,
      "step": 15860
    },
    {
      "entropy": 5.49515929222107,
      "epoch": 1.781827371258494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.3072,
      "mean_token_accuracy": 0.19043599367141723,
      "num_tokens": 35415862.0,
      "step": 15865
    },
    {
      "entropy": 5.427636623382568,
      "epoch": 1.7823889481664514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.22,
      "mean_token_accuracy": 0.19370841532945632,
      "num_tokens": 35426696.0,
      "step": 15870
    },
    {
      "entropy": 5.492132997512817,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.2803,
      "mean_token_accuracy": 0.19535786509513856,
      "num_tokens": 35438624.0,
      "step": 15875
    },
    {
      "entropy": 5.583156633377075,
      "epoch": 1.7835121019823665,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.3933,
      "mean_token_accuracy": 0.1896327555179596,
      "num_tokens": 35449841.0,
      "step": 15880
    },
    {
      "entropy": 5.679481792449951,
      "epoch": 1.784073678890324,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18230774253606796,
      "num_tokens": 35462081.0,
      "step": 15885
    },
    {
      "entropy": 5.437917423248291,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.1847,
      "mean_token_accuracy": 0.20067851096391678,
      "num_tokens": 35472834.0,
      "step": 15890
    },
    {
      "entropy": 5.461607217788696,
      "epoch": 1.785196832706239,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.1429,
      "mean_token_accuracy": 0.20348993986845015,
      "num_tokens": 35482975.0,
      "step": 15895
    },
    {
      "entropy": 5.392340517044067,
      "epoch": 1.7857584096141967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.2522,
      "mean_token_accuracy": 0.19824526607990264,
      "num_tokens": 35494470.0,
      "step": 15900
    },
    {
      "entropy": 5.408385181427002,
      "epoch": 1.7863199865221542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.199,
      "mean_token_accuracy": 0.19572625756263734,
      "num_tokens": 35505464.0,
      "step": 15905
    },
    {
      "entropy": 5.428882598876953,
      "epoch": 1.7868815634301116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1960815206170082,
      "num_tokens": 35517609.0,
      "step": 15910
    },
    {
      "entropy": 5.457346963882446,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.1663,
      "mean_token_accuracy": 0.19810595065355302,
      "num_tokens": 35528994.0,
      "step": 15915
    },
    {
      "entropy": 5.498655128479004,
      "epoch": 1.788004717246027,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.3601,
      "mean_token_accuracy": 0.18583411127328872,
      "num_tokens": 35541044.0,
      "step": 15920
    },
    {
      "entropy": 5.401097106933594,
      "epoch": 1.7885662941539844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.1783,
      "mean_token_accuracy": 0.2050304338335991,
      "num_tokens": 35552660.0,
      "step": 15925
    },
    {
      "entropy": 5.543985891342163,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.2892,
      "mean_token_accuracy": 0.18554833233356477,
      "num_tokens": 35565144.0,
      "step": 15930
    },
    {
      "entropy": 5.499610090255738,
      "epoch": 1.7896894479698995,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19383473843336105,
      "num_tokens": 35577067.0,
      "step": 15935
    },
    {
      "entropy": 5.408252334594726,
      "epoch": 1.7902510248778571,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19355790317058563,
      "num_tokens": 35588712.0,
      "step": 15940
    },
    {
      "entropy": 5.494270086288452,
      "epoch": 1.7908126017858146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.1873,
      "mean_token_accuracy": 0.19448510706424713,
      "num_tokens": 35599212.0,
      "step": 15945
    },
    {
      "entropy": 5.500825452804565,
      "epoch": 1.791374178693772,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.24,
      "mean_token_accuracy": 0.20169495046138763,
      "num_tokens": 35609397.0,
      "step": 15950
    },
    {
      "entropy": 5.444353532791138,
      "epoch": 1.7919357556017297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19417669177055358,
      "num_tokens": 35620615.0,
      "step": 15955
    },
    {
      "entropy": 5.504662752151489,
      "epoch": 1.7924973325096873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19662481546401978,
      "num_tokens": 35632684.0,
      "step": 15960
    },
    {
      "entropy": 5.451061630249024,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.125,
      "learning_rate": 0.000468671452146028,
      "loss": 5.27,
      "mean_token_accuracy": 0.1996163472533226,
      "num_tokens": 35643896.0,
      "step": 15965
    },
    {
      "entropy": 5.5074351787567135,
      "epoch": 1.7936204863256022,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.2564,
      "mean_token_accuracy": 0.19853272438049316,
      "num_tokens": 35655752.0,
      "step": 15970
    },
    {
      "entropy": 5.398624134063721,
      "epoch": 1.7941820632335599,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468630567805468,
      "loss": 5.1545,
      "mean_token_accuracy": 0.19439476281404494,
      "num_tokens": 35665870.0,
      "step": 15975
    },
    {
      "entropy": 5.4185138702392575,
      "epoch": 1.7947436401415173,
      "grad_norm": 1.125,
      "learning_rate": 0.000468610116386954,
      "loss": 5.3174,
      "mean_token_accuracy": 0.19464239329099656,
      "num_tokens": 35677792.0,
      "step": 15980
    },
    {
      "entropy": 5.4135995388031,
      "epoch": 1.7953052170494748,
      "grad_norm": 1.125,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19553089141845703,
      "num_tokens": 35689059.0,
      "step": 15985
    },
    {
      "entropy": 5.543794345855713,
      "epoch": 1.7958667939574324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.3074,
      "mean_token_accuracy": 0.18938867300748824,
      "num_tokens": 35699953.0,
      "step": 15990
    },
    {
      "entropy": 5.529967403411865,
      "epoch": 1.79642837086539,
      "grad_norm": 1.125,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19743361026048661,
      "num_tokens": 35710894.0,
      "step": 15995
    },
    {
      "entropy": 5.441607904434204,
      "epoch": 1.7969899477733475,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.2778,
      "mean_token_accuracy": 0.1943812847137451,
      "num_tokens": 35722440.0,
      "step": 16000
    },
    {
      "entropy": 5.451500701904297,
      "epoch": 1.797551524681305,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.282,
      "mean_token_accuracy": 0.19846251755952835,
      "num_tokens": 35733768.0,
      "step": 16005
    },
    {
      "entropy": 5.510002946853637,
      "epoch": 1.7981131015892626,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19835041761398314,
      "num_tokens": 35744211.0,
      "step": 16010
    },
    {
      "entropy": 5.4808416843414305,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19498499482870102,
      "num_tokens": 35755498.0,
      "step": 16015
    },
    {
      "entropy": 5.5295497417449955,
      "epoch": 1.7992362554051777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19888902455568314,
      "num_tokens": 35767442.0,
      "step": 16020
    },
    {
      "entropy": 5.436308908462524,
      "epoch": 1.7997978323131352,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.2343,
      "mean_token_accuracy": 0.19678854644298555,
      "num_tokens": 35778040.0,
      "step": 16025
    },
    {
      "entropy": 5.4519213199615475,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.1788,
      "mean_token_accuracy": 0.20557676255702972,
      "num_tokens": 35788139.0,
      "step": 16030
    },
    {
      "entropy": 5.4884552478790285,
      "epoch": 1.8009209861290505,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.2338,
      "mean_token_accuracy": 0.19548669308423997,
      "num_tokens": 35799077.0,
      "step": 16035
    },
    {
      "entropy": 5.480828952789307,
      "epoch": 1.801482563037008,
      "grad_norm": 1.125,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.2878,
      "mean_token_accuracy": 0.19201088696718216,
      "num_tokens": 35809748.0,
      "step": 16040
    },
    {
      "entropy": 5.55096321105957,
      "epoch": 1.8020441399449654,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.3038,
      "mean_token_accuracy": 0.19516561478376387,
      "num_tokens": 35821621.0,
      "step": 16045
    },
    {
      "entropy": 5.522806835174561,
      "epoch": 1.802605716852923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.2519,
      "mean_token_accuracy": 0.1982177034020424,
      "num_tokens": 35833497.0,
      "step": 16050
    },
    {
      "entropy": 5.539760637283325,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.125,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.3593,
      "mean_token_accuracy": 0.19536954164505005,
      "num_tokens": 35844266.0,
      "step": 16055
    },
    {
      "entropy": 5.587708520889282,
      "epoch": 1.8037288706688381,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19403456151485443,
      "num_tokens": 35855260.0,
      "step": 16060
    },
    {
      "entropy": 5.535833501815796,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.2961,
      "mean_token_accuracy": 0.20049406588077545,
      "num_tokens": 35865950.0,
      "step": 16065
    },
    {
      "entropy": 5.503913021087646,
      "epoch": 1.8048520244847532,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.445,
      "mean_token_accuracy": 0.18648439794778823,
      "num_tokens": 35877028.0,
      "step": 16070
    },
    {
      "entropy": 5.492651748657226,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19604576826095582,
      "num_tokens": 35887831.0,
      "step": 16075
    },
    {
      "entropy": 5.5385983943939205,
      "epoch": 1.8059751783006681,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.2629,
      "mean_token_accuracy": 0.19517093896865845,
      "num_tokens": 35897946.0,
      "step": 16080
    },
    {
      "entropy": 5.603180360794068,
      "epoch": 1.8065367552086258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.4066,
      "mean_token_accuracy": 0.19121986776590347,
      "num_tokens": 35910973.0,
      "step": 16085
    },
    {
      "entropy": 5.517197322845459,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1891381949186325,
      "num_tokens": 35921478.0,
      "step": 16090
    },
    {
      "entropy": 5.503784275054931,
      "epoch": 1.8076599090245409,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.2078,
      "mean_token_accuracy": 0.20318155437707902,
      "num_tokens": 35932211.0,
      "step": 16095
    },
    {
      "entropy": 5.4260437965393065,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.3806,
      "mean_token_accuracy": 0.19121943563222885,
      "num_tokens": 35944350.0,
      "step": 16100
    },
    {
      "entropy": 5.521867036819458,
      "epoch": 1.808783062840456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.3439,
      "mean_token_accuracy": 0.18876016587018968,
      "num_tokens": 35955590.0,
      "step": 16105
    },
    {
      "entropy": 5.442447137832642,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004680762176322701,
      "loss": 5.1196,
      "mean_token_accuracy": 0.21004122346639634,
      "num_tokens": 35965872.0,
      "step": 16110
    },
    {
      "entropy": 5.381133699417115,
      "epoch": 1.809906216656371,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1935344398021698,
      "num_tokens": 35976851.0,
      "step": 16115
    },
    {
      "entropy": 5.4849670886993405,
      "epoch": 1.8104677935643285,
      "grad_norm": 1.125,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19965680986642836,
      "num_tokens": 35986984.0,
      "step": 16120
    },
    {
      "entropy": 5.505698776245117,
      "epoch": 1.8110293704722862,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.2773,
      "mean_token_accuracy": 0.19423348158597947,
      "num_tokens": 35999349.0,
      "step": 16125
    },
    {
      "entropy": 5.449534463882446,
      "epoch": 1.8115909473802438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.2896,
      "mean_token_accuracy": 0.20000939965248107,
      "num_tokens": 36011122.0,
      "step": 16130
    },
    {
      "entropy": 5.513085412979126,
      "epoch": 1.8121525242882013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046797306799879,
      "loss": 5.2164,
      "mean_token_accuracy": 0.19834259301424026,
      "num_tokens": 36022729.0,
      "step": 16135
    },
    {
      "entropy": 5.4503247261047365,
      "epoch": 1.8127141011961587,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19925459176301957,
      "num_tokens": 36034042.0,
      "step": 16140
    },
    {
      "entropy": 5.467282438278199,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.2626,
      "mean_token_accuracy": 0.1887116625905037,
      "num_tokens": 36045357.0,
      "step": 16145
    },
    {
      "entropy": 5.350513744354248,
      "epoch": 1.813837255012074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004679111044813987,
      "loss": 5.0236,
      "mean_token_accuracy": 0.20837678462266923,
      "num_tokens": 36055180.0,
      "step": 16150
    },
    {
      "entropy": 5.3760003566741945,
      "epoch": 1.8143988319200315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.176,
      "mean_token_accuracy": 0.20194455981254578,
      "num_tokens": 36065574.0,
      "step": 16155
    },
    {
      "entropy": 5.428332757949829,
      "epoch": 1.814960408827989,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.2658,
      "mean_token_accuracy": 0.1963086485862732,
      "num_tokens": 36077316.0,
      "step": 16160
    },
    {
      "entropy": 5.414357376098633,
      "epoch": 1.8155219857359466,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.2145,
      "mean_token_accuracy": 0.19487461745738982,
      "num_tokens": 36088735.0,
      "step": 16165
    },
    {
      "entropy": 5.44441294670105,
      "epoch": 1.8160835626439042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000467828400468182,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19373282343149184,
      "num_tokens": 36098806.0,
      "step": 16170
    },
    {
      "entropy": 5.493496513366699,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19467793405056,
      "num_tokens": 36110333.0,
      "step": 16175
    },
    {
      "entropy": 5.5141997814178465,
      "epoch": 1.8172067164598191,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.2599,
      "mean_token_accuracy": 0.19326572716236115,
      "num_tokens": 36121719.0,
      "step": 16180
    },
    {
      "entropy": 5.447924518585205,
      "epoch": 1.8177682933677768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.1539,
      "mean_token_accuracy": 0.2053838238120079,
      "num_tokens": 36132129.0,
      "step": 16185
    },
    {
      "entropy": 5.5286133766174315,
      "epoch": 1.8183298702757342,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.4494,
      "mean_token_accuracy": 0.1912141978740692,
      "num_tokens": 36145189.0,
      "step": 16190
    },
    {
      "entropy": 5.450881576538086,
      "epoch": 1.8188914471836917,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1981054201722145,
      "num_tokens": 36157896.0,
      "step": 16195
    },
    {
      "entropy": 5.522910785675049,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19353700876235963,
      "num_tokens": 36168842.0,
      "step": 16200
    },
    {
      "entropy": 5.356402206420898,
      "epoch": 1.820014600999607,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20072351694107055,
      "num_tokens": 36180075.0,
      "step": 16205
    },
    {
      "entropy": 5.407615089416504,
      "epoch": 1.8205761779075644,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.196,
      "mean_token_accuracy": 0.2022756114602089,
      "num_tokens": 36190589.0,
      "step": 16210
    },
    {
      "entropy": 5.39691276550293,
      "epoch": 1.8211377548155219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046764195732846013,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20610568076372146,
      "num_tokens": 36201145.0,
      "step": 16215
    },
    {
      "entropy": 5.43857479095459,
      "epoch": 1.8216993317234795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19864938110113145,
      "num_tokens": 36213243.0,
      "step": 16220
    },
    {
      "entropy": 5.353356170654297,
      "epoch": 1.8222609086314372,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.1938,
      "mean_token_accuracy": 0.2008663147687912,
      "num_tokens": 36224491.0,
      "step": 16225
    },
    {
      "entropy": 5.436663484573364,
      "epoch": 1.8228224855393946,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19710069447755812,
      "num_tokens": 36236006.0,
      "step": 16230
    },
    {
      "entropy": 5.587708473205566,
      "epoch": 1.823384062447352,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046755893420355,
      "loss": 5.3377,
      "mean_token_accuracy": 0.1891933336853981,
      "num_tokens": 36246925.0,
      "step": 16235
    },
    {
      "entropy": 5.518993234634399,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.2537,
      "mean_token_accuracy": 0.19068574011325837,
      "num_tokens": 36258449.0,
      "step": 16240
    },
    {
      "entropy": 5.467200994491577,
      "epoch": 1.8245072162632674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.2591,
      "mean_token_accuracy": 0.19682479798793792,
      "num_tokens": 36270820.0,
      "step": 16245
    },
    {
      "entropy": 5.498639965057373,
      "epoch": 1.8250687931712248,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.3254,
      "mean_token_accuracy": 0.19491575211286544,
      "num_tokens": 36282346.0,
      "step": 16250
    },
    {
      "entropy": 5.545383739471435,
      "epoch": 1.8256303700791823,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18515999615192413,
      "num_tokens": 36294111.0,
      "step": 16255
    },
    {
      "entropy": 5.48102822303772,
      "epoch": 1.82619194698714,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.2159,
      "mean_token_accuracy": 0.1969439998269081,
      "num_tokens": 36304899.0,
      "step": 16260
    },
    {
      "entropy": 5.483670806884765,
      "epoch": 1.8267535238950976,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.3663,
      "mean_token_accuracy": 0.18919069468975067,
      "num_tokens": 36316272.0,
      "step": 16265
    },
    {
      "entropy": 5.444512605667114,
      "epoch": 1.8273151008030548,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20142329186201097,
      "num_tokens": 36326624.0,
      "step": 16270
    },
    {
      "entropy": 5.44782829284668,
      "epoch": 1.8278766777110125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.1643,
      "mean_token_accuracy": 0.20240344554185868,
      "num_tokens": 36338178.0,
      "step": 16275
    },
    {
      "entropy": 5.51089334487915,
      "epoch": 1.8284382546189701,
      "grad_norm": 1.125,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1959345057606697,
      "num_tokens": 36349993.0,
      "step": 16280
    },
    {
      "entropy": 5.500933694839477,
      "epoch": 1.8289998315269276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.327,
      "mean_token_accuracy": 0.19261912554502486,
      "num_tokens": 36361002.0,
      "step": 16285
    },
    {
      "entropy": 5.463829231262207,
      "epoch": 1.829561408434885,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19682866632938384,
      "num_tokens": 36371688.0,
      "step": 16290
    },
    {
      "entropy": 5.455495691299438,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20718118399381638,
      "num_tokens": 36382683.0,
      "step": 16295
    },
    {
      "entropy": 5.526200532913208,
      "epoch": 1.8306845622508003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.3446,
      "mean_token_accuracy": 0.18963032066822053,
      "num_tokens": 36394170.0,
      "step": 16300
    },
    {
      "entropy": 5.527466106414795,
      "epoch": 1.8312461391587578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19231263548135757,
      "num_tokens": 36405613.0,
      "step": 16305
    },
    {
      "entropy": 5.506656646728516,
      "epoch": 1.8318077160667152,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1976760745048523,
      "num_tokens": 36416282.0,
      "step": 16310
    },
    {
      "entropy": 5.527156734466553,
      "epoch": 1.8323692929746729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.3443,
      "mean_token_accuracy": 0.18781709671020508,
      "num_tokens": 36427624.0,
      "step": 16315
    },
    {
      "entropy": 5.4442009925842285,
      "epoch": 1.8329308698826305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20778766721487046,
      "num_tokens": 36438230.0,
      "step": 16320
    },
    {
      "entropy": 5.486078929901123,
      "epoch": 1.833492446790588,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19529445916414262,
      "num_tokens": 36448854.0,
      "step": 16325
    },
    {
      "entropy": 5.440536022186279,
      "epoch": 1.8340540236985454,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1929825648665428,
      "num_tokens": 36460426.0,
      "step": 16330
    },
    {
      "entropy": 5.515605926513672,
      "epoch": 1.834615600606503,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.2658,
      "mean_token_accuracy": 0.1934250235557556,
      "num_tokens": 36470914.0,
      "step": 16335
    },
    {
      "entropy": 5.434169721603394,
      "epoch": 1.8351771775144607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.1673,
      "mean_token_accuracy": 0.19939337074756622,
      "num_tokens": 36480753.0,
      "step": 16340
    },
    {
      "entropy": 5.380618000030518,
      "epoch": 1.8357387544224182,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.1686,
      "mean_token_accuracy": 0.20086667835712432,
      "num_tokens": 36492858.0,
      "step": 16345
    },
    {
      "entropy": 5.355233383178711,
      "epoch": 1.8363003313303756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046707964881554005,
      "loss": 5.1304,
      "mean_token_accuracy": 0.2012440785765648,
      "num_tokens": 36503912.0,
      "step": 16350
    },
    {
      "entropy": 5.408527183532715,
      "epoch": 1.8368619082383333,
      "grad_norm": 1.125,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.2159,
      "mean_token_accuracy": 0.1951673373579979,
      "num_tokens": 36515410.0,
      "step": 16355
    },
    {
      "entropy": 5.460459232330322,
      "epoch": 1.837423485146291,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.1603,
      "mean_token_accuracy": 0.20181371420621871,
      "num_tokens": 36526310.0,
      "step": 16360
    },
    {
      "entropy": 5.425773429870605,
      "epoch": 1.8379850620542482,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046701689463761947,
      "loss": 5.1158,
      "mean_token_accuracy": 0.2011399671435356,
      "num_tokens": 36537081.0,
      "step": 16365
    },
    {
      "entropy": 5.418057584762574,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19751242697238922,
      "num_tokens": 36549419.0,
      "step": 16370
    },
    {
      "entropy": 5.3919353008270265,
      "epoch": 1.8391082158701635,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.2001,
      "mean_token_accuracy": 0.20440282970666884,
      "num_tokens": 36560842.0,
      "step": 16375
    },
    {
      "entropy": 5.434959363937378,
      "epoch": 1.839669792778121,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.175,
      "mean_token_accuracy": 0.2019817352294922,
      "num_tokens": 36571717.0,
      "step": 16380
    },
    {
      "entropy": 5.455682373046875,
      "epoch": 1.8402313696860784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.209,
      "mean_token_accuracy": 0.19455764889717103,
      "num_tokens": 36581709.0,
      "step": 16385
    },
    {
      "entropy": 5.471096420288086,
      "epoch": 1.840792946594036,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.2246,
      "mean_token_accuracy": 0.196385957300663,
      "num_tokens": 36593033.0,
      "step": 16390
    },
    {
      "entropy": 5.522565937042236,
      "epoch": 1.8413545235019937,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.3249,
      "mean_token_accuracy": 0.193872332572937,
      "num_tokens": 36603605.0,
      "step": 16395
    },
    {
      "entropy": 5.505653142929077,
      "epoch": 1.8419161004099511,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.3304,
      "mean_token_accuracy": 0.19150075018405915,
      "num_tokens": 36616219.0,
      "step": 16400
    },
    {
      "entropy": 5.552918195724487,
      "epoch": 1.8424776773179086,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19474121928215027,
      "num_tokens": 36627777.0,
      "step": 16405
    },
    {
      "entropy": 5.484918260574341,
      "epoch": 1.8430392542258662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.2341,
      "mean_token_accuracy": 0.19648625403642656,
      "num_tokens": 36638792.0,
      "step": 16410
    },
    {
      "entropy": 5.479833507537842,
      "epoch": 1.843600831133824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19887581318616868,
      "num_tokens": 36650218.0,
      "step": 16415
    },
    {
      "entropy": 5.480641794204712,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1968456357717514,
      "num_tokens": 36660987.0,
      "step": 16420
    },
    {
      "entropy": 5.427464437484741,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.1976,
      "mean_token_accuracy": 0.19036560505628586,
      "num_tokens": 36671270.0,
      "step": 16425
    },
    {
      "entropy": 5.456794357299804,
      "epoch": 1.8452855618576964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.1999,
      "mean_token_accuracy": 0.2033836618065834,
      "num_tokens": 36682484.0,
      "step": 16430
    },
    {
      "entropy": 5.519985580444336,
      "epoch": 1.845847138765654,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.2992,
      "mean_token_accuracy": 0.19730817377567292,
      "num_tokens": 36693723.0,
      "step": 16435
    },
    {
      "entropy": 5.359445381164551,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004667022987327732,
      "loss": 5.0972,
      "mean_token_accuracy": 0.2054307758808136,
      "num_tokens": 36704166.0,
      "step": 16440
    },
    {
      "entropy": 5.391143417358398,
      "epoch": 1.846970292581569,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20609794855117797,
      "num_tokens": 36714139.0,
      "step": 16445
    },
    {
      "entropy": 5.44873685836792,
      "epoch": 1.8475318694895266,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.1883,
      "mean_token_accuracy": 0.20303950160741807,
      "num_tokens": 36724273.0,
      "step": 16450
    },
    {
      "entropy": 5.462536144256592,
      "epoch": 1.8480934463974843,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.2619,
      "mean_token_accuracy": 0.1917540907859802,
      "num_tokens": 36735176.0,
      "step": 16455
    },
    {
      "entropy": 5.40107274055481,
      "epoch": 1.8486550233054415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.2029,
      "mean_token_accuracy": 0.1995821252465248,
      "num_tokens": 36746170.0,
      "step": 16460
    },
    {
      "entropy": 5.551182985305786,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.3621,
      "mean_token_accuracy": 0.191077621281147,
      "num_tokens": 36756969.0,
      "step": 16465
    },
    {
      "entropy": 5.548001861572265,
      "epoch": 1.8497781771213568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.2211,
      "mean_token_accuracy": 0.20138726979494095,
      "num_tokens": 36767892.0,
      "step": 16470
    },
    {
      "entropy": 5.452150917053222,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.2244,
      "mean_token_accuracy": 0.19823097884655,
      "num_tokens": 36778967.0,
      "step": 16475
    },
    {
      "entropy": 5.337763929367066,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004665339525502236,
      "loss": 5.0952,
      "mean_token_accuracy": 0.199875046312809,
      "num_tokens": 36789204.0,
      "step": 16480
    },
    {
      "entropy": 5.4093445301055905,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.1592,
      "mean_token_accuracy": 0.2065848022699356,
      "num_tokens": 36798990.0,
      "step": 16485
    },
    {
      "entropy": 5.511639404296875,
      "epoch": 1.852024484753187,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19570061564445496,
      "num_tokens": 36809988.0,
      "step": 16490
    },
    {
      "entropy": 5.433776426315307,
      "epoch": 1.8525860616611445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.1953,
      "mean_token_accuracy": 0.20575449764728546,
      "num_tokens": 36821579.0,
      "step": 16495
    },
    {
      "entropy": 5.534322500228882,
      "epoch": 1.853147638569102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.3331,
      "mean_token_accuracy": 0.19465134739875795,
      "num_tokens": 36833141.0,
      "step": 16500
    },
    {
      "entropy": 5.5263941287994385,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.3276,
      "mean_token_accuracy": 0.19896939396858215,
      "num_tokens": 36843347.0,
      "step": 16505
    },
    {
      "entropy": 5.539216756820679,
      "epoch": 1.8542707923850172,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.3526,
      "mean_token_accuracy": 0.18484846949577333,
      "num_tokens": 36854459.0,
      "step": 16510
    },
    {
      "entropy": 5.400459146499633,
      "epoch": 1.8548323692929747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004663863294567462,
      "loss": 5.0778,
      "mean_token_accuracy": 0.1987343817949295,
      "num_tokens": 36864761.0,
      "step": 16515
    },
    {
      "entropy": 5.580765295028686,
      "epoch": 1.8553939462009321,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.3756,
      "mean_token_accuracy": 0.19152718931436538,
      "num_tokens": 36875413.0,
      "step": 16520
    },
    {
      "entropy": 5.457789564132691,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19583977311849593,
      "num_tokens": 36885350.0,
      "step": 16525
    },
    {
      "entropy": 5.453854990005493,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.1563,
      "mean_token_accuracy": 0.2017476513981819,
      "num_tokens": 36895355.0,
      "step": 16530
    },
    {
      "entropy": 5.464024496078491,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.242,
      "mean_token_accuracy": 0.20229275822639464,
      "num_tokens": 36907454.0,
      "step": 16535
    },
    {
      "entropy": 5.442661333084106,
      "epoch": 1.8576402538327623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19481302052736282,
      "num_tokens": 36919199.0,
      "step": 16540
    },
    {
      "entropy": 5.563452243804932,
      "epoch": 1.85820183074072,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.3332,
      "mean_token_accuracy": 0.19176400154829026,
      "num_tokens": 36930499.0,
      "step": 16545
    },
    {
      "entropy": 5.501461410522461,
      "epoch": 1.8587634076486776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.2615,
      "mean_token_accuracy": 0.19176605194807053,
      "num_tokens": 36943082.0,
      "step": 16550
    },
    {
      "entropy": 5.422438716888427,
      "epoch": 1.859324984556635,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1977447509765625,
      "num_tokens": 36954591.0,
      "step": 16555
    },
    {
      "entropy": 5.397451639175415,
      "epoch": 1.8598865614645925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.2676,
      "mean_token_accuracy": 0.19991008788347245,
      "num_tokens": 36967028.0,
      "step": 16560
    },
    {
      "entropy": 5.463958406448365,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19127349704504013,
      "num_tokens": 36978716.0,
      "step": 16565
    },
    {
      "entropy": 5.4492768287658695,
      "epoch": 1.8610097152805076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.1756,
      "mean_token_accuracy": 0.19747342467308043,
      "num_tokens": 36989416.0,
      "step": 16570
    },
    {
      "entropy": 5.459396409988403,
      "epoch": 1.861571292188465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.2306,
      "mean_token_accuracy": 0.19828828424215317,
      "num_tokens": 37002197.0,
      "step": 16575
    },
    {
      "entropy": 5.610602617263794,
      "epoch": 1.8621328690964227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.3399,
      "mean_token_accuracy": 0.18506837785243987,
      "num_tokens": 37012977.0,
      "step": 16580
    },
    {
      "entropy": 5.487373018264771,
      "epoch": 1.8626944460043804,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.2507,
      "mean_token_accuracy": 0.19659968614578247,
      "num_tokens": 37024141.0,
      "step": 16585
    },
    {
      "entropy": 5.402905750274658,
      "epoch": 1.8632560229123378,
      "grad_norm": 1.125,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20519924610853196,
      "num_tokens": 37034471.0,
      "step": 16590
    },
    {
      "entropy": 5.420420789718628,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.2434,
      "mean_token_accuracy": 0.19732611179351806,
      "num_tokens": 37046375.0,
      "step": 16595
    },
    {
      "entropy": 5.470990324020386,
      "epoch": 1.864379176728253,
      "grad_norm": 1.125,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.2569,
      "mean_token_accuracy": 0.20421495139598847,
      "num_tokens": 37056617.0,
      "step": 16600
    },
    {
      "entropy": 5.387459993362427,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000466005356663438,
      "loss": 5.088,
      "mean_token_accuracy": 0.20210430324077605,
      "num_tokens": 37066477.0,
      "step": 16605
    },
    {
      "entropy": 5.425457239151001,
      "epoch": 1.865502330544168,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.2596,
      "mean_token_accuracy": 0.19388592392206191,
      "num_tokens": 37077150.0,
      "step": 16610
    },
    {
      "entropy": 5.525846862792969,
      "epoch": 1.8660639074521255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.3116,
      "mean_token_accuracy": 0.19255971312522888,
      "num_tokens": 37089961.0,
      "step": 16615
    },
    {
      "entropy": 5.5383423328399655,
      "epoch": 1.8666254843600831,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1939225673675537,
      "num_tokens": 37100779.0,
      "step": 16620
    },
    {
      "entropy": 5.440025424957275,
      "epoch": 1.8671870612680408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.1586,
      "mean_token_accuracy": 0.2000011220574379,
      "num_tokens": 37111293.0,
      "step": 16625
    },
    {
      "entropy": 5.478985166549682,
      "epoch": 1.8677486381759982,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.2768,
      "mean_token_accuracy": 0.1982982188463211,
      "num_tokens": 37123087.0,
      "step": 16630
    },
    {
      "entropy": 5.448623752593994,
      "epoch": 1.8683102150839557,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.2451,
      "mean_token_accuracy": 0.19441010802984238,
      "num_tokens": 37135227.0,
      "step": 16635
    },
    {
      "entropy": 5.439109516143799,
      "epoch": 1.8688717919919133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19959462881088258,
      "num_tokens": 37146970.0,
      "step": 16640
    },
    {
      "entropy": 5.500585651397705,
      "epoch": 1.869433368899871,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.2133,
      "mean_token_accuracy": 0.19257651418447494,
      "num_tokens": 37158144.0,
      "step": 16645
    },
    {
      "entropy": 5.508302307128906,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.3449,
      "mean_token_accuracy": 0.1924276128411293,
      "num_tokens": 37169265.0,
      "step": 16650
    },
    {
      "entropy": 5.515935468673706,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.125,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.2926,
      "mean_token_accuracy": 0.1949426531791687,
      "num_tokens": 37179186.0,
      "step": 16655
    },
    {
      "entropy": 5.492211532592774,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.222,
      "mean_token_accuracy": 0.20232529193162918,
      "num_tokens": 37188906.0,
      "step": 16660
    },
    {
      "entropy": 5.365943431854248,
      "epoch": 1.871679676531701,
      "grad_norm": 1.0,
      "learning_rate": 0.000465750279855493,
      "loss": 5.1551,
      "mean_token_accuracy": 0.20616481304168702,
      "num_tokens": 37200879.0,
      "step": 16665
    },
    {
      "entropy": 5.434335803985595,
      "epoch": 1.8722412534396584,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.1712,
      "mean_token_accuracy": 0.2032892018556595,
      "num_tokens": 37211648.0,
      "step": 16670
    },
    {
      "entropy": 5.58739857673645,
      "epoch": 1.872802830347616,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.3711,
      "mean_token_accuracy": 0.19582848697900773,
      "num_tokens": 37223164.0,
      "step": 16675
    },
    {
      "entropy": 5.509513425827026,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19520914256572724,
      "num_tokens": 37233860.0,
      "step": 16680
    },
    {
      "entropy": 5.524698543548584,
      "epoch": 1.8739259841635312,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1982952684164047,
      "num_tokens": 37245431.0,
      "step": 16685
    },
    {
      "entropy": 5.493672466278076,
      "epoch": 1.8744875610714886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.3235,
      "mean_token_accuracy": 0.19340966492891312,
      "num_tokens": 37256394.0,
      "step": 16690
    },
    {
      "entropy": 5.460845470428467,
      "epoch": 1.8750491379794463,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.2901,
      "mean_token_accuracy": 0.19576283246278764,
      "num_tokens": 37268367.0,
      "step": 16695
    },
    {
      "entropy": 5.4677431106567385,
      "epoch": 1.875610714887404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.263,
      "mean_token_accuracy": 0.1967165470123291,
      "num_tokens": 37279019.0,
      "step": 16700
    },
    {
      "entropy": 5.541481828689575,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.2057,
      "mean_token_accuracy": 0.19804907441139222,
      "num_tokens": 37289806.0,
      "step": 16705
    },
    {
      "entropy": 5.390263175964355,
      "epoch": 1.8767338687033188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.2392,
      "mean_token_accuracy": 0.19549369364976882,
      "num_tokens": 37300607.0,
      "step": 16710
    },
    {
      "entropy": 5.443153142929077,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.2956,
      "mean_token_accuracy": 0.19506974667310714,
      "num_tokens": 37311966.0,
      "step": 16715
    },
    {
      "entropy": 5.396130084991455,
      "epoch": 1.8778570225192341,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1985211431980133,
      "num_tokens": 37324459.0,
      "step": 16720
    },
    {
      "entropy": 5.469950866699219,
      "epoch": 1.8784185994271916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.2292,
      "mean_token_accuracy": 0.196366611123085,
      "num_tokens": 37335276.0,
      "step": 16725
    },
    {
      "entropy": 5.4683925151824955,
      "epoch": 1.878980176335149,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20398791283369064,
      "num_tokens": 37347274.0,
      "step": 16730
    },
    {
      "entropy": 5.460577440261841,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.2755,
      "mean_token_accuracy": 0.1909830614924431,
      "num_tokens": 37359789.0,
      "step": 16735
    },
    {
      "entropy": 5.435467672348023,
      "epoch": 1.8801033301510643,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.216,
      "mean_token_accuracy": 0.20039468854665757,
      "num_tokens": 37369851.0,
      "step": 16740
    },
    {
      "entropy": 5.545345497131348,
      "epoch": 1.8806649070590218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1927112430334091,
      "num_tokens": 37381008.0,
      "step": 16745
    },
    {
      "entropy": 5.466767311096191,
      "epoch": 1.8812264839669792,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.2072,
      "mean_token_accuracy": 0.1926458492875099,
      "num_tokens": 37392110.0,
      "step": 16750
    },
    {
      "entropy": 5.517423677444458,
      "epoch": 1.8817880608749369,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.2452,
      "mean_token_accuracy": 0.2020352378487587,
      "num_tokens": 37402995.0,
      "step": 16755
    },
    {
      "entropy": 5.487637567520141,
      "epoch": 1.8823496377828943,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.2272,
      "mean_token_accuracy": 0.2001844674348831,
      "num_tokens": 37415072.0,
      "step": 16760
    },
    {
      "entropy": 5.511497974395752,
      "epoch": 1.8829112146908518,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1877359688282013,
      "num_tokens": 37427375.0,
      "step": 16765
    },
    {
      "entropy": 5.407660722732544,
      "epoch": 1.8834727915988094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.1757,
      "mean_token_accuracy": 0.20452356338500977,
      "num_tokens": 37436829.0,
      "step": 16770
    },
    {
      "entropy": 5.523489952087402,
      "epoch": 1.884034368506767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.2843,
      "mean_token_accuracy": 0.18755460083484649,
      "num_tokens": 37447875.0,
      "step": 16775
    },
    {
      "entropy": 5.392043781280518,
      "epoch": 1.8845959454147245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.1482,
      "mean_token_accuracy": 0.19770222753286362,
      "num_tokens": 37458408.0,
      "step": 16780
    },
    {
      "entropy": 5.402614879608154,
      "epoch": 1.885157522322682,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046523750363497,
      "loss": 5.2147,
      "mean_token_accuracy": 0.19525437653064728,
      "num_tokens": 37468437.0,
      "step": 16785
    },
    {
      "entropy": 5.562240648269653,
      "epoch": 1.8857190992306396,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.3083,
      "mean_token_accuracy": 0.19734027981758118,
      "num_tokens": 37479359.0,
      "step": 16790
    },
    {
      "entropy": 5.472897243499756,
      "epoch": 1.8862806761385973,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.2345,
      "mean_token_accuracy": 0.20217179656028747,
      "num_tokens": 37491321.0,
      "step": 16795
    },
    {
      "entropy": 5.463063144683838,
      "epoch": 1.8868422530465547,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19329507052898406,
      "num_tokens": 37501828.0,
      "step": 16800
    },
    {
      "entropy": 5.570747756958008,
      "epoch": 1.8874038299545122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.386,
      "mean_token_accuracy": 0.18740827441215516,
      "num_tokens": 37514525.0,
      "step": 16805
    },
    {
      "entropy": 5.503907585144043,
      "epoch": 1.8879654068624698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.1532,
      "mean_token_accuracy": 0.19461924135684966,
      "num_tokens": 37525588.0,
      "step": 16810
    },
    {
      "entropy": 5.43377423286438,
      "epoch": 1.8885269837704275,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000465108764060676,
      "loss": 5.2276,
      "mean_token_accuracy": 0.1948937341570854,
      "num_tokens": 37537941.0,
      "step": 16815
    },
    {
      "entropy": 5.450218152999878,
      "epoch": 1.889088560678385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.2439,
      "mean_token_accuracy": 0.19229685813188552,
      "num_tokens": 37549663.0,
      "step": 16820
    },
    {
      "entropy": 5.524704837799073,
      "epoch": 1.8896501375863424,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.3005,
      "mean_token_accuracy": 0.19765808284282685,
      "num_tokens": 37559793.0,
      "step": 16825
    },
    {
      "entropy": 5.431095838546753,
      "epoch": 1.8902117144943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.1753,
      "mean_token_accuracy": 0.1964648738503456,
      "num_tokens": 37570203.0,
      "step": 16830
    },
    {
      "entropy": 5.414950895309448,
      "epoch": 1.8907732914022577,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.2703,
      "mean_token_accuracy": 0.19715822339057923,
      "num_tokens": 37581858.0,
      "step": 16835
    },
    {
      "entropy": 5.457394742965699,
      "epoch": 1.8913348683102151,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.2163,
      "mean_token_accuracy": 0.19874331653118132,
      "num_tokens": 37592311.0,
      "step": 16840
    },
    {
      "entropy": 5.516688299179077,
      "epoch": 1.8918964452181726,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.3581,
      "mean_token_accuracy": 0.18968896567821503,
      "num_tokens": 37604721.0,
      "step": 16845
    },
    {
      "entropy": 5.580503273010254,
      "epoch": 1.8924580221261302,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1888548105955124,
      "num_tokens": 37616044.0,
      "step": 16850
    },
    {
      "entropy": 5.508128023147583,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19902894347906114,
      "num_tokens": 37626501.0,
      "step": 16855
    },
    {
      "entropy": 5.539801597595215,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1941537395119667,
      "num_tokens": 37637861.0,
      "step": 16860
    },
    {
      "entropy": 5.413544797897339,
      "epoch": 1.8941427528500028,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20064548254013062,
      "num_tokens": 37649288.0,
      "step": 16865
    },
    {
      "entropy": 5.380209016799927,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20221962481737138,
      "num_tokens": 37659700.0,
      "step": 16870
    },
    {
      "entropy": 5.473927736282349,
      "epoch": 1.8952659066659179,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.2907,
      "mean_token_accuracy": 0.1938436284661293,
      "num_tokens": 37671739.0,
      "step": 16875
    },
    {
      "entropy": 5.474382638931274,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.1796,
      "mean_token_accuracy": 0.20720221996307372,
      "num_tokens": 37682250.0,
      "step": 16880
    },
    {
      "entropy": 5.454179382324218,
      "epoch": 1.896389060481833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046480752453711,
      "loss": 5.2504,
      "mean_token_accuracy": 0.20065705180168153,
      "num_tokens": 37694052.0,
      "step": 16885
    },
    {
      "entropy": 5.539553689956665,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.3349,
      "mean_token_accuracy": 0.1960662394762039,
      "num_tokens": 37704818.0,
      "step": 16890
    },
    {
      "entropy": 5.598157691955566,
      "epoch": 1.897512214297748,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.3799,
      "mean_token_accuracy": 0.18657071888446808,
      "num_tokens": 37717679.0,
      "step": 16895
    },
    {
      "entropy": 5.439261960983276,
      "epoch": 1.8980737912057055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19970800578594208,
      "num_tokens": 37728711.0,
      "step": 16900
    },
    {
      "entropy": 5.453038501739502,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.214,
      "mean_token_accuracy": 0.19857898205518723,
      "num_tokens": 37740082.0,
      "step": 16905
    },
    {
      "entropy": 5.524498796463012,
      "epoch": 1.8991969450216208,
      "grad_norm": 1.171875,
      "learning_rate": 0.000464699651826876,
      "loss": 5.3461,
      "mean_token_accuracy": 0.18871774524450302,
      "num_tokens": 37750715.0,
      "step": 16910
    },
    {
      "entropy": 5.52142186164856,
      "epoch": 1.8997585219295783,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.3137,
      "mean_token_accuracy": 0.1888110488653183,
      "num_tokens": 37763230.0,
      "step": 16915
    },
    {
      "entropy": 5.4834840297698975,
      "epoch": 1.9003200988375357,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19617471992969512,
      "num_tokens": 37773621.0,
      "step": 16920
    },
    {
      "entropy": 5.484920692443848,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19421189576387404,
      "num_tokens": 37785068.0,
      "step": 16925
    },
    {
      "entropy": 5.435989189147949,
      "epoch": 1.901443252653451,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.2612,
      "mean_token_accuracy": 0.2018831804394722,
      "num_tokens": 37796925.0,
      "step": 16930
    },
    {
      "entropy": 5.4933570384979244,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.2847,
      "mean_token_accuracy": 0.1948854923248291,
      "num_tokens": 37808291.0,
      "step": 16935
    },
    {
      "entropy": 5.4373047828674315,
      "epoch": 1.902566406469366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004645700052585267,
      "loss": 5.1193,
      "mean_token_accuracy": 0.21289151310920715,
      "num_tokens": 37819337.0,
      "step": 16940
    },
    {
      "entropy": 5.52640380859375,
      "epoch": 1.9031279833773236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.3188,
      "mean_token_accuracy": 0.19600376933813096,
      "num_tokens": 37831832.0,
      "step": 16945
    },
    {
      "entropy": 5.4921496391296385,
      "epoch": 1.903689560285281,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.2322,
      "mean_token_accuracy": 0.19636641293764115,
      "num_tokens": 37842586.0,
      "step": 16950
    },
    {
      "entropy": 5.435227298736573,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.1579,
      "mean_token_accuracy": 0.2063504546880722,
      "num_tokens": 37852198.0,
      "step": 16955
    },
    {
      "entropy": 5.446521186828614,
      "epoch": 1.9048127141011961,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.2065,
      "mean_token_accuracy": 0.1997394159436226,
      "num_tokens": 37862967.0,
      "step": 16960
    },
    {
      "entropy": 5.446404647827149,
      "epoch": 1.9053742910091538,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.3262,
      "mean_token_accuracy": 0.187962144613266,
      "num_tokens": 37874501.0,
      "step": 16965
    },
    {
      "entropy": 5.452405071258545,
      "epoch": 1.9059358679171112,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004644401413948279,
      "loss": 5.0901,
      "mean_token_accuracy": 0.20705420672893524,
      "num_tokens": 37886404.0,
      "step": 16970
    },
    {
      "entropy": 5.553581809997558,
      "epoch": 1.9064974448250687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.286,
      "mean_token_accuracy": 0.195398685336113,
      "num_tokens": 37897344.0,
      "step": 16975
    },
    {
      "entropy": 5.379161643981933,
      "epoch": 1.9070590217330263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.152,
      "mean_token_accuracy": 0.20995626151561736,
      "num_tokens": 37909226.0,
      "step": 16980
    },
    {
      "entropy": 5.484614324569702,
      "epoch": 1.907620598640984,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19569606482982635,
      "num_tokens": 37920578.0,
      "step": 16985
    },
    {
      "entropy": 5.516542768478393,
      "epoch": 1.9081821755489414,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1942470669746399,
      "num_tokens": 37931814.0,
      "step": 16990
    },
    {
      "entropy": 5.460321283340454,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19769106805324554,
      "num_tokens": 37942345.0,
      "step": 16995
    },
    {
      "entropy": 5.3960329532623295,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046431006038463653,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20145775228738785,
      "num_tokens": 37953027.0,
      "step": 17000
    },
    {
      "entropy": 5.474249935150146,
      "epoch": 1.9098669062728142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.2799,
      "mean_token_accuracy": 0.19447172284126282,
      "num_tokens": 37963290.0,
      "step": 17005
    },
    {
      "entropy": 5.499614906311035,
      "epoch": 1.9104284831807716,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.3712,
      "mean_token_accuracy": 0.18781432658433914,
      "num_tokens": 37974424.0,
      "step": 17010
    },
    {
      "entropy": 5.598958158493042,
      "epoch": 1.910990060088729,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.365,
      "mean_token_accuracy": 0.19186062961816788,
      "num_tokens": 37987993.0,
      "step": 17015
    },
    {
      "entropy": 5.462058448791504,
      "epoch": 1.9115516369966867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.1461,
      "mean_token_accuracy": 0.2007363885641098,
      "num_tokens": 37999340.0,
      "step": 17020
    },
    {
      "entropy": 5.416390752792358,
      "epoch": 1.9121132139046444,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.212,
      "mean_token_accuracy": 0.1935009092092514,
      "num_tokens": 38009737.0,
      "step": 17025
    },
    {
      "entropy": 5.463556098937988,
      "epoch": 1.9126747908126018,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.209,
      "mean_token_accuracy": 0.19952737987041474,
      "num_tokens": 38021283.0,
      "step": 17030
    },
    {
      "entropy": 5.575936651229858,
      "epoch": 1.9132363677205593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19020410180091857,
      "num_tokens": 38034044.0,
      "step": 17035
    },
    {
      "entropy": 5.480965566635132,
      "epoch": 1.913797944628517,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20085353553295135,
      "num_tokens": 38044585.0,
      "step": 17040
    },
    {
      "entropy": 5.405561113357544,
      "epoch": 1.9143595215364744,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.1526,
      "mean_token_accuracy": 0.1980004131793976,
      "num_tokens": 38055771.0,
      "step": 17045
    },
    {
      "entropy": 5.435069799423218,
      "epoch": 1.9149210984444318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.2282,
      "mean_token_accuracy": 0.1987907186150551,
      "num_tokens": 38066344.0,
      "step": 17050
    },
    {
      "entropy": 5.508805561065674,
      "epoch": 1.9154826753523895,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.1705,
      "mean_token_accuracy": 0.20338296592235566,
      "num_tokens": 38077785.0,
      "step": 17055
    },
    {
      "entropy": 5.435651159286499,
      "epoch": 1.9160442522603471,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19283691346645354,
      "num_tokens": 38089023.0,
      "step": 17060
    },
    {
      "entropy": 5.468222332000733,
      "epoch": 1.9166058291683046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046402747397356666,
      "loss": 5.1455,
      "mean_token_accuracy": 0.2024591252207756,
      "num_tokens": 38099428.0,
      "step": 17065
    },
    {
      "entropy": 5.433711338043213,
      "epoch": 1.917167406076262,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.1586,
      "mean_token_accuracy": 0.19864461421966553,
      "num_tokens": 38110757.0,
      "step": 17070
    },
    {
      "entropy": 5.478124761581421,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19660273492336272,
      "num_tokens": 38121442.0,
      "step": 17075
    },
    {
      "entropy": 5.4433839321136475,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.2251,
      "mean_token_accuracy": 0.20031900107860565,
      "num_tokens": 38132917.0,
      "step": 17080
    },
    {
      "entropy": 5.401716613769532,
      "epoch": 1.9188521368001348,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19604706019163132,
      "num_tokens": 38143188.0,
      "step": 17085
    },
    {
      "entropy": 5.565263891220093,
      "epoch": 1.9194137137080922,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.3361,
      "mean_token_accuracy": 0.19754895865917205,
      "num_tokens": 38154300.0,
      "step": 17090
    },
    {
      "entropy": 5.53109393119812,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.3315,
      "mean_token_accuracy": 0.19322279542684556,
      "num_tokens": 38165697.0,
      "step": 17095
    },
    {
      "entropy": 5.479623174667358,
      "epoch": 1.9205368675240075,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.2268,
      "mean_token_accuracy": 0.1981937736272812,
      "num_tokens": 38176601.0,
      "step": 17100
    },
    {
      "entropy": 5.494045972824097,
      "epoch": 1.921098444431965,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.3215,
      "mean_token_accuracy": 0.19902921169996263,
      "num_tokens": 38188334.0,
      "step": 17105
    },
    {
      "entropy": 5.507667446136475,
      "epoch": 1.9216600213399224,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.2507,
      "mean_token_accuracy": 0.19804709404706955,
      "num_tokens": 38199483.0,
      "step": 17110
    },
    {
      "entropy": 5.524251270294189,
      "epoch": 1.92222159824788,
      "grad_norm": 1.125,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.3363,
      "mean_token_accuracy": 0.19040213227272035,
      "num_tokens": 38211391.0,
      "step": 17115
    },
    {
      "entropy": 5.4890039443969725,
      "epoch": 1.9227831751558377,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1998496562242508,
      "num_tokens": 38223415.0,
      "step": 17120
    },
    {
      "entropy": 5.4596755027771,
      "epoch": 1.9233447520637952,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.1803,
      "mean_token_accuracy": 0.20113856196403504,
      "num_tokens": 38235093.0,
      "step": 17125
    },
    {
      "entropy": 5.463783931732178,
      "epoch": 1.9239063289717526,
      "grad_norm": 1.125,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.2996,
      "mean_token_accuracy": 0.187471242249012,
      "num_tokens": 38247192.0,
      "step": 17130
    },
    {
      "entropy": 5.4878153800964355,
      "epoch": 1.9244679058797103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.282,
      "mean_token_accuracy": 0.19151819348335267,
      "num_tokens": 38257965.0,
      "step": 17135
    },
    {
      "entropy": 5.580102920532227,
      "epoch": 1.9250294827876677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000463700148899063,
      "loss": 5.3683,
      "mean_token_accuracy": 0.18687620460987092,
      "num_tokens": 38268529.0,
      "step": 17140
    },
    {
      "entropy": 5.463663864135742,
      "epoch": 1.9255910596956252,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1919669434428215,
      "num_tokens": 38279480.0,
      "step": 17145
    },
    {
      "entropy": 5.488632726669311,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.3035,
      "mean_token_accuracy": 0.1978561446070671,
      "num_tokens": 38291670.0,
      "step": 17150
    },
    {
      "entropy": 5.405059671401977,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046363452158532857,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21451079249382018,
      "num_tokens": 38303073.0,
      "step": 17155
    },
    {
      "entropy": 5.4871845722198485,
      "epoch": 1.927275790419498,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.2504,
      "mean_token_accuracy": 0.19328262209892272,
      "num_tokens": 38313915.0,
      "step": 17160
    },
    {
      "entropy": 5.464030742645264,
      "epoch": 1.9278373673274554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.2125,
      "mean_token_accuracy": 0.1906324401497841,
      "num_tokens": 38324845.0,
      "step": 17165
    },
    {
      "entropy": 5.467322254180909,
      "epoch": 1.928398944235413,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.249,
      "mean_token_accuracy": 0.1999796837568283,
      "num_tokens": 38335096.0,
      "step": 17170
    },
    {
      "entropy": 5.500621318817139,
      "epoch": 1.9289605211433707,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.2263,
      "mean_token_accuracy": 0.19897073209285737,
      "num_tokens": 38345559.0,
      "step": 17175
    },
    {
      "entropy": 5.422845697402954,
      "epoch": 1.9295220980513281,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.1452,
      "mean_token_accuracy": 0.21408585608005523,
      "num_tokens": 38356318.0,
      "step": 17180
    },
    {
      "entropy": 5.420642328262329,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1991479814052582,
      "num_tokens": 38367246.0,
      "step": 17185
    },
    {
      "entropy": 5.432358312606811,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19524916559457778,
      "num_tokens": 38378767.0,
      "step": 17190
    },
    {
      "entropy": 5.447893142700195,
      "epoch": 1.931206828775201,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.1825,
      "mean_token_accuracy": 0.1989078029990196,
      "num_tokens": 38390414.0,
      "step": 17195
    },
    {
      "entropy": 5.471879339218139,
      "epoch": 1.9317684056831583,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20094675570726395,
      "num_tokens": 38401569.0,
      "step": 17200
    },
    {
      "entropy": 5.577297782897949,
      "epoch": 1.9323299825911158,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.3271,
      "mean_token_accuracy": 0.19419706463813782,
      "num_tokens": 38413561.0,
      "step": 17205
    },
    {
      "entropy": 5.494162225723267,
      "epoch": 1.9328915594990734,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.2541,
      "mean_token_accuracy": 0.20286345183849336,
      "num_tokens": 38423965.0,
      "step": 17210
    },
    {
      "entropy": 5.408056259155273,
      "epoch": 1.933453136407031,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19720540642738343,
      "num_tokens": 38434158.0,
      "step": 17215
    },
    {
      "entropy": 5.422565126419068,
      "epoch": 1.9340147133149885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.2198,
      "mean_token_accuracy": 0.19582920521497726,
      "num_tokens": 38445449.0,
      "step": 17220
    },
    {
      "entropy": 5.520533037185669,
      "epoch": 1.934576290222946,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.3184,
      "mean_token_accuracy": 0.19835469573736192,
      "num_tokens": 38457601.0,
      "step": 17225
    },
    {
      "entropy": 5.488399171829224,
      "epoch": 1.9351378671309036,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20097000598907472,
      "num_tokens": 38468631.0,
      "step": 17230
    },
    {
      "entropy": 5.405942296981811,
      "epoch": 1.935699444038861,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19931226074695588,
      "num_tokens": 38479567.0,
      "step": 17235
    },
    {
      "entropy": 5.3931702136993405,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.1875,
      "learning_rate": 0.000463261612974268,
      "loss": 5.2171,
      "mean_token_accuracy": 0.2005762130022049,
      "num_tokens": 38489934.0,
      "step": 17240
    },
    {
      "entropy": 5.586167812347412,
      "epoch": 1.9368225978547762,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18913425058126448,
      "num_tokens": 38502214.0,
      "step": 17245
    },
    {
      "entropy": 5.560330724716186,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.227,
      "mean_token_accuracy": 0.19509009569883345,
      "num_tokens": 38513915.0,
      "step": 17250
    },
    {
      "entropy": 5.468756866455078,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.1914,
      "mean_token_accuracy": 0.2018933430314064,
      "num_tokens": 38525039.0,
      "step": 17255
    },
    {
      "entropy": 5.36847620010376,
      "epoch": 1.9385073285786487,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.1809,
      "mean_token_accuracy": 0.2076302394270897,
      "num_tokens": 38537264.0,
      "step": 17260
    },
    {
      "entropy": 5.481237077713013,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.2738,
      "mean_token_accuracy": 0.1979181796312332,
      "num_tokens": 38547926.0,
      "step": 17265
    },
    {
      "entropy": 5.478799676895141,
      "epoch": 1.939630482394564,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19916255921125411,
      "num_tokens": 38558745.0,
      "step": 17270
    },
    {
      "entropy": 5.439381742477417,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.15625,
      "learning_rate": 0.000463107558650873,
      "loss": 5.1869,
      "mean_token_accuracy": 0.2005079433321953,
      "num_tokens": 38568190.0,
      "step": 17275
    },
    {
      "entropy": 5.512634944915772,
      "epoch": 1.940753636210479,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.2943,
      "mean_token_accuracy": 0.19765460789203643,
      "num_tokens": 38578932.0,
      "step": 17280
    },
    {
      "entropy": 5.4692127227783205,
      "epoch": 1.9413152131184366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.249,
      "mean_token_accuracy": 0.19438978284597397,
      "num_tokens": 38590501.0,
      "step": 17285
    },
    {
      "entropy": 5.441666126251221,
      "epoch": 1.9418767900263942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20501628518104553,
      "num_tokens": 38600498.0,
      "step": 17290
    },
    {
      "entropy": 5.403835201263428,
      "epoch": 1.9424383669343517,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000463019395826327,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19831580221652984,
      "num_tokens": 38611991.0,
      "step": 17295
    },
    {
      "entropy": 5.408307123184204,
      "epoch": 1.9429999438423091,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.1835,
      "mean_token_accuracy": 0.2026299938559532,
      "num_tokens": 38622284.0,
      "step": 17300
    },
    {
      "entropy": 5.498871850967407,
      "epoch": 1.9435615207502668,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.2866,
      "mean_token_accuracy": 0.19303154647350312,
      "num_tokens": 38633777.0,
      "step": 17305
    },
    {
      "entropy": 5.523743343353272,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.2462,
      "mean_token_accuracy": 0.1925231248140335,
      "num_tokens": 38644360.0,
      "step": 17310
    },
    {
      "entropy": 5.442286348342895,
      "epoch": 1.9446846745661819,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.246,
      "mean_token_accuracy": 0.19642017930746078,
      "num_tokens": 38655242.0,
      "step": 17315
    },
    {
      "entropy": 5.419105339050293,
      "epoch": 1.9452462514741393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046290905760218193,
      "loss": 5.1301,
      "mean_token_accuracy": 0.2068805143237114,
      "num_tokens": 38667792.0,
      "step": 17320
    },
    {
      "entropy": 5.368481588363648,
      "epoch": 1.945807828382097,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046288697200524964,
      "loss": 5.1216,
      "mean_token_accuracy": 0.2080197423696518,
      "num_tokens": 38677735.0,
      "step": 17325
    },
    {
      "entropy": 5.484059238433838,
      "epoch": 1.9463694052900544,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19770711064338684,
      "num_tokens": 38689556.0,
      "step": 17330
    },
    {
      "entropy": 5.452941513061523,
      "epoch": 1.9469309821980119,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20585621297359466,
      "num_tokens": 38700556.0,
      "step": 17335
    },
    {
      "entropy": 5.426540517807007,
      "epoch": 1.9474925591059695,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19898008704185485,
      "num_tokens": 38711019.0,
      "step": 17340
    },
    {
      "entropy": 5.406993341445923,
      "epoch": 1.9480541360139272,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.214,
      "mean_token_accuracy": 0.198685884475708,
      "num_tokens": 38722228.0,
      "step": 17345
    },
    {
      "entropy": 5.408995151519775,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000462776454298748,
      "loss": 5.1257,
      "mean_token_accuracy": 0.21163479685783387,
      "num_tokens": 38732183.0,
      "step": 17350
    },
    {
      "entropy": 5.396948766708374,
      "epoch": 1.949177289829842,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.1217,
      "mean_token_accuracy": 0.20836301594972612,
      "num_tokens": 38743659.0,
      "step": 17355
    },
    {
      "entropy": 5.488188123703003,
      "epoch": 1.9497388667377997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19503349363803862,
      "num_tokens": 38755032.0,
      "step": 17360
    },
    {
      "entropy": 5.464019870758056,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.1879,
      "mean_token_accuracy": 0.20382087528705597,
      "num_tokens": 38766126.0,
      "step": 17365
    },
    {
      "entropy": 5.500922536849975,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.3234,
      "mean_token_accuracy": 0.19810323566198348,
      "num_tokens": 38776921.0,
      "step": 17370
    },
    {
      "entropy": 5.442402410507202,
      "epoch": 1.9514235974616723,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.185,
      "mean_token_accuracy": 0.1986000120639801,
      "num_tokens": 38788884.0,
      "step": 17375
    },
    {
      "entropy": 5.562372922897339,
      "epoch": 1.95198517436963,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1904572382569313,
      "num_tokens": 38800505.0,
      "step": 17380
    },
    {
      "entropy": 5.472084617614746,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.1463,
      "mean_token_accuracy": 0.19802062064409257,
      "num_tokens": 38811132.0,
      "step": 17385
    },
    {
      "entropy": 5.521532678604126,
      "epoch": 1.953108328185545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.2385,
      "mean_token_accuracy": 0.20619104504585267,
      "num_tokens": 38821349.0,
      "step": 17390
    },
    {
      "entropy": 5.486487293243409,
      "epoch": 1.9536699050935025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19598303586244584,
      "num_tokens": 38832970.0,
      "step": 17395
    },
    {
      "entropy": 5.427934741973877,
      "epoch": 1.9542314820014601,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.2211,
      "mean_token_accuracy": 0.20310211777687073,
      "num_tokens": 38844220.0,
      "step": 17400
    },
    {
      "entropy": 5.4370650291442875,
      "epoch": 1.9547930589094178,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.207,
      "mean_token_accuracy": 0.19337136745452882,
      "num_tokens": 38854848.0,
      "step": 17405
    },
    {
      "entropy": 5.51825213432312,
      "epoch": 1.9553546358173752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.3288,
      "mean_token_accuracy": 0.19359671771526338,
      "num_tokens": 38864827.0,
      "step": 17410
    },
    {
      "entropy": 5.483699893951416,
      "epoch": 1.9559162127253327,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20050750374794007,
      "num_tokens": 38875914.0,
      "step": 17415
    },
    {
      "entropy": 5.3950018882751465,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.1512,
      "mean_token_accuracy": 0.2009650155901909,
      "num_tokens": 38886565.0,
      "step": 17420
    },
    {
      "entropy": 5.5279758930206295,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.125,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.423,
      "mean_token_accuracy": 0.18694986402988434,
      "num_tokens": 38897319.0,
      "step": 17425
    },
    {
      "entropy": 5.459879922866821,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000462421793621278,
      "loss": 5.1626,
      "mean_token_accuracy": 0.2020094558596611,
      "num_tokens": 38908824.0,
      "step": 17430
    },
    {
      "entropy": 5.4830629348754885,
      "epoch": 1.9581625203571629,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462399576574456,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19632184207439424,
      "num_tokens": 38921800.0,
      "step": 17435
    },
    {
      "entropy": 5.473407459259033,
      "epoch": 1.9587240972651205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.2244,
      "mean_token_accuracy": 0.19653329402208328,
      "num_tokens": 38932284.0,
      "step": 17440
    },
    {
      "entropy": 5.449905920028686,
      "epoch": 1.959285674173078,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.233,
      "mean_token_accuracy": 0.20368038564920427,
      "num_tokens": 38943632.0,
      "step": 17445
    },
    {
      "entropy": 5.49415397644043,
      "epoch": 1.9598472510810354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.1703,
      "mean_token_accuracy": 0.20023696422576903,
      "num_tokens": 38954701.0,
      "step": 17450
    },
    {
      "entropy": 5.592452764511108,
      "epoch": 1.960408827988993,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19546638876199723,
      "num_tokens": 38965625.0,
      "step": 17455
    },
    {
      "entropy": 5.485908651351929,
      "epoch": 1.9609704048969507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19822994321584703,
      "num_tokens": 38976989.0,
      "step": 17460
    },
    {
      "entropy": 5.394016981124878,
      "epoch": 1.9615319818049082,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.255,
      "mean_token_accuracy": 0.20444424897432328,
      "num_tokens": 38987661.0,
      "step": 17465
    },
    {
      "entropy": 5.405404329299927,
      "epoch": 1.9620935587128656,
      "grad_norm": 1.125,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.2201,
      "mean_token_accuracy": 0.20042105317115783,
      "num_tokens": 38998014.0,
      "step": 17470
    },
    {
      "entropy": 5.433750772476197,
      "epoch": 1.9626551356208233,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20771558284759523,
      "num_tokens": 39008843.0,
      "step": 17475
    },
    {
      "entropy": 5.498402547836304,
      "epoch": 1.963216712528781,
      "grad_norm": 1.125,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.2869,
      "mean_token_accuracy": 0.19607864320278168,
      "num_tokens": 39020173.0,
      "step": 17480
    },
    {
      "entropy": 5.47962384223938,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.3027,
      "mean_token_accuracy": 0.18916794508695603,
      "num_tokens": 39033087.0,
      "step": 17485
    },
    {
      "entropy": 5.548334455490112,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19969939440488815,
      "num_tokens": 39045252.0,
      "step": 17490
    },
    {
      "entropy": 5.534248733520508,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.3803,
      "mean_token_accuracy": 0.19075103104114532,
      "num_tokens": 39056755.0,
      "step": 17495
    },
    {
      "entropy": 5.449520206451416,
      "epoch": 1.9654630201606111,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.1537,
      "mean_token_accuracy": 0.20327087342739106,
      "num_tokens": 39066993.0,
      "step": 17500
    },
    {
      "entropy": 5.457622623443603,
      "epoch": 1.9660245970685686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.2252,
      "mean_token_accuracy": 0.20211263000965118,
      "num_tokens": 39078671.0,
      "step": 17505
    },
    {
      "entropy": 5.496820974349975,
      "epoch": 1.966586173976526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.2195,
      "mean_token_accuracy": 0.1970143586397171,
      "num_tokens": 39089563.0,
      "step": 17510
    },
    {
      "entropy": 5.543226623535157,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.2716,
      "mean_token_accuracy": 0.1950434923171997,
      "num_tokens": 39099837.0,
      "step": 17515
    },
    {
      "entropy": 5.532700252532959,
      "epoch": 1.9677093277924411,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.2697,
      "mean_token_accuracy": 0.19125046133995055,
      "num_tokens": 39112061.0,
      "step": 17520
    },
    {
      "entropy": 5.4774268627166744,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.125,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.2143,
      "mean_token_accuracy": 0.2042147174477577,
      "num_tokens": 39122777.0,
      "step": 17525
    },
    {
      "entropy": 5.415557765960694,
      "epoch": 1.9688324816083562,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.1719,
      "mean_token_accuracy": 0.20116323083639145,
      "num_tokens": 39134827.0,
      "step": 17530
    },
    {
      "entropy": 5.439550447463989,
      "epoch": 1.9693940585163139,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19726840108633042,
      "num_tokens": 39144352.0,
      "step": 17535
    },
    {
      "entropy": 5.430553293228149,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.1568,
      "mean_token_accuracy": 0.2000689134001732,
      "num_tokens": 39155610.0,
      "step": 17540
    },
    {
      "entropy": 5.403349590301514,
      "epoch": 1.9705172123322288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.1961,
      "mean_token_accuracy": 0.20000596195459366,
      "num_tokens": 39167058.0,
      "step": 17545
    },
    {
      "entropy": 5.4965376377105715,
      "epoch": 1.9710787892401864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.3163,
      "mean_token_accuracy": 0.19153826385736467,
      "num_tokens": 39178797.0,
      "step": 17550
    },
    {
      "entropy": 5.462180423736572,
      "epoch": 1.971640366148144,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.1562,
      "mean_token_accuracy": 0.2074221134185791,
      "num_tokens": 39188980.0,
      "step": 17555
    },
    {
      "entropy": 5.437407207489014,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1956959381699562,
      "num_tokens": 39200058.0,
      "step": 17560
    },
    {
      "entropy": 5.433461666107178,
      "epoch": 1.972763519964059,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.2446,
      "mean_token_accuracy": 0.19840811789035798,
      "num_tokens": 39211369.0,
      "step": 17565
    },
    {
      "entropy": 5.456657552719117,
      "epoch": 1.9733250968720166,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.281,
      "mean_token_accuracy": 0.19754408895969391,
      "num_tokens": 39222438.0,
      "step": 17570
    },
    {
      "entropy": 5.495665073394775,
      "epoch": 1.9738866737799743,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.2572,
      "mean_token_accuracy": 0.19137495756149292,
      "num_tokens": 39233047.0,
      "step": 17575
    },
    {
      "entropy": 5.478396368026734,
      "epoch": 1.9744482506879317,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.1886,
      "mean_token_accuracy": 0.19990155398845671,
      "num_tokens": 39244547.0,
      "step": 17580
    },
    {
      "entropy": 5.468042230606079,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1995212808251381,
      "num_tokens": 39255549.0,
      "step": 17585
    },
    {
      "entropy": 5.383805847167968,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000461707892069342,
      "loss": 5.1491,
      "mean_token_accuracy": 0.201088647544384,
      "num_tokens": 39266527.0,
      "step": 17590
    },
    {
      "entropy": 5.43171238899231,
      "epoch": 1.9761329814118045,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.2509,
      "mean_token_accuracy": 0.19938703626394272,
      "num_tokens": 39278013.0,
      "step": 17595
    },
    {
      "entropy": 5.475345325469971,
      "epoch": 1.976694558319762,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.2134,
      "mean_token_accuracy": 0.20027145445346833,
      "num_tokens": 39289349.0,
      "step": 17600
    },
    {
      "entropy": 5.415403127670288,
      "epoch": 1.9772561352277194,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.1983,
      "mean_token_accuracy": 0.2033061057329178,
      "num_tokens": 39301213.0,
      "step": 17605
    },
    {
      "entropy": 5.380880069732666,
      "epoch": 1.977817712135677,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004616182258676765,
      "loss": 5.0803,
      "mean_token_accuracy": 0.20803125947713852,
      "num_tokens": 39310960.0,
      "step": 17610
    },
    {
      "entropy": 5.478777551651001,
      "epoch": 1.9783792890436345,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.2758,
      "mean_token_accuracy": 0.19571359753608703,
      "num_tokens": 39322603.0,
      "step": 17615
    },
    {
      "entropy": 5.492423868179321,
      "epoch": 1.978940865951592,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.2037,
      "mean_token_accuracy": 0.2020836740732193,
      "num_tokens": 39333811.0,
      "step": 17620
    },
    {
      "entropy": 5.536708736419678,
      "epoch": 1.9795024428595496,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.2468,
      "mean_token_accuracy": 0.196746826171875,
      "num_tokens": 39345218.0,
      "step": 17625
    },
    {
      "entropy": 5.39954400062561,
      "epoch": 1.9800640197675072,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046152846459406765,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20189063251018524,
      "num_tokens": 39356777.0,
      "step": 17630
    },
    {
      "entropy": 5.374990081787109,
      "epoch": 1.9806255966754647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.1884,
      "mean_token_accuracy": 0.2019709512591362,
      "num_tokens": 39368288.0,
      "step": 17635
    },
    {
      "entropy": 5.507363176345825,
      "epoch": 1.9811871735834221,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.2031,
      "mean_token_accuracy": 0.2046121507883072,
      "num_tokens": 39379398.0,
      "step": 17640
    },
    {
      "entropy": 5.41421799659729,
      "epoch": 1.9817487504913798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.206,
      "mean_token_accuracy": 0.19975281357765198,
      "num_tokens": 39391129.0,
      "step": 17645
    },
    {
      "entropy": 5.501135921478271,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.125,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1867342248558998,
      "num_tokens": 39402442.0,
      "step": 17650
    },
    {
      "entropy": 5.589143705368042,
      "epoch": 1.9828719043072949,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19941292554140091,
      "num_tokens": 39413044.0,
      "step": 17655
    },
    {
      "entropy": 5.486483812332153,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.2048,
      "mean_token_accuracy": 0.2027856007218361,
      "num_tokens": 39423739.0,
      "step": 17660
    },
    {
      "entropy": 5.464673566818237,
      "epoch": 1.98399505812321,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000461371153735959,
      "loss": 5.3457,
      "mean_token_accuracy": 0.19283718168735503,
      "num_tokens": 39436199.0,
      "step": 17665
    },
    {
      "entropy": 5.506896734237671,
      "epoch": 1.9845566350311676,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.2234,
      "mean_token_accuracy": 0.202064910531044,
      "num_tokens": 39449302.0,
      "step": 17670
    },
    {
      "entropy": 5.503790330886841,
      "epoch": 1.985118211939125,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.2806,
      "mean_token_accuracy": 0.1987873300909996,
      "num_tokens": 39460723.0,
      "step": 17675
    },
    {
      "entropy": 5.412898874282837,
      "epoch": 1.9856797888470825,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.1852,
      "mean_token_accuracy": 0.20379892736673355,
      "num_tokens": 39471251.0,
      "step": 17680
    },
    {
      "entropy": 5.48502836227417,
      "epoch": 1.9862413657550402,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20433451384305953,
      "num_tokens": 39482613.0,
      "step": 17685
    },
    {
      "entropy": 5.436874055862427,
      "epoch": 1.9868029426629978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.2146,
      "mean_token_accuracy": 0.20829022228717803,
      "num_tokens": 39494626.0,
      "step": 17690
    },
    {
      "entropy": 5.380635070800781,
      "epoch": 1.9873645195709553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1982518494129181,
      "num_tokens": 39505302.0,
      "step": 17695
    },
    {
      "entropy": 5.458251237869263,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046121355210552626,
      "loss": 5.1284,
      "mean_token_accuracy": 0.19946794509887694,
      "num_tokens": 39515854.0,
      "step": 17700
    },
    {
      "entropy": 5.505066871643066,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.2991,
      "mean_token_accuracy": 0.19136724919080733,
      "num_tokens": 39527504.0,
      "step": 17705
    },
    {
      "entropy": 5.495616340637207,
      "epoch": 1.989049250294828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.2007,
      "mean_token_accuracy": 0.19955624788999557,
      "num_tokens": 39538885.0,
      "step": 17710
    },
    {
      "entropy": 5.456606817245484,
      "epoch": 1.9896108272027853,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.1857,
      "mean_token_accuracy": 0.206529837846756,
      "num_tokens": 39549401.0,
      "step": 17715
    },
    {
      "entropy": 5.42933669090271,
      "epoch": 1.990172404110743,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.2423,
      "mean_token_accuracy": 0.19985499680042268,
      "num_tokens": 39560814.0,
      "step": 17720
    },
    {
      "entropy": 5.435828971862793,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.2092,
      "mean_token_accuracy": 0.1969259038567543,
      "num_tokens": 39571490.0,
      "step": 17725
    },
    {
      "entropy": 5.377770185470581,
      "epoch": 1.991295557926658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.1284,
      "mean_token_accuracy": 0.20340136736631392,
      "num_tokens": 39582815.0,
      "step": 17730
    },
    {
      "entropy": 5.419006204605102,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.2253,
      "mean_token_accuracy": 0.200182768702507,
      "num_tokens": 39593712.0,
      "step": 17735
    },
    {
      "entropy": 5.492513036727905,
      "epoch": 1.9924187117425731,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20106678158044816,
      "num_tokens": 39604504.0,
      "step": 17740
    },
    {
      "entropy": 5.4835577487945555,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20139239579439164,
      "num_tokens": 39614971.0,
      "step": 17745
    },
    {
      "entropy": 5.443051433563232,
      "epoch": 1.9935418655584882,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000460987903005792,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20598647743463516,
      "num_tokens": 39626647.0,
      "step": 17750
    },
    {
      "entropy": 5.48618369102478,
      "epoch": 1.9941034424664457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19547720700502397,
      "num_tokens": 39636432.0,
      "step": 17755
    },
    {
      "entropy": 5.509604740142822,
      "epoch": 1.9946650193744033,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.242,
      "mean_token_accuracy": 0.19769977629184723,
      "num_tokens": 39647055.0,
      "step": 17760
    },
    {
      "entropy": 5.4818892002105715,
      "epoch": 1.995226596282361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20348577201366425,
      "num_tokens": 39657891.0,
      "step": 17765
    },
    {
      "entropy": 5.336728525161743,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004608974775116888,
      "loss": 5.0962,
      "mean_token_accuracy": 0.2095554456114769,
      "num_tokens": 39668023.0,
      "step": 17770
    },
    {
      "entropy": 5.423732805252075,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.213,
      "mean_token_accuracy": 0.20259709507226945,
      "num_tokens": 39678398.0,
      "step": 17775
    },
    {
      "entropy": 5.429116821289062,
      "epoch": 1.9969113270062335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19839489758014678,
      "num_tokens": 39690619.0,
      "step": 17780
    },
    {
      "entropy": 5.409654188156128,
      "epoch": 1.9974729039141912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608295962385911,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20759865939617156,
      "num_tokens": 39701690.0,
      "step": 17785
    },
    {
      "entropy": 5.460873937606811,
      "epoch": 1.9980344808221486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.288,
      "mean_token_accuracy": 0.1898223027586937,
      "num_tokens": 39713878.0,
      "step": 17790
    },
    {
      "entropy": 5.498700046539307,
      "epoch": 1.998596057730106,
      "grad_norm": 1.046875,
      "learning_rate": 0.000460784312470893,
      "loss": 5.2149,
      "mean_token_accuracy": 0.20086885690689088,
      "num_tokens": 39725164.0,
      "step": 17795
    },
    {
      "entropy": 5.446836137771607,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.1486,
      "mean_token_accuracy": 0.197618767619133,
      "num_tokens": 39735354.0,
      "step": 17800
    },
    {
      "entropy": 5.5237993717193605,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000460739005041417,
      "loss": 5.2979,
      "mean_token_accuracy": 0.19139618277549744,
      "num_tokens": 39746618.0,
      "step": 17805
    },
    {
      "entropy": 5.478542115953234,
      "epoch": 2.000224630763183,
      "grad_norm": 1.125,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.2336,
      "mean_token_accuracy": 0.1988481299744712,
      "num_tokens": 39755697.0,
      "step": 17810
    },
    {
      "entropy": 5.375217008590698,
      "epoch": 2.0007862076711405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19753584563732146,
      "num_tokens": 39766311.0,
      "step": 17815
    },
    {
      "entropy": 5.493737888336182,
      "epoch": 2.001347784579098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.2503,
      "mean_token_accuracy": 0.19518130272626877,
      "num_tokens": 39777504.0,
      "step": 17820
    },
    {
      "entropy": 5.506360960006714,
      "epoch": 2.001909361487056,
      "grad_norm": 1.125,
      "learning_rate": 0.00046064831922021534,
      "loss": 5.1152,
      "mean_token_accuracy": 0.1954638808965683,
      "num_tokens": 39789190.0,
      "step": 17825
    },
    {
      "entropy": 5.456617164611816,
      "epoch": 2.002470938395013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004606256329853235,
      "loss": 5.1164,
      "mean_token_accuracy": 0.19629822969436644,
      "num_tokens": 39801534.0,
      "step": 17830
    },
    {
      "entropy": 5.480358552932739,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.1745,
      "mean_token_accuracy": 0.19565867334604264,
      "num_tokens": 39812353.0,
      "step": 17835
    },
    {
      "entropy": 5.422595739364624,
      "epoch": 2.0035940922109283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004605802427850861,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2053135707974434,
      "num_tokens": 39823313.0,
      "step": 17840
    },
    {
      "entropy": 5.358608102798462,
      "epoch": 2.004155669118886,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046055753882118564,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21111321151256562,
      "num_tokens": 39833966.0,
      "step": 17845
    },
    {
      "entropy": 5.429851007461548,
      "epoch": 2.004717246026843,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.1694,
      "mean_token_accuracy": 0.20227143168449402,
      "num_tokens": 39843963.0,
      "step": 17850
    },
    {
      "entropy": 5.4828473091125485,
      "epoch": 2.005278822934801,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.1418,
      "mean_token_accuracy": 0.20275248140096663,
      "num_tokens": 39855237.0,
      "step": 17855
    },
    {
      "entropy": 5.473034334182739,
      "epoch": 2.0058403998427585,
      "grad_norm": 1.0,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.2772,
      "mean_token_accuracy": 0.1957951307296753,
      "num_tokens": 39867270.0,
      "step": 17860
    },
    {
      "entropy": 5.504860591888428,
      "epoch": 2.006401976750716,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.1878,
      "mean_token_accuracy": 0.20437454134225846,
      "num_tokens": 39878103.0,
      "step": 17865
    },
    {
      "entropy": 5.427053928375244,
      "epoch": 2.0069635536586734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000460443930392788,
      "loss": 5.1095,
      "mean_token_accuracy": 0.20045692771673201,
      "num_tokens": 39888945.0,
      "step": 17870
    },
    {
      "entropy": 5.5109058856964115,
      "epoch": 2.007525130566631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19451324194669722,
      "num_tokens": 39899848.0,
      "step": 17875
    },
    {
      "entropy": 5.465785312652588,
      "epoch": 2.0080867074745887,
      "grad_norm": 1.125,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19846513122320175,
      "num_tokens": 39911824.0,
      "step": 17880
    },
    {
      "entropy": 5.437666034698486,
      "epoch": 2.0086482843825464,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004603756944746813,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20339842289686202,
      "num_tokens": 39923257.0,
      "step": 17885
    },
    {
      "entropy": 5.376050567626953,
      "epoch": 2.0092098612905036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046035293736281185,
      "loss": 5.0813,
      "mean_token_accuracy": 0.20711201876401902,
      "num_tokens": 39935312.0,
      "step": 17890
    },
    {
      "entropy": 5.380534362792969,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004603301743492331,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20278966277837754,
      "num_tokens": 39946468.0,
      "step": 17895
    },
    {
      "entropy": 5.292778491973877,
      "epoch": 2.010333015106419,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004603074054346696,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21465659737586976,
      "num_tokens": 39957517.0,
      "step": 17900
    },
    {
      "entropy": 5.366285181045532,
      "epoch": 2.0108945920143766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004602846306198465,
      "loss": 5.1095,
      "mean_token_accuracy": 0.2004966050386429,
      "num_tokens": 39968385.0,
      "step": 17905
    },
    {
      "entropy": 5.3870128154754635,
      "epoch": 2.011456168922334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004602618499054889,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2043771043419838,
      "num_tokens": 39979393.0,
      "step": 17910
    },
    {
      "entropy": 5.452572631835937,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20247626453638076,
      "num_tokens": 39989709.0,
      "step": 17915
    },
    {
      "entropy": 5.444457292556763,
      "epoch": 2.012579322738249,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046021627078107177,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2018706366419792,
      "num_tokens": 40000472.0,
      "step": 17920
    },
    {
      "entropy": 5.294516372680664,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046019347237246343,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2092535302042961,
      "num_tokens": 40011062.0,
      "step": 17925
    },
    {
      "entropy": 5.4835052490234375,
      "epoch": 2.013702476554164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004601706680672232,
      "loss": 5.2223,
      "mean_token_accuracy": 0.1911832183599472,
      "num_tokens": 40022430.0,
      "step": 17930
    },
    {
      "entropy": 5.378408432006836,
      "epoch": 2.0142640534621217,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004601478578660771,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20309155732393264,
      "num_tokens": 40033077.0,
      "step": 17935
    },
    {
      "entropy": 5.419387531280518,
      "epoch": 2.0148256303700793,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004601250417697513,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20527995228767396,
      "num_tokens": 40044258.0,
      "step": 17940
    },
    {
      "entropy": 5.4610981941223145,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19742362797260285,
      "num_tokens": 40055876.0,
      "step": 17945
    },
    {
      "entropy": 5.3891064643859865,
      "epoch": 2.015948784185994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004600793918944671,
      "loss": 5.0561,
      "mean_token_accuracy": 0.21038345545530318,
      "num_tokens": 40066691.0,
      "step": 17950
    },
    {
      "entropy": 5.377702474594116,
      "epoch": 2.016510361093952,
      "grad_norm": 1.25,
      "learning_rate": 0.0004600565581169621,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2023324579000473,
      "num_tokens": 40077335.0,
      "step": 17955
    },
    {
      "entropy": 5.377427053451538,
      "epoch": 2.0170719380019095,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.1843,
      "mean_token_accuracy": 0.19583686143159867,
      "num_tokens": 40089053.0,
      "step": 17960
    },
    {
      "entropy": 5.410895347595215,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046001087288586164,
      "loss": 5.064,
      "mean_token_accuracy": 0.21343138813972473,
      "num_tokens": 40099371.0,
      "step": 17965
    },
    {
      "entropy": 5.44838547706604,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19507868885993956,
      "num_tokens": 40110274.0,
      "step": 17970
    },
    {
      "entropy": 5.447851467132568,
      "epoch": 2.018756668725782,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19923406094312668,
      "num_tokens": 40120364.0,
      "step": 17975
    },
    {
      "entropy": 5.411220073699951,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.1498,
      "mean_token_accuracy": 0.2032131403684616,
      "num_tokens": 40132400.0,
      "step": 17980
    },
    {
      "entropy": 5.461653137207032,
      "epoch": 2.019879822541697,
      "grad_norm": 1.25,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19805140495300294,
      "num_tokens": 40142820.0,
      "step": 17985
    },
    {
      "entropy": 5.452293825149536,
      "epoch": 2.0204413994496546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19969666600227357,
      "num_tokens": 40153620.0,
      "step": 17990
    },
    {
      "entropy": 5.413561582565308,
      "epoch": 2.0210029763576123,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004598736758362198,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20340935736894608,
      "num_tokens": 40164250.0,
      "step": 17995
    },
    {
      "entropy": 5.411157274246216,
      "epoch": 2.02156455326557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045985078905445734,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20011104941368102,
      "num_tokens": 40175809.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.321115857459603,
      "eval_loss": 5.305892467498779,
      "eval_mean_token_accuracy": 0.20272016978606694,
      "eval_num_tokens": 40175809.0,
      "eval_runtime": 59.3033,
      "eval_samples_per_second": 522.956,
      "eval_steps_per_second": 65.376,
      "step": 18000
    },
    {
      "entropy": 5.342676639556885,
      "epoch": 2.022126130173527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045982789638697414,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2111600697040558,
      "num_tokens": 40187855.0,
      "step": 18005
    },
    {
      "entropy": 5.4333912372589115,
      "epoch": 2.022687707081485,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045980499783449897,
      "loss": 5.0945,
      "mean_token_accuracy": 0.20565036237239837,
      "num_tokens": 40198802.0,
      "step": 18010
    },
    {
      "entropy": 5.5392069816589355,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19172475039958953,
      "num_tokens": 40210636.0,
      "step": 18015
    },
    {
      "entropy": 5.48834981918335,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.177,
      "mean_token_accuracy": 0.20160965770483016,
      "num_tokens": 40221216.0,
      "step": 18020
    },
    {
      "entropy": 5.357290506362915,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004597362668744141,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21414878219366074,
      "num_tokens": 40231988.0,
      "step": 18025
    },
    {
      "entropy": 5.464912080764771,
      "epoch": 2.024934014713315,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.3246,
      "mean_token_accuracy": 0.1910130575299263,
      "num_tokens": 40243789.0,
      "step": 18030
    },
    {
      "entropy": 5.443649530410767,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20498851090669631,
      "num_tokens": 40254771.0,
      "step": 18035
    },
    {
      "entropy": 5.445785856246948,
      "epoch": 2.02605716852923,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045966748297566085,
      "loss": 5.1119,
      "mean_token_accuracy": 0.20620031654834747,
      "num_tokens": 40265444.0,
      "step": 18040
    },
    {
      "entropy": 5.415645790100098,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.132,
      "mean_token_accuracy": 0.20005106925964355,
      "num_tokens": 40277146.0,
      "step": 18045
    },
    {
      "entropy": 5.396540737152099,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.183,
      "mean_token_accuracy": 0.206864395737648,
      "num_tokens": 40286493.0,
      "step": 18050
    },
    {
      "entropy": 5.450122547149658,
      "epoch": 2.027741899253103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.1386,
      "mean_token_accuracy": 0.20539827197790145,
      "num_tokens": 40297209.0,
      "step": 18055
    },
    {
      "entropy": 5.372835731506347,
      "epoch": 2.02830347616106,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045957568879568736,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20194728523492814,
      "num_tokens": 40308666.0,
      "step": 18060
    },
    {
      "entropy": 5.4655797481536865,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.2179,
      "mean_token_accuracy": 0.19341700226068498,
      "num_tokens": 40321145.0,
      "step": 18065
    },
    {
      "entropy": 5.511399030685425,
      "epoch": 2.0294266299769754,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.14,
      "mean_token_accuracy": 0.20271724164485933,
      "num_tokens": 40332418.0,
      "step": 18070
    },
    {
      "entropy": 5.383347415924073,
      "epoch": 2.029988206884933,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045950678145005456,
      "loss": 5.0994,
      "mean_token_accuracy": 0.2037181079387665,
      "num_tokens": 40342838.0,
      "step": 18075
    },
    {
      "entropy": 5.344696950912476,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.125,
      "learning_rate": 0.0004594838005843333,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20574889779090882,
      "num_tokens": 40352940.0,
      "step": 18080
    },
    {
      "entropy": 5.551378393173218,
      "epoch": 2.031111360700848,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.1909,
      "mean_token_accuracy": 0.19839725643396378,
      "num_tokens": 40364273.0,
      "step": 18085
    },
    {
      "entropy": 5.464349460601807,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045943782123151556,
      "loss": 5.0678,
      "mean_token_accuracy": 0.2119649678468704,
      "num_tokens": 40374617.0,
      "step": 18090
    },
    {
      "entropy": 5.406306219100952,
      "epoch": 2.0322345145167633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19823758900165558,
      "num_tokens": 40385398.0,
      "step": 18095
    },
    {
      "entropy": 5.546410942077637,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.263,
      "mean_token_accuracy": 0.19752512574195863,
      "num_tokens": 40397030.0,
      "step": 18100
    },
    {
      "entropy": 5.546636247634888,
      "epoch": 2.033357668332678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.2165,
      "mean_token_accuracy": 0.19881228655576705,
      "num_tokens": 40408027.0,
      "step": 18105
    },
    {
      "entropy": 5.5013751029968265,
      "epoch": 2.033919245240636,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1980292335152626,
      "num_tokens": 40418755.0,
      "step": 18110
    },
    {
      "entropy": 5.4994605541229244,
      "epoch": 2.034480822148593,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.1846,
      "mean_token_accuracy": 0.1996479034423828,
      "num_tokens": 40430694.0,
      "step": 18115
    },
    {
      "entropy": 5.474099493026733,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20339277535676956,
      "num_tokens": 40442019.0,
      "step": 18120
    },
    {
      "entropy": 5.468339729309082,
      "epoch": 2.0356039759645084,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1907990410923958,
      "num_tokens": 40454120.0,
      "step": 18125
    },
    {
      "entropy": 5.468143033981323,
      "epoch": 2.036165552872466,
      "grad_norm": 1.0,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1962733805179596,
      "num_tokens": 40467445.0,
      "step": 18130
    },
    {
      "entropy": 5.396000623703003,
      "epoch": 2.0367271297804233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045923062353615167,
      "loss": 5.0807,
      "mean_token_accuracy": 0.19919672310352327,
      "num_tokens": 40477512.0,
      "step": 18135
    },
    {
      "entropy": 5.377988529205322,
      "epoch": 2.037288706688381,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045920757223034833,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2078073188662529,
      "num_tokens": 40488209.0,
      "step": 18140
    },
    {
      "entropy": 5.503016757965088,
      "epoch": 2.0378502835963386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.1584,
      "mean_token_accuracy": 0.20099363029003142,
      "num_tokens": 40499793.0,
      "step": 18145
    },
    {
      "entropy": 5.492732429504395,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.2591,
      "mean_token_accuracy": 0.18891579508781434,
      "num_tokens": 40510598.0,
      "step": 18150
    },
    {
      "entropy": 5.550028085708618,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.1737,
      "mean_token_accuracy": 0.20601840019226075,
      "num_tokens": 40521289.0,
      "step": 18155
    },
    {
      "entropy": 5.514279079437256,
      "epoch": 2.039535014320211,
      "grad_norm": 2.09375,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.2287,
      "mean_token_accuracy": 0.20284309089183808,
      "num_tokens": 40532048.0,
      "step": 18160
    },
    {
      "entropy": 5.421380043029785,
      "epoch": 2.0400965912281688,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.219,
      "mean_token_accuracy": 0.202123162150383,
      "num_tokens": 40543699.0,
      "step": 18165
    },
    {
      "entropy": 5.511491966247559,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.125,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.2368,
      "mean_token_accuracy": 0.20196765661239624,
      "num_tokens": 40554626.0,
      "step": 18170
    },
    {
      "entropy": 5.497847557067871,
      "epoch": 2.0412197450440837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045904604890411476,
      "loss": 5.1873,
      "mean_token_accuracy": 0.19653425216674805,
      "num_tokens": 40567046.0,
      "step": 18175
    },
    {
      "entropy": 5.581865406036377,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19376737773418426,
      "num_tokens": 40578472.0,
      "step": 18180
    },
    {
      "entropy": 5.43798828125,
      "epoch": 2.042342898859999,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.1616,
      "mean_token_accuracy": 0.20614684820175172,
      "num_tokens": 40589012.0,
      "step": 18185
    },
    {
      "entropy": 5.367080783843994,
      "epoch": 2.0429044757679566,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045897673670528086,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20400851219892502,
      "num_tokens": 40600807.0,
      "step": 18190
    },
    {
      "entropy": 5.541237449645996,
      "epoch": 2.043466052675914,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.2549,
      "mean_token_accuracy": 0.19673992544412613,
      "num_tokens": 40612578.0,
      "step": 18195
    },
    {
      "entropy": 5.422832822799682,
      "epoch": 2.0440276295838715,
      "grad_norm": 1.125,
      "learning_rate": 0.0004589304992820454,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21189106851816178,
      "num_tokens": 40623504.0,
      "step": 18200
    },
    {
      "entropy": 5.4539172649383545,
      "epoch": 2.044589206491829,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000458907371785432,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20232882499694824,
      "num_tokens": 40633676.0,
      "step": 18205
    },
    {
      "entropy": 5.46436653137207,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19831877201795578,
      "num_tokens": 40645991.0,
      "step": 18210
    },
    {
      "entropy": 5.406685876846313,
      "epoch": 2.045712360307744,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045886109922589564,
      "loss": 5.0581,
      "mean_token_accuracy": 0.20341494381427766,
      "num_tokens": 40656617.0,
      "step": 18215
    },
    {
      "entropy": 5.4644303798675535,
      "epoch": 2.0462739372157017,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.1888,
      "mean_token_accuracy": 0.2039027139544487,
      "num_tokens": 40669708.0,
      "step": 18220
    },
    {
      "entropy": 5.429202508926392,
      "epoch": 2.0468355141236594,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.1642,
      "mean_token_accuracy": 0.19802865236997605,
      "num_tokens": 40680248.0,
      "step": 18225
    },
    {
      "entropy": 5.529597806930542,
      "epoch": 2.0473970910316166,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.2533,
      "mean_token_accuracy": 0.1969241663813591,
      "num_tokens": 40692704.0,
      "step": 18230
    },
    {
      "entropy": 5.526028156280518,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.1539,
      "mean_token_accuracy": 0.2051002413034439,
      "num_tokens": 40703156.0,
      "step": 18235
    },
    {
      "entropy": 5.4386255741119385,
      "epoch": 2.048520244847532,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045874531539130204,
      "loss": 5.117,
      "mean_token_accuracy": 0.2031269609928131,
      "num_tokens": 40713996.0,
      "step": 18240
    },
    {
      "entropy": 5.371656608581543,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045872214106986684,
      "loss": 5.0362,
      "mean_token_accuracy": 0.2048846647143364,
      "num_tokens": 40724532.0,
      "step": 18245
    },
    {
      "entropy": 5.483946418762207,
      "epoch": 2.049643398663447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.2416,
      "mean_token_accuracy": 0.19358222484588622,
      "num_tokens": 40736045.0,
      "step": 18250
    },
    {
      "entropy": 5.337462568283081,
      "epoch": 2.0502049755714045,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004586757748783817,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20495191663503648,
      "num_tokens": 40747398.0,
      "step": 18255
    },
    {
      "entropy": 5.323112201690674,
      "epoch": 2.050766552479362,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045865258300980817,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21282731890678405,
      "num_tokens": 40758219.0,
      "step": 18260
    },
    {
      "entropy": 5.454240703582764,
      "epoch": 2.05132812938732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045862938529366514,
      "loss": 5.09,
      "mean_token_accuracy": 0.20576892495155336,
      "num_tokens": 40770002.0,
      "step": 18265
    },
    {
      "entropy": 5.378674936294556,
      "epoch": 2.051889706295277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004586061817306913,
      "loss": 5.065,
      "mean_token_accuracy": 0.2059107542037964,
      "num_tokens": 40780087.0,
      "step": 18270
    },
    {
      "entropy": 5.4030139446258545,
      "epoch": 2.0524512832032347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19829004555940627,
      "num_tokens": 40792225.0,
      "step": 18275
    },
    {
      "entropy": 5.499179172515869,
      "epoch": 2.0530128601111923,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.2646,
      "mean_token_accuracy": 0.1952887460589409,
      "num_tokens": 40804437.0,
      "step": 18280
    },
    {
      "entropy": 5.409182548522949,
      "epoch": 2.05357443701915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004585365359681736,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21883408427238465,
      "num_tokens": 40814625.0,
      "step": 18285
    },
    {
      "entropy": 5.3476338386535645,
      "epoch": 2.054136013927107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045851330902526633,
      "loss": 5.0638,
      "mean_token_accuracy": 0.20834892839193345,
      "num_tokens": 40825843.0,
      "step": 18290
    },
    {
      "entropy": 5.419632816314698,
      "epoch": 2.054697590835065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.241,
      "mean_token_accuracy": 0.19405852556228637,
      "num_tokens": 40837146.0,
      "step": 18295
    },
    {
      "entropy": 5.560157108306885,
      "epoch": 2.0552591677430225,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.317,
      "mean_token_accuracy": 0.1893691748380661,
      "num_tokens": 40849109.0,
      "step": 18300
    },
    {
      "entropy": 5.447722721099853,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.109375,
      "learning_rate": 0.000458443593140694,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20514724403619766,
      "num_tokens": 40861168.0,
      "step": 18305
    },
    {
      "entropy": 5.456511831283569,
      "epoch": 2.0563823215589374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.1933,
      "mean_token_accuracy": 0.20332808643579484,
      "num_tokens": 40873403.0,
      "step": 18310
    },
    {
      "entropy": 5.419351530075073,
      "epoch": 2.056943898466895,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004583970866785035,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20136093050241471,
      "num_tokens": 40883434.0,
      "step": 18315
    },
    {
      "entropy": 5.448317337036133,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20032256245613098,
      "num_tokens": 40894742.0,
      "step": 18320
    },
    {
      "entropy": 5.359233283996582,
      "epoch": 2.05806705228281,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004583505568585757,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20630411952733993,
      "num_tokens": 40905233.0,
      "step": 18325
    },
    {
      "entropy": 5.384528303146363,
      "epoch": 2.0586286291907676,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045832728319131205,
      "loss": 5.0912,
      "mean_token_accuracy": 0.2106534093618393,
      "num_tokens": 40915897.0,
      "step": 18330
    },
    {
      "entropy": 5.392107009887695,
      "epoch": 2.0591902060987253,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004583040036868367,
      "loss": 5.1009,
      "mean_token_accuracy": 0.1980993553996086,
      "num_tokens": 40928178.0,
      "step": 18335
    },
    {
      "entropy": 5.419974660873413,
      "epoch": 2.059751783006683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.197,
      "mean_token_accuracy": 0.19679423570632934,
      "num_tokens": 40938988.0,
      "step": 18340
    },
    {
      "entropy": 5.402058029174805,
      "epoch": 2.06031335991464,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045825742716921553,
      "loss": 5.0603,
      "mean_token_accuracy": 0.2110185816884041,
      "num_tokens": 40949671.0,
      "step": 18345
    },
    {
      "entropy": 5.333623218536377,
      "epoch": 2.060874936822598,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004582341301575529,
      "loss": 5.024,
      "mean_token_accuracy": 0.20600511878728867,
      "num_tokens": 40960818.0,
      "step": 18350
    },
    {
      "entropy": 5.516381692886353,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19057075828313827,
      "num_tokens": 40973866.0,
      "step": 18355
    },
    {
      "entropy": 5.427972030639649,
      "epoch": 2.061998090638513,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.1414,
      "mean_token_accuracy": 0.19878353029489518,
      "num_tokens": 40983355.0,
      "step": 18360
    },
    {
      "entropy": 5.40568733215332,
      "epoch": 2.0625596675464704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045816420412005826,
      "loss": 5.0394,
      "mean_token_accuracy": 0.20313866138458253,
      "num_tokens": 40994045.0,
      "step": 18365
    },
    {
      "entropy": 5.344874572753906,
      "epoch": 2.063121244454428,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045814088377586505,
      "loss": 5.0601,
      "mean_token_accuracy": 0.21148548722267152,
      "num_tokens": 41003974.0,
      "step": 18370
    },
    {
      "entropy": 5.418494415283203,
      "epoch": 2.0636828213623857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.2131,
      "mean_token_accuracy": 0.20303902626037598,
      "num_tokens": 41015262.0,
      "step": 18375
    },
    {
      "entropy": 5.526286554336548,
      "epoch": 2.0642443982703433,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.2442,
      "mean_token_accuracy": 0.19860045462846757,
      "num_tokens": 41027103.0,
      "step": 18380
    },
    {
      "entropy": 5.481543779373169,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20718812793493271,
      "num_tokens": 41037240.0,
      "step": 18385
    },
    {
      "entropy": 5.492831611633301,
      "epoch": 2.065367552086258,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.1288,
      "mean_token_accuracy": 0.2053864374756813,
      "num_tokens": 41048258.0,
      "step": 18390
    },
    {
      "entropy": 5.467458629608155,
      "epoch": 2.065929128994216,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19467367082834244,
      "num_tokens": 41058204.0,
      "step": 18395
    },
    {
      "entropy": 5.5256345748901365,
      "epoch": 2.066490705902173,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19870574325323104,
      "num_tokens": 41070019.0,
      "step": 18400
    },
    {
      "entropy": 5.524305391311645,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.1885,
      "mean_token_accuracy": 0.20117858648300171,
      "num_tokens": 41081503.0,
      "step": 18405
    },
    {
      "entropy": 5.471236801147461,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.1433,
      "mean_token_accuracy": 0.19625366777181624,
      "num_tokens": 41091959.0,
      "step": 18410
    },
    {
      "entropy": 5.44686918258667,
      "epoch": 2.068175436626046,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19635324776172638,
      "num_tokens": 41103193.0,
      "step": 18415
    },
    {
      "entropy": 5.514858627319336,
      "epoch": 2.0687370135340033,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.2382,
      "mean_token_accuracy": 0.2046363964676857,
      "num_tokens": 41115886.0,
      "step": 18420
    },
    {
      "entropy": 5.456508922576904,
      "epoch": 2.069298590441961,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045788397528896224,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20412856042385102,
      "num_tokens": 41126868.0,
      "step": 18425
    },
    {
      "entropy": 5.498434734344483,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20957268327474593,
      "num_tokens": 41136417.0,
      "step": 18430
    },
    {
      "entropy": 5.462555646896362,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.2289,
      "mean_token_accuracy": 0.19695803970098497,
      "num_tokens": 41149515.0,
      "step": 18435
    },
    {
      "entropy": 5.480551862716675,
      "epoch": 2.0709833211658335,
      "grad_norm": 1.109375,
      "learning_rate": 0.000457813787011288,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20340424925088882,
      "num_tokens": 41160023.0,
      "step": 18440
    },
    {
      "entropy": 5.3311539649963375,
      "epoch": 2.071544898073791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577903792760127,
      "loss": 5.015,
      "mean_token_accuracy": 0.2142372488975525,
      "num_tokens": 41170909.0,
      "step": 18445
    },
    {
      "entropy": 5.474853134155273,
      "epoch": 2.072106474981749,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.2234,
      "mean_token_accuracy": 0.19887248575687408,
      "num_tokens": 41181960.0,
      "step": 18450
    },
    {
      "entropy": 5.537634372711182,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.2323,
      "mean_token_accuracy": 0.1956140100955963,
      "num_tokens": 41192292.0,
      "step": 18455
    },
    {
      "entropy": 5.509148359298706,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20363277643918992,
      "num_tokens": 41203826.0,
      "step": 18460
    },
    {
      "entropy": 5.395159101486206,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.1844,
      "mean_token_accuracy": 0.203907510638237,
      "num_tokens": 41215095.0,
      "step": 18465
    },
    {
      "entropy": 5.459643125534058,
      "epoch": 2.074352782613579,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045767325331280044,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20108969360589982,
      "num_tokens": 41224953.0,
      "step": 18470
    },
    {
      "entropy": 5.464155530929565,
      "epoch": 2.0749143595215367,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.1939,
      "mean_token_accuracy": 0.19282402992248535,
      "num_tokens": 41235479.0,
      "step": 18475
    },
    {
      "entropy": 5.477154350280761,
      "epoch": 2.075475936429494,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20027188062667847,
      "num_tokens": 41248520.0,
      "step": 18480
    },
    {
      "entropy": 5.565487384796143,
      "epoch": 2.0760375133374516,
      "grad_norm": 1.0625,
      "learning_rate": 0.000457602907932259,
      "loss": 5.2896,
      "mean_token_accuracy": 0.19452620595693587,
      "num_tokens": 41259112.0,
      "step": 18485
    },
    {
      "entropy": 5.431046628952027,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045757944784278803,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20431412160396575,
      "num_tokens": 41269554.0,
      "step": 18490
    },
    {
      "entropy": 5.419406127929688,
      "epoch": 2.0771606671533664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.1737,
      "mean_token_accuracy": 0.2000341922044754,
      "num_tokens": 41281339.0,
      "step": 18495
    },
    {
      "entropy": 5.378335285186767,
      "epoch": 2.077722244061324,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045753251022439195,
      "loss": 5.0698,
      "mean_token_accuracy": 0.21373364478349685,
      "num_tokens": 41291629.0,
      "step": 18500
    },
    {
      "entropy": 5.557219362258911,
      "epoch": 2.0782838209692818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20161436796188353,
      "num_tokens": 41302162.0,
      "step": 18505
    },
    {
      "entropy": 5.40969409942627,
      "epoch": 2.0788453978772394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.1543,
      "mean_token_accuracy": 0.20534624606370927,
      "num_tokens": 41313667.0,
      "step": 18510
    },
    {
      "entropy": 5.391064834594727,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045746206020937276,
      "loss": 5.1113,
      "mean_token_accuracy": 0.19790396988391876,
      "num_tokens": 41325063.0,
      "step": 18515
    },
    {
      "entropy": 5.434782218933106,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20320819318294525,
      "num_tokens": 41335974.0,
      "step": 18520
    },
    {
      "entropy": 5.448341941833496,
      "epoch": 2.080530128601112,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045741506448313327,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20061599016189574,
      "num_tokens": 41347469.0,
      "step": 18525
    },
    {
      "entropy": 5.528663539886475,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.2057,
      "mean_token_accuracy": 0.20094305127859116,
      "num_tokens": 41358418.0,
      "step": 18530
    },
    {
      "entropy": 5.439347124099731,
      "epoch": 2.081653282417027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.1837,
      "mean_token_accuracy": 0.20304896384477616,
      "num_tokens": 41368859.0,
      "step": 18535
    },
    {
      "entropy": 5.453055143356323,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.2435,
      "mean_token_accuracy": 0.19476781487464906,
      "num_tokens": 41380929.0,
      "step": 18540
    },
    {
      "entropy": 5.4635162353515625,
      "epoch": 2.082776436232942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004573210033386464,
      "loss": 5.0651,
      "mean_token_accuracy": 0.2080380767583847,
      "num_tokens": 41392190.0,
      "step": 18545
    },
    {
      "entropy": 5.517323398590088,
      "epoch": 2.0833380131409,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20934794694185258,
      "num_tokens": 41402646.0,
      "step": 18550
    },
    {
      "entropy": 5.43314905166626,
      "epoch": 2.083899590048857,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.1509,
      "mean_token_accuracy": 0.2019234076142311,
      "num_tokens": 41415179.0,
      "step": 18555
    },
    {
      "entropy": 5.417903470993042,
      "epoch": 2.0844611669568147,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.2,
      "mean_token_accuracy": 0.19754811972379685,
      "num_tokens": 41426695.0,
      "step": 18560
    },
    {
      "entropy": 5.39660415649414,
      "epoch": 2.0850227438647724,
      "grad_norm": 1.125,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.1422,
      "mean_token_accuracy": 0.19991028308868408,
      "num_tokens": 41438078.0,
      "step": 18565
    },
    {
      "entropy": 5.49943060874939,
      "epoch": 2.08558432077273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1975611314177513,
      "num_tokens": 41448862.0,
      "step": 18570
    },
    {
      "entropy": 5.394349050521851,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045717973748176695,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21039617210626602,
      "num_tokens": 41458528.0,
      "step": 18575
    },
    {
      "entropy": 5.361552572250366,
      "epoch": 2.086707474588645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045715617286555516,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20048311054706575,
      "num_tokens": 41469553.0,
      "step": 18580
    },
    {
      "entropy": 5.4895062923431395,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1961825355887413,
      "num_tokens": 41482327.0,
      "step": 18585
    },
    {
      "entropy": 5.423326110839843,
      "epoch": 2.08783062840456,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004571090262341125,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21568716764450074,
      "num_tokens": 41493126.0,
      "step": 18590
    },
    {
      "entropy": 5.462135696411133,
      "epoch": 2.0883922053125175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.2227,
      "mean_token_accuracy": 0.19605877697467805,
      "num_tokens": 41505597.0,
      "step": 18595
    },
    {
      "entropy": 5.472991895675659,
      "epoch": 2.088953782220475,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.2494,
      "mean_token_accuracy": 0.1958896115422249,
      "num_tokens": 41517296.0,
      "step": 18600
    },
    {
      "entropy": 5.4867640972137455,
      "epoch": 2.089515359128433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20461816638708114,
      "num_tokens": 41529900.0,
      "step": 18605
    },
    {
      "entropy": 5.509304237365723,
      "epoch": 2.09007693603639,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20411134660243987,
      "num_tokens": 41540641.0,
      "step": 18610
    },
    {
      "entropy": 5.430690717697144,
      "epoch": 2.0906385129443477,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.1592,
      "mean_token_accuracy": 0.1952848806977272,
      "num_tokens": 41551335.0,
      "step": 18615
    },
    {
      "entropy": 5.386656427383423,
      "epoch": 2.0912000898523053,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20962560176849365,
      "num_tokens": 41562265.0,
      "step": 18620
    },
    {
      "entropy": 5.4246316909790036,
      "epoch": 2.091761666760263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004569438304132031,
      "loss": 5.047,
      "mean_token_accuracy": 0.20731533616781234,
      "num_tokens": 41574758.0,
      "step": 18625
    },
    {
      "entropy": 5.3818775653839115,
      "epoch": 2.09232324366822,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045692020783155376,
      "loss": 5.0218,
      "mean_token_accuracy": 0.20591089278459548,
      "num_tokens": 41585007.0,
      "step": 18630
    },
    {
      "entropy": 5.353652048110962,
      "epoch": 2.092884820576178,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004568965794574943,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20557635724544526,
      "num_tokens": 41596026.0,
      "step": 18635
    },
    {
      "entropy": 5.466334581375122,
      "epoch": 2.0934463974841355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.2112,
      "mean_token_accuracy": 0.20319215953350067,
      "num_tokens": 41608440.0,
      "step": 18640
    },
    {
      "entropy": 5.475106048583984,
      "epoch": 2.094007974392093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045684930533515506,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20157624930143356,
      "num_tokens": 41620467.0,
      "step": 18645
    },
    {
      "entropy": 5.37600622177124,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045682565958838045,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20145183801651,
      "num_tokens": 41630918.0,
      "step": 18650
    },
    {
      "entropy": 5.3887773036956785,
      "epoch": 2.095131128208008,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004568020080522063,
      "loss": 5.1062,
      "mean_token_accuracy": 0.19668225198984146,
      "num_tokens": 41641148.0,
      "step": 18655
    },
    {
      "entropy": 5.469299411773681,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.1611,
      "mean_token_accuracy": 0.2015403240919113,
      "num_tokens": 41652148.0,
      "step": 18660
    },
    {
      "entropy": 5.399859428405762,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045675468761467204,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20415159165859223,
      "num_tokens": 41662275.0,
      "step": 18665
    },
    {
      "entropy": 5.472311973571777,
      "epoch": 2.0968158589318806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19498195499181747,
      "num_tokens": 41673141.0,
      "step": 18670
    },
    {
      "entropy": 5.562878036499024,
      "epoch": 2.0973774358398383,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19360179752111434,
      "num_tokens": 41683473.0,
      "step": 18675
    },
    {
      "entropy": 5.323286437988282,
      "epoch": 2.097939012747796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045668366355670697,
      "loss": 4.9456,
      "mean_token_accuracy": 0.20927121192216874,
      "num_tokens": 41693847.0,
      "step": 18680
    },
    {
      "entropy": 5.407052087783813,
      "epoch": 2.098500589655753,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.1971,
      "mean_token_accuracy": 0.20385240912437438,
      "num_tokens": 41705810.0,
      "step": 18685
    },
    {
      "entropy": 5.428302955627442,
      "epoch": 2.099062166563711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045663628525908237,
      "loss": 5.0532,
      "mean_token_accuracy": 0.20395218580961227,
      "num_tokens": 41717160.0,
      "step": 18690
    },
    {
      "entropy": 5.4835858821868895,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.2142,
      "mean_token_accuracy": 0.200162111222744,
      "num_tokens": 41728631.0,
      "step": 18695
    },
    {
      "entropy": 5.515122652053833,
      "epoch": 2.100185320379626,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1960071235895157,
      "num_tokens": 41738933.0,
      "step": 18700
    },
    {
      "entropy": 5.539283609390258,
      "epoch": 2.1007468972875833,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.2546,
      "mean_token_accuracy": 0.20130504220724105,
      "num_tokens": 41750386.0,
      "step": 18705
    },
    {
      "entropy": 5.420370292663574,
      "epoch": 2.101308474195541,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20445092320442199,
      "num_tokens": 41760711.0,
      "step": 18710
    },
    {
      "entropy": 5.4341185092926025,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.2115,
      "mean_token_accuracy": 0.19421566128730774,
      "num_tokens": 41772147.0,
      "step": 18715
    },
    {
      "entropy": 5.489456415176392,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045649401158948664,
      "loss": 5.0851,
      "mean_token_accuracy": 0.2066085547208786,
      "num_tokens": 41783265.0,
      "step": 18720
    },
    {
      "entropy": 5.4553502082824705,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004564702790808758,
      "loss": 5.0709,
      "mean_token_accuracy": 0.20633191168308257,
      "num_tokens": 41793977.0,
      "step": 18725
    },
    {
      "entropy": 5.431276035308838,
      "epoch": 2.103554781827371,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004564465407941809,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20670644640922547,
      "num_tokens": 41804790.0,
      "step": 18730
    },
    {
      "entropy": 5.326567792892456,
      "epoch": 2.104116358735329,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045642279673015787,
      "loss": 5.0885,
      "mean_token_accuracy": 0.19919336289167405,
      "num_tokens": 41815611.0,
      "step": 18735
    },
    {
      "entropy": 5.374368286132812,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.1267,
      "mean_token_accuracy": 0.19925709962844848,
      "num_tokens": 41826935.0,
      "step": 18740
    },
    {
      "entropy": 5.438326406478882,
      "epoch": 2.1052395125512438,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004563752912731512,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20564910024404526,
      "num_tokens": 41838932.0,
      "step": 18745
    },
    {
      "entropy": 5.342984628677368,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.1491,
      "mean_token_accuracy": 0.1996524453163147,
      "num_tokens": 41850771.0,
      "step": 18750
    },
    {
      "entropy": 5.364787626266479,
      "epoch": 2.106362666367159,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045632776271590624,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20850711017847062,
      "num_tokens": 41861515.0,
      "step": 18755
    },
    {
      "entropy": 5.412940645217896,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004563039897765859,
      "loss": 4.9982,
      "mean_token_accuracy": 0.21647325456142424,
      "num_tokens": 41872195.0,
      "step": 18760
    },
    {
      "entropy": 5.370642948150635,
      "epoch": 2.107485820183074,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004562802110644764,
      "loss": 5.0935,
      "mean_token_accuracy": 0.2083800658583641,
      "num_tokens": 41882496.0,
      "step": 18765
    },
    {
      "entropy": 5.51829423904419,
      "epoch": 2.1080473970910316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1927931562066078,
      "num_tokens": 41893078.0,
      "step": 18770
    },
    {
      "entropy": 5.539067697525025,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.1821,
      "mean_token_accuracy": 0.20135900229215623,
      "num_tokens": 41905062.0,
      "step": 18775
    },
    {
      "entropy": 5.409191608428955,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.1486,
      "mean_token_accuracy": 0.20454021692276,
      "num_tokens": 41916086.0,
      "step": 18780
    },
    {
      "entropy": 5.423175621032715,
      "epoch": 2.109732127814904,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.199,
      "mean_token_accuracy": 0.2041832149028778,
      "num_tokens": 41925466.0,
      "step": 18785
    },
    {
      "entropy": 5.495348691940308,
      "epoch": 2.110293704722862,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19579385071992875,
      "num_tokens": 41936567.0,
      "step": 18790
    },
    {
      "entropy": 5.391596174240112,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045613741760565385,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2067500650882721,
      "num_tokens": 41947021.0,
      "step": 18795
    },
    {
      "entropy": 5.491127681732178,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.1489,
      "mean_token_accuracy": 0.19715440422296523,
      "num_tokens": 41957929.0,
      "step": 18800
    },
    {
      "entropy": 5.431910514831543,
      "epoch": 2.1119784354467344,
      "grad_norm": 1.125,
      "learning_rate": 0.000456089773638076,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2119723841547966,
      "num_tokens": 41968677.0,
      "step": 18805
    },
    {
      "entropy": 5.430100631713867,
      "epoch": 2.112540012354692,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.2284,
      "mean_token_accuracy": 0.200178100168705,
      "num_tokens": 41980330.0,
      "step": 18810
    },
    {
      "entropy": 5.422392892837524,
      "epoch": 2.1131015892626497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004560421066066237,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20078005194664,
      "num_tokens": 41991258.0,
      "step": 18815
    },
    {
      "entropy": 5.518226575851441,
      "epoch": 2.113663166170607,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.2183,
      "mean_token_accuracy": 0.20122957080602646,
      "num_tokens": 42002679.0,
      "step": 18820
    },
    {
      "entropy": 5.434555959701538,
      "epoch": 2.1142247430785646,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045599441651736805,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20533353090286255,
      "num_tokens": 42012845.0,
      "step": 18825
    },
    {
      "entropy": 5.458740615844727,
      "epoch": 2.1147863199865222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.1744,
      "mean_token_accuracy": 0.19988572001457214,
      "num_tokens": 42024461.0,
      "step": 18830
    },
    {
      "entropy": 5.472439956665039,
      "epoch": 2.11534789689448,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004559467033763829,
      "loss": 5.1641,
      "mean_token_accuracy": 0.19594511538743972,
      "num_tokens": 42036996.0,
      "step": 18835
    },
    {
      "entropy": 5.492838907241821,
      "epoch": 2.115909473802437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.2189,
      "mean_token_accuracy": 0.20225370973348616,
      "num_tokens": 42048731.0,
      "step": 18840
    },
    {
      "entropy": 5.508620548248291,
      "epoch": 2.1164710507103948,
      "grad_norm": 1.125,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20715158581733703,
      "num_tokens": 42060396.0,
      "step": 18845
    },
    {
      "entropy": 5.34435806274414,
      "epoch": 2.1170326276183524,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004558750904562063,
      "loss": 5.065,
      "mean_token_accuracy": 0.2060753136873245,
      "num_tokens": 42072037.0,
      "step": 18850
    },
    {
      "entropy": 5.374680376052856,
      "epoch": 2.11759420452631,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004558512079635345,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21232302635908126,
      "num_tokens": 42082577.0,
      "step": 18855
    },
    {
      "entropy": 5.351109647750855,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004558273197124899,
      "loss": 5.048,
      "mean_token_accuracy": 0.21000571250915528,
      "num_tokens": 42093978.0,
      "step": 18860
    },
    {
      "entropy": 5.437092351913452,
      "epoch": 2.118717358342225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1936510145664215,
      "num_tokens": 42105773.0,
      "step": 18865
    },
    {
      "entropy": 5.434174871444702,
      "epoch": 2.1192789352501826,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20186442285776138,
      "num_tokens": 42116232.0,
      "step": 18870
    },
    {
      "entropy": 5.438861465454101,
      "epoch": 2.11984051215814,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20004648268222808,
      "num_tokens": 42127813.0,
      "step": 18875
    },
    {
      "entropy": 5.392910099029541,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045573170913980094,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20497957020998,
      "num_tokens": 42138054.0,
      "step": 18880
    },
    {
      "entropy": 5.48740520477295,
      "epoch": 2.120963665974055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.1723,
      "mean_token_accuracy": 0.1970529392361641,
      "num_tokens": 42149575.0,
      "step": 18885
    },
    {
      "entropy": 5.44726448059082,
      "epoch": 2.121525242882013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.1766,
      "mean_token_accuracy": 0.2020699843764305,
      "num_tokens": 42160824.0,
      "step": 18890
    },
    {
      "entropy": 5.456457281112671,
      "epoch": 2.12208681978997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.1995,
      "mean_token_accuracy": 0.20040919482707978,
      "num_tokens": 42171088.0,
      "step": 18895
    },
    {
      "entropy": 5.397271776199341,
      "epoch": 2.1226483966979277,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.1521,
      "mean_token_accuracy": 0.2001300647854805,
      "num_tokens": 42182112.0,
      "step": 18900
    },
    {
      "entropy": 5.45868763923645,
      "epoch": 2.1232099736058854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004556120664518838,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21389970928430557,
      "num_tokens": 42193563.0,
      "step": 18905
    },
    {
      "entropy": 5.448945569992065,
      "epoch": 2.123771550513843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20248157680034637,
      "num_tokens": 42204891.0,
      "step": 18910
    },
    {
      "entropy": 5.459609270095825,
      "epoch": 2.1243331274218002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.1825,
      "mean_token_accuracy": 0.20771038085222243,
      "num_tokens": 42217226.0,
      "step": 18915
    },
    {
      "entropy": 5.480522537231446,
      "epoch": 2.124894704329758,
      "grad_norm": 1.15625,
      "learning_rate": 0.000455540211823939,
      "loss": 5.1579,
      "mean_token_accuracy": 0.19973548352718354,
      "num_tokens": 42228134.0,
      "step": 18920
    },
    {
      "entropy": 5.466844320297241,
      "epoch": 2.1254562812377156,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.1708,
      "mean_token_accuracy": 0.19729188978672027,
      "num_tokens": 42239868.0,
      "step": 18925
    },
    {
      "entropy": 5.501995801925659,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19998051822185517,
      "num_tokens": 42250297.0,
      "step": 18930
    },
    {
      "entropy": 5.418631935119629,
      "epoch": 2.1265794350536305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045546830545976203,
      "loss": 4.9213,
      "mean_token_accuracy": 0.21900530755519867,
      "num_tokens": 42261263.0,
      "step": 18935
    },
    {
      "entropy": 5.44773416519165,
      "epoch": 2.127141011961588,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20379654169082642,
      "num_tokens": 42272357.0,
      "step": 18940
    },
    {
      "entropy": 5.463847923278808,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19887278974056244,
      "num_tokens": 42283630.0,
      "step": 18945
    },
    {
      "entropy": 5.468612623214722,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.2215,
      "mean_token_accuracy": 0.19014732539653778,
      "num_tokens": 42294981.0,
      "step": 18950
    },
    {
      "entropy": 5.440510082244873,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045537234986455964,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20265614837408066,
      "num_tokens": 42304877.0,
      "step": 18955
    },
    {
      "entropy": 5.449815654754639,
      "epoch": 2.1293873195934183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.2044,
      "mean_token_accuracy": 0.19880924224853516,
      "num_tokens": 42316325.0,
      "step": 18960
    },
    {
      "entropy": 5.424876499176025,
      "epoch": 2.129948896501376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1977984607219696,
      "num_tokens": 42327437.0,
      "step": 18965
    },
    {
      "entropy": 5.327878475189209,
      "epoch": 2.130510473409333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004553003228626665,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2110733762383461,
      "num_tokens": 42337742.0,
      "step": 18970
    },
    {
      "entropy": 5.424053001403808,
      "epoch": 2.131072050317291,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.1295,
      "mean_token_accuracy": 0.2034108117222786,
      "num_tokens": 42348511.0,
      "step": 18975
    },
    {
      "entropy": 5.41993522644043,
      "epoch": 2.1316336272252485,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004552522761559694,
      "loss": 5.1041,
      "mean_token_accuracy": 0.2035752847790718,
      "num_tokens": 42358977.0,
      "step": 18980
    },
    {
      "entropy": 5.47494535446167,
      "epoch": 2.132195204133206,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.2033,
      "mean_token_accuracy": 0.2011194109916687,
      "num_tokens": 42371609.0,
      "step": 18985
    },
    {
      "entropy": 5.487996816635132,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20524742007255553,
      "num_tokens": 42383150.0,
      "step": 18990
    },
    {
      "entropy": 5.456073999404907,
      "epoch": 2.133318357949121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004551801630530692,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20199889689683914,
      "num_tokens": 42395150.0,
      "step": 18995
    },
    {
      "entropy": 5.459486722946167,
      "epoch": 2.1338799348570787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.1704,
      "mean_token_accuracy": 0.19895248115062714,
      "num_tokens": 42406348.0,
      "step": 19000
    },
    {
      "entropy": 5.495809078216553,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.1709,
      "mean_token_accuracy": 0.1946457713842392,
      "num_tokens": 42417183.0,
      "step": 19005
    },
    {
      "entropy": 5.447253894805908,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045510799831711366,
      "loss": 5.1003,
      "mean_token_accuracy": 0.21330567449331284,
      "num_tokens": 42427805.0,
      "step": 19010
    },
    {
      "entropy": 5.44368577003479,
      "epoch": 2.1355646655809513,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045508393193469,
      "loss": 5.107,
      "mean_token_accuracy": 0.19952069073915482,
      "num_tokens": 42439521.0,
      "step": 19015
    },
    {
      "entropy": 5.526159000396729,
      "epoch": 2.136126242488909,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19805433452129365,
      "num_tokens": 42451201.0,
      "step": 19020
    },
    {
      "entropy": 5.472752714157105,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004550357819687825,
      "loss": 5.121,
      "mean_token_accuracy": 0.2027232512831688,
      "num_tokens": 42462315.0,
      "step": 19025
    },
    {
      "entropy": 5.591591310501099,
      "epoch": 2.137249396304824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.319,
      "mean_token_accuracy": 0.18874753266572952,
      "num_tokens": 42473993.0,
      "step": 19030
    },
    {
      "entropy": 5.590986633300782,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.3004,
      "mean_token_accuracy": 0.19695394486188889,
      "num_tokens": 42484307.0,
      "step": 19035
    },
    {
      "entropy": 5.490777206420899,
      "epoch": 2.138372550120739,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19738707989454268,
      "num_tokens": 42497305.0,
      "step": 19040
    },
    {
      "entropy": 5.482597923278808,
      "epoch": 2.138934127028697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.1918,
      "mean_token_accuracy": 0.1993209093809128,
      "num_tokens": 42508307.0,
      "step": 19045
    },
    {
      "entropy": 5.511053085327148,
      "epoch": 2.139495703936654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19159839451313018,
      "num_tokens": 42519946.0,
      "step": 19050
    },
    {
      "entropy": 5.40810375213623,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045489119451778645,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20282476991415024,
      "num_tokens": 42530737.0,
      "step": 19055
    },
    {
      "entropy": 5.332473802566528,
      "epoch": 2.1406188577525693,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004548670765574908,
      "loss": 5.0351,
      "mean_token_accuracy": 0.20721573382616043,
      "num_tokens": 42542184.0,
      "step": 19060
    },
    {
      "entropy": 5.4416738986969,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.1605,
      "mean_token_accuracy": 0.19836343228816986,
      "num_tokens": 42553816.0,
      "step": 19065
    },
    {
      "entropy": 5.536497449874878,
      "epoch": 2.141742011568484,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.1809,
      "mean_token_accuracy": 0.2028401955962181,
      "num_tokens": 42564685.0,
      "step": 19070
    },
    {
      "entropy": 5.420281267166137,
      "epoch": 2.142303588476442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.1138,
      "mean_token_accuracy": 0.19693809300661086,
      "num_tokens": 42574864.0,
      "step": 19075
    },
    {
      "entropy": 5.410591459274292,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004547705474536152,
      "loss": 5.1006,
      "mean_token_accuracy": 0.19923099428415297,
      "num_tokens": 42585214.0,
      "step": 19080
    },
    {
      "entropy": 5.495586776733399,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19812824130058287,
      "num_tokens": 42596017.0,
      "step": 19085
    },
    {
      "entropy": 5.453445863723755,
      "epoch": 2.1439883192003144,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.12,
      "mean_token_accuracy": 0.19889673888683318,
      "num_tokens": 42607903.0,
      "step": 19090
    },
    {
      "entropy": 5.411392307281494,
      "epoch": 2.144549896108272,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20624203085899354,
      "num_tokens": 42618690.0,
      "step": 19095
    },
    {
      "entropy": 5.427911186218262,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045467392676632016,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2038321852684021,
      "num_tokens": 42629655.0,
      "step": 19100
    },
    {
      "entropy": 5.436651659011841,
      "epoch": 2.145673049924187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.2117,
      "mean_token_accuracy": 0.20069251209497452,
      "num_tokens": 42641214.0,
      "step": 19105
    },
    {
      "entropy": 5.506593036651611,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.2505,
      "mean_token_accuracy": 0.19528359025716782,
      "num_tokens": 42653829.0,
      "step": 19110
    },
    {
      "entropy": 5.501069784164429,
      "epoch": 2.1467962037401023,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1994260460138321,
      "num_tokens": 42665245.0,
      "step": 19115
    },
    {
      "entropy": 5.4487401962280275,
      "epoch": 2.14735778064806,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004545772145448284,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2001232087612152,
      "num_tokens": 42676409.0,
      "step": 19120
    },
    {
      "entropy": 5.3757185459136965,
      "epoch": 2.147919357556017,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045455302219301,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21238219887018203,
      "num_tokens": 42686586.0,
      "step": 19125
    },
    {
      "entropy": 5.359950590133667,
      "epoch": 2.148480934463975,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545288241241538,
      "loss": 5.0593,
      "mean_token_accuracy": 0.205304753780365,
      "num_tokens": 42696597.0,
      "step": 19130
    },
    {
      "entropy": 5.510161638259888,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.2398,
      "mean_token_accuracy": 0.20513799488544465,
      "num_tokens": 42707447.0,
      "step": 19135
    },
    {
      "entropy": 5.569417095184326,
      "epoch": 2.14960408827989,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.2434,
      "mean_token_accuracy": 0.19274359047412873,
      "num_tokens": 42717533.0,
      "step": 19140
    },
    {
      "entropy": 5.52343339920044,
      "epoch": 2.1501656651878474,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.2365,
      "mean_token_accuracy": 0.19718028604984283,
      "num_tokens": 42730563.0,
      "step": 19145
    },
    {
      "entropy": 5.430177402496338,
      "epoch": 2.150727242095805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.1106,
      "mean_token_accuracy": 0.20820065587759018,
      "num_tokens": 42740301.0,
      "step": 19150
    },
    {
      "entropy": 5.434264612197876,
      "epoch": 2.1512888190037627,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.1681,
      "mean_token_accuracy": 0.2004183828830719,
      "num_tokens": 42752296.0,
      "step": 19155
    },
    {
      "entropy": 5.372303581237793,
      "epoch": 2.15185039591172,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045438351569638055,
      "loss": 5.0297,
      "mean_token_accuracy": 0.2091466799378395,
      "num_tokens": 42762530.0,
      "step": 19160
    },
    {
      "entropy": 5.421622610092163,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045435927762984277,
      "loss": 5.0937,
      "mean_token_accuracy": 0.20492661148309707,
      "num_tokens": 42773059.0,
      "step": 19165
    },
    {
      "entropy": 5.432879638671875,
      "epoch": 2.152973549727635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000454335033852436,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20628607273101807,
      "num_tokens": 42783832.0,
      "step": 19170
    },
    {
      "entropy": 5.458415174484253,
      "epoch": 2.153535126635593,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.2261,
      "mean_token_accuracy": 0.20346362590789796,
      "num_tokens": 42794565.0,
      "step": 19175
    },
    {
      "entropy": 5.5305615901947025,
      "epoch": 2.15409670354355,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.2683,
      "mean_token_accuracy": 0.19414251148700715,
      "num_tokens": 42805536.0,
      "step": 19180
    },
    {
      "entropy": 5.45842490196228,
      "epoch": 2.1546582804515078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.1709,
      "mean_token_accuracy": 0.2008275121450424,
      "num_tokens": 42816718.0,
      "step": 19185
    },
    {
      "entropy": 5.465261745452881,
      "epoch": 2.1552198573594654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.1533,
      "mean_token_accuracy": 0.20271989703178406,
      "num_tokens": 42830183.0,
      "step": 19190
    },
    {
      "entropy": 5.423596811294556,
      "epoch": 2.155781434267423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20599999725818635,
      "num_tokens": 42841241.0,
      "step": 19195
    },
    {
      "entropy": 5.577798175811767,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.2192,
      "mean_token_accuracy": 0.20121305137872697,
      "num_tokens": 42852775.0,
      "step": 19200
    },
    {
      "entropy": 5.624221086502075,
      "epoch": 2.156904588083338,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.3359,
      "mean_token_accuracy": 0.19158112108707429,
      "num_tokens": 42863947.0,
      "step": 19205
    },
    {
      "entropy": 5.430298233032227,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.184,
      "mean_token_accuracy": 0.20596070289611818,
      "num_tokens": 42874794.0,
      "step": 19210
    },
    {
      "entropy": 5.457150936126709,
      "epoch": 2.1580277418992533,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.1634,
      "mean_token_accuracy": 0.201410935819149,
      "num_tokens": 42886038.0,
      "step": 19215
    },
    {
      "entropy": 5.454669713973999,
      "epoch": 2.1585893188072105,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19576012194156647,
      "num_tokens": 42897243.0,
      "step": 19220
    },
    {
      "entropy": 5.437304878234864,
      "epoch": 2.159150895715168,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004540679756045125,
      "loss": 5.0714,
      "mean_token_accuracy": 0.20804495215415955,
      "num_tokens": 42908802.0,
      "step": 19225
    },
    {
      "entropy": 5.512308835983276,
      "epoch": 2.159712472623126,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.2601,
      "mean_token_accuracy": 0.19065648466348648,
      "num_tokens": 42921156.0,
      "step": 19230
    },
    {
      "entropy": 5.3409342765808105,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.125,
      "learning_rate": 0.00045401934540854937,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20856190621852874,
      "num_tokens": 42931336.0,
      "step": 19235
    },
    {
      "entropy": 5.44011697769165,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20694223344326018,
      "num_tokens": 42942312.0,
      "step": 19240
    },
    {
      "entropy": 5.463801765441895,
      "epoch": 2.1613972033469984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.215,
      "mean_token_accuracy": 0.20644889920949935,
      "num_tokens": 42953429.0,
      "step": 19245
    },
    {
      "entropy": 5.376381969451904,
      "epoch": 2.161958780254956,
      "grad_norm": 1.171875,
      "learning_rate": 0.000453946357366201,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20865806192159653,
      "num_tokens": 42963354.0,
      "step": 19250
    },
    {
      "entropy": 5.529055786132813,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.221,
      "mean_token_accuracy": 0.20140518099069596,
      "num_tokens": 42974643.0,
      "step": 19255
    },
    {
      "entropy": 5.463166189193726,
      "epoch": 2.163081934070871,
      "grad_norm": 1.125,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.1989,
      "mean_token_accuracy": 0.2009876325726509,
      "num_tokens": 42985821.0,
      "step": 19260
    },
    {
      "entropy": 5.470652151107788,
      "epoch": 2.1636435109788286,
      "grad_norm": 1.0,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19452736228704454,
      "num_tokens": 42998258.0,
      "step": 19265
    },
    {
      "entropy": 5.449978971481324,
      "epoch": 2.1642050878867862,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.125,
      "mean_token_accuracy": 0.20073011368513108,
      "num_tokens": 43009117.0,
      "step": 19270
    },
    {
      "entropy": 5.408582592010498,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.1316,
      "mean_token_accuracy": 0.2023191884160042,
      "num_tokens": 43020437.0,
      "step": 19275
    },
    {
      "entropy": 5.489286994934082,
      "epoch": 2.165328241702701,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.1882,
      "mean_token_accuracy": 0.19996050745248795,
      "num_tokens": 43031722.0,
      "step": 19280
    },
    {
      "entropy": 5.434236526489258,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045377585254998743,
      "loss": 5.1086,
      "mean_token_accuracy": 0.21206972002983093,
      "num_tokens": 43042957.0,
      "step": 19285
    },
    {
      "entropy": 5.409184885025025,
      "epoch": 2.1664513955186164,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.1764,
      "mean_token_accuracy": 0.19841236174106597,
      "num_tokens": 43054086.0,
      "step": 19290
    },
    {
      "entropy": 5.557629632949829,
      "epoch": 2.1670129724265736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.277,
      "mean_token_accuracy": 0.19409868568181993,
      "num_tokens": 43066259.0,
      "step": 19295
    },
    {
      "entropy": 5.37485933303833,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045370269365742143,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2103622615337372,
      "num_tokens": 43076837.0,
      "step": 19300
    },
    {
      "entropy": 5.363848924636841,
      "epoch": 2.168136126242489,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045367829597893776,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20678746104240417,
      "num_tokens": 43087802.0,
      "step": 19305
    },
    {
      "entropy": 5.423171138763427,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045365389261126793,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20238826423883438,
      "num_tokens": 43097562.0,
      "step": 19310
    },
    {
      "entropy": 5.345750617980957,
      "epoch": 2.169259280058404,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045362948355518883,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20817989110946655,
      "num_tokens": 43107973.0,
      "step": 19315
    },
    {
      "entropy": 5.322268772125244,
      "epoch": 2.1698208569663615,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004536050688114777,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2148093357682228,
      "num_tokens": 43118923.0,
      "step": 19320
    },
    {
      "entropy": 5.391568899154663,
      "epoch": 2.170382433874319,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045358064838091187,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20451484322547914,
      "num_tokens": 43130904.0,
      "step": 19325
    },
    {
      "entropy": 5.472601747512817,
      "epoch": 2.170944010782277,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045355622226426903,
      "loss": 5.1179,
      "mean_token_accuracy": 0.1967015326023102,
      "num_tokens": 43142293.0,
      "step": 19330
    },
    {
      "entropy": 5.474063920974731,
      "epoch": 2.171505587690234,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20847147852182388,
      "num_tokens": 43154340.0,
      "step": 19335
    },
    {
      "entropy": 5.421504926681519,
      "epoch": 2.1720671645981917,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000453507352975863,
      "loss": 5.1276,
      "mean_token_accuracy": 0.2093738466501236,
      "num_tokens": 43165924.0,
      "step": 19340
    },
    {
      "entropy": 5.33529577255249,
      "epoch": 2.1726287415061494,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045348290980565595,
      "loss": 5.0823,
      "mean_token_accuracy": 0.2024005651473999,
      "num_tokens": 43177394.0,
      "step": 19345
    },
    {
      "entropy": 5.402228498458863,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045345846095248376,
      "loss": 5.0617,
      "mean_token_accuracy": 0.2059154614806175,
      "num_tokens": 43188043.0,
      "step": 19350
    },
    {
      "entropy": 5.49878249168396,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.2633,
      "mean_token_accuracy": 0.1958798125386238,
      "num_tokens": 43199466.0,
      "step": 19355
    },
    {
      "entropy": 5.389993667602539,
      "epoch": 2.174313472230022,
      "grad_norm": 1.25,
      "learning_rate": 0.00045340954620035827,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20503222346305847,
      "num_tokens": 43210673.0,
      "step": 19360
    },
    {
      "entropy": 5.459617710113525,
      "epoch": 2.1748750491379796,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20423915684223176,
      "num_tokens": 43221263.0,
      "step": 19365
    },
    {
      "entropy": 5.477347183227539,
      "epoch": 2.175436626045937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20773029774427415,
      "num_tokens": 43232590.0,
      "step": 19370
    },
    {
      "entropy": 5.52472882270813,
      "epoch": 2.1759982029538945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19878347367048263,
      "num_tokens": 43243693.0,
      "step": 19375
    },
    {
      "entropy": 5.485237693786621,
      "epoch": 2.176559779861852,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045331164853479084,
      "loss": 5.1264,
      "mean_token_accuracy": 0.2032419040799141,
      "num_tokens": 43254526.0,
      "step": 19380
    },
    {
      "entropy": 5.555027914047241,
      "epoch": 2.17712135676981,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.2165,
      "mean_token_accuracy": 0.20115191340446473,
      "num_tokens": 43264490.0,
      "step": 19385
    },
    {
      "entropy": 5.448733234405518,
      "epoch": 2.177682933677767,
      "grad_norm": 1.125,
      "learning_rate": 0.00045326266563381734,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20167769938707353,
      "num_tokens": 43275012.0,
      "step": 19390
    },
    {
      "entropy": 5.423792457580566,
      "epoch": 2.1782445105857247,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045323816566901205,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20693261325359344,
      "num_tokens": 43285194.0,
      "step": 19395
    },
    {
      "entropy": 5.436718845367432,
      "epoch": 2.1788060874936823,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20652735829353333,
      "num_tokens": 43298159.0,
      "step": 19400
    },
    {
      "entropy": 5.476078653335572,
      "epoch": 2.17936766440164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.2199,
      "mean_token_accuracy": 0.20245183110237122,
      "num_tokens": 43309664.0,
      "step": 19405
    },
    {
      "entropy": 5.466586685180664,
      "epoch": 2.179929241309597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.1953,
      "mean_token_accuracy": 0.19685198366641998,
      "num_tokens": 43320399.0,
      "step": 19410
    },
    {
      "entropy": 5.406252002716064,
      "epoch": 2.180490818217555,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045314010906587096,
      "loss": 5.0155,
      "mean_token_accuracy": 0.20388790518045424,
      "num_tokens": 43331510.0,
      "step": 19415
    },
    {
      "entropy": 5.513053512573242,
      "epoch": 2.1810523951255125,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.2401,
      "mean_token_accuracy": 0.20349526554346084,
      "num_tokens": 43342844.0,
      "step": 19420
    },
    {
      "entropy": 5.371338891983032,
      "epoch": 2.18161397203347,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004530910467288769,
      "loss": 5.0994,
      "mean_token_accuracy": 0.19907493740320206,
      "num_tokens": 43352556.0,
      "step": 19425
    },
    {
      "entropy": 5.382122325897217,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045306650705425727,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20880690962076187,
      "num_tokens": 43363357.0,
      "step": 19430
    },
    {
      "entropy": 5.4958654880523685,
      "epoch": 2.182737125849385,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20085732787847518,
      "num_tokens": 43374697.0,
      "step": 19435
    },
    {
      "entropy": 5.522832632064819,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.2239,
      "mean_token_accuracy": 0.19652022123336793,
      "num_tokens": 43386826.0,
      "step": 19440
    },
    {
      "entropy": 5.41835298538208,
      "epoch": 2.1838602796653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19774671792984008,
      "num_tokens": 43398352.0,
      "step": 19445
    },
    {
      "entropy": 5.46236629486084,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20672448724508286,
      "num_tokens": 43409509.0,
      "step": 19450
    },
    {
      "entropy": 5.403712463378906,
      "epoch": 2.1849834334812153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045294372365119276,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20152269005775453,
      "num_tokens": 43420195.0,
      "step": 19455
    },
    {
      "entropy": 5.454810953140258,
      "epoch": 2.185545010389173,
      "grad_norm": 1.25,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.1976,
      "mean_token_accuracy": 0.20226840525865555,
      "num_tokens": 43430599.0,
      "step": 19460
    },
    {
      "entropy": 5.382679986953735,
      "epoch": 2.18610658729713,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004528945706239122,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2047211453318596,
      "num_tokens": 43442212.0,
      "step": 19465
    },
    {
      "entropy": 5.39938530921936,
      "epoch": 2.186668164205088,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045286998561353236,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20168321579694748,
      "num_tokens": 43452631.0,
      "step": 19470
    },
    {
      "entropy": 5.393381214141845,
      "epoch": 2.1872297411130455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.135,
      "mean_token_accuracy": 0.2065647765994072,
      "num_tokens": 43463323.0,
      "step": 19475
    },
    {
      "entropy": 5.411081790924072,
      "epoch": 2.187791318021003,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.1906,
      "mean_token_accuracy": 0.19413543343544007,
      "num_tokens": 43474451.0,
      "step": 19480
    },
    {
      "entropy": 5.40665373802185,
      "epoch": 2.1883528949289603,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045279619660482925,
      "loss": 5.049,
      "mean_token_accuracy": 0.2050679862499237,
      "num_tokens": 43485948.0,
      "step": 19485
    },
    {
      "entropy": 5.513670444488525,
      "epoch": 2.188914471836918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.1395,
      "mean_token_accuracy": 0.2081505224108696,
      "num_tokens": 43498955.0,
      "step": 19490
    },
    {
      "entropy": 5.462107992172241,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.125,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.0986,
      "mean_token_accuracy": 0.21066556572914125,
      "num_tokens": 43509416.0,
      "step": 19495
    },
    {
      "entropy": 5.3991227626800535,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.1346,
      "mean_token_accuracy": 0.2058756023645401,
      "num_tokens": 43520387.0,
      "step": 19500
    },
    {
      "entropy": 5.378425455093383,
      "epoch": 2.1905992025607905,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.1408,
      "mean_token_accuracy": 0.2011588528752327,
      "num_tokens": 43532006.0,
      "step": 19505
    },
    {
      "entropy": 5.381918573379517,
      "epoch": 2.191160779468748,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004526731017121111,
      "loss": 5.0554,
      "mean_token_accuracy": 0.2088074192404747,
      "num_tokens": 43542303.0,
      "step": 19510
    },
    {
      "entropy": 5.509841871261597,
      "epoch": 2.191722356376706,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.2338,
      "mean_token_accuracy": 0.19600335359573365,
      "num_tokens": 43553711.0,
      "step": 19515
    },
    {
      "entropy": 5.386949729919434,
      "epoch": 2.1922839332846635,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045262382414927235,
      "loss": 5.087,
      "mean_token_accuracy": 0.20398693531751633,
      "num_tokens": 43564488.0,
      "step": 19520
    },
    {
      "entropy": 5.415289497375488,
      "epoch": 2.1928455101926208,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045259917688404336,
      "loss": 5.1071,
      "mean_token_accuracy": 0.20185879468917847,
      "num_tokens": 43575414.0,
      "step": 19525
    },
    {
      "entropy": 5.369929265975952,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004525745239639877,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20495269149541856,
      "num_tokens": 43586789.0,
      "step": 19530
    },
    {
      "entropy": 5.4115245819091795,
      "epoch": 2.193968664008536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045254986538989045,
      "loss": 5.0789,
      "mean_token_accuracy": 0.2147873729467392,
      "num_tokens": 43597423.0,
      "step": 19535
    },
    {
      "entropy": 5.412087249755859,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004525252011625366,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20439665019512177,
      "num_tokens": 43606703.0,
      "step": 19540
    },
    {
      "entropy": 5.481432294845581,
      "epoch": 2.195091817824451,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19984448403120042,
      "num_tokens": 43617869.0,
      "step": 19545
    },
    {
      "entropy": 5.532707166671753,
      "epoch": 2.1956533947324086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.2931,
      "mean_token_accuracy": 0.20148752629756927,
      "num_tokens": 43629102.0,
      "step": 19550
    },
    {
      "entropy": 5.465566396713257,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004524511745687895,
      "loss": 5.182,
      "mean_token_accuracy": 0.20886197239160537,
      "num_tokens": 43640421.0,
      "step": 19555
    },
    {
      "entropy": 5.457819938659668,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004524264877362643,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20673491507768632,
      "num_tokens": 43650516.0,
      "step": 19560
    },
    {
      "entropy": 5.469805765151977,
      "epoch": 2.197338125456281,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045240179525441094,
      "loss": 5.1107,
      "mean_token_accuracy": 0.2096590980887413,
      "num_tokens": 43661439.0,
      "step": 19565
    },
    {
      "entropy": 5.436346387863159,
      "epoch": 2.197899702364239,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.185,
      "mean_token_accuracy": 0.2025914564728737,
      "num_tokens": 43672137.0,
      "step": 19570
    },
    {
      "entropy": 5.444410514831543,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.1727,
      "mean_token_accuracy": 0.19978250563144684,
      "num_tokens": 43683612.0,
      "step": 19575
    },
    {
      "entropy": 5.432463073730469,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.1131,
      "mean_token_accuracy": 0.2059452474117279,
      "num_tokens": 43694331.0,
      "step": 19580
    },
    {
      "entropy": 5.478057479858398,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.2169,
      "mean_token_accuracy": 0.19790075719356537,
      "num_tokens": 43706572.0,
      "step": 19585
    },
    {
      "entropy": 5.391623020172119,
      "epoch": 2.200146009996069,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045227824813274517,
      "loss": 5.0849,
      "mean_token_accuracy": 0.2133317321538925,
      "num_tokens": 43718472.0,
      "step": 19590
    },
    {
      "entropy": 5.469793748855591,
      "epoch": 2.2007075869040267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.2013,
      "mean_token_accuracy": 0.20271628350019455,
      "num_tokens": 43729642.0,
      "step": 19595
    },
    {
      "entropy": 5.479317569732666,
      "epoch": 2.201269163811984,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19823446571826936,
      "num_tokens": 43739846.0,
      "step": 19600
    },
    {
      "entropy": 5.382539463043213,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045220405211815327,
      "loss": 5.0376,
      "mean_token_accuracy": 0.20500393807888032,
      "num_tokens": 43750265.0,
      "step": 19605
    },
    {
      "entropy": 5.440049457550049,
      "epoch": 2.202392317627899,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.1407,
      "mean_token_accuracy": 0.1977613463997841,
      "num_tokens": 43761245.0,
      "step": 19610
    },
    {
      "entropy": 5.522872447967529,
      "epoch": 2.202953894535857,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.151,
      "mean_token_accuracy": 0.20322937816381453,
      "num_tokens": 43772677.0,
      "step": 19615
    },
    {
      "entropy": 5.490469741821289,
      "epoch": 2.203515471443814,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.2224,
      "mean_token_accuracy": 0.19638148099184036,
      "num_tokens": 43783128.0,
      "step": 19620
    },
    {
      "entropy": 5.378346920013428,
      "epoch": 2.2040770483517718,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.1505,
      "mean_token_accuracy": 0.2051555410027504,
      "num_tokens": 43795323.0,
      "step": 19625
    },
    {
      "entropy": 5.495058298110962,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.2309,
      "mean_token_accuracy": 0.19798728674650193,
      "num_tokens": 43806917.0,
      "step": 19630
    },
    {
      "entropy": 5.578366374969482,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.2118,
      "mean_token_accuracy": 0.2005753293633461,
      "num_tokens": 43818963.0,
      "step": 19635
    },
    {
      "entropy": 5.39679536819458,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004520307306422844,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20616210103034974,
      "num_tokens": 43830163.0,
      "step": 19640
    },
    {
      "entropy": 5.359626770019531,
      "epoch": 2.206323355983602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045200594787822795,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20827651470899583,
      "num_tokens": 43840434.0,
      "step": 19645
    },
    {
      "entropy": 5.326594877243042,
      "epoch": 2.2068849328915596,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004519811594782335,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21284874975681306,
      "num_tokens": 43851086.0,
      "step": 19650
    },
    {
      "entropy": 5.438270330429077,
      "epoch": 2.207446509799517,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.1779,
      "mean_token_accuracy": 0.1964749351143837,
      "num_tokens": 43862834.0,
      "step": 19655
    },
    {
      "entropy": 5.41281156539917,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19788086116313935,
      "num_tokens": 43874868.0,
      "step": 19660
    },
    {
      "entropy": 5.498284864425659,
      "epoch": 2.208569663615432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.1537,
      "mean_token_accuracy": 0.20734975934028627,
      "num_tokens": 43886175.0,
      "step": 19665
    },
    {
      "entropy": 5.468722200393676,
      "epoch": 2.20913124052339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045188194953466363,
      "loss": 5.1127,
      "mean_token_accuracy": 0.20881798565387727,
      "num_tokens": 43896489.0,
      "step": 19670
    },
    {
      "entropy": 5.39941668510437,
      "epoch": 2.209692817431347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045185713296682175,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20671319514513015,
      "num_tokens": 43907759.0,
      "step": 19675
    },
    {
      "entropy": 5.396782732009887,
      "epoch": 2.2102543943393047,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.1138,
      "mean_token_accuracy": 0.201300710439682,
      "num_tokens": 43918136.0,
      "step": 19680
    },
    {
      "entropy": 5.474825096130371,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.1749,
      "mean_token_accuracy": 0.19686809033155442,
      "num_tokens": 43929662.0,
      "step": 19685
    },
    {
      "entropy": 5.466724920272827,
      "epoch": 2.21137754815522,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20807749927043914,
      "num_tokens": 43939599.0,
      "step": 19690
    },
    {
      "entropy": 5.400431394577026,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045175781039136133,
      "loss": 5.0902,
      "mean_token_accuracy": 0.2035491183400154,
      "num_tokens": 43949728.0,
      "step": 19695
    },
    {
      "entropy": 5.467465591430664,
      "epoch": 2.212500701971135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.2018,
      "mean_token_accuracy": 0.20216406881809235,
      "num_tokens": 43961147.0,
      "step": 19700
    },
    {
      "entropy": 5.495803165435791,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.1897,
      "mean_token_accuracy": 0.20225719809532167,
      "num_tokens": 43972541.0,
      "step": 19705
    },
    {
      "entropy": 5.446324872970581,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19894807785749435,
      "num_tokens": 43983540.0,
      "step": 19710
    },
    {
      "entropy": 5.404163122177124,
      "epoch": 2.2141854326950074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004516583977673087,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21014162600040437,
      "num_tokens": 43994255.0,
      "step": 19715
    },
    {
      "entropy": 5.527474355697632,
      "epoch": 2.214747009602965,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.2451,
      "mean_token_accuracy": 0.19958435446023942,
      "num_tokens": 44004364.0,
      "step": 19720
    },
    {
      "entropy": 5.458494710922241,
      "epoch": 2.2153085865109228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.2164,
      "mean_token_accuracy": 0.20162972211837768,
      "num_tokens": 44015487.0,
      "step": 19725
    },
    {
      "entropy": 5.521120691299439,
      "epoch": 2.21587016341888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.188,
      "mean_token_accuracy": 0.20811489671468736,
      "num_tokens": 44026974.0,
      "step": 19730
    },
    {
      "entropy": 5.482464218139649,
      "epoch": 2.2164317403268377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.1441,
      "mean_token_accuracy": 0.20208192765712737,
      "num_tokens": 44038718.0,
      "step": 19735
    },
    {
      "entropy": 5.390679168701172,
      "epoch": 2.2169933172347953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004515340054335696,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2050388216972351,
      "num_tokens": 44048821.0,
      "step": 19740
    },
    {
      "entropy": 5.472933292388916,
      "epoch": 2.217554894142753,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19522123485803605,
      "num_tokens": 44060973.0,
      "step": 19745
    },
    {
      "entropy": 5.593734645843506,
      "epoch": 2.21811647105071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.1967,
      "mean_token_accuracy": 0.20391546189785004,
      "num_tokens": 44073595.0,
      "step": 19750
    },
    {
      "entropy": 5.401270771026612,
      "epoch": 2.218678047958668,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.1146,
      "mean_token_accuracy": 0.2045004889369011,
      "num_tokens": 44084860.0,
      "step": 19755
    },
    {
      "entropy": 5.425442790985107,
      "epoch": 2.2192396248666255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.224,
      "mean_token_accuracy": 0.1957052990794182,
      "num_tokens": 44096026.0,
      "step": 19760
    },
    {
      "entropy": 5.473255777359009,
      "epoch": 2.219801201774583,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.1392,
      "mean_token_accuracy": 0.2014969527721405,
      "num_tokens": 44107981.0,
      "step": 19765
    },
    {
      "entropy": 5.4604541778564455,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.1949,
      "mean_token_accuracy": 0.19833221584558486,
      "num_tokens": 44119589.0,
      "step": 19770
    },
    {
      "entropy": 5.420510101318359,
      "epoch": 2.220924355590498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.1749,
      "mean_token_accuracy": 0.20289915651082993,
      "num_tokens": 44131094.0,
      "step": 19775
    },
    {
      "entropy": 5.441696643829346,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.1622,
      "mean_token_accuracy": 0.19940573573112488,
      "num_tokens": 44141348.0,
      "step": 19780
    },
    {
      "entropy": 5.579134464263916,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19988732784986496,
      "num_tokens": 44152085.0,
      "step": 19785
    },
    {
      "entropy": 5.468767595291138,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045128479929690756,
      "loss": 5.088,
      "mean_token_accuracy": 0.21069650053977967,
      "num_tokens": 44163518.0,
      "step": 19790
    },
    {
      "entropy": 5.380663156509399,
      "epoch": 2.2231706632223283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.1334,
      "mean_token_accuracy": 0.19633735716342926,
      "num_tokens": 44175513.0,
      "step": 19795
    },
    {
      "entropy": 5.469863700866699,
      "epoch": 2.223732240130286,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.1319,
      "mean_token_accuracy": 0.2018109068274498,
      "num_tokens": 44186660.0,
      "step": 19800
    },
    {
      "entropy": 5.483963251113892,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.213,
      "mean_token_accuracy": 0.19745038002729415,
      "num_tokens": 44197613.0,
      "step": 19805
    },
    {
      "entropy": 5.407143449783325,
      "epoch": 2.224855393946201,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.1335,
      "mean_token_accuracy": 0.1987305238842964,
      "num_tokens": 44209469.0,
      "step": 19810
    },
    {
      "entropy": 5.473362445831299,
      "epoch": 2.2254169708541585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1981020152568817,
      "num_tokens": 44220672.0,
      "step": 19815
    },
    {
      "entropy": 5.480352973937988,
      "epoch": 2.225978547762116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1967556044459343,
      "num_tokens": 44232867.0,
      "step": 19820
    },
    {
      "entropy": 5.4493903636932375,
      "epoch": 2.2265401246700733,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20165464133024216,
      "num_tokens": 44245567.0,
      "step": 19825
    },
    {
      "entropy": 5.435141754150391,
      "epoch": 2.227101701578031,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045108503021523737,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20610821545124053,
      "num_tokens": 44257449.0,
      "step": 19830
    },
    {
      "entropy": 5.459766530990601,
      "epoch": 2.2276632784859887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045106003384190604,
      "loss": 5.2096,
      "mean_token_accuracy": 0.197022745013237,
      "num_tokens": 44268890.0,
      "step": 19835
    },
    {
      "entropy": 5.532184362411499,
      "epoch": 2.2282248553939463,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19489402323961258,
      "num_tokens": 44280719.0,
      "step": 19840
    },
    {
      "entropy": 5.549978542327881,
      "epoch": 2.2287864323019035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.2616,
      "mean_token_accuracy": 0.19264942705631255,
      "num_tokens": 44292567.0,
      "step": 19845
    },
    {
      "entropy": 5.465829467773437,
      "epoch": 2.229348009209861,
      "grad_norm": 1.25,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1947641223669052,
      "num_tokens": 44304177.0,
      "step": 19850
    },
    {
      "entropy": 5.437126684188843,
      "epoch": 2.229909586117819,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045095999229834437,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2043955832719803,
      "num_tokens": 44315142.0,
      "step": 19855
    },
    {
      "entropy": 5.431200265884399,
      "epoch": 2.2304711630257765,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.249,
      "mean_token_accuracy": 0.20043786913156508,
      "num_tokens": 44327457.0,
      "step": 19860
    },
    {
      "entropy": 5.552132654190063,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.23,
      "mean_token_accuracy": 0.19535116851329803,
      "num_tokens": 44338978.0,
      "step": 19865
    },
    {
      "entropy": 5.5329993724822994,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.2273,
      "mean_token_accuracy": 0.201108855009079,
      "num_tokens": 44349121.0,
      "step": 19870
    },
    {
      "entropy": 5.397964572906494,
      "epoch": 2.232155893749649,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508598611126359,
      "loss": 5.1004,
      "mean_token_accuracy": 0.2088798239827156,
      "num_tokens": 44360753.0,
      "step": 19875
    },
    {
      "entropy": 5.479589796066284,
      "epoch": 2.2327174706576067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.1646,
      "mean_token_accuracy": 0.19993786960840226,
      "num_tokens": 44371977.0,
      "step": 19880
    },
    {
      "entropy": 5.42446346282959,
      "epoch": 2.233279047565564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045080976191991587,
      "loss": 5.0802,
      "mean_token_accuracy": 0.2039664462208748,
      "num_tokens": 44383077.0,
      "step": 19885
    },
    {
      "entropy": 5.457240438461303,
      "epoch": 2.2338406244735216,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004507847039263805,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2009204939007759,
      "num_tokens": 44394827.0,
      "step": 19890
    },
    {
      "entropy": 5.401480388641358,
      "epoch": 2.2344022013814793,
      "grad_norm": 1.203125,
      "learning_rate": 0.000450759640335792,
      "loss": 5.0796,
      "mean_token_accuracy": 0.208608914911747,
      "num_tokens": 44406594.0,
      "step": 19895
    },
    {
      "entropy": 5.3781944751739506,
      "epoch": 2.234963778289437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507345711489484,
      "loss": 5.0789,
      "mean_token_accuracy": 0.2009037658572197,
      "num_tokens": 44417842.0,
      "step": 19900
    },
    {
      "entropy": 5.47165584564209,
      "epoch": 2.235525355197394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.123,
      "mean_token_accuracy": 0.20028728544712066,
      "num_tokens": 44428234.0,
      "step": 19905
    },
    {
      "entropy": 5.384031295776367,
      "epoch": 2.236086932105352,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004506844159896888,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20572000443935395,
      "num_tokens": 44437504.0,
      "step": 19910
    },
    {
      "entropy": 5.457863998413086,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.1303,
      "mean_token_accuracy": 0.2080046519637108,
      "num_tokens": 44447873.0,
      "step": 19915
    },
    {
      "entropy": 5.464411640167237,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.2087,
      "mean_token_accuracy": 0.20213390290737152,
      "num_tokens": 44458990.0,
      "step": 19920
    },
    {
      "entropy": 5.466749906539917,
      "epoch": 2.2377716628292244,
      "grad_norm": 1.125,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.1104,
      "mean_token_accuracy": 0.20302632004022597,
      "num_tokens": 44469374.0,
      "step": 19925
    },
    {
      "entropy": 5.447497749328614,
      "epoch": 2.238333239737182,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20520556420087815,
      "num_tokens": 44479629.0,
      "step": 19930
    },
    {
      "entropy": 5.50403060913086,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.207,
      "mean_token_accuracy": 0.1963550940155983,
      "num_tokens": 44490603.0,
      "step": 19935
    },
    {
      "entropy": 5.3765510559082035,
      "epoch": 2.239456393553097,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004505338162848382,
      "loss": 5.073,
      "mean_token_accuracy": 0.2119968131184578,
      "num_tokens": 44501527.0,
      "step": 19940
    },
    {
      "entropy": 5.514767837524414,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19897997081279756,
      "num_tokens": 44512619.0,
      "step": 19945
    },
    {
      "entropy": 5.569184970855713,
      "epoch": 2.240579547369012,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19585637599229813,
      "num_tokens": 44524639.0,
      "step": 19950
    },
    {
      "entropy": 5.417506980895996,
      "epoch": 2.24114112427697,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.1243,
      "mean_token_accuracy": 0.21007796376943588,
      "num_tokens": 44535013.0,
      "step": 19955
    },
    {
      "entropy": 5.557058954238892,
      "epoch": 2.241702701184927,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.233,
      "mean_token_accuracy": 0.19973451793193817,
      "num_tokens": 44546838.0,
      "step": 19960
    },
    {
      "entropy": 5.443006944656372,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20768779814243316,
      "num_tokens": 44557984.0,
      "step": 19965
    },
    {
      "entropy": 5.442243766784668,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045038301537367374,
      "loss": 5.078,
      "mean_token_accuracy": 0.20558004230260848,
      "num_tokens": 44569029.0,
      "step": 19970
    },
    {
      "entropy": 5.426868629455567,
      "epoch": 2.2433874319088,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.1766,
      "mean_token_accuracy": 0.206084306538105,
      "num_tokens": 44578804.0,
      "step": 19975
    },
    {
      "entropy": 5.4826044082641605,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045033270372064123,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20685162395238876,
      "num_tokens": 44589362.0,
      "step": 19980
    },
    {
      "entropy": 5.4772521495819095,
      "epoch": 2.244510585724715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.1628,
      "mean_token_accuracy": 0.20179336816072463,
      "num_tokens": 44600137.0,
      "step": 19985
    },
    {
      "entropy": 5.413036584854126,
      "epoch": 2.2450721626326726,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.1514,
      "mean_token_accuracy": 0.21040593832731247,
      "num_tokens": 44611707.0,
      "step": 19990
    },
    {
      "entropy": 5.448164176940918,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.1368,
      "mean_token_accuracy": 0.2031902328133583,
      "num_tokens": 44621141.0,
      "step": 19995
    },
    {
      "entropy": 5.486722564697265,
      "epoch": 2.2461953164485875,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.1966,
      "mean_token_accuracy": 0.20007037818431855,
      "num_tokens": 44632043.0,
      "step": 20000
    },
    {
      "entropy": 5.515257024765015,
      "epoch": 2.246756893356545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.211,
      "mean_token_accuracy": 0.19908837527036666,
      "num_tokens": 44643905.0,
      "step": 20005
    },
    {
      "entropy": 5.40823450088501,
      "epoch": 2.247318470264503,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.1291,
      "mean_token_accuracy": 0.19696681797504426,
      "num_tokens": 44655105.0,
      "step": 20010
    },
    {
      "entropy": 5.408110475540161,
      "epoch": 2.24788004717246,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045015643712327465,
      "loss": 5.0951,
      "mean_token_accuracy": 0.20303869098424912,
      "num_tokens": 44667096.0,
      "step": 20015
    },
    {
      "entropy": 5.4923171043396,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.157,
      "mean_token_accuracy": 0.20112112462520598,
      "num_tokens": 44678206.0,
      "step": 20020
    },
    {
      "entropy": 5.425653123855591,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004501060250328419,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20635404735803603,
      "num_tokens": 44688619.0,
      "step": 20025
    },
    {
      "entropy": 5.392659759521484,
      "epoch": 2.249564777896333,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20224724858999252,
      "num_tokens": 44700757.0,
      "step": 20030
    },
    {
      "entropy": 5.34596815109253,
      "epoch": 2.2501263548042907,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004500555906406343,
      "loss": 5.003,
      "mean_token_accuracy": 0.21299721002578736,
      "num_tokens": 44710822.0,
      "step": 20035
    },
    {
      "entropy": 5.329952096939087,
      "epoch": 2.250687931712248,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004500303650833722,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2127438023686409,
      "num_tokens": 44722163.0,
      "step": 20040
    },
    {
      "entropy": 5.501678800582885,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.2003,
      "mean_token_accuracy": 0.1995450437068939,
      "num_tokens": 44734271.0,
      "step": 20045
    },
    {
      "entropy": 5.520967864990235,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.153,
      "mean_token_accuracy": 0.20573473870754241,
      "num_tokens": 44745345.0,
      "step": 20050
    },
    {
      "entropy": 5.448700714111328,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.1245,
      "mean_token_accuracy": 0.1958379179239273,
      "num_tokens": 44756986.0,
      "step": 20055
    },
    {
      "entropy": 5.366703033447266,
      "epoch": 2.252934239344078,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004499294071320112,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20842144787311553,
      "num_tokens": 44767259.0,
      "step": 20060
    },
    {
      "entropy": 5.403236007690429,
      "epoch": 2.2534958162520358,
      "grad_norm": 1.125,
      "learning_rate": 0.000449904153717611,
      "loss": 5.1658,
      "mean_token_accuracy": 0.2043090581893921,
      "num_tokens": 44778884.0,
      "step": 20065
    },
    {
      "entropy": 5.445444107055664,
      "epoch": 2.2540573931599934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004498788947341947,
      "loss": 5.0721,
      "mean_token_accuracy": 0.2082044005393982,
      "num_tokens": 44789216.0,
      "step": 20070
    },
    {
      "entropy": 5.337964963912964,
      "epoch": 2.2546189700679506,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004498536301825665,
      "loss": 5.024,
      "mean_token_accuracy": 0.2110508233308792,
      "num_tokens": 44799357.0,
      "step": 20075
    },
    {
      "entropy": 5.377265071868896,
      "epoch": 2.2551805469759083,
      "grad_norm": 1.03125,
      "learning_rate": 0.000449828360063531,
      "loss": 5.1641,
      "mean_token_accuracy": 0.20232641696929932,
      "num_tokens": 44811040.0,
      "step": 20080
    },
    {
      "entropy": 5.475870895385742,
      "epoch": 2.255742123883866,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044980308437789267,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19621861577033997,
      "num_tokens": 44822103.0,
      "step": 20085
    },
    {
      "entropy": 5.451181411743164,
      "epoch": 2.2563037007918236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.127,
      "mean_token_accuracy": 0.19825687557458876,
      "num_tokens": 44834712.0,
      "step": 20090
    },
    {
      "entropy": 5.434058809280396,
      "epoch": 2.256865277699781,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044975251631002694,
      "loss": 5.0669,
      "mean_token_accuracy": 0.20046408474445343,
      "num_tokens": 44845412.0,
      "step": 20095
    },
    {
      "entropy": 5.388689088821411,
      "epoch": 2.2574268546077385,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.1421,
      "mean_token_accuracy": 0.2022656410932541,
      "num_tokens": 44857720.0,
      "step": 20100
    },
    {
      "entropy": 5.388620758056641,
      "epoch": 2.257988431515696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.113,
      "mean_token_accuracy": 0.20880619436502457,
      "num_tokens": 44868253.0,
      "step": 20105
    },
    {
      "entropy": 5.472629404067993,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044967662247883295,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1986072525382042,
      "num_tokens": 44878525.0,
      "step": 20110
    },
    {
      "entropy": 5.365305233001709,
      "epoch": 2.259111585331611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044965131341048454,
      "loss": 5.079,
      "mean_token_accuracy": 0.20988816767930984,
      "num_tokens": 44889565.0,
      "step": 20115
    },
    {
      "entropy": 5.3405157089233395,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004496259987811707,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20656754821538925,
      "num_tokens": 44901191.0,
      "step": 20120
    },
    {
      "entropy": 5.412547588348389,
      "epoch": 2.2602347391475264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004496006785916973,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21028562039136886,
      "num_tokens": 44911483.0,
      "step": 20125
    },
    {
      "entropy": 5.5000035762786865,
      "epoch": 2.260796316055484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.2279,
      "mean_token_accuracy": 0.2033305749297142,
      "num_tokens": 44923537.0,
      "step": 20130
    },
    {
      "entropy": 5.494990682601928,
      "epoch": 2.2613578929634413,
      "grad_norm": 1.125,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.1746,
      "mean_token_accuracy": 0.19797872006893158,
      "num_tokens": 44935780.0,
      "step": 20135
    },
    {
      "entropy": 5.489409494400024,
      "epoch": 2.261919469871399,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.1553,
      "mean_token_accuracy": 0.20243533551692963,
      "num_tokens": 44946879.0,
      "step": 20140
    },
    {
      "entropy": 5.4829188823699955,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.1437,
      "mean_token_accuracy": 0.20465155690908432,
      "num_tokens": 44958385.0,
      "step": 20145
    },
    {
      "entropy": 5.450573396682739,
      "epoch": 2.263042623687314,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.2525,
      "mean_token_accuracy": 0.19533410370349885,
      "num_tokens": 44970597.0,
      "step": 20150
    },
    {
      "entropy": 5.454568576812744,
      "epoch": 2.2636042005952715,
      "grad_norm": 1.125,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.178,
      "mean_token_accuracy": 0.20250202864408492,
      "num_tokens": 44983123.0,
      "step": 20155
    },
    {
      "entropy": 5.4483579158782955,
      "epoch": 2.264165777503229,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044942328164866,
      "loss": 5.0974,
      "mean_token_accuracy": 0.19975200891494752,
      "num_tokens": 44993378.0,
      "step": 20160
    },
    {
      "entropy": 5.399605560302734,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004493979170069486,
      "loss": 5.0704,
      "mean_token_accuracy": 0.1999097764492035,
      "num_tokens": 45004905.0,
      "step": 20165
    },
    {
      "entropy": 5.380496597290039,
      "epoch": 2.265288931319144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20543215721845626,
      "num_tokens": 45016640.0,
      "step": 20170
    },
    {
      "entropy": 5.487663221359253,
      "epoch": 2.2658505082271017,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19706982225179673,
      "num_tokens": 45028285.0,
      "step": 20175
    },
    {
      "entropy": 5.501610612869262,
      "epoch": 2.2664120851350593,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19833078980445862,
      "num_tokens": 45038924.0,
      "step": 20180
    },
    {
      "entropy": 5.482291555404663,
      "epoch": 2.266973662043017,
      "grad_norm": 1.125,
      "learning_rate": 0.00044929640291920865,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20344796925783157,
      "num_tokens": 45050483.0,
      "step": 20185
    },
    {
      "entropy": 5.513018989562989,
      "epoch": 2.267535238950974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004492710105210904,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19606169760227204,
      "num_tokens": 45063242.0,
      "step": 20190
    },
    {
      "entropy": 5.3799660205841064,
      "epoch": 2.268096815858932,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.155,
      "mean_token_accuracy": 0.2003306970000267,
      "num_tokens": 45074057.0,
      "step": 20195
    },
    {
      "entropy": 5.414817190170288,
      "epoch": 2.2686583927668895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.2061,
      "mean_token_accuracy": 0.19339478462934495,
      "num_tokens": 45085488.0,
      "step": 20200
    },
    {
      "entropy": 5.51557149887085,
      "epoch": 2.2692199696748467,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.2609,
      "mean_token_accuracy": 0.19702999889850617,
      "num_tokens": 45096303.0,
      "step": 20205
    },
    {
      "entropy": 5.461765432357788,
      "epoch": 2.2697815465828044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044916938544813936,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20492334961891173,
      "num_tokens": 45106411.0,
      "step": 20210
    },
    {
      "entropy": 5.462426328659058,
      "epoch": 2.270343123490762,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20202705413103103,
      "num_tokens": 45117355.0,
      "step": 20215
    },
    {
      "entropy": 5.434532356262207,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044911853963470295,
      "loss": 5.074,
      "mean_token_accuracy": 0.21141235232353212,
      "num_tokens": 45127612.0,
      "step": 20220
    },
    {
      "entropy": 5.5082543849945065,
      "epoch": 2.2714662773066774,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.2505,
      "mean_token_accuracy": 0.20010817795991898,
      "num_tokens": 45138495.0,
      "step": 20225
    },
    {
      "entropy": 5.426121997833252,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044906767164525936,
      "loss": 5.0897,
      "mean_token_accuracy": 0.21036437153816223,
      "num_tokens": 45149727.0,
      "step": 20230
    },
    {
      "entropy": 5.414049911499023,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20095557421445848,
      "num_tokens": 45160921.0,
      "step": 20235
    },
    {
      "entropy": 5.396750402450562,
      "epoch": 2.27315100803055,
      "grad_norm": 1.0,
      "learning_rate": 0.0004490167814862872,
      "loss": 5.068,
      "mean_token_accuracy": 0.20542089641094208,
      "num_tokens": 45172607.0,
      "step": 20240
    },
    {
      "entropy": 5.5406591415405275,
      "epoch": 2.273712584938507,
      "grad_norm": 1.125,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.2551,
      "mean_token_accuracy": 0.20312247425317764,
      "num_tokens": 45186330.0,
      "step": 20245
    },
    {
      "entropy": 5.329312181472778,
      "epoch": 2.274274161846465,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004489658691642681,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2148903951048851,
      "num_tokens": 45198239.0,
      "step": 20250
    },
    {
      "entropy": 5.314783096313477,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044894040469414207,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20988162457942963,
      "num_tokens": 45209295.0,
      "step": 20255
    },
    {
      "entropy": 5.430111074447632,
      "epoch": 2.27539731566238,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20581330060958863,
      "num_tokens": 45220210.0,
      "step": 20260
    },
    {
      "entropy": 5.513519191741944,
      "epoch": 2.2759588925703373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20108020156621934,
      "num_tokens": 45231809.0,
      "step": 20265
    },
    {
      "entropy": 5.445365953445434,
      "epoch": 2.276520469478295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21129992455244065,
      "num_tokens": 45242847.0,
      "step": 20270
    },
    {
      "entropy": 5.338790941238403,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044883849143844914,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21474968791007995,
      "num_tokens": 45253357.0,
      "step": 20275
    },
    {
      "entropy": 5.320821952819824,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004488129992847848,
      "loss": 4.9942,
      "mean_token_accuracy": 0.20881016701459884,
      "num_tokens": 45264607.0,
      "step": 20280
    },
    {
      "entropy": 5.493633079528808,
      "epoch": 2.2782052002021675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.1753,
      "mean_token_accuracy": 0.20363058149814606,
      "num_tokens": 45275211.0,
      "step": 20285
    },
    {
      "entropy": 5.396355867385864,
      "epoch": 2.278766777110125,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004487619983754483,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20995334386825562,
      "num_tokens": 45286665.0,
      "step": 20290
    },
    {
      "entropy": 5.397709560394287,
      "epoch": 2.279328354018083,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.1663,
      "mean_token_accuracy": 0.2040874645113945,
      "num_tokens": 45298066.0,
      "step": 20295
    },
    {
      "entropy": 5.401922607421875,
      "epoch": 2.27988993092604,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1969730570912361,
      "num_tokens": 45308799.0,
      "step": 20300
    },
    {
      "entropy": 5.440746736526489,
      "epoch": 2.2804515078339977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004486854555186037,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21088502556085587,
      "num_tokens": 45319932.0,
      "step": 20305
    },
    {
      "entropy": 5.4936261653900145,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.16,
      "mean_token_accuracy": 0.20557641834020615,
      "num_tokens": 45330008.0,
      "step": 20310
    },
    {
      "entropy": 5.432768106460571,
      "epoch": 2.281574661649913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.1782,
      "mean_token_accuracy": 0.19589841216802598,
      "num_tokens": 45340877.0,
      "step": 20315
    },
    {
      "entropy": 5.430566358566284,
      "epoch": 2.2821362385578707,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2034818485379219,
      "num_tokens": 45351956.0,
      "step": 20320
    },
    {
      "entropy": 5.505562162399292,
      "epoch": 2.282697815465828,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.219,
      "mean_token_accuracy": 0.19854362457990646,
      "num_tokens": 45363133.0,
      "step": 20325
    },
    {
      "entropy": 5.402228689193725,
      "epoch": 2.2832593923737856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004485577734971247,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2046315237879753,
      "num_tokens": 45373917.0,
      "step": 20330
    },
    {
      "entropy": 5.471015977859497,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.125,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.1818,
      "mean_token_accuracy": 0.202048559486866,
      "num_tokens": 45385575.0,
      "step": 20335
    },
    {
      "entropy": 5.419047737121582,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044850666199981244,
      "loss": 5.0747,
      "mean_token_accuracy": 0.20632156878709793,
      "num_tokens": 45396615.0,
      "step": 20340
    },
    {
      "entropy": 5.452276515960693,
      "epoch": 2.284944123097658,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044848109796397035,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20425934344530106,
      "num_tokens": 45408931.0,
      "step": 20345
    },
    {
      "entropy": 5.493478202819825,
      "epoch": 2.285505700005616,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.1806,
      "mean_token_accuracy": 0.19871591329574584,
      "num_tokens": 45419993.0,
      "step": 20350
    },
    {
      "entropy": 5.45655255317688,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.2168,
      "mean_token_accuracy": 0.19570968747138978,
      "num_tokens": 45431889.0,
      "step": 20355
    },
    {
      "entropy": 5.385016918182373,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2062792956829071,
      "num_tokens": 45444233.0,
      "step": 20360
    },
    {
      "entropy": 5.399848318099975,
      "epoch": 2.2871904307294884,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004483787865916898,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20416233390569688,
      "num_tokens": 45455743.0,
      "step": 20365
    },
    {
      "entropy": 5.389883375167846,
      "epoch": 2.287752007637446,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044835319494546366,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20174830704927443,
      "num_tokens": 45466272.0,
      "step": 20370
    },
    {
      "entropy": 5.4742837905883786,
      "epoch": 2.2883135845454037,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044832759777960456,
      "loss": 5.1164,
      "mean_token_accuracy": 0.2006078764796257,
      "num_tokens": 45477828.0,
      "step": 20375
    },
    {
      "entropy": 5.4513427257537845,
      "epoch": 2.288875161453361,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004483019950949276,
      "loss": 5.1228,
      "mean_token_accuracy": 0.19857906848192214,
      "num_tokens": 45489315.0,
      "step": 20380
    },
    {
      "entropy": 5.347049140930176,
      "epoch": 2.2894367383613186,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004482763868922478,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21041259169578552,
      "num_tokens": 45499836.0,
      "step": 20385
    },
    {
      "entropy": 5.4062036037445065,
      "epoch": 2.289998315269276,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20286243855953218,
      "num_tokens": 45511106.0,
      "step": 20390
    },
    {
      "entropy": 5.409636354446411,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004482251539361418,
      "loss": 5.1086,
      "mean_token_accuracy": 0.2008928507566452,
      "num_tokens": 45521314.0,
      "step": 20395
    },
    {
      "entropy": 5.532837772369385,
      "epoch": 2.291121469085191,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20846298038959504,
      "num_tokens": 45531675.0,
      "step": 20400
    },
    {
      "entropy": 5.4842345237731935,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.182,
      "mean_token_accuracy": 0.20274983048439027,
      "num_tokens": 45542064.0,
      "step": 20405
    },
    {
      "entropy": 5.406715536117554,
      "epoch": 2.2922446229011064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20416887253522872,
      "num_tokens": 45553764.0,
      "step": 20410
    },
    {
      "entropy": 5.527762603759766,
      "epoch": 2.292806199809064,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044812262184378503,
      "loss": 5.1981,
      "mean_token_accuracy": 0.2025418683886528,
      "num_tokens": 45565302.0,
      "step": 20415
    },
    {
      "entropy": 5.339626455307007,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044809697503792647,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21230603456497193,
      "num_tokens": 45577165.0,
      "step": 20420
    },
    {
      "entropy": 5.41700119972229,
      "epoch": 2.293929353624979,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044807132272059306,
      "loss": 5.0864,
      "mean_token_accuracy": 0.20804390907287598,
      "num_tokens": 45587771.0,
      "step": 20425
    },
    {
      "entropy": 5.439433813095093,
      "epoch": 2.2944909305329366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004480456648926016,
      "loss": 5.1223,
      "mean_token_accuracy": 0.21467435508966445,
      "num_tokens": 45598815.0,
      "step": 20430
    },
    {
      "entropy": 5.4032196521759035,
      "epoch": 2.295052507440894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480200015547691,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20760321021080017,
      "num_tokens": 45609007.0,
      "step": 20435
    },
    {
      "entropy": 5.419494676589966,
      "epoch": 2.2956140843488515,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.1907,
      "mean_token_accuracy": 0.19920454621315004,
      "num_tokens": 45621323.0,
      "step": 20440
    },
    {
      "entropy": 5.36867995262146,
      "epoch": 2.296175661256809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004479686583528494,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2121911406517029,
      "num_tokens": 45631895.0,
      "step": 20445
    },
    {
      "entropy": 5.476343727111816,
      "epoch": 2.296737238164767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20749529153108598,
      "num_tokens": 45643204.0,
      "step": 20450
    },
    {
      "entropy": 5.462383794784546,
      "epoch": 2.297298815072724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20414634495973588,
      "num_tokens": 45654607.0,
      "step": 20455
    },
    {
      "entropy": 5.343553733825684,
      "epoch": 2.2978603919806817,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004478916022465958,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20763460099697112,
      "num_tokens": 45666746.0,
      "step": 20460
    },
    {
      "entropy": 5.3788408756256105,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004478659058668827,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20930723398923873,
      "num_tokens": 45676663.0,
      "step": 20465
    },
    {
      "entropy": 5.3930426120758055,
      "epoch": 2.298983545796597,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044784020398305204,
      "loss": 5.0583,
      "mean_token_accuracy": 0.20310064256191254,
      "num_tokens": 45688746.0,
      "step": 20470
    },
    {
      "entropy": 5.534788799285889,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19727590233087539,
      "num_tokens": 45700199.0,
      "step": 20475
    },
    {
      "entropy": 5.461199808120727,
      "epoch": 2.300106699612512,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044778878370631183,
      "loss": 5.0975,
      "mean_token_accuracy": 0.197740438580513,
      "num_tokens": 45710771.0,
      "step": 20480
    },
    {
      "entropy": 5.477123880386353,
      "epoch": 2.3006682765204696,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.2738,
      "mean_token_accuracy": 0.20002095252275467,
      "num_tokens": 45721957.0,
      "step": 20485
    },
    {
      "entropy": 5.410528373718262,
      "epoch": 2.301229853428427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.1345,
      "mean_token_accuracy": 0.19851591885089875,
      "num_tokens": 45733904.0,
      "step": 20490
    },
    {
      "entropy": 5.4797749519348145,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20586278438568115,
      "num_tokens": 45744004.0,
      "step": 20495
    },
    {
      "entropy": 5.429679298400879,
      "epoch": 2.302353007244342,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004476858771394412,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21419067978858947,
      "num_tokens": 45755130.0,
      "step": 20500
    },
    {
      "entropy": 5.43867506980896,
      "epoch": 2.3029145841522998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.2285,
      "mean_token_accuracy": 0.2003798231482506,
      "num_tokens": 45767353.0,
      "step": 20505
    },
    {
      "entropy": 5.520587825775147,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000447634390862417,
      "loss": 5.2957,
      "mean_token_accuracy": 0.19349124580621718,
      "num_tokens": 45779266.0,
      "step": 20510
    },
    {
      "entropy": 5.560569190979004,
      "epoch": 2.3040377379682146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.2447,
      "mean_token_accuracy": 0.19961845576763154,
      "num_tokens": 45790601.0,
      "step": 20515
    },
    {
      "entropy": 5.432533597946167,
      "epoch": 2.3045993148761723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044758288259840904,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2029760792851448,
      "num_tokens": 45801449.0,
      "step": 20520
    },
    {
      "entropy": 5.330210494995117,
      "epoch": 2.30516089178413,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475571202233363,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20565571039915084,
      "num_tokens": 45811737.0,
      "step": 20525
    },
    {
      "entropy": 5.436011457443238,
      "epoch": 2.305722468692087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.1017,
      "mean_token_accuracy": 0.2086130827665329,
      "num_tokens": 45822955.0,
      "step": 20530
    },
    {
      "entropy": 5.455626344680786,
      "epoch": 2.306284045600045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.1957,
      "mean_token_accuracy": 0.20599274784326554,
      "num_tokens": 45834697.0,
      "step": 20535
    },
    {
      "entropy": 5.403077077865601,
      "epoch": 2.3068456225080025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004474798001356859,
      "loss": 5.1165,
      "mean_token_accuracy": 0.19859466701745987,
      "num_tokens": 45846869.0,
      "step": 20540
    },
    {
      "entropy": 5.401674222946167,
      "epoch": 2.30740719941596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000447454015788394,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20572948902845384,
      "num_tokens": 45858171.0,
      "step": 20545
    },
    {
      "entropy": 5.42176570892334,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004474282259500992,
      "loss": 5.0928,
      "mean_token_accuracy": 0.20348020046949386,
      "num_tokens": 45868715.0,
      "step": 20550
    },
    {
      "entropy": 5.4375592231750485,
      "epoch": 2.308530353231875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000447402430621623,
      "loss": 5.0823,
      "mean_token_accuracy": 0.21181593537330629,
      "num_tokens": 45880569.0,
      "step": 20555
    },
    {
      "entropy": 5.443147325515747,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004473766298037865,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20375711023807525,
      "num_tokens": 45890452.0,
      "step": 20560
    },
    {
      "entropy": 5.499129438400269,
      "epoch": 2.3096535070477904,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.1987,
      "mean_token_accuracy": 0.1998823270201683,
      "num_tokens": 45903125.0,
      "step": 20565
    },
    {
      "entropy": 5.407380628585815,
      "epoch": 2.3102150839557476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044732501170331886,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20521142929792405,
      "num_tokens": 45914961.0,
      "step": 20570
    },
    {
      "entropy": 5.4881196975708,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.2333,
      "mean_token_accuracy": 0.2012341096997261,
      "num_tokens": 45927198.0,
      "step": 20575
    },
    {
      "entropy": 5.484736537933349,
      "epoch": 2.311338237771663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.193,
      "mean_token_accuracy": 0.20932626724243164,
      "num_tokens": 45937937.0,
      "step": 20580
    },
    {
      "entropy": 5.538526439666748,
      "epoch": 2.31189981467962,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.1459,
      "mean_token_accuracy": 0.19765277504920958,
      "num_tokens": 45948983.0,
      "step": 20585
    },
    {
      "entropy": 5.451184225082398,
      "epoch": 2.312461391587578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1980422854423523,
      "num_tokens": 45960674.0,
      "step": 20590
    },
    {
      "entropy": 5.428760814666748,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1996896296739578,
      "num_tokens": 45971260.0,
      "step": 20595
    },
    {
      "entropy": 5.5292823791503904,
      "epoch": 2.313584545403493,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.1505,
      "mean_token_accuracy": 0.2068763941526413,
      "num_tokens": 45983942.0,
      "step": 20600
    },
    {
      "entropy": 5.475135469436646,
      "epoch": 2.314146122311451,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.189,
      "mean_token_accuracy": 0.1958257347345352,
      "num_tokens": 45995983.0,
      "step": 20605
    },
    {
      "entropy": 5.418174982070923,
      "epoch": 2.314707699219408,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044711831989142534,
      "loss": 5.092,
      "mean_token_accuracy": 0.20370739102363586,
      "num_tokens": 46007074.0,
      "step": 20610
    },
    {
      "entropy": 5.4555549144744875,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.1555,
      "mean_token_accuracy": 0.2058824837207794,
      "num_tokens": 46018376.0,
      "step": 20615
    },
    {
      "entropy": 5.444838523864746,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004470665921188519,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20470893383026123,
      "num_tokens": 46029192.0,
      "step": 20620
    },
    {
      "entropy": 5.393720579147339,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044704072001415285,
      "loss": 5.1548,
      "mean_token_accuracy": 0.2058490812778473,
      "num_tokens": 46040735.0,
      "step": 20625
    },
    {
      "entropy": 5.363713550567627,
      "epoch": 2.316954006851238,
      "grad_norm": 1.125,
      "learning_rate": 0.00044701484243161055,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2093999281525612,
      "num_tokens": 46050859.0,
      "step": 20630
    },
    {
      "entropy": 5.349782466888428,
      "epoch": 2.317515583759196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000446988959372049,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20917047411203385,
      "num_tokens": 46060704.0,
      "step": 20635
    },
    {
      "entropy": 5.362698698043824,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.125,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.1182,
      "mean_token_accuracy": 0.2047714814543724,
      "num_tokens": 46072317.0,
      "step": 20640
    },
    {
      "entropy": 5.519372749328613,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1977435290813446,
      "num_tokens": 46083823.0,
      "step": 20645
    },
    {
      "entropy": 5.547029685974121,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.1796,
      "mean_token_accuracy": 0.20172752439975739,
      "num_tokens": 46093920.0,
      "step": 20650
    },
    {
      "entropy": 5.424487447738647,
      "epoch": 2.319761891391026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004468853723800948,
      "loss": 5.1026,
      "mean_token_accuracy": 0.21028678268194198,
      "num_tokens": 46104618.0,
      "step": 20655
    },
    {
      "entropy": 5.381516170501709,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004468594619478021,
      "loss": 4.995,
      "mean_token_accuracy": 0.21107428073883056,
      "num_tokens": 46114748.0,
      "step": 20660
    },
    {
      "entropy": 5.436009788513184,
      "epoch": 2.320885045206941,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20895057618618013,
      "num_tokens": 46125125.0,
      "step": 20665
    },
    {
      "entropy": 5.323436641693116,
      "epoch": 2.3214466221148986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004468076246678261,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2040531173348427,
      "num_tokens": 46135877.0,
      "step": 20670
    },
    {
      "entropy": 5.485434103012085,
      "epoch": 2.3220081990228563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19618668854236604,
      "num_tokens": 46146769.0,
      "step": 20675
    },
    {
      "entropy": 5.43642225265503,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20431993901729584,
      "num_tokens": 46158408.0,
      "step": 20680
    },
    {
      "entropy": 5.533440589904785,
      "epoch": 2.323131352838771,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19789442867040635,
      "num_tokens": 46169961.0,
      "step": 20685
    },
    {
      "entropy": 5.5149883270263675,
      "epoch": 2.323692929746729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20298723876476288,
      "num_tokens": 46181750.0,
      "step": 20690
    },
    {
      "entropy": 5.432737255096436,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004466779357499612,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20784169733524321,
      "num_tokens": 46192745.0,
      "step": 20695
    },
    {
      "entropy": 5.385699129104614,
      "epoch": 2.324816083562644,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.1443,
      "mean_token_accuracy": 0.20713905692100526,
      "num_tokens": 46203267.0,
      "step": 20700
    },
    {
      "entropy": 5.425318336486816,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.1223,
      "mean_token_accuracy": 0.2082781419157982,
      "num_tokens": 46214966.0,
      "step": 20705
    },
    {
      "entropy": 5.37998251914978,
      "epoch": 2.325939237378559,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044660005679713027,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20619067698717117,
      "num_tokens": 46225755.0,
      "step": 20710
    },
    {
      "entropy": 5.43026328086853,
      "epoch": 2.3265008142865167,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.1657,
      "mean_token_accuracy": 0.2067641317844391,
      "num_tokens": 46237091.0,
      "step": 20715
    },
    {
      "entropy": 5.465959119796753,
      "epoch": 2.327062391194474,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.1562,
      "mean_token_accuracy": 0.20297016501426696,
      "num_tokens": 46249250.0,
      "step": 20720
    },
    {
      "entropy": 5.441180801391601,
      "epoch": 2.3276239681024316,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044652212866998914,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20590269863605498,
      "num_tokens": 46259866.0,
      "step": 20725
    },
    {
      "entropy": 5.41418309211731,
      "epoch": 2.328185545010389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044649614170384354,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21372539848089217,
      "num_tokens": 46270888.0,
      "step": 20730
    },
    {
      "entropy": 5.40954999923706,
      "epoch": 2.328747121918347,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.136,
      "mean_token_accuracy": 0.20617228895425796,
      "num_tokens": 46281109.0,
      "step": 20735
    },
    {
      "entropy": 5.406164693832397,
      "epoch": 2.329308698826304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044644415139086913,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2161534085869789,
      "num_tokens": 46292171.0,
      "step": 20740
    },
    {
      "entropy": 5.413169527053833,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004464181480456955,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2029654785990715,
      "num_tokens": 46302919.0,
      "step": 20745
    },
    {
      "entropy": 5.311221170425415,
      "epoch": 2.3304318526422194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004463921392425017,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21227649748325347,
      "num_tokens": 46313199.0,
      "step": 20750
    },
    {
      "entropy": 5.490975570678711,
      "epoch": 2.330993429550177,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.1682,
      "mean_token_accuracy": 0.20500044524669647,
      "num_tokens": 46324316.0,
      "step": 20755
    },
    {
      "entropy": 5.42360668182373,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.1333,
      "mean_token_accuracy": 0.19894029945135117,
      "num_tokens": 46335767.0,
      "step": 20760
    },
    {
      "entropy": 5.372499132156372,
      "epoch": 2.332116583366092,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004463140800930804,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2019847124814987,
      "num_tokens": 46347840.0,
      "step": 20765
    },
    {
      "entropy": 5.401471328735352,
      "epoch": 2.3326781602740496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.2156,
      "mean_token_accuracy": 0.20026473551988602,
      "num_tokens": 46359766.0,
      "step": 20770
    },
    {
      "entropy": 5.496323347091675,
      "epoch": 2.333239737182007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044626201338521834,
      "loss": 5.1211,
      "mean_token_accuracy": 0.2147686406970024,
      "num_tokens": 46371426.0,
      "step": 20775
    },
    {
      "entropy": 5.483458709716797,
      "epoch": 2.3338013140899645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.1667,
      "mean_token_accuracy": 0.20436545163393022,
      "num_tokens": 46382257.0,
      "step": 20780
    },
    {
      "entropy": 5.438762044906616,
      "epoch": 2.334362890997922,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20005451142787933,
      "num_tokens": 46393689.0,
      "step": 20785
    },
    {
      "entropy": 5.419998693466186,
      "epoch": 2.33492446790588,
      "grad_norm": 1.125,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.1764,
      "mean_token_accuracy": 0.20407554209232331,
      "num_tokens": 46404856.0,
      "step": 20790
    },
    {
      "entropy": 5.570212554931641,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.1697,
      "mean_token_accuracy": 0.197558256983757,
      "num_tokens": 46415395.0,
      "step": 20795
    },
    {
      "entropy": 5.564279842376709,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.2006,
      "mean_token_accuracy": 0.1974046438932419,
      "num_tokens": 46426773.0,
      "step": 20800
    },
    {
      "entropy": 5.355666637420654,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004461056824149979,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2046341374516487,
      "num_tokens": 46438206.0,
      "step": 20805
    },
    {
      "entropy": 5.401529598236084,
      "epoch": 2.33717077553771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000446079608180204,
      "loss": 5.1575,
      "mean_token_accuracy": 0.20468377321958542,
      "num_tokens": 46449922.0,
      "step": 20810
    },
    {
      "entropy": 5.438352012634278,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.125,
      "learning_rate": 0.0004460535284981692,
      "loss": 5.0886,
      "mean_token_accuracy": 0.2068081885576248,
      "num_tokens": 46461715.0,
      "step": 20815
    },
    {
      "entropy": 5.434967184066773,
      "epoch": 2.338293929353625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004460274433697235,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20653095543384553,
      "num_tokens": 46473136.0,
      "step": 20820
    },
    {
      "entropy": 5.37169508934021,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044600135279569784,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21270499229431153,
      "num_tokens": 46483644.0,
      "step": 20825
    },
    {
      "entropy": 5.464639043807983,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.1614,
      "mean_token_accuracy": 0.2070650801062584,
      "num_tokens": 46494896.0,
      "step": 20830
    },
    {
      "entropy": 5.470511770248413,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004459491553142291,
      "loss": 5.1559,
      "mean_token_accuracy": 0.20189064741134644,
      "num_tokens": 46505415.0,
      "step": 20835
    },
    {
      "entropy": 5.409118700027466,
      "epoch": 2.340540236985455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.1278,
      "mean_token_accuracy": 0.2101516142487526,
      "num_tokens": 46516280.0,
      "step": 20840
    },
    {
      "entropy": 5.339461994171143,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.109375,
      "learning_rate": 0.000445896936060411,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21017466485500336,
      "num_tokens": 46526456.0,
      "step": 20845
    },
    {
      "entropy": 5.474007940292358,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20065757185220717,
      "num_tokens": 46537547.0,
      "step": 20850
    },
    {
      "entropy": 5.374297952651977,
      "epoch": 2.3422249677093276,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044584469504089407,
      "loss": 5.0514,
      "mean_token_accuracy": 0.2099015712738037,
      "num_tokens": 46549424.0,
      "step": 20855
    },
    {
      "entropy": 5.443913841247559,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004458185663710777,
      "loss": 5.0837,
      "mean_token_accuracy": 0.20638689547777175,
      "num_tokens": 46560234.0,
      "step": 20860
    },
    {
      "entropy": 5.443848466873169,
      "epoch": 2.343348121525243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.1186,
      "mean_token_accuracy": 0.201383376121521,
      "num_tokens": 46570847.0,
      "step": 20865
    },
    {
      "entropy": 5.340955400466919,
      "epoch": 2.3439096984332,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044576629271548884,
      "loss": 5.1095,
      "mean_token_accuracy": 0.2021559715270996,
      "num_tokens": 46582379.0,
      "step": 20870
    },
    {
      "entropy": 5.408187294006348,
      "epoch": 2.344471275341158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.1045,
      "mean_token_accuracy": 0.19638585895299912,
      "num_tokens": 46594669.0,
      "step": 20875
    },
    {
      "entropy": 5.299769401550293,
      "epoch": 2.3450328522491155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004457139973108403,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21712994277477266,
      "num_tokens": 46605834.0,
      "step": 20880
    },
    {
      "entropy": 5.386976528167724,
      "epoch": 2.345594429157073,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.115,
      "mean_token_accuracy": 0.20739735066890716,
      "num_tokens": 46616684.0,
      "step": 20885
    },
    {
      "entropy": 5.450810432434082,
      "epoch": 2.346156006065031,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20875076949596405,
      "num_tokens": 46628045.0,
      "step": 20890
    },
    {
      "entropy": 5.401967525482178,
      "epoch": 2.346717582972988,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.1111,
      "mean_token_accuracy": 0.20266646146774292,
      "num_tokens": 46639455.0,
      "step": 20895
    },
    {
      "entropy": 5.559433126449585,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.3275,
      "mean_token_accuracy": 0.19512481838464737,
      "num_tokens": 46651025.0,
      "step": 20900
    },
    {
      "entropy": 5.488357830047607,
      "epoch": 2.3478407367889034,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.1212,
      "mean_token_accuracy": 0.20403787791728972,
      "num_tokens": 46661859.0,
      "step": 20905
    },
    {
      "entropy": 5.472838401794434,
      "epoch": 2.3484023136968606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044555698066915385,
      "loss": 5.1046,
      "mean_token_accuracy": 0.2062048837542534,
      "num_tokens": 46673441.0,
      "step": 20910
    },
    {
      "entropy": 5.4575324058532715,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004455307922169092,
      "loss": 5.185,
      "mean_token_accuracy": 0.20046877562999726,
      "num_tokens": 46684959.0,
      "step": 20915
    },
    {
      "entropy": 5.463108539581299,
      "epoch": 2.349525467512776,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.1897,
      "mean_token_accuracy": 0.21050703525543213,
      "num_tokens": 46694940.0,
      "step": 20920
    },
    {
      "entropy": 5.426689386367798,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004454783990239542,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21039489954710006,
      "num_tokens": 46706184.0,
      "step": 20925
    },
    {
      "entropy": 5.438544273376465,
      "epoch": 2.350648621328691,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20452527850866317,
      "num_tokens": 46718267.0,
      "step": 20930
    },
    {
      "entropy": 5.434673738479614,
      "epoch": 2.3512101982366485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.1412,
      "mean_token_accuracy": 0.2029553547501564,
      "num_tokens": 46729293.0,
      "step": 20935
    },
    {
      "entropy": 5.418290853500366,
      "epoch": 2.351771775144606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20700334012508392,
      "num_tokens": 46740214.0,
      "step": 20940
    },
    {
      "entropy": 5.487683725357056,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.1649,
      "mean_token_accuracy": 0.20408000648021699,
      "num_tokens": 46751269.0,
      "step": 20945
    },
    {
      "entropy": 5.374183988571167,
      "epoch": 2.352894928960521,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004453473210644503,
      "loss": 4.9544,
      "mean_token_accuracy": 0.2161452889442444,
      "num_tokens": 46761851.0,
      "step": 20950
    },
    {
      "entropy": 5.3502898693084715,
      "epoch": 2.3534565058684787,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004453210891974355,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20845457017421723,
      "num_tokens": 46773441.0,
      "step": 20955
    },
    {
      "entropy": 5.328681230545044,
      "epoch": 2.3540180827764363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044529485190733113,
      "loss": 5.0807,
      "mean_token_accuracy": 0.20720956325531006,
      "num_tokens": 46785016.0,
      "step": 20960
    },
    {
      "entropy": 5.452505350112915,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.1403,
      "mean_token_accuracy": 0.19733494967222215,
      "num_tokens": 46796059.0,
      "step": 20965
    },
    {
      "entropy": 5.37146406173706,
      "epoch": 2.355141236592351,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044524236106119563,
      "loss": 5.0512,
      "mean_token_accuracy": 0.20761397182941438,
      "num_tokens": 46805725.0,
      "step": 20970
    },
    {
      "entropy": 5.402361822128296,
      "epoch": 2.355702813500309,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044521610750683586,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20883367508649825,
      "num_tokens": 46816300.0,
      "step": 20975
    },
    {
      "entropy": 5.411523389816284,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004451898485327291,
      "loss": 5.0879,
      "mean_token_accuracy": 0.19941720962524415,
      "num_tokens": 46827498.0,
      "step": 20980
    },
    {
      "entropy": 5.410050582885742,
      "epoch": 2.356825967316224,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451635841397116,
      "loss": 5.1061,
      "mean_token_accuracy": 0.19964056015014647,
      "num_tokens": 46838416.0,
      "step": 20985
    },
    {
      "entropy": 5.456202602386474,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20411793291568756,
      "num_tokens": 46849371.0,
      "step": 20990
    },
    {
      "entropy": 5.40569429397583,
      "epoch": 2.357949121132139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20762857049703598,
      "num_tokens": 46860095.0,
      "step": 20995
    },
    {
      "entropy": 5.49809455871582,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1987141713500023,
      "num_tokens": 46871217.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.279021420797368,
      "eval_loss": 5.261969089508057,
      "eval_mean_token_accuracy": 0.206022280616446,
      "eval_num_tokens": 46871217.0,
      "eval_runtime": 59.594,
      "eval_samples_per_second": 520.405,
      "eval_steps_per_second": 65.057,
      "step": 21000
    },
    {
      "entropy": 5.395120859146118,
      "epoch": 2.359072274948054,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.1425,
      "mean_token_accuracy": 0.21045327335596084,
      "num_tokens": 46881224.0,
      "step": 21005
    },
    {
      "entropy": 5.44067153930664,
      "epoch": 2.3596338518560116,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.1257,
      "mean_token_accuracy": 0.20409146696329117,
      "num_tokens": 46892673.0,
      "step": 21010
    },
    {
      "entropy": 5.584213829040527,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.2219,
      "mean_token_accuracy": 0.2042550638318062,
      "num_tokens": 46902753.0,
      "step": 21015
    },
    {
      "entropy": 5.3496802806854244,
      "epoch": 2.360757005671927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044497958172935536,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21060204058885573,
      "num_tokens": 46914093.0,
      "step": 21020
    },
    {
      "entropy": 5.398322629928589,
      "epoch": 2.361318582579884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.1287,
      "mean_token_accuracy": 0.21044951379299165,
      "num_tokens": 46924946.0,
      "step": 21025
    },
    {
      "entropy": 5.4322679996490475,
      "epoch": 2.361880159487842,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19735558629035949,
      "num_tokens": 46936756.0,
      "step": 21030
    },
    {
      "entropy": 5.514412069320679,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.1891,
      "mean_token_accuracy": 0.20125218182802201,
      "num_tokens": 46948552.0,
      "step": 21035
    },
    {
      "entropy": 5.522393369674683,
      "epoch": 2.363003313303757,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.1512,
      "mean_token_accuracy": 0.20165485739707947,
      "num_tokens": 46960926.0,
      "step": 21040
    },
    {
      "entropy": 5.4943528175354,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.1291,
      "mean_token_accuracy": 0.19872175008058549,
      "num_tokens": 46972657.0,
      "step": 21045
    },
    {
      "entropy": 5.447904586791992,
      "epoch": 2.364126467119672,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20387446284294128,
      "num_tokens": 46984585.0,
      "step": 21050
    },
    {
      "entropy": 5.376103115081787,
      "epoch": 2.3646880440276297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20121333450078965,
      "num_tokens": 46996052.0,
      "step": 21055
    },
    {
      "entropy": 5.484797382354737,
      "epoch": 2.365249620935587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044476896854422785,
      "loss": 5.0643,
      "mean_token_accuracy": 0.21161414682865143,
      "num_tokens": 47006185.0,
      "step": 21060
    },
    {
      "entropy": 5.423520565032959,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.125,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20293399393558503,
      "num_tokens": 47016647.0,
      "step": 21065
    },
    {
      "entropy": 5.449479627609253,
      "epoch": 2.366372774751502,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.1592,
      "mean_token_accuracy": 0.2039691463112831,
      "num_tokens": 47026847.0,
      "step": 21070
    },
    {
      "entropy": 5.437782049179077,
      "epoch": 2.36693435165946,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.1863,
      "mean_token_accuracy": 0.2026365265250206,
      "num_tokens": 47037948.0,
      "step": 21075
    },
    {
      "entropy": 5.480746603012085,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.0885,
      "mean_token_accuracy": 0.2067458525300026,
      "num_tokens": 47047751.0,
      "step": 21080
    },
    {
      "entropy": 5.400516939163208,
      "epoch": 2.3680575054753747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.107,
      "mean_token_accuracy": 0.20717793852090835,
      "num_tokens": 47059589.0,
      "step": 21085
    },
    {
      "entropy": 5.392293405532837,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20066543966531752,
      "num_tokens": 47070060.0,
      "step": 21090
    },
    {
      "entropy": 5.509277200698852,
      "epoch": 2.36918065929129,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.169,
      "mean_token_accuracy": 0.20607393831014634,
      "num_tokens": 47081419.0,
      "step": 21095
    },
    {
      "entropy": 5.438107585906982,
      "epoch": 2.3697422361992473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.1349,
      "mean_token_accuracy": 0.19580455124378204,
      "num_tokens": 47093701.0,
      "step": 21100
    },
    {
      "entropy": 5.431586265563965,
      "epoch": 2.370303813107205,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.1064,
      "mean_token_accuracy": 0.20724297314882278,
      "num_tokens": 47104930.0,
      "step": 21105
    },
    {
      "entropy": 5.421691608428955,
      "epoch": 2.3708653900151626,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004445052156173796,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20100560486316682,
      "num_tokens": 47115109.0,
      "step": 21110
    },
    {
      "entropy": 5.40271782875061,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004444788106267161,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20566214323043824,
      "num_tokens": 47125386.0,
      "step": 21115
    },
    {
      "entropy": 5.472434711456299,
      "epoch": 2.3719885438310775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20435019284486772,
      "num_tokens": 47136066.0,
      "step": 21120
    },
    {
      "entropy": 5.438594627380371,
      "epoch": 2.372550120739035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004444259844574167,
      "loss": 5.1849,
      "mean_token_accuracy": 0.20177771598100663,
      "num_tokens": 47147551.0,
      "step": 21125
    },
    {
      "entropy": 5.453580093383789,
      "epoch": 2.373111697646993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20676442384719848,
      "num_tokens": 47158798.0,
      "step": 21130
    },
    {
      "entropy": 5.549594736099243,
      "epoch": 2.3736732745549505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004443731367097613,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19992907494306564,
      "num_tokens": 47169519.0,
      "step": 21135
    },
    {
      "entropy": 5.434387922286987,
      "epoch": 2.3742348514629077,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044434670474615334,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20294055938720704,
      "num_tokens": 47180147.0,
      "step": 21140
    },
    {
      "entropy": 5.464947462081909,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.1839,
      "mean_token_accuracy": 0.20280699133872987,
      "num_tokens": 47191832.0,
      "step": 21145
    },
    {
      "entropy": 5.40901780128479,
      "epoch": 2.375358005278823,
      "grad_norm": 1.140625,
      "learning_rate": 0.000444293824643585,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2094426214694977,
      "num_tokens": 47201979.0,
      "step": 21150
    },
    {
      "entropy": 5.444290685653686,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.2159,
      "mean_token_accuracy": 0.2014210745692253,
      "num_tokens": 47212271.0,
      "step": 21155
    },
    {
      "entropy": 5.412382555007935,
      "epoch": 2.376481159094738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004442409229794925,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20491573661565782,
      "num_tokens": 47223412.0,
      "step": 21160
    },
    {
      "entropy": 5.532047891616822,
      "epoch": 2.3770427360026956,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.2508,
      "mean_token_accuracy": 0.19323180317878724,
      "num_tokens": 47235919.0,
      "step": 21165
    },
    {
      "entropy": 5.46102614402771,
      "epoch": 2.377604312910653,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004441879997606138,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20039450824260713,
      "num_tokens": 47246568.0,
      "step": 21170
    },
    {
      "entropy": 5.451126480102539,
      "epoch": 2.378165889818611,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20310898274183273,
      "num_tokens": 47259474.0,
      "step": 21175
    },
    {
      "entropy": 5.425491189956665,
      "epoch": 2.378727466726568,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.1266,
      "mean_token_accuracy": 0.20608357340097427,
      "num_tokens": 47270906.0,
      "step": 21180
    },
    {
      "entropy": 5.451981782913208,
      "epoch": 2.3792890436345258,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.204,
      "mean_token_accuracy": 0.18999551832675934,
      "num_tokens": 47283341.0,
      "step": 21185
    },
    {
      "entropy": 5.513774013519287,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19948486983776093,
      "num_tokens": 47294590.0,
      "step": 21190
    },
    {
      "entropy": 5.3335737705230715,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004440555974554673,
      "loss": 4.976,
      "mean_token_accuracy": 0.2088215783238411,
      "num_tokens": 47305111.0,
      "step": 21195
    },
    {
      "entropy": 5.412742948532104,
      "epoch": 2.3809737743583983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20277796536684037,
      "num_tokens": 47316145.0,
      "step": 21200
    },
    {
      "entropy": 5.44914059638977,
      "epoch": 2.381535351266356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004440025988479349,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20805713087320327,
      "num_tokens": 47327085.0,
      "step": 21205
    },
    {
      "entropy": 5.502685451507569,
      "epoch": 2.3820969281743136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.2379,
      "mean_token_accuracy": 0.2042923927307129,
      "num_tokens": 47338308.0,
      "step": 21210
    },
    {
      "entropy": 5.450416564941406,
      "epoch": 2.382658505082271,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.2087,
      "mean_token_accuracy": 0.2038563758134842,
      "num_tokens": 47348717.0,
      "step": 21215
    },
    {
      "entropy": 5.467160415649414,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044392306058043453,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2136983022093773,
      "num_tokens": 47359804.0,
      "step": 21220
    },
    {
      "entropy": 5.46223087310791,
      "epoch": 2.383781658898186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.1269,
      "mean_token_accuracy": 0.19665953069925307,
      "num_tokens": 47370589.0,
      "step": 21225
    },
    {
      "entropy": 5.399850606918335,
      "epoch": 2.384343235806144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004438700081744819,
      "loss": 5.0725,
      "mean_token_accuracy": 0.20427145808935165,
      "num_tokens": 47381586.0,
      "step": 21230
    },
    {
      "entropy": 5.3411805629730225,
      "epoch": 2.384904812714101,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044384347390575337,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2174593672156334,
      "num_tokens": 47394064.0,
      "step": 21235
    },
    {
      "entropy": 5.403311777114868,
      "epoch": 2.3854663896220587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20397615879774095,
      "num_tokens": 47405371.0,
      "step": 21240
    },
    {
      "entropy": 5.349968719482422,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.1213,
      "mean_token_accuracy": 0.20951241254806519,
      "num_tokens": 47416446.0,
      "step": 21245
    },
    {
      "entropy": 5.377676630020142,
      "epoch": 2.3865895434379736,
      "grad_norm": 1.125,
      "learning_rate": 0.0004437638388466979,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2169431984424591,
      "num_tokens": 47426588.0,
      "step": 21250
    },
    {
      "entropy": 5.444718742370606,
      "epoch": 2.3871511203459312,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19430748671293258,
      "num_tokens": 47438015.0,
      "step": 21255
    },
    {
      "entropy": 5.539596462249756,
      "epoch": 2.387712697253889,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.2917,
      "mean_token_accuracy": 0.19131196439266204,
      "num_tokens": 47448953.0,
      "step": 21260
    },
    {
      "entropy": 5.476703214645386,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.125,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.1189,
      "mean_token_accuracy": 0.2077726423740387,
      "num_tokens": 47459669.0,
      "step": 21265
    },
    {
      "entropy": 5.53345513343811,
      "epoch": 2.3888358510698042,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.1904,
      "mean_token_accuracy": 0.19939847737550737,
      "num_tokens": 47471018.0,
      "step": 21270
    },
    {
      "entropy": 5.519071912765503,
      "epoch": 2.3893974279777614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20761818140745164,
      "num_tokens": 47481613.0,
      "step": 21275
    },
    {
      "entropy": 5.393090295791626,
      "epoch": 2.389959004885719,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004436044236667603,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21119425147771836,
      "num_tokens": 47491245.0,
      "step": 21280
    },
    {
      "entropy": 5.377611112594605,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004435778356756618,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21309246122837067,
      "num_tokens": 47502308.0,
      "step": 21285
    },
    {
      "entropy": 5.3581384181976315,
      "epoch": 2.391082158701634,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044355124231697967,
      "loss": 5.068,
      "mean_token_accuracy": 0.20996348261833192,
      "num_tokens": 47512799.0,
      "step": 21290
    },
    {
      "entropy": 5.469184398651123,
      "epoch": 2.3916437356095916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.1815,
      "mean_token_accuracy": 0.20514436364173888,
      "num_tokens": 47524740.0,
      "step": 21295
    },
    {
      "entropy": 5.525358629226685,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1979128509759903,
      "num_tokens": 47535785.0,
      "step": 21300
    },
    {
      "entropy": 5.40638837814331,
      "epoch": 2.392766889425507,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044347143004389955,
      "loss": 4.988,
      "mean_token_accuracy": 0.21589091569185256,
      "num_tokens": 47545698.0,
      "step": 21305
    },
    {
      "entropy": 5.388017272949218,
      "epoch": 2.393328466333464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.1779,
      "mean_token_accuracy": 0.20811548084020615,
      "num_tokens": 47559786.0,
      "step": 21310
    },
    {
      "entropy": 5.438652467727661,
      "epoch": 2.393890043241422,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.0993,
      "mean_token_accuracy": 0.2090171530842781,
      "num_tokens": 47570267.0,
      "step": 21315
    },
    {
      "entropy": 5.538932800292969,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004433915694930594,
      "loss": 5.2351,
      "mean_token_accuracy": 0.19925165176391602,
      "num_tokens": 47580462.0,
      "step": 21320
    },
    {
      "entropy": 5.397080421447754,
      "epoch": 2.395013197057337,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004433649385850099,
      "loss": 5.0805,
      "mean_token_accuracy": 0.2061771422624588,
      "num_tokens": 47590651.0,
      "step": 21325
    },
    {
      "entropy": 5.493697881698608,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.2718,
      "mean_token_accuracy": 0.2020375743508339,
      "num_tokens": 47602263.0,
      "step": 21330
    },
    {
      "entropy": 5.473259925842285,
      "epoch": 2.396136350873252,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004433116606873443,
      "loss": 5.0912,
      "mean_token_accuracy": 0.21028023958206177,
      "num_tokens": 47613570.0,
      "step": 21335
    },
    {
      "entropy": 5.43745436668396,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20452479869127274,
      "num_tokens": 47624915.0,
      "step": 21340
    },
    {
      "entropy": 5.4431085109710695,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004432583613532446,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20427016764879227,
      "num_tokens": 47635974.0,
      "step": 21345
    },
    {
      "entropy": 5.431233692169189,
      "epoch": 2.3978210815971246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004432317036496533,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20478383153676988,
      "num_tokens": 47647312.0,
      "step": 21350
    },
    {
      "entropy": 5.355186128616333,
      "epoch": 2.3983826585050823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044320504058949916,
      "loss": 4.9962,
      "mean_token_accuracy": 0.208903731405735,
      "num_tokens": 47659224.0,
      "step": 21355
    },
    {
      "entropy": 5.407533168792725,
      "epoch": 2.39894423541304,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20750807821750641,
      "num_tokens": 47670720.0,
      "step": 21360
    },
    {
      "entropy": 5.454542207717895,
      "epoch": 2.3995058123209976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20211936831474303,
      "num_tokens": 47680646.0,
      "step": 21365
    },
    {
      "entropy": 5.5213581085205075,
      "epoch": 2.400067389228955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.1302,
      "mean_token_accuracy": 0.20552841126918792,
      "num_tokens": 47692155.0,
      "step": 21370
    },
    {
      "entropy": 5.562925672531128,
      "epoch": 2.4006289661369125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.2551,
      "mean_token_accuracy": 0.19760272800922393,
      "num_tokens": 47703662.0,
      "step": 21375
    },
    {
      "entropy": 5.440453386306762,
      "epoch": 2.40119054304487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044307164497000813,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20608555674552917,
      "num_tokens": 47713617.0,
      "step": 21380
    },
    {
      "entropy": 5.417771244049073,
      "epoch": 2.4017521199528273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044304494978831204,
      "loss": 5.072,
      "mean_token_accuracy": 0.2095030814409256,
      "num_tokens": 47725051.0,
      "step": 21385
    },
    {
      "entropy": 5.347665309906006,
      "epoch": 2.402313696860785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004430182492559997,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21397167146205903,
      "num_tokens": 47735023.0,
      "step": 21390
    },
    {
      "entropy": 5.404210948944092,
      "epoch": 2.4028752737687427,
      "grad_norm": 1.0,
      "learning_rate": 0.000442991543373921,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20779025703668594,
      "num_tokens": 47746252.0,
      "step": 21395
    },
    {
      "entropy": 5.413584995269775,
      "epoch": 2.4034368506767003,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.1863,
      "mean_token_accuracy": 0.20031809955835342,
      "num_tokens": 47756980.0,
      "step": 21400
    },
    {
      "entropy": 5.451322746276856,
      "epoch": 2.4039984275846575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.1562,
      "mean_token_accuracy": 0.2083822473883629,
      "num_tokens": 47767633.0,
      "step": 21405
    },
    {
      "entropy": 5.367067909240722,
      "epoch": 2.404560004492615,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004429113936375918,
      "loss": 5.0722,
      "mean_token_accuracy": 0.20575738549232483,
      "num_tokens": 47779346.0,
      "step": 21410
    },
    {
      "entropy": 5.564115524291992,
      "epoch": 2.405121581400573,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.2066,
      "mean_token_accuracy": 0.20203225761651994,
      "num_tokens": 47791039.0,
      "step": 21415
    },
    {
      "entropy": 5.47951922416687,
      "epoch": 2.4056831583085305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.2231,
      "mean_token_accuracy": 0.20087585896253585,
      "num_tokens": 47802096.0,
      "step": 21420
    },
    {
      "entropy": 5.476530456542969,
      "epoch": 2.4062447352164877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19578009843826294,
      "num_tokens": 47814087.0,
      "step": 21425
    },
    {
      "entropy": 5.394873380661011,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004428044524773651,
      "loss": 5.0938,
      "mean_token_accuracy": 0.2036154329776764,
      "num_tokens": 47824656.0,
      "step": 21430
    },
    {
      "entropy": 5.405338382720947,
      "epoch": 2.407367889032403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.1031,
      "mean_token_accuracy": 0.21432308703660966,
      "num_tokens": 47836045.0,
      "step": 21435
    },
    {
      "entropy": 5.438067245483398,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20589010417461395,
      "num_tokens": 47846494.0,
      "step": 21440
    },
    {
      "entropy": 5.505574512481689,
      "epoch": 2.408491042848318,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044272419050314833,
      "loss": 5.2003,
      "mean_token_accuracy": 0.20183025300502777,
      "num_tokens": 47858689.0,
      "step": 21445
    },
    {
      "entropy": 5.3836852550506595,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044269742582980017,
      "loss": 5.078,
      "mean_token_accuracy": 0.206304931640625,
      "num_tokens": 47870534.0,
      "step": 21450
    },
    {
      "entropy": 5.419252634048462,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.171875,
      "learning_rate": 0.000442670655816901,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20267007648944854,
      "num_tokens": 47880354.0,
      "step": 21455
    },
    {
      "entropy": 5.461186313629151,
      "epoch": 2.410175773572191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20323858708143233,
      "num_tokens": 47892528.0,
      "step": 21460
    },
    {
      "entropy": 5.4954393863677975,
      "epoch": 2.410737350480148,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.1886,
      "mean_token_accuracy": 0.210203917324543,
      "num_tokens": 47904653.0,
      "step": 21465
    },
    {
      "entropy": 5.506584692001343,
      "epoch": 2.411298927388106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19999172389507294,
      "num_tokens": 47916452.0,
      "step": 21470
    },
    {
      "entropy": 5.492001676559449,
      "epoch": 2.4118605042960635,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.1056,
      "mean_token_accuracy": 0.20237177908420562,
      "num_tokens": 47927746.0,
      "step": 21475
    },
    {
      "entropy": 5.407006120681762,
      "epoch": 2.4124220812040207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.156,
      "mean_token_accuracy": 0.19586678594350815,
      "num_tokens": 47941057.0,
      "step": 21480
    },
    {
      "entropy": 5.462095880508423,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.1905,
      "mean_token_accuracy": 0.20257679969072342,
      "num_tokens": 47951912.0,
      "step": 21485
    },
    {
      "entropy": 5.463425016403198,
      "epoch": 2.413545235019936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.1317,
      "mean_token_accuracy": 0.2050954058766365,
      "num_tokens": 47963483.0,
      "step": 21490
    },
    {
      "entropy": 5.548700952529908,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.2191,
      "mean_token_accuracy": 0.2001659780740738,
      "num_tokens": 47975047.0,
      "step": 21495
    },
    {
      "entropy": 5.492776203155517,
      "epoch": 2.414668388835851,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.1874,
      "mean_token_accuracy": 0.1982165604829788,
      "num_tokens": 47986432.0,
      "step": 21500
    },
    {
      "entropy": 5.424514389038086,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044240266220020455,
      "loss": 5.0822,
      "mean_token_accuracy": 0.20726712197065353,
      "num_tokens": 47997283.0,
      "step": 21505
    },
    {
      "entropy": 5.427126026153564,
      "epoch": 2.415791542651766,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20072503983974457,
      "num_tokens": 48008353.0,
      "step": 21510
    },
    {
      "entropy": 5.4532266616821286,
      "epoch": 2.416353119559724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.1569,
      "mean_token_accuracy": 0.2036217913031578,
      "num_tokens": 48020065.0,
      "step": 21515
    },
    {
      "entropy": 5.427692794799805,
      "epoch": 2.416914696467681,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.1603,
      "mean_token_accuracy": 0.20426976829767227,
      "num_tokens": 48030062.0,
      "step": 21520
    },
    {
      "entropy": 5.3935215950012205,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004422953154610838,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1980382889509201,
      "num_tokens": 48039639.0,
      "step": 21525
    },
    {
      "entropy": 5.411022090911866,
      "epoch": 2.4180378502835964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.1569,
      "mean_token_accuracy": 0.19955521821975708,
      "num_tokens": 48050428.0,
      "step": 21530
    },
    {
      "entropy": 5.435058355331421,
      "epoch": 2.418599427191554,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19878015518188477,
      "num_tokens": 48061214.0,
      "step": 21535
    },
    {
      "entropy": 5.384264421463013,
      "epoch": 2.4191610040995113,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044221474947290205,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21001414358615875,
      "num_tokens": 48071983.0,
      "step": 21540
    },
    {
      "entropy": 5.417275905609131,
      "epoch": 2.419722581007469,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004421878834940029,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20901309102773666,
      "num_tokens": 48083411.0,
      "step": 21545
    },
    {
      "entropy": 5.456182241439819,
      "epoch": 2.4202841579154266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1976591408252716,
      "num_tokens": 48094718.0,
      "step": 21550
    },
    {
      "entropy": 5.502328777313233,
      "epoch": 2.4208457348233843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20300680547952651,
      "num_tokens": 48106663.0,
      "step": 21555
    },
    {
      "entropy": 5.4638549327850345,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.1565,
      "mean_token_accuracy": 0.20490562319755554,
      "num_tokens": 48117232.0,
      "step": 21560
    },
    {
      "entropy": 5.453782463073731,
      "epoch": 2.421968888639299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.1523,
      "mean_token_accuracy": 0.21219206154346465,
      "num_tokens": 48128817.0,
      "step": 21565
    },
    {
      "entropy": 5.445317935943604,
      "epoch": 2.422530465547257,
      "grad_norm": 1.25,
      "learning_rate": 0.0004420534737667157,
      "loss": 5.1051,
      "mean_token_accuracy": 0.21150525510311127,
      "num_tokens": 48140893.0,
      "step": 21570
    },
    {
      "entropy": 5.368730926513672,
      "epoch": 2.423092042455214,
      "grad_norm": 1.125,
      "learning_rate": 0.00044202657586069153,
      "loss": 4.9909,
      "mean_token_accuracy": 0.20952087938785552,
      "num_tokens": 48152661.0,
      "step": 21575
    },
    {
      "entropy": 5.449160766601563,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20704046487808228,
      "num_tokens": 48163342.0,
      "step": 21580
    },
    {
      "entropy": 5.455794334411621,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.1615,
      "mean_token_accuracy": 0.2024209901690483,
      "num_tokens": 48175321.0,
      "step": 21585
    },
    {
      "entropy": 5.497385025024414,
      "epoch": 2.424776773179087,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.1863,
      "mean_token_accuracy": 0.19844339191913604,
      "num_tokens": 48186847.0,
      "step": 21590
    },
    {
      "entropy": 5.499751043319702,
      "epoch": 2.4253383500870442,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.248,
      "mean_token_accuracy": 0.2030159503221512,
      "num_tokens": 48198840.0,
      "step": 21595
    },
    {
      "entropy": 5.516138029098511,
      "epoch": 2.425899926995002,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.2238,
      "mean_token_accuracy": 0.1940855786204338,
      "num_tokens": 48209948.0,
      "step": 21600
    },
    {
      "entropy": 5.449101352691651,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044186507677252345,
      "loss": 5.128,
      "mean_token_accuracy": 0.2117145448923111,
      "num_tokens": 48220125.0,
      "step": 21605
    },
    {
      "entropy": 5.439281558990478,
      "epoch": 2.427023080810917,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.1164,
      "mean_token_accuracy": 0.20603690892457963,
      "num_tokens": 48230736.0,
      "step": 21610
    },
    {
      "entropy": 5.331849193572998,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004418112012296,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21426064372062684,
      "num_tokens": 48241387.0,
      "step": 21615
    },
    {
      "entropy": 5.445859289169311,
      "epoch": 2.428146234626832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044178425549070934,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20343160331249238,
      "num_tokens": 48252279.0,
      "step": 21620
    },
    {
      "entropy": 5.479264736175537,
      "epoch": 2.4287078115347898,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.148,
      "mean_token_accuracy": 0.2076702520251274,
      "num_tokens": 48264235.0,
      "step": 21625
    },
    {
      "entropy": 5.3917234420776365,
      "epoch": 2.4292693884427474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044173034808236,
      "loss": 5.0201,
      "mean_token_accuracy": 0.21223850846290587,
      "num_tokens": 48276194.0,
      "step": 21630
    },
    {
      "entropy": 5.465225791931152,
      "epoch": 2.4298309653507046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.1761,
      "mean_token_accuracy": 0.20470261126756667,
      "num_tokens": 48287129.0,
      "step": 21635
    },
    {
      "entropy": 5.388080883026123,
      "epoch": 2.4303925422586623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.1184,
      "mean_token_accuracy": 0.20450163036584854,
      "num_tokens": 48298850.0,
      "step": 21640
    },
    {
      "entropy": 5.454375839233398,
      "epoch": 2.43095411916662,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.1582,
      "mean_token_accuracy": 0.19934479147195816,
      "num_tokens": 48309968.0,
      "step": 21645
    },
    {
      "entropy": 5.464831018447876,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.1062,
      "mean_token_accuracy": 0.2045288473367691,
      "num_tokens": 48320419.0,
      "step": 21650
    },
    {
      "entropy": 5.455358695983887,
      "epoch": 2.432077272982535,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044159548667323126,
      "loss": 5.1704,
      "mean_token_accuracy": 0.2042033240199089,
      "num_tokens": 48332611.0,
      "step": 21655
    },
    {
      "entropy": 5.408911323547363,
      "epoch": 2.4326388498904925,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.1541,
      "mean_token_accuracy": 0.2038664251565933,
      "num_tokens": 48344551.0,
      "step": 21660
    },
    {
      "entropy": 5.490611791610718,
      "epoch": 2.43320042679845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20480212569236755,
      "num_tokens": 48355142.0,
      "step": 21665
    },
    {
      "entropy": 5.399710416793823,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20340823531150817,
      "num_tokens": 48367342.0,
      "step": 21670
    },
    {
      "entropy": 5.480062627792359,
      "epoch": 2.434323580614365,
      "grad_norm": 1.125,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.1724,
      "mean_token_accuracy": 0.2054501950740814,
      "num_tokens": 48378608.0,
      "step": 21675
    },
    {
      "entropy": 5.513961505889893,
      "epoch": 2.4348851575223227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20667513310909272,
      "num_tokens": 48390455.0,
      "step": 21680
    },
    {
      "entropy": 5.472140550613403,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.1295,
      "mean_token_accuracy": 0.2044661223888397,
      "num_tokens": 48401723.0,
      "step": 21685
    },
    {
      "entropy": 5.441765069961548,
      "epoch": 2.4360083113382376,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.1551,
      "mean_token_accuracy": 0.20014991164207457,
      "num_tokens": 48414208.0,
      "step": 21690
    },
    {
      "entropy": 5.4366865158081055,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20694406032562257,
      "num_tokens": 48425978.0,
      "step": 21695
    },
    {
      "entropy": 5.399993753433227,
      "epoch": 2.437131465154153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004413524020277388,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2193100556731224,
      "num_tokens": 48435988.0,
      "step": 21700
    },
    {
      "entropy": 5.446127986907959,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004413253661275875,
      "loss": 5.1063,
      "mean_token_accuracy": 0.2069665253162384,
      "num_tokens": 48447217.0,
      "step": 21705
    },
    {
      "entropy": 5.452662944793701,
      "epoch": 2.438254618970068,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044129832493157195,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20426540076732635,
      "num_tokens": 48458597.0,
      "step": 21710
    },
    {
      "entropy": 5.374740123748779,
      "epoch": 2.4388161958780255,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044127127844055335,
      "loss": 5.11,
      "mean_token_accuracy": 0.20425433665513992,
      "num_tokens": 48470601.0,
      "step": 21715
    },
    {
      "entropy": 5.4078230381011965,
      "epoch": 2.439377772785983,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044124422665539273,
      "loss": 5.1206,
      "mean_token_accuracy": 0.20039352923631668,
      "num_tokens": 48481570.0,
      "step": 21720
    },
    {
      "entropy": 5.5286335945129395,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.2562,
      "mean_token_accuracy": 0.19942024052143098,
      "num_tokens": 48492146.0,
      "step": 21725
    },
    {
      "entropy": 5.4845287799835205,
      "epoch": 2.440500926601898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.1263,
      "mean_token_accuracy": 0.19758029133081437,
      "num_tokens": 48504644.0,
      "step": 21730
    },
    {
      "entropy": 5.39531569480896,
      "epoch": 2.4410625035098557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20379807502031327,
      "num_tokens": 48516405.0,
      "step": 21735
    },
    {
      "entropy": 5.384009790420532,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20238864421844482,
      "num_tokens": 48528759.0,
      "step": 21740
    },
    {
      "entropy": 5.550750398635865,
      "epoch": 2.442185657325771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.2601,
      "mean_token_accuracy": 0.19456628412008287,
      "num_tokens": 48540731.0,
      "step": 21745
    },
    {
      "entropy": 5.52672438621521,
      "epoch": 2.442747234233728,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.1808,
      "mean_token_accuracy": 0.20640503764152526,
      "num_tokens": 48553642.0,
      "step": 21750
    },
    {
      "entropy": 5.547429466247559,
      "epoch": 2.443308811141686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.1798,
      "mean_token_accuracy": 0.20320207625627518,
      "num_tokens": 48564596.0,
      "step": 21755
    },
    {
      "entropy": 5.415605640411377,
      "epoch": 2.4438703880496435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000441027621888392,
      "loss": 5.1796,
      "mean_token_accuracy": 0.19893087446689606,
      "num_tokens": 48574521.0,
      "step": 21760
    },
    {
      "entropy": 5.377556943893433,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044100052249473393,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20565551668405532,
      "num_tokens": 48586299.0,
      "step": 21765
    },
    {
      "entropy": 5.466115808486938,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.1909,
      "mean_token_accuracy": 0.2067903012037277,
      "num_tokens": 48598366.0,
      "step": 21770
    },
    {
      "entropy": 5.4472582817077635,
      "epoch": 2.445555118773516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.1021,
      "mean_token_accuracy": 0.2037637263536453,
      "num_tokens": 48609659.0,
      "step": 21775
    },
    {
      "entropy": 5.388309049606323,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.125,
      "learning_rate": 0.00044091919260408585,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21275570690631868,
      "num_tokens": 48621623.0,
      "step": 21780
    },
    {
      "entropy": 5.389882230758667,
      "epoch": 2.446678272589431,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004408920720735225,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20395049899816514,
      "num_tokens": 48632711.0,
      "step": 21785
    },
    {
      "entropy": 5.4382179260253904,
      "epoch": 2.4472398494973886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044086494626089127,
      "loss": 5.1523,
      "mean_token_accuracy": 0.20303974896669388,
      "num_tokens": 48644518.0,
      "step": 21790
    },
    {
      "entropy": 5.444531488418579,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.1454,
      "mean_token_accuracy": 0.20105010122060776,
      "num_tokens": 48655261.0,
      "step": 21795
    },
    {
      "entropy": 5.4188947677612305,
      "epoch": 2.448363003313304,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044081067879288,
      "loss": 5.0863,
      "mean_token_accuracy": 0.21279276907444,
      "num_tokens": 48666701.0,
      "step": 21800
    },
    {
      "entropy": 5.369130706787109,
      "epoch": 2.448924580221261,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004407835371392279,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21357204467058183,
      "num_tokens": 48677919.0,
      "step": 21805
    },
    {
      "entropy": 5.367187070846557,
      "epoch": 2.449486157129219,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004407563902069636,
      "loss": 5.1092,
      "mean_token_accuracy": 0.21640799045562745,
      "num_tokens": 48688680.0,
      "step": 21810
    },
    {
      "entropy": 5.516146326065064,
      "epoch": 2.4500477340371765,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20733316689729692,
      "num_tokens": 48701149.0,
      "step": 21815
    },
    {
      "entropy": 5.450657510757447,
      "epoch": 2.450609310945134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.1083,
      "mean_token_accuracy": 0.20583403557538987,
      "num_tokens": 48712236.0,
      "step": 21820
    },
    {
      "entropy": 5.346473026275635,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004406749177471425,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21158252656459808,
      "num_tokens": 48723923.0,
      "step": 21825
    },
    {
      "entropy": 5.359855079650879,
      "epoch": 2.451732464761049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044064774970907516,
      "loss": 5.0689,
      "mean_token_accuracy": 0.2073149174451828,
      "num_tokens": 48734925.0,
      "step": 21830
    },
    {
      "entropy": 5.393504476547241,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004406205763967193,
      "loss": 5.037,
      "mean_token_accuracy": 0.21102525889873505,
      "num_tokens": 48745758.0,
      "step": 21835
    },
    {
      "entropy": 5.352279806137085,
      "epoch": 2.4528556185769643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004405933978109398,
      "loss": 5.0662,
      "mean_token_accuracy": 0.2082151010632515,
      "num_tokens": 48756899.0,
      "step": 21840
    },
    {
      "entropy": 5.466966676712036,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.1032,
      "mean_token_accuracy": 0.20761791318655015,
      "num_tokens": 48766805.0,
      "step": 21845
    },
    {
      "entropy": 5.536985588073731,
      "epoch": 2.453978772392879,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044053902482257265,
      "loss": 5.1547,
      "mean_token_accuracy": 0.20659345686435698,
      "num_tokens": 48779360.0,
      "step": 21850
    },
    {
      "entropy": 5.419006204605102,
      "epoch": 2.454540349300837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000440511830421716,
      "loss": 5.117,
      "mean_token_accuracy": 0.20663078129291534,
      "num_tokens": 48789836.0,
      "step": 21855
    },
    {
      "entropy": 5.331317281723022,
      "epoch": 2.455101926208794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.0956,
      "mean_token_accuracy": 0.2009486138820648,
      "num_tokens": 48801327.0,
      "step": 21860
    },
    {
      "entropy": 5.355624628067017,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044045742581098624,
      "loss": 5.0571,
      "mean_token_accuracy": 0.2127911627292633,
      "num_tokens": 48812521.0,
      "step": 21865
    },
    {
      "entropy": 5.395951986312866,
      "epoch": 2.4562250800247094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004404302156028453,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20303827971220018,
      "num_tokens": 48824068.0,
      "step": 21870
    },
    {
      "entropy": 5.42915530204773,
      "epoch": 2.456786656932667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.1253,
      "mean_token_accuracy": 0.20128611475229263,
      "num_tokens": 48835183.0,
      "step": 21875
    },
    {
      "entropy": 5.444581174850464,
      "epoch": 2.4573482338406243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.1491,
      "mean_token_accuracy": 0.2077243909239769,
      "num_tokens": 48846616.0,
      "step": 21880
    },
    {
      "entropy": 5.4088372707366945,
      "epoch": 2.457909810748582,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20397455394268035,
      "num_tokens": 48858159.0,
      "step": 21885
    },
    {
      "entropy": 5.394100761413574,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004403213221053258,
      "loss": 5.0571,
      "mean_token_accuracy": 0.2102770373225212,
      "num_tokens": 48870363.0,
      "step": 21890
    },
    {
      "entropy": 5.46568455696106,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.1838,
      "mean_token_accuracy": 0.2042328342795372,
      "num_tokens": 48882036.0,
      "step": 21895
    },
    {
      "entropy": 5.477778387069702,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20722760111093522,
      "num_tokens": 48893741.0,
      "step": 21900
    },
    {
      "entropy": 5.553566884994507,
      "epoch": 2.460156118380412,
      "grad_norm": 1.125,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19680130034685134,
      "num_tokens": 48906260.0,
      "step": 21905
    },
    {
      "entropy": 5.518086624145508,
      "epoch": 2.46071769528837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19904158115386963,
      "num_tokens": 48918078.0,
      "step": 21910
    },
    {
      "entropy": 5.419825029373169,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.1351,
      "mean_token_accuracy": 0.204119910299778,
      "num_tokens": 48928999.0,
      "step": 21915
    },
    {
      "entropy": 5.3604944229125975,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004401578239595501,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20505352169275284,
      "num_tokens": 48940587.0,
      "step": 21920
    },
    {
      "entropy": 5.452903699874878,
      "epoch": 2.4624024260122424,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.2182,
      "mean_token_accuracy": 0.2068462550640106,
      "num_tokens": 48951632.0,
      "step": 21925
    },
    {
      "entropy": 5.5242125511169435,
      "epoch": 2.4629640029202,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.2927,
      "mean_token_accuracy": 0.19909486323595046,
      "num_tokens": 48963578.0,
      "step": 21930
    },
    {
      "entropy": 5.456844091415405,
      "epoch": 2.4635255798281577,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20137947499752046,
      "num_tokens": 48974690.0,
      "step": 21935
    },
    {
      "entropy": 5.511993408203125,
      "epoch": 2.464087156736115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20190346986055374,
      "num_tokens": 48985871.0,
      "step": 21940
    },
    {
      "entropy": 5.38631911277771,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004400214308828384,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21150250285863875,
      "num_tokens": 48996475.0,
      "step": 21945
    },
    {
      "entropy": 5.481264543533325,
      "epoch": 2.46521031055203,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.2175,
      "mean_token_accuracy": 0.20560684204101562,
      "num_tokens": 49008900.0,
      "step": 21950
    },
    {
      "entropy": 5.358253002166748,
      "epoch": 2.4657718874599874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004399668368656481,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20484982430934906,
      "num_tokens": 49020583.0,
      "step": 21955
    },
    {
      "entropy": 5.526344347000122,
      "epoch": 2.466333464367945,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.2004,
      "mean_token_accuracy": 0.19531797468662263,
      "num_tokens": 49031363.0,
      "step": 21960
    },
    {
      "entropy": 5.434241533279419,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.1831,
      "mean_token_accuracy": 0.20617378503084183,
      "num_tokens": 49043404.0,
      "step": 21965
    },
    {
      "entropy": 5.5109570026397705,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.09375,
      "learning_rate": 0.000439884906447476,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20614171028137207,
      "num_tokens": 49054240.0,
      "step": 21970
    },
    {
      "entropy": 5.398666477203369,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000439857585806927,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21299918740987778,
      "num_tokens": 49064730.0,
      "step": 21975
    },
    {
      "entropy": 5.466164493560791,
      "epoch": 2.4685797719997753,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20345753580331802,
      "num_tokens": 49075218.0,
      "step": 21980
    },
    {
      "entropy": 5.43324613571167,
      "epoch": 2.469141348907733,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.1025,
      "mean_token_accuracy": 0.21175386905670165,
      "num_tokens": 49086182.0,
      "step": 21985
    },
    {
      "entropy": 5.491301250457764,
      "epoch": 2.4697029258156906,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043977559239398353,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19301327019929887,
      "num_tokens": 49099054.0,
      "step": 21990
    },
    {
      "entropy": 5.390465641021729,
      "epoch": 2.470264502723648,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.1325,
      "mean_token_accuracy": 0.21004735827445983,
      "num_tokens": 49110234.0,
      "step": 21995
    },
    {
      "entropy": 5.407011270523071,
      "epoch": 2.4708260796316055,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004397209038846436,
      "loss": 5.0904,
      "mean_token_accuracy": 0.2064321056008339,
      "num_tokens": 49121884.0,
      "step": 22000
    },
    {
      "entropy": 5.409144735336303,
      "epoch": 2.471387656539563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043969355176237207,
      "loss": 5.0694,
      "mean_token_accuracy": 0.2075664684176445,
      "num_tokens": 49133491.0,
      "step": 22005
    },
    {
      "entropy": 5.356835889816284,
      "epoch": 2.471949233447521,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004396661943961941,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20151276588439943,
      "num_tokens": 49145895.0,
      "step": 22010
    },
    {
      "entropy": 5.520007753372193,
      "epoch": 2.472510810355478,
      "grad_norm": 1.125,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20306777060031891,
      "num_tokens": 49157840.0,
      "step": 22015
    },
    {
      "entropy": 5.4553635597229,
      "epoch": 2.4730723872634357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.1444,
      "mean_token_accuracy": 0.20980408936738967,
      "num_tokens": 49170391.0,
      "step": 22020
    },
    {
      "entropy": 5.443445301055908,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.0625,
      "learning_rate": 0.000439584090842932,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1985890179872513,
      "num_tokens": 49181721.0,
      "step": 22025
    },
    {
      "entropy": 5.46344895362854,
      "epoch": 2.474195541079351,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.132,
      "mean_token_accuracy": 0.21011772751808167,
      "num_tokens": 49192137.0,
      "step": 22030
    },
    {
      "entropy": 5.502632570266724,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.1763,
      "mean_token_accuracy": 0.20808822512626649,
      "num_tokens": 49203260.0,
      "step": 22035
    },
    {
      "entropy": 5.404366779327392,
      "epoch": 2.475318694895266,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004395019401258794,
      "loss": 5.0601,
      "mean_token_accuracy": 0.21174711883068084,
      "num_tokens": 49214799.0,
      "step": 22040
    },
    {
      "entropy": 5.385938596725464,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043947454607675464,
      "loss": 5.0548,
      "mean_token_accuracy": 0.21290966272354125,
      "num_tokens": 49224595.0,
      "step": 22045
    },
    {
      "entropy": 5.43797082901001,
      "epoch": 2.476441848711181,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.1253,
      "mean_token_accuracy": 0.2089976891875267,
      "num_tokens": 49235190.0,
      "step": 22050
    },
    {
      "entropy": 5.366400957107544,
      "epoch": 2.4770034256191384,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043941974226857763,
      "loss": 5.0898,
      "mean_token_accuracy": 0.2071092650294304,
      "num_tokens": 49247103.0,
      "step": 22055
    },
    {
      "entropy": 5.454441165924072,
      "epoch": 2.477565002527096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.1727,
      "mean_token_accuracy": 0.2003272920846939,
      "num_tokens": 49259874.0,
      "step": 22060
    },
    {
      "entropy": 5.3326475620269775,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004393649175196472,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2106897711753845,
      "num_tokens": 49271059.0,
      "step": 22065
    },
    {
      "entropy": 5.414147853851318,
      "epoch": 2.478688156343011,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004393374972945815,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2114626243710518,
      "num_tokens": 49283142.0,
      "step": 22070
    },
    {
      "entropy": 5.437171983718872,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20922687649726868,
      "num_tokens": 49294022.0,
      "step": 22075
    },
    {
      "entropy": 5.377180194854736,
      "epoch": 2.4798113101589263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.1193,
      "mean_token_accuracy": 0.20738238096237183,
      "num_tokens": 49305849.0,
      "step": 22080
    },
    {
      "entropy": 5.441099548339844,
      "epoch": 2.480372887066884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043925520522745953,
      "loss": 5.1151,
      "mean_token_accuracy": 0.20262175649404526,
      "num_tokens": 49316957.0,
      "step": 22085
    },
    {
      "entropy": 5.478968763351441,
      "epoch": 2.480934463974841,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004392277640773553,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2071602925658226,
      "num_tokens": 49329040.0,
      "step": 22090
    },
    {
      "entropy": 5.408207702636719,
      "epoch": 2.481496040882799,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.127,
      "mean_token_accuracy": 0.20847557187080384,
      "num_tokens": 49340037.0,
      "step": 22095
    },
    {
      "entropy": 5.352603530883789,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.0983,
      "mean_token_accuracy": 0.19999921470880508,
      "num_tokens": 49350059.0,
      "step": 22100
    },
    {
      "entropy": 5.386012363433838,
      "epoch": 2.482619194698714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043914540925608426,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21297772228717804,
      "num_tokens": 49360880.0,
      "step": 22105
    },
    {
      "entropy": 5.520684051513672,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004391179471949213,
      "loss": 5.0803,
      "mean_token_accuracy": 0.20770995169878007,
      "num_tokens": 49370725.0,
      "step": 22110
    },
    {
      "entropy": 5.462919521331787,
      "epoch": 2.483742348514629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.0987,
      "mean_token_accuracy": 0.20494619756937027,
      "num_tokens": 49380991.0,
      "step": 22115
    },
    {
      "entropy": 5.32874755859375,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043906300739673265,
      "loss": 5.0741,
      "mean_token_accuracy": 0.2099584385752678,
      "num_tokens": 49390802.0,
      "step": 22120
    },
    {
      "entropy": 5.421877431869507,
      "epoch": 2.4848655023305444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043903552966145624,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20770589411258697,
      "num_tokens": 49401507.0,
      "step": 22125
    },
    {
      "entropy": 5.454907369613648,
      "epoch": 2.4854270792385016,
      "grad_norm": 1.140625,
      "learning_rate": 0.000439008046703225,
      "loss": 5.1923,
      "mean_token_accuracy": 0.19612319469451905,
      "num_tokens": 49412022.0,
      "step": 22130
    },
    {
      "entropy": 5.4542498111724855,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438980558522914,
      "loss": 5.0729,
      "mean_token_accuracy": 0.2123567432165146,
      "num_tokens": 49423294.0,
      "step": 22135
    },
    {
      "entropy": 5.400702285766601,
      "epoch": 2.486550233054417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004389530651213983,
      "loss": 5.0429,
      "mean_token_accuracy": 0.20341117531061173,
      "num_tokens": 49433915.0,
      "step": 22140
    },
    {
      "entropy": 5.329306364059448,
      "epoch": 2.487111809962374,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.136,
      "mean_token_accuracy": 0.20248204171657563,
      "num_tokens": 49445122.0,
      "step": 22145
    },
    {
      "entropy": 5.5616011142730715,
      "epoch": 2.487673386870332,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.2711,
      "mean_token_accuracy": 0.20364955067634583,
      "num_tokens": 49456276.0,
      "step": 22150
    },
    {
      "entropy": 5.467766952514649,
      "epoch": 2.4882349637782895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20588837563991547,
      "num_tokens": 49467346.0,
      "step": 22155
    },
    {
      "entropy": 5.4073625087738035,
      "epoch": 2.488796540686247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004388430393208005,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21020291298627852,
      "num_tokens": 49477635.0,
      "step": 22160
    },
    {
      "entropy": 5.421252155303955,
      "epoch": 2.4893581175942043,
      "grad_norm": 1.0,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20207604616880417,
      "num_tokens": 49491060.0,
      "step": 22165
    },
    {
      "entropy": 5.467432165145874,
      "epoch": 2.489919694502162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21027807295322418,
      "num_tokens": 49502551.0,
      "step": 22170
    },
    {
      "entropy": 5.426984071731567,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004387604651906152,
      "loss": 5.1017,
      "mean_token_accuracy": 0.19871904104948043,
      "num_tokens": 49514505.0,
      "step": 22175
    },
    {
      "entropy": 5.356157016754151,
      "epoch": 2.4910428483180773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004387329300509913,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2095242202281952,
      "num_tokens": 49526301.0,
      "step": 22180
    },
    {
      "entropy": 5.356016731262207,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.125,
      "learning_rate": 0.00043870538969804733,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20761689096689223,
      "num_tokens": 49536546.0,
      "step": 22185
    },
    {
      "entropy": 5.459936857223511,
      "epoch": 2.492166002133992,
      "grad_norm": 1.125,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.1998,
      "mean_token_accuracy": 0.1975899115204811,
      "num_tokens": 49548046.0,
      "step": 22190
    },
    {
      "entropy": 5.403985691070557,
      "epoch": 2.49272757904195,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043865029335570725,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20942748188972474,
      "num_tokens": 49560001.0,
      "step": 22195
    },
    {
      "entropy": 5.389744615554809,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004386227373680655,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21312234401702881,
      "num_tokens": 49571111.0,
      "step": 22200
    },
    {
      "entropy": 5.359477710723877,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004385951761706122,
      "loss": 5.0529,
      "mean_token_accuracy": 0.20459167659282684,
      "num_tokens": 49582888.0,
      "step": 22205
    },
    {
      "entropy": 5.391813707351685,
      "epoch": 2.4944123097658224,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043856760976422506,
      "loss": 5.0747,
      "mean_token_accuracy": 0.2094276189804077,
      "num_tokens": 49593710.0,
      "step": 22210
    },
    {
      "entropy": 5.441859197616577,
      "epoch": 2.49497388667378,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.1397,
      "mean_token_accuracy": 0.20599220395088197,
      "num_tokens": 49604263.0,
      "step": 22215
    },
    {
      "entropy": 5.527809953689575,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20015251487493516,
      "num_tokens": 49615152.0,
      "step": 22220
    },
    {
      "entropy": 5.481729793548584,
      "epoch": 2.496097040489695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.2138,
      "mean_token_accuracy": 0.19845256209373474,
      "num_tokens": 49627778.0,
      "step": 22225
    },
    {
      "entropy": 5.459508180618286,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.1042,
      "mean_token_accuracy": 0.206694333255291,
      "num_tokens": 49637701.0,
      "step": 22230
    },
    {
      "entropy": 5.392733383178711,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004384296996290076,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21463416367769242,
      "num_tokens": 49647680.0,
      "step": 22235
    },
    {
      "entropy": 5.468190860748291,
      "epoch": 2.4977817712135675,
      "grad_norm": 1.0,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.2261,
      "mean_token_accuracy": 0.20166214406490326,
      "num_tokens": 49659419.0,
      "step": 22240
    },
    {
      "entropy": 5.395883274078369,
      "epoch": 2.498343348121525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.1162,
      "mean_token_accuracy": 0.21457273811101912,
      "num_tokens": 49671131.0,
      "step": 22245
    },
    {
      "entropy": 5.375246953964234,
      "epoch": 2.498904925029483,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043834689109687077,
      "loss": 5.086,
      "mean_token_accuracy": 0.2061410814523697,
      "num_tokens": 49681748.0,
      "step": 22250
    },
    {
      "entropy": 5.507851266860962,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.1336,
      "mean_token_accuracy": 0.21268114298582078,
      "num_tokens": 49693321.0,
      "step": 22255
    },
    {
      "entropy": 5.446102428436279,
      "epoch": 2.500028078845398,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000438291659402173,
      "loss": 5.1305,
      "mean_token_accuracy": 0.21337852478027344,
      "num_tokens": 49704419.0,
      "step": 22260
    },
    {
      "entropy": 5.431684303283691,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.1386,
      "mean_token_accuracy": 0.20521537214517593,
      "num_tokens": 49715692.0,
      "step": 22265
    },
    {
      "entropy": 5.449667882919312,
      "epoch": 2.501151232661313,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19907229840755464,
      "num_tokens": 49726324.0,
      "step": 22270
    },
    {
      "entropy": 5.353047847747803,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043820877286780224,
      "loss": 5.0578,
      "mean_token_accuracy": 0.214020773768425,
      "num_tokens": 49737069.0,
      "step": 22275
    },
    {
      "entropy": 5.393963384628296,
      "epoch": 2.502274386477228,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.075,
      "mean_token_accuracy": 0.20897442549467088,
      "num_tokens": 49748507.0,
      "step": 22280
    },
    {
      "entropy": 5.421612644195557,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004381534891936024,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20387994945049287,
      "num_tokens": 49759017.0,
      "step": 22285
    },
    {
      "entropy": 5.495273923873901,
      "epoch": 2.503397540293143,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.1444,
      "mean_token_accuracy": 0.2037773162126541,
      "num_tokens": 49771353.0,
      "step": 22290
    },
    {
      "entropy": 5.520669603347779,
      "epoch": 2.5039591172011004,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19268622100353242,
      "num_tokens": 49782337.0,
      "step": 22295
    },
    {
      "entropy": 5.432688188552857,
      "epoch": 2.504520694109058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043807052472297626,
      "loss": 5.0501,
      "mean_token_accuracy": 0.20838258415460587,
      "num_tokens": 49794212.0,
      "step": 22300
    },
    {
      "entropy": 5.386861944198609,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004380428595138024,
      "loss": 5.0921,
      "mean_token_accuracy": 0.2110753610730171,
      "num_tokens": 49804602.0,
      "step": 22305
    },
    {
      "entropy": 5.4752496719360355,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.2374,
      "mean_token_accuracy": 0.19616419076919556,
      "num_tokens": 49816277.0,
      "step": 22310
    },
    {
      "entropy": 5.444352912902832,
      "epoch": 2.506205424832931,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043798751352229227,
      "loss": 5.1232,
      "mean_token_accuracy": 0.21200831532478331,
      "num_tokens": 49827851.0,
      "step": 22315
    },
    {
      "entropy": 5.399543190002442,
      "epoch": 2.5067670017408883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.1349,
      "mean_token_accuracy": 0.20161833614110947,
      "num_tokens": 49840492.0,
      "step": 22320
    },
    {
      "entropy": 5.30247278213501,
      "epoch": 2.507328578648846,
      "grad_norm": 1.125,
      "learning_rate": 0.00043793214677243967,
      "loss": 5.0103,
      "mean_token_accuracy": 0.20935969203710555,
      "num_tokens": 49851427.0,
      "step": 22325
    },
    {
      "entropy": 5.402151012420655,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004379044556153384,
      "loss": 5.0731,
      "mean_token_accuracy": 0.20953814089298248,
      "num_tokens": 49862883.0,
      "step": 22330
    },
    {
      "entropy": 5.426407766342163,
      "epoch": 2.508451732464761,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043787675927129625,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20087360143661498,
      "num_tokens": 49874120.0,
      "step": 22335
    },
    {
      "entropy": 5.515376710891724,
      "epoch": 2.5090133093727185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043784905774119483,
      "loss": 5.143,
      "mean_token_accuracy": 0.20761257112026216,
      "num_tokens": 49885961.0,
      "step": 22340
    },
    {
      "entropy": 5.349850511550903,
      "epoch": 2.509574886280676,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043782135102591626,
      "loss": 5.0866,
      "mean_token_accuracy": 0.21052897423505784,
      "num_tokens": 49896821.0,
      "step": 22345
    },
    {
      "entropy": 5.287903833389282,
      "epoch": 2.510136463188634,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004377936391263427,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2109289586544037,
      "num_tokens": 49908148.0,
      "step": 22350
    },
    {
      "entropy": 5.45786681175232,
      "epoch": 2.5106980400965915,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.1299,
      "mean_token_accuracy": 0.2077123999595642,
      "num_tokens": 49919487.0,
      "step": 22355
    },
    {
      "entropy": 5.397446727752685,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004377381997778403,
      "loss": 5.0792,
      "mean_token_accuracy": 0.209236016869545,
      "num_tokens": 49930722.0,
      "step": 22360
    },
    {
      "entropy": 5.466025733947754,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.198,
      "mean_token_accuracy": 0.2013266533613205,
      "num_tokens": 49941917.0,
      "step": 22365
    },
    {
      "entropy": 5.447283506393433,
      "epoch": 2.512382770820464,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.1382,
      "mean_token_accuracy": 0.19877417981624604,
      "num_tokens": 49952681.0,
      "step": 22370
    },
    {
      "entropy": 5.44276795387268,
      "epoch": 2.5129443477284212,
      "grad_norm": 1.125,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.1378,
      "mean_token_accuracy": 0.21178530752658845,
      "num_tokens": 49965542.0,
      "step": 22375
    },
    {
      "entropy": 5.445027923583984,
      "epoch": 2.513505924636379,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.1814,
      "mean_token_accuracy": 0.20130502432584763,
      "num_tokens": 49977506.0,
      "step": 22380
    },
    {
      "entropy": 5.437851762771606,
      "epoch": 2.5140675015443366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004375995107432077,
      "loss": 5.1353,
      "mean_token_accuracy": 0.20334738343954087,
      "num_tokens": 49988674.0,
      "step": 22385
    },
    {
      "entropy": 5.459364795684815,
      "epoch": 2.5146290784522938,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20566024631261826,
      "num_tokens": 50001618.0,
      "step": 22390
    },
    {
      "entropy": 5.405025815963745,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004375439988825508,
      "loss": 5.087,
      "mean_token_accuracy": 0.2120325371623039,
      "num_tokens": 50012556.0,
      "step": 22395
    },
    {
      "entropy": 5.411772537231445,
      "epoch": 2.515752232268209,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004375162351885845,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21520023792982101,
      "num_tokens": 50023399.0,
      "step": 22400
    },
    {
      "entropy": 5.37677583694458,
      "epoch": 2.5163138091761668,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043748846632003823,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2090368539094925,
      "num_tokens": 50035209.0,
      "step": 22405
    },
    {
      "entropy": 5.402634477615356,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20410651415586473,
      "num_tokens": 50046383.0,
      "step": 22410
    },
    {
      "entropy": 5.384183931350708,
      "epoch": 2.5174369629920816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.1379,
      "mean_token_accuracy": 0.2091060161590576,
      "num_tokens": 50058785.0,
      "step": 22415
    },
    {
      "entropy": 5.50031590461731,
      "epoch": 2.5179985399000393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043740512867576224,
      "loss": 5.1251,
      "mean_token_accuracy": 0.20307649224996566,
      "num_tokens": 50069629.0,
      "step": 22420
    },
    {
      "entropy": 5.473536968231201,
      "epoch": 2.518560116807997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004373773391177395,
      "loss": 5.1095,
      "mean_token_accuracy": 0.20870732218027116,
      "num_tokens": 50080923.0,
      "step": 22425
    },
    {
      "entropy": 5.378355693817139,
      "epoch": 2.519121693715954,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004373495443895593,
      "loss": 5.0521,
      "mean_token_accuracy": 0.21295883059501647,
      "num_tokens": 50092133.0,
      "step": 22430
    },
    {
      "entropy": 5.342506551742554,
      "epoch": 2.519683270623912,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004373217444921066,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21462577432394028,
      "num_tokens": 50104001.0,
      "step": 22435
    },
    {
      "entropy": 5.447042036056518,
      "epoch": 2.5202448475318695,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.1391,
      "mean_token_accuracy": 0.21015462577342986,
      "num_tokens": 50115908.0,
      "step": 22440
    },
    {
      "entropy": 5.468999767303467,
      "epoch": 2.520806424439827,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043726612919292476,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20074061900377274,
      "num_tokens": 50127380.0,
      "step": 22445
    },
    {
      "entropy": 5.481263303756714,
      "epoch": 2.521368001347785,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004372383137929663,
      "loss": 5.084,
      "mean_token_accuracy": 0.20823401659727098,
      "num_tokens": 50138458.0,
      "step": 22450
    },
    {
      "entropy": 5.502782249450684,
      "epoch": 2.521929578255742,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.2114,
      "mean_token_accuracy": 0.19980818629264832,
      "num_tokens": 50148814.0,
      "step": 22455
    },
    {
      "entropy": 5.538178014755249,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19601669609546662,
      "num_tokens": 50159233.0,
      "step": 22460
    },
    {
      "entropy": 5.408986902236938,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20489413887262345,
      "num_tokens": 50170752.0,
      "step": 22465
    },
    {
      "entropy": 5.3562246799469,
      "epoch": 2.5236143089796146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004371270005446833,
      "loss": 5.098,
      "mean_token_accuracy": 0.2090272054076195,
      "num_tokens": 50183759.0,
      "step": 22470
    },
    {
      "entropy": 5.41733717918396,
      "epoch": 2.5241758858875722,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004370991593249305,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20793443024158478,
      "num_tokens": 50194091.0,
      "step": 22475
    },
    {
      "entropy": 5.492463397979736,
      "epoch": 2.52473746279553,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20651806741952897,
      "num_tokens": 50204591.0,
      "step": 22480
    },
    {
      "entropy": 5.381657838821411,
      "epoch": 2.525299039703487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004370434614024113,
      "loss": 5.0578,
      "mean_token_accuracy": 0.21137502491474153,
      "num_tokens": 50215809.0,
      "step": 22485
    },
    {
      "entropy": 5.362876033782959,
      "epoch": 2.525860616611445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20291368663311005,
      "num_tokens": 50227038.0,
      "step": 22490
    },
    {
      "entropy": 5.422391176223755,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004369877428417856,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2133182629942894,
      "num_tokens": 50237001.0,
      "step": 22495
    },
    {
      "entropy": 5.445967435836792,
      "epoch": 2.52698377042736,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004369598758244003,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2148035928606987,
      "num_tokens": 50248389.0,
      "step": 22500
    },
    {
      "entropy": 5.3945839405059814,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043693200365014985,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20744817852973937,
      "num_tokens": 50259676.0,
      "step": 22505
    },
    {
      "entropy": 5.475585222244263,
      "epoch": 2.528106924243275,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.2184,
      "mean_token_accuracy": 0.2037392109632492,
      "num_tokens": 50270707.0,
      "step": 22510
    },
    {
      "entropy": 5.387564659118652,
      "epoch": 2.5286685011512327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043687624383460306,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21642952114343644,
      "num_tokens": 50282187.0,
      "step": 22515
    },
    {
      "entropy": 5.4640970706939695,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.156,
      "mean_token_accuracy": 0.21247451454401017,
      "num_tokens": 50293672.0,
      "step": 22520
    },
    {
      "entropy": 5.379787445068359,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004368204634022469,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20856668353080748,
      "num_tokens": 50305235.0,
      "step": 22525
    },
    {
      "entropy": 5.465430450439453,
      "epoch": 2.530353231875105,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20288207083940507,
      "num_tokens": 50316178.0,
      "step": 22530
    },
    {
      "entropy": 5.384532451629639,
      "epoch": 2.530914808783063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004367646623601856,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2108073726296425,
      "num_tokens": 50326713.0,
      "step": 22535
    },
    {
      "entropy": 5.458239650726318,
      "epoch": 2.5314763856910205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1993677020072937,
      "num_tokens": 50338648.0,
      "step": 22540
    },
    {
      "entropy": 5.476779651641846,
      "epoch": 2.532037962598978,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004367088407155261,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21403953582048416,
      "num_tokens": 50350134.0,
      "step": 22545
    },
    {
      "entropy": 5.406435489654541,
      "epoch": 2.5325995395069354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004366809221694437,
      "loss": 5.0413,
      "mean_token_accuracy": 0.21183970719575881,
      "num_tokens": 50361101.0,
      "step": 22550
    },
    {
      "entropy": 5.306814241409302,
      "epoch": 2.533161116414893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043665299847537805,
      "loss": 5.0581,
      "mean_token_accuracy": 0.2086237773299217,
      "num_tokens": 50372376.0,
      "step": 22555
    },
    {
      "entropy": 5.383921098709107,
      "epoch": 2.5337226933228507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043662506963421815,
      "loss": 5.0968,
      "mean_token_accuracy": 0.20655709952116014,
      "num_tokens": 50383494.0,
      "step": 22560
    },
    {
      "entropy": 5.477561187744141,
      "epoch": 2.534284270230808,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.1266,
      "mean_token_accuracy": 0.20617068111896514,
      "num_tokens": 50394720.0,
      "step": 22565
    },
    {
      "entropy": 5.545948839187622,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.1971,
      "mean_token_accuracy": 0.20396039187908171,
      "num_tokens": 50407109.0,
      "step": 22570
    },
    {
      "entropy": 5.5147747039794925,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.078125,
      "learning_rate": 0.000436541252237067,
      "loss": 5.2378,
      "mean_token_accuracy": 0.2009627938270569,
      "num_tokens": 50418664.0,
      "step": 22575
    },
    {
      "entropy": 5.4135674953460695,
      "epoch": 2.5359690009546805,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004365133028164248,
      "loss": 5.0454,
      "mean_token_accuracy": 0.20754648000001907,
      "num_tokens": 50431069.0,
      "step": 22580
    },
    {
      "entropy": 5.427354049682617,
      "epoch": 2.536530577862638,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364853482531363,
      "loss": 5.0957,
      "mean_token_accuracy": 0.2093221962451935,
      "num_tokens": 50441484.0,
      "step": 22585
    },
    {
      "entropy": 5.344646453857422,
      "epoch": 2.537092154770596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043645738854809177,
      "loss": 5.0437,
      "mean_token_accuracy": 0.20868187695741652,
      "num_tokens": 50452594.0,
      "step": 22590
    },
    {
      "entropy": 5.4867760181427006,
      "epoch": 2.5376537316785535,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.1773,
      "mean_token_accuracy": 0.20168371498584747,
      "num_tokens": 50463342.0,
      "step": 22595
    },
    {
      "entropy": 5.511192703247071,
      "epoch": 2.538215308586511,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20015522688627244,
      "num_tokens": 50475164.0,
      "step": 22600
    },
    {
      "entropy": 5.483790540695191,
      "epoch": 2.5387768854944683,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.1732,
      "mean_token_accuracy": 0.19870745092630387,
      "num_tokens": 50486793.0,
      "step": 22605
    },
    {
      "entropy": 5.362153577804565,
      "epoch": 2.539338462402426,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20863587856292726,
      "num_tokens": 50499202.0,
      "step": 22610
    },
    {
      "entropy": 5.561712312698364,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19596510380506516,
      "num_tokens": 50511801.0,
      "step": 22615
    },
    {
      "entropy": 5.40680890083313,
      "epoch": 2.540461616218341,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043628952239081113,
      "loss": 5.0804,
      "mean_token_accuracy": 0.19609894156455993,
      "num_tokens": 50522918.0,
      "step": 22620
    },
    {
      "entropy": 5.367985582351684,
      "epoch": 2.5410231931262985,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043626152671840984,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21014861464500428,
      "num_tokens": 50534027.0,
      "step": 22625
    },
    {
      "entropy": 5.3989551067352295,
      "epoch": 2.541584770034256,
      "grad_norm": 1.1875,
      "learning_rate": 0.000436233525911379,
      "loss": 5.046,
      "mean_token_accuracy": 0.21116897314786912,
      "num_tokens": 50543814.0,
      "step": 22630
    },
    {
      "entropy": 5.376401472091675,
      "epoch": 2.542146346942214,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004362055199706101,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21158151030540467,
      "num_tokens": 50553818.0,
      "step": 22635
    },
    {
      "entropy": 5.457902240753174,
      "epoch": 2.5427079238501715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.2256,
      "mean_token_accuracy": 0.19589451998472213,
      "num_tokens": 50564526.0,
      "step": 22640
    },
    {
      "entropy": 5.42753119468689,
      "epoch": 2.5432695007581287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043614949269142546,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21333532333374022,
      "num_tokens": 50576410.0,
      "step": 22645
    },
    {
      "entropy": 5.412558126449585,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043612147135479356,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21246462017297746,
      "num_tokens": 50586378.0,
      "step": 22650
    },
    {
      "entropy": 5.422483682632446,
      "epoch": 2.544392654574044,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.1398,
      "mean_token_accuracy": 0.20535697489976884,
      "num_tokens": 50597157.0,
      "step": 22655
    },
    {
      "entropy": 5.377931594848633,
      "epoch": 2.5449542314820013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004360654132919118,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2127812013030052,
      "num_tokens": 50607390.0,
      "step": 22660
    },
    {
      "entropy": 5.361377429962158,
      "epoch": 2.545515808389959,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004360373765674467,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21487368047237396,
      "num_tokens": 50618397.0,
      "step": 22665
    },
    {
      "entropy": 5.492807626724243,
      "epoch": 2.5460773852979166,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043600933471548927,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19758753031492232,
      "num_tokens": 50631186.0,
      "step": 22670
    },
    {
      "entropy": 5.431328010559082,
      "epoch": 2.546638962205874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004359812877369321,
      "loss": 5.1444,
      "mean_token_accuracy": 0.2049618899822235,
      "num_tokens": 50642761.0,
      "step": 22675
    },
    {
      "entropy": 5.394762372970581,
      "epoch": 2.5472005391138315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043595323563266815,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21461237072944642,
      "num_tokens": 50653374.0,
      "step": 22680
    },
    {
      "entropy": 5.37497091293335,
      "epoch": 2.547762116021789,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00043592517840359093,
      "loss": 5.0686,
      "mean_token_accuracy": 0.209019236266613,
      "num_tokens": 50664331.0,
      "step": 22685
    },
    {
      "entropy": 5.453075885772705,
      "epoch": 2.548323692929747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004358971160505936,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20302470326423644,
      "num_tokens": 50675011.0,
      "step": 22690
    },
    {
      "entropy": 5.302299356460571,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043586904857456963,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21480513960123063,
      "num_tokens": 50685139.0,
      "step": 22695
    },
    {
      "entropy": 5.4733726978302,
      "epoch": 2.5494468467456617,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043584097597641275,
      "loss": 5.0853,
      "mean_token_accuracy": 0.2080165907740593,
      "num_tokens": 50696610.0,
      "step": 22700
    },
    {
      "entropy": 5.438340854644776,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043581289825701683,
      "loss": 5.0998,
      "mean_token_accuracy": 0.21179667562246324,
      "num_tokens": 50708662.0,
      "step": 22705
    },
    {
      "entropy": 5.307219457626343,
      "epoch": 2.550570000561577,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004357848154172759,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21297313570976256,
      "num_tokens": 50719457.0,
      "step": 22710
    },
    {
      "entropy": 5.366083097457886,
      "epoch": 2.5511315774695342,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043575672745808404,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21175528019666673,
      "num_tokens": 50730038.0,
      "step": 22715
    },
    {
      "entropy": 5.4547240257263185,
      "epoch": 2.551693154377492,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.1069,
      "mean_token_accuracy": 0.2113250568509102,
      "num_tokens": 50742327.0,
      "step": 22720
    },
    {
      "entropy": 5.472136831283569,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.1452,
      "mean_token_accuracy": 0.2050308555364609,
      "num_tokens": 50753056.0,
      "step": 22725
    },
    {
      "entropy": 5.423031044006348,
      "epoch": 2.552816308193407,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20696534663438798,
      "num_tokens": 50763961.0,
      "step": 22730
    },
    {
      "entropy": 5.421393585205078,
      "epoch": 2.553377885101365,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043564432444469686,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21738698780536653,
      "num_tokens": 50774978.0,
      "step": 22735
    },
    {
      "entropy": 5.4066469192504885,
      "epoch": 2.553939462009322,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20424332916736604,
      "num_tokens": 50786198.0,
      "step": 22740
    },
    {
      "entropy": 5.345152425765991,
      "epoch": 2.5545010389172798,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004355880922445584,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20320806354284288,
      "num_tokens": 50797685.0,
      "step": 22745
    },
    {
      "entropy": 5.3020072937011715,
      "epoch": 2.5550626158252374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043555996847426077,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21797598898410797,
      "num_tokens": 50808518.0,
      "step": 22750
    },
    {
      "entropy": 5.395802211761475,
      "epoch": 2.5556241927331946,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.1262,
      "mean_token_accuracy": 0.20272241532802582,
      "num_tokens": 50819241.0,
      "step": 22755
    },
    {
      "entropy": 5.5134941101074215,
      "epoch": 2.5561857696411523,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.179,
      "mean_token_accuracy": 0.1975128248333931,
      "num_tokens": 50830334.0,
      "step": 22760
    },
    {
      "entropy": 5.555348873138428,
      "epoch": 2.55674734654911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004354755664932006,
      "loss": 5.1377,
      "mean_token_accuracy": 0.2103369116783142,
      "num_tokens": 50840738.0,
      "step": 22765
    },
    {
      "entropy": 5.493203020095825,
      "epoch": 2.557308923457067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19741234183311462,
      "num_tokens": 50853397.0,
      "step": 22770
    },
    {
      "entropy": 5.490702390670776,
      "epoch": 2.557870500365025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.114,
      "mean_token_accuracy": 0.20745037347078324,
      "num_tokens": 50865786.0,
      "step": 22775
    },
    {
      "entropy": 5.47097544670105,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.1329,
      "mean_token_accuracy": 0.2120119020342827,
      "num_tokens": 50876029.0,
      "step": 22780
    },
    {
      "entropy": 5.383950042724609,
      "epoch": 2.55899365418094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043536295898817767,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21356012374162675,
      "num_tokens": 50886941.0,
      "step": 22785
    },
    {
      "entropy": 5.4245100021362305,
      "epoch": 2.559555231088898,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004353347943446336,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20498018264770507,
      "num_tokens": 50898231.0,
      "step": 22790
    },
    {
      "entropy": 5.4215860843658445,
      "epoch": 2.560116807996855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043530662459596744,
      "loss": 5.1288,
      "mean_token_accuracy": 0.2015076458454132,
      "num_tokens": 50908282.0,
      "step": 22795
    },
    {
      "entropy": 5.429770660400391,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004352784497430762,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21625035703182222,
      "num_tokens": 50918462.0,
      "step": 22800
    },
    {
      "entropy": 5.423698949813843,
      "epoch": 2.5612399618127704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20477923452854158,
      "num_tokens": 50929160.0,
      "step": 22805
    },
    {
      "entropy": 5.45960578918457,
      "epoch": 2.5618015387207276,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.2856,
      "mean_token_accuracy": 0.19807161390781403,
      "num_tokens": 50943361.0,
      "step": 22810
    },
    {
      "entropy": 5.505819511413574,
      "epoch": 2.5623631156286852,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.2022,
      "mean_token_accuracy": 0.2015761911869049,
      "num_tokens": 50954515.0,
      "step": 22815
    },
    {
      "entropy": 5.588121557235718,
      "epoch": 2.562924692536643,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.1822,
      "mean_token_accuracy": 0.199455626308918,
      "num_tokens": 50965237.0,
      "step": 22820
    },
    {
      "entropy": 5.408406209945679,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.125,
      "learning_rate": 0.0004351374989466477,
      "loss": 5.0524,
      "mean_token_accuracy": 0.20998081862926482,
      "num_tokens": 50976406.0,
      "step": 22825
    },
    {
      "entropy": 5.265512132644654,
      "epoch": 2.5640478463525582,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043510929348725065,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21002364605665208,
      "num_tokens": 50987214.0,
      "step": 22830
    },
    {
      "entropy": 5.387618827819824,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043508108292991156,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21090883314609526,
      "num_tokens": 50997668.0,
      "step": 22835
    },
    {
      "entropy": 5.466152286529541,
      "epoch": 2.565171000168473,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004350528672755287,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21331877559423446,
      "num_tokens": 51008358.0,
      "step": 22840
    },
    {
      "entropy": 5.42615122795105,
      "epoch": 2.5657325770764308,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004350246465250005,
      "loss": 5.0837,
      "mean_token_accuracy": 0.21067432761192323,
      "num_tokens": 51019410.0,
      "step": 22845
    },
    {
      "entropy": 5.518824815750122,
      "epoch": 2.566294153984388,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043499642067922544,
      "loss": 5.2347,
      "mean_token_accuracy": 0.19578413516283036,
      "num_tokens": 51032307.0,
      "step": 22850
    },
    {
      "entropy": 5.528751039505005,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.125,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20211681276559829,
      "num_tokens": 51043176.0,
      "step": 22855
    },
    {
      "entropy": 5.462810134887695,
      "epoch": 2.5674173078003033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043493995370553,
      "loss": 5.1321,
      "mean_token_accuracy": 0.20021130740642548,
      "num_tokens": 51053977.0,
      "step": 22860
    },
    {
      "entropy": 5.42155909538269,
      "epoch": 2.5679788847082605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.1153,
      "mean_token_accuracy": 0.21119359880685806,
      "num_tokens": 51064801.0,
      "step": 22865
    },
    {
      "entropy": 5.4349987506866455,
      "epoch": 2.568540461616218,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043488346636163405,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21323357224464418,
      "num_tokens": 51075400.0,
      "step": 22870
    },
    {
      "entropy": 5.474390268325806,
      "epoch": 2.569102038524176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19865864366292954,
      "num_tokens": 51087077.0,
      "step": 22875
    },
    {
      "entropy": 5.375650310516358,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043482695865473207,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20475028157234193,
      "num_tokens": 51099798.0,
      "step": 22880
    },
    {
      "entropy": 5.493290090560913,
      "epoch": 2.570225192340091,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.2267,
      "mean_token_accuracy": 0.19780304580926894,
      "num_tokens": 51111880.0,
      "step": 22885
    },
    {
      "entropy": 5.488579797744751,
      "epoch": 2.5707867692480484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.1072,
      "mean_token_accuracy": 0.20333519130945205,
      "num_tokens": 51122358.0,
      "step": 22890
    },
    {
      "entropy": 5.44717378616333,
      "epoch": 2.571348346156006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.1304,
      "mean_token_accuracy": 0.21601888984441758,
      "num_tokens": 51132775.0,
      "step": 22895
    },
    {
      "entropy": 5.409942579269409,
      "epoch": 2.5719099230639637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.1878,
      "mean_token_accuracy": 0.20439240634441375,
      "num_tokens": 51144974.0,
      "step": 22900
    },
    {
      "entropy": 5.432405996322632,
      "epoch": 2.572471499971921,
      "grad_norm": 1.046875,
      "learning_rate": 0.000434685600346561,
      "loss": 5.0328,
      "mean_token_accuracy": 0.2171677306294441,
      "num_tokens": 51154700.0,
      "step": 22905
    },
    {
      "entropy": 5.491173219680786,
      "epoch": 2.5730330768798786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000434657313427971,
      "loss": 5.1565,
      "mean_token_accuracy": 0.20412685722112656,
      "num_tokens": 51166910.0,
      "step": 22910
    },
    {
      "entropy": 5.404850625991822,
      "epoch": 2.5735946537878363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004346290214258301,
      "loss": 5.0389,
      "mean_token_accuracy": 0.20770311653614043,
      "num_tokens": 51177620.0,
      "step": 22915
    },
    {
      "entropy": 5.460544490814209,
      "epoch": 2.574156230695794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043460072434103934,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2089150667190552,
      "num_tokens": 51187894.0,
      "step": 22920
    },
    {
      "entropy": 5.3946911811828615,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004345724221744996,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19766069650650026,
      "num_tokens": 51199625.0,
      "step": 22925
    },
    {
      "entropy": 5.368243074417114,
      "epoch": 2.575279384511709,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043454411492711197,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20217340141534806,
      "num_tokens": 51211932.0,
      "step": 22930
    },
    {
      "entropy": 5.4355552196502686,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20393133163452148,
      "num_tokens": 51223196.0,
      "step": 22935
    },
    {
      "entropy": 5.468527936935425,
      "epoch": 2.576402538327624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.1051,
      "mean_token_accuracy": 0.2014297604560852,
      "num_tokens": 51234246.0,
      "step": 22940
    },
    {
      "entropy": 5.468591976165771,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043445916270887626,
      "loss": 5.0785,
      "mean_token_accuracy": 0.20755950957536698,
      "num_tokens": 51245056.0,
      "step": 22945
    },
    {
      "entropy": 5.470414638519287,
      "epoch": 2.577525692143539,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20821897983551024,
      "num_tokens": 51256513.0,
      "step": 22950
    },
    {
      "entropy": 5.339639902114868,
      "epoch": 2.5780872690514967,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043440250250900854,
      "loss": 5.0904,
      "mean_token_accuracy": 0.2077845439314842,
      "num_tokens": 51267609.0,
      "step": 22955
    },
    {
      "entropy": 5.480114459991455,
      "epoch": 2.578648845959454,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.2171,
      "mean_token_accuracy": 0.20123790949583054,
      "num_tokens": 51277940.0,
      "step": 22960
    },
    {
      "entropy": 5.627141761779785,
      "epoch": 2.5792104228674115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.2213,
      "mean_token_accuracy": 0.2063843846321106,
      "num_tokens": 51288853.0,
      "step": 22965
    },
    {
      "entropy": 5.459927177429199,
      "epoch": 2.579771999775369,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1996838629245758,
      "num_tokens": 51299197.0,
      "step": 22970
    },
    {
      "entropy": 5.449153280258178,
      "epoch": 2.580333576683327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19712938517332076,
      "num_tokens": 51310821.0,
      "step": 22975
    },
    {
      "entropy": 5.536267232894898,
      "epoch": 2.5808951535912845,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043426076320497557,
      "loss": 5.131,
      "mean_token_accuracy": 0.1990828335285187,
      "num_tokens": 51321805.0,
      "step": 22980
    },
    {
      "entropy": 5.370523262023926,
      "epoch": 2.5814567304992417,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043423240012778437,
      "loss": 5.0351,
      "mean_token_accuracy": 0.2095471516251564,
      "num_tokens": 51332647.0,
      "step": 22985
    },
    {
      "entropy": 5.344236755371094,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004342040319805717,
      "loss": 5.0176,
      "mean_token_accuracy": 0.2128847286105156,
      "num_tokens": 51344715.0,
      "step": 22990
    },
    {
      "entropy": 5.475342226028443,
      "epoch": 2.582579884315157,
      "grad_norm": 1.078125,
      "learning_rate": 0.000434175658764241,
      "loss": 5.2053,
      "mean_token_accuracy": 0.20574932545423508,
      "num_tokens": 51355481.0,
      "step": 22995
    },
    {
      "entropy": 5.45399055480957,
      "epoch": 2.5831414612231143,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043414728047969553,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21697865277528763,
      "num_tokens": 51367098.0,
      "step": 23000
    },
    {
      "entropy": 5.452412891387939,
      "epoch": 2.583703038131072,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.2267,
      "mean_token_accuracy": 0.19664254784584045,
      "num_tokens": 51379948.0,
      "step": 23005
    },
    {
      "entropy": 5.395167684555053,
      "epoch": 2.5842646150390296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043409050870957505,
      "loss": 5.0385,
      "mean_token_accuracy": 0.208560773730278,
      "num_tokens": 51391675.0,
      "step": 23010
    },
    {
      "entropy": 5.382825183868408,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004340621152258075,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21071346700191498,
      "num_tokens": 51401809.0,
      "step": 23015
    },
    {
      "entropy": 5.399500608444214,
      "epoch": 2.585387768854945,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20611492693424224,
      "num_tokens": 51413184.0,
      "step": 23020
    },
    {
      "entropy": 5.445734167098999,
      "epoch": 2.585949345762902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.1342,
      "mean_token_accuracy": 0.20165777802467347,
      "num_tokens": 51424751.0,
      "step": 23025
    },
    {
      "entropy": 5.405837869644165,
      "epoch": 2.58651092267086,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004339769043905255,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21013531535863877,
      "num_tokens": 51436550.0,
      "step": 23030
    },
    {
      "entropy": 5.457635402679443,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20473264455795287,
      "num_tokens": 51447691.0,
      "step": 23035
    },
    {
      "entropy": 5.38881402015686,
      "epoch": 2.5876340764867747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043392007185606533,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2066597744822502,
      "num_tokens": 51459715.0,
      "step": 23040
    },
    {
      "entropy": 5.391453599929809,
      "epoch": 2.5881956533947323,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004338916479982678,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21278662979602814,
      "num_tokens": 51470541.0,
      "step": 23045
    },
    {
      "entropy": 5.40009503364563,
      "epoch": 2.58875723030269,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043386321908129856,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21344232857227324,
      "num_tokens": 51480724.0,
      "step": 23050
    },
    {
      "entropy": 5.551987743377685,
      "epoch": 2.589318807210647,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19754238575696945,
      "num_tokens": 51492204.0,
      "step": 23055
    },
    {
      "entropy": 5.488556575775147,
      "epoch": 2.589880384118605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20013070702552796,
      "num_tokens": 51503354.0,
      "step": 23060
    },
    {
      "entropy": 5.462402582168579,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.1337,
      "mean_token_accuracy": 0.2124890297651291,
      "num_tokens": 51513563.0,
      "step": 23065
    },
    {
      "entropy": 5.4714590549469,
      "epoch": 2.59100353793452,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2051364928483963,
      "num_tokens": 51524996.0,
      "step": 23070
    },
    {
      "entropy": 5.349925518035889,
      "epoch": 2.591565114842478,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004337209986405639,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21556390821933746,
      "num_tokens": 51536822.0,
      "step": 23075
    },
    {
      "entropy": 5.30804328918457,
      "epoch": 2.592126691750435,
      "grad_norm": 1.140625,
      "learning_rate": 0.000433692539387579,
      "loss": 5.054,
      "mean_token_accuracy": 0.21537428945302964,
      "num_tokens": 51547545.0,
      "step": 23080
    },
    {
      "entropy": 5.421136569976807,
      "epoch": 2.5926882686583927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.1228,
      "mean_token_accuracy": 0.20881434231996537,
      "num_tokens": 51560721.0,
      "step": 23085
    },
    {
      "entropy": 5.412398052215576,
      "epoch": 2.5932498455663504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043363560572401953,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21050462275743484,
      "num_tokens": 51573226.0,
      "step": 23090
    },
    {
      "entropy": 5.403572130203247,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20711091309785842,
      "num_tokens": 51584338.0,
      "step": 23095
    },
    {
      "entropy": 5.419375562667847,
      "epoch": 2.5943729993822653,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20598181933164597,
      "num_tokens": 51593979.0,
      "step": 23100
    },
    {
      "entropy": 5.403251075744629,
      "epoch": 2.594934576290223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004335501673483033,
      "loss": 5.0543,
      "mean_token_accuracy": 0.21524862349033355,
      "num_tokens": 51604713.0,
      "step": 23105
    },
    {
      "entropy": 5.42981653213501,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043352167779192423,
      "loss": 5.1559,
      "mean_token_accuracy": 0.20573045760393144,
      "num_tokens": 51616034.0,
      "step": 23110
    },
    {
      "entropy": 5.285714197158813,
      "epoch": 2.5960577301061383,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043349318318815363,
      "loss": 4.9976,
      "mean_token_accuracy": 0.20969287157058716,
      "num_tokens": 51627551.0,
      "step": 23115
    },
    {
      "entropy": 5.41580114364624,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043346468353789853,
      "loss": 5.1133,
      "mean_token_accuracy": 0.21085019558668136,
      "num_tokens": 51639625.0,
      "step": 23120
    },
    {
      "entropy": 5.42709493637085,
      "epoch": 2.597180883922053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004334361788420664,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2161325767636299,
      "num_tokens": 51649918.0,
      "step": 23125
    },
    {
      "entropy": 5.365367460250854,
      "epoch": 2.597742460830011,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043340766910156497,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2209593877196312,
      "num_tokens": 51659527.0,
      "step": 23130
    },
    {
      "entropy": 5.392098903656006,
      "epoch": 2.598304037737968,
      "grad_norm": 1.125,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.1953,
      "mean_token_accuracy": 0.20087363868951796,
      "num_tokens": 51670704.0,
      "step": 23135
    },
    {
      "entropy": 5.398945951461792,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043335063449018515,
      "loss": 5.0736,
      "mean_token_accuracy": 0.21201589107513427,
      "num_tokens": 51681606.0,
      "step": 23140
    },
    {
      "entropy": 5.51498851776123,
      "epoch": 2.5994271915538834,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20540947169065477,
      "num_tokens": 51692592.0,
      "step": 23145
    },
    {
      "entropy": 5.419532299041748,
      "epoch": 2.599988768461841,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2059289962053299,
      "num_tokens": 51704990.0,
      "step": 23150
    },
    {
      "entropy": 5.4582329273223875,
      "epoch": 2.6005503453697982,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043326504476079425,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2071886643767357,
      "num_tokens": 51715934.0,
      "step": 23155
    },
    {
      "entropy": 5.457715082168579,
      "epoch": 2.601111922277756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19857268333435057,
      "num_tokens": 51727314.0,
      "step": 23160
    },
    {
      "entropy": 5.4128447532653805,
      "epoch": 2.6016734991857136,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004332079597435842,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2165367364883423,
      "num_tokens": 51738322.0,
      "step": 23165
    },
    {
      "entropy": 5.318795824050904,
      "epoch": 2.602235076093671,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004331794096784206,
      "loss": 5.0239,
      "mean_token_accuracy": 0.211628657579422,
      "num_tokens": 51749221.0,
      "step": 23170
    },
    {
      "entropy": 5.34723596572876,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004331508545767633,
      "loss": 5.0816,
      "mean_token_accuracy": 0.20517460405826568,
      "num_tokens": 51760568.0,
      "step": 23175
    },
    {
      "entropy": 5.494160413742065,
      "epoch": 2.603358229909586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20936444252729416,
      "num_tokens": 51772652.0,
      "step": 23180
    },
    {
      "entropy": 5.477021408081055,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20930712819099426,
      "num_tokens": 51783739.0,
      "step": 23185
    },
    {
      "entropy": 5.408077001571655,
      "epoch": 2.604481383725501,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.1237,
      "mean_token_accuracy": 0.2082753598690033,
      "num_tokens": 51795521.0,
      "step": 23190
    },
    {
      "entropy": 5.354010772705078,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004330365838233827,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21345844864845276,
      "num_tokens": 51805041.0,
      "step": 23195
    },
    {
      "entropy": 5.443048810958862,
      "epoch": 2.6056045375414163,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.1683,
      "mean_token_accuracy": 0.20651041269302367,
      "num_tokens": 51815157.0,
      "step": 23200
    },
    {
      "entropy": 5.484793376922608,
      "epoch": 2.606166114449374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.1828,
      "mean_token_accuracy": 0.20419644117355346,
      "num_tokens": 51826760.0,
      "step": 23205
    },
    {
      "entropy": 5.564723348617553,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.2909,
      "mean_token_accuracy": 0.1972733110189438,
      "num_tokens": 51838018.0,
      "step": 23210
    },
    {
      "entropy": 5.539481925964355,
      "epoch": 2.607289268265289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.1669,
      "mean_token_accuracy": 0.2018720328807831,
      "num_tokens": 51848217.0,
      "step": 23215
    },
    {
      "entropy": 5.427230405807495,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043289363216969973,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21510244458913802,
      "num_tokens": 51857537.0,
      "step": 23220
    },
    {
      "entropy": 5.424886655807495,
      "epoch": 2.608412422081204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043286502675313757,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21123616844415666,
      "num_tokens": 51867961.0,
      "step": 23225
    },
    {
      "entropy": 5.444334840774536,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.1914,
      "mean_token_accuracy": 0.19916372895240783,
      "num_tokens": 51879868.0,
      "step": 23230
    },
    {
      "entropy": 5.4494178771972654,
      "epoch": 2.609535575897119,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20073457211256027,
      "num_tokens": 51890849.0,
      "step": 23235
    },
    {
      "entropy": 5.471737480163574,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.140625,
      "learning_rate": 0.000432779180348193,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20553481429815293,
      "num_tokens": 51901193.0,
      "step": 23240
    },
    {
      "entropy": 5.44538631439209,
      "epoch": 2.6106587297130344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043275055483116254,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20792739987373351,
      "num_tokens": 51912551.0,
      "step": 23245
    },
    {
      "entropy": 5.5036476135253904,
      "epoch": 2.6112203066209916,
      "grad_norm": 1.25,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.2725,
      "mean_token_accuracy": 0.19145291596651076,
      "num_tokens": 51923970.0,
      "step": 23250
    },
    {
      "entropy": 5.331946277618409,
      "epoch": 2.6117818835289492,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004326932887294967,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21095839142799377,
      "num_tokens": 51936667.0,
      "step": 23255
    },
    {
      "entropy": 5.429032897949218,
      "epoch": 2.612343460436907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004326646481466848,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21409022361040114,
      "num_tokens": 51947479.0,
      "step": 23260
    },
    {
      "entropy": 5.450981903076172,
      "epoch": 2.6129050373448646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.1433,
      "mean_token_accuracy": 0.21211277544498444,
      "num_tokens": 51958669.0,
      "step": 23265
    },
    {
      "entropy": 5.337259101867676,
      "epoch": 2.613466614252822,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043260735192166217,
      "loss": 4.985,
      "mean_token_accuracy": 0.2169222965836525,
      "num_tokens": 51969273.0,
      "step": 23270
    },
    {
      "entropy": 5.420536136627197,
      "epoch": 2.6140281911607794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000432578696281276,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20688224732875823,
      "num_tokens": 51981220.0,
      "step": 23275
    },
    {
      "entropy": 5.414618444442749,
      "epoch": 2.614589768068737,
      "grad_norm": 1.140625,
      "learning_rate": 0.000432550035623523,
      "loss": 5.158,
      "mean_token_accuracy": 0.20361594706773758,
      "num_tokens": 51992768.0,
      "step": 23280
    },
    {
      "entropy": 5.424482536315918,
      "epoch": 2.6151513449766943,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004325213699493157,
      "loss": 5.0634,
      "mean_token_accuracy": 0.2025006517767906,
      "num_tokens": 52002754.0,
      "step": 23285
    },
    {
      "entropy": 5.436583518981934,
      "epoch": 2.615712921884652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004324926992595669,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2032313749194145,
      "num_tokens": 52014503.0,
      "step": 23290
    },
    {
      "entropy": 5.473859977722168,
      "epoch": 2.6162744987926096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20786860585212708,
      "num_tokens": 52026151.0,
      "step": 23295
    },
    {
      "entropy": 5.411765766143799,
      "epoch": 2.6168360757005673,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004324353428370965,
      "loss": 5.1136,
      "mean_token_accuracy": 0.19949520230293274,
      "num_tokens": 52037704.0,
      "step": 23300
    },
    {
      "entropy": 5.398798084259033,
      "epoch": 2.617397652608525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.1253,
      "mean_token_accuracy": 0.2071276366710663,
      "num_tokens": 52049326.0,
      "step": 23305
    },
    {
      "entropy": 5.467378807067871,
      "epoch": 2.617959229516482,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20899081230163574,
      "num_tokens": 52060030.0,
      "step": 23310
    },
    {
      "entropy": 5.554237508773804,
      "epoch": 2.61852080642444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000432349270609657,
      "loss": 5.2597,
      "mean_token_accuracy": 0.20273399353027344,
      "num_tokens": 52071744.0,
      "step": 23315
    },
    {
      "entropy": 5.493950843811035,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.125,
      "learning_rate": 0.00043232056984583527,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20528416335582733,
      "num_tokens": 52082443.0,
      "step": 23320
    },
    {
      "entropy": 5.335069799423218,
      "epoch": 2.6196439602403547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004322918640728656,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20854045450687408,
      "num_tokens": 52094211.0,
      "step": 23325
    },
    {
      "entropy": 5.359192943572998,
      "epoch": 2.6202055371483124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043226315329166203,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21080244928598404,
      "num_tokens": 52104802.0,
      "step": 23330
    },
    {
      "entropy": 5.526046466827393,
      "epoch": 2.62076711405627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043223443750313873,
      "loss": 5.1517,
      "mean_token_accuracy": 0.20148580968379975,
      "num_tokens": 52115459.0,
      "step": 23335
    },
    {
      "entropy": 5.4361655712127686,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.1701,
      "mean_token_accuracy": 0.20079960972070693,
      "num_tokens": 52125572.0,
      "step": 23340
    },
    {
      "entropy": 5.358213472366333,
      "epoch": 2.621890267872185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.0977,
      "mean_token_accuracy": 0.20843746662139892,
      "num_tokens": 52137405.0,
      "step": 23345
    },
    {
      "entropy": 5.456938695907593,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.1807,
      "mean_token_accuracy": 0.2014411672949791,
      "num_tokens": 52148131.0,
      "step": 23350
    },
    {
      "entropy": 5.461906290054321,
      "epoch": 2.6230134216881003,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004321195242941361,
      "loss": 5.0924,
      "mean_token_accuracy": 0.2069147750735283,
      "num_tokens": 52159361.0,
      "step": 23355
    },
    {
      "entropy": 5.406369352340699,
      "epoch": 2.623574998596058,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043209078348273175,
      "loss": 5.0818,
      "mean_token_accuracy": 0.203237184882164,
      "num_tokens": 52170521.0,
      "step": 23360
    },
    {
      "entropy": 5.469378709793091,
      "epoch": 2.624136575504015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.1965,
      "mean_token_accuracy": 0.20749760121107103,
      "num_tokens": 52182069.0,
      "step": 23365
    },
    {
      "entropy": 5.382222843170166,
      "epoch": 2.624698152411973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043203328685534396,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21757508367300032,
      "num_tokens": 52191867.0,
      "step": 23370
    },
    {
      "entropy": 5.495793676376342,
      "epoch": 2.6252597293199305,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.2707,
      "mean_token_accuracy": 0.19041595458984376,
      "num_tokens": 52204511.0,
      "step": 23375
    },
    {
      "entropy": 5.436718463897705,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043197577022795327,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2090335950255394,
      "num_tokens": 52214276.0,
      "step": 23380
    },
    {
      "entropy": 5.422986650466919,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.0902,
      "mean_token_accuracy": 0.21197400093078614,
      "num_tokens": 52225388.0,
      "step": 23385
    },
    {
      "entropy": 5.417079639434815,
      "epoch": 2.626944460043803,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004319182336078852,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21496820598840713,
      "num_tokens": 52235300.0,
      "step": 23390
    },
    {
      "entropy": 5.493234252929687,
      "epoch": 2.6275060369517607,
      "grad_norm": 1.109375,
      "learning_rate": 0.000431889457802887,
      "loss": 5.2586,
      "mean_token_accuracy": 0.19737498015165328,
      "num_tokens": 52248392.0,
      "step": 23395
    },
    {
      "entropy": 5.459042882919311,
      "epoch": 2.6280676138597183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2030751034617424,
      "num_tokens": 52259839.0,
      "step": 23400
    },
    {
      "entropy": 5.455615472793579,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.1209,
      "mean_token_accuracy": 0.19735388308763505,
      "num_tokens": 52271195.0,
      "step": 23405
    },
    {
      "entropy": 5.420063257217407,
      "epoch": 2.629190767675633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21192841231822968,
      "num_tokens": 52282445.0,
      "step": 23410
    },
    {
      "entropy": 5.479402828216553,
      "epoch": 2.629752344583591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043177430463784685,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20441534519195556,
      "num_tokens": 52294144.0,
      "step": 23415
    },
    {
      "entropy": 5.364819669723511,
      "epoch": 2.630313921491548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043174550386490816,
      "loss": 5.1011,
      "mean_token_accuracy": 0.2093067318201065,
      "num_tokens": 52304510.0,
      "step": 23420
    },
    {
      "entropy": 5.413105583190918,
      "epoch": 2.6308754983995057,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004317166981011319,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20868104547262192,
      "num_tokens": 52314507.0,
      "step": 23425
    },
    {
      "entropy": 5.3814328670501705,
      "epoch": 2.6314370753074634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043168788734743516,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21835426837205887,
      "num_tokens": 52325104.0,
      "step": 23430
    },
    {
      "entropy": 5.4760137557983395,
      "epoch": 2.631998652215421,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.1543,
      "mean_token_accuracy": 0.19868220537900924,
      "num_tokens": 52336037.0,
      "step": 23435
    },
    {
      "entropy": 5.3826331615448,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004316302508739499,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2118873968720436,
      "num_tokens": 52346901.0,
      "step": 23440
    },
    {
      "entropy": 5.423587465286255,
      "epoch": 2.633121806031336,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21157545000314712,
      "num_tokens": 52359159.0,
      "step": 23445
    },
    {
      "entropy": 5.372445058822632,
      "epoch": 2.6336833829392936,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043157259445179304,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2109259620308876,
      "num_tokens": 52369274.0,
      "step": 23450
    },
    {
      "entropy": 5.312423896789551,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004315437587622574,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21556084007024764,
      "num_tokens": 52381360.0,
      "step": 23455
    },
    {
      "entropy": 5.387463665008545,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.1117,
      "mean_token_accuracy": 0.20508803129196168,
      "num_tokens": 52391550.0,
      "step": 23460
    },
    {
      "entropy": 5.402850770950318,
      "epoch": 2.635368113663166,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043148607243086255,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2121645137667656,
      "num_tokens": 52401339.0,
      "step": 23465
    },
    {
      "entropy": 5.475887441635132,
      "epoch": 2.635929690571124,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2047210842370987,
      "num_tokens": 52412062.0,
      "step": 23470
    },
    {
      "entropy": 5.457107019424439,
      "epoch": 2.636491267479081,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004314283661691587,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20480019897222518,
      "num_tokens": 52422524.0,
      "step": 23475
    },
    {
      "entropy": 5.4966268062591555,
      "epoch": 2.6370528443870387,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20226750075817107,
      "num_tokens": 52433422.0,
      "step": 23480
    },
    {
      "entropy": 5.492078876495361,
      "epoch": 2.6376144212949963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20072555541992188,
      "num_tokens": 52443879.0,
      "step": 23485
    },
    {
      "entropy": 5.457697343826294,
      "epoch": 2.638175998202954,
      "grad_norm": 1.015625,
      "learning_rate": 0.000431341769423352,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20372862368822098,
      "num_tokens": 52455610.0,
      "step": 23490
    },
    {
      "entropy": 5.5050102233886715,
      "epoch": 2.6387375751109117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.1849,
      "mean_token_accuracy": 0.20738497376441956,
      "num_tokens": 52467392.0,
      "step": 23495
    },
    {
      "entropy": 5.460610961914062,
      "epoch": 2.639299152018869,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043128401336803616,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20494269132614135,
      "num_tokens": 52478812.0,
      "step": 23500
    },
    {
      "entropy": 5.4652838706970215,
      "epoch": 2.6398607289268265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.1276,
      "mean_token_accuracy": 0.21047438383102418,
      "num_tokens": 52488837.0,
      "step": 23505
    },
    {
      "entropy": 5.4736635208129885,
      "epoch": 2.640422305834784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.1978,
      "mean_token_accuracy": 0.20212417989969253,
      "num_tokens": 52501656.0,
      "step": 23510
    },
    {
      "entropy": 5.383529138565064,
      "epoch": 2.6409838827427414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043119734196628507,
      "loss": 5.052,
      "mean_token_accuracy": 0.21037476509809494,
      "num_tokens": 52512280.0,
      "step": 23515
    },
    {
      "entropy": 5.395903396606445,
      "epoch": 2.641545459650699,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043116844155104025,
      "loss": 5.1118,
      "mean_token_accuracy": 0.21258897036314012,
      "num_tokens": 52523664.0,
      "step": 23520
    },
    {
      "entropy": 5.4381482124328615,
      "epoch": 2.6421070365586568,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.1461,
      "mean_token_accuracy": 0.20472173839807511,
      "num_tokens": 52534913.0,
      "step": 23525
    },
    {
      "entropy": 5.354604244232178,
      "epoch": 2.6426686134666144,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004311106258040795,
      "loss": 5.0418,
      "mean_token_accuracy": 0.21551703959703444,
      "num_tokens": 52545576.0,
      "step": 23530
    },
    {
      "entropy": 5.494649124145508,
      "epoch": 2.643230190374572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20229222923517226,
      "num_tokens": 52556136.0,
      "step": 23535
    },
    {
      "entropy": 5.474274063110352,
      "epoch": 2.6437917672825293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004310527901746275,
      "loss": 5.1509,
      "mean_token_accuracy": 0.2036298468708992,
      "num_tokens": 52567425.0,
      "step": 23540
    },
    {
      "entropy": 5.379021835327149,
      "epoch": 2.644353344190487,
      "grad_norm": 1.0625,
      "learning_rate": 0.000431023864906269,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20859957188367845,
      "num_tokens": 52578040.0,
      "step": 23545
    },
    {
      "entropy": 5.388794612884522,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043099493467005015,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21423546969890594,
      "num_tokens": 52588261.0,
      "step": 23550
    },
    {
      "entropy": 5.378135299682617,
      "epoch": 2.645476498006402,
      "grad_norm": 1.03125,
      "learning_rate": 0.000430965999466892,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20821608006954193,
      "num_tokens": 52599522.0,
      "step": 23555
    },
    {
      "entropy": 5.477059459686279,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.2273,
      "mean_token_accuracy": 0.20048934519290923,
      "num_tokens": 52610244.0,
      "step": 23560
    },
    {
      "entropy": 5.529667758941651,
      "epoch": 2.646599651822317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.1774,
      "mean_token_accuracy": 0.20618942528963088,
      "num_tokens": 52621745.0,
      "step": 23565
    },
    {
      "entropy": 5.423194217681885,
      "epoch": 2.6471612287302744,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.1378,
      "mean_token_accuracy": 0.20901388674974442,
      "num_tokens": 52633089.0,
      "step": 23570
    },
    {
      "entropy": 5.458262014389038,
      "epoch": 2.647722805638232,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.1478,
      "mean_token_accuracy": 0.20306558459997176,
      "num_tokens": 52644911.0,
      "step": 23575
    },
    {
      "entropy": 5.407061576843262,
      "epoch": 2.6482843825461897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043082124897926463,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2145857557654381,
      "num_tokens": 52656958.0,
      "step": 23580
    },
    {
      "entropy": 5.488681745529175,
      "epoch": 2.6488459594541474,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.1706,
      "mean_token_accuracy": 0.20456071197986603,
      "num_tokens": 52668988.0,
      "step": 23585
    },
    {
      "entropy": 5.398256063461304,
      "epoch": 2.649407536362105,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.0885,
      "mean_token_accuracy": 0.20276652574539183,
      "num_tokens": 52679948.0,
      "step": 23590
    },
    {
      "entropy": 5.410170364379883,
      "epoch": 2.6499691132700622,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004307343391424055,
      "loss": 5.0761,
      "mean_token_accuracy": 0.2093397244811058,
      "num_tokens": 52690249.0,
      "step": 23595
    },
    {
      "entropy": 5.43637866973877,
      "epoch": 2.65053069017802,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.1827,
      "mean_token_accuracy": 0.2046618014574051,
      "num_tokens": 52702766.0,
      "step": 23600
    },
    {
      "entropy": 5.494542741775513,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.1342,
      "mean_token_accuracy": 0.20970163494348526,
      "num_tokens": 52714252.0,
      "step": 23605
    },
    {
      "entropy": 5.53400354385376,
      "epoch": 2.651653843993935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.1953,
      "mean_token_accuracy": 0.20609711110591888,
      "num_tokens": 52724748.0,
      "step": 23610
    },
    {
      "entropy": 5.373533153533936,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.1431,
      "mean_token_accuracy": 0.2032534137368202,
      "num_tokens": 52735636.0,
      "step": 23615
    },
    {
      "entropy": 5.520676517486573,
      "epoch": 2.65277699780985,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.2793,
      "mean_token_accuracy": 0.19937525540590287,
      "num_tokens": 52747812.0,
      "step": 23620
    },
    {
      "entropy": 5.45426836013794,
      "epoch": 2.6533385747178078,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004305603856102808,
      "loss": 5.0761,
      "mean_token_accuracy": 0.2132926493883133,
      "num_tokens": 52759322.0,
      "step": 23625
    },
    {
      "entropy": 5.403883647918701,
      "epoch": 2.6539001516257654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004305313760136799,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20728790909051895,
      "num_tokens": 52770238.0,
      "step": 23630
    },
    {
      "entropy": 5.440744066238404,
      "epoch": 2.6544617285337226,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.082,
      "mean_token_accuracy": 0.2092401549220085,
      "num_tokens": 52781243.0,
      "step": 23635
    },
    {
      "entropy": 5.453002071380615,
      "epoch": 2.6550233054416803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004304733419648638,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20533337593078613,
      "num_tokens": 52791341.0,
      "step": 23640
    },
    {
      "entropy": 5.4701573848724365,
      "epoch": 2.655584882349638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1995821177959442,
      "num_tokens": 52803322.0,
      "step": 23645
    },
    {
      "entropy": 5.363886880874634,
      "epoch": 2.656146459257595,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004304152881147219,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2115256205201149,
      "num_tokens": 52813561.0,
      "step": 23650
    },
    {
      "entropy": 5.417882871627808,
      "epoch": 2.656708036165553,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.1568,
      "mean_token_accuracy": 0.2046465903520584,
      "num_tokens": 52824150.0,
      "step": 23655
    },
    {
      "entropy": 5.326371192932129,
      "epoch": 2.6572696130735105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000430357214470648,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21735884994268417,
      "num_tokens": 52835013.0,
      "step": 23660
    },
    {
      "entropy": 5.455588006973267,
      "epoch": 2.6578311899814677,
      "grad_norm": 1.0,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.1796,
      "mean_token_accuracy": 0.2102593794465065,
      "num_tokens": 52846682.0,
      "step": 23665
    },
    {
      "entropy": 5.300525856018067,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043029912104003827,
      "loss": 5.0453,
      "mean_token_accuracy": 0.20207959711551665,
      "num_tokens": 52857561.0,
      "step": 23670
    },
    {
      "entropy": 5.410378456115723,
      "epoch": 2.658954343797383,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043027006690709457,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20814764648675918,
      "num_tokens": 52868833.0,
      "step": 23675
    },
    {
      "entropy": 5.526295185089111,
      "epoch": 2.6595159207053407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004302410078302918,
      "loss": 5.172,
      "mean_token_accuracy": 0.20282174199819564,
      "num_tokens": 52879725.0,
      "step": 23680
    },
    {
      "entropy": 5.535358667373657,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043021194381055505,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19760958701372147,
      "num_tokens": 52890585.0,
      "step": 23685
    },
    {
      "entropy": 5.512133836746216,
      "epoch": 2.6606390745212556,
      "grad_norm": 1.0,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.2141,
      "mean_token_accuracy": 0.19680009335279464,
      "num_tokens": 52902816.0,
      "step": 23690
    },
    {
      "entropy": 5.56955394744873,
      "epoch": 2.6612006514292132,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19939966797828673,
      "num_tokens": 52914523.0,
      "step": 23695
    },
    {
      "entropy": 5.4777790069580075,
      "epoch": 2.661762228337171,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20963690280914307,
      "num_tokens": 52924953.0,
      "step": 23700
    },
    {
      "entropy": 5.349331569671631,
      "epoch": 2.662323805245128,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043009563832077977,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20861204266548156,
      "num_tokens": 52935435.0,
      "step": 23705
    },
    {
      "entropy": 5.427224016189575,
      "epoch": 2.662885382153086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.1326,
      "mean_token_accuracy": 0.20777385532855988,
      "num_tokens": 52946969.0,
      "step": 23710
    },
    {
      "entropy": 5.399440622329712,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.147,
      "mean_token_accuracy": 0.2036915346980095,
      "num_tokens": 52958634.0,
      "step": 23715
    },
    {
      "entropy": 5.421158504486084,
      "epoch": 2.664008535969001,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004300083573480031,
      "loss": 5.0228,
      "mean_token_accuracy": 0.2136502131819725,
      "num_tokens": 52968412.0,
      "step": 23720
    },
    {
      "entropy": 5.351476764678955,
      "epoch": 2.6645701128769588,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042997925381812313,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20661198496818542,
      "num_tokens": 52978653.0,
      "step": 23725
    },
    {
      "entropy": 5.3989983081817625,
      "epoch": 2.665131689784916,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042995014535364363,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21125713139772415,
      "num_tokens": 52988518.0,
      "step": 23730
    },
    {
      "entropy": 5.390887784957886,
      "epoch": 2.6656932666928737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004299210319554914,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20542672425508499,
      "num_tokens": 52999271.0,
      "step": 23735
    },
    {
      "entropy": 5.426955080032348,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.1456,
      "mean_token_accuracy": 0.2142097160220146,
      "num_tokens": 53010530.0,
      "step": 23740
    },
    {
      "entropy": 5.407283639907837,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20760780274868013,
      "num_tokens": 53021631.0,
      "step": 23745
    },
    {
      "entropy": 5.433384370803833,
      "epoch": 2.667377997416746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042983366216826907,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20958445221185684,
      "num_tokens": 53033763.0,
      "step": 23750
    },
    {
      "entropy": 5.400677347183228,
      "epoch": 2.667939574324704,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.1284,
      "mean_token_accuracy": 0.20233191996812822,
      "num_tokens": 53045839.0,
      "step": 23755
    },
    {
      "entropy": 5.430856704711914,
      "epoch": 2.668501151232661,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.1202,
      "mean_token_accuracy": 0.20373202860355377,
      "num_tokens": 53057295.0,
      "step": 23760
    },
    {
      "entropy": 5.615190744400024,
      "epoch": 2.6690627281406187,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.3261,
      "mean_token_accuracy": 0.19458230435848237,
      "num_tokens": 53068595.0,
      "step": 23765
    },
    {
      "entropy": 5.437267637252807,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042971710010347594,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20453943759202958,
      "num_tokens": 53080170.0,
      "step": 23770
    },
    {
      "entropy": 5.396388149261474,
      "epoch": 2.670185881956534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042968794726932634,
      "loss": 5.0423,
      "mean_token_accuracy": 0.2084454908967018,
      "num_tokens": 53090894.0,
      "step": 23775
    },
    {
      "entropy": 5.428504610061646,
      "epoch": 2.6707474588644917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.1806,
      "mean_token_accuracy": 0.20200824737548828,
      "num_tokens": 53103732.0,
      "step": 23780
    },
    {
      "entropy": 5.485531139373779,
      "epoch": 2.671309035772449,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20904125720262529,
      "num_tokens": 53115237.0,
      "step": 23785
    },
    {
      "entropy": 5.470984601974488,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042960045921864627,
      "loss": 5.0703,
      "mean_token_accuracy": 0.20819331109523773,
      "num_tokens": 53125638.0,
      "step": 23790
    },
    {
      "entropy": 5.366934299468994,
      "epoch": 2.6724321895883643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004295712866887709,
      "loss": 5.0428,
      "mean_token_accuracy": 0.21460684686899184,
      "num_tokens": 53136153.0,
      "step": 23795
    },
    {
      "entropy": 5.3888102054595945,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.14,
      "mean_token_accuracy": 0.20648224651813507,
      "num_tokens": 53146562.0,
      "step": 23800
    },
    {
      "entropy": 5.554017829895019,
      "epoch": 2.673555343404279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004295129268651198,
      "loss": 5.2169,
      "mean_token_accuracy": 0.20183952450752257,
      "num_tokens": 53158460.0,
      "step": 23805
    },
    {
      "entropy": 5.441367816925049,
      "epoch": 2.674116920312237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004294837395732025,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20881316512823106,
      "num_tokens": 53170576.0,
      "step": 23810
    },
    {
      "entropy": 5.3978125095367435,
      "epoch": 2.6746784972201945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042945454736246293,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21231072396039963,
      "num_tokens": 53181658.0,
      "step": 23815
    },
    {
      "entropy": 5.466558599472046,
      "epoch": 2.675240074128152,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.1686,
      "mean_token_accuracy": 0.20002918541431428,
      "num_tokens": 53193983.0,
      "step": 23820
    },
    {
      "entropy": 5.406112957000732,
      "epoch": 2.6758016510361093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042939614818823547,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2031721517443657,
      "num_tokens": 53205603.0,
      "step": 23825
    },
    {
      "entropy": 5.375855779647827,
      "epoch": 2.676363227944067,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004293669412266069,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21377104073762893,
      "num_tokens": 53216295.0,
      "step": 23830
    },
    {
      "entropy": 5.414020347595215,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004293377293498751,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2135077029466629,
      "num_tokens": 53227055.0,
      "step": 23835
    },
    {
      "entropy": 5.443935918807983,
      "epoch": 2.677486381759982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042930851255897014,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20695868581533433,
      "num_tokens": 53239050.0,
      "step": 23840
    },
    {
      "entropy": 5.417085361480713,
      "epoch": 2.6780479586679395,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004292792908548223,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21170464009046555,
      "num_tokens": 53249551.0,
      "step": 23845
    },
    {
      "entropy": 5.421101808547974,
      "epoch": 2.678609535575897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042925006423836197,
      "loss": 5.1825,
      "mean_token_accuracy": 0.1989559069275856,
      "num_tokens": 53261645.0,
      "step": 23850
    },
    {
      "entropy": 5.462469291687012,
      "epoch": 2.6791711124838544,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.1889,
      "mean_token_accuracy": 0.19979670643806458,
      "num_tokens": 53272601.0,
      "step": 23855
    },
    {
      "entropy": 5.488002920150757,
      "epoch": 2.679732689391812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20662403702735901,
      "num_tokens": 53283190.0,
      "step": 23860
    },
    {
      "entropy": 5.455269145965576,
      "epoch": 2.6802942662997697,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004291623549244129,
      "loss": 5.085,
      "mean_token_accuracy": 0.20648755580186845,
      "num_tokens": 53294374.0,
      "step": 23865
    },
    {
      "entropy": 5.47508282661438,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.2113,
      "mean_token_accuracy": 0.20011205971240997,
      "num_tokens": 53305609.0,
      "step": 23870
    },
    {
      "entropy": 5.323021984100341,
      "epoch": 2.681417420115685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042910385750394964,
      "loss": 5.05,
      "mean_token_accuracy": 0.21304287761449814,
      "num_tokens": 53316251.0,
      "step": 23875
    },
    {
      "entropy": 5.3883397579193115,
      "epoch": 2.6819789970236423,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004290746014331623,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20981789827346803,
      "num_tokens": 53326798.0,
      "step": 23880
    },
    {
      "entropy": 5.482421255111694,
      "epoch": 2.6825405739316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042904534045658,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20399740636348723,
      "num_tokens": 53337586.0,
      "step": 23885
    },
    {
      "entropy": 5.490723323822022,
      "epoch": 2.6831021508395576,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20513687580823897,
      "num_tokens": 53348517.0,
      "step": 23890
    },
    {
      "entropy": 5.439955139160157,
      "epoch": 2.683663727747515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004289868037897571,
      "loss": 5.105,
      "mean_token_accuracy": 0.2101793736219406,
      "num_tokens": 53358937.0,
      "step": 23895
    },
    {
      "entropy": 5.40477819442749,
      "epoch": 2.6842253046554725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004289575281013802,
      "loss": 5.0735,
      "mean_token_accuracy": 0.21647287607192994,
      "num_tokens": 53370007.0,
      "step": 23900
    },
    {
      "entropy": 5.387922048568726,
      "epoch": 2.68478688156343,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000428928247510936,
      "loss": 4.9683,
      "mean_token_accuracy": 0.21555528789758682,
      "num_tokens": 53380913.0,
      "step": 23905
    },
    {
      "entropy": 5.394782686233521,
      "epoch": 2.685348458471388,
      "grad_norm": 1.125,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20735053569078446,
      "num_tokens": 53392526.0,
      "step": 23910
    },
    {
      "entropy": 5.396281766891479,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20746338069438935,
      "num_tokens": 53404522.0,
      "step": 23915
    },
    {
      "entropy": 5.37363076210022,
      "epoch": 2.6864716122873027,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042884037633652276,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21111266314983368,
      "num_tokens": 53415156.0,
      "step": 23920
    },
    {
      "entropy": 5.452668190002441,
      "epoch": 2.6870331891952604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042881107614713355,
      "loss": 5.119,
      "mean_token_accuracy": 0.19550672620534898,
      "num_tokens": 53426297.0,
      "step": 23925
    },
    {
      "entropy": 5.573445796966553,
      "epoch": 2.687594766103218,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042878177106034,
      "loss": 5.2037,
      "mean_token_accuracy": 0.19766824394464494,
      "num_tokens": 53438415.0,
      "step": 23930
    },
    {
      "entropy": 5.456334590911865,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20160286724567414,
      "num_tokens": 53450103.0,
      "step": 23935
    },
    {
      "entropy": 5.399632263183594,
      "epoch": 2.688717919919133,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042872314619827234,
      "loss": 5.0296,
      "mean_token_accuracy": 0.20934877544641495,
      "num_tokens": 53461440.0,
      "step": 23940
    },
    {
      "entropy": 5.48495192527771,
      "epoch": 2.6892794968270906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000428693826424865,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19940572530031203,
      "num_tokens": 53472369.0,
      "step": 23945
    },
    {
      "entropy": 5.472682237625122,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004286645017577866,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20963944643735885,
      "num_tokens": 53483739.0,
      "step": 23950
    },
    {
      "entropy": 5.475051355361939,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.1713,
      "mean_token_accuracy": 0.20393347293138503,
      "num_tokens": 53494131.0,
      "step": 23955
    },
    {
      "entropy": 5.450966024398804,
      "epoch": 2.690964227550963,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.1671,
      "mean_token_accuracy": 0.199500147998333,
      "num_tokens": 53504923.0,
      "step": 23960
    },
    {
      "entropy": 5.470095539093018,
      "epoch": 2.6915258044589208,
      "grad_norm": 1.046875,
      "learning_rate": 0.000428576498403863,
      "loss": 5.1709,
      "mean_token_accuracy": 0.19616492539644242,
      "num_tokens": 53516660.0,
      "step": 23965
    },
    {
      "entropy": 5.414212846755982,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.03125,
      "learning_rate": 0.000428547154171439,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21818125993013382,
      "num_tokens": 53527617.0,
      "step": 23970
    },
    {
      "entropy": 5.407492542266846,
      "epoch": 2.6926489582748356,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20258679389953613,
      "num_tokens": 53537830.0,
      "step": 23975
    },
    {
      "entropy": 5.485403776168823,
      "epoch": 2.6932105351827933,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19991962611675262,
      "num_tokens": 53549981.0,
      "step": 23980
    },
    {
      "entropy": 5.456025266647339,
      "epoch": 2.693772112090751,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004284590921438997,
      "loss": 5.1021,
      "mean_token_accuracy": 0.20904364287853242,
      "num_tokens": 53559941.0,
      "step": 23985
    },
    {
      "entropy": 5.362939500808716,
      "epoch": 2.694333688998708,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042842972836107954,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2077724128961563,
      "num_tokens": 53571611.0,
      "step": 23990
    },
    {
      "entropy": 5.474068450927734,
      "epoch": 2.694895265906666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004284003596929973,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20877393633127211,
      "num_tokens": 53582425.0,
      "step": 23995
    },
    {
      "entropy": 5.48139443397522,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042837098614058815,
      "loss": 5.1904,
      "mean_token_accuracy": 0.20107323974370955,
      "num_tokens": 53593576.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.264745267912811,
      "eval_loss": 5.228978633880615,
      "eval_mean_token_accuracy": 0.20873110280255697,
      "eval_num_tokens": 53593576.0,
      "eval_runtime": 53.2664,
      "eval_samples_per_second": 582.225,
      "eval_steps_per_second": 72.785,
      "step": 24000
    },
    {
      "entropy": 5.4566246509552006,
      "epoch": 2.696018419722581,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.128,
      "mean_token_accuracy": 0.20537324696779252,
      "num_tokens": 53603768.0,
      "step": 24005
    },
    {
      "entropy": 5.440035486221314,
      "epoch": 2.696579996630539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042831222438653016,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21177463829517365,
      "num_tokens": 53614464.0,
      "step": 24010
    },
    {
      "entropy": 5.469366073608398,
      "epoch": 2.697141573538496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.1904,
      "mean_token_accuracy": 0.20033243149518967,
      "num_tokens": 53625641.0,
      "step": 24015
    },
    {
      "entropy": 5.471464347839356,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20710937231779097,
      "num_tokens": 53636741.0,
      "step": 24020
    },
    {
      "entropy": 5.459600353240967,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20821496546268464,
      "num_tokens": 53647295.0,
      "step": 24025
    },
    {
      "entropy": 5.422929096221924,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000428194642307653,
      "loss": 5.0984,
      "mean_token_accuracy": 0.21573614925146103,
      "num_tokens": 53658371.0,
      "step": 24030
    },
    {
      "entropy": 5.414579200744629,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.1287,
      "mean_token_accuracy": 0.2082317814230919,
      "num_tokens": 53669534.0,
      "step": 24035
    },
    {
      "entropy": 5.470734739303589,
      "epoch": 2.699949458078284,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042813582199780913,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20070925056934358,
      "num_tokens": 53681247.0,
      "step": 24040
    },
    {
      "entropy": 5.492310571670532,
      "epoch": 2.700511034986241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.1625,
      "mean_token_accuracy": 0.19946877658367157,
      "num_tokens": 53692197.0,
      "step": 24045
    },
    {
      "entropy": 5.4120701313018795,
      "epoch": 2.701072611894199,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042807698218434964,
      "loss": 5.0902,
      "mean_token_accuracy": 0.20528240352869034,
      "num_tokens": 53703734.0,
      "step": 24050
    },
    {
      "entropy": 5.4512107372283936,
      "epoch": 2.7016341888021564,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20667815655469896,
      "num_tokens": 53714888.0,
      "step": 24055
    },
    {
      "entropy": 5.440464019775391,
      "epoch": 2.702195765710114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042801812287476836,
      "loss": 5.0833,
      "mean_token_accuracy": 0.20726208090782167,
      "num_tokens": 53726231.0,
      "step": 24060
    },
    {
      "entropy": 5.452973651885986,
      "epoch": 2.7027573426180718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.1309,
      "mean_token_accuracy": 0.20522993355989455,
      "num_tokens": 53737777.0,
      "step": 24065
    },
    {
      "entropy": 5.4332499504089355,
      "epoch": 2.703318919526029,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.1047,
      "mean_token_accuracy": 0.2047575905919075,
      "num_tokens": 53748277.0,
      "step": 24070
    },
    {
      "entropy": 5.417534303665161,
      "epoch": 2.7038804964339866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042792979737156717,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20893512517213822,
      "num_tokens": 53759205.0,
      "step": 24075
    },
    {
      "entropy": 5.456623649597168,
      "epoch": 2.7044420733419443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.1381,
      "mean_token_accuracy": 0.1971593901515007,
      "num_tokens": 53770984.0,
      "step": 24080
    },
    {
      "entropy": 5.399263763427735,
      "epoch": 2.7050036502499015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004278708893544837,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21074981838464737,
      "num_tokens": 53782092.0,
      "step": 24085
    },
    {
      "entropy": 5.430385732650757,
      "epoch": 2.705565227157859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.0993,
      "mean_token_accuracy": 0.20223431289196014,
      "num_tokens": 53794136.0,
      "step": 24090
    },
    {
      "entropy": 5.469253921508789,
      "epoch": 2.706126804065817,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.1444,
      "mean_token_accuracy": 0.2013072982430458,
      "num_tokens": 53804612.0,
      "step": 24095
    },
    {
      "entropy": 5.3991382122039795,
      "epoch": 2.7066883809737745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.0821,
      "mean_token_accuracy": 0.2028146728873253,
      "num_tokens": 53815850.0,
      "step": 24100
    },
    {
      "entropy": 5.405112600326538,
      "epoch": 2.707249957881732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004277530149182014,
      "loss": 5.0626,
      "mean_token_accuracy": 0.2114321395754814,
      "num_tokens": 53827050.0,
      "step": 24105
    },
    {
      "entropy": 5.45383358001709,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.1521,
      "mean_token_accuracy": 0.2024977445602417,
      "num_tokens": 53837697.0,
      "step": 24110
    },
    {
      "entropy": 5.39859037399292,
      "epoch": 2.708373111697647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042769404851401524,
      "loss": 5.0876,
      "mean_token_accuracy": 0.2064247488975525,
      "num_tokens": 53849278.0,
      "step": 24115
    },
    {
      "entropy": 5.457406330108642,
      "epoch": 2.7089346886056047,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.1627,
      "mean_token_accuracy": 0.2008380874991417,
      "num_tokens": 53861951.0,
      "step": 24120
    },
    {
      "entropy": 5.491641283035278,
      "epoch": 2.709496265513562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042763506266247834,
      "loss": 5.01,
      "mean_token_accuracy": 0.21592295914888382,
      "num_tokens": 53873647.0,
      "step": 24125
    },
    {
      "entropy": 5.3927130699157715,
      "epoch": 2.7100578424215196,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004276055624463009,
      "loss": 5.0721,
      "mean_token_accuracy": 0.20682420134544371,
      "num_tokens": 53885063.0,
      "step": 24130
    },
    {
      "entropy": 5.398015785217285,
      "epoch": 2.7106194193294773,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042757605737110324,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20371937602758408,
      "num_tokens": 53896955.0,
      "step": 24135
    },
    {
      "entropy": 5.37937445640564,
      "epoch": 2.7111809962374345,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004275465474378246,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21476583182811737,
      "num_tokens": 53907862.0,
      "step": 24140
    },
    {
      "entropy": 5.460904836654663,
      "epoch": 2.711742573145392,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20618511587381363,
      "num_tokens": 53918918.0,
      "step": 24145
    },
    {
      "entropy": 5.446791553497315,
      "epoch": 2.71230415005335,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.1349,
      "mean_token_accuracy": 0.20653711557388305,
      "num_tokens": 53930488.0,
      "step": 24150
    },
    {
      "entropy": 5.50650806427002,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004274579884989008,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20637646168470383,
      "num_tokens": 53942314.0,
      "step": 24155
    },
    {
      "entropy": 5.451483106613159,
      "epoch": 2.713427303869265,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042742845914269645,
      "loss": 5.156,
      "mean_token_accuracy": 0.2225661799311638,
      "num_tokens": 53951915.0,
      "step": 24160
    },
    {
      "entropy": 5.4005388736724855,
      "epoch": 2.7139888807772223,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.1399,
      "mean_token_accuracy": 0.21265121400356293,
      "num_tokens": 53962815.0,
      "step": 24165
    },
    {
      "entropy": 5.375163173675537,
      "epoch": 2.71455045768518,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042736938587108447,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2101823642849922,
      "num_tokens": 53972740.0,
      "step": 24170
    },
    {
      "entropy": 5.457304859161377,
      "epoch": 2.7151120345931377,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2050396218895912,
      "num_tokens": 53983632.0,
      "step": 24175
    },
    {
      "entropy": 5.486858558654785,
      "epoch": 2.715673611501095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.1719,
      "mean_token_accuracy": 0.20705492347478865,
      "num_tokens": 53994085.0,
      "step": 24180
    },
    {
      "entropy": 5.443900537490845,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.125,
      "learning_rate": 0.000427280739579766,
      "loss": 5.074,
      "mean_token_accuracy": 0.21638650000095366,
      "num_tokens": 54004106.0,
      "step": 24185
    },
    {
      "entropy": 5.323067760467529,
      "epoch": 2.71679676531701,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042725118111738305,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21599481999874115,
      "num_tokens": 54014582.0,
      "step": 24190
    },
    {
      "entropy": 5.470783519744873,
      "epoch": 2.717358342224968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.1151,
      "mean_token_accuracy": 0.20614746510982512,
      "num_tokens": 54025626.0,
      "step": 24195
    },
    {
      "entropy": 5.464997911453247,
      "epoch": 2.7179199191329255,
      "grad_norm": 1.125,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.1427,
      "mean_token_accuracy": 0.20668617486953736,
      "num_tokens": 54036974.0,
      "step": 24200
    },
    {
      "entropy": 5.451187992095948,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20522376894950867,
      "num_tokens": 54048404.0,
      "step": 24205
    },
    {
      "entropy": 5.439457845687866,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.171,
      "mean_token_accuracy": 0.2101066827774048,
      "num_tokens": 54059210.0,
      "step": 24210
    },
    {
      "entropy": 5.476410293579102,
      "epoch": 2.719604649856798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20608214735984803,
      "num_tokens": 54070406.0,
      "step": 24215
    },
    {
      "entropy": 5.406866359710693,
      "epoch": 2.7201662267647553,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042707372857356615,
      "loss": 5.09,
      "mean_token_accuracy": 0.20155340284109116,
      "num_tokens": 54081497.0,
      "step": 24220
    },
    {
      "entropy": 5.333953428268432,
      "epoch": 2.720727803672713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004270441361967994,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22134315073490143,
      "num_tokens": 54090934.0,
      "step": 24225
    },
    {
      "entropy": 5.292768859863282,
      "epoch": 2.7212893805806706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004270145389788883,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21836819499731064,
      "num_tokens": 54102011.0,
      "step": 24230
    },
    {
      "entropy": 5.477162504196167,
      "epoch": 2.721850957488628,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.1673,
      "mean_token_accuracy": 0.2071761280298233,
      "num_tokens": 54113066.0,
      "step": 24235
    },
    {
      "entropy": 5.386420965194702,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004269553300234031,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21336116790771484,
      "num_tokens": 54124291.0,
      "step": 24240
    },
    {
      "entropy": 5.509314346313476,
      "epoch": 2.722974111304543,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042692571828771423,
      "loss": 5.2382,
      "mean_token_accuracy": 0.1983680620789528,
      "num_tokens": 54136147.0,
      "step": 24245
    },
    {
      "entropy": 5.48182110786438,
      "epoch": 2.723535688212501,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042689610171465147,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20453879684209825,
      "num_tokens": 54147953.0,
      "step": 24250
    },
    {
      "entropy": 5.398506498336792,
      "epoch": 2.7240972651204585,
      "grad_norm": 1.03125,
      "learning_rate": 0.000426866480305158,
      "loss": 5.083,
      "mean_token_accuracy": 0.21315261572599412,
      "num_tokens": 54159498.0,
      "step": 24255
    },
    {
      "entropy": 5.474379444122315,
      "epoch": 2.7246588420284157,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.077,
      "mean_token_accuracy": 0.2068095788359642,
      "num_tokens": 54170380.0,
      "step": 24260
    },
    {
      "entropy": 5.437729597091675,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20423548817634582,
      "num_tokens": 54181105.0,
      "step": 24265
    },
    {
      "entropy": 5.595462942123413,
      "epoch": 2.725781995844331,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19992727935314178,
      "num_tokens": 54192142.0,
      "step": 24270
    },
    {
      "entropy": 5.384733104705811,
      "epoch": 2.7263435727522882,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004267479463217415,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20343977957963943,
      "num_tokens": 54202716.0,
      "step": 24275
    },
    {
      "entropy": 5.321061754226685,
      "epoch": 2.726905149660246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042671830074424453,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21796441376209258,
      "num_tokens": 54213035.0,
      "step": 24280
    },
    {
      "entropy": 5.379977369308472,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000426688650335978,
      "loss": 5.0463,
      "mean_token_accuracy": 0.20563970357179642,
      "num_tokens": 54223999.0,
      "step": 24285
    },
    {
      "entropy": 5.413325071334839,
      "epoch": 2.728028303476161,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042665899509788593,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21054293513298034,
      "num_tokens": 54234393.0,
      "step": 24290
    },
    {
      "entropy": 5.372700643539429,
      "epoch": 2.728589880384119,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004266293350309128,
      "loss": 5.064,
      "mean_token_accuracy": 0.21209191530942917,
      "num_tokens": 54244787.0,
      "step": 24295
    },
    {
      "entropy": 5.435005950927734,
      "epoch": 2.729151457292076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.1285,
      "mean_token_accuracy": 0.2021153002977371,
      "num_tokens": 54256033.0,
      "step": 24300
    },
    {
      "entropy": 5.499114751815796,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042657000041410026,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21078140139579774,
      "num_tokens": 54266796.0,
      "step": 24305
    },
    {
      "entropy": 5.504781866073609,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.1797,
      "mean_token_accuracy": 0.19927702099084854,
      "num_tokens": 54279049.0,
      "step": 24310
    },
    {
      "entropy": 5.42168493270874,
      "epoch": 2.7308361880159486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.12,
      "mean_token_accuracy": 0.20550689995288848,
      "num_tokens": 54290544.0,
      "step": 24315
    },
    {
      "entropy": 5.47007417678833,
      "epoch": 2.7313977649239063,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20969361960887908,
      "num_tokens": 54300905.0,
      "step": 24320
    },
    {
      "entropy": 5.410911989212036,
      "epoch": 2.731959341831864,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042645127327546405,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21228172332048417,
      "num_tokens": 54313019.0,
      "step": 24325
    },
    {
      "entropy": 5.461646366119385,
      "epoch": 2.732520918739821,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20129934549331666,
      "num_tokens": 54324034.0,
      "step": 24330
    },
    {
      "entropy": 5.3681559562683105,
      "epoch": 2.733082495647779,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004263918807687616,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21071934849023818,
      "num_tokens": 54335335.0,
      "step": 24335
    },
    {
      "entropy": 5.479773044586182,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004263621772843733,
      "loss": 5.1669,
      "mean_token_accuracy": 0.20655296593904496,
      "num_tokens": 54345659.0,
      "step": 24340
    },
    {
      "entropy": 5.538153457641601,
      "epoch": 2.734205649463694,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.1673,
      "mean_token_accuracy": 0.19851786494255066,
      "num_tokens": 54357959.0,
      "step": 24345
    },
    {
      "entropy": 5.406682538986206,
      "epoch": 2.734767226371652,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20787941366434098,
      "num_tokens": 54368779.0,
      "step": 24350
    },
    {
      "entropy": 5.444190120697021,
      "epoch": 2.735328803279609,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.136,
      "mean_token_accuracy": 0.20359240025281905,
      "num_tokens": 54381726.0,
      "step": 24355
    },
    {
      "entropy": 5.4878990173339846,
      "epoch": 2.7358903801875667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.1169,
      "mean_token_accuracy": 0.2079822063446045,
      "num_tokens": 54393201.0,
      "step": 24360
    },
    {
      "entropy": 5.476948499679565,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042621358758989584,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2154635712504387,
      "num_tokens": 54404178.0,
      "step": 24365
    },
    {
      "entropy": 5.315859937667847,
      "epoch": 2.7370135340034816,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261838552031169,
      "loss": 4.967,
      "mean_token_accuracy": 0.21475684642791748,
      "num_tokens": 54414963.0,
      "step": 24370
    },
    {
      "entropy": 5.328534507751465,
      "epoch": 2.7375751109114392,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042615411800258523,
      "loss": 4.986,
      "mean_token_accuracy": 0.2213350936770439,
      "num_tokens": 54425333.0,
      "step": 24375
    },
    {
      "entropy": 5.452703809738159,
      "epoch": 2.738136687819397,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.2022,
      "mean_token_accuracy": 0.20067482441663742,
      "num_tokens": 54436326.0,
      "step": 24380
    },
    {
      "entropy": 5.41722617149353,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.0737,
      "mean_token_accuracy": 0.21049940437078477,
      "num_tokens": 54446664.0,
      "step": 24385
    },
    {
      "entropy": 5.446616792678833,
      "epoch": 2.739259841635312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.0793,
      "mean_token_accuracy": 0.21610225439071656,
      "num_tokens": 54457135.0,
      "step": 24390
    },
    {
      "entropy": 5.292341184616089,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042603512108187224,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2067853406071663,
      "num_tokens": 54468248.0,
      "step": 24395
    },
    {
      "entropy": 5.390209817886353,
      "epoch": 2.740382995451227,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042600535982678347,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2175549015402794,
      "num_tokens": 54478676.0,
      "step": 24400
    },
    {
      "entropy": 5.393702554702759,
      "epoch": 2.7409445723591848,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042597559376362544,
      "loss": 5.0664,
      "mean_token_accuracy": 0.19907500743865966,
      "num_tokens": 54490342.0,
      "step": 24405
    },
    {
      "entropy": 5.3996148109436035,
      "epoch": 2.741506149267142,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004259458228933459,
      "loss": 5.0678,
      "mean_token_accuracy": 0.20970346480607988,
      "num_tokens": 54501543.0,
      "step": 24410
    },
    {
      "entropy": 5.405019283294678,
      "epoch": 2.7420677261750996,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042591604721689297,
      "loss": 5.0867,
      "mean_token_accuracy": 0.20407354682683945,
      "num_tokens": 54511777.0,
      "step": 24415
    },
    {
      "entropy": 5.477287530899048,
      "epoch": 2.7426293030830573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004258862667352145,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21167310923337937,
      "num_tokens": 54523255.0,
      "step": 24420
    },
    {
      "entropy": 5.391017293930053,
      "epoch": 2.7431908799910145,
      "grad_norm": 1.125,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.1104,
      "mean_token_accuracy": 0.20442616194486618,
      "num_tokens": 54534158.0,
      "step": 24425
    },
    {
      "entropy": 5.488002681732178,
      "epoch": 2.743752456898972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000425826691359974,
      "loss": 5.1695,
      "mean_token_accuracy": 0.20408032238483428,
      "num_tokens": 54545735.0,
      "step": 24430
    },
    {
      "entropy": 5.513960886001587,
      "epoch": 2.74431403380693,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20767418295145035,
      "num_tokens": 54557112.0,
      "step": 24435
    },
    {
      "entropy": 5.575674247741699,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.25,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.1707,
      "mean_token_accuracy": 0.203592249751091,
      "num_tokens": 54568600.0,
      "step": 24440
    },
    {
      "entropy": 5.41973352432251,
      "epoch": 2.745437187622845,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000425737292281633,
      "loss": 5.047,
      "mean_token_accuracy": 0.21083619445562363,
      "num_tokens": 54579008.0,
      "step": 24445
    },
    {
      "entropy": 5.369713497161865,
      "epoch": 2.7459987645308024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.1577,
      "mean_token_accuracy": 0.2035520628094673,
      "num_tokens": 54590306.0,
      "step": 24450
    },
    {
      "entropy": 5.438015651702881,
      "epoch": 2.74656034143876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20103263407945632,
      "num_tokens": 54600999.0,
      "step": 24455
    },
    {
      "entropy": 5.48735933303833,
      "epoch": 2.7471219183467177,
      "grad_norm": 1.125,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20745571851730346,
      "num_tokens": 54612977.0,
      "step": 24460
    },
    {
      "entropy": 5.430905675888061,
      "epoch": 2.747683495254675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004256180263214687,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2206955835223198,
      "num_tokens": 54623597.0,
      "step": 24465
    },
    {
      "entropy": 5.329027366638184,
      "epoch": 2.7482450721626326,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042558819783971295,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2219969391822815,
      "num_tokens": 54634618.0,
      "step": 24470
    },
    {
      "entropy": 5.402643489837646,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.1026,
      "mean_token_accuracy": 0.21027155071496964,
      "num_tokens": 54645291.0,
      "step": 24475
    },
    {
      "entropy": 5.425688171386719,
      "epoch": 2.749368225978548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.1486,
      "mean_token_accuracy": 0.20178811848163605,
      "num_tokens": 54657294.0,
      "step": 24480
    },
    {
      "entropy": 5.461173582077026,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042549868362952556,
      "loss": 5.0722,
      "mean_token_accuracy": 0.20499977916479112,
      "num_tokens": 54668627.0,
      "step": 24485
    },
    {
      "entropy": 5.45303201675415,
      "epoch": 2.750491379794463,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042546883597432327,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2072302669286728,
      "num_tokens": 54679763.0,
      "step": 24490
    },
    {
      "entropy": 5.414893913269043,
      "epoch": 2.7510529567024204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042543898352813485,
      "loss": 5.074,
      "mean_token_accuracy": 0.21302530467510222,
      "num_tokens": 54690476.0,
      "step": 24495
    },
    {
      "entropy": 5.468398237228394,
      "epoch": 2.751614533610378,
      "grad_norm": 1.0625,
      "learning_rate": 0.000425409126291911,
      "loss": 5.133,
      "mean_token_accuracy": 0.2126018524169922,
      "num_tokens": 54701383.0,
      "step": 24500
    },
    {
      "entropy": 5.439736557006836,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004253792642666021,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21369703859090805,
      "num_tokens": 54712268.0,
      "step": 24505
    },
    {
      "entropy": 5.373271083831787,
      "epoch": 2.752737687426293,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042534939745315914,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21130848079919815,
      "num_tokens": 54722834.0,
      "step": 24510
    },
    {
      "entropy": 5.419625329971313,
      "epoch": 2.7532992643342507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000425319525852533,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21300197392702103,
      "num_tokens": 54734628.0,
      "step": 24515
    },
    {
      "entropy": 5.335749626159668,
      "epoch": 2.753860841242208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004252896494656749,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21787977069616318,
      "num_tokens": 54745927.0,
      "step": 24520
    },
    {
      "entropy": 5.371732997894287,
      "epoch": 2.7544224181501655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004252597682935362,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21040948033332824,
      "num_tokens": 54757822.0,
      "step": 24525
    },
    {
      "entropy": 5.386055421829224,
      "epoch": 2.754983995058123,
      "grad_norm": 1.078125,
      "learning_rate": 0.000425229882337068,
      "loss": 5.0798,
      "mean_token_accuracy": 0.21604579985141753,
      "num_tokens": 54768964.0,
      "step": 24530
    },
    {
      "entropy": 5.408090734481812,
      "epoch": 2.755545571966081,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042519999159722216,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20771943777799606,
      "num_tokens": 54779471.0,
      "step": 24535
    },
    {
      "entropy": 5.42512993812561,
      "epoch": 2.7561071488740385,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042517009607495035,
      "loss": 5.024,
      "mean_token_accuracy": 0.2126488447189331,
      "num_tokens": 54790426.0,
      "step": 24540
    },
    {
      "entropy": 5.324272632598877,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21275423020124434,
      "num_tokens": 54801765.0,
      "step": 24545
    },
    {
      "entropy": 5.400925779342652,
      "epoch": 2.7572303026899534,
      "grad_norm": 1.125,
      "learning_rate": 0.0004251102906869364,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20487612932920457,
      "num_tokens": 54814272.0,
      "step": 24550
    },
    {
      "entropy": 5.442745733261108,
      "epoch": 2.757791879597911,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20957311391830444,
      "num_tokens": 54825610.0,
      "step": 24555
    },
    {
      "entropy": 5.53571572303772,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.1115,
      "mean_token_accuracy": 0.2109365850687027,
      "num_tokens": 54835546.0,
      "step": 24560
    },
    {
      "entropy": 5.435815048217774,
      "epoch": 2.758915033413826,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.1268,
      "mean_token_accuracy": 0.209674634039402,
      "num_tokens": 54845653.0,
      "step": 24565
    },
    {
      "entropy": 5.418931865692139,
      "epoch": 2.7594766103217836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042499062256368987,
      "loss": 5.06,
      "mean_token_accuracy": 0.2147188663482666,
      "num_tokens": 54855514.0,
      "step": 24570
    },
    {
      "entropy": 5.381489992141724,
      "epoch": 2.7600381872297413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20673631131649017,
      "num_tokens": 54866131.0,
      "step": 24575
    },
    {
      "entropy": 5.377145004272461,
      "epoch": 2.760599764137699,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004249307598436984,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2160503476858139,
      "num_tokens": 54876842.0,
      "step": 24580
    },
    {
      "entropy": 5.4714292049407955,
      "epoch": 2.761161341045656,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.1861,
      "mean_token_accuracy": 0.20001502484083175,
      "num_tokens": 54888036.0,
      "step": 24585
    },
    {
      "entropy": 5.449224185943604,
      "epoch": 2.761722917953614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042487087802829275,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2089020386338234,
      "num_tokens": 54898501.0,
      "step": 24590
    },
    {
      "entropy": 5.437549352645874,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004248409299621931,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20961105823516846,
      "num_tokens": 54909054.0,
      "step": 24595
    },
    {
      "entropy": 5.443773794174194,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004248109771251,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2034783512353897,
      "num_tokens": 54919572.0,
      "step": 24600
    },
    {
      "entropy": 5.520026922225952,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.1203,
      "mean_token_accuracy": 0.2069982945919037,
      "num_tokens": 54930929.0,
      "step": 24605
    },
    {
      "entropy": 5.393429231643677,
      "epoch": 2.763969225585444,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004247510571417487,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21322440952062607,
      "num_tokens": 54941317.0,
      "step": 24610
    },
    {
      "entropy": 5.4037638187408445,
      "epoch": 2.764530802493401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004247210899973985,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21803749799728395,
      "num_tokens": 54951203.0,
      "step": 24615
    },
    {
      "entropy": 5.402284812927246,
      "epoch": 2.765092379401359,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20650673508644105,
      "num_tokens": 54961485.0,
      "step": 24620
    },
    {
      "entropy": 5.404002904891968,
      "epoch": 2.7656539563093165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.077,
      "mean_token_accuracy": 0.2080414593219757,
      "num_tokens": 54973152.0,
      "step": 24625
    },
    {
      "entropy": 5.426065254211426,
      "epoch": 2.766215533217274,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042463115996509984,
      "loss": 5.0852,
      "mean_token_accuracy": 0.2089973956346512,
      "num_tokens": 54984863.0,
      "step": 24630
    },
    {
      "entropy": 5.45213475227356,
      "epoch": 2.766777110125232,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.0909,
      "mean_token_accuracy": 0.20593977868556976,
      "num_tokens": 54997382.0,
      "step": 24635
    },
    {
      "entropy": 5.492494773864746,
      "epoch": 2.767338687033189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.12,
      "mean_token_accuracy": 0.20525524914264678,
      "num_tokens": 55007686.0,
      "step": 24640
    },
    {
      "entropy": 5.45709457397461,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.1743,
      "mean_token_accuracy": 0.2022397741675377,
      "num_tokens": 55018776.0,
      "step": 24645
    },
    {
      "entropy": 5.423551416397094,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004245111865594273,
      "loss": 5.1462,
      "mean_token_accuracy": 0.2047217771410942,
      "num_tokens": 55031566.0,
      "step": 24650
    },
    {
      "entropy": 5.3992531299591064,
      "epoch": 2.7690234177570616,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042448118130438583,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22122963070869445,
      "num_tokens": 55043321.0,
      "step": 24655
    },
    {
      "entropy": 5.388670444488525,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042445117128980553,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21433960646390915,
      "num_tokens": 55052655.0,
      "step": 24660
    },
    {
      "entropy": 5.3742739200592045,
      "epoch": 2.770146571572977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20731837153434754,
      "num_tokens": 55063317.0,
      "step": 24665
    },
    {
      "entropy": 5.357176828384399,
      "epoch": 2.7707081484809346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042439113698585103,
      "loss": 5.031,
      "mean_token_accuracy": 0.2081143870949745,
      "num_tokens": 55074916.0,
      "step": 24670
    },
    {
      "entropy": 5.392309188842773,
      "epoch": 2.7712697253888923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004243611126983882,
      "loss": 5.0394,
      "mean_token_accuracy": 0.21162784695625306,
      "num_tokens": 55086456.0,
      "step": 24675
    },
    {
      "entropy": 5.519128894805908,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1950395181775093,
      "num_tokens": 55097268.0,
      "step": 24680
    },
    {
      "entropy": 5.459381866455078,
      "epoch": 2.772392879204807,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042430104985727155,
      "loss": 5.0465,
      "mean_token_accuracy": 0.20819493234157563,
      "num_tokens": 55107451.0,
      "step": 24685
    },
    {
      "entropy": 5.461480951309204,
      "epoch": 2.772954456112765,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042427101130553,
      "loss": 5.0761,
      "mean_token_accuracy": 0.20967528223991394,
      "num_tokens": 55117781.0,
      "step": 24690
    },
    {
      "entropy": 5.514344215393066,
      "epoch": 2.773516033020722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.1366,
      "mean_token_accuracy": 0.20413052141666413,
      "num_tokens": 55129519.0,
      "step": 24695
    },
    {
      "entropy": 5.347922325134277,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004242109199444629,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20814354419708253,
      "num_tokens": 55138964.0,
      "step": 24700
    },
    {
      "entropy": 5.437012386322022,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.1051,
      "mean_token_accuracy": 0.20684029906988144,
      "num_tokens": 55149953.0,
      "step": 24705
    },
    {
      "entropy": 5.469457197189331,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.125,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20007234811782837,
      "num_tokens": 55162208.0,
      "step": 24710
    },
    {
      "entropy": 5.452560901641846,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004241207472732531,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20340191423892975,
      "num_tokens": 55174014.0,
      "step": 24715
    },
    {
      "entropy": 5.490642070770264,
      "epoch": 2.77632391756051,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.1887,
      "mean_token_accuracy": 0.19798362255096436,
      "num_tokens": 55185173.0,
      "step": 24720
    },
    {
      "entropy": 5.4969648838043215,
      "epoch": 2.7768854944684676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.1463,
      "mean_token_accuracy": 0.2107707753777504,
      "num_tokens": 55198404.0,
      "step": 24725
    },
    {
      "entropy": 5.3858277797698975,
      "epoch": 2.777447071376425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042403053186948237,
      "loss": 5.066,
      "mean_token_accuracy": 0.2064301088452339,
      "num_tokens": 55209126.0,
      "step": 24730
    },
    {
      "entropy": 5.402226066589355,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004240004505765688,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20742869824171067,
      "num_tokens": 55220311.0,
      "step": 24735
    },
    {
      "entropy": 5.381696128845215,
      "epoch": 2.77857022519234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004239703645394231,
      "loss": 5.0557,
      "mean_token_accuracy": 0.21231485158205032,
      "num_tokens": 55230525.0,
      "step": 24740
    },
    {
      "entropy": 5.532253646850586,
      "epoch": 2.7791318021002978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.2063,
      "mean_token_accuracy": 0.20173334181308747,
      "num_tokens": 55242084.0,
      "step": 24745
    },
    {
      "entropy": 5.459042596817016,
      "epoch": 2.779693379008255,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042391017823626716,
      "loss": 5.0501,
      "mean_token_accuracy": 0.20643051266670226,
      "num_tokens": 55253677.0,
      "step": 24750
    },
    {
      "entropy": 5.348340272903442,
      "epoch": 2.7802549559162126,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004238800779721734,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21499786674976348,
      "num_tokens": 55265118.0,
      "step": 24755
    },
    {
      "entropy": 5.360972166061401,
      "epoch": 2.7808165328241703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042384997296768007,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21303496807813643,
      "num_tokens": 55277119.0,
      "step": 24760
    },
    {
      "entropy": 5.4453967094421385,
      "epoch": 2.781378109732128,
      "grad_norm": 1.078125,
      "learning_rate": 0.000423819863223746,
      "loss": 5.041,
      "mean_token_accuracy": 0.21168179363012313,
      "num_tokens": 55287828.0,
      "step": 24765
    },
    {
      "entropy": 5.520501756668091,
      "epoch": 2.7819396866400856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.1776,
      "mean_token_accuracy": 0.2049567297101021,
      "num_tokens": 55299115.0,
      "step": 24770
    },
    {
      "entropy": 5.381559610366821,
      "epoch": 2.782501263548043,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004237596295213901,
      "loss": 5.0684,
      "mean_token_accuracy": 0.2069260984659195,
      "num_tokens": 55310532.0,
      "step": 24775
    },
    {
      "entropy": 5.4478761672973635,
      "epoch": 2.7830628404560005,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004237295055648862,
      "loss": 5.151,
      "mean_token_accuracy": 0.20224139839410782,
      "num_tokens": 55321127.0,
      "step": 24780
    },
    {
      "entropy": 5.504236173629761,
      "epoch": 2.783624417363958,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.0912,
      "mean_token_accuracy": 0.218447507917881,
      "num_tokens": 55331791.0,
      "step": 24785
    },
    {
      "entropy": 5.455613851547241,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20550192147493362,
      "num_tokens": 55342724.0,
      "step": 24790
    },
    {
      "entropy": 5.377570390701294,
      "epoch": 2.784747571179873,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004236391052855811,
      "loss": 5.0545,
      "mean_token_accuracy": 0.20663844794034958,
      "num_tokens": 55354471.0,
      "step": 24795
    },
    {
      "entropy": 5.440036916732788,
      "epoch": 2.7853091480878307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042360896239241306,
      "loss": 5.0813,
      "mean_token_accuracy": 0.20794207453727723,
      "num_tokens": 55366273.0,
      "step": 24800
    },
    {
      "entropy": 5.412883234024048,
      "epoch": 2.785870724995788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042357881476747807,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20865770876407624,
      "num_tokens": 55377384.0,
      "step": 24805
    },
    {
      "entropy": 5.396326398849487,
      "epoch": 2.7864323019037456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000423548662411736,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2070961207151413,
      "num_tokens": 55387348.0,
      "step": 24810
    },
    {
      "entropy": 5.38743405342102,
      "epoch": 2.7869938788117032,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004235185053261469,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21779301166534423,
      "num_tokens": 55398484.0,
      "step": 24815
    },
    {
      "entropy": 5.41792254447937,
      "epoch": 2.787555455719661,
      "grad_norm": 1.15625,
      "learning_rate": 0.000423488343511671,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21162512004375458,
      "num_tokens": 55409347.0,
      "step": 24820
    },
    {
      "entropy": 5.369329738616943,
      "epoch": 2.7881170326276186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004234581769692686,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20759902745485306,
      "num_tokens": 55421048.0,
      "step": 24825
    },
    {
      "entropy": 5.348599243164062,
      "epoch": 2.788678609535576,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004234280056999003,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2178695484995842,
      "num_tokens": 55431189.0,
      "step": 24830
    },
    {
      "entropy": 5.4659933090209964,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004233978297045267,
      "loss": 5.1019,
      "mean_token_accuracy": 0.2051868185400963,
      "num_tokens": 55442695.0,
      "step": 24835
    },
    {
      "entropy": 5.555420351028443,
      "epoch": 2.789801763351491,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.1261,
      "mean_token_accuracy": 0.20582433342933654,
      "num_tokens": 55454954.0,
      "step": 24840
    },
    {
      "entropy": 5.455147981643677,
      "epoch": 2.7903633402594483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20844314247369766,
      "num_tokens": 55466270.0,
      "step": 24845
    },
    {
      "entropy": 5.483736801147461,
      "epoch": 2.790924917167406,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.168,
      "mean_token_accuracy": 0.20762641727924347,
      "num_tokens": 55476925.0,
      "step": 24850
    },
    {
      "entropy": 5.424216604232788,
      "epoch": 2.7914864940753636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004232770784821984,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2076100677251816,
      "num_tokens": 55489705.0,
      "step": 24855
    },
    {
      "entropy": 5.504993295669555,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.1933,
      "mean_token_accuracy": 0.19971868246793748,
      "num_tokens": 55501362.0,
      "step": 24860
    },
    {
      "entropy": 5.463320922851563,
      "epoch": 2.792609647891279,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.155,
      "mean_token_accuracy": 0.20492017418146133,
      "num_tokens": 55512754.0,
      "step": 24865
    },
    {
      "entropy": 5.428063917160034,
      "epoch": 2.793171224799236,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042318646548949177,
      "loss": 5.0734,
      "mean_token_accuracy": 0.2005606010556221,
      "num_tokens": 55524057.0,
      "step": 24870
    },
    {
      "entropy": 5.422874784469604,
      "epoch": 2.793732801707194,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042315625172067794,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2119520127773285,
      "num_tokens": 55534768.0,
      "step": 24875
    },
    {
      "entropy": 5.529557609558106,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.1548,
      "mean_token_accuracy": 0.20033008158206939,
      "num_tokens": 55547135.0,
      "step": 24880
    },
    {
      "entropy": 5.457588529586792,
      "epoch": 2.7948559555231087,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20453896820545198,
      "num_tokens": 55559638.0,
      "step": 24885
    },
    {
      "entropy": 5.433962202072143,
      "epoch": 2.7954175324310664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2136334151029587,
      "num_tokens": 55570985.0,
      "step": 24890
    },
    {
      "entropy": 5.366181373596191,
      "epoch": 2.795979109339024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004230353494815203,
      "loss": 5.0482,
      "mean_token_accuracy": 0.20984814316034317,
      "num_tokens": 55581701.0,
      "step": 24895
    },
    {
      "entropy": 5.437272262573242,
      "epoch": 2.7965406862469813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004230051121355672,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2086991012096405,
      "num_tokens": 55592891.0,
      "step": 24900
    },
    {
      "entropy": 5.368270063400269,
      "epoch": 2.797102263154939,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21459896266460418,
      "num_tokens": 55603762.0,
      "step": 24905
    },
    {
      "entropy": 5.419996833801269,
      "epoch": 2.7976638400628966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.1016,
      "mean_token_accuracy": 0.2074057623744011,
      "num_tokens": 55614450.0,
      "step": 24910
    },
    {
      "entropy": 5.3711591243743895,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004229143718263186,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21090243011713028,
      "num_tokens": 55625240.0,
      "step": 24915
    },
    {
      "entropy": 5.3747786521911625,
      "epoch": 2.798786993878812,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004228841156359829,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21655580550432205,
      "num_tokens": 55635646.0,
      "step": 24920
    },
    {
      "entropy": 5.360005855560303,
      "epoch": 2.799348570786769,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004228538547369596,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21808647811412812,
      "num_tokens": 55646669.0,
      "step": 24925
    },
    {
      "entropy": 5.5630405902862545,
      "epoch": 2.799910147694727,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1984929621219635,
      "num_tokens": 55658792.0,
      "step": 24930
    },
    {
      "entropy": 5.448081350326538,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21051573157310485,
      "num_tokens": 55670037.0,
      "step": 24935
    },
    {
      "entropy": 5.461284923553467,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004227630437974,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20497117340564727,
      "num_tokens": 55680405.0,
      "step": 24940
    },
    {
      "entropy": 5.431045246124268,
      "epoch": 2.8015948784185993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.0668,
      "mean_token_accuracy": 0.210676808655262,
      "num_tokens": 55692199.0,
      "step": 24945
    },
    {
      "entropy": 5.3435344219207765,
      "epoch": 2.802156455326557,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042270247964525755,
      "loss": 5.0389,
      "mean_token_accuracy": 0.2135771095752716,
      "num_tokens": 55702654.0,
      "step": 24950
    },
    {
      "entropy": 5.392324495315552,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2149198606610298,
      "num_tokens": 55712710.0,
      "step": 24955
    },
    {
      "entropy": 5.476361465454102,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.0683,
      "mean_token_accuracy": 0.20856136232614517,
      "num_tokens": 55723429.0,
      "step": 24960
    },
    {
      "entropy": 5.4523883819580075,
      "epoch": 2.8038411860504295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.1231,
      "mean_token_accuracy": 0.21022663712501527,
      "num_tokens": 55735113.0,
      "step": 24965
    },
    {
      "entropy": 5.2966924667358395,
      "epoch": 2.804402762958387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000422581294913838,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21630571484565736,
      "num_tokens": 55745857.0,
      "step": 24970
    },
    {
      "entropy": 5.373834133148193,
      "epoch": 2.804964339866345,
      "grad_norm": 1.125,
      "learning_rate": 0.000422550986980957,
      "loss": 4.9837,
      "mean_token_accuracy": 0.2109202891588211,
      "num_tokens": 55756856.0,
      "step": 24975
    },
    {
      "entropy": 5.495521402359008,
      "epoch": 2.805525916774302,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.1393,
      "mean_token_accuracy": 0.2043999746441841,
      "num_tokens": 55768827.0,
      "step": 24980
    },
    {
      "entropy": 5.423760986328125,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004224903570219181,
      "loss": 5.0731,
      "mean_token_accuracy": 0.2063637286424637,
      "num_tokens": 55780360.0,
      "step": 24985
    },
    {
      "entropy": 5.432508039474487,
      "epoch": 2.8066490705902174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042246003499769077,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20684473365545272,
      "num_tokens": 55791874.0,
      "step": 24990
    },
    {
      "entropy": 5.392110013961792,
      "epoch": 2.8072106474981746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042242970827827874,
      "loss": 5.068,
      "mean_token_accuracy": 0.2107057124376297,
      "num_tokens": 55802689.0,
      "step": 24995
    },
    {
      "entropy": 5.390161991119385,
      "epoch": 2.8077722244061323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004223993768646476,
      "loss": 5.0834,
      "mean_token_accuracy": 0.2129926338791847,
      "num_tokens": 55813945.0,
      "step": 25000
    },
    {
      "entropy": 5.392459726333618,
      "epoch": 2.80833380131409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004223690407577631,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21135923862457276,
      "num_tokens": 55826403.0,
      "step": 25005
    },
    {
      "entropy": 5.375484228134155,
      "epoch": 2.8088953782220476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042233869995859127,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21635821759700774,
      "num_tokens": 55836858.0,
      "step": 25010
    },
    {
      "entropy": 5.408909749984741,
      "epoch": 2.8094569551300053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042230835446809803,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21661725491285325,
      "num_tokens": 55847664.0,
      "step": 25015
    },
    {
      "entropy": 5.381688117980957,
      "epoch": 2.8100185320379625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222780042872496,
      "loss": 5.0456,
      "mean_token_accuracy": 0.2103533685207367,
      "num_tokens": 55858333.0,
      "step": 25020
    },
    {
      "entropy": 5.396325349807739,
      "epoch": 2.81058010894592,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004222476494170125,
      "loss": 5.047,
      "mean_token_accuracy": 0.21342923790216445,
      "num_tokens": 55868819.0,
      "step": 25025
    },
    {
      "entropy": 5.366071033477783,
      "epoch": 2.811141685853878,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222172898583531,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2117368310689926,
      "num_tokens": 55879133.0,
      "step": 25030
    },
    {
      "entropy": 5.357093286514282,
      "epoch": 2.811703262761835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004221869256122381,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21276918351650237,
      "num_tokens": 55890362.0,
      "step": 25035
    },
    {
      "entropy": 5.436327934265137,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004221565566796343,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20895051062107087,
      "num_tokens": 55899920.0,
      "step": 25040
    },
    {
      "entropy": 5.428067255020141,
      "epoch": 2.8128264165777503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004221261830615086,
      "loss": 5.1068,
      "mean_token_accuracy": 0.21027365177869797,
      "num_tokens": 55911598.0,
      "step": 25045
    },
    {
      "entropy": 5.306889295578003,
      "epoch": 2.813387993485708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004220958047588282,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2179407775402069,
      "num_tokens": 55922514.0,
      "step": 25050
    },
    {
      "entropy": 5.3752306461334225,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004220654217725603,
      "loss": 5.026,
      "mean_token_accuracy": 0.21824088245630263,
      "num_tokens": 55932796.0,
      "step": 25055
    },
    {
      "entropy": 5.390900039672852,
      "epoch": 2.814511147301623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.1164,
      "mean_token_accuracy": 0.21280603408813475,
      "num_tokens": 55944337.0,
      "step": 25060
    },
    {
      "entropy": 5.501917552947998,
      "epoch": 2.8150727242095805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000422004641753132,
      "loss": 5.2095,
      "mean_token_accuracy": 0.2033474177122116,
      "num_tokens": 55955871.0,
      "step": 25065
    },
    {
      "entropy": 5.499521541595459,
      "epoch": 2.815634301117538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20398508608341218,
      "num_tokens": 55966091.0,
      "step": 25070
    },
    {
      "entropy": 5.4468787670135494,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042194384301096457,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21080018132925032,
      "num_tokens": 55976755.0,
      "step": 25075
    },
    {
      "entropy": 5.412540531158447,
      "epoch": 2.816757454933453,
      "grad_norm": 1.0,
      "learning_rate": 0.00042191343662127337,
      "loss": 5.1403,
      "mean_token_accuracy": 0.2085412859916687,
      "num_tokens": 55987579.0,
      "step": 25080
    },
    {
      "entropy": 5.419413375854492,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042188302555380144,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20746956914663314,
      "num_tokens": 55997027.0,
      "step": 25085
    },
    {
      "entropy": 5.404504489898682,
      "epoch": 2.817880608749368,
      "grad_norm": 1.1875,
      "learning_rate": 0.000421852609809517,
      "loss": 5.0956,
      "mean_token_accuracy": 0.20655363500118257,
      "num_tokens": 56008928.0,
      "step": 25090
    },
    {
      "entropy": 5.415930509567261,
      "epoch": 2.8184421856573256,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004218221893893884,
      "loss": 5.0922,
      "mean_token_accuracy": 0.21128398329019546,
      "num_tokens": 56021313.0,
      "step": 25095
    },
    {
      "entropy": 5.429538440704346,
      "epoch": 2.8190037625652833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042179176429438454,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2103274419903755,
      "num_tokens": 56032002.0,
      "step": 25100
    },
    {
      "entropy": 5.414218473434448,
      "epoch": 2.819565339473241,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004217613345254739,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2206183671951294,
      "num_tokens": 56042905.0,
      "step": 25105
    },
    {
      "entropy": 5.418306159973144,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042173090008362535,
      "loss": 5.0228,
      "mean_token_accuracy": 0.20768200904130935,
      "num_tokens": 56052507.0,
      "step": 25110
    },
    {
      "entropy": 5.365984010696411,
      "epoch": 2.820688493289156,
      "grad_norm": 1.125,
      "learning_rate": 0.00042170046096980807,
      "loss": 5.105,
      "mean_token_accuracy": 0.20700279474258423,
      "num_tokens": 56065250.0,
      "step": 25115
    },
    {
      "entropy": 5.380886650085449,
      "epoch": 2.8212500701971135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042167001718499124,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21230120062828065,
      "num_tokens": 56075858.0,
      "step": 25120
    },
    {
      "entropy": 5.440095853805542,
      "epoch": 2.821811647105071,
      "grad_norm": 1.125,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.117,
      "mean_token_accuracy": 0.20615849643945694,
      "num_tokens": 56086647.0,
      "step": 25125
    },
    {
      "entropy": 5.495005035400391,
      "epoch": 2.8223732240130284,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19398128092288972,
      "num_tokens": 56098124.0,
      "step": 25130
    },
    {
      "entropy": 5.442276573181152,
      "epoch": 2.822934800920986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004215786578142375,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2106737956404686,
      "num_tokens": 56108505.0,
      "step": 25135
    },
    {
      "entropy": 5.430901432037354,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042154819535511733,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21518958508968353,
      "num_tokens": 56118423.0,
      "step": 25140
    },
    {
      "entropy": 5.469357490539551,
      "epoch": 2.8240579547369014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.0913,
      "mean_token_accuracy": 0.21460759937763213,
      "num_tokens": 56130346.0,
      "step": 25145
    },
    {
      "entropy": 5.355208539962769,
      "epoch": 2.824619531644859,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.1276,
      "mean_token_accuracy": 0.21244096010923386,
      "num_tokens": 56141044.0,
      "step": 25150
    },
    {
      "entropy": 5.378180837631225,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004214567799847289,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21881273984909058,
      "num_tokens": 56151786.0,
      "step": 25155
    },
    {
      "entropy": 5.432152223587036,
      "epoch": 2.825742685460774,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004214262988668242,
      "loss": 5.0657,
      "mean_token_accuracy": 0.20793017297983168,
      "num_tokens": 56164196.0,
      "step": 25160
    },
    {
      "entropy": 5.44922513961792,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.0917,
      "mean_token_accuracy": 0.2125035583972931,
      "num_tokens": 56176259.0,
      "step": 25165
    },
    {
      "entropy": 5.4347686767578125,
      "epoch": 2.8268658392766888,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.1089,
      "mean_token_accuracy": 0.19996924251317977,
      "num_tokens": 56186617.0,
      "step": 25170
    },
    {
      "entropy": 5.320649576187134,
      "epoch": 2.8274274161846464,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042133482754337236,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21559496968984604,
      "num_tokens": 56197609.0,
      "step": 25175
    },
    {
      "entropy": 5.348799800872802,
      "epoch": 2.827988993092604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042130432778221186,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21698755919933319,
      "num_tokens": 56207490.0,
      "step": 25180
    },
    {
      "entropy": 5.392520332336426,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042127382336266467,
      "loss": 5.0497,
      "mean_token_accuracy": 0.2083284616470337,
      "num_tokens": 56218167.0,
      "step": 25185
    },
    {
      "entropy": 5.342153787612915,
      "epoch": 2.829112146908519,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004212433142857024,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21406321078538895,
      "num_tokens": 56229199.0,
      "step": 25190
    },
    {
      "entropy": 5.424461412429809,
      "epoch": 2.8296737238164766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042121280055229624,
      "loss": 5.1895,
      "mean_token_accuracy": 0.20959295630455016,
      "num_tokens": 56240165.0,
      "step": 25195
    },
    {
      "entropy": 5.504427242279053,
      "epoch": 2.8302353007244343,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004211822821634177,
      "loss": 5.2185,
      "mean_token_accuracy": 0.20209736078977586,
      "num_tokens": 56252447.0,
      "step": 25200
    },
    {
      "entropy": 5.455720281600952,
      "epoch": 2.830796877632392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.0764,
      "mean_token_accuracy": 0.2123867467045784,
      "num_tokens": 56263116.0,
      "step": 25205
    },
    {
      "entropy": 5.417319917678833,
      "epoch": 2.831358454540349,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.1177,
      "mean_token_accuracy": 0.20371393710374833,
      "num_tokens": 56276257.0,
      "step": 25210
    },
    {
      "entropy": 5.37516417503357,
      "epoch": 2.831920031448307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042109069907366646,
      "loss": 5.0275,
      "mean_token_accuracy": 0.20809692442417144,
      "num_tokens": 56287865.0,
      "step": 25215
    },
    {
      "entropy": 5.440722560882568,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.0,
      "learning_rate": 0.00042106016207261754,
      "loss": 5.0541,
      "mean_token_accuracy": 0.21161699593067168,
      "num_tokens": 56298122.0,
      "step": 25220
    },
    {
      "entropy": 5.467601108551025,
      "epoch": 2.8330431852642217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.1634,
      "mean_token_accuracy": 0.20722975134849547,
      "num_tokens": 56309455.0,
      "step": 25225
    },
    {
      "entropy": 5.45652928352356,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.1377,
      "mean_token_accuracy": 0.2037941798567772,
      "num_tokens": 56319804.0,
      "step": 25230
    },
    {
      "entropy": 5.400773668289185,
      "epoch": 2.834166339080137,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042096852316968756,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21616577506065368,
      "num_tokens": 56330136.0,
      "step": 25235
    },
    {
      "entropy": 5.547188425064087,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.2416,
      "mean_token_accuracy": 0.2006765842437744,
      "num_tokens": 56341861.0,
      "step": 25240
    },
    {
      "entropy": 5.3975800514221195,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2085522636771202,
      "num_tokens": 56352016.0,
      "step": 25245
    },
    {
      "entropy": 5.344671201705933,
      "epoch": 2.8358510698040096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004208768424375089,
      "loss": 4.9896,
      "mean_token_accuracy": 0.20886125266551972,
      "num_tokens": 56363740.0,
      "step": 25250
    },
    {
      "entropy": 5.413744163513184,
      "epoch": 2.8364126467119672,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042084627290260124,
      "loss": 5.015,
      "mean_token_accuracy": 0.2172751948237419,
      "num_tokens": 56375080.0,
      "step": 25255
    },
    {
      "entropy": 5.426944875717163,
      "epoch": 2.836974223619925,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004208156987238918,
      "loss": 5.1518,
      "mean_token_accuracy": 0.20408067256212234,
      "num_tokens": 56385858.0,
      "step": 25260
    },
    {
      "entropy": 5.491457319259643,
      "epoch": 2.837535800527882,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.1003,
      "mean_token_accuracy": 0.2093324050307274,
      "num_tokens": 56394423.0,
      "step": 25265
    },
    {
      "entropy": 5.395686435699463,
      "epoch": 2.83809737743584,
      "grad_norm": 1.125,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.082,
      "mean_token_accuracy": 0.20782612562179564,
      "num_tokens": 56405843.0,
      "step": 25270
    },
    {
      "entropy": 5.398968696594238,
      "epoch": 2.8386589543437974,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042072394833468765,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2046208620071411,
      "num_tokens": 56417757.0,
      "step": 25275
    },
    {
      "entropy": 5.483303499221802,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.09375,
      "learning_rate": 0.000420693355590507,
      "loss": 5.101,
      "mean_token_accuracy": 0.20583642423152923,
      "num_tokens": 56429937.0,
      "step": 25280
    },
    {
      "entropy": 5.375780010223389,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042066275820739333,
      "loss": 5.0687,
      "mean_token_accuracy": 0.21505827009677886,
      "num_tokens": 56441114.0,
      "step": 25285
    },
    {
      "entropy": 5.360804224014283,
      "epoch": 2.84034368506767,
      "grad_norm": 1.078125,
      "learning_rate": 0.000420632156186321,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20424392372369765,
      "num_tokens": 56451839.0,
      "step": 25290
    },
    {
      "entropy": 5.4408293724060055,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.125,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.1237,
      "mean_token_accuracy": 0.20438411235809326,
      "num_tokens": 56463384.0,
      "step": 25295
    },
    {
      "entropy": 5.479838800430298,
      "epoch": 2.8414668388835853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042057093823419795,
      "loss": 5.058,
      "mean_token_accuracy": 0.20130124390125276,
      "num_tokens": 56474905.0,
      "step": 25300
    },
    {
      "entropy": 5.432728099822998,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042054032230509636,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20566768646240235,
      "num_tokens": 56485319.0,
      "step": 25305
    },
    {
      "entropy": 5.419349718093872,
      "epoch": 2.8425899926995,
      "grad_norm": 1.125,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.1457,
      "mean_token_accuracy": 0.2015333130955696,
      "num_tokens": 56496064.0,
      "step": 25310
    },
    {
      "entropy": 5.378847694396972,
      "epoch": 2.843151569607458,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20579174906015396,
      "num_tokens": 56507167.0,
      "step": 25315
    },
    {
      "entropy": 5.5308290958404545,
      "epoch": 2.843713146515415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004204484467173297,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20846358835697174,
      "num_tokens": 56518400.0,
      "step": 25320
    },
    {
      "entropy": 5.532485246658325,
      "epoch": 2.8442747234233727,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.1975,
      "mean_token_accuracy": 0.20681436657905578,
      "num_tokens": 56529630.0,
      "step": 25325
    },
    {
      "entropy": 5.482250165939331,
      "epoch": 2.8448363003313304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.0963,
      "mean_token_accuracy": 0.20767994374036788,
      "num_tokens": 56539520.0,
      "step": 25330
    },
    {
      "entropy": 5.437097549438477,
      "epoch": 2.845397877239288,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004203565294493494,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2145366281270981,
      "num_tokens": 56550677.0,
      "step": 25335
    },
    {
      "entropy": 5.401930809020996,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.1345,
      "mean_token_accuracy": 0.20786912590265275,
      "num_tokens": 56561694.0,
      "step": 25340
    },
    {
      "entropy": 5.452916812896729,
      "epoch": 2.846521031055203,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.0951,
      "mean_token_accuracy": 0.20399759113788604,
      "num_tokens": 56572322.0,
      "step": 25345
    },
    {
      "entropy": 5.473609352111817,
      "epoch": 2.8470826079631606,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20242003500461578,
      "num_tokens": 56583198.0,
      "step": 25350
    },
    {
      "entropy": 5.4217695713043215,
      "epoch": 2.8476441848711183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.1666,
      "mean_token_accuracy": 0.21060453355312347,
      "num_tokens": 56594732.0,
      "step": 25355
    },
    {
      "entropy": 5.455605411529541,
      "epoch": 2.8482057617790755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000420203241451565,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20967237949371337,
      "num_tokens": 56605907.0,
      "step": 25360
    },
    {
      "entropy": 5.525914049148559,
      "epoch": 2.848767338687033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042017256997812016,
      "loss": 5.171,
      "mean_token_accuracy": 0.20431709587574004,
      "num_tokens": 56616841.0,
      "step": 25365
    },
    {
      "entropy": 5.446662712097168,
      "epoch": 2.849328915594991,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.1093,
      "mean_token_accuracy": 0.2035503253340721,
      "num_tokens": 56628044.0,
      "step": 25370
    },
    {
      "entropy": 5.427626895904541,
      "epoch": 2.849890492502948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20415647327899933,
      "num_tokens": 56638350.0,
      "step": 25375
    },
    {
      "entropy": 5.453616428375244,
      "epoch": 2.8504520694109057,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004200805278275873,
      "loss": 5.0671,
      "mean_token_accuracy": 0.2088095635175705,
      "num_tokens": 56649923.0,
      "step": 25380
    },
    {
      "entropy": 5.336795377731323,
      "epoch": 2.8510136463188633,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004200498378705997,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21508701592683793,
      "num_tokens": 56661210.0,
      "step": 25385
    },
    {
      "entropy": 5.530411958694458,
      "epoch": 2.851575223226821,
      "grad_norm": 1.125,
      "learning_rate": 0.0004200191432951692,
      "loss": 5.1903,
      "mean_token_accuracy": 0.19953951835632325,
      "num_tokens": 56673106.0,
      "step": 25390
    },
    {
      "entropy": 5.433280324935913,
      "epoch": 2.8521368001347787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.0897,
      "mean_token_accuracy": 0.21068581193685532,
      "num_tokens": 56684703.0,
      "step": 25395
    },
    {
      "entropy": 5.4512251853942875,
      "epoch": 2.852698377042736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041995774029288807,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20749808996915817,
      "num_tokens": 56696881.0,
      "step": 25400
    },
    {
      "entropy": 5.397800159454346,
      "epoch": 2.8532599539506935,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041992703186799273,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21742852926254272,
      "num_tokens": 56707372.0,
      "step": 25405
    },
    {
      "entropy": 5.3473118305206295,
      "epoch": 2.853821530858651,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041989631882856447,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21552158445119857,
      "num_tokens": 56717954.0,
      "step": 25410
    },
    {
      "entropy": 5.410594367980957,
      "epoch": 2.8543831077666084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20644324719905854,
      "num_tokens": 56728798.0,
      "step": 25415
    },
    {
      "entropy": 5.4540966033935545,
      "epoch": 2.854944684674566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004198348789100211,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20912293642759322,
      "num_tokens": 56740095.0,
      "step": 25420
    },
    {
      "entropy": 5.395148229598999,
      "epoch": 2.8555062615825237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041980415203286223,
      "loss": 5.0195,
      "mean_token_accuracy": 0.20521110892295838,
      "num_tokens": 56751002.0,
      "step": 25425
    },
    {
      "entropy": 5.3969789981842045,
      "epoch": 2.8560678384904814,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000419773420545083,
      "loss": 5.0416,
      "mean_token_accuracy": 0.2122397467494011,
      "num_tokens": 56762244.0,
      "step": 25430
    },
    {
      "entropy": 5.460066938400269,
      "epoch": 2.856629415398439,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.12,
      "mean_token_accuracy": 0.20771208256483079,
      "num_tokens": 56772918.0,
      "step": 25435
    },
    {
      "entropy": 5.319263315200805,
      "epoch": 2.8571909923063963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004197119437415776,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21692632585763932,
      "num_tokens": 56784069.0,
      "step": 25440
    },
    {
      "entropy": 5.359051370620728,
      "epoch": 2.857752569214354,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004196811984278089,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21318614780902861,
      "num_tokens": 56795187.0,
      "step": 25445
    },
    {
      "entropy": 5.422108364105225,
      "epoch": 2.8583141461223116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004196504485073347,
      "loss": 5.0694,
      "mean_token_accuracy": 0.20756106972694396,
      "num_tokens": 56806766.0,
      "step": 25450
    },
    {
      "entropy": 5.464345264434814,
      "epoch": 2.858875723030269,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20768696963787078,
      "num_tokens": 56817872.0,
      "step": 25455
    },
    {
      "entropy": 5.484509658813477,
      "epoch": 2.8594372999382265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.1736,
      "mean_token_accuracy": 0.20331741273403167,
      "num_tokens": 56828291.0,
      "step": 25460
    },
    {
      "entropy": 5.418423366546631,
      "epoch": 2.859998876846184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004195581711154709,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21263631135225297,
      "num_tokens": 56839165.0,
      "step": 25465
    },
    {
      "entropy": 5.387149667739868,
      "epoch": 2.8605604537541414,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041952740277796717,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22038833498954774,
      "num_tokens": 56850754.0,
      "step": 25470
    },
    {
      "entropy": 5.342712926864624,
      "epoch": 2.861122030662099,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.0922,
      "mean_token_accuracy": 0.21170541644096375,
      "num_tokens": 56862798.0,
      "step": 25475
    },
    {
      "entropy": 5.340575695037842,
      "epoch": 2.8616836075700567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004194658522985139,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2168458640575409,
      "num_tokens": 56873479.0,
      "step": 25480
    },
    {
      "entropy": 5.447553586959839,
      "epoch": 2.8622451844780143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004194350701585241,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20874096006155013,
      "num_tokens": 56883924.0,
      "step": 25485
    },
    {
      "entropy": 5.44626932144165,
      "epoch": 2.862806761385972,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.0791,
      "mean_token_accuracy": 0.21461231857538224,
      "num_tokens": 56895622.0,
      "step": 25490
    },
    {
      "entropy": 5.403787899017334,
      "epoch": 2.8633683382939292,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004193734920829188,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20768801122903824,
      "num_tokens": 56906504.0,
      "step": 25495
    },
    {
      "entropy": 5.343988466262817,
      "epoch": 2.863929915201887,
      "grad_norm": 1.0625,
      "learning_rate": 0.000419342696149264,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21133945733308793,
      "num_tokens": 56917228.0,
      "step": 25500
    },
    {
      "entropy": 5.435186862945557,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004193118956196818,
      "loss": 5.062,
      "mean_token_accuracy": 0.2064625769853592,
      "num_tokens": 56927981.0,
      "step": 25505
    },
    {
      "entropy": 5.385399436950683,
      "epoch": 2.8650530690178018,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004192810904951528,
      "loss": 5.0679,
      "mean_token_accuracy": 0.20595361590385436,
      "num_tokens": 56938552.0,
      "step": 25510
    },
    {
      "entropy": 5.385594987869263,
      "epoch": 2.8656146459257594,
      "grad_norm": 1.03125,
      "learning_rate": 0.000419250280776658,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21326264441013337,
      "num_tokens": 56949902.0,
      "step": 25515
    },
    {
      "entropy": 5.48624701499939,
      "epoch": 2.866176222833717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.1738,
      "mean_token_accuracy": 0.21045093089342118,
      "num_tokens": 56961335.0,
      "step": 25520
    },
    {
      "entropy": 5.387441778182984,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004191886475616947,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2172359749674797,
      "num_tokens": 56972542.0,
      "step": 25525
    },
    {
      "entropy": 5.408881664276123,
      "epoch": 2.8672993766496324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041915782406718875,
      "loss": 5.0993,
      "mean_token_accuracy": 0.21220753192901612,
      "num_tokens": 56983838.0,
      "step": 25530
    },
    {
      "entropy": 5.42573037147522,
      "epoch": 2.8678609535575896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004191269959826418,
      "loss": 5.0744,
      "mean_token_accuracy": 0.21078961044549943,
      "num_tokens": 56996125.0,
      "step": 25535
    },
    {
      "entropy": 5.447035312652588,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041909616330903544,
      "loss": 5.0774,
      "mean_token_accuracy": 0.2090054988861084,
      "num_tokens": 57007599.0,
      "step": 25540
    },
    {
      "entropy": 5.419825506210327,
      "epoch": 2.868984107373505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041906532604735127,
      "loss": 5.0716,
      "mean_token_accuracy": 0.2108407884836197,
      "num_tokens": 57018102.0,
      "step": 25545
    },
    {
      "entropy": 5.419286155700684,
      "epoch": 2.869545684281462,
      "grad_norm": 1.125,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20637976229190827,
      "num_tokens": 57029253.0,
      "step": 25550
    },
    {
      "entropy": 5.495482158660889,
      "epoch": 2.87010726118942,
      "grad_norm": 1.125,
      "learning_rate": 0.00041900363776367747,
      "loss": 5.1283,
      "mean_token_accuracy": 0.2021507054567337,
      "num_tokens": 57040485.0,
      "step": 25555
    },
    {
      "entropy": 5.379592752456665,
      "epoch": 2.8706688380973775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004189727867436519,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2202585220336914,
      "num_tokens": 57051340.0,
      "step": 25560
    },
    {
      "entropy": 5.3074281215667725,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041894193113947703,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2086580976843834,
      "num_tokens": 57061726.0,
      "step": 25565
    },
    {
      "entropy": 5.407056093215942,
      "epoch": 2.8717919919132924,
      "grad_norm": 1.125,
      "learning_rate": 0.00041891107095213523,
      "loss": 5.0824,
      "mean_token_accuracy": 0.206282339990139,
      "num_tokens": 57073367.0,
      "step": 25570
    },
    {
      "entropy": 5.495729398727417,
      "epoch": 2.87235356882125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.1971,
      "mean_token_accuracy": 0.21119094789028167,
      "num_tokens": 57085379.0,
      "step": 25575
    },
    {
      "entropy": 5.500995874404907,
      "epoch": 2.8729151457292077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004188493368318814,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20952188372612,
      "num_tokens": 57096151.0,
      "step": 25580
    },
    {
      "entropy": 5.441212749481201,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000418818462900935,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20424191355705262,
      "num_tokens": 57107378.0,
      "step": 25585
    },
    {
      "entropy": 5.410331535339355,
      "epoch": 2.8740382995451226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20845856964588166,
      "num_tokens": 57119814.0,
      "step": 25590
    },
    {
      "entropy": 5.595893287658692,
      "epoch": 2.8745998764530802,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20636238008737565,
      "num_tokens": 57132070.0,
      "step": 25595
    },
    {
      "entropy": 5.543785810470581,
      "epoch": 2.875161453361038,
      "grad_norm": 1.140625,
      "learning_rate": 0.000418725813636615,
      "loss": 5.1842,
      "mean_token_accuracy": 0.2007378250360489,
      "num_tokens": 57143216.0,
      "step": 25600
    },
    {
      "entropy": 5.356467580795288,
      "epoch": 2.875723030268995,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.0868,
      "mean_token_accuracy": 0.2117350682616234,
      "num_tokens": 57154485.0,
      "step": 25605
    },
    {
      "entropy": 5.437893724441528,
      "epoch": 2.876284607176953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20444340407848358,
      "num_tokens": 57164738.0,
      "step": 25610
    },
    {
      "entropy": 5.494603824615479,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.1233,
      "mean_token_accuracy": 0.2047460690140724,
      "num_tokens": 57175790.0,
      "step": 25615
    },
    {
      "entropy": 5.417849922180176,
      "epoch": 2.877407760992868,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.1785,
      "mean_token_accuracy": 0.20332439839839936,
      "num_tokens": 57187201.0,
      "step": 25620
    },
    {
      "entropy": 5.4206883907318115,
      "epoch": 2.8779693379008258,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041857130668348705,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21272264420986176,
      "num_tokens": 57197591.0,
      "step": 25625
    },
    {
      "entropy": 5.512492227554321,
      "epoch": 2.878530914808783,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2098799765110016,
      "num_tokens": 57210194.0,
      "step": 25630
    },
    {
      "entropy": 5.484561395645142,
      "epoch": 2.8790924917167406,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004185094718957848,
      "loss": 5.0911,
      "mean_token_accuracy": 0.2030409097671509,
      "num_tokens": 57220752.0,
      "step": 25635
    },
    {
      "entropy": 5.430163383483887,
      "epoch": 2.8796540686246983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041847854764734556,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21704408526420593,
      "num_tokens": 57233109.0,
      "step": 25640
    },
    {
      "entropy": 5.395926141738892,
      "epoch": 2.8802156455326555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.1685,
      "mean_token_accuracy": 0.20503151416778564,
      "num_tokens": 57244548.0,
      "step": 25645
    },
    {
      "entropy": 5.419839525222779,
      "epoch": 2.880777222440613,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041841668544621396,
      "loss": 5.0608,
      "mean_token_accuracy": 0.21475321501493455,
      "num_tokens": 57255328.0,
      "step": 25650
    },
    {
      "entropy": 5.441213464736938,
      "epoch": 2.881338799348571,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004183857474954915,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20347959846258162,
      "num_tokens": 57268890.0,
      "step": 25655
    },
    {
      "entropy": 5.497204351425171,
      "epoch": 2.881900376256528,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20581125617027282,
      "num_tokens": 57278505.0,
      "step": 25660
    },
    {
      "entropy": 5.512013578414917,
      "epoch": 2.8824619531644857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20444290488958358,
      "num_tokens": 57289499.0,
      "step": 25665
    },
    {
      "entropy": 5.567802619934082,
      "epoch": 2.8830235300724434,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.2786,
      "mean_token_accuracy": 0.20638787150382995,
      "num_tokens": 57300970.0,
      "step": 25670
    },
    {
      "entropy": 5.423797607421875,
      "epoch": 2.883585106980401,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041826195004787616,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22225052118301392,
      "num_tokens": 57311997.0,
      "step": 25675
    },
    {
      "entropy": 5.460712051391601,
      "epoch": 2.8841466838883587,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.1247,
      "mean_token_accuracy": 0.20363961011171341,
      "num_tokens": 57323385.0,
      "step": 25680
    },
    {
      "entropy": 5.416005277633667,
      "epoch": 2.884708260796316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004182000239510301,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20132754892110824,
      "num_tokens": 57335389.0,
      "step": 25685
    },
    {
      "entropy": 5.489405488967895,
      "epoch": 2.8852698377042736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20470672249794006,
      "num_tokens": 57346843.0,
      "step": 25690
    },
    {
      "entropy": 5.478165102005005,
      "epoch": 2.8858314146122312,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.142,
      "mean_token_accuracy": 0.20753378719091414,
      "num_tokens": 57358253.0,
      "step": 25695
    },
    {
      "entropy": 5.485944128036499,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004181071006116442,
      "loss": 5.164,
      "mean_token_accuracy": 0.20432851910591127,
      "num_tokens": 57369204.0,
      "step": 25700
    },
    {
      "entropy": 5.408002853393555,
      "epoch": 2.886954568428146,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041807611705069593,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21851354092359543,
      "num_tokens": 57380214.0,
      "step": 25705
    },
    {
      "entropy": 5.431133794784546,
      "epoch": 2.887516145336104,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041804512893414865,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20473915487527847,
      "num_tokens": 57391848.0,
      "step": 25710
    },
    {
      "entropy": 5.532349014282227,
      "epoch": 2.8880777222440615,
      "grad_norm": 1.125,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.1706,
      "mean_token_accuracy": 0.206267249584198,
      "num_tokens": 57403026.0,
      "step": 25715
    },
    {
      "entropy": 5.409966087341308,
      "epoch": 2.888639299152019,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20506458580493928,
      "num_tokens": 57414223.0,
      "step": 25720
    },
    {
      "entropy": 5.393777656555176,
      "epoch": 2.8892008760599763,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004179521372607798,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21456553936004638,
      "num_tokens": 57424661.0,
      "step": 25725
    },
    {
      "entropy": 5.391867971420288,
      "epoch": 2.889762452967934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041792113093170447,
      "loss": 5.0374,
      "mean_token_accuracy": 0.20940565764904023,
      "num_tokens": 57435859.0,
      "step": 25730
    },
    {
      "entropy": 5.402270555496216,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2069082349538803,
      "num_tokens": 57447784.0,
      "step": 25735
    },
    {
      "entropy": 5.426554822921753,
      "epoch": 2.890885606783849,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004178591046225485,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20467803180217742,
      "num_tokens": 57458881.0,
      "step": 25740
    },
    {
      "entropy": 5.403743743896484,
      "epoch": 2.8914471836918065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.0851,
      "mean_token_accuracy": 0.2060352772474289,
      "num_tokens": 57470287.0,
      "step": 25745
    },
    {
      "entropy": 5.484283113479615,
      "epoch": 2.892008760599764,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20617766082286834,
      "num_tokens": 57481723.0,
      "step": 25750
    },
    {
      "entropy": 5.504418802261353,
      "epoch": 2.8925703375077214,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041776603104611427,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20658947378396988,
      "num_tokens": 57492804.0,
      "step": 25755
    },
    {
      "entropy": 5.367903804779052,
      "epoch": 2.893131914415679,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041773499742786736,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21405443847179412,
      "num_tokens": 57504148.0,
      "step": 25760
    },
    {
      "entropy": 5.428642416000367,
      "epoch": 2.8936934913236367,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.1176,
      "mean_token_accuracy": 0.21061230599880218,
      "num_tokens": 57516026.0,
      "step": 25765
    },
    {
      "entropy": 5.452574110031128,
      "epoch": 2.8942550682315944,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.1103,
      "mean_token_accuracy": 0.19943454414606093,
      "num_tokens": 57526882.0,
      "step": 25770
    },
    {
      "entropy": 5.382192707061767,
      "epoch": 2.894816645139552,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041764186930865377,
      "loss": 5.1058,
      "mean_token_accuracy": 0.21129528135061265,
      "num_tokens": 57537875.0,
      "step": 25775
    },
    {
      "entropy": 5.440432739257813,
      "epoch": 2.8953782220475093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004176108175173863,
      "loss": 5.0556,
      "mean_token_accuracy": 0.2105400636792183,
      "num_tokens": 57548143.0,
      "step": 25780
    },
    {
      "entropy": 5.4083963394165036,
      "epoch": 2.895939798955467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004175797611853351,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21154009848833083,
      "num_tokens": 57558127.0,
      "step": 25785
    },
    {
      "entropy": 5.457000303268432,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.015625,
      "learning_rate": 0.000417548700313489,
      "loss": 5.1395,
      "mean_token_accuracy": 0.21226473301649093,
      "num_tokens": 57570028.0,
      "step": 25790
    },
    {
      "entropy": 5.405595397949218,
      "epoch": 2.897062952771382,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417517634902837,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2121260717511177,
      "num_tokens": 57581089.0,
      "step": 25795
    },
    {
      "entropy": 5.512363624572754,
      "epoch": 2.8976245296793395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.1912,
      "mean_token_accuracy": 0.20525837689638138,
      "num_tokens": 57592555.0,
      "step": 25800
    },
    {
      "entropy": 5.42119870185852,
      "epoch": 2.898186106587297,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041745549046907205,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20505211800336837,
      "num_tokens": 57603492.0,
      "step": 25805
    },
    {
      "entropy": 5.400435829162598,
      "epoch": 2.898747683495255,
      "grad_norm": 1.125,
      "learning_rate": 0.00041742441144793774,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2095305800437927,
      "num_tokens": 57613477.0,
      "step": 25810
    },
    {
      "entropy": 5.4144776344299315,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004173933278919549,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20417742133140565,
      "num_tokens": 57623703.0,
      "step": 25815
    },
    {
      "entropy": 5.4997282981872555,
      "epoch": 2.8998708373111697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004173622398021133,
      "loss": 5.0787,
      "mean_token_accuracy": 0.2118339419364929,
      "num_tokens": 57637172.0,
      "step": 25820
    },
    {
      "entropy": 5.474332618713379,
      "epoch": 2.9004324142191273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041733114717940273,
      "loss": 5.0997,
      "mean_token_accuracy": 0.2053397923707962,
      "num_tokens": 57648224.0,
      "step": 25825
    },
    {
      "entropy": 5.438064002990723,
      "epoch": 2.900993991127085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004173000500248133,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21917988061904908,
      "num_tokens": 57659093.0,
      "step": 25830
    },
    {
      "entropy": 5.471279144287109,
      "epoch": 2.901555568035042,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.1872,
      "mean_token_accuracy": 0.2076654925942421,
      "num_tokens": 57671399.0,
      "step": 25835
    },
    {
      "entropy": 5.528289175033569,
      "epoch": 2.902117144943,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.1032,
      "mean_token_accuracy": 0.21161250472068788,
      "num_tokens": 57682150.0,
      "step": 25840
    },
    {
      "entropy": 5.437770032882691,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041720673137967356,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20806330740451812,
      "num_tokens": 57693232.0,
      "step": 25845
    },
    {
      "entropy": 5.356527853012085,
      "epoch": 2.9032402987589148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041717561610747124,
      "loss": 5.0513,
      "mean_token_accuracy": 0.20941588282585144,
      "num_tokens": 57703422.0,
      "step": 25850
    },
    {
      "entropy": 5.421836090087891,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.125,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2063567891716957,
      "num_tokens": 57714454.0,
      "step": 25855
    },
    {
      "entropy": 5.38168740272522,
      "epoch": 2.90436345257483,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004171133719832774,
      "loss": 4.9475,
      "mean_token_accuracy": 0.22147613614797593,
      "num_tokens": 57726243.0,
      "step": 25860
    },
    {
      "entropy": 5.360598802566528,
      "epoch": 2.9049250294827877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041708224313326747,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2132922500371933,
      "num_tokens": 57737882.0,
      "step": 25865
    },
    {
      "entropy": 5.486487960815429,
      "epoch": 2.9054866063907454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000417051109759304,
      "loss": 5.129,
      "mean_token_accuracy": 0.20622992217540742,
      "num_tokens": 57750205.0,
      "step": 25870
    },
    {
      "entropy": 5.4093146324157715,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.125,
      "learning_rate": 0.000417019971862378,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21517722308635712,
      "num_tokens": 57761762.0,
      "step": 25875
    },
    {
      "entropy": 5.4142461776733395,
      "epoch": 2.9066097602066603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.081,
      "mean_token_accuracy": 0.20336637496948243,
      "num_tokens": 57771280.0,
      "step": 25880
    },
    {
      "entropy": 5.423370218276977,
      "epoch": 2.907171337114618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041695768250360466,
      "loss": 5.076,
      "mean_token_accuracy": 0.20784443169832229,
      "num_tokens": 57784099.0,
      "step": 25885
    },
    {
      "entropy": 5.421542119979859,
      "epoch": 2.907732914022575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20408418476581575,
      "num_tokens": 57795544.0,
      "step": 25890
    },
    {
      "entropy": 5.408284521102905,
      "epoch": 2.908294490930533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004168953750648809,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21849903017282485,
      "num_tokens": 57805963.0,
      "step": 25895
    },
    {
      "entropy": 5.484272193908692,
      "epoch": 2.9088560678384905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.146,
      "mean_token_accuracy": 0.20299637466669082,
      "num_tokens": 57815281.0,
      "step": 25900
    },
    {
      "entropy": 5.46207218170166,
      "epoch": 2.909417644746448,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20505156815052034,
      "num_tokens": 57826172.0,
      "step": 25905
    },
    {
      "entropy": 5.459085464477539,
      "epoch": 2.909979221654406,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041680188002424755,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21026309132575988,
      "num_tokens": 57836617.0,
      "step": 25910
    },
    {
      "entropy": 5.437853527069092,
      "epoch": 2.910540798562363,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004167707059793262,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2124670445919037,
      "num_tokens": 57847636.0,
      "step": 25915
    },
    {
      "entropy": 5.52723388671875,
      "epoch": 2.9111023754703207,
      "grad_norm": 1.125,
      "learning_rate": 0.00041673952742037057,
      "loss": 5.1809,
      "mean_token_accuracy": 0.20481421947479247,
      "num_tokens": 57858816.0,
      "step": 25920
    },
    {
      "entropy": 5.456689691543579,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.1153,
      "mean_token_accuracy": 0.2059694141149521,
      "num_tokens": 57870749.0,
      "step": 25925
    },
    {
      "entropy": 5.391223192214966,
      "epoch": 2.9122255292862356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.14,
      "mean_token_accuracy": 0.20486689507961273,
      "num_tokens": 57883454.0,
      "step": 25930
    },
    {
      "entropy": 5.420502853393555,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004166459646692262,
      "loss": 5.0618,
      "mean_token_accuracy": 0.20838523656129837,
      "num_tokens": 57894498.0,
      "step": 25935
    },
    {
      "entropy": 5.355986642837524,
      "epoch": 2.913348683102151,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004166147680640623,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21845532357692718,
      "num_tokens": 57905332.0,
      "step": 25940
    },
    {
      "entropy": 5.356674242019653,
      "epoch": 2.913910260010108,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041658356694982936,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2169194445014,
      "num_tokens": 57916456.0,
      "step": 25945
    },
    {
      "entropy": 5.413755702972412,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.1063,
      "mean_token_accuracy": 0.21528584957122804,
      "num_tokens": 57927615.0,
      "step": 25950
    },
    {
      "entropy": 5.355228042602539,
      "epoch": 2.9150334138260234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004165211511981299,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21651007682085038,
      "num_tokens": 57937771.0,
      "step": 25955
    },
    {
      "entropy": 5.401184511184693,
      "epoch": 2.915594990733981,
      "grad_norm": 1.0,
      "learning_rate": 0.0004164899365626507,
      "loss": 4.9547,
      "mean_token_accuracy": 0.20760841816663742,
      "num_tokens": 57948564.0,
      "step": 25960
    },
    {
      "entropy": 5.372656536102295,
      "epoch": 2.9161565676419388,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004164587174220771,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2127408728003502,
      "num_tokens": 57959758.0,
      "step": 25965
    },
    {
      "entropy": 5.456940031051635,
      "epoch": 2.916718144549896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2080602988600731,
      "num_tokens": 57970988.0,
      "step": 25970
    },
    {
      "entropy": 5.353135108947754,
      "epoch": 2.9172797214578536,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041639626562962293,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22431057542562485,
      "num_tokens": 57981056.0,
      "step": 25975
    },
    {
      "entropy": 5.360879707336426,
      "epoch": 2.9178412983658113,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.0808,
      "mean_token_accuracy": 0.2140895962715149,
      "num_tokens": 57992092.0,
      "step": 25980
    },
    {
      "entropy": 5.470174074172974,
      "epoch": 2.9184028752737685,
      "grad_norm": 1.078125,
      "learning_rate": 0.000416333795828721,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21645472347736358,
      "num_tokens": 58003000.0,
      "step": 25985
    },
    {
      "entropy": 5.386347341537475,
      "epoch": 2.918964452181726,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004163025541775885,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21076276004314423,
      "num_tokens": 58013932.0,
      "step": 25990
    },
    {
      "entropy": 5.479140329360962,
      "epoch": 2.919526029089684,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.1306,
      "mean_token_accuracy": 0.20881747752428054,
      "num_tokens": 58024481.0,
      "step": 25995
    },
    {
      "entropy": 5.491335678100586,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.125,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20946117788553237,
      "num_tokens": 58035753.0,
      "step": 26000
    },
    {
      "entropy": 5.565003108978272,
      "epoch": 2.920649182905599,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.2461,
      "mean_token_accuracy": 0.2038903772830963,
      "num_tokens": 58046599.0,
      "step": 26005
    },
    {
      "entropy": 5.4701783657073975,
      "epoch": 2.9212107598135564,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041617754259172665,
      "loss": 5.1479,
      "mean_token_accuracy": 0.20732745081186293,
      "num_tokens": 58058151.0,
      "step": 26010
    },
    {
      "entropy": 5.492415857315064,
      "epoch": 2.921772336721514,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.1208,
      "mean_token_accuracy": 0.20721977353096008,
      "num_tokens": 58070116.0,
      "step": 26015
    },
    {
      "entropy": 5.478378963470459,
      "epoch": 2.9223339136294717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.0736,
      "mean_token_accuracy": 0.210087414085865,
      "num_tokens": 58081089.0,
      "step": 26020
    },
    {
      "entropy": 5.380297994613647,
      "epoch": 2.922895490537429,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041608373669979696,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21423850506544112,
      "num_tokens": 58092033.0,
      "step": 26025
    },
    {
      "entropy": 5.386036825180054,
      "epoch": 2.9234570674453866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004160524590835042,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21063886433839799,
      "num_tokens": 58102864.0,
      "step": 26030
    },
    {
      "entropy": 5.369965648651123,
      "epoch": 2.9240186443533442,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041602117697604664,
      "loss": 5.0616,
      "mean_token_accuracy": 0.20118890404701234,
      "num_tokens": 58114607.0,
      "step": 26035
    },
    {
      "entropy": 5.428013658523559,
      "epoch": 2.9245802212613015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041598989037842,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20538741499185562,
      "num_tokens": 58125935.0,
      "step": 26040
    },
    {
      "entropy": 5.408543825149536,
      "epoch": 2.925141798169259,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004159585992916206,
      "loss": 5.046,
      "mean_token_accuracy": 0.21360983401536943,
      "num_tokens": 58136439.0,
      "step": 26045
    },
    {
      "entropy": 5.45725383758545,
      "epoch": 2.925703375077217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20408802181482316,
      "num_tokens": 58147778.0,
      "step": 26050
    },
    {
      "entropy": 5.491113805770874,
      "epoch": 2.9262649519851744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041589600365448884,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21362184435129167,
      "num_tokens": 58159290.0,
      "step": 26055
    },
    {
      "entropy": 5.471893358230591,
      "epoch": 2.926826528893132,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.1081,
      "mean_token_accuracy": 0.20282102674245833,
      "num_tokens": 58169917.0,
      "step": 26060
    },
    {
      "entropy": 5.393456220626831,
      "epoch": 2.9273881058010893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004158333900726235,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2135949969291687,
      "num_tokens": 58180905.0,
      "step": 26065
    },
    {
      "entropy": 5.404960012435913,
      "epoch": 2.927949682709047,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.0687,
      "mean_token_accuracy": 0.2131287693977356,
      "num_tokens": 58192312.0,
      "step": 26070
    },
    {
      "entropy": 5.397861099243164,
      "epoch": 2.9285112596170046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004157707585539994,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2159481331706047,
      "num_tokens": 58204739.0,
      "step": 26075
    },
    {
      "entropy": 5.417920255661011,
      "epoch": 2.929072836524962,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041573943607089536,
      "loss": 5.0456,
      "mean_token_accuracy": 0.211036516726017,
      "num_tokens": 58215663.0,
      "step": 26080
    },
    {
      "entropy": 5.41449236869812,
      "epoch": 2.9296344134329195,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004157081091065932,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20894305408000946,
      "num_tokens": 58226480.0,
      "step": 26085
    },
    {
      "entropy": 5.599746036529541,
      "epoch": 2.930195990340877,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.3423,
      "mean_token_accuracy": 0.19856517016887665,
      "num_tokens": 58238087.0,
      "step": 26090
    },
    {
      "entropy": 5.475650930404663,
      "epoch": 2.930757567248835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000415645441738384,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20367458313703538,
      "num_tokens": 58248738.0,
      "step": 26095
    },
    {
      "entropy": 5.38551959991455,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.46875,
      "learning_rate": 0.00041561410133647244,
      "loss": 5.0883,
      "mean_token_accuracy": 0.20835185497999192,
      "num_tokens": 58260440.0,
      "step": 26100
    },
    {
      "entropy": 5.4868165969848635,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20485747009515762,
      "num_tokens": 58272704.0,
      "step": 26105
    },
    {
      "entropy": 5.460170555114746,
      "epoch": 2.9324422979727074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041555140710202484,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20732046216726302,
      "num_tokens": 58284254.0,
      "step": 26110
    },
    {
      "entropy": 5.459529685974121,
      "epoch": 2.933003874880665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20705208480358123,
      "num_tokens": 58296472.0,
      "step": 26115
    },
    {
      "entropy": 5.387962532043457,
      "epoch": 2.9335654517886223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004154886949667321,
      "loss": 5.028,
      "mean_token_accuracy": 0.21463914811611176,
      "num_tokens": 58307734.0,
      "step": 26120
    },
    {
      "entropy": 5.420158720016479,
      "epoch": 2.93412702869658,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041545733218876466,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2137228414416313,
      "num_tokens": 58318329.0,
      "step": 26125
    },
    {
      "entropy": 5.357426977157592,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004154259649385813,
      "loss": 5.0373,
      "mean_token_accuracy": 0.2103509858250618,
      "num_tokens": 58330036.0,
      "step": 26130
    },
    {
      "entropy": 5.403976535797119,
      "epoch": 2.935250182512495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041539459321718076,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21634389758110045,
      "num_tokens": 58340303.0,
      "step": 26135
    },
    {
      "entropy": 5.366495084762573,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004153632170255619,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21734954416751862,
      "num_tokens": 58352394.0,
      "step": 26140
    },
    {
      "entropy": 5.457917976379394,
      "epoch": 2.93637333632841,
      "grad_norm": 2.171875,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.2049,
      "mean_token_accuracy": 0.20125900954008102,
      "num_tokens": 58364605.0,
      "step": 26145
    },
    {
      "entropy": 5.4505202293396,
      "epoch": 2.936934913236368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004153004512356656,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21446591913700103,
      "num_tokens": 58375514.0,
      "step": 26150
    },
    {
      "entropy": 5.4604551792144775,
      "epoch": 2.9374964901443255,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004152690616393866,
      "loss": 5.0806,
      "mean_token_accuracy": 0.21050941199064255,
      "num_tokens": 58387135.0,
      "step": 26155
    },
    {
      "entropy": 5.405916595458985,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041523766757688634,
      "loss": 5.0753,
      "mean_token_accuracy": 0.20663321763277054,
      "num_tokens": 58399077.0,
      "step": 26160
    },
    {
      "entropy": 5.5281898021698,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.2054,
      "mean_token_accuracy": 0.21059574484825133,
      "num_tokens": 58409897.0,
      "step": 26165
    },
    {
      "entropy": 5.3822447776794435,
      "epoch": 2.939181220868198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004151748660572203,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2198258548974991,
      "num_tokens": 58421189.0,
      "step": 26170
    },
    {
      "entropy": 5.354880142211914,
      "epoch": 2.939742797776155,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004151434586020541,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20622098445892334,
      "num_tokens": 58432258.0,
      "step": 26175
    },
    {
      "entropy": 5.462463474273681,
      "epoch": 2.940304374684113,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.1372,
      "mean_token_accuracy": 0.21158119142055512,
      "num_tokens": 58444288.0,
      "step": 26180
    },
    {
      "entropy": 5.495795869827271,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041508063030605585,
      "loss": 5.0668,
      "mean_token_accuracy": 0.21693544685840607,
      "num_tokens": 58453367.0,
      "step": 26185
    },
    {
      "entropy": 5.478635978698731,
      "epoch": 2.941427528500028,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041504920946722413,
      "loss": 5.2186,
      "mean_token_accuracy": 0.20242106169462204,
      "num_tokens": 58464862.0,
      "step": 26190
    },
    {
      "entropy": 5.424022817611695,
      "epoch": 2.941989105407986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.0844,
      "mean_token_accuracy": 0.22019710540771484,
      "num_tokens": 58476353.0,
      "step": 26195
    },
    {
      "entropy": 5.426383399963379,
      "epoch": 2.942550682315943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20733167827129365,
      "num_tokens": 58488330.0,
      "step": 26200
    },
    {
      "entropy": 5.444040727615357,
      "epoch": 2.9431122592239007,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041495492019940467,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21406294107437135,
      "num_tokens": 58500004.0,
      "step": 26205
    },
    {
      "entropy": 5.4153424263000485,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004149234815296925,
      "loss": 5.0695,
      "mean_token_accuracy": 0.2041459634900093,
      "num_tokens": 58511433.0,
      "step": 26210
    },
    {
      "entropy": 5.3686676025390625,
      "epoch": 2.9442354130398156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041489203840476243,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21089380383491516,
      "num_tokens": 58522234.0,
      "step": 26215
    },
    {
      "entropy": 5.476142120361328,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20723894834518433,
      "num_tokens": 58534370.0,
      "step": 26220
    },
    {
      "entropy": 5.504309511184692,
      "epoch": 2.945358566855731,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004148291387932533,
      "loss": 5.0583,
      "mean_token_accuracy": 0.2119843378663063,
      "num_tokens": 58545859.0,
      "step": 26225
    },
    {
      "entropy": 5.356972074508667,
      "epoch": 2.9459201437636886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041479768230867696,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2225022241473198,
      "num_tokens": 58557159.0,
      "step": 26230
    },
    {
      "entropy": 5.342748689651489,
      "epoch": 2.946481720671646,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041476622137288823,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20449849367141723,
      "num_tokens": 58568985.0,
      "step": 26235
    },
    {
      "entropy": 5.447239923477173,
      "epoch": 2.9470432975796035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041473475598688886,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20970015823841096,
      "num_tokens": 58579897.0,
      "step": 26240
    },
    {
      "entropy": 5.402147912979126,
      "epoch": 2.947604874487561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004147032861516805,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21835739463567733,
      "num_tokens": 58590042.0,
      "step": 26245
    },
    {
      "entropy": 5.414743947982788,
      "epoch": 2.948166451395519,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.1328,
      "mean_token_accuracy": 0.21502893716096877,
      "num_tokens": 58601645.0,
      "step": 26250
    },
    {
      "entropy": 5.4197252750396725,
      "epoch": 2.948728028303476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.1369,
      "mean_token_accuracy": 0.21094148606061935,
      "num_tokens": 58613881.0,
      "step": 26255
    },
    {
      "entropy": 5.40075569152832,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041460884996082363,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21373191773891448,
      "num_tokens": 58623430.0,
      "step": 26260
    },
    {
      "entropy": 5.4030622959136965,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041457736233880166,
      "loss": 5.0028,
      "mean_token_accuracy": 0.20568868964910508,
      "num_tokens": 58635289.0,
      "step": 26265
    },
    {
      "entropy": 5.333928394317627,
      "epoch": 2.9504127590273486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004145458702725823,
      "loss": 5.0619,
      "mean_token_accuracy": 0.20832496434450148,
      "num_tokens": 58646086.0,
      "step": 26270
    },
    {
      "entropy": 5.396041822433472,
      "epoch": 2.9509743359353062,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041451437376316825,
      "loss": 5.0439,
      "mean_token_accuracy": 0.20373013168573378,
      "num_tokens": 58657662.0,
      "step": 26275
    },
    {
      "entropy": 5.453104305267334,
      "epoch": 2.951535912843264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004144828728115624,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21530164778232574,
      "num_tokens": 58668054.0,
      "step": 26280
    },
    {
      "entropy": 5.49367938041687,
      "epoch": 2.9520974897512215,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.174,
      "mean_token_accuracy": 0.20659656524658204,
      "num_tokens": 58679285.0,
      "step": 26285
    },
    {
      "entropy": 5.465607452392578,
      "epoch": 2.952659066659179,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.1682,
      "mean_token_accuracy": 0.2140774205327034,
      "num_tokens": 58691878.0,
      "step": 26290
    },
    {
      "entropy": 5.305894947052002,
      "epoch": 2.9532206435671364,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004143883433136252,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21775573939085008,
      "num_tokens": 58702130.0,
      "step": 26295
    },
    {
      "entropy": 5.423879861831665,
      "epoch": 2.953782220475094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143568246032838,
      "loss": 5.0413,
      "mean_token_accuracy": 0.21920413374900818,
      "num_tokens": 58712872.0,
      "step": 26300
    },
    {
      "entropy": 5.430193710327148,
      "epoch": 2.9543437973830518,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.1289,
      "mean_token_accuracy": 0.20435429811477662,
      "num_tokens": 58724783.0,
      "step": 26305
    },
    {
      "entropy": 5.484325981140136,
      "epoch": 2.954905374291009,
      "grad_norm": 1.125,
      "learning_rate": 0.00041429377387207865,
      "loss": 5.1365,
      "mean_token_accuracy": 0.2040925070643425,
      "num_tokens": 58736473.0,
      "step": 26310
    },
    {
      "entropy": 5.484816408157348,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20189605206251143,
      "num_tokens": 58748223.0,
      "step": 26315
    },
    {
      "entropy": 5.336780595779419,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004142307054002025,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21798297762870789,
      "num_tokens": 58759876.0,
      "step": 26320
    },
    {
      "entropy": 5.438477325439453,
      "epoch": 2.956590105014882,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.1223,
      "mean_token_accuracy": 0.21275402009487152,
      "num_tokens": 58771288.0,
      "step": 26325
    },
    {
      "entropy": 5.500085306167603,
      "epoch": 2.957151681922839,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041416761919568784,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20584044903516768,
      "num_tokens": 58781664.0,
      "step": 26330
    },
    {
      "entropy": 5.443375968933106,
      "epoch": 2.957713258830797,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20681860893964768,
      "num_tokens": 58792808.0,
      "step": 26335
    },
    {
      "entropy": 5.3595293998718265,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004141045152665695,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21123780459165573,
      "num_tokens": 58804455.0,
      "step": 26340
    },
    {
      "entropy": 5.363130283355713,
      "epoch": 2.958836412646712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041407295665779533,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21596324294805527,
      "num_tokens": 58814718.0,
      "step": 26345
    },
    {
      "entropy": 5.449704885482788,
      "epoch": 2.9593979895546694,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.1243,
      "mean_token_accuracy": 0.20717872083187103,
      "num_tokens": 58825649.0,
      "step": 26350
    },
    {
      "entropy": 5.455759525299072,
      "epoch": 2.959959566462627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.1468,
      "mean_token_accuracy": 0.20453344881534577,
      "num_tokens": 58836529.0,
      "step": 26355
    },
    {
      "entropy": 5.42989068031311,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20624094307422638,
      "num_tokens": 58847879.0,
      "step": 26360
    },
    {
      "entropy": 5.498843955993652,
      "epoch": 2.961082720278542,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004139466779513805,
      "loss": 5.2321,
      "mean_token_accuracy": 0.20076417475938796,
      "num_tokens": 58859283.0,
      "step": 26365
    },
    {
      "entropy": 5.46402645111084,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20609248727560042,
      "num_tokens": 58870529.0,
      "step": 26370
    },
    {
      "entropy": 5.457849311828613,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041388351204945534,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21319345533847808,
      "num_tokens": 58880828.0,
      "step": 26375
    },
    {
      "entropy": 5.5263814449310305,
      "epoch": 2.962767451002415,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.1708,
      "mean_token_accuracy": 0.21277730613946916,
      "num_tokens": 58892121.0,
      "step": 26380
    },
    {
      "entropy": 5.42065110206604,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20636169463396073,
      "num_tokens": 58902768.0,
      "step": 26385
    },
    {
      "entropy": 5.47159252166748,
      "epoch": 2.9638906048183298,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.1231,
      "mean_token_accuracy": 0.2102830946445465,
      "num_tokens": 58913404.0,
      "step": 26390
    },
    {
      "entropy": 5.5348732471466064,
      "epoch": 2.9644521817262874,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20840201079845427,
      "num_tokens": 58925086.0,
      "step": 26395
    },
    {
      "entropy": 5.453754234313965,
      "epoch": 2.965013758634245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137255199254112,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21082539856433868,
      "num_tokens": 58935425.0,
      "step": 26400
    },
    {
      "entropy": 5.4408351421356205,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.125,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.1074,
      "mean_token_accuracy": 0.2046562522649765,
      "num_tokens": 58945528.0,
      "step": 26405
    },
    {
      "entropy": 5.424080228805542,
      "epoch": 2.96613691245016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041366229214922926,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21184166818857192,
      "num_tokens": 58957008.0,
      "step": 26410
    },
    {
      "entropy": 5.485074424743653,
      "epoch": 2.9666984893581176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004136306716380432,
      "loss": 5.1093,
      "mean_token_accuracy": 0.20259469151496887,
      "num_tokens": 58968648.0,
      "step": 26415
    },
    {
      "entropy": 5.448821306228638,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041359904671280285,
      "loss": 5.1001,
      "mean_token_accuracy": 0.20371476858854293,
      "num_tokens": 58979794.0,
      "step": 26420
    },
    {
      "entropy": 5.38902735710144,
      "epoch": 2.9678216431740325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004135674173745151,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21516756266355513,
      "num_tokens": 58990534.0,
      "step": 26425
    },
    {
      "entropy": 5.4008814811706545,
      "epoch": 2.96838322008199,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041353578362418695,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20997032076120375,
      "num_tokens": 59001717.0,
      "step": 26430
    },
    {
      "entropy": 5.3625281810760494,
      "epoch": 2.968944796989948,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004135041454628258,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21180500984191894,
      "num_tokens": 59012118.0,
      "step": 26435
    },
    {
      "entropy": 5.477153301239014,
      "epoch": 2.9695063738979055,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004134725028914389,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20583583563566207,
      "num_tokens": 59023940.0,
      "step": 26440
    },
    {
      "entropy": 5.498946237564087,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041344085591103383,
      "loss": 5.0397,
      "mean_token_accuracy": 0.212878055870533,
      "num_tokens": 59035147.0,
      "step": 26445
    },
    {
      "entropy": 5.351807117462158,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004134092045226182,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2194119319319725,
      "num_tokens": 59045531.0,
      "step": 26450
    },
    {
      "entropy": 5.434279918670654,
      "epoch": 2.971191104621778,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041337754872719984,
      "loss": 5.2037,
      "mean_token_accuracy": 0.2097873017191887,
      "num_tokens": 59056364.0,
      "step": 26455
    },
    {
      "entropy": 5.396720123291016,
      "epoch": 2.9717526815297353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004133458885257866,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2125670164823532,
      "num_tokens": 59066504.0,
      "step": 26460
    },
    {
      "entropy": 5.376542711257935,
      "epoch": 2.972314258437693,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041331422391938663,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21188508868217468,
      "num_tokens": 59077888.0,
      "step": 26465
    },
    {
      "entropy": 5.485977840423584,
      "epoch": 2.9728758353456506,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004132825549090082,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21588545739650727,
      "num_tokens": 59088222.0,
      "step": 26470
    },
    {
      "entropy": 5.490508031845093,
      "epoch": 2.9734374122536082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004132508814956595,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20614744871854782,
      "num_tokens": 59099302.0,
      "step": 26475
    },
    {
      "entropy": 5.426647663116455,
      "epoch": 2.973998989161566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004132192036803491,
      "loss": 5.0517,
      "mean_token_accuracy": 0.2099557787179947,
      "num_tokens": 59109238.0,
      "step": 26480
    },
    {
      "entropy": 5.357282400131226,
      "epoch": 2.974560566069523,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041318752146408575,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21598994880914688,
      "num_tokens": 59119795.0,
      "step": 26485
    },
    {
      "entropy": 5.378308439254761,
      "epoch": 2.975122142977481,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004131558348478781,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20368223190307616,
      "num_tokens": 59129649.0,
      "step": 26490
    },
    {
      "entropy": 5.4586601734161375,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.134,
      "mean_token_accuracy": 0.20293068438768386,
      "num_tokens": 59140802.0,
      "step": 26495
    },
    {
      "entropy": 5.448359584808349,
      "epoch": 2.9762452967933957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041309244841966565,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21158026903867722,
      "num_tokens": 59151333.0,
      "step": 26500
    },
    {
      "entropy": 5.460981464385986,
      "epoch": 2.9768068737013533,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.0731,
      "mean_token_accuracy": 0.2069780334830284,
      "num_tokens": 59163414.0,
      "step": 26505
    },
    {
      "entropy": 5.458379507064819,
      "epoch": 2.977368450609311,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041302904440378494,
      "loss": 5.1001,
      "mean_token_accuracy": 0.21037240624427794,
      "num_tokens": 59173884.0,
      "step": 26510
    },
    {
      "entropy": 5.481905460357666,
      "epoch": 2.9779300275172687,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.1231,
      "mean_token_accuracy": 0.20370368659496307,
      "num_tokens": 59185263.0,
      "step": 26515
    },
    {
      "entropy": 5.546029424667358,
      "epoch": 2.978491604425226,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041296562280831106,
      "loss": 5.2035,
      "mean_token_accuracy": 0.19763565063476562,
      "num_tokens": 59196279.0,
      "step": 26520
    },
    {
      "entropy": 5.345974063873291,
      "epoch": 2.9790531813331835,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041293390542075095,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22098766416311263,
      "num_tokens": 59207537.0,
      "step": 26525
    },
    {
      "entropy": 5.329504728317261,
      "epoch": 2.979614758241141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004129021836413217,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21289030611515045,
      "num_tokens": 59218583.0,
      "step": 26530
    },
    {
      "entropy": 5.373044729232788,
      "epoch": 2.980176335149099,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004128704574710334,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2138034999370575,
      "num_tokens": 59228638.0,
      "step": 26535
    },
    {
      "entropy": 5.540080118179321,
      "epoch": 2.980737912057056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.1969,
      "mean_token_accuracy": 0.2085081711411476,
      "num_tokens": 59240574.0,
      "step": 26540
    },
    {
      "entropy": 5.425041007995605,
      "epoch": 2.9812994889650137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041280699196192074,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21467933803796768,
      "num_tokens": 59252986.0,
      "step": 26545
    },
    {
      "entropy": 5.390027379989624,
      "epoch": 2.9818610658729714,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.1224,
      "mean_token_accuracy": 0.2020873710513115,
      "num_tokens": 59263034.0,
      "step": 26550
    },
    {
      "entropy": 5.464475154876709,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20463671386241913,
      "num_tokens": 59275170.0,
      "step": 26555
    },
    {
      "entropy": 5.480004167556762,
      "epoch": 2.9829842196888863,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000412711760792068,
      "loss": 5.1256,
      "mean_token_accuracy": 0.21040759682655336,
      "num_tokens": 59287168.0,
      "step": 26560
    },
    {
      "entropy": 5.48776650428772,
      "epoch": 2.983545796596844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.1534,
      "mean_token_accuracy": 0.2117634266614914,
      "num_tokens": 59298043.0,
      "step": 26565
    },
    {
      "entropy": 5.373422336578369,
      "epoch": 2.9841073735048016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041264825141983547,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21255938857793807,
      "num_tokens": 59310454.0,
      "step": 26570
    },
    {
      "entropy": 5.5297438621521,
      "epoch": 2.9846689504127593,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20575196146965027,
      "num_tokens": 59323275.0,
      "step": 26575
    },
    {
      "entropy": 5.385955142974853,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004125847245165084,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21084896922111512,
      "num_tokens": 59334003.0,
      "step": 26580
    },
    {
      "entropy": 5.367092847824097,
      "epoch": 2.985792104228674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004125529544932126,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2117108955979347,
      "num_tokens": 59345403.0,
      "step": 26585
    },
    {
      "entropy": 5.408843088150024,
      "epoch": 2.986353681136632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.1063,
      "mean_token_accuracy": 0.204322712123394,
      "num_tokens": 59357457.0,
      "step": 26590
    },
    {
      "entropy": 5.408827590942383,
      "epoch": 2.986915258044589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004124894013084145,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22160053700208665,
      "num_tokens": 59369059.0,
      "step": 26595
    },
    {
      "entropy": 5.424338340759277,
      "epoch": 2.9874768349525467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041245761814893576,
      "loss": 5.0399,
      "mean_token_accuracy": 0.2124126598238945,
      "num_tokens": 59380700.0,
      "step": 26600
    },
    {
      "entropy": 5.394623136520385,
      "epoch": 2.9880384118605043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004124258306127531,
      "loss": 5.0215,
      "mean_token_accuracy": 0.20597931891679763,
      "num_tokens": 59391504.0,
      "step": 26605
    },
    {
      "entropy": 5.4545880317687985,
      "epoch": 2.988599988768462,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20888252258300782,
      "num_tokens": 59402648.0,
      "step": 26610
    },
    {
      "entropy": 5.47133994102478,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004123622424143248,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20643875896930694,
      "num_tokens": 59413842.0,
      "step": 26615
    },
    {
      "entropy": 5.509388256072998,
      "epoch": 2.989723142584377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041233044175410387,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20456794500350953,
      "num_tokens": 59425630.0,
      "step": 26620
    },
    {
      "entropy": 5.421006107330323,
      "epoch": 2.9902847194923345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041229863672122844,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21247529089450837,
      "num_tokens": 59437140.0,
      "step": 26625
    },
    {
      "entropy": 5.378793096542358,
      "epoch": 2.990846296400292,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004122668273167111,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21692056506872176,
      "num_tokens": 59447629.0,
      "step": 26630
    },
    {
      "entropy": 5.375456285476685,
      "epoch": 2.9914078733082494,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004122350135415648,
      "loss": 5.094,
      "mean_token_accuracy": 0.20378992557525635,
      "num_tokens": 59458469.0,
      "step": 26635
    },
    {
      "entropy": 5.398623609542847,
      "epoch": 2.991969450216207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004122031953968024,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21028599739074708,
      "num_tokens": 59468864.0,
      "step": 26640
    },
    {
      "entropy": 5.365701389312744,
      "epoch": 2.9925310271241647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041217137288343704,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21057586818933488,
      "num_tokens": 59480689.0,
      "step": 26645
    },
    {
      "entropy": 5.493232536315918,
      "epoch": 2.993092604032122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412139546002482,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20955168455839157,
      "num_tokens": 59491510.0,
      "step": 26650
    },
    {
      "entropy": 5.444991159439087,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20688148438930512,
      "num_tokens": 59502838.0,
      "step": 26655
    },
    {
      "entropy": 5.453610610961914,
      "epoch": 2.9942157578480373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004120758791418564,
      "loss": 5.0867,
      "mean_token_accuracy": 0.214074245095253,
      "num_tokens": 59514356.0,
      "step": 26660
    },
    {
      "entropy": 5.405602502822876,
      "epoch": 2.994777334755995,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004120440391642131,
      "loss": 5.0567,
      "mean_token_accuracy": 0.20425528287887573,
      "num_tokens": 59525933.0,
      "step": 26665
    },
    {
      "entropy": 5.522029399871826,
      "epoch": 2.9953389116639526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20350732058286666,
      "num_tokens": 59536856.0,
      "step": 26670
    },
    {
      "entropy": 5.460264587402344,
      "epoch": 2.99590048857191,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21336050778627397,
      "num_tokens": 59546023.0,
      "step": 26675
    },
    {
      "entropy": 5.392319917678833,
      "epoch": 2.9964620654798675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004119484930541267,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2168182134628296,
      "num_tokens": 59557515.0,
      "step": 26680
    },
    {
      "entropy": 5.3842895984649655,
      "epoch": 2.997023642387825,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.1152,
      "mean_token_accuracy": 0.2038549706339836,
      "num_tokens": 59570198.0,
      "step": 26685
    },
    {
      "entropy": 5.466392230987549,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.1757,
      "mean_token_accuracy": 0.2049131527543068,
      "num_tokens": 59580365.0,
      "step": 26690
    },
    {
      "entropy": 5.386739587783813,
      "epoch": 2.99814679620374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004118529076996029,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21258673667907715,
      "num_tokens": 59591304.0,
      "step": 26695
    },
    {
      "entropy": 5.479853820800781,
      "epoch": 2.9987083731116977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.1736,
      "mean_token_accuracy": 0.201378932595253,
      "num_tokens": 59603005.0,
      "step": 26700
    },
    {
      "entropy": 5.4104677677154545,
      "epoch": 2.9992699500196554,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041178916234098134,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21422571390867234,
      "num_tokens": 59614211.0,
      "step": 26705
    },
    {
      "entropy": 5.360387945175171,
      "epoch": 2.999831526927613,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004117572831280332,
      "loss": 5.0319,
      "mean_token_accuracy": 0.20936227440834046,
      "num_tokens": 59624162.0,
      "step": 26710
    },
    {
      "entropy": 5.376235644022624,
      "epoch": 3.0003369461447744,
      "grad_norm": 1.125,
      "learning_rate": 0.00041172539956068,
      "loss": 5.0196,
      "mean_token_accuracy": 0.20490415228737724,
      "num_tokens": 59633165.0,
      "step": 26715
    },
    {
      "entropy": 5.536422252655029,
      "epoch": 3.000898523052732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041169351163993705,
      "loss": 5.0847,
      "mean_token_accuracy": 0.20264710187911988,
      "num_tokens": 59645130.0,
      "step": 26720
    },
    {
      "entropy": 5.397323560714722,
      "epoch": 3.0014600999606897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2271607279777527,
      "num_tokens": 59656826.0,
      "step": 26725
    },
    {
      "entropy": 5.4490326881408695,
      "epoch": 3.002021676868647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004116297227423435,
      "loss": 5.1039,
      "mean_token_accuracy": 0.2125353455543518,
      "num_tokens": 59667743.0,
      "step": 26730
    },
    {
      "entropy": 5.3049205303192135,
      "epoch": 3.0025832537766046,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.878,
      "mean_token_accuracy": 0.22438380867242813,
      "num_tokens": 59679750.0,
      "step": 26735
    },
    {
      "entropy": 5.308299779891968,
      "epoch": 3.0031448306845623,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004115659164433765,
      "loss": 4.955,
      "mean_token_accuracy": 0.21746192276477813,
      "num_tokens": 59689113.0,
      "step": 26740
    },
    {
      "entropy": 5.371446895599365,
      "epoch": 3.00370640759252,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041153400677091745,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21709476113319398,
      "num_tokens": 59701282.0,
      "step": 26745
    },
    {
      "entropy": 5.3990943908691404,
      "epoch": 3.004267984500477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041150209275116266,
      "loss": 5.0119,
      "mean_token_accuracy": 0.20761092752218246,
      "num_tokens": 59713536.0,
      "step": 26750
    },
    {
      "entropy": 5.508176946640015,
      "epoch": 3.004829561408435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041147017438512834,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20768249928951263,
      "num_tokens": 59723818.0,
      "step": 26755
    },
    {
      "entropy": 5.532594347000122,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041143825167383073,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20945964008569717,
      "num_tokens": 59735486.0,
      "step": 26760
    },
    {
      "entropy": 5.466030645370483,
      "epoch": 3.00595271522435,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004114063246182863,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2075786754488945,
      "num_tokens": 59746006.0,
      "step": 26765
    },
    {
      "entropy": 5.478054809570312,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041137439321951147,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2149529665708542,
      "num_tokens": 59756346.0,
      "step": 26770
    },
    {
      "entropy": 5.391297197341919,
      "epoch": 3.007075869040265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041134245747852314,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21939416527748107,
      "num_tokens": 59768038.0,
      "step": 26775
    },
    {
      "entropy": 5.453972864151001,
      "epoch": 3.0076374459482227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004113105173963382,
      "loss": 5.0521,
      "mean_token_accuracy": 0.20971456170082092,
      "num_tokens": 59779211.0,
      "step": 26780
    },
    {
      "entropy": 5.380724716186523,
      "epoch": 3.0081990228561803,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004112785729739733,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21123794466257095,
      "num_tokens": 59790777.0,
      "step": 26785
    },
    {
      "entropy": 5.467867231369018,
      "epoch": 3.0087605997641376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041124662421244593,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20663250237703323,
      "num_tokens": 59802030.0,
      "step": 26790
    },
    {
      "entropy": 5.471771144866944,
      "epoch": 3.0093221766720952,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004112146711127732,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21010865420103073,
      "num_tokens": 59814616.0,
      "step": 26795
    },
    {
      "entropy": 5.382354402542115,
      "epoch": 3.009883753580053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004111827136759725,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2171312615275383,
      "num_tokens": 59825797.0,
      "step": 26800
    },
    {
      "entropy": 5.3722094058990475,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004111507519030613,
      "loss": 5.0395,
      "mean_token_accuracy": 0.2143327474594116,
      "num_tokens": 59836699.0,
      "step": 26805
    },
    {
      "entropy": 5.455108547210694,
      "epoch": 3.0110069073959678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.0846,
      "mean_token_accuracy": 0.210118892788887,
      "num_tokens": 59848575.0,
      "step": 26810
    },
    {
      "entropy": 5.401218414306641,
      "epoch": 3.0115684843039254,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004110868153529786,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21696003824472426,
      "num_tokens": 59860471.0,
      "step": 26815
    },
    {
      "entropy": 5.4486504077911375,
      "epoch": 3.012130061211883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004110548405778428,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21465688943862915,
      "num_tokens": 59871759.0,
      "step": 26820
    },
    {
      "entropy": 5.461083602905274,
      "epoch": 3.0126916381198403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004110228614706681,
      "loss": 4.9927,
      "mean_token_accuracy": 0.21088467240333558,
      "num_tokens": 59884375.0,
      "step": 26825
    },
    {
      "entropy": 5.43704047203064,
      "epoch": 3.013253215027798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041099087803247276,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2124039739370346,
      "num_tokens": 59895606.0,
      "step": 26830
    },
    {
      "entropy": 5.4541562557220455,
      "epoch": 3.0138147919357556,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004109588902642751,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20936283320188523,
      "num_tokens": 59908090.0,
      "step": 26835
    },
    {
      "entropy": 5.44768328666687,
      "epoch": 3.0143763688437133,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004109268981670936,
      "loss": 5.0533,
      "mean_token_accuracy": 0.2094138741493225,
      "num_tokens": 59919218.0,
      "step": 26840
    },
    {
      "entropy": 5.461885643005371,
      "epoch": 3.0149379457516705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000410894901741947,
      "loss": 5.1299,
      "mean_token_accuracy": 0.21293011903762818,
      "num_tokens": 59931195.0,
      "step": 26845
    },
    {
      "entropy": 5.406977415084839,
      "epoch": 3.015499522659628,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000410862900989854,
      "loss": 5.0706,
      "mean_token_accuracy": 0.2137948140501976,
      "num_tokens": 59943653.0,
      "step": 26850
    },
    {
      "entropy": 5.414356470108032,
      "epoch": 3.016061099567586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004108308959118336,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21293594390153886,
      "num_tokens": 59954677.0,
      "step": 26855
    },
    {
      "entropy": 5.488589811325073,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004107988865089047,
      "loss": 5.094,
      "mean_token_accuracy": 0.21058952063322067,
      "num_tokens": 59966135.0,
      "step": 26860
    },
    {
      "entropy": 5.375816917419433,
      "epoch": 3.0171842533835007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041076687278208664,
      "loss": 4.894,
      "mean_token_accuracy": 0.22137296199798584,
      "num_tokens": 59977828.0,
      "step": 26865
    },
    {
      "entropy": 5.35514612197876,
      "epoch": 3.0177458302914584,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21573070287704468,
      "num_tokens": 59989226.0,
      "step": 26870
    },
    {
      "entropy": 5.443681049346924,
      "epoch": 3.018307407199416,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041070283236086024,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22013031095266342,
      "num_tokens": 60000461.0,
      "step": 26875
    },
    {
      "entropy": 5.447268056869507,
      "epoch": 3.0188689841073737,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.149,
      "mean_token_accuracy": 0.20763793736696243,
      "num_tokens": 60012364.0,
      "step": 26880
    },
    {
      "entropy": 5.419142770767212,
      "epoch": 3.019430561015331,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041063877465631084,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2143715053796768,
      "num_tokens": 60022913.0,
      "step": 26885
    },
    {
      "entropy": 5.467012119293213,
      "epoch": 3.0199921379232886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004106067393253394,
      "loss": 4.9989,
      "mean_token_accuracy": 0.22040078639984131,
      "num_tokens": 60034407.0,
      "step": 26890
    },
    {
      "entropy": 5.383430147171021,
      "epoch": 3.0205537148312462,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004105746996765968,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20721897184848787,
      "num_tokens": 60044063.0,
      "step": 26895
    },
    {
      "entropy": 5.4207329750061035,
      "epoch": 3.021115291739204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004105426557111031,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22014065980911254,
      "num_tokens": 60054432.0,
      "step": 26900
    },
    {
      "entropy": 5.426228332519531,
      "epoch": 3.021676868647161,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004105106074298787,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21505020260810853,
      "num_tokens": 60065876.0,
      "step": 26905
    },
    {
      "entropy": 5.402584457397461,
      "epoch": 3.022238445555119,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000410478554833944,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21240318715572357,
      "num_tokens": 60076108.0,
      "step": 26910
    },
    {
      "entropy": 5.354814338684082,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041044649792431963,
      "loss": 4.985,
      "mean_token_accuracy": 0.2076695367693901,
      "num_tokens": 60087386.0,
      "step": 26915
    },
    {
      "entropy": 5.408126974105835,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.9695,
      "mean_token_accuracy": 0.22371392250061034,
      "num_tokens": 60099017.0,
      "step": 26920
    },
    {
      "entropy": 5.4229754447937015,
      "epoch": 3.0239231762789913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21625877618789674,
      "num_tokens": 60110676.0,
      "step": 26925
    },
    {
      "entropy": 5.341418504714966,
      "epoch": 3.024484753186949,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2240448072552681,
      "num_tokens": 60120807.0,
      "step": 26930
    },
    {
      "entropy": 5.4319874286651615,
      "epoch": 3.0250463300949066,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041031822716934086,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2098930299282074,
      "num_tokens": 60132174.0,
      "step": 26935
    },
    {
      "entropy": 5.492434978485107,
      "epoch": 3.025607907002864,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041028614870658106,
      "loss": 5.1469,
      "mean_token_accuracy": 0.20722096413373947,
      "num_tokens": 60144311.0,
      "step": 26940
    },
    {
      "entropy": 5.445954513549805,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004102540659362579,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20750462114810944,
      "num_tokens": 60155368.0,
      "step": 26945
    },
    {
      "entropy": 5.469135761260986,
      "epoch": 3.026731060818779,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041022197885939285,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20452825278043746,
      "num_tokens": 60167633.0,
      "step": 26950
    },
    {
      "entropy": 5.382994270324707,
      "epoch": 3.027292637726737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041018988747700745,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21661615520715713,
      "num_tokens": 60180596.0,
      "step": 26955
    },
    {
      "entropy": 5.480962133407592,
      "epoch": 3.027854214634694,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004101577917901236,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2072099357843399,
      "num_tokens": 60191811.0,
      "step": 26960
    },
    {
      "entropy": 5.4987955570220945,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041012569179976326,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21832088083028794,
      "num_tokens": 60202575.0,
      "step": 26965
    },
    {
      "entropy": 5.380842638015747,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004100935875069485,
      "loss": 4.976,
      "mean_token_accuracy": 0.2133752301335335,
      "num_tokens": 60212541.0,
      "step": 26970
    },
    {
      "entropy": 5.375646162033081,
      "epoch": 3.029538945358567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004100614789127015,
      "loss": 5.0351,
      "mean_token_accuracy": 0.22028861194849014,
      "num_tokens": 60223018.0,
      "step": 26975
    },
    {
      "entropy": 5.459233665466309,
      "epoch": 3.0301005222665243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004100293660180447,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2109009563922882,
      "num_tokens": 60234104.0,
      "step": 26980
    },
    {
      "entropy": 5.378936338424682,
      "epoch": 3.030662099174482,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040999724882400046,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2121136426925659,
      "num_tokens": 60245119.0,
      "step": 26985
    },
    {
      "entropy": 5.413856792449951,
      "epoch": 3.0312236760824396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004099651273315915,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22019678205251694,
      "num_tokens": 60255984.0,
      "step": 26990
    },
    {
      "entropy": 5.463854169845581,
      "epoch": 3.0317852529903973,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004099330015418405,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2221771866083145,
      "num_tokens": 60266764.0,
      "step": 26995
    },
    {
      "entropy": 5.383301496505737,
      "epoch": 3.0323468298983545,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040990087145577044,
      "loss": 4.9857,
      "mean_token_accuracy": 0.21601632237434387,
      "num_tokens": 60278284.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.172328568316625,
      "eval_loss": 5.195724010467529,
      "eval_mean_token_accuracy": 0.21233731101615114,
      "eval_num_tokens": 60278284.0,
      "eval_runtime": 58.8573,
      "eval_samples_per_second": 526.919,
      "eval_steps_per_second": 65.871,
      "step": 27000
    },
    {
      "entropy": 5.335855531692505,
      "epoch": 3.032908406806312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2194489523768425,
      "num_tokens": 60287993.0,
      "step": 27005
    },
    {
      "entropy": 5.435783576965332,
      "epoch": 3.03346998371427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004098365983987652,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20498273074626921,
      "num_tokens": 60300040.0,
      "step": 27010
    },
    {
      "entropy": 5.436420583724976,
      "epoch": 3.034031560622227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040980445542987653,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21091993004083634,
      "num_tokens": 60311008.0,
      "step": 27015
    },
    {
      "entropy": 5.380270385742188,
      "epoch": 3.0345931375301847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21245717406272888,
      "num_tokens": 60321071.0,
      "step": 27020
    },
    {
      "entropy": 5.431918430328369,
      "epoch": 3.0351547144381423,
      "grad_norm": 1.0,
      "learning_rate": 0.00040974015661644433,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20774587094783784,
      "num_tokens": 60333075.0,
      "step": 27025
    },
    {
      "entropy": 5.493340873718262,
      "epoch": 3.0357162913461,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004097080007739481,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20727697163820266,
      "num_tokens": 60343393.0,
      "step": 27030
    },
    {
      "entropy": 5.443082714080811,
      "epoch": 3.036277868254057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040967584064229695,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20412653386592866,
      "num_tokens": 60353981.0,
      "step": 27035
    },
    {
      "entropy": 5.378370571136474,
      "epoch": 3.036839445162015,
      "grad_norm": 1.125,
      "learning_rate": 0.00040964367622251473,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22033199965953826,
      "num_tokens": 60364544.0,
      "step": 27040
    },
    {
      "entropy": 5.46098895072937,
      "epoch": 3.0374010220699725,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004096115075156257,
      "loss": 5.0852,
      "mean_token_accuracy": 0.2139773726463318,
      "num_tokens": 60376031.0,
      "step": 27045
    },
    {
      "entropy": 5.457260990142823,
      "epoch": 3.03796259897793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.0982,
      "mean_token_accuracy": 0.2045975387096405,
      "num_tokens": 60387320.0,
      "step": 27050
    },
    {
      "entropy": 5.441420602798462,
      "epoch": 3.0385241758858874,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004095471572446242,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2143941804766655,
      "num_tokens": 60397345.0,
      "step": 27055
    },
    {
      "entropy": 5.443764495849609,
      "epoch": 3.039085752793845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004095149756825606,
      "loss": 5.0696,
      "mean_token_accuracy": 0.20568637251853944,
      "num_tokens": 60408194.0,
      "step": 27060
    },
    {
      "entropy": 5.475212478637696,
      "epoch": 3.0396473297018027,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004094827898374881,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20896393656730652,
      "num_tokens": 60418928.0,
      "step": 27065
    },
    {
      "entropy": 5.400350904464721,
      "epoch": 3.0402089066097604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004094505997104314,
      "loss": 4.91,
      "mean_token_accuracy": 0.21988089978694916,
      "num_tokens": 60428819.0,
      "step": 27070
    },
    {
      "entropy": 5.321444702148438,
      "epoch": 3.0407704835177176,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21725768297910691,
      "num_tokens": 60440575.0,
      "step": 27075
    },
    {
      "entropy": 5.419401693344116,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040938620661446526,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20909214168787002,
      "num_tokens": 60452977.0,
      "step": 27080
    },
    {
      "entropy": 5.464405155181884,
      "epoch": 3.041893637333633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004093540036476062,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2090132474899292,
      "num_tokens": 60463552.0,
      "step": 27085
    },
    {
      "entropy": 5.498808908462524,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040932179640286356,
      "loss": 5.0685,
      "mean_token_accuracy": 0.2112014576792717,
      "num_tokens": 60474674.0,
      "step": 27090
    },
    {
      "entropy": 5.404291439056396,
      "epoch": 3.043016791149548,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004092895848812628,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21905385553836823,
      "num_tokens": 60486140.0,
      "step": 27095
    },
    {
      "entropy": 5.395531845092774,
      "epoch": 3.0435783680575055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040925736908382953,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21972471177577974,
      "num_tokens": 60497386.0,
      "step": 27100
    },
    {
      "entropy": 5.499067640304565,
      "epoch": 3.044139944965463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004092251490115896,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20135437399148942,
      "num_tokens": 60508535.0,
      "step": 27105
    },
    {
      "entropy": 5.448532295227051,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040919292466556895,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20305000692605973,
      "num_tokens": 60520791.0,
      "step": 27110
    },
    {
      "entropy": 5.4129047870635985,
      "epoch": 3.045263098781378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004091606960467934,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2108118563890457,
      "num_tokens": 60533676.0,
      "step": 27115
    },
    {
      "entropy": 5.453302621841431,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040912846315628934,
      "loss": 5.0505,
      "mean_token_accuracy": 0.2098599597811699,
      "num_tokens": 60545216.0,
      "step": 27120
    },
    {
      "entropy": 5.327354574203492,
      "epoch": 3.0463862525972933,
      "grad_norm": 1.046875,
      "learning_rate": 0.000409096225995083,
      "loss": 4.9565,
      "mean_token_accuracy": 0.21978838443756105,
      "num_tokens": 60554730.0,
      "step": 27125
    },
    {
      "entropy": 5.30461015701294,
      "epoch": 3.0469478295052506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.8974,
      "mean_token_accuracy": 0.21576722711324692,
      "num_tokens": 60565494.0,
      "step": 27130
    },
    {
      "entropy": 5.417676162719727,
      "epoch": 3.047509406413208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040903173886466934,
      "loss": 5.029,
      "mean_token_accuracy": 0.2117558553814888,
      "num_tokens": 60576249.0,
      "step": 27135
    },
    {
      "entropy": 5.4016087532043455,
      "epoch": 3.048070983321166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2087640717625618,
      "num_tokens": 60587253.0,
      "step": 27140
    },
    {
      "entropy": 5.429412412643432,
      "epoch": 3.0486325602291235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040896723466376557,
      "loss": 5.0842,
      "mean_token_accuracy": 0.2084420770406723,
      "num_tokens": 60598158.0,
      "step": 27145
    },
    {
      "entropy": 5.421510457992554,
      "epoch": 3.0491941371370808,
      "grad_norm": 1.0,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22299146354198457,
      "num_tokens": 60608812.0,
      "step": 27150
    },
    {
      "entropy": 5.401135635375977,
      "epoch": 3.0497557140450384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004089027134005871,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22080521285533905,
      "num_tokens": 60619264.0,
      "step": 27155
    },
    {
      "entropy": 5.3051646709442135,
      "epoch": 3.050317290952996,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004088704463732127,
      "loss": 4.894,
      "mean_token_accuracy": 0.22044012546539307,
      "num_tokens": 60629688.0,
      "step": 27160
    },
    {
      "entropy": 5.38848876953125,
      "epoch": 3.0508788678609537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004088381750833514,
      "loss": 4.986,
      "mean_token_accuracy": 0.21420982331037522,
      "num_tokens": 60640790.0,
      "step": 27165
    },
    {
      "entropy": 5.443771076202393,
      "epoch": 3.051440444768911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004088058995320307,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21418090015649796,
      "num_tokens": 60651952.0,
      "step": 27170
    },
    {
      "entropy": 5.529635858535767,
      "epoch": 3.0520020216768686,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004087736197202784,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20857079476118087,
      "num_tokens": 60665388.0,
      "step": 27175
    },
    {
      "entropy": 5.336746597290039,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040874133564912195,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22352350056171416,
      "num_tokens": 60676407.0,
      "step": 27180
    },
    {
      "entropy": 5.34404182434082,
      "epoch": 3.053125175492784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040870904731958965,
      "loss": 4.9518,
      "mean_token_accuracy": 0.21728930473327637,
      "num_tokens": 60686796.0,
      "step": 27185
    },
    {
      "entropy": 5.4583127975463865,
      "epoch": 3.053686752400741,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004086767547327094,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20573580712080003,
      "num_tokens": 60697770.0,
      "step": 27190
    },
    {
      "entropy": 5.449737215042115,
      "epoch": 3.054248329308699,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040864445788950935,
      "loss": 4.991,
      "mean_token_accuracy": 0.2127147436141968,
      "num_tokens": 60709059.0,
      "step": 27195
    },
    {
      "entropy": 5.39676513671875,
      "epoch": 3.0548099062166565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000408612156791018,
      "loss": 4.987,
      "mean_token_accuracy": 0.21364986151456833,
      "num_tokens": 60720228.0,
      "step": 27200
    },
    {
      "entropy": 5.476347398757935,
      "epoch": 3.0553714831246137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004085798514382637,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20978060513734817,
      "num_tokens": 60730658.0,
      "step": 27205
    },
    {
      "entropy": 5.419907665252685,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040854754183227516,
      "loss": 4.996,
      "mean_token_accuracy": 0.21580075323581696,
      "num_tokens": 60741757.0,
      "step": 27210
    },
    {
      "entropy": 5.403694534301758,
      "epoch": 3.056494636940529,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004085152279740811,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21310523599386216,
      "num_tokens": 60755452.0,
      "step": 27215
    },
    {
      "entropy": 5.450388193130493,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040848290986471037,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21242587864398957,
      "num_tokens": 60765715.0,
      "step": 27220
    },
    {
      "entropy": 5.449524211883545,
      "epoch": 3.057617790756444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.0834,
      "mean_token_accuracy": 0.2113735094666481,
      "num_tokens": 60777021.0,
      "step": 27225
    },
    {
      "entropy": 5.351037549972534,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040841826089655525,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2132211834192276,
      "num_tokens": 60787309.0,
      "step": 27230
    },
    {
      "entropy": 5.4086240291595455,
      "epoch": 3.0587409445723592,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040838593003982926,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21154128909111022,
      "num_tokens": 60797566.0,
      "step": 27235
    },
    {
      "entropy": 5.466561889648437,
      "epoch": 3.059302521480317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004083535949360436,
      "loss": 5.0294,
      "mean_token_accuracy": 0.2179264545440674,
      "num_tokens": 60808213.0,
      "step": 27240
    },
    {
      "entropy": 5.451826667785644,
      "epoch": 3.059864098388274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004083212555862277,
      "loss": 5.0772,
      "mean_token_accuracy": 0.20937979817390442,
      "num_tokens": 60820656.0,
      "step": 27245
    },
    {
      "entropy": 5.416002321243286,
      "epoch": 3.0604256752962318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004082889119914113,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2165740102529526,
      "num_tokens": 60831572.0,
      "step": 27250
    },
    {
      "entropy": 5.335677719116211,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004082565641526242,
      "loss": 4.923,
      "mean_token_accuracy": 0.22145743668079376,
      "num_tokens": 60842443.0,
      "step": 27255
    },
    {
      "entropy": 5.37428560256958,
      "epoch": 3.061548829112147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040822421207089654,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21780834197998047,
      "num_tokens": 60853427.0,
      "step": 27260
    },
    {
      "entropy": 5.477810478210449,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004081918557472582,
      "loss": 5.1143,
      "mean_token_accuracy": 0.2089615896344185,
      "num_tokens": 60864517.0,
      "step": 27265
    },
    {
      "entropy": 5.4783447265625,
      "epoch": 3.062671982928062,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004081594951827396,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21635268479585648,
      "num_tokens": 60874280.0,
      "step": 27270
    },
    {
      "entropy": 5.44776496887207,
      "epoch": 3.0632335598360196,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040812713037837096,
      "loss": 5.034,
      "mean_token_accuracy": 0.21340147256851197,
      "num_tokens": 60885656.0,
      "step": 27275
    },
    {
      "entropy": 5.458360910415649,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040809476133518284,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21605492979288102,
      "num_tokens": 60896856.0,
      "step": 27280
    },
    {
      "entropy": 5.319024467468262,
      "epoch": 3.0643567136519345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040806238805420595,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2142007365822792,
      "num_tokens": 60907464.0,
      "step": 27285
    },
    {
      "entropy": 5.435304832458496,
      "epoch": 3.064918290559892,
      "grad_norm": 1.0625,
      "learning_rate": 0.000408030010536471,
      "loss": 5.0556,
      "mean_token_accuracy": 0.20736779719591142,
      "num_tokens": 60918321.0,
      "step": 27290
    },
    {
      "entropy": 5.49639663696289,
      "epoch": 3.06547986746785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040799762878300895,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20380956381559373,
      "num_tokens": 60928823.0,
      "step": 27295
    },
    {
      "entropy": 5.562618064880371,
      "epoch": 3.066041444375807,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040796524279485084,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20801119804382323,
      "num_tokens": 60940161.0,
      "step": 27300
    },
    {
      "entropy": 5.430185985565186,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040793285257302776,
      "loss": 5.0986,
      "mean_token_accuracy": 0.2050876423716545,
      "num_tokens": 60951968.0,
      "step": 27305
    },
    {
      "entropy": 5.500172710418701,
      "epoch": 3.0671645981917224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000407900458118571,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20382774472236634,
      "num_tokens": 60963619.0,
      "step": 27310
    },
    {
      "entropy": 5.477764368057251,
      "epoch": 3.06772617509968,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004078680594325123,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21857806593179702,
      "num_tokens": 60974615.0,
      "step": 27315
    },
    {
      "entropy": 5.413342571258545,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.171875,
      "learning_rate": 0.000407835656515883,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20464511960744858,
      "num_tokens": 60986943.0,
      "step": 27320
    },
    {
      "entropy": 5.397714757919312,
      "epoch": 3.068849328915595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040780324936971486,
      "loss": 5.1396,
      "mean_token_accuracy": 0.2113993525505066,
      "num_tokens": 60998306.0,
      "step": 27325
    },
    {
      "entropy": 5.501650142669678,
      "epoch": 3.0694109058235526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004077708379950397,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2071204110980034,
      "num_tokens": 61010102.0,
      "step": 27330
    },
    {
      "entropy": 5.499847364425659,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004077384223928896,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21283882707357407,
      "num_tokens": 61019474.0,
      "step": 27335
    },
    {
      "entropy": 5.394285726547241,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004077060025642967,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2191701278090477,
      "num_tokens": 61030401.0,
      "step": 27340
    },
    {
      "entropy": 5.509185028076172,
      "epoch": 3.071095636547425,
      "grad_norm": 1.125,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20790794640779495,
      "num_tokens": 61042139.0,
      "step": 27345
    },
    {
      "entropy": 5.352402925491333,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004076411502319114,
      "loss": 4.9496,
      "mean_token_accuracy": 0.2153944656252861,
      "num_tokens": 61052468.0,
      "step": 27350
    },
    {
      "entropy": 5.439441871643067,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040760871773018405,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21428596675395967,
      "num_tokens": 61063625.0,
      "step": 27355
    },
    {
      "entropy": 5.345438051223755,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004075762810061437,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2190405920147896,
      "num_tokens": 61074246.0,
      "step": 27360
    },
    {
      "entropy": 5.470649337768554,
      "epoch": 3.0733419441792553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000407543840060823,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20358238220214844,
      "num_tokens": 61086096.0,
      "step": 27365
    },
    {
      "entropy": 5.471615743637085,
      "epoch": 3.073903521087213,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004075113948952551,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21867919713258743,
      "num_tokens": 61097176.0,
      "step": 27370
    },
    {
      "entropy": 5.503083372116089,
      "epoch": 3.0744650979951706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004074789455104731,
      "loss": 5.1421,
      "mean_token_accuracy": 0.21230696141719818,
      "num_tokens": 61107865.0,
      "step": 27375
    },
    {
      "entropy": 5.465560007095337,
      "epoch": 3.075026674903128,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040744649190750995,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2151864290237427,
      "num_tokens": 61118682.0,
      "step": 27380
    },
    {
      "entropy": 5.452250051498413,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004074140340873992,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22157640010118484,
      "num_tokens": 61130210.0,
      "step": 27385
    },
    {
      "entropy": 5.347085952758789,
      "epoch": 3.076149828719043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004073815720511743,
      "loss": 5.0731,
      "mean_token_accuracy": 0.2126825988292694,
      "num_tokens": 61141216.0,
      "step": 27390
    },
    {
      "entropy": 5.401501321792603,
      "epoch": 3.0767114056270004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040734910579986875,
      "loss": 5.07,
      "mean_token_accuracy": 0.207163967192173,
      "num_tokens": 61153233.0,
      "step": 27395
    },
    {
      "entropy": 5.4468787670135494,
      "epoch": 3.077272982534958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040731663533451624,
      "loss": 4.976,
      "mean_token_accuracy": 0.2176486149430275,
      "num_tokens": 61163044.0,
      "step": 27400
    },
    {
      "entropy": 5.405003356933594,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040728416065615087,
      "loss": 5.055,
      "mean_token_accuracy": 0.20984024405479432,
      "num_tokens": 61174595.0,
      "step": 27405
    },
    {
      "entropy": 5.35715012550354,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004072516817658065,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21411350965499878,
      "num_tokens": 61185215.0,
      "step": 27410
    },
    {
      "entropy": 5.337227439880371,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040721919866451724,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22290213257074357,
      "num_tokens": 61195504.0,
      "step": 27415
    },
    {
      "entropy": 5.3478759765625,
      "epoch": 3.0795192901667883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2163272589445114,
      "num_tokens": 61206040.0,
      "step": 27420
    },
    {
      "entropy": 5.443912410736084,
      "epoch": 3.080080867074746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20911585688591003,
      "num_tokens": 61217797.0,
      "step": 27425
    },
    {
      "entropy": 5.461352682113647,
      "epoch": 3.0806424439827036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040712172410532397,
      "loss": 5.0745,
      "mean_token_accuracy": 0.2091207355260849,
      "num_tokens": 61230158.0,
      "step": 27430
    },
    {
      "entropy": 5.472966861724854,
      "epoch": 3.081204020890661,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004070892241705994,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21655634492635728,
      "num_tokens": 61241072.0,
      "step": 27435
    },
    {
      "entropy": 5.321994590759277,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22058464884757994,
      "num_tokens": 61251057.0,
      "step": 27440
    },
    {
      "entropy": 5.308177661895752,
      "epoch": 3.082327174706576,
      "grad_norm": 1.03125,
      "learning_rate": 0.000407024211684869,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22090588510036469,
      "num_tokens": 61262729.0,
      "step": 27445
    },
    {
      "entropy": 5.355594062805176,
      "epoch": 3.082888751614534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040699169913593306,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22143718749284744,
      "num_tokens": 61274665.0,
      "step": 27450
    },
    {
      "entropy": 5.4329752922058105,
      "epoch": 3.083450328522491,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004069591823843302,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21400293111801147,
      "num_tokens": 61285244.0,
      "step": 27455
    },
    {
      "entropy": 5.415931558609008,
      "epoch": 3.0840119054304487,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004069266614310959,
      "loss": 5.035,
      "mean_token_accuracy": 0.2096407562494278,
      "num_tokens": 61295739.0,
      "step": 27460
    },
    {
      "entropy": 5.416792678833008,
      "epoch": 3.0845734823384063,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040689413627726544,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21824028491973876,
      "num_tokens": 61305964.0,
      "step": 27465
    },
    {
      "entropy": 5.4731526374816895,
      "epoch": 3.085135059246364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.0922,
      "mean_token_accuracy": 0.21077089607715607,
      "num_tokens": 61317004.0,
      "step": 27470
    },
    {
      "entropy": 5.467885875701905,
      "epoch": 3.085696636154321,
      "grad_norm": 1.0,
      "learning_rate": 0.00040682907337195894,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20023516267538072,
      "num_tokens": 61329014.0,
      "step": 27475
    },
    {
      "entropy": 5.351541566848755,
      "epoch": 3.086258213062279,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004067965356225545,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21282070726156235,
      "num_tokens": 61340522.0,
      "step": 27480
    },
    {
      "entropy": 5.473064231872558,
      "epoch": 3.0868197899702365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040676399367669733,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2137558713555336,
      "num_tokens": 61351595.0,
      "step": 27485
    },
    {
      "entropy": 5.4911253452301025,
      "epoch": 3.0873813668781938,
      "grad_norm": 1.0,
      "learning_rate": 0.0004067314475354234,
      "loss": 5.0759,
      "mean_token_accuracy": 0.21383837461471558,
      "num_tokens": 61362432.0,
      "step": 27490
    },
    {
      "entropy": 5.400733137130738,
      "epoch": 3.0879429437861514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040669889719976916,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21858576834201812,
      "num_tokens": 61372717.0,
      "step": 27495
    },
    {
      "entropy": 5.417714977264405,
      "epoch": 3.088504520694109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.9937,
      "mean_token_accuracy": 0.20640729665756224,
      "num_tokens": 61383150.0,
      "step": 27500
    },
    {
      "entropy": 5.4416156768798825,
      "epoch": 3.0890660976020667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040663378394946534,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20024597495794297,
      "num_tokens": 61395538.0,
      "step": 27505
    },
    {
      "entropy": 5.460188531875611,
      "epoch": 3.089627674510024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004066012210368889,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20947265177965163,
      "num_tokens": 61407523.0,
      "step": 27510
    },
    {
      "entropy": 5.436410903930664,
      "epoch": 3.0901892514179816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004065686539340786,
      "loss": 4.9977,
      "mean_token_accuracy": 0.2063826337456703,
      "num_tokens": 61418680.0,
      "step": 27515
    },
    {
      "entropy": 5.371262073516846,
      "epoch": 3.0907508283259393,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040653608264207134,
      "loss": 5.0135,
      "mean_token_accuracy": 0.2087418779730797,
      "num_tokens": 61430006.0,
      "step": 27520
    },
    {
      "entropy": 5.2948486328125,
      "epoch": 3.091312405233897,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040650350716190417,
      "loss": 4.8526,
      "mean_token_accuracy": 0.22751565724611283,
      "num_tokens": 61440564.0,
      "step": 27525
    },
    {
      "entropy": 5.3188512325286865,
      "epoch": 3.091873982141854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004064709274946143,
      "loss": 4.8676,
      "mean_token_accuracy": 0.215931499004364,
      "num_tokens": 61451589.0,
      "step": 27530
    },
    {
      "entropy": 5.287192964553833,
      "epoch": 3.092435559049812,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004064383436412392,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2215173602104187,
      "num_tokens": 61462177.0,
      "step": 27535
    },
    {
      "entropy": 5.46807861328125,
      "epoch": 3.0929971359577695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004064057556028162,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21214713305234909,
      "num_tokens": 61472626.0,
      "step": 27540
    },
    {
      "entropy": 5.392845821380615,
      "epoch": 3.093558712865727,
      "grad_norm": 1.03125,
      "learning_rate": 0.000406373163380383,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2184429571032524,
      "num_tokens": 61483839.0,
      "step": 27545
    },
    {
      "entropy": 5.366174697875977,
      "epoch": 3.0941202897736844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040634056697497726,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2246927410364151,
      "num_tokens": 61495099.0,
      "step": 27550
    },
    {
      "entropy": 5.39782943725586,
      "epoch": 3.094681866681642,
      "grad_norm": 1.0625,
      "learning_rate": 0.000406307966387637,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21744150519371033,
      "num_tokens": 61506902.0,
      "step": 27555
    },
    {
      "entropy": 5.433969974517822,
      "epoch": 3.0952434435895997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040627536161940016,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20614322274923325,
      "num_tokens": 61519357.0,
      "step": 27560
    },
    {
      "entropy": 5.440590190887451,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.125,
      "learning_rate": 0.0004062427526713049,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22205025106668472,
      "num_tokens": 61529510.0,
      "step": 27565
    },
    {
      "entropy": 5.299580335617065,
      "epoch": 3.0963665974055146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22692011147737504,
      "num_tokens": 61539679.0,
      "step": 27570
    },
    {
      "entropy": 5.34421181678772,
      "epoch": 3.0969281743134722,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004061775222396922,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22017921805381774,
      "num_tokens": 61549488.0,
      "step": 27575
    },
    {
      "entropy": 5.406993675231933,
      "epoch": 3.09748975122143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004061449007582519,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21528475880622863,
      "num_tokens": 61561167.0,
      "step": 27580
    },
    {
      "entropy": 5.420262575149536,
      "epoch": 3.098051328129387,
      "grad_norm": 1.0,
      "learning_rate": 0.000406112275101107,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22029442191123963,
      "num_tokens": 61572730.0,
      "step": 27585
    },
    {
      "entropy": 5.351977872848511,
      "epoch": 3.0986129050373448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040607964526929655,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2167113795876503,
      "num_tokens": 61583846.0,
      "step": 27590
    },
    {
      "entropy": 5.369285774230957,
      "epoch": 3.0991744819453024,
      "grad_norm": 1.125,
      "learning_rate": 0.00040604701126385924,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2142985373735428,
      "num_tokens": 61594360.0,
      "step": 27595
    },
    {
      "entropy": 5.424808597564697,
      "epoch": 3.09973605885326,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004060143730858342,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21798200905323029,
      "num_tokens": 61606750.0,
      "step": 27600
    },
    {
      "entropy": 5.41988525390625,
      "epoch": 3.1002976357612173,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040598173073626086,
      "loss": 5.0381,
      "mean_token_accuracy": 0.20717861652374267,
      "num_tokens": 61618238.0,
      "step": 27605
    },
    {
      "entropy": 5.45361967086792,
      "epoch": 3.100859212669175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004059490842161783,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20451831668615342,
      "num_tokens": 61629568.0,
      "step": 27610
    },
    {
      "entropy": 5.4549966812133786,
      "epoch": 3.1014207895771326,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040591643352662615,
      "loss": 5.0948,
      "mean_token_accuracy": 0.2146727114915848,
      "num_tokens": 61640103.0,
      "step": 27615
    },
    {
      "entropy": 5.399542379379272,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000405883778668644,
      "loss": 4.927,
      "mean_token_accuracy": 0.22378535121679305,
      "num_tokens": 61651392.0,
      "step": 27620
    },
    {
      "entropy": 5.451713800430298,
      "epoch": 3.1025439433930475,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040585111964327157,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20275364965200424,
      "num_tokens": 61664932.0,
      "step": 27625
    },
    {
      "entropy": 5.372363424301147,
      "epoch": 3.103105520301005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.9437,
      "mean_token_accuracy": 0.21398891806602477,
      "num_tokens": 61675719.0,
      "step": 27630
    },
    {
      "entropy": 5.467622137069702,
      "epoch": 3.103667097208963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004057857890945155,
      "loss": 5.1085,
      "mean_token_accuracy": 0.21035877317190171,
      "num_tokens": 61687298.0,
      "step": 27635
    },
    {
      "entropy": 5.461323642730713,
      "epoch": 3.1042286741169205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040575311757321203,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21143337786197663,
      "num_tokens": 61697943.0,
      "step": 27640
    },
    {
      "entropy": 5.4669764041900635,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040572044188867863,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21212414652109146,
      "num_tokens": 61708897.0,
      "step": 27645
    },
    {
      "entropy": 5.440397310256958,
      "epoch": 3.1053518279328354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.123,
      "mean_token_accuracy": 0.21437063813209534,
      "num_tokens": 61720409.0,
      "step": 27650
    },
    {
      "entropy": 5.47587218284607,
      "epoch": 3.105913404840793,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004056550780340835,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20902468413114547,
      "num_tokens": 61731378.0,
      "step": 27655
    },
    {
      "entropy": 5.3964790344238285,
      "epoch": 3.1064749817487507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004056223898661031,
      "loss": 4.9589,
      "mean_token_accuracy": 0.20996004194021226,
      "num_tokens": 61742713.0,
      "step": 27660
    },
    {
      "entropy": 5.456196117401123,
      "epoch": 3.107036558656708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040558969753905514,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21571034491062163,
      "num_tokens": 61753561.0,
      "step": 27665
    },
    {
      "entropy": 5.466312026977539,
      "epoch": 3.1075981355646656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004055570010539806,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21355261653661728,
      "num_tokens": 61764966.0,
      "step": 27670
    },
    {
      "entropy": 5.4417191505432125,
      "epoch": 3.1081597124726232,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004055243004119204,
      "loss": 5.0627,
      "mean_token_accuracy": 0.20770394951105117,
      "num_tokens": 61776988.0,
      "step": 27675
    },
    {
      "entropy": 5.395893907546997,
      "epoch": 3.108721289380581,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040549159561391595,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20836199223995208,
      "num_tokens": 61788327.0,
      "step": 27680
    },
    {
      "entropy": 5.356692123413086,
      "epoch": 3.109282866288538,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004054588866610085,
      "loss": 4.9366,
      "mean_token_accuracy": 0.21955997198820115,
      "num_tokens": 61799872.0,
      "step": 27685
    },
    {
      "entropy": 5.400307559967041,
      "epoch": 3.1098444431964958,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040542617355423947,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21557986736297607,
      "num_tokens": 61810736.0,
      "step": 27690
    },
    {
      "entropy": 5.471657133102417,
      "epoch": 3.1104060201044534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040539345629465043,
      "loss": 5.1164,
      "mean_token_accuracy": 0.2072710871696472,
      "num_tokens": 61821236.0,
      "step": 27695
    },
    {
      "entropy": 5.526703596115112,
      "epoch": 3.1109675970124107,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040536073488328325,
      "loss": 5.0836,
      "mean_token_accuracy": 0.20846316516399382,
      "num_tokens": 61832282.0,
      "step": 27700
    },
    {
      "entropy": 5.547054672241211,
      "epoch": 3.1115291739203683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.0913,
      "mean_token_accuracy": 0.20865544974803923,
      "num_tokens": 61842376.0,
      "step": 27705
    },
    {
      "entropy": 5.42389817237854,
      "epoch": 3.112090750828326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004052952796093817,
      "loss": 5.0425,
      "mean_token_accuracy": 0.21642351299524307,
      "num_tokens": 61852227.0,
      "step": 27710
    },
    {
      "entropy": 5.472598218917847,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.125,
      "learning_rate": 0.00040526254574893154,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21374387592077254,
      "num_tokens": 61863119.0,
      "step": 27715
    },
    {
      "entropy": 5.358550405502319,
      "epoch": 3.113213904644241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040522980774087143,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21833953708410264,
      "num_tokens": 61874036.0,
      "step": 27720
    },
    {
      "entropy": 5.331623601913452,
      "epoch": 3.1137754815521985,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004051970655862437,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2176857441663742,
      "num_tokens": 61884744.0,
      "step": 27725
    },
    {
      "entropy": 5.428954124450684,
      "epoch": 3.114337058460156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004051643192860908,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2068003833293915,
      "num_tokens": 61896866.0,
      "step": 27730
    },
    {
      "entropy": 5.395007753372193,
      "epoch": 3.114898635368114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040513156884145564,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22498347908258437,
      "num_tokens": 61907384.0,
      "step": 27735
    },
    {
      "entropy": 5.269737672805786,
      "epoch": 3.115460212276071,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004050988142533808,
      "loss": 4.8374,
      "mean_token_accuracy": 0.22247960418462753,
      "num_tokens": 61917719.0,
      "step": 27740
    },
    {
      "entropy": 5.341945934295654,
      "epoch": 3.1160217891840287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004050660555229093,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2172018766403198,
      "num_tokens": 61928159.0,
      "step": 27745
    },
    {
      "entropy": 5.40667610168457,
      "epoch": 3.1165833660919864,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004050332926510841,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2168777346611023,
      "num_tokens": 61938748.0,
      "step": 27750
    },
    {
      "entropy": 5.356906461715698,
      "epoch": 3.117144942999944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.8902,
      "mean_token_accuracy": 0.22732681035995483,
      "num_tokens": 61950429.0,
      "step": 27755
    },
    {
      "entropy": 5.375272798538208,
      "epoch": 3.1177065199079013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21136504411697388,
      "num_tokens": 61961156.0,
      "step": 27760
    },
    {
      "entropy": 5.388326644897461,
      "epoch": 3.118268096815859,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040493497919791906,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2144156739115715,
      "num_tokens": 61972673.0,
      "step": 27765
    },
    {
      "entropy": 5.379097890853882,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040490219977111237,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21844060122966766,
      "num_tokens": 61984481.0,
      "step": 27770
    },
    {
      "entropy": 5.389525604248047,
      "epoch": 3.1193912506317742,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004048694162081693,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21967082023620604,
      "num_tokens": 61994919.0,
      "step": 27775
    },
    {
      "entropy": 5.442720746994018,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004048366285101336,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22043980807065963,
      "num_tokens": 62005303.0,
      "step": 27780
    },
    {
      "entropy": 5.34294581413269,
      "epoch": 3.120514404447689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004048038366780494,
      "loss": 4.85,
      "mean_token_accuracy": 0.22302231192588806,
      "num_tokens": 62015104.0,
      "step": 27785
    },
    {
      "entropy": 5.2856879234313965,
      "epoch": 3.121075981355647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004047710407129606,
      "loss": 4.94,
      "mean_token_accuracy": 0.21046449691057206,
      "num_tokens": 62025577.0,
      "step": 27790
    },
    {
      "entropy": 5.522017765045166,
      "epoch": 3.121637558263604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.1644,
      "mean_token_accuracy": 0.20111519992351531,
      "num_tokens": 62037001.0,
      "step": 27795
    },
    {
      "entropy": 5.514863443374634,
      "epoch": 3.1221991351715617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004047054363879466,
      "loss": 5.0822,
      "mean_token_accuracy": 0.207161246240139,
      "num_tokens": 62049520.0,
      "step": 27800
    },
    {
      "entropy": 5.422874546051025,
      "epoch": 3.1227607120795193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040467262803011025,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21386611610651016,
      "num_tokens": 62061650.0,
      "step": 27805
    },
    {
      "entropy": 5.400363922119141,
      "epoch": 3.123322288987477,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.0893,
      "mean_token_accuracy": 0.21154745370149614,
      "num_tokens": 62074433.0,
      "step": 27810
    },
    {
      "entropy": 5.405375862121582,
      "epoch": 3.123883865895434,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040460699892900197,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21904438436031343,
      "num_tokens": 62084885.0,
      "step": 27815
    },
    {
      "entropy": 5.442333459854126,
      "epoch": 3.124445442803392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004045741781878198,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21123866885900497,
      "num_tokens": 62096847.0,
      "step": 27820
    },
    {
      "entropy": 5.469074058532715,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040454135332094543,
      "loss": 5.0771,
      "mean_token_accuracy": 0.21321185380220414,
      "num_tokens": 62108131.0,
      "step": 27825
    },
    {
      "entropy": 5.42802996635437,
      "epoch": 3.125568596619307,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004045085243294242,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2161060929298401,
      "num_tokens": 62118450.0,
      "step": 27830
    },
    {
      "entropy": 5.440286493301391,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004044756912143013,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21139421910047532,
      "num_tokens": 62130702.0,
      "step": 27835
    },
    {
      "entropy": 5.452565622329712,
      "epoch": 3.126691750435222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040444285397662225,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20879294723272324,
      "num_tokens": 62142863.0,
      "step": 27840
    },
    {
      "entropy": 5.369851732254029,
      "epoch": 3.1272533273431797,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004044100126174324,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2124473199248314,
      "num_tokens": 62153787.0,
      "step": 27845
    },
    {
      "entropy": 5.450869131088257,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040437716713777756,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21128561794757844,
      "num_tokens": 62166862.0,
      "step": 27850
    },
    {
      "entropy": 5.46034345626831,
      "epoch": 3.1283764811590946,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004043443175387036,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21591718941926957,
      "num_tokens": 62177431.0,
      "step": 27855
    },
    {
      "entropy": 5.436140060424805,
      "epoch": 3.1289380580670523,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040431146382125644,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21251683980226516,
      "num_tokens": 62189465.0,
      "step": 27860
    },
    {
      "entropy": 5.422050046920776,
      "epoch": 3.12949963497501,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000404278605986482,
      "loss": 4.9184,
      "mean_token_accuracy": 0.223202383518219,
      "num_tokens": 62201327.0,
      "step": 27865
    },
    {
      "entropy": 5.3635852336883545,
      "epoch": 3.1300612118829676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004042457440354267,
      "loss": 5.0078,
      "mean_token_accuracy": 0.20764648765325547,
      "num_tokens": 62212689.0,
      "step": 27870
    },
    {
      "entropy": 5.346079683303833,
      "epoch": 3.130622788790925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004042128779691368,
      "loss": 4.99,
      "mean_token_accuracy": 0.21255695670843125,
      "num_tokens": 62223771.0,
      "step": 27875
    },
    {
      "entropy": 5.4889294624328615,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004041800077886587,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2127087965607643,
      "num_tokens": 62234137.0,
      "step": 27880
    },
    {
      "entropy": 5.440011358261108,
      "epoch": 3.13174594260684,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040414713349503907,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2135622024536133,
      "num_tokens": 62244978.0,
      "step": 27885
    },
    {
      "entropy": 5.45301923751831,
      "epoch": 3.1323075195147974,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040411425508932457,
      "loss": 5.0967,
      "mean_token_accuracy": 0.20775532722473145,
      "num_tokens": 62257364.0,
      "step": 27890
    },
    {
      "entropy": 5.414375162124633,
      "epoch": 3.132869096422755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040408137257256225,
      "loss": 4.9003,
      "mean_token_accuracy": 0.21896794438362122,
      "num_tokens": 62268135.0,
      "step": 27895
    },
    {
      "entropy": 5.435984563827515,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.125,
      "learning_rate": 0.0004040484859457988,
      "loss": 5.0687,
      "mean_token_accuracy": 0.21572374552488327,
      "num_tokens": 62279096.0,
      "step": 27900
    },
    {
      "entropy": 5.394611978530884,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040401559521008155,
      "loss": 5.0169,
      "mean_token_accuracy": 0.209776808321476,
      "num_tokens": 62289640.0,
      "step": 27905
    },
    {
      "entropy": 5.480691814422608,
      "epoch": 3.1345538271466276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040398270036645775,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20384691506624222,
      "num_tokens": 62301085.0,
      "step": 27910
    },
    {
      "entropy": 5.312471055984497,
      "epoch": 3.135115404054585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004039498014159748,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22294693440198898,
      "num_tokens": 62312667.0,
      "step": 27915
    },
    {
      "entropy": 5.430387258529663,
      "epoch": 3.135676980962543,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040391689835968004,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21135587096214295,
      "num_tokens": 62324640.0,
      "step": 27920
    },
    {
      "entropy": 5.45161247253418,
      "epoch": 3.1362385578705005,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004038839911986213,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21053556948900223,
      "num_tokens": 62335940.0,
      "step": 27925
    },
    {
      "entropy": 5.50938196182251,
      "epoch": 3.1368001347784578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004038510799338462,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21540090292692185,
      "num_tokens": 62346745.0,
      "step": 27930
    },
    {
      "entropy": 5.454267406463623,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004038181645664029,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2086929827928543,
      "num_tokens": 62357694.0,
      "step": 27935
    },
    {
      "entropy": 5.4863440990448,
      "epoch": 3.137923288594373,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20506281703710555,
      "num_tokens": 62370621.0,
      "step": 27940
    },
    {
      "entropy": 5.402229833602905,
      "epoch": 3.1384848655023307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004037523215277033,
      "loss": 5.0299,
      "mean_token_accuracy": 0.21254054009914397,
      "num_tokens": 62381248.0,
      "step": 27945
    },
    {
      "entropy": 5.493071556091309,
      "epoch": 3.139046442410288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004037193938585436,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21032768934965135,
      "num_tokens": 62392102.0,
      "step": 27950
    },
    {
      "entropy": 5.480816268920899,
      "epoch": 3.1396080193182456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004036864620909085,
      "loss": 5.0499,
      "mean_token_accuracy": 0.22029622346162797,
      "num_tokens": 62403254.0,
      "step": 27955
    },
    {
      "entropy": 5.389600467681885,
      "epoch": 3.1401695962262033,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22424625903367995,
      "num_tokens": 62413686.0,
      "step": 27960
    },
    {
      "entropy": 5.424889373779297,
      "epoch": 3.140731173134161,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040362058626440637,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21307966560125352,
      "num_tokens": 62424119.0,
      "step": 27965
    },
    {
      "entropy": 5.385415887832641,
      "epoch": 3.141292750042118,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040358764220763685,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2181812882423401,
      "num_tokens": 62434900.0,
      "step": 27970
    },
    {
      "entropy": 5.379421520233154,
      "epoch": 3.141854326950076,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040355469405658687,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21478084921836854,
      "num_tokens": 62446766.0,
      "step": 27975
    },
    {
      "entropy": 5.382782697677612,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004035217418123056,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2157361015677452,
      "num_tokens": 62457863.0,
      "step": 27980
    },
    {
      "entropy": 5.464255428314209,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040348878547584227,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20278729796409606,
      "num_tokens": 62470380.0,
      "step": 27985
    },
    {
      "entropy": 5.468150854110718,
      "epoch": 3.1435390576739484,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040345582504824624,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21071893125772476,
      "num_tokens": 62482135.0,
      "step": 27990
    },
    {
      "entropy": 5.546390438079834,
      "epoch": 3.144100634581906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004034228605305668,
      "loss": 5.2136,
      "mean_token_accuracy": 0.20365926176309584,
      "num_tokens": 62494409.0,
      "step": 27995
    },
    {
      "entropy": 5.47317476272583,
      "epoch": 3.1446622114898637,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040338989192385377,
      "loss": 5.1237,
      "mean_token_accuracy": 0.2060320720076561,
      "num_tokens": 62506247.0,
      "step": 28000
    },
    {
      "entropy": 5.498948621749878,
      "epoch": 3.145223788397821,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004033569192291567,
      "loss": 5.0946,
      "mean_token_accuracy": 0.19931576102972032,
      "num_tokens": 62517580.0,
      "step": 28005
    },
    {
      "entropy": 5.44220986366272,
      "epoch": 3.1457853653057786,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004033239424475256,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21397986263036728,
      "num_tokens": 62529086.0,
      "step": 28010
    },
    {
      "entropy": 5.375871467590332,
      "epoch": 3.1463469422137362,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040329096158001033,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21499669402837754,
      "num_tokens": 62540381.0,
      "step": 28015
    },
    {
      "entropy": 5.46193413734436,
      "epoch": 3.146908519121694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004032579766276612,
      "loss": 5.06,
      "mean_token_accuracy": 0.20177818238735198,
      "num_tokens": 62551756.0,
      "step": 28020
    },
    {
      "entropy": 5.455435466766358,
      "epoch": 3.147470096029651,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040322498759152824,
      "loss": 5.0838,
      "mean_token_accuracy": 0.2054230034351349,
      "num_tokens": 62563666.0,
      "step": 28025
    },
    {
      "entropy": 5.528929424285889,
      "epoch": 3.1480316729376088,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.1892,
      "mean_token_accuracy": 0.20221561640501023,
      "num_tokens": 62576326.0,
      "step": 28030
    },
    {
      "entropy": 5.457661485671997,
      "epoch": 3.1485932498455664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004031589972721129,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21856063455343247,
      "num_tokens": 62586911.0,
      "step": 28035
    },
    {
      "entropy": 5.349528646469116,
      "epoch": 3.149154826753524,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004031259959909317,
      "loss": 5.0327,
      "mean_token_accuracy": 0.2206551119685173,
      "num_tokens": 62596694.0,
      "step": 28040
    },
    {
      "entropy": 5.444238471984863,
      "epoch": 3.1497164036614813,
      "grad_norm": 1.046875,
      "learning_rate": 0.000403092990630169,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21760139167308806,
      "num_tokens": 62607576.0,
      "step": 28045
    },
    {
      "entropy": 5.407065200805664,
      "epoch": 3.150277980569439,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2165343642234802,
      "num_tokens": 62618015.0,
      "step": 28050
    },
    {
      "entropy": 5.377551555633545,
      "epoch": 3.1508395574773966,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004030269676741032,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21087802797555924,
      "num_tokens": 62629317.0,
      "step": 28055
    },
    {
      "entropy": 5.405185604095459,
      "epoch": 3.1514011343853543,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004029939500809023,
      "loss": 5.055,
      "mean_token_accuracy": 0.21015580147504806,
      "num_tokens": 62640077.0,
      "step": 28060
    },
    {
      "entropy": 5.598571872711181,
      "epoch": 3.1519627112933115,
      "grad_norm": 1.125,
      "learning_rate": 0.0004029609284123243,
      "loss": 5.2353,
      "mean_token_accuracy": 0.2033708930015564,
      "num_tokens": 62652399.0,
      "step": 28065
    },
    {
      "entropy": 5.4816220760345455,
      "epoch": 3.152524288201269,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040292790266942074,
      "loss": 4.993,
      "mean_token_accuracy": 0.21563391983509064,
      "num_tokens": 62663489.0,
      "step": 28070
    },
    {
      "entropy": 5.389800548553467,
      "epoch": 3.153085865109227,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040289487285324324,
      "loss": 5.083,
      "mean_token_accuracy": 0.2093673050403595,
      "num_tokens": 62674975.0,
      "step": 28075
    },
    {
      "entropy": 5.5295600414276125,
      "epoch": 3.153647442017184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20245572179555893,
      "num_tokens": 62686608.0,
      "step": 28080
    },
    {
      "entropy": 5.45195198059082,
      "epoch": 3.1542090189251417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004028288010052728,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21647129356861114,
      "num_tokens": 62697844.0,
      "step": 28085
    },
    {
      "entropy": 5.357226276397705,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004027957589755837,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21761917918920518,
      "num_tokens": 62709092.0,
      "step": 28090
    },
    {
      "entropy": 5.337656545639038,
      "epoch": 3.155332172741057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040276271287682813,
      "loss": 4.9662,
      "mean_token_accuracy": 0.2147550791501999,
      "num_tokens": 62720047.0,
      "step": 28095
    },
    {
      "entropy": 5.467298603057861,
      "epoch": 3.1558937496490143,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040272966271005816,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21456197053194045,
      "num_tokens": 62729952.0,
      "step": 28100
    },
    {
      "entropy": 5.523375558853149,
      "epoch": 3.156455326556972,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040269660847632626,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20704346746206284,
      "num_tokens": 62741961.0,
      "step": 28105
    },
    {
      "entropy": 5.387214803695679,
      "epoch": 3.1570169034649296,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040266355017668475,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21233806014060974,
      "num_tokens": 62753552.0,
      "step": 28110
    },
    {
      "entropy": 5.317467737197876,
      "epoch": 3.1575784803728872,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004026304878121864,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2213462308049202,
      "num_tokens": 62765239.0,
      "step": 28115
    },
    {
      "entropy": 5.325391387939453,
      "epoch": 3.1581400572808445,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040259742138388366,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2130701258778572,
      "num_tokens": 62776975.0,
      "step": 28120
    },
    {
      "entropy": 5.372756767272949,
      "epoch": 3.158701634188802,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040256435089282966,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2153546154499054,
      "num_tokens": 62787435.0,
      "step": 28125
    },
    {
      "entropy": 5.4165397644042965,
      "epoch": 3.15926321109676,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004025312763400772,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21461483389139174,
      "num_tokens": 62799014.0,
      "step": 28130
    },
    {
      "entropy": 5.419067239761352,
      "epoch": 3.1598247880047174,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004024981977266794,
      "loss": 5.0864,
      "mean_token_accuracy": 0.2022975727915764,
      "num_tokens": 62810196.0,
      "step": 28135
    },
    {
      "entropy": 5.340655565261841,
      "epoch": 3.1603863649126747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004024651150536896,
      "loss": 4.9008,
      "mean_token_accuracy": 0.21777146309614182,
      "num_tokens": 62821416.0,
      "step": 28140
    },
    {
      "entropy": 5.396743631362915,
      "epoch": 3.1609479418206323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004024320283221611,
      "loss": 4.9475,
      "mean_token_accuracy": 0.21768246442079545,
      "num_tokens": 62831736.0,
      "step": 28145
    },
    {
      "entropy": 5.3657938003540036,
      "epoch": 3.16150951872859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040239893753314735,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2187177538871765,
      "num_tokens": 62842017.0,
      "step": 28150
    },
    {
      "entropy": 5.482996988296509,
      "epoch": 3.1620710956365476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004023658426877021,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20857616811990737,
      "num_tokens": 62853594.0,
      "step": 28155
    },
    {
      "entropy": 5.396644067764282,
      "epoch": 3.162632672544505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.9648,
      "mean_token_accuracy": 0.21733091175556182,
      "num_tokens": 62864831.0,
      "step": 28160
    },
    {
      "entropy": 5.5280767440795895,
      "epoch": 3.1631942494524625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.117,
      "mean_token_accuracy": 0.21210750192403793,
      "num_tokens": 62877557.0,
      "step": 28165
    },
    {
      "entropy": 5.459180736541748,
      "epoch": 3.16375582636042,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040226653382331484,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21279616057872772,
      "num_tokens": 62889162.0,
      "step": 28170
    },
    {
      "entropy": 5.343330669403076,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000402233422762682,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21946400701999663,
      "num_tokens": 62899376.0,
      "step": 28175
    },
    {
      "entropy": 5.391782569885254,
      "epoch": 3.164878980176335,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.9961,
      "mean_token_accuracy": 0.21644318699836732,
      "num_tokens": 62912448.0,
      "step": 28180
    },
    {
      "entropy": 5.36465015411377,
      "epoch": 3.1654405570842927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004021671884889862,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2164289817214012,
      "num_tokens": 62924021.0,
      "step": 28185
    },
    {
      "entropy": 5.512688493728637,
      "epoch": 3.1660021339922504,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004021340652780321,
      "loss": 5.0701,
      "mean_token_accuracy": 0.2124418243765831,
      "num_tokens": 62935330.0,
      "step": 28190
    },
    {
      "entropy": 5.421286678314209,
      "epoch": 3.1665637109002076,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004021009380190801,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2154337301850319,
      "num_tokens": 62947211.0,
      "step": 28195
    },
    {
      "entropy": 5.501026201248169,
      "epoch": 3.1671252878081653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040206780671318495,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20753404647111892,
      "num_tokens": 62959012.0,
      "step": 28200
    },
    {
      "entropy": 5.466755199432373,
      "epoch": 3.167686864716123,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004020346713614015,
      "loss": 5.1173,
      "mean_token_accuracy": 0.21250320822000504,
      "num_tokens": 62970608.0,
      "step": 28205
    },
    {
      "entropy": 5.524542331695557,
      "epoch": 3.1682484416240806,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040200153196478486,
      "loss": 5.0954,
      "mean_token_accuracy": 0.21454194635152818,
      "num_tokens": 62982613.0,
      "step": 28210
    },
    {
      "entropy": 5.332536411285401,
      "epoch": 3.168810018532038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004019683885243902,
      "loss": 4.9467,
      "mean_token_accuracy": 0.20898162871599196,
      "num_tokens": 62994394.0,
      "step": 28215
    },
    {
      "entropy": 5.41992449760437,
      "epoch": 3.1693715954399955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004019352410412728,
      "loss": 5.1017,
      "mean_token_accuracy": 0.2108816236257553,
      "num_tokens": 63007017.0,
      "step": 28220
    },
    {
      "entropy": 5.470665788650512,
      "epoch": 3.169933172347953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040190208951648807,
      "loss": 5.0216,
      "mean_token_accuracy": 0.20901652425527573,
      "num_tokens": 63018325.0,
      "step": 28225
    },
    {
      "entropy": 5.547572422027588,
      "epoch": 3.170494749255911,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004018689339510916,
      "loss": 5.0979,
      "mean_token_accuracy": 0.2128029376268387,
      "num_tokens": 63030694.0,
      "step": 28230
    },
    {
      "entropy": 5.369363117218017,
      "epoch": 3.171056326163868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040183577434613904,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2123667776584625,
      "num_tokens": 63040889.0,
      "step": 28235
    },
    {
      "entropy": 5.279773235321045,
      "epoch": 3.1716179030718257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004018026107026863,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22588375359773635,
      "num_tokens": 63051385.0,
      "step": 28240
    },
    {
      "entropy": 5.380295610427856,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004017694430217892,
      "loss": 5.0278,
      "mean_token_accuracy": 0.210318686068058,
      "num_tokens": 63061989.0,
      "step": 28245
    },
    {
      "entropy": 5.5026038646697994,
      "epoch": 3.172741056887741,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004017362713045039,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20864275395870208,
      "num_tokens": 63073031.0,
      "step": 28250
    },
    {
      "entropy": 5.411127710342408,
      "epoch": 3.173302633795698,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040170309555188655,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2239553898572922,
      "num_tokens": 63083674.0,
      "step": 28255
    },
    {
      "entropy": 5.386812591552735,
      "epoch": 3.173864210703656,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004016699157649936,
      "loss": 5.0043,
      "mean_token_accuracy": 0.22718640714883803,
      "num_tokens": 63094916.0,
      "step": 28260
    },
    {
      "entropy": 5.410656833648682,
      "epoch": 3.1744257876116135,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20505140572786332,
      "num_tokens": 63106307.0,
      "step": 28265
    },
    {
      "entropy": 5.362738132476807,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040160354409260625,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22006263881921767,
      "num_tokens": 63115871.0,
      "step": 28270
    },
    {
      "entropy": 5.315191221237183,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004015703522092254,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22114246040582658,
      "num_tokens": 63126282.0,
      "step": 28275
    },
    {
      "entropy": 5.346062421798706,
      "epoch": 3.176110518335486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004015371562957954,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21972213983535765,
      "num_tokens": 63138061.0,
      "step": 28280
    },
    {
      "entropy": 5.438321161270141,
      "epoch": 3.1766720952434437,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004015039563533732,
      "loss": 5.0381,
      "mean_token_accuracy": 0.2125880762934685,
      "num_tokens": 63148478.0,
      "step": 28285
    },
    {
      "entropy": 5.4842764377594,
      "epoch": 3.177233672151401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004014707523830161,
      "loss": 5.0646,
      "mean_token_accuracy": 0.214061376452446,
      "num_tokens": 63161205.0,
      "step": 28290
    },
    {
      "entropy": 5.450049638748169,
      "epoch": 3.1777952490593586,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004014375443857811,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21372186094522477,
      "num_tokens": 63173532.0,
      "step": 28295
    },
    {
      "entropy": 5.458025312423706,
      "epoch": 3.1783568259673163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004014043323627257,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20726367086172104,
      "num_tokens": 63185254.0,
      "step": 28300
    },
    {
      "entropy": 5.486159086227417,
      "epoch": 3.178918402875274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004013711163149072,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20449285358190536,
      "num_tokens": 63196481.0,
      "step": 28305
    },
    {
      "entropy": 5.41504921913147,
      "epoch": 3.179479979783231,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040133789624338353,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21987544596195222,
      "num_tokens": 63207411.0,
      "step": 28310
    },
    {
      "entropy": 5.403178310394287,
      "epoch": 3.180041556691189,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040130467214921216,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21374777406454087,
      "num_tokens": 63219097.0,
      "step": 28315
    },
    {
      "entropy": 5.438876056671143,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.03125,
      "learning_rate": 0.000401271444033451,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20942820310592652,
      "num_tokens": 63230530.0,
      "step": 28320
    },
    {
      "entropy": 5.436724281311035,
      "epoch": 3.181164710507104,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004012382118971582,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21910807341337205,
      "num_tokens": 63241918.0,
      "step": 28325
    },
    {
      "entropy": 5.499464797973633,
      "epoch": 3.1817262874150614,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004012049757413917,
      "loss": 5.1157,
      "mean_token_accuracy": 0.20743403881788253,
      "num_tokens": 63253821.0,
      "step": 28330
    },
    {
      "entropy": 5.430166482925415,
      "epoch": 3.182287864323019,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040117173556720976,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21087329387664794,
      "num_tokens": 63265463.0,
      "step": 28335
    },
    {
      "entropy": 5.4027036190032955,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040113849137567094,
      "loss": 4.9556,
      "mean_token_accuracy": 0.21656334549188613,
      "num_tokens": 63276010.0,
      "step": 28340
    },
    {
      "entropy": 5.407425355911255,
      "epoch": 3.1834110181389343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004011052431678336,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21229837387800216,
      "num_tokens": 63287179.0,
      "step": 28345
    },
    {
      "entropy": 5.402667617797851,
      "epoch": 3.1839725950468916,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004010719909447564,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2169952943921089,
      "num_tokens": 63298453.0,
      "step": 28350
    },
    {
      "entropy": 5.405023431777954,
      "epoch": 3.1845341719548492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004010387347074981,
      "loss": 4.9811,
      "mean_token_accuracy": 0.20661901086568832,
      "num_tokens": 63310130.0,
      "step": 28355
    },
    {
      "entropy": 5.410324478149414,
      "epoch": 3.185095748862807,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040100547445711765,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21489404141902924,
      "num_tokens": 63319809.0,
      "step": 28360
    },
    {
      "entropy": 5.399167251586914,
      "epoch": 3.185657325770764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040097221019467393,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21052730828523636,
      "num_tokens": 63331263.0,
      "step": 28365
    },
    {
      "entropy": 5.397555875778198,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2152826175093651,
      "num_tokens": 63341983.0,
      "step": 28370
    },
    {
      "entropy": 5.519799613952637,
      "epoch": 3.1867804795866794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040090566963783375,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21906624883413314,
      "num_tokens": 63352856.0,
      "step": 28375
    },
    {
      "entropy": 5.466302442550659,
      "epoch": 3.187342056494637,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040087239334555603,
      "loss": 5.0669,
      "mean_token_accuracy": 0.20824160873889924,
      "num_tokens": 63364893.0,
      "step": 28380
    },
    {
      "entropy": 5.428958320617676,
      "epoch": 3.1879036334025943,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040083911304545234,
      "loss": 4.966,
      "mean_token_accuracy": 0.21581951081752776,
      "num_tokens": 63375222.0,
      "step": 28385
    },
    {
      "entropy": 5.30457649230957,
      "epoch": 3.188465210310552,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040080582873858254,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2118071958422661,
      "num_tokens": 63387056.0,
      "step": 28390
    },
    {
      "entropy": 5.444035530090332,
      "epoch": 3.1890267872185096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004007725404260064,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20854805558919906,
      "num_tokens": 63397475.0,
      "step": 28395
    },
    {
      "entropy": 5.448030614852906,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004007392481087837,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21646847277879716,
      "num_tokens": 63407920.0,
      "step": 28400
    },
    {
      "entropy": 5.341136693954468,
      "epoch": 3.1901499410344245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004007059517879747,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22834170311689378,
      "num_tokens": 63418537.0,
      "step": 28405
    },
    {
      "entropy": 5.326443529129028,
      "epoch": 3.190711517942382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004006726514646393,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2132164791226387,
      "num_tokens": 63429844.0,
      "step": 28410
    },
    {
      "entropy": 5.4479546546936035,
      "epoch": 3.19127309485034,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040063934713983803,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21333355456590652,
      "num_tokens": 63441417.0,
      "step": 28415
    },
    {
      "entropy": 5.36895523071289,
      "epoch": 3.1918346717582975,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.9031,
      "mean_token_accuracy": 0.21294616013765336,
      "num_tokens": 63451772.0,
      "step": 28420
    },
    {
      "entropy": 5.389113521575927,
      "epoch": 3.1923962486662547,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040057272649007944,
      "loss": 4.9474,
      "mean_token_accuracy": 0.2167169541120529,
      "num_tokens": 63462683.0,
      "step": 28425
    },
    {
      "entropy": 5.464451169967651,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040053941016724326,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21137323826551438,
      "num_tokens": 63473627.0,
      "step": 28430
    },
    {
      "entropy": 5.445599889755249,
      "epoch": 3.19351940248217,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.108,
      "mean_token_accuracy": 0.20769158005714417,
      "num_tokens": 63485554.0,
      "step": 28435
    },
    {
      "entropy": 5.511378192901612,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004004727655309614,
      "loss": 5.0781,
      "mean_token_accuracy": 0.2142342060804367,
      "num_tokens": 63495881.0,
      "step": 28440
    },
    {
      "entropy": 5.483813858032226,
      "epoch": 3.194642556298085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040043943721963766,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2034783437848091,
      "num_tokens": 63506639.0,
      "step": 28445
    },
    {
      "entropy": 5.45351619720459,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.125,
      "learning_rate": 0.00040040610491427357,
      "loss": 5.0708,
      "mean_token_accuracy": 0.208072666823864,
      "num_tokens": 63518482.0,
      "step": 28450
    },
    {
      "entropy": 5.3761622428894045,
      "epoch": 3.1957657101140002,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004003727686159305,
      "loss": 4.9472,
      "mean_token_accuracy": 0.21810166984796525,
      "num_tokens": 63529019.0,
      "step": 28455
    },
    {
      "entropy": 5.421195936203003,
      "epoch": 3.1963272870219575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040033942832566984,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21354663819074632,
      "num_tokens": 63540330.0,
      "step": 28460
    },
    {
      "entropy": 5.344333600997925,
      "epoch": 3.196888863929915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004003060840445531,
      "loss": 4.9904,
      "mean_token_accuracy": 0.20880550742149354,
      "num_tokens": 63552689.0,
      "step": 28465
    },
    {
      "entropy": 5.371374130249023,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004002727357736422,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2187417432665825,
      "num_tokens": 63562954.0,
      "step": 28470
    },
    {
      "entropy": 5.495130205154419,
      "epoch": 3.1980120177458304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040023938351399865,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20913564413785934,
      "num_tokens": 63574697.0,
      "step": 28475
    },
    {
      "entropy": 5.430255174636841,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004002060272666846,
      "loss": 5.0323,
      "mean_token_accuracy": 0.2259202316403389,
      "num_tokens": 63585060.0,
      "step": 28480
    },
    {
      "entropy": 5.416520261764527,
      "epoch": 3.1991351715617453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004001726670327621,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22098149806261064,
      "num_tokens": 63596749.0,
      "step": 28485
    },
    {
      "entropy": 5.348002338409424,
      "epoch": 3.199696748469703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004001393028132933,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2235645219683647,
      "num_tokens": 63607629.0,
      "step": 28490
    },
    {
      "entropy": 5.359988021850586,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040010593460934065,
      "loss": 4.934,
      "mean_token_accuracy": 0.22007778584957122,
      "num_tokens": 63617354.0,
      "step": 28495
    },
    {
      "entropy": 5.426018238067627,
      "epoch": 3.200819902285618,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040007256242196644,
      "loss": 5.0006,
      "mean_token_accuracy": 0.21406710594892503,
      "num_tokens": 63629320.0,
      "step": 28500
    },
    {
      "entropy": 5.448759317398071,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040003918625223335,
      "loss": 5.0446,
      "mean_token_accuracy": 0.20624848902225495,
      "num_tokens": 63640453.0,
      "step": 28505
    },
    {
      "entropy": 5.415848445892334,
      "epoch": 3.201943056101533,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040000580610120406,
      "loss": 4.962,
      "mean_token_accuracy": 0.22090501934289933,
      "num_tokens": 63652484.0,
      "step": 28510
    },
    {
      "entropy": 5.42688021659851,
      "epoch": 3.202504633009491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003999724219699415,
      "loss": 5.0366,
      "mean_token_accuracy": 0.22798023819923402,
      "num_tokens": 63663084.0,
      "step": 28515
    },
    {
      "entropy": 5.386060380935669,
      "epoch": 3.203066209917448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003999390338595085,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21400052905082703,
      "num_tokens": 63674023.0,
      "step": 28520
    },
    {
      "entropy": 5.443847608566284,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039990564177096817,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21281450986862183,
      "num_tokens": 63685370.0,
      "step": 28525
    },
    {
      "entropy": 5.471294164657593,
      "epoch": 3.2041893637333634,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21101570576429368,
      "num_tokens": 63696550.0,
      "step": 28530
    },
    {
      "entropy": 5.5487010955810545,
      "epoch": 3.204750940641321,
      "grad_norm": 1.0,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.1082,
      "mean_token_accuracy": 0.211043144762516,
      "num_tokens": 63708040.0,
      "step": 28535
    },
    {
      "entropy": 5.440803670883179,
      "epoch": 3.2053125175492783,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003998054416473363,
      "loss": 5.0507,
      "mean_token_accuracy": 0.208536496758461,
      "num_tokens": 63719704.0,
      "step": 28540
    },
    {
      "entropy": 5.387283086776733,
      "epoch": 3.205874094457236,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039977203365700026,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21241639107465743,
      "num_tokens": 63730111.0,
      "step": 28545
    },
    {
      "entropy": 5.467095804214478,
      "epoch": 3.2064356713651936,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003997386216938743,
      "loss": 5.007,
      "mean_token_accuracy": 0.22357982099056245,
      "num_tokens": 63741054.0,
      "step": 28550
    },
    {
      "entropy": 5.4580206871032715,
      "epoch": 3.206997248273151,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039970520575902216,
      "loss": 5.028,
      "mean_token_accuracy": 0.21286799460649491,
      "num_tokens": 63752598.0,
      "step": 28555
    },
    {
      "entropy": 5.457810735702514,
      "epoch": 3.2075588251811085,
      "grad_norm": 1.0625,
      "learning_rate": 0.000399671785853508,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21802017390727996,
      "num_tokens": 63763927.0,
      "step": 28560
    },
    {
      "entropy": 5.437067461013794,
      "epoch": 3.208120402089066,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039963836197839584,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22059844434261322,
      "num_tokens": 63775072.0,
      "step": 28565
    },
    {
      "entropy": 5.411277484893799,
      "epoch": 3.208681978997024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039960493413474977,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21490006595849992,
      "num_tokens": 63786044.0,
      "step": 28570
    },
    {
      "entropy": 5.4001145362854,
      "epoch": 3.209243555904981,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039957150232363433,
      "loss": 5.0151,
      "mean_token_accuracy": 0.2170064389705658,
      "num_tokens": 63797889.0,
      "step": 28575
    },
    {
      "entropy": 5.463947200775147,
      "epoch": 3.2098051328129387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000399538066546114,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20948139131069182,
      "num_tokens": 63809567.0,
      "step": 28580
    },
    {
      "entropy": 5.498517942428589,
      "epoch": 3.2103667097208963,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039950462680325336,
      "loss": 5.0362,
      "mean_token_accuracy": 0.2127929225564003,
      "num_tokens": 63819617.0,
      "step": 28585
    },
    {
      "entropy": 5.516414737701416,
      "epoch": 3.210928286628854,
      "grad_norm": 1.078125,
      "learning_rate": 0.000399471183096117,
      "loss": 5.1866,
      "mean_token_accuracy": 0.20432811975479126,
      "num_tokens": 63831535.0,
      "step": 28590
    },
    {
      "entropy": 5.485160684585571,
      "epoch": 3.211489863536811,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003994377354257698,
      "loss": 5.1036,
      "mean_token_accuracy": 0.2115199163556099,
      "num_tokens": 63842629.0,
      "step": 28595
    },
    {
      "entropy": 5.53062334060669,
      "epoch": 3.212051440444769,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039940428379327703,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21195934265851973,
      "num_tokens": 63853677.0,
      "step": 28600
    },
    {
      "entropy": 5.485795164108277,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003993708281997036,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20449895709753035,
      "num_tokens": 63865536.0,
      "step": 28605
    },
    {
      "entropy": 5.463983392715454,
      "epoch": 3.213174594260684,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003993373686461147,
      "loss": 5.0165,
      "mean_token_accuracy": 0.20770811587572097,
      "num_tokens": 63875849.0,
      "step": 28610
    },
    {
      "entropy": 5.496801424026489,
      "epoch": 3.2137361711686414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003993039051335758,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21616479009389877,
      "num_tokens": 63887222.0,
      "step": 28615
    },
    {
      "entropy": 5.435396957397461,
      "epoch": 3.214297748076599,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003992704376631523,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21589087396860124,
      "num_tokens": 63897969.0,
      "step": 28620
    },
    {
      "entropy": 5.526981592178345,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003992369662359099,
      "loss": 5.1367,
      "mean_token_accuracy": 0.2170506700873375,
      "num_tokens": 63908913.0,
      "step": 28625
    },
    {
      "entropy": 5.44479455947876,
      "epoch": 3.2154209018925144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2119178518652916,
      "num_tokens": 63919269.0,
      "step": 28630
    },
    {
      "entropy": 5.368867301940918,
      "epoch": 3.2159824788004716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039917001151523144,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2100215435028076,
      "num_tokens": 63930624.0,
      "step": 28635
    },
    {
      "entropy": 5.3790723323822025,
      "epoch": 3.2165440557084293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003991365282239272,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21249576061964034,
      "num_tokens": 63941571.0,
      "step": 28640
    },
    {
      "entropy": 5.509418296813965,
      "epoch": 3.217105632616387,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039910304098006775,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21433897614479064,
      "num_tokens": 63953221.0,
      "step": 28645
    },
    {
      "entropy": 5.468585777282715,
      "epoch": 3.217667209524344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003990695497847193,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2169075906276703,
      "num_tokens": 63965076.0,
      "step": 28650
    },
    {
      "entropy": 5.419501495361328,
      "epoch": 3.218228786432302,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003990360546389484,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22324344217777253,
      "num_tokens": 63976601.0,
      "step": 28655
    },
    {
      "entropy": 5.356812334060669,
      "epoch": 3.2187903633402595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039900255554382126,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21781306564807892,
      "num_tokens": 63987757.0,
      "step": 28660
    },
    {
      "entropy": 5.4383667469024655,
      "epoch": 3.219351940248217,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039896905250040473,
      "loss": 5.0643,
      "mean_token_accuracy": 0.21138158440589905,
      "num_tokens": 64000104.0,
      "step": 28665
    },
    {
      "entropy": 5.413522863388062,
      "epoch": 3.2199135171561744,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039893554550976545,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22130339592695236,
      "num_tokens": 64010773.0,
      "step": 28670
    },
    {
      "entropy": 5.416013956069946,
      "epoch": 3.220475094064132,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003989020345729705,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21477470248937608,
      "num_tokens": 64022269.0,
      "step": 28675
    },
    {
      "entropy": 5.430961084365845,
      "epoch": 3.2210366709720897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039886851969108653,
      "loss": 4.9969,
      "mean_token_accuracy": 0.212725268304348,
      "num_tokens": 64033622.0,
      "step": 28680
    },
    {
      "entropy": 5.360050821304322,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039883500086518086,
      "loss": 5.0111,
      "mean_token_accuracy": 0.2167672857642174,
      "num_tokens": 64044243.0,
      "step": 28685
    },
    {
      "entropy": 5.4628503799438475,
      "epoch": 3.2221598247880046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039880147809632084,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21339345276355742,
      "num_tokens": 64054452.0,
      "step": 28690
    },
    {
      "entropy": 5.462273693084716,
      "epoch": 3.222721401695962,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039876795138557363,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21345402747392656,
      "num_tokens": 64065743.0,
      "step": 28695
    },
    {
      "entropy": 5.488807153701782,
      "epoch": 3.22328297860392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003987344207340069,
      "loss": 5.0578,
      "mean_token_accuracy": 0.2094692051410675,
      "num_tokens": 64077359.0,
      "step": 28700
    },
    {
      "entropy": 5.402265930175782,
      "epoch": 3.2238445555118775,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039870088614268824,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21857446879148484,
      "num_tokens": 64088281.0,
      "step": 28705
    },
    {
      "entropy": 5.40048041343689,
      "epoch": 3.2244061324198348,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039866734761268534,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21073488593101503,
      "num_tokens": 64099600.0,
      "step": 28710
    },
    {
      "entropy": 5.514808273315429,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039863380514506614,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2120948001742363,
      "num_tokens": 64110381.0,
      "step": 28715
    },
    {
      "entropy": 5.450271797180176,
      "epoch": 3.22552928623575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039860025874089865,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2055472493171692,
      "num_tokens": 64120878.0,
      "step": 28720
    },
    {
      "entropy": 5.315389633178711,
      "epoch": 3.2260908631437077,
      "grad_norm": 1.140625,
      "learning_rate": 0.000398566708401251,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2273039072751999,
      "num_tokens": 64131427.0,
      "step": 28725
    },
    {
      "entropy": 5.428261184692383,
      "epoch": 3.226652440051665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039853315412719144,
      "loss": 5.0819,
      "mean_token_accuracy": 0.21062883734703064,
      "num_tokens": 64142833.0,
      "step": 28730
    },
    {
      "entropy": 5.332300424575806,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003984995959197883,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2167263612151146,
      "num_tokens": 64152405.0,
      "step": 28735
    },
    {
      "entropy": 5.420177316665649,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.949,
      "mean_token_accuracy": 0.21374204754829407,
      "num_tokens": 64164398.0,
      "step": 28740
    },
    {
      "entropy": 5.425990772247315,
      "epoch": 3.2283371707755375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039843246770922557,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20983479619026185,
      "num_tokens": 64175313.0,
      "step": 28745
    },
    {
      "entropy": 5.331839752197266,
      "epoch": 3.228898747683495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003983988977082033,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2178475081920624,
      "num_tokens": 64185866.0,
      "step": 28750
    },
    {
      "entropy": 5.452063703536988,
      "epoch": 3.229460324591453,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039836532377811237,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21540938913822175,
      "num_tokens": 64197126.0,
      "step": 28755
    },
    {
      "entropy": 5.377417135238647,
      "epoch": 3.2300219014994105,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003983317459200217,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2120041221380234,
      "num_tokens": 64209271.0,
      "step": 28760
    },
    {
      "entropy": 5.465100288391113,
      "epoch": 3.2305834784073677,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039829816413500035,
      "loss": 5.0948,
      "mean_token_accuracy": 0.21517219990491868,
      "num_tokens": 64221200.0,
      "step": 28765
    },
    {
      "entropy": 5.46867938041687,
      "epoch": 3.2311450553153254,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003982645784241176,
      "loss": 5.055,
      "mean_token_accuracy": 0.20714413076639177,
      "num_tokens": 64230835.0,
      "step": 28770
    },
    {
      "entropy": 5.470195627212524,
      "epoch": 3.231706632223283,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003982309887884429,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22255949229001998,
      "num_tokens": 64242243.0,
      "step": 28775
    },
    {
      "entropy": 5.466567754745483,
      "epoch": 3.2322682091312407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.9917,
      "mean_token_accuracy": 0.21328160762786866,
      "num_tokens": 64253362.0,
      "step": 28780
    },
    {
      "entropy": 5.40196590423584,
      "epoch": 3.232829786039198,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039816379774699573,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21590474396944045,
      "num_tokens": 64264462.0,
      "step": 28785
    },
    {
      "entropy": 5.3532859802246096,
      "epoch": 3.2333913629471556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039813019634336264,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21418488025665283,
      "num_tokens": 64275153.0,
      "step": 28790
    },
    {
      "entropy": 5.4002691268920895,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003980965910192163,
      "loss": 4.9611,
      "mean_token_accuracy": 0.2228725418448448,
      "num_tokens": 64288082.0,
      "step": 28795
    },
    {
      "entropy": 5.420742845535278,
      "epoch": 3.234514516763071,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039806298177562675,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21591437757015228,
      "num_tokens": 64298872.0,
      "step": 28800
    },
    {
      "entropy": 5.496512985229492,
      "epoch": 3.235076093671028,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039802936861366416,
      "loss": 5.1263,
      "mean_token_accuracy": 0.205428446829319,
      "num_tokens": 64310245.0,
      "step": 28805
    },
    {
      "entropy": 5.508009815216065,
      "epoch": 3.2356376705789858,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003979957515343988,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2098827213048935,
      "num_tokens": 64323681.0,
      "step": 28810
    },
    {
      "entropy": 5.382147169113159,
      "epoch": 3.2361992474869434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000397962130538901,
      "loss": 5.021,
      "mean_token_accuracy": 0.21670915335416793,
      "num_tokens": 64335265.0,
      "step": 28815
    },
    {
      "entropy": 5.507946348190307,
      "epoch": 3.236760824394901,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039792850562824124,
      "loss": 5.0811,
      "mean_token_accuracy": 0.21099757999181748,
      "num_tokens": 64346279.0,
      "step": 28820
    },
    {
      "entropy": 5.4747072696685795,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003978948768034902,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21896374374628066,
      "num_tokens": 64357511.0,
      "step": 28825
    },
    {
      "entropy": 5.4181290626525875,
      "epoch": 3.237883978210816,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039786124406571854,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2107859045267105,
      "num_tokens": 64367863.0,
      "step": 28830
    },
    {
      "entropy": 5.382227087020874,
      "epoch": 3.2384455551187736,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039782760741599734,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2152082607150078,
      "num_tokens": 64379746.0,
      "step": 28835
    },
    {
      "entropy": 5.395762491226196,
      "epoch": 3.239007132026731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003977939668553973,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22157296538352966,
      "num_tokens": 64390370.0,
      "step": 28840
    },
    {
      "entropy": 5.49207763671875,
      "epoch": 3.2395687089346885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039776032238498995,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20772936642169954,
      "num_tokens": 64402620.0,
      "step": 28845
    },
    {
      "entropy": 5.479757261276245,
      "epoch": 3.240130285842646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003977266740058462,
      "loss": 5.029,
      "mean_token_accuracy": 0.22012848109006883,
      "num_tokens": 64412969.0,
      "step": 28850
    },
    {
      "entropy": 5.440990877151489,
      "epoch": 3.240691862750604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003976930217190376,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21169232875108718,
      "num_tokens": 64423679.0,
      "step": 28855
    },
    {
      "entropy": 5.461258888244629,
      "epoch": 3.241253439658561,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003976593655256356,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21258900165557862,
      "num_tokens": 64436124.0,
      "step": 28860
    },
    {
      "entropy": 5.4018639087677,
      "epoch": 3.2418150165665187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21532561630010605,
      "num_tokens": 64446494.0,
      "step": 28865
    },
    {
      "entropy": 5.413084030151367,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.109375,
      "learning_rate": 0.000397592041423338,
      "loss": 4.9318,
      "mean_token_accuracy": 0.21794043928384782,
      "num_tokens": 64456152.0,
      "step": 28870
    },
    {
      "entropy": 5.463778972625732,
      "epoch": 3.242938170382434,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039755837351658605,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21271355599164962,
      "num_tokens": 64468004.0,
      "step": 28875
    },
    {
      "entropy": 5.416790628433228,
      "epoch": 3.2434997472903913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003975247017075279,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21852547228336333,
      "num_tokens": 64478506.0,
      "step": 28880
    },
    {
      "entropy": 5.373204898834229,
      "epoch": 3.244061324198349,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003974910259972358,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2176134929060936,
      "num_tokens": 64489379.0,
      "step": 28885
    },
    {
      "entropy": 5.424749422073364,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003974573463867819,
      "loss": 5.0477,
      "mean_token_accuracy": 0.22029167413711548,
      "num_tokens": 64499500.0,
      "step": 28890
    },
    {
      "entropy": 5.3508507251739506,
      "epoch": 3.2451844780142642,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039742366287723855,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22550860345363616,
      "num_tokens": 64510869.0,
      "step": 28895
    },
    {
      "entropy": 5.3724846839904785,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039738997546967834,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20674297958612442,
      "num_tokens": 64521581.0,
      "step": 28900
    },
    {
      "entropy": 5.480036640167237,
      "epoch": 3.246307631830179,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039735628416517394,
      "loss": 5.1048,
      "mean_token_accuracy": 0.21076753586530686,
      "num_tokens": 64532436.0,
      "step": 28905
    },
    {
      "entropy": 5.494428539276123,
      "epoch": 3.246869208738137,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039732258896479786,
      "loss": 5.047,
      "mean_token_accuracy": 0.21273422837257386,
      "num_tokens": 64543546.0,
      "step": 28910
    },
    {
      "entropy": 5.544923210144043,
      "epoch": 3.2474307856460944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003972888898696231,
      "loss": 5.1616,
      "mean_token_accuracy": 0.20611661225557326,
      "num_tokens": 64554964.0,
      "step": 28915
    },
    {
      "entropy": 5.473891544342041,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039725518688072276,
      "loss": 5.011,
      "mean_token_accuracy": 0.21139386147260666,
      "num_tokens": 64564082.0,
      "step": 28920
    },
    {
      "entropy": 5.446129560470581,
      "epoch": 3.2485539394620093,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039722147999916976,
      "loss": 5.1062,
      "mean_token_accuracy": 0.2097248375415802,
      "num_tokens": 64575377.0,
      "step": 28925
    },
    {
      "entropy": 5.445741748809814,
      "epoch": 3.249115516369967,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003971877692260375,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2145071431994438,
      "num_tokens": 64587686.0,
      "step": 28930
    },
    {
      "entropy": 5.426993751525879,
      "epoch": 3.249677093277924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039715405456239924,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21313527822494507,
      "num_tokens": 64598501.0,
      "step": 28935
    },
    {
      "entropy": 5.4173805713653564,
      "epoch": 3.250238670185882,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003971203360093286,
      "loss": 4.9899,
      "mean_token_accuracy": 0.21616103500127792,
      "num_tokens": 64609341.0,
      "step": 28940
    },
    {
      "entropy": 5.493397808074951,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039708661356789896,
      "loss": 5.0741,
      "mean_token_accuracy": 0.21159220933914186,
      "num_tokens": 64621613.0,
      "step": 28945
    },
    {
      "entropy": 5.354087972640992,
      "epoch": 3.251361824001797,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039705288723918415,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21282264441251755,
      "num_tokens": 64633358.0,
      "step": 28950
    },
    {
      "entropy": 5.378477716445923,
      "epoch": 3.2519234009097544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.961,
      "mean_token_accuracy": 0.21354991495609282,
      "num_tokens": 64644799.0,
      "step": 28955
    },
    {
      "entropy": 5.395093011856079,
      "epoch": 3.252484977817712,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000396985422924195,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21939303874969482,
      "num_tokens": 64657028.0,
      "step": 28960
    },
    {
      "entropy": 5.5271824359893795,
      "epoch": 3.2530465547256697,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039695168494006857,
      "loss": 5.1398,
      "mean_token_accuracy": 0.20488000959157943,
      "num_tokens": 64668572.0,
      "step": 28965
    },
    {
      "entropy": 5.322952699661255,
      "epoch": 3.2536081316336274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003969179430729531,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21910068839788438,
      "num_tokens": 64679168.0,
      "step": 28970
    },
    {
      "entropy": 5.463303422927856,
      "epoch": 3.2541697085415846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003968841973239231,
      "loss": 5.0782,
      "mean_token_accuracy": 0.2083178475499153,
      "num_tokens": 64691571.0,
      "step": 28975
    },
    {
      "entropy": 5.486023235321045,
      "epoch": 3.2547312854495423,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000396850447694053,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20937778800725937,
      "num_tokens": 64703659.0,
      "step": 28980
    },
    {
      "entropy": 5.434529113769531,
      "epoch": 3.2552928623575,
      "grad_norm": 1.25,
      "learning_rate": 0.0003968166941844174,
      "loss": 4.9391,
      "mean_token_accuracy": 0.22553416043519975,
      "num_tokens": 64715121.0,
      "step": 28985
    },
    {
      "entropy": 5.411077213287354,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.140625,
      "learning_rate": 0.000396782936796091,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20817241966724395,
      "num_tokens": 64726289.0,
      "step": 28990
    },
    {
      "entropy": 5.39320936203003,
      "epoch": 3.256416016173415,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003967491755301487,
      "loss": 4.974,
      "mean_token_accuracy": 0.21612541228532792,
      "num_tokens": 64737378.0,
      "step": 28995
    },
    {
      "entropy": 5.516737794876098,
      "epoch": 3.2569775930813725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.1936,
      "mean_token_accuracy": 0.2053195580840111,
      "num_tokens": 64749203.0,
      "step": 29000
    },
    {
      "entropy": 5.374837875366211,
      "epoch": 3.25753916998933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039668164136971615,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21770531833171844,
      "num_tokens": 64759514.0,
      "step": 29005
    },
    {
      "entropy": 5.411973905563355,
      "epoch": 3.258100746897288,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003966478684773763,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21819112747907637,
      "num_tokens": 64769370.0,
      "step": 29010
    },
    {
      "entropy": 5.400743389129639,
      "epoch": 3.258662323805245,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039661409171172117,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21682276129722594,
      "num_tokens": 64779930.0,
      "step": 29015
    },
    {
      "entropy": 5.428154039382934,
      "epoch": 3.2592239007132027,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039658031107382615,
      "loss": 5.0663,
      "mean_token_accuracy": 0.20910998284816742,
      "num_tokens": 64791726.0,
      "step": 29020
    },
    {
      "entropy": 5.348386192321778,
      "epoch": 3.2597854776211603,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039654652656476695,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22944832891225814,
      "num_tokens": 64802778.0,
      "step": 29025
    },
    {
      "entropy": 5.378180646896363,
      "epoch": 3.2603470545291175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003965127381856191,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20952530950307846,
      "num_tokens": 64814664.0,
      "step": 29030
    },
    {
      "entropy": 5.404183340072632,
      "epoch": 3.260908631437075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039647894593745865,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21753921061754228,
      "num_tokens": 64825775.0,
      "step": 29035
    },
    {
      "entropy": 5.497480630874634,
      "epoch": 3.261470208345033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039644514982136136,
      "loss": 5.1351,
      "mean_token_accuracy": 0.2103094592690468,
      "num_tokens": 64837376.0,
      "step": 29040
    },
    {
      "entropy": 5.507981824874878,
      "epoch": 3.2620317852529905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21178942322731018,
      "num_tokens": 64847941.0,
      "step": 29045
    },
    {
      "entropy": 5.506442642211914,
      "epoch": 3.2625933621609478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000396377545989661,
      "loss": 5.1067,
      "mean_token_accuracy": 0.2136112332344055,
      "num_tokens": 64858793.0,
      "step": 29050
    },
    {
      "entropy": 5.436602306365967,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039634373827621047,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21404997110366822,
      "num_tokens": 64870233.0,
      "step": 29055
    },
    {
      "entropy": 5.405445528030396,
      "epoch": 3.263716515976863,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21553783118724823,
      "num_tokens": 64881875.0,
      "step": 29060
    },
    {
      "entropy": 5.372291946411133,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039627611125949086,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21904713958501815,
      "num_tokens": 64892837.0,
      "step": 29065
    },
    {
      "entropy": 5.5447532653808596,
      "epoch": 3.264839669792778,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.1125,
      "mean_token_accuracy": 0.2083963230252266,
      "num_tokens": 64904773.0,
      "step": 29070
    },
    {
      "entropy": 5.496504592895508,
      "epoch": 3.2654012467007356,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039620846879685767,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21139702200889587,
      "num_tokens": 64916596.0,
      "step": 29075
    },
    {
      "entropy": 5.430698108673096,
      "epoch": 3.2659628236086933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039617464177601564,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21089435815811158,
      "num_tokens": 64927736.0,
      "step": 29080
    },
    {
      "entropy": 5.457288026809692,
      "epoch": 3.266524400516651,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20659130662679673,
      "num_tokens": 64940426.0,
      "step": 29085
    },
    {
      "entropy": 5.367679309844971,
      "epoch": 3.267085977424608,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039610697616066604,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21906974017620087,
      "num_tokens": 64951034.0,
      "step": 29090
    },
    {
      "entropy": 5.5262775897979735,
      "epoch": 3.267647554332566,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039607313756831296,
      "loss": 5.0708,
      "mean_token_accuracy": 0.20957130044698716,
      "num_tokens": 64962900.0,
      "step": 29095
    },
    {
      "entropy": 5.4312803745269775,
      "epoch": 3.2682091312405235,
      "grad_norm": 1.125,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2231535479426384,
      "num_tokens": 64974820.0,
      "step": 29100
    },
    {
      "entropy": 5.328769159317017,
      "epoch": 3.268770708148481,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21337043792009353,
      "num_tokens": 64985964.0,
      "step": 29105
    },
    {
      "entropy": 5.377073907852173,
      "epoch": 3.2693322850564384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039597159866546996,
      "loss": 4.9611,
      "mean_token_accuracy": 0.21359928101301193,
      "num_tokens": 64996772.0,
      "step": 29110
    },
    {
      "entropy": 5.369845390319824,
      "epoch": 3.269893861964396,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003959377446595201,
      "loss": 4.9248,
      "mean_token_accuracy": 0.21512921303510665,
      "num_tokens": 65007296.0,
      "step": 29115
    },
    {
      "entropy": 5.418950319290161,
      "epoch": 3.2704554388723537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039590388680286543,
      "loss": 4.9673,
      "mean_token_accuracy": 0.22312796860933304,
      "num_tokens": 65018384.0,
      "step": 29120
    },
    {
      "entropy": 5.392879247665405,
      "epoch": 3.271017015780311,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003958700250965842,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22160020917654039,
      "num_tokens": 65029042.0,
      "step": 29125
    },
    {
      "entropy": 5.4121901512146,
      "epoch": 3.2715785926882686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003958361595417544,
      "loss": 5.0311,
      "mean_token_accuracy": 0.21048536896705627,
      "num_tokens": 65040792.0,
      "step": 29130
    },
    {
      "entropy": 5.420088768005371,
      "epoch": 3.2721401695962262,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003958022901394545,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22813891768455505,
      "num_tokens": 65051702.0,
      "step": 29135
    },
    {
      "entropy": 5.390988111495972,
      "epoch": 3.272701746504184,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003957684168907628,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21755065619945527,
      "num_tokens": 65063288.0,
      "step": 29140
    },
    {
      "entropy": 5.4745889663696286,
      "epoch": 3.273263323412141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039573453979675793,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20115756541490554,
      "num_tokens": 65074913.0,
      "step": 29145
    },
    {
      "entropy": 5.414552497863769,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039570065885851844,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2115033373236656,
      "num_tokens": 65085327.0,
      "step": 29150
    },
    {
      "entropy": 5.42893795967102,
      "epoch": 3.2743864772280564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003956667740771231,
      "loss": 4.98,
      "mean_token_accuracy": 0.21657072603702546,
      "num_tokens": 65095709.0,
      "step": 29155
    },
    {
      "entropy": 5.512595176696777,
      "epoch": 3.274948054136014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003956328854536509,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21554239690303803,
      "num_tokens": 65106786.0,
      "step": 29160
    },
    {
      "entropy": 5.4029261589050295,
      "epoch": 3.2755096310439713,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039559899298918094,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21302748173475267,
      "num_tokens": 65117506.0,
      "step": 29165
    },
    {
      "entropy": 5.3112856388092045,
      "epoch": 3.276071207951929,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.8779,
      "mean_token_accuracy": 0.22637743949890138,
      "num_tokens": 65127830.0,
      "step": 29170
    },
    {
      "entropy": 5.395195007324219,
      "epoch": 3.2766327848598866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000395531196541564,
      "loss": 5.0588,
      "mean_token_accuracy": 0.21388299465179444,
      "num_tokens": 65139562.0,
      "step": 29175
    },
    {
      "entropy": 5.449921035766602,
      "epoch": 3.2771943617678443,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039549729256057583,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21650593727827072,
      "num_tokens": 65150799.0,
      "step": 29180
    },
    {
      "entropy": 5.416740226745605,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000395463384742907,
      "loss": 5.01,
      "mean_token_accuracy": 0.21392360478639602,
      "num_tokens": 65161458.0,
      "step": 29185
    },
    {
      "entropy": 5.39469952583313,
      "epoch": 3.278317515583759,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039542947308963733,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21742458790540695,
      "num_tokens": 65172098.0,
      "step": 29190
    },
    {
      "entropy": 5.473291063308716,
      "epoch": 3.278879092491717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003953955576018466,
      "loss": 5.0318,
      "mean_token_accuracy": 0.213050477206707,
      "num_tokens": 65183668.0,
      "step": 29195
    },
    {
      "entropy": 5.406983470916748,
      "epoch": 3.2794406693996745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003953616382806145,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2193588897585869,
      "num_tokens": 65193920.0,
      "step": 29200
    },
    {
      "entropy": 5.43871169090271,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039532771512702124,
      "loss": 5.0513,
      "mean_token_accuracy": 0.2135640174150467,
      "num_tokens": 65203397.0,
      "step": 29205
    },
    {
      "entropy": 5.383724594116211,
      "epoch": 3.2805638232155894,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003952937881421468,
      "loss": 4.9653,
      "mean_token_accuracy": 0.216984124481678,
      "num_tokens": 65213821.0,
      "step": 29210
    },
    {
      "entropy": 5.3553437232971195,
      "epoch": 3.281125400123547,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21935902982950212,
      "num_tokens": 65224034.0,
      "step": 29215
    },
    {
      "entropy": 5.344850635528564,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003952259226828759,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21396186351776122,
      "num_tokens": 65234879.0,
      "step": 29220
    },
    {
      "entropy": 5.439653730392456,
      "epoch": 3.282248553939462,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039519198421064005,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21850355565547944,
      "num_tokens": 65246530.0,
      "step": 29225
    },
    {
      "entropy": 5.374012136459351,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039515804191144496,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21915081590414048,
      "num_tokens": 65258102.0,
      "step": 29230
    },
    {
      "entropy": 5.397857332229615,
      "epoch": 3.2833717077553772,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039512409578637114,
      "loss": 4.9372,
      "mean_token_accuracy": 0.22472615242004396,
      "num_tokens": 65269396.0,
      "step": 29235
    },
    {
      "entropy": 5.366357135772705,
      "epoch": 3.2839332846633345,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039509014583649947,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2247446969151497,
      "num_tokens": 65280004.0,
      "step": 29240
    },
    {
      "entropy": 5.405739498138428,
      "epoch": 3.284494861571292,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039505619206291106,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21743328720331193,
      "num_tokens": 65291564.0,
      "step": 29245
    },
    {
      "entropy": 5.371134805679321,
      "epoch": 3.2850564384792498,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22021564841270447,
      "num_tokens": 65303032.0,
      "step": 29250
    },
    {
      "entropy": 5.348583507537842,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003949882730489082,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22286177575588226,
      "num_tokens": 65313658.0,
      "step": 29255
    },
    {
      "entropy": 5.533181238174438,
      "epoch": 3.2861795922951647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039495430781065645,
      "loss": 5.102,
      "mean_token_accuracy": 0.21376049667596816,
      "num_tokens": 65324593.0,
      "step": 29260
    },
    {
      "entropy": 5.481296730041504,
      "epoch": 3.2867411692031223,
      "grad_norm": 1.234375,
      "learning_rate": 0.000394920338753013,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21674527525901793,
      "num_tokens": 65335059.0,
      "step": 29265
    },
    {
      "entropy": 5.520305919647217,
      "epoch": 3.28730274611108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003948863658770594,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20930522084236144,
      "num_tokens": 65347082.0,
      "step": 29270
    },
    {
      "entropy": 5.3251735210418705,
      "epoch": 3.2878643230190376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2154395505785942,
      "num_tokens": 65358644.0,
      "step": 29275
    },
    {
      "entropy": 5.4165442943572994,
      "epoch": 3.288425899926995,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21540917456150055,
      "num_tokens": 65369652.0,
      "step": 29280
    },
    {
      "entropy": 5.435080242156983,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22125717997550964,
      "num_tokens": 65380196.0,
      "step": 29285
    },
    {
      "entropy": 5.396737098693848,
      "epoch": 3.28954905374291,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039475043621178005,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2179188013076782,
      "num_tokens": 65391135.0,
      "step": 29290
    },
    {
      "entropy": 5.491714715957642,
      "epoch": 3.290110630650868,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003947164442605041,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21185840368270875,
      "num_tokens": 65402330.0,
      "step": 29295
    },
    {
      "entropy": 5.428884887695313,
      "epoch": 3.290672207558825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003946824484974101,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2130390226840973,
      "num_tokens": 65414111.0,
      "step": 29300
    },
    {
      "entropy": 5.439313793182373,
      "epoch": 3.2912337844667827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003946484489235805,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2157433271408081,
      "num_tokens": 65424760.0,
      "step": 29305
    },
    {
      "entropy": 5.35621132850647,
      "epoch": 3.2917953613747404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039461444554009787,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21953096091747284,
      "num_tokens": 65435213.0,
      "step": 29310
    },
    {
      "entropy": 5.332256317138672,
      "epoch": 3.2923569382826976,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000394580438348045,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21234426498413086,
      "num_tokens": 65446141.0,
      "step": 29315
    },
    {
      "entropy": 5.438039827346802,
      "epoch": 3.2929185151906553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039454642734850447,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21504078805446625,
      "num_tokens": 65457634.0,
      "step": 29320
    },
    {
      "entropy": 5.433530235290528,
      "epoch": 3.293480092098613,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003945124125425594,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21185438334941864,
      "num_tokens": 65468457.0,
      "step": 29325
    },
    {
      "entropy": 5.484934282302857,
      "epoch": 3.2940416690065706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003944783939312928,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21652575731277465,
      "num_tokens": 65479938.0,
      "step": 29330
    },
    {
      "entropy": 5.4590747356414795,
      "epoch": 3.294603245914528,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003944443715157877,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20421258509159088,
      "num_tokens": 65492030.0,
      "step": 29335
    },
    {
      "entropy": 5.510568046569825,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003944103452971275,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2139256328344345,
      "num_tokens": 65503385.0,
      "step": 29340
    },
    {
      "entropy": 5.3607734680175785,
      "epoch": 3.295726399730443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2229817256331444,
      "num_tokens": 65514027.0,
      "step": 29345
    },
    {
      "entropy": 5.373022937774659,
      "epoch": 3.296287976638401,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039434228145467554,
      "loss": 4.937,
      "mean_token_accuracy": 0.21461528390645981,
      "num_tokens": 65525653.0,
      "step": 29350
    },
    {
      "entropy": 5.2895488262176515,
      "epoch": 3.296849553546358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000394308243833051,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2191789746284485,
      "num_tokens": 65536040.0,
      "step": 29355
    },
    {
      "entropy": 5.5137265682220455,
      "epoch": 3.2974111304543157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039427420241260557,
      "loss": 5.1132,
      "mean_token_accuracy": 0.20921690315008162,
      "num_tokens": 65547505.0,
      "step": 29360
    },
    {
      "entropy": 5.490051746368408,
      "epoch": 3.2979727073622733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21897171884775163,
      "num_tokens": 65557565.0,
      "step": 29365
    },
    {
      "entropy": 5.383626174926758,
      "epoch": 3.298534284270231,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039420610817958836,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21506858766078948,
      "num_tokens": 65568729.0,
      "step": 29370
    },
    {
      "entropy": 5.423231363296509,
      "epoch": 3.299095861178188,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003941720553691846,
      "loss": 5.108,
      "mean_token_accuracy": 0.20944231152534484,
      "num_tokens": 65580664.0,
      "step": 29375
    },
    {
      "entropy": 5.356070184707642,
      "epoch": 3.299657438086146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003941379987642963,
      "loss": 4.8888,
      "mean_token_accuracy": 0.22268005907535554,
      "num_tokens": 65591348.0,
      "step": 29380
    },
    {
      "entropy": 5.458045768737793,
      "epoch": 3.3002190149941035,
      "grad_norm": 1.140625,
      "learning_rate": 0.000394103938366008,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21538207083940505,
      "num_tokens": 65603553.0,
      "step": 29385
    },
    {
      "entropy": 5.373459243774414,
      "epoch": 3.300780591902061,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003940698741754041,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22100254893302917,
      "num_tokens": 65613903.0,
      "step": 29390
    },
    {
      "entropy": 5.393213748931885,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003940358061935692,
      "loss": 5.0164,
      "mean_token_accuracy": 0.22040466964244843,
      "num_tokens": 65624862.0,
      "step": 29395
    },
    {
      "entropy": 5.365935897827148,
      "epoch": 3.301903745717976,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003940017344215881,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2186554715037346,
      "num_tokens": 65636828.0,
      "step": 29400
    },
    {
      "entropy": 5.481935834884643,
      "epoch": 3.3024653226259337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039396765886054555,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2153605580329895,
      "num_tokens": 65648251.0,
      "step": 29405
    },
    {
      "entropy": 5.5379472255706785,
      "epoch": 3.303026899533891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039393357951152667,
      "loss": 5.1728,
      "mean_token_accuracy": 0.20836376249790192,
      "num_tokens": 65659715.0,
      "step": 29410
    },
    {
      "entropy": 5.475959348678589,
      "epoch": 3.3035884764418486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039389949637561646,
      "loss": 5.0306,
      "mean_token_accuracy": 0.2104856163263321,
      "num_tokens": 65670844.0,
      "step": 29415
    },
    {
      "entropy": 5.328753995895386,
      "epoch": 3.3041500533498063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039386540945390017,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2238621547818184,
      "num_tokens": 65682404.0,
      "step": 29420
    },
    {
      "entropy": 5.420098543167114,
      "epoch": 3.304711630257764,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003938313187474631,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21847616285085678,
      "num_tokens": 65692826.0,
      "step": 29425
    },
    {
      "entropy": 5.456115007400513,
      "epoch": 3.3052732071657216,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003937972242573908,
      "loss": 5.046,
      "mean_token_accuracy": 0.21802508533000947,
      "num_tokens": 65704139.0,
      "step": 29430
    },
    {
      "entropy": 5.378921556472778,
      "epoch": 3.305834784073679,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003937631259847688,
      "loss": 4.9766,
      "mean_token_accuracy": 0.21859926283359526,
      "num_tokens": 65716282.0,
      "step": 29435
    },
    {
      "entropy": 5.482021427154541,
      "epoch": 3.3063963609816365,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20913704186677934,
      "num_tokens": 65728086.0,
      "step": 29440
    },
    {
      "entropy": 5.5110602378845215,
      "epoch": 3.306957937889594,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039369491809621867,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20704671293497084,
      "num_tokens": 65740584.0,
      "step": 29445
    },
    {
      "entropy": 5.435218095779419,
      "epoch": 3.3075195147975514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003936608084824623,
      "loss": 5.0436,
      "mean_token_accuracy": 0.21423669457435607,
      "num_tokens": 65751479.0,
      "step": 29450
    },
    {
      "entropy": 5.378938484191894,
      "epoch": 3.308081091705509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039362669509049975,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21425842493772507,
      "num_tokens": 65762073.0,
      "step": 29455
    },
    {
      "entropy": 5.354738378524781,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039359257792141717,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21838144361972808,
      "num_tokens": 65772808.0,
      "step": 29460
    },
    {
      "entropy": 5.433453226089478,
      "epoch": 3.3092042455214243,
      "grad_norm": 1.046875,
      "learning_rate": 0.000393558456976301,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21307091861963273,
      "num_tokens": 65783507.0,
      "step": 29465
    },
    {
      "entropy": 5.443253707885742,
      "epoch": 3.3097658224293816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003935243322562375,
      "loss": 5.0145,
      "mean_token_accuracy": 0.21805361062288284,
      "num_tokens": 65794250.0,
      "step": 29470
    },
    {
      "entropy": 5.3888288021087645,
      "epoch": 3.310327399337339,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003934902037623133,
      "loss": 4.9911,
      "mean_token_accuracy": 0.2108955428004265,
      "num_tokens": 65805487.0,
      "step": 29475
    },
    {
      "entropy": 5.4464270114898685,
      "epoch": 3.310888976245297,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039345607149561513,
      "loss": 5.0417,
      "mean_token_accuracy": 0.2110534578561783,
      "num_tokens": 65818264.0,
      "step": 29480
    },
    {
      "entropy": 5.49127426147461,
      "epoch": 3.3114505531532545,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.1285,
      "mean_token_accuracy": 0.21572498679161073,
      "num_tokens": 65829686.0,
      "step": 29485
    },
    {
      "entropy": 5.417027473449707,
      "epoch": 3.3120121300612118,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003933877956482438,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22380726933479309,
      "num_tokens": 65840800.0,
      "step": 29490
    },
    {
      "entropy": 5.489288377761841,
      "epoch": 3.3125737069691694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039335365206974465,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21522391736507415,
      "num_tokens": 65853634.0,
      "step": 29495
    },
    {
      "entropy": 5.395300960540771,
      "epoch": 3.313135283877127,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003933195047228193,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21463861912488938,
      "num_tokens": 65864754.0,
      "step": 29500
    },
    {
      "entropy": 5.3390357971191404,
      "epoch": 3.3136968607850843,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039328535360855503,
      "loss": 5.038,
      "mean_token_accuracy": 0.21348643600940703,
      "num_tokens": 65876609.0,
      "step": 29505
    },
    {
      "entropy": 5.403463506698609,
      "epoch": 3.314258437693042,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039325119872803924,
      "loss": 4.9949,
      "mean_token_accuracy": 0.20859327912330627,
      "num_tokens": 65887583.0,
      "step": 29510
    },
    {
      "entropy": 5.456968212127686,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2238749459385872,
      "num_tokens": 65898131.0,
      "step": 29515
    },
    {
      "entropy": 5.393203449249268,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21651228219270707,
      "num_tokens": 65909353.0,
      "step": 29520
    },
    {
      "entropy": 5.4003394603729244,
      "epoch": 3.315943168416915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039314871149985834,
      "loss": 5.003,
      "mean_token_accuracy": 0.2157943144440651,
      "num_tokens": 65920709.0,
      "step": 29525
    },
    {
      "entropy": 5.446651124954224,
      "epoch": 3.316504745324872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003931145415652126,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21788393408060075,
      "num_tokens": 65932231.0,
      "step": 29530
    },
    {
      "entropy": 5.545429611206055,
      "epoch": 3.31706632223283,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20646177530288695,
      "num_tokens": 65942629.0,
      "step": 29535
    },
    {
      "entropy": 5.469950485229492,
      "epoch": 3.3176278991407875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003930461904145709,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21872362345457078,
      "num_tokens": 65954385.0,
      "step": 29540
    },
    {
      "entropy": 5.450983762741089,
      "epoch": 3.3181894760487447,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003930120092007512,
      "loss": 5.008,
      "mean_token_accuracy": 0.21920231431722642,
      "num_tokens": 65966283.0,
      "step": 29545
    },
    {
      "entropy": 5.5032764911651615,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003929778242293834,
      "loss": 5.1683,
      "mean_token_accuracy": 0.20289764702320098,
      "num_tokens": 65978594.0,
      "step": 29550
    },
    {
      "entropy": 5.341289758682251,
      "epoch": 3.31931262986466,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.9234,
      "mean_token_accuracy": 0.21703745275735856,
      "num_tokens": 65989398.0,
      "step": 29555
    },
    {
      "entropy": 5.368479633331299,
      "epoch": 3.3198742067726177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039290944301835723,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21716476678848268,
      "num_tokens": 66000545.0,
      "step": 29560
    },
    {
      "entropy": 5.418311548233032,
      "epoch": 3.320435783680575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039287524678087617,
      "loss": 5.007,
      "mean_token_accuracy": 0.2158213347196579,
      "num_tokens": 66011638.0,
      "step": 29565
    },
    {
      "entropy": 5.389225149154663,
      "epoch": 3.3209973605885326,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039284104679020154,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21039177626371383,
      "num_tokens": 66023284.0,
      "step": 29570
    },
    {
      "entropy": 5.385278511047363,
      "epoch": 3.3215589374964902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2148028254508972,
      "num_tokens": 66034099.0,
      "step": 29575
    },
    {
      "entropy": 5.36187252998352,
      "epoch": 3.322120514404448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003927726355536275,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22337814271450043,
      "num_tokens": 66043885.0,
      "step": 29580
    },
    {
      "entropy": 5.439000940322876,
      "epoch": 3.322682091312405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039273842430990634,
      "loss": 5.053,
      "mean_token_accuracy": 0.20969699770212175,
      "num_tokens": 66054190.0,
      "step": 29585
    },
    {
      "entropy": 5.432321977615357,
      "epoch": 3.3232436682203628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003927042093173481,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2135960191488266,
      "num_tokens": 66065254.0,
      "step": 29590
    },
    {
      "entropy": 5.450256729125977,
      "epoch": 3.3238052451283204,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003926699905770421,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20881291925907136,
      "num_tokens": 66076582.0,
      "step": 29595
    },
    {
      "entropy": 5.502733516693115,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000392635768090078,
      "loss": 5.0527,
      "mean_token_accuracy": 0.21130347549915313,
      "num_tokens": 66086410.0,
      "step": 29600
    },
    {
      "entropy": 5.486304903030396,
      "epoch": 3.3249283989442353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039260154185754545,
      "loss": 5.0836,
      "mean_token_accuracy": 0.20438562333583832,
      "num_tokens": 66098393.0,
      "step": 29605
    },
    {
      "entropy": 5.463166284561157,
      "epoch": 3.325489975852193,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003925673118805341,
      "loss": 5.0322,
      "mean_token_accuracy": 0.2143481358885765,
      "num_tokens": 66108077.0,
      "step": 29610
    },
    {
      "entropy": 5.304306554794311,
      "epoch": 3.3260515527601506,
      "grad_norm": 1.0625,
      "learning_rate": 0.000392533078160134,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2336338371038437,
      "num_tokens": 66118008.0,
      "step": 29615
    },
    {
      "entropy": 5.427421379089355,
      "epoch": 3.3266131296681083,
      "grad_norm": 1.0,
      "learning_rate": 0.00039249884069743514,
      "loss": 5.1031,
      "mean_token_accuracy": 0.21110040992498397,
      "num_tokens": 66128697.0,
      "step": 29620
    },
    {
      "entropy": 5.48628568649292,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003924645994935276,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21168190836906434,
      "num_tokens": 66139800.0,
      "step": 29625
    },
    {
      "entropy": 5.410335922241211,
      "epoch": 3.327736283484023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003924303545495017,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2253279522061348,
      "num_tokens": 66151202.0,
      "step": 29630
    },
    {
      "entropy": 5.396734952926636,
      "epoch": 3.328297860391981,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039239610586644765,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21415654867887496,
      "num_tokens": 66162159.0,
      "step": 29635
    },
    {
      "entropy": 5.440955591201782,
      "epoch": 3.328859437299938,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003923618534454561,
      "loss": 5.0081,
      "mean_token_accuracy": 0.2145876854658127,
      "num_tokens": 66174492.0,
      "step": 29640
    },
    {
      "entropy": 5.4303121089935305,
      "epoch": 3.3294210142078957,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003923275972876176,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21926959604024887,
      "num_tokens": 66186617.0,
      "step": 29645
    },
    {
      "entropy": 5.374270868301392,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22160379588603973,
      "num_tokens": 66198001.0,
      "step": 29650
    },
    {
      "entropy": 5.337052822113037,
      "epoch": 3.330544168023811,
      "grad_norm": 1.078125,
      "learning_rate": 0.000392259073765763,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21483289301395417,
      "num_tokens": 66208017.0,
      "step": 29655
    },
    {
      "entropy": 5.327909374237061,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039222480640392854,
      "loss": 4.956,
      "mean_token_accuracy": 0.21465105712413787,
      "num_tokens": 66219798.0,
      "step": 29660
    },
    {
      "entropy": 5.433707666397095,
      "epoch": 3.331667321839726,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039219053530961087,
      "loss": 5.0128,
      "mean_token_accuracy": 0.22140001654624938,
      "num_tokens": 66230662.0,
      "step": 29665
    },
    {
      "entropy": 5.477301406860351,
      "epoch": 3.3322288987476836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21949291676282884,
      "num_tokens": 66240787.0,
      "step": 29670
    },
    {
      "entropy": 5.355583524703979,
      "epoch": 3.3327904756556412,
      "grad_norm": 1.125,
      "learning_rate": 0.0003921219819278905,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2201547384262085,
      "num_tokens": 66250937.0,
      "step": 29675
    },
    {
      "entropy": 5.456735610961914,
      "epoch": 3.3333520525635985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003920876996426706,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20781537592411042,
      "num_tokens": 66261322.0,
      "step": 29680
    },
    {
      "entropy": 5.362344264984131,
      "epoch": 3.333913629471556,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21716698557138442,
      "num_tokens": 66271307.0,
      "step": 29685
    },
    {
      "entropy": 5.348518800735474,
      "epoch": 3.334475206379514,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22157902717590333,
      "num_tokens": 66283067.0,
      "step": 29690
    },
    {
      "entropy": 5.419895076751709,
      "epoch": 3.335036783287471,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003919848304226711,
      "loss": 5.0347,
      "mean_token_accuracy": 0.20748198181390762,
      "num_tokens": 66295853.0,
      "step": 29695
    },
    {
      "entropy": 5.457744026184082,
      "epoch": 3.3355983601954287,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003919505332315307,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2140216648578644,
      "num_tokens": 66307573.0,
      "step": 29700
    },
    {
      "entropy": 5.404476690292358,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039191623231663985,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22511793226003646,
      "num_tokens": 66318100.0,
      "step": 29705
    },
    {
      "entropy": 5.4496969223022464,
      "epoch": 3.336721514011344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000391881927679091,
      "loss": 4.9857,
      "mean_token_accuracy": 0.21446900963783264,
      "num_tokens": 66328918.0,
      "step": 29710
    },
    {
      "entropy": 5.393119525909424,
      "epoch": 3.3372830909193016,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039184761931997607,
      "loss": 5.0753,
      "mean_token_accuracy": 0.20741646289825438,
      "num_tokens": 66341124.0,
      "step": 29715
    },
    {
      "entropy": 5.410408449172974,
      "epoch": 3.337844667827259,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003918133072403878,
      "loss": 5.02,
      "mean_token_accuracy": 0.209653577208519,
      "num_tokens": 66351033.0,
      "step": 29720
    },
    {
      "entropy": 5.389892196655273,
      "epoch": 3.3384062447352165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039177899144141835,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22044313549995423,
      "num_tokens": 66363338.0,
      "step": 29725
    },
    {
      "entropy": 5.357401514053345,
      "epoch": 3.338967821643174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003917446719241606,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21771716177463532,
      "num_tokens": 66373716.0,
      "step": 29730
    },
    {
      "entropy": 5.444851493835449,
      "epoch": 3.3395293985511314,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22041223496198653,
      "num_tokens": 66384448.0,
      "step": 29735
    },
    {
      "entropy": 5.369354677200318,
      "epoch": 3.340090975459089,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039167602173915104,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22240135818719864,
      "num_tokens": 66395662.0,
      "step": 29740
    },
    {
      "entropy": 5.302375555038452,
      "epoch": 3.3406525523670467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22484293580055237,
      "num_tokens": 66407856.0,
      "step": 29745
    },
    {
      "entropy": 5.4706995487213135,
      "epoch": 3.3412141292750044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039160735669410246,
      "loss": 5.1088,
      "mean_token_accuracy": 0.20073889940977097,
      "num_tokens": 66419324.0,
      "step": 29750
    },
    {
      "entropy": 5.37638258934021,
      "epoch": 3.3417757061829616,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003915730186017964,
      "loss": 4.9049,
      "mean_token_accuracy": 0.21775493770837784,
      "num_tokens": 66430720.0,
      "step": 29755
    },
    {
      "entropy": 5.431620407104492,
      "epoch": 3.3423372830909193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039153867679776023,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2105632469058037,
      "num_tokens": 66441736.0,
      "step": 29760
    },
    {
      "entropy": 5.471164321899414,
      "epoch": 3.342898859998877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003915043312830873,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21568836122751237,
      "num_tokens": 66451851.0,
      "step": 29765
    },
    {
      "entropy": 5.5476487159729,
      "epoch": 3.3434604369068346,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003914699820588714,
      "loss": 5.1056,
      "mean_token_accuracy": 0.20627558529376983,
      "num_tokens": 66463105.0,
      "step": 29770
    },
    {
      "entropy": 5.504495429992676,
      "epoch": 3.344022013814792,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039143562912620603,
      "loss": 5.1171,
      "mean_token_accuracy": 0.21096610724925996,
      "num_tokens": 66474613.0,
      "step": 29775
    },
    {
      "entropy": 5.421213817596436,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.078125,
      "learning_rate": 0.000391401272486185,
      "loss": 5.0722,
      "mean_token_accuracy": 0.20721800178289412,
      "num_tokens": 66486783.0,
      "step": 29780
    },
    {
      "entropy": 5.538176441192627,
      "epoch": 3.345145167630707,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.0761,
      "mean_token_accuracy": 0.2179429978132248,
      "num_tokens": 66498581.0,
      "step": 29785
    },
    {
      "entropy": 5.407577562332153,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039133254808845216,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22372017353773116,
      "num_tokens": 66509480.0,
      "step": 29790
    },
    {
      "entropy": 5.387586641311645,
      "epoch": 3.346268321446622,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039129818033292843,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21841578781604767,
      "num_tokens": 66521831.0,
      "step": 29795
    },
    {
      "entropy": 5.422393178939819,
      "epoch": 3.3468298983545797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003912638088744256,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21810179054737092,
      "num_tokens": 66533900.0,
      "step": 29800
    },
    {
      "entropy": 5.545636892318726,
      "epoch": 3.3473914752625373,
      "grad_norm": 1.125,
      "learning_rate": 0.000391229433714038,
      "loss": 5.1335,
      "mean_token_accuracy": 0.2063618689775467,
      "num_tokens": 66543848.0,
      "step": 29805
    },
    {
      "entropy": 5.460444688796997,
      "epoch": 3.347953052170495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2159655511379242,
      "num_tokens": 66554113.0,
      "step": 29810
    },
    {
      "entropy": 5.480855798721313,
      "epoch": 3.348514629078452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003911606722919866,
      "loss": 5.0525,
      "mean_token_accuracy": 0.2100195050239563,
      "num_tokens": 66565202.0,
      "step": 29815
    },
    {
      "entropy": 5.465921926498413,
      "epoch": 3.34907620598641,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003911262860325123,
      "loss": 5.0975,
      "mean_token_accuracy": 0.20638331174850463,
      "num_tokens": 66575930.0,
      "step": 29820
    },
    {
      "entropy": 5.4703105926513675,
      "epoch": 3.3496377828943675,
      "grad_norm": 1.125,
      "learning_rate": 0.0003910918960755319,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21666872203350068,
      "num_tokens": 66585422.0,
      "step": 29825
    },
    {
      "entropy": 5.46008563041687,
      "epoch": 3.3501993598023248,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003910575024221406,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21480216532945634,
      "num_tokens": 66597395.0,
      "step": 29830
    },
    {
      "entropy": 5.432285308837891,
      "epoch": 3.3507609367102824,
      "grad_norm": 1.0,
      "learning_rate": 0.00039102310507343335,
      "loss": 5.0092,
      "mean_token_accuracy": 0.22011426836252213,
      "num_tokens": 66608464.0,
      "step": 29835
    },
    {
      "entropy": 5.457088279724121,
      "epoch": 3.35132251361824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003909887040305055,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2086101457476616,
      "num_tokens": 66619930.0,
      "step": 29840
    },
    {
      "entropy": 5.384872674942017,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003909542992944523,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21955328285694123,
      "num_tokens": 66632833.0,
      "step": 29845
    },
    {
      "entropy": 5.420553350448609,
      "epoch": 3.352445667434155,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003909198908663692,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21721287369728087,
      "num_tokens": 66643641.0,
      "step": 29850
    },
    {
      "entropy": 5.5144672870635985,
      "epoch": 3.3530072443421126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039088547874735187,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21013445109128953,
      "num_tokens": 66653863.0,
      "step": 29855
    },
    {
      "entropy": 5.3663698673248295,
      "epoch": 3.3535688212500703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039085106293849594,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22164923697710037,
      "num_tokens": 66665885.0,
      "step": 29860
    },
    {
      "entropy": 5.481834125518799,
      "epoch": 3.354130398158028,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039081664344089724,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21479262113571168,
      "num_tokens": 66676584.0,
      "step": 29865
    },
    {
      "entropy": 5.367331552505493,
      "epoch": 3.354691975065985,
      "grad_norm": 1.125,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.871,
      "mean_token_accuracy": 0.22526373118162155,
      "num_tokens": 66687882.0,
      "step": 29870
    },
    {
      "entropy": 5.422545862197876,
      "epoch": 3.355253551973943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003907477933838553,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2110013797879219,
      "num_tokens": 66699665.0,
      "step": 29875
    },
    {
      "entropy": 5.406256437301636,
      "epoch": 3.3558151288819005,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20774501711130142,
      "num_tokens": 66711455.0,
      "step": 29880
    },
    {
      "entropy": 5.3545623302459715,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039067892858499496,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22320693880319595,
      "num_tokens": 66722369.0,
      "step": 29885
    },
    {
      "entropy": 5.404347610473633,
      "epoch": 3.3569382826978154,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003906444906601237,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21509175300598143,
      "num_tokens": 66733200.0,
      "step": 29890
    },
    {
      "entropy": 5.404410362243652,
      "epoch": 3.357499859605773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000390610049053087,
      "loss": 4.9496,
      "mean_token_accuracy": 0.2195542111992836,
      "num_tokens": 66744650.0,
      "step": 29895
    },
    {
      "entropy": 5.44504714012146,
      "epoch": 3.3580614365137307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039057560376498147,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2080956816673279,
      "num_tokens": 66755712.0,
      "step": 29900
    },
    {
      "entropy": 5.449199676513672,
      "epoch": 3.3586230134216883,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003905411547969038,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2204502284526825,
      "num_tokens": 66765855.0,
      "step": 29905
    },
    {
      "entropy": 5.457657384872436,
      "epoch": 3.3591845903296456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003905067021499511,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2230652838945389,
      "num_tokens": 66776870.0,
      "step": 29910
    },
    {
      "entropy": 5.387476968765259,
      "epoch": 3.359746167237603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22289518713951112,
      "num_tokens": 66787226.0,
      "step": 29915
    },
    {
      "entropy": 5.36427640914917,
      "epoch": 3.360307744145561,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003904377858238081,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2173761248588562,
      "num_tokens": 66797352.0,
      "step": 29920
    },
    {
      "entropy": 5.471642732620239,
      "epoch": 3.360869321053518,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003904033221468122,
      "loss": 5.1145,
      "mean_token_accuracy": 0.2090977355837822,
      "num_tokens": 66809227.0,
      "step": 29925
    },
    {
      "entropy": 5.46846342086792,
      "epoch": 3.3614308979614758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003903688547953297,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21588925272226334,
      "num_tokens": 66821196.0,
      "step": 29930
    },
    {
      "entropy": 5.472156095504761,
      "epoch": 3.3619924748694334,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003903343837704582,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21322200894355775,
      "num_tokens": 66831958.0,
      "step": 29935
    },
    {
      "entropy": 5.441942119598389,
      "epoch": 3.362554051777391,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003902999090732951,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21331294924020766,
      "num_tokens": 66843727.0,
      "step": 29940
    },
    {
      "entropy": 5.382746028900146,
      "epoch": 3.3631156286853483,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039026543070493823,
      "loss": 4.997,
      "mean_token_accuracy": 0.21804955154657363,
      "num_tokens": 66854756.0,
      "step": 29945
    },
    {
      "entropy": 5.327332592010498,
      "epoch": 3.363677205593306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22045520097017288,
      "num_tokens": 66865615.0,
      "step": 29950
    },
    {
      "entropy": 5.3945897102355955,
      "epoch": 3.3642387825012636,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003901964629590344,
      "loss": 5.0126,
      "mean_token_accuracy": 0.20975813269615173,
      "num_tokens": 66877262.0,
      "step": 29955
    },
    {
      "entropy": 5.52663288116455,
      "epoch": 3.3648003594092213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003901619735836833,
      "loss": 5.1006,
      "mean_token_accuracy": 0.20858699530363084,
      "num_tokens": 66889405.0,
      "step": 29960
    },
    {
      "entropy": 5.412596988677978,
      "epoch": 3.3653619363171785,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039012748054153044,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22184085100889206,
      "num_tokens": 66900013.0,
      "step": 29965
    },
    {
      "entropy": 5.3628870964050295,
      "epoch": 3.365923513225136,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003900929838336738,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22317400723695754,
      "num_tokens": 66909731.0,
      "step": 29970
    },
    {
      "entropy": 5.317621803283691,
      "epoch": 3.366485090133094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21928259879350662,
      "num_tokens": 66919731.0,
      "step": 29975
    },
    {
      "entropy": 5.397083854675293,
      "epoch": 3.367046667041051,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003900239794252434,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2143187031149864,
      "num_tokens": 66930169.0,
      "step": 29980
    },
    {
      "entropy": 5.43265814781189,
      "epoch": 3.3676082439490087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003899894717268667,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22038244754076003,
      "num_tokens": 66942010.0,
      "step": 29985
    },
    {
      "entropy": 5.439857339859008,
      "epoch": 3.3681698208569664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003899549603671807,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2208854377269745,
      "num_tokens": 66953972.0,
      "step": 29990
    },
    {
      "entropy": 5.428193664550781,
      "epoch": 3.368731397764924,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038992044534728404,
      "loss": 5.0787,
      "mean_token_accuracy": 0.21117445826530457,
      "num_tokens": 66965548.0,
      "step": 29995
    },
    {
      "entropy": 5.419106292724609,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038988592666827594,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21269267350435256,
      "num_tokens": 66976414.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.263197124388712,
      "eval_loss": 5.1703057289123535,
      "eval_mean_token_accuracy": 0.21479068723211037,
      "eval_num_tokens": 66976414.0,
      "eval_runtime": 53.5448,
      "eval_samples_per_second": 579.198,
      "eval_steps_per_second": 72.407,
      "step": 30000
    },
    {
      "entropy": 5.356390857696534,
      "epoch": 3.369854551580839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003898514043312553,
      "loss": 4.9584,
      "mean_token_accuracy": 0.21750566214323044,
      "num_tokens": 66987392.0,
      "step": 30005
    },
    {
      "entropy": 5.574415969848633,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038981687833732146,
      "loss": 5.154,
      "mean_token_accuracy": 0.20670795142650605,
      "num_tokens": 67000444.0,
      "step": 30010
    },
    {
      "entropy": 5.4609535217285154,
      "epoch": 3.3709777053967542,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003897823486875737,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21463242471218108,
      "num_tokens": 67012124.0,
      "step": 30015
    },
    {
      "entropy": 5.496159076690674,
      "epoch": 3.3715392823047114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003897478153831114,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21203296333551408,
      "num_tokens": 67024179.0,
      "step": 30020
    },
    {
      "entropy": 5.438608646392822,
      "epoch": 3.372100859212669,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21883501261472701,
      "num_tokens": 67035876.0,
      "step": 30025
    },
    {
      "entropy": 5.361820030212402,
      "epoch": 3.3726624361206268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038967873781444167,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2189247727394104,
      "num_tokens": 67046499.0,
      "step": 30030
    },
    {
      "entropy": 5.387604761123657,
      "epoch": 3.3732240130285844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038964419355243365,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21360039263963698,
      "num_tokens": 67058357.0,
      "step": 30035
    },
    {
      "entropy": 5.4415830135345455,
      "epoch": 3.3737855899365417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003896096456401101,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20744526833295823,
      "num_tokens": 67069753.0,
      "step": 30040
    },
    {
      "entropy": 5.404859781265259,
      "epoch": 3.3743471668444993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21883049011230468,
      "num_tokens": 67080185.0,
      "step": 30045
    },
    {
      "entropy": 5.400932836532593,
      "epoch": 3.374908743752457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21718937903642654,
      "num_tokens": 67091382.0,
      "step": 30050
    },
    {
      "entropy": 5.531070327758789,
      "epoch": 3.3754703206604146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003895059800122466,
      "loss": 5.1771,
      "mean_token_accuracy": 0.19447798579931258,
      "num_tokens": 67102952.0,
      "step": 30055
    },
    {
      "entropy": 5.444001007080078,
      "epoch": 3.376031897568372,
      "grad_norm": 1.0625,
      "learning_rate": 0.000389471417509662,
      "loss": 4.9284,
      "mean_token_accuracy": 0.21277776956558228,
      "num_tokens": 67113498.0,
      "step": 30060
    },
    {
      "entropy": 5.418158340454101,
      "epoch": 3.3765934744763295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000389436851362263,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21807664036750793,
      "num_tokens": 67124591.0,
      "step": 30065
    },
    {
      "entropy": 5.421072578430175,
      "epoch": 3.377155051384287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038940228157115036,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2102195754647255,
      "num_tokens": 67135627.0,
      "step": 30070
    },
    {
      "entropy": 5.373394775390625,
      "epoch": 3.3777166282922444,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038936770813742473,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21811100989580154,
      "num_tokens": 67146084.0,
      "step": 30075
    },
    {
      "entropy": 5.404456090927124,
      "epoch": 3.378278205200202,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038933313106218694,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21629900187253953,
      "num_tokens": 67157749.0,
      "step": 30080
    },
    {
      "entropy": 5.479741859436035,
      "epoch": 3.3788397821081597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038929855034653776,
      "loss": 5.1045,
      "mean_token_accuracy": 0.21472135782241822,
      "num_tokens": 67169386.0,
      "step": 30085
    },
    {
      "entropy": 5.370194292068481,
      "epoch": 3.3794013590161174,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038926396599157855,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21454079151153566,
      "num_tokens": 67181436.0,
      "step": 30090
    },
    {
      "entropy": 5.418502378463745,
      "epoch": 3.379962935924075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003892293779984103,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2262387290596962,
      "num_tokens": 67192576.0,
      "step": 30095
    },
    {
      "entropy": 5.544774293899536,
      "epoch": 3.3805245128320323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.1194,
      "mean_token_accuracy": 0.2097189709544182,
      "num_tokens": 67205083.0,
      "step": 30100
    },
    {
      "entropy": 5.325855159759522,
      "epoch": 3.38108608973999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22101180702447892,
      "num_tokens": 67216499.0,
      "step": 30105
    },
    {
      "entropy": 5.326472425460816,
      "epoch": 3.3816476666479476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22122545242309571,
      "num_tokens": 67228390.0,
      "step": 30110
    },
    {
      "entropy": 5.42553186416626,
      "epoch": 3.382209243555905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000389090989665675,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21733855456113815,
      "num_tokens": 67239655.0,
      "step": 30115
    },
    {
      "entropy": 5.413579988479614,
      "epoch": 3.3827708204638625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003890563834979835,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21064104735851288,
      "num_tokens": 67251712.0,
      "step": 30120
    },
    {
      "entropy": 5.43354549407959,
      "epoch": 3.38333239737182,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003890217736986926,
      "loss": 4.985,
      "mean_token_accuracy": 0.21767974346876146,
      "num_tokens": 67262494.0,
      "step": 30125
    },
    {
      "entropy": 5.41361927986145,
      "epoch": 3.383893974279778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038898716026890414,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2226088210940361,
      "num_tokens": 67273179.0,
      "step": 30130
    },
    {
      "entropy": 5.406420755386352,
      "epoch": 3.384455551187735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003889525432097203,
      "loss": 5.0266,
      "mean_token_accuracy": 0.22028181254863738,
      "num_tokens": 67284031.0,
      "step": 30135
    },
    {
      "entropy": 5.4914538860321045,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003889179225222432,
      "loss": 5.1181,
      "mean_token_accuracy": 0.20829652547836303,
      "num_tokens": 67295728.0,
      "step": 30140
    },
    {
      "entropy": 5.448974561691284,
      "epoch": 3.3855787050036503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038888329820757536,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20924435704946517,
      "num_tokens": 67306880.0,
      "step": 30145
    },
    {
      "entropy": 5.410636472702026,
      "epoch": 3.386140281911608,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21838705837726594,
      "num_tokens": 67317849.0,
      "step": 30150
    },
    {
      "entropy": 5.360437536239624,
      "epoch": 3.386701858819565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22095607817173005,
      "num_tokens": 67329421.0,
      "step": 30155
    },
    {
      "entropy": 5.430154275894165,
      "epoch": 3.387263435727523,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038877940351145186,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21530342996120452,
      "num_tokens": 67340072.0,
      "step": 30160
    },
    {
      "entropy": 5.391668081283569,
      "epoch": 3.3878250126354805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.957,
      "mean_token_accuracy": 0.2175828367471695,
      "num_tokens": 67351687.0,
      "step": 30165
    },
    {
      "entropy": 5.41436996459961,
      "epoch": 3.3883865895434377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2164614036679268,
      "num_tokens": 67362978.0,
      "step": 30170
    },
    {
      "entropy": 5.337440824508667,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.09375,
      "learning_rate": 0.000388675476210306,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2272678554058075,
      "num_tokens": 67374142.0,
      "step": 30175
    },
    {
      "entropy": 5.401316547393799,
      "epoch": 3.389509743359353,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038864082653617744,
      "loss": 5.0258,
      "mean_token_accuracy": 0.20643302351236342,
      "num_tokens": 67385196.0,
      "step": 30180
    },
    {
      "entropy": 5.3670713901519775,
      "epoch": 3.3900713202673107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038860617324368085,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2223791465163231,
      "num_tokens": 67395830.0,
      "step": 30185
    },
    {
      "entropy": 5.4027430534362795,
      "epoch": 3.3906328971752684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003885715163339196,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21243101209402085,
      "num_tokens": 67405950.0,
      "step": 30190
    },
    {
      "entropy": 5.418920230865479,
      "epoch": 3.3911944740832256,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9994,
      "mean_token_accuracy": 0.21518580615520477,
      "num_tokens": 67417993.0,
      "step": 30195
    },
    {
      "entropy": 5.46838755607605,
      "epoch": 3.3917560509911833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22321170419454575,
      "num_tokens": 67429222.0,
      "step": 30200
    },
    {
      "entropy": 5.473378038406372,
      "epoch": 3.392317627899141,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003884675239120838,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21326005309820176,
      "num_tokens": 67439870.0,
      "step": 30205
    },
    {
      "entropy": 5.4111741065979,
      "epoch": 3.392879204807098,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038843285254430005,
      "loss": 4.8778,
      "mean_token_accuracy": 0.2207608252763748,
      "num_tokens": 67450491.0,
      "step": 30210
    },
    {
      "entropy": 5.458513832092285,
      "epoch": 3.393440781715056,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003883981775647705,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21297233998775483,
      "num_tokens": 67461432.0,
      "step": 30215
    },
    {
      "entropy": 5.449087142944336,
      "epoch": 3.3940023586230135,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003883634989745988,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21804586350917815,
      "num_tokens": 67472614.0,
      "step": 30220
    },
    {
      "entropy": 5.418417930603027,
      "epoch": 3.394563935530971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038832881677488944,
      "loss": 5.013,
      "mean_token_accuracy": 0.21874067634344102,
      "num_tokens": 67484383.0,
      "step": 30225
    },
    {
      "entropy": 5.412888383865356,
      "epoch": 3.3951255124389284,
      "grad_norm": 1.125,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.979,
      "mean_token_accuracy": 0.21304466128349303,
      "num_tokens": 67495697.0,
      "step": 30230
    },
    {
      "entropy": 5.369659757614135,
      "epoch": 3.395687089346886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003882594415512746,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21882841736078262,
      "num_tokens": 67507937.0,
      "step": 30235
    },
    {
      "entropy": 5.382726526260376,
      "epoch": 3.3962486662548437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21943603008985518,
      "num_tokens": 67519649.0,
      "step": 30240
    },
    {
      "entropy": 5.572783422470093,
      "epoch": 3.3968102431628013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003881900519027618,
      "loss": 5.096,
      "mean_token_accuracy": 0.20984088629484177,
      "num_tokens": 67532301.0,
      "step": 30245
    },
    {
      "entropy": 5.458900308609008,
      "epoch": 3.3973718200707586,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003881553516719303,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21826161742210387,
      "num_tokens": 67545106.0,
      "step": 30250
    },
    {
      "entropy": 5.43557071685791,
      "epoch": 3.397933396978716,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038812064783818847,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20697888880968093,
      "num_tokens": 67556188.0,
      "step": 30255
    },
    {
      "entropy": 5.387828636169433,
      "epoch": 3.398494973886674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.9382,
      "mean_token_accuracy": 0.21820231676101684,
      "num_tokens": 67567707.0,
      "step": 30260
    },
    {
      "entropy": 5.393126916885376,
      "epoch": 3.399056550794631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000388051229366394,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2123485118150711,
      "num_tokens": 67578334.0,
      "step": 30265
    },
    {
      "entropy": 5.310758256912232,
      "epoch": 3.3996181277025888,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038801651473055167,
      "loss": 4.8113,
      "mean_token_accuracy": 0.22361154556274415,
      "num_tokens": 67588896.0,
      "step": 30270
    },
    {
      "entropy": 5.398510313034057,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003879817964962197,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21246001422405242,
      "num_tokens": 67599032.0,
      "step": 30275
    },
    {
      "entropy": 5.402614545822144,
      "epoch": 3.400741281518504,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003879470746645035,
      "loss": 5.0677,
      "mean_token_accuracy": 0.2059406280517578,
      "num_tokens": 67610095.0,
      "step": 30280
    },
    {
      "entropy": 5.473405504226685,
      "epoch": 3.4013028584264617,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038791234923650866,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20767418146133423,
      "num_tokens": 67622039.0,
      "step": 30285
    },
    {
      "entropy": 5.505795478820801,
      "epoch": 3.401864435334419,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038787762021334077,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21120986044406892,
      "num_tokens": 67634123.0,
      "step": 30290
    },
    {
      "entropy": 5.4455015659332275,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003878428875961057,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21945976465940475,
      "num_tokens": 67644696.0,
      "step": 30295
    },
    {
      "entropy": 5.431408882141113,
      "epoch": 3.4029875891503343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038780815138590935,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2028679668903351,
      "num_tokens": 67656193.0,
      "step": 30300
    },
    {
      "entropy": 5.37232141494751,
      "epoch": 3.4035491660582915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003877734115838577,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21898456066846847,
      "num_tokens": 67668261.0,
      "step": 30305
    },
    {
      "entropy": 5.455403566360474,
      "epoch": 3.404110742966249,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038773866819105684,
      "loss": 5.042,
      "mean_token_accuracy": 0.2099032387137413,
      "num_tokens": 67678264.0,
      "step": 30310
    },
    {
      "entropy": 5.47145791053772,
      "epoch": 3.404672319874207,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003877039212086131,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21684155464172364,
      "num_tokens": 67688455.0,
      "step": 30315
    },
    {
      "entropy": 5.364530992507935,
      "epoch": 3.4052338967821645,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.8983,
      "mean_token_accuracy": 0.219798281788826,
      "num_tokens": 67700970.0,
      "step": 30320
    },
    {
      "entropy": 5.438500881195068,
      "epoch": 3.4057954736901217,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038763441647922246,
      "loss": 5.0406,
      "mean_token_accuracy": 0.21601363122463227,
      "num_tokens": 67711835.0,
      "step": 30325
    },
    {
      "entropy": 5.359304475784302,
      "epoch": 3.4063570505980794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003875996587344885,
      "loss": 4.9907,
      "mean_token_accuracy": 0.21662724614143372,
      "num_tokens": 67722915.0,
      "step": 30330
    },
    {
      "entropy": 5.483043003082275,
      "epoch": 3.406918627506037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038756489740453785,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21697765737771987,
      "num_tokens": 67734301.0,
      "step": 30335
    },
    {
      "entropy": 5.469515609741211,
      "epoch": 3.4074802044139947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038753013249047706,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21445566415786743,
      "num_tokens": 67745763.0,
      "step": 30340
    },
    {
      "entropy": 5.465812349319458,
      "epoch": 3.408041781321952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003874953639934134,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21650987565517427,
      "num_tokens": 67757241.0,
      "step": 30345
    },
    {
      "entropy": 5.423696422576905,
      "epoch": 3.4086033582299096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003874605919144536,
      "loss": 5.022,
      "mean_token_accuracy": 0.21691177785396576,
      "num_tokens": 67768231.0,
      "step": 30350
    },
    {
      "entropy": 5.458590507507324,
      "epoch": 3.4091649351378672,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20559905767440795,
      "num_tokens": 67780735.0,
      "step": 30355
    },
    {
      "entropy": 5.4007415771484375,
      "epoch": 3.4097265120458244,
      "grad_norm": 1.0,
      "learning_rate": 0.0003873910370152747,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22381912618875505,
      "num_tokens": 67791817.0,
      "step": 30360
    },
    {
      "entropy": 5.467533731460572,
      "epoch": 3.410288088953782,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003873562541972702,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2169603541493416,
      "num_tokens": 67803591.0,
      "step": 30365
    },
    {
      "entropy": 5.439803743362427,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003873214678017991,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21671829968690873,
      "num_tokens": 67813250.0,
      "step": 30370
    },
    {
      "entropy": 5.3959414005279545,
      "epoch": 3.4114112427696974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003872866778299688,
      "loss": 4.999,
      "mean_token_accuracy": 0.2156549334526062,
      "num_tokens": 67825042.0,
      "step": 30375
    },
    {
      "entropy": 5.488438653945923,
      "epoch": 3.411972819677655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003872518842828872,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20476576536893845,
      "num_tokens": 67837535.0,
      "step": 30380
    },
    {
      "entropy": 5.522025299072266,
      "epoch": 3.4125343965856123,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038721708716166203,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21846868842840195,
      "num_tokens": 67848950.0,
      "step": 30385
    },
    {
      "entropy": 5.45252833366394,
      "epoch": 3.41309597349357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003871822864674013,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21058784425258636,
      "num_tokens": 67861488.0,
      "step": 30390
    },
    {
      "entropy": 5.433585596084595,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038714748220121305,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2153114855289459,
      "num_tokens": 67872627.0,
      "step": 30395
    },
    {
      "entropy": 5.416838312149048,
      "epoch": 3.414219127309485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003871126743642055,
      "loss": 5.0172,
      "mean_token_accuracy": 0.20840884000062943,
      "num_tokens": 67884033.0,
      "step": 30400
    },
    {
      "entropy": 5.408342266082764,
      "epoch": 3.4147807042174425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038707786295748693,
      "loss": 5.0171,
      "mean_token_accuracy": 0.20950771123170853,
      "num_tokens": 67895696.0,
      "step": 30405
    },
    {
      "entropy": 5.474095630645752,
      "epoch": 3.4153422811254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22693520933389663,
      "num_tokens": 67906328.0,
      "step": 30410
    },
    {
      "entropy": 5.385289430618286,
      "epoch": 3.415903858033358,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22586987763643265,
      "num_tokens": 67917258.0,
      "step": 30415
    },
    {
      "entropy": 5.36063141822815,
      "epoch": 3.416465434941315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003869734073301495,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21352871358394623,
      "num_tokens": 67927431.0,
      "step": 30420
    },
    {
      "entropy": 5.466087388992309,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038693858165567195,
      "loss": 5.0482,
      "mean_token_accuracy": 0.213152776658535,
      "num_tokens": 67938801.0,
      "step": 30425
    },
    {
      "entropy": 5.39945878982544,
      "epoch": 3.4175885887572304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22163398563861847,
      "num_tokens": 67949690.0,
      "step": 30430
    },
    {
      "entropy": 5.362391567230224,
      "epoch": 3.418150165665188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038686891961532227,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2183550015091896,
      "num_tokens": 67959542.0,
      "step": 30435
    },
    {
      "entropy": 5.339684963226318,
      "epoch": 3.4187117425731453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.8654,
      "mean_token_accuracy": 0.21930772066116333,
      "num_tokens": 67969759.0,
      "step": 30440
    },
    {
      "entropy": 5.386876392364502,
      "epoch": 3.419273319481103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038679924332717363,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21790457814931868,
      "num_tokens": 67982302.0,
      "step": 30445
    },
    {
      "entropy": 5.4488931655883786,
      "epoch": 3.4198348963890606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038676439984294776,
      "loss": 5.0767,
      "mean_token_accuracy": 0.21134627014398574,
      "num_tokens": 67994452.0,
      "step": 30450
    },
    {
      "entropy": 5.46291708946228,
      "epoch": 3.420396473297018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038672955280009996,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2086002141237259,
      "num_tokens": 68006445.0,
      "step": 30455
    },
    {
      "entropy": 5.440454578399658,
      "epoch": 3.4209580502049755,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038669470219973996,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2147781565785408,
      "num_tokens": 68017720.0,
      "step": 30460
    },
    {
      "entropy": 5.4213377952575685,
      "epoch": 3.421519627112933,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038665984804297734,
      "loss": 4.956,
      "mean_token_accuracy": 0.21842214316129685,
      "num_tokens": 68028199.0,
      "step": 30465
    },
    {
      "entropy": 5.475848770141601,
      "epoch": 3.4220812040208908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003866249903309218,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2092217594385147,
      "num_tokens": 68040171.0,
      "step": 30470
    },
    {
      "entropy": 5.4729479312896725,
      "epoch": 3.4226427809288484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003865901290646833,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2155963435769081,
      "num_tokens": 68050484.0,
      "step": 30475
    },
    {
      "entropy": 5.429516839981079,
      "epoch": 3.4232043578368057,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038655526424537175,
      "loss": 5.0467,
      "mean_token_accuracy": 0.21701580435037612,
      "num_tokens": 68062110.0,
      "step": 30480
    },
    {
      "entropy": 5.474966716766358,
      "epoch": 3.4237659347447633,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038652039587409734,
      "loss": 5.0767,
      "mean_token_accuracy": 0.2147672414779663,
      "num_tokens": 68074890.0,
      "step": 30485
    },
    {
      "entropy": 5.422239494323731,
      "epoch": 3.424327511652721,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003864855239519703,
      "loss": 5.001,
      "mean_token_accuracy": 0.21114905029535294,
      "num_tokens": 68084909.0,
      "step": 30490
    },
    {
      "entropy": 5.331094646453858,
      "epoch": 3.424889088560678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038645064848010093,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2196807771921158,
      "num_tokens": 68097114.0,
      "step": 30495
    },
    {
      "entropy": 5.4503552436828615,
      "epoch": 3.425450665468636,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038641576945959964,
      "loss": 5.0284,
      "mean_token_accuracy": 0.22285671830177306,
      "num_tokens": 68108377.0,
      "step": 30500
    },
    {
      "entropy": 5.382394742965698,
      "epoch": 3.4260122423765935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038638088689157713,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21735697984695435,
      "num_tokens": 68118859.0,
      "step": 30505
    },
    {
      "entropy": 5.382855415344238,
      "epoch": 3.426573819284551,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038634600077714395,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22645294070243835,
      "num_tokens": 68130151.0,
      "step": 30510
    },
    {
      "entropy": 5.362254333496094,
      "epoch": 3.4271353961925084,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003863111111174109,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21665270179510115,
      "num_tokens": 68142514.0,
      "step": 30515
    },
    {
      "entropy": 5.450138139724731,
      "epoch": 3.427696973100466,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038627621791348886,
      "loss": 5.0925,
      "mean_token_accuracy": 0.2096511721611023,
      "num_tokens": 68153751.0,
      "step": 30520
    },
    {
      "entropy": 5.470752334594726,
      "epoch": 3.4282585500084237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038624132116648897,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21724725216627122,
      "num_tokens": 68164885.0,
      "step": 30525
    },
    {
      "entropy": 5.384390115737915,
      "epoch": 3.4288201269163814,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003862064208775223,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21908168494701385,
      "num_tokens": 68176320.0,
      "step": 30530
    },
    {
      "entropy": 5.433656215667725,
      "epoch": 3.4293817038243386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038617151704769994,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2226571410894394,
      "num_tokens": 68186665.0,
      "step": 30535
    },
    {
      "entropy": 5.480368280410767,
      "epoch": 3.4299432807322963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003861366096781334,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20761403888463975,
      "num_tokens": 68198011.0,
      "step": 30540
    },
    {
      "entropy": 5.501778697967529,
      "epoch": 3.430504857640254,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003861016987699342,
      "loss": 5.1338,
      "mean_token_accuracy": 0.21084092557430267,
      "num_tokens": 68209659.0,
      "step": 30545
    },
    {
      "entropy": 5.440423154830933,
      "epoch": 3.431066434548211,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21906651556491852,
      "num_tokens": 68220424.0,
      "step": 30550
    },
    {
      "entropy": 5.424066495895386,
      "epoch": 3.431628011456169,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.964,
      "mean_token_accuracy": 0.21535649746656418,
      "num_tokens": 68231413.0,
      "step": 30555
    },
    {
      "entropy": 5.415523624420166,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038599694482465606,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22540453672409058,
      "num_tokens": 68242435.0,
      "step": 30560
    },
    {
      "entropy": 5.415911722183227,
      "epoch": 3.432751165272084,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038596201977304263,
      "loss": 4.9899,
      "mean_token_accuracy": 0.22242345660924911,
      "num_tokens": 68254462.0,
      "step": 30565
    },
    {
      "entropy": 5.422529888153076,
      "epoch": 3.433312742180042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038592709118835547,
      "loss": 5.036,
      "mean_token_accuracy": 0.21381750255823134,
      "num_tokens": 68265842.0,
      "step": 30570
    },
    {
      "entropy": 5.4009528160095215,
      "epoch": 3.433874319087999,
      "grad_norm": 1.0,
      "learning_rate": 0.00038589215907170664,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21514428853988649,
      "num_tokens": 68277117.0,
      "step": 30575
    },
    {
      "entropy": 5.470483016967774,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003858572234242086,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21176849454641342,
      "num_tokens": 68287869.0,
      "step": 30580
    },
    {
      "entropy": 5.4607765674591064,
      "epoch": 3.4349974729039143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21739738285541535,
      "num_tokens": 68300711.0,
      "step": 30585
    },
    {
      "entropy": 5.479371070861816,
      "epoch": 3.4355590498118715,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003857873415411138,
      "loss": 5.09,
      "mean_token_accuracy": 0.2162911370396614,
      "num_tokens": 68311997.0,
      "step": 30590
    },
    {
      "entropy": 5.393675470352173,
      "epoch": 3.436120626719829,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003857523953077423,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21748077273368835,
      "num_tokens": 68322129.0,
      "step": 30595
    },
    {
      "entropy": 5.349336242675781,
      "epoch": 3.436682203627787,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038571744554797156,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21900523751974105,
      "num_tokens": 68334202.0,
      "step": 30600
    },
    {
      "entropy": 5.405133485794067,
      "epoch": 3.4372437805357445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038568249226291433,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21997382640838622,
      "num_tokens": 68345370.0,
      "step": 30605
    },
    {
      "entropy": 5.367954587936401,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038564753545368375,
      "loss": 4.867,
      "mean_token_accuracy": 0.22071742564439772,
      "num_tokens": 68355703.0,
      "step": 30610
    },
    {
      "entropy": 5.39345064163208,
      "epoch": 3.4383669343516594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038561257512139264,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2047760233283043,
      "num_tokens": 68367412.0,
      "step": 30615
    },
    {
      "entropy": 5.424029588699341,
      "epoch": 3.438928511259617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038557761126715426,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21954822093248366,
      "num_tokens": 68378074.0,
      "step": 30620
    },
    {
      "entropy": 5.465312576293945,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003855426438920818,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2104610875248909,
      "num_tokens": 68388730.0,
      "step": 30625
    },
    {
      "entropy": 5.40934910774231,
      "epoch": 3.440051665075532,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21681826710700988,
      "num_tokens": 68399172.0,
      "step": 30630
    },
    {
      "entropy": 5.438844823837281,
      "epoch": 3.4406132419834896,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038547269858388853,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21505960673093796,
      "num_tokens": 68411704.0,
      "step": 30635
    },
    {
      "entropy": 5.365919351577759,
      "epoch": 3.4411748188914473,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003854377206529947,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21580222845077515,
      "num_tokens": 68422774.0,
      "step": 30640
    },
    {
      "entropy": 5.377093696594239,
      "epoch": 3.4417363957994045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000385402739205721,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2278720960021019,
      "num_tokens": 68432964.0,
      "step": 30645
    },
    {
      "entropy": 5.411448431015015,
      "epoch": 3.442297972707362,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003853677542431813,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2115648478269577,
      "num_tokens": 68445875.0,
      "step": 30650
    },
    {
      "entropy": 5.502961254119873,
      "epoch": 3.44285954961532,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003853327657664895,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20910585969686507,
      "num_tokens": 68457972.0,
      "step": 30655
    },
    {
      "entropy": 5.399326658248901,
      "epoch": 3.4434211265232775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003852977737767596,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21948170214891433,
      "num_tokens": 68468825.0,
      "step": 30660
    },
    {
      "entropy": 5.363817882537842,
      "epoch": 3.443982703431235,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003852627782751059,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2256024569272995,
      "num_tokens": 68479427.0,
      "step": 30665
    },
    {
      "entropy": 5.337355613708496,
      "epoch": 3.4445442803391924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.9209,
      "mean_token_accuracy": 0.21892691552639007,
      "num_tokens": 68490542.0,
      "step": 30670
    },
    {
      "entropy": 5.32402868270874,
      "epoch": 3.44510585724715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003851927767404839,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2213971272110939,
      "num_tokens": 68501396.0,
      "step": 30675
    },
    {
      "entropy": 5.486931943893433,
      "epoch": 3.4456674341551077,
      "grad_norm": 1.0,
      "learning_rate": 0.0003851577707097445,
      "loss": 5.1282,
      "mean_token_accuracy": 0.20419915318489074,
      "num_tokens": 68512710.0,
      "step": 30680
    },
    {
      "entropy": 5.441511726379394,
      "epoch": 3.446229011063065,
      "grad_norm": 1.15625,
      "learning_rate": 0.000385122761171539,
      "loss": 4.9919,
      "mean_token_accuracy": 0.20937654078006745,
      "num_tokens": 68524289.0,
      "step": 30685
    },
    {
      "entropy": 5.420687103271485,
      "epoch": 3.4467905879710226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038508774812698207,
      "loss": 4.9751,
      "mean_token_accuracy": 0.21640805006027222,
      "num_tokens": 68536783.0,
      "step": 30690
    },
    {
      "entropy": 5.404241323471069,
      "epoch": 3.44735216487898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003850527315771885,
      "loss": 5.0436,
      "mean_token_accuracy": 0.21436649560928345,
      "num_tokens": 68548113.0,
      "step": 30695
    },
    {
      "entropy": 5.404731321334839,
      "epoch": 3.447913741786938,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038501771152327333,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21887895613908767,
      "num_tokens": 68560234.0,
      "step": 30700
    },
    {
      "entropy": 5.465876483917237,
      "epoch": 3.448475318694895,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038498268796635155,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22018066942691802,
      "num_tokens": 68571712.0,
      "step": 30705
    },
    {
      "entropy": 5.387991714477539,
      "epoch": 3.4490368956028528,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.948,
      "mean_token_accuracy": 0.22060882300138474,
      "num_tokens": 68583166.0,
      "step": 30710
    },
    {
      "entropy": 5.337613821029663,
      "epoch": 3.4495984725108104,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23412090241909028,
      "num_tokens": 68594026.0,
      "step": 30715
    },
    {
      "entropy": 5.282212448120117,
      "epoch": 3.450160049418768,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003848775962886987,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21850949078798293,
      "num_tokens": 68604046.0,
      "step": 30720
    },
    {
      "entropy": 5.309293556213379,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003848425587309032,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2235302060842514,
      "num_tokens": 68615575.0,
      "step": 30725
    },
    {
      "entropy": 5.468441343307495,
      "epoch": 3.451283203234683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038480751767567807,
      "loss": 5.08,
      "mean_token_accuracy": 0.21781038343906403,
      "num_tokens": 68626805.0,
      "step": 30730
    },
    {
      "entropy": 5.567868328094482,
      "epoch": 3.4518447801426406,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000384772473124139,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21050034910440446,
      "num_tokens": 68638802.0,
      "step": 30735
    },
    {
      "entropy": 5.475788402557373,
      "epoch": 3.452406357050598,
      "grad_norm": 1.0,
      "learning_rate": 0.0003847374250774018,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20988640040159226,
      "num_tokens": 68651137.0,
      "step": 30740
    },
    {
      "entropy": 5.352191400527954,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038470237353658237,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22320826202630997,
      "num_tokens": 68661268.0,
      "step": 30745
    },
    {
      "entropy": 5.285566949844361,
      "epoch": 3.453529510866513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038466731850279683,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2181159719824791,
      "num_tokens": 68672334.0,
      "step": 30750
    },
    {
      "entropy": 5.399474573135376,
      "epoch": 3.454091087774471,
      "grad_norm": 1.0,
      "learning_rate": 0.00038463225997716137,
      "loss": 5.0316,
      "mean_token_accuracy": 0.217561212182045,
      "num_tokens": 68683765.0,
      "step": 30755
    },
    {
      "entropy": 5.578990936279297,
      "epoch": 3.4546526646824285,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038459719796079227,
      "loss": 5.1322,
      "mean_token_accuracy": 0.20988909900188446,
      "num_tokens": 68694969.0,
      "step": 30760
    },
    {
      "entropy": 5.5022448062896725,
      "epoch": 3.4552142415903857,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003845621324548059,
      "loss": 5.1126,
      "mean_token_accuracy": 0.21021440923213958,
      "num_tokens": 68708586.0,
      "step": 30765
    },
    {
      "entropy": 5.496584796905518,
      "epoch": 3.4557758184983434,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003845270634603186,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21199397295713424,
      "num_tokens": 68719649.0,
      "step": 30770
    },
    {
      "entropy": 5.4262556552886965,
      "epoch": 3.456337395406301,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038449199097844724,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21387517601251602,
      "num_tokens": 68729491.0,
      "step": 30775
    },
    {
      "entropy": 5.371114730834961,
      "epoch": 3.4568989723142582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003844569150103084,
      "loss": 4.9461,
      "mean_token_accuracy": 0.21702685058116913,
      "num_tokens": 68741087.0,
      "step": 30780
    },
    {
      "entropy": 5.431599521636963,
      "epoch": 3.457460549222216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003844218355570189,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22231292426586152,
      "num_tokens": 68752699.0,
      "step": 30785
    },
    {
      "entropy": 5.424840450286865,
      "epoch": 3.4580221261301736,
      "grad_norm": 1.0,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.944,
      "mean_token_accuracy": 0.2214082360267639,
      "num_tokens": 68763066.0,
      "step": 30790
    },
    {
      "entropy": 5.401832056045532,
      "epoch": 3.4585837030381312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003843516661994559,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21503449082374573,
      "num_tokens": 68773878.0,
      "step": 30795
    },
    {
      "entropy": 5.349673795700073,
      "epoch": 3.4591452799460884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003843165762974167,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21698655039072037,
      "num_tokens": 68783864.0,
      "step": 30800
    },
    {
      "entropy": 5.362696504592895,
      "epoch": 3.459706856854046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038428148291469515,
      "loss": 4.945,
      "mean_token_accuracy": 0.22170832604169846,
      "num_tokens": 68794280.0,
      "step": 30805
    },
    {
      "entropy": 5.488753843307495,
      "epoch": 3.4602684337620038,
      "grad_norm": 1.109375,
      "learning_rate": 0.000384246386052409,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2091240331530571,
      "num_tokens": 68806125.0,
      "step": 30810
    },
    {
      "entropy": 5.428026866912842,
      "epoch": 3.4608300106699614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003842112857116754,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2193845435976982,
      "num_tokens": 68816742.0,
      "step": 30815
    },
    {
      "entropy": 5.376478242874145,
      "epoch": 3.4613915875779186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003841761818936122,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2169645071029663,
      "num_tokens": 68828504.0,
      "step": 30820
    },
    {
      "entropy": 5.510113191604614,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000384141074599337,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20825639218091965,
      "num_tokens": 68839993.0,
      "step": 30825
    },
    {
      "entropy": 5.371119165420533,
      "epoch": 3.462514741393834,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.8987,
      "mean_token_accuracy": 0.2229189857840538,
      "num_tokens": 68850395.0,
      "step": 30830
    },
    {
      "entropy": 5.413512325286865,
      "epoch": 3.463076318301791,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003840708495866221,
      "loss": 4.8945,
      "mean_token_accuracy": 0.21755548864603041,
      "num_tokens": 68860621.0,
      "step": 30835
    },
    {
      "entropy": 5.383293771743775,
      "epoch": 3.463637895209749,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003840357318704183,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21234843581914903,
      "num_tokens": 68872838.0,
      "step": 30840
    },
    {
      "entropy": 5.445636844635009,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003840006106824746,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22131307721138,
      "num_tokens": 68883372.0,
      "step": 30845
    },
    {
      "entropy": 5.447606992721558,
      "epoch": 3.464761049025664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003839654860239091,
      "loss": 4.975,
      "mean_token_accuracy": 0.2181721031665802,
      "num_tokens": 68894148.0,
      "step": 30850
    },
    {
      "entropy": 5.442621564865112,
      "epoch": 3.465322625933622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003839303578958404,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21519084870815278,
      "num_tokens": 68904381.0,
      "step": 30855
    },
    {
      "entropy": 5.444068050384521,
      "epoch": 3.465884202841579,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038389522629938676,
      "loss": 5.1147,
      "mean_token_accuracy": 0.2078919768333435,
      "num_tokens": 68916354.0,
      "step": 30860
    },
    {
      "entropy": 5.371511697769165,
      "epoch": 3.4664457797495367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21838167309761047,
      "num_tokens": 68927551.0,
      "step": 30865
    },
    {
      "entropy": 5.50264253616333,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003838249527057995,
      "loss": 5.0592,
      "mean_token_accuracy": 0.21237000226974487,
      "num_tokens": 68939528.0,
      "step": 30870
    },
    {
      "entropy": 5.397599840164185,
      "epoch": 3.4675689335654516,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038378981071090346,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21957029551267623,
      "num_tokens": 68951634.0,
      "step": 30875
    },
    {
      "entropy": 5.415715742111206,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22302155792713166,
      "num_tokens": 68962259.0,
      "step": 30880
    },
    {
      "entropy": 5.441117238998413,
      "epoch": 3.468692087381367,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038371951633050106,
      "loss": 5.0695,
      "mean_token_accuracy": 0.2182522565126419,
      "num_tokens": 68973185.0,
      "step": 30885
    },
    {
      "entropy": 5.489957046508789,
      "epoch": 3.4692536642893246,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038368436394723294,
      "loss": 5.0465,
      "mean_token_accuracy": 0.212929305434227,
      "num_tokens": 68984052.0,
      "step": 30890
    },
    {
      "entropy": 5.4447794437408445,
      "epoch": 3.469815241197282,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038364920810341255,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21555945575237273,
      "num_tokens": 68993730.0,
      "step": 30895
    },
    {
      "entropy": 5.4664875030517575,
      "epoch": 3.4703768181052395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003836140488001592,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21444083601236344,
      "num_tokens": 69006229.0,
      "step": 30900
    },
    {
      "entropy": 5.428348636627197,
      "epoch": 3.470938395013197,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038357888603859236,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21534684747457505,
      "num_tokens": 69018518.0,
      "step": 30905
    },
    {
      "entropy": 5.4729437828063965,
      "epoch": 3.471499971921155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038354371981983175,
      "loss": 5.0508,
      "mean_token_accuracy": 0.2061849057674408,
      "num_tokens": 69031113.0,
      "step": 30910
    },
    {
      "entropy": 5.53304967880249,
      "epoch": 3.472061548829112,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003835085501449969,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20745449215173722,
      "num_tokens": 69042038.0,
      "step": 30915
    },
    {
      "entropy": 5.3679076671600345,
      "epoch": 3.4726231257370697,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038347337701520785,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22882012873888016,
      "num_tokens": 69053675.0,
      "step": 30920
    },
    {
      "entropy": 5.435304641723633,
      "epoch": 3.4731847026450273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038343820043158436,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21398995667696,
      "num_tokens": 69064475.0,
      "step": 30925
    },
    {
      "entropy": 5.36908106803894,
      "epoch": 3.4737462795529845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003834030203952465,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2163706675171852,
      "num_tokens": 69076069.0,
      "step": 30930
    },
    {
      "entropy": 5.470829677581787,
      "epoch": 3.474307856460942,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003833678369073144,
      "loss": 5.078,
      "mean_token_accuracy": 0.21276405751705169,
      "num_tokens": 69087144.0,
      "step": 30935
    },
    {
      "entropy": 5.403851127624511,
      "epoch": 3.4748694333689,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.9375,
      "mean_token_accuracy": 0.21580549329519272,
      "num_tokens": 69098035.0,
      "step": 30940
    },
    {
      "entropy": 5.439182901382447,
      "epoch": 3.4754310102768575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003832974595811488,
      "loss": 5.0515,
      "mean_token_accuracy": 0.21672852486371993,
      "num_tokens": 69111722.0,
      "step": 30945
    },
    {
      "entropy": 5.420649290084839,
      "epoch": 3.475992587184815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000383262265745156,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21818263679742814,
      "num_tokens": 69122228.0,
      "step": 30950
    },
    {
      "entropy": 5.4255486011505125,
      "epoch": 3.4765541640927724,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038322706846205073,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21685288697481156,
      "num_tokens": 69134150.0,
      "step": 30955
    },
    {
      "entropy": 5.346400308609009,
      "epoch": 3.47711574100073,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038319186773295364,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20927387923002244,
      "num_tokens": 69145757.0,
      "step": 30960
    },
    {
      "entropy": 5.483669710159302,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003831566635589856,
      "loss": 5.1176,
      "mean_token_accuracy": 0.2150464415550232,
      "num_tokens": 69157545.0,
      "step": 30965
    },
    {
      "entropy": 5.4539107322692875,
      "epoch": 3.478238894816645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003831214559412674,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21749078631401061,
      "num_tokens": 69169070.0,
      "step": 30970
    },
    {
      "entropy": 5.395230388641357,
      "epoch": 3.4788004717246026,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038308624488092005,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22328922897577286,
      "num_tokens": 69178903.0,
      "step": 30975
    },
    {
      "entropy": 5.315260982513427,
      "epoch": 3.4793620486325603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003830510303790647,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2185083359479904,
      "num_tokens": 69190489.0,
      "step": 30980
    },
    {
      "entropy": 5.320816421508789,
      "epoch": 3.479923625540518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038301581243682263,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21544672846794127,
      "num_tokens": 69200574.0,
      "step": 30985
    },
    {
      "entropy": 5.415171432495117,
      "epoch": 3.480485202448475,
      "grad_norm": 1.125,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22449058592319487,
      "num_tokens": 69210597.0,
      "step": 30990
    },
    {
      "entropy": 5.506035184860229,
      "epoch": 3.481046779356433,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038294536623566407,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21020745486021042,
      "num_tokens": 69221392.0,
      "step": 30995
    },
    {
      "entropy": 5.299217605590821,
      "epoch": 3.4816083562643905,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003829101379789905,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21528760194778443,
      "num_tokens": 69232825.0,
      "step": 31000
    },
    {
      "entropy": 5.35230598449707,
      "epoch": 3.482169933172348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21447892636060714,
      "num_tokens": 69243989.0,
      "step": 31005
    },
    {
      "entropy": 5.428092813491821,
      "epoch": 3.4827315100803053,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003828396711590631,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21239812672138214,
      "num_tokens": 69255828.0,
      "step": 31010
    },
    {
      "entropy": 5.458231019973755,
      "epoch": 3.483293086988263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038280443259805307,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21483456790447236,
      "num_tokens": 69266496.0,
      "step": 31015
    },
    {
      "entropy": 5.4364013671875,
      "epoch": 3.4838546638962207,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003827691906045081,
      "loss": 5.068,
      "mean_token_accuracy": 0.20956310778856277,
      "num_tokens": 69277884.0,
      "step": 31020
    },
    {
      "entropy": 5.5103111743927,
      "epoch": 3.484416240804178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003827339451795503,
      "loss": 5.1546,
      "mean_token_accuracy": 0.2057715907692909,
      "num_tokens": 69290660.0,
      "step": 31025
    },
    {
      "entropy": 5.563466310501099,
      "epoch": 3.4849778177121356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003826986963243018,
      "loss": 5.1306,
      "mean_token_accuracy": 0.21825169175863265,
      "num_tokens": 69302310.0,
      "step": 31030
    },
    {
      "entropy": 5.460573625564575,
      "epoch": 3.485539394620093,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003826634440398852,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21462943255901337,
      "num_tokens": 69313804.0,
      "step": 31035
    },
    {
      "entropy": 5.39689269065857,
      "epoch": 3.486100971528051,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21420642286539077,
      "num_tokens": 69325372.0,
      "step": 31040
    },
    {
      "entropy": 5.394873332977295,
      "epoch": 3.4866625484360085,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038259292918803696,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2129154086112976,
      "num_tokens": 69335835.0,
      "step": 31045
    },
    {
      "entropy": 5.419984626770019,
      "epoch": 3.4872241253439658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003825576666228506,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21720386296510696,
      "num_tokens": 69346895.0,
      "step": 31050
    },
    {
      "entropy": 5.432080316543579,
      "epoch": 3.4877857022519234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003825224006329864,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21676453202962875,
      "num_tokens": 69357649.0,
      "step": 31055
    },
    {
      "entropy": 5.4293121814727785,
      "epoch": 3.488347279159881,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003824871312195673,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21316093504428862,
      "num_tokens": 69367973.0,
      "step": 31060
    },
    {
      "entropy": 5.42748613357544,
      "epoch": 3.4889088560678383,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003824518583837163,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2237251654267311,
      "num_tokens": 69379411.0,
      "step": 31065
    },
    {
      "entropy": 5.489446973800659,
      "epoch": 3.489470432975796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003824165821265564,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21316212862730027,
      "num_tokens": 69391261.0,
      "step": 31070
    },
    {
      "entropy": 5.386063528060913,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22153668105602264,
      "num_tokens": 69401712.0,
      "step": 31075
    },
    {
      "entropy": 5.345809841156006,
      "epoch": 3.4905935867917113,
      "grad_norm": 1.09375,
      "learning_rate": 0.000382346019352803,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22436301857233049,
      "num_tokens": 69412612.0,
      "step": 31080
    },
    {
      "entropy": 5.4454811096191404,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003823107328384562,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21856861263513566,
      "num_tokens": 69424042.0,
      "step": 31085
    },
    {
      "entropy": 5.423348093032837,
      "epoch": 3.491716740607626,
      "grad_norm": 1.125,
      "learning_rate": 0.0003822754429072942,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21231066733598708,
      "num_tokens": 69434667.0,
      "step": 31090
    },
    {
      "entropy": 5.498107242584228,
      "epoch": 3.492278317515584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003822401495604403,
      "loss": 5.0975,
      "mean_token_accuracy": 0.20932940244674683,
      "num_tokens": 69447015.0,
      "step": 31095
    },
    {
      "entropy": 5.392149353027344,
      "epoch": 3.4928398944235415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038220485279901855,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2115733429789543,
      "num_tokens": 69457585.0,
      "step": 31100
    },
    {
      "entropy": 5.3993528366088865,
      "epoch": 3.4934014713314987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038216955262415273,
      "loss": 5.0108,
      "mean_token_accuracy": 0.2127808779478073,
      "num_tokens": 69467751.0,
      "step": 31105
    },
    {
      "entropy": 5.507439517974854,
      "epoch": 3.4939630482394564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038213424903696673,
      "loss": 5.0921,
      "mean_token_accuracy": 0.21104004383087158,
      "num_tokens": 69478898.0,
      "step": 31110
    },
    {
      "entropy": 5.487381649017334,
      "epoch": 3.494524625147414,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038209894203858477,
      "loss": 5.0496,
      "mean_token_accuracy": 0.2152410015463829,
      "num_tokens": 69488367.0,
      "step": 31115
    },
    {
      "entropy": 5.403996849060059,
      "epoch": 3.4950862020553712,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038206363163013093,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21753977686166764,
      "num_tokens": 69500446.0,
      "step": 31120
    },
    {
      "entropy": 5.445170927047729,
      "epoch": 3.495647778963329,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038202831781272944,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21119256168603898,
      "num_tokens": 69512383.0,
      "step": 31125
    },
    {
      "entropy": 5.425660610198975,
      "epoch": 3.4962093558712866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2234874337911606,
      "num_tokens": 69523238.0,
      "step": 31130
    },
    {
      "entropy": 5.5791748046875,
      "epoch": 3.4967709327792442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038195767995558174,
      "loss": 5.1465,
      "mean_token_accuracy": 0.20802054554224014,
      "num_tokens": 69535196.0,
      "step": 31135
    },
    {
      "entropy": 5.458070182800293,
      "epoch": 3.497332509687202,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22067631036043167,
      "num_tokens": 69546896.0,
      "step": 31140
    },
    {
      "entropy": 5.340889978408813,
      "epoch": 3.497894086595159,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000381887028476138,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21409590393304825,
      "num_tokens": 69558230.0,
      "step": 31145
    },
    {
      "entropy": 5.337574863433838,
      "epoch": 3.4984556635031168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038185169763086703,
      "loss": 4.954,
      "mean_token_accuracy": 0.22123309075832367,
      "num_tokens": 69568891.0,
      "step": 31150
    },
    {
      "entropy": 5.4423729419708256,
      "epoch": 3.4990172404110744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003818163633833966,
      "loss": 5.0597,
      "mean_token_accuracy": 0.2114477887749672,
      "num_tokens": 69579988.0,
      "step": 31155
    },
    {
      "entropy": 5.409789609909057,
      "epoch": 3.4995788173190316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2182759538292885,
      "num_tokens": 69590714.0,
      "step": 31160
    },
    {
      "entropy": 5.368581771850586,
      "epoch": 3.5001403942269893,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003817456846863575,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2292957827448845,
      "num_tokens": 69601755.0,
      "step": 31165
    },
    {
      "entropy": 5.389016246795654,
      "epoch": 3.500701971134947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21880873590707778,
      "num_tokens": 69612862.0,
      "step": 31170
    },
    {
      "entropy": 5.517245149612426,
      "epoch": 3.5012635480429046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003816749923940226,
      "loss": 5.0541,
      "mean_token_accuracy": 0.2140080899000168,
      "num_tokens": 69624746.0,
      "step": 31175
    },
    {
      "entropy": 5.421370983123779,
      "epoch": 3.5018251249508623,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038163964115243276,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22295431792736053,
      "num_tokens": 69634477.0,
      "step": 31180
    },
    {
      "entropy": 5.398225021362305,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038160428651539535,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2136620745062828,
      "num_tokens": 69645913.0,
      "step": 31185
    },
    {
      "entropy": 5.374249887466431,
      "epoch": 3.502948278766777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038156892848403606,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21470312476158143,
      "num_tokens": 69657727.0,
      "step": 31190
    },
    {
      "entropy": 5.314769983291626,
      "epoch": 3.503509855674735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003815335670594808,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24425848871469497,
      "num_tokens": 69668649.0,
      "step": 31195
    },
    {
      "entropy": 5.433824157714843,
      "epoch": 3.504071432582692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038149820224285546,
      "loss": 5.0507,
      "mean_token_accuracy": 0.21189334988594055,
      "num_tokens": 69680286.0,
      "step": 31200
    },
    {
      "entropy": 5.42534646987915,
      "epoch": 3.5046330094906497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000381462834035286,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21395964920520782,
      "num_tokens": 69689801.0,
      "step": 31205
    },
    {
      "entropy": 5.41913423538208,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22003531605005264,
      "num_tokens": 69700898.0,
      "step": 31210
    },
    {
      "entropy": 5.413319301605225,
      "epoch": 3.5057561633065646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2231561005115509,
      "num_tokens": 69711969.0,
      "step": 31215
    },
    {
      "entropy": 5.439508390426636,
      "epoch": 3.5063177402145222,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003813567090781752,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21809143275022508,
      "num_tokens": 69723458.0,
      "step": 31220
    },
    {
      "entropy": 5.404107999801636,
      "epoch": 3.50687931712248,
      "grad_norm": 0.984375,
      "learning_rate": 0.000381321327318092,
      "loss": 4.987,
      "mean_token_accuracy": 0.21304131895303727,
      "num_tokens": 69735527.0,
      "step": 31225
    },
    {
      "entropy": 5.498193836212158,
      "epoch": 3.5074408940304376,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003812859421726964,
      "loss": 5.0982,
      "mean_token_accuracy": 0.21628180742263795,
      "num_tokens": 69748829.0,
      "step": 31230
    },
    {
      "entropy": 5.408562278747558,
      "epoch": 3.5080024709383952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038125055364311526,
      "loss": 4.891,
      "mean_token_accuracy": 0.2192994087934494,
      "num_tokens": 69758970.0,
      "step": 31235
    },
    {
      "entropy": 5.438550758361816,
      "epoch": 3.5085640478463525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003812151617304753,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21217085868120195,
      "num_tokens": 69770082.0,
      "step": 31240
    },
    {
      "entropy": 5.403321313858032,
      "epoch": 3.50912562475431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2124886229634285,
      "num_tokens": 69781386.0,
      "step": 31245
    },
    {
      "entropy": 5.467429065704346,
      "epoch": 3.5096872016622678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038114436776052644,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2138720527291298,
      "num_tokens": 69792571.0,
      "step": 31250
    },
    {
      "entropy": 5.4944905757904055,
      "epoch": 3.510248778570225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003811089657054717,
      "loss": 5.0643,
      "mean_token_accuracy": 0.2161405488848686,
      "num_tokens": 69804737.0,
      "step": 31255
    },
    {
      "entropy": 5.476451301574707,
      "epoch": 3.5108103554781827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2171425700187683,
      "num_tokens": 69816584.0,
      "step": 31260
    },
    {
      "entropy": 5.321476984024048,
      "epoch": 3.5113719323861403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003810381514608378,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23117675483226777,
      "num_tokens": 69828657.0,
      "step": 31265
    },
    {
      "entropy": 5.4390788078308105,
      "epoch": 3.511933509294098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003810027392735133,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21632836014032364,
      "num_tokens": 69839780.0,
      "step": 31270
    },
    {
      "entropy": 5.438322639465332,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003809673237110204,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21740121394395828,
      "num_tokens": 69850452.0,
      "step": 31275
    },
    {
      "entropy": 5.436842632293701,
      "epoch": 3.513056663110013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003809319047744869,
      "loss": 5.046,
      "mean_token_accuracy": 0.21542335897684098,
      "num_tokens": 69861840.0,
      "step": 31280
    },
    {
      "entropy": 5.4727802753448485,
      "epoch": 3.5136182400179705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.9848,
      "mean_token_accuracy": 0.21578046530485154,
      "num_tokens": 69872653.0,
      "step": 31285
    },
    {
      "entropy": 5.3888819217681885,
      "epoch": 3.514179816925928,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038086105678380896,
      "loss": 5.026,
      "mean_token_accuracy": 0.21862777471542358,
      "num_tokens": 69883381.0,
      "step": 31290
    },
    {
      "entropy": 5.480158281326294,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038082562773192026,
      "loss": 5.1066,
      "mean_token_accuracy": 0.20951550155878068,
      "num_tokens": 69894732.0,
      "step": 31295
    },
    {
      "entropy": 5.533979177474976,
      "epoch": 3.515302970741843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038079019531050257,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21539557725191116,
      "num_tokens": 69905645.0,
      "step": 31300
    },
    {
      "entropy": 5.508505201339721,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038075475952068405,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21099125146865844,
      "num_tokens": 69917176.0,
      "step": 31305
    },
    {
      "entropy": 5.461016511917114,
      "epoch": 3.516426124557758,
      "grad_norm": 1.0625,
      "learning_rate": 0.000380719320363593,
      "loss": 5.1036,
      "mean_token_accuracy": 0.21461637318134308,
      "num_tokens": 69928923.0,
      "step": 31310
    },
    {
      "entropy": 5.373480033874512,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038068387784035774,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2151376411318779,
      "num_tokens": 69938981.0,
      "step": 31315
    },
    {
      "entropy": 5.384507465362549,
      "epoch": 3.5175492783736733,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003806484319521069,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21687886565923692,
      "num_tokens": 69949265.0,
      "step": 31320
    },
    {
      "entropy": 5.340601062774658,
      "epoch": 3.518110855281631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038061298269996893,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22113083004951478,
      "num_tokens": 69961399.0,
      "step": 31325
    },
    {
      "entropy": 5.428219747543335,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2182676985859871,
      "num_tokens": 69971601.0,
      "step": 31330
    },
    {
      "entropy": 5.377566576004028,
      "epoch": 3.519234009097546,
      "grad_norm": 1.125,
      "learning_rate": 0.0003805420741085469,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21151263117790223,
      "num_tokens": 69982375.0,
      "step": 31335
    },
    {
      "entropy": 5.445303678512573,
      "epoch": 3.5197955860055035,
      "grad_norm": 1.125,
      "learning_rate": 0.0003805066147715205,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21879090070724488,
      "num_tokens": 69994787.0,
      "step": 31340
    },
    {
      "entropy": 5.53015775680542,
      "epoch": 3.520357162913461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003804711520751226,
      "loss": 5.1745,
      "mean_token_accuracy": 0.20432191491127014,
      "num_tokens": 70006077.0,
      "step": 31345
    },
    {
      "entropy": 5.441598987579345,
      "epoch": 3.5209187398214183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038043568602048227,
      "loss": 4.978,
      "mean_token_accuracy": 0.21724890619516374,
      "num_tokens": 70016655.0,
      "step": 31350
    },
    {
      "entropy": 5.454234504699707,
      "epoch": 3.521480316729376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003804002166087289,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22453716695308684,
      "num_tokens": 70026857.0,
      "step": 31355
    },
    {
      "entropy": 5.388972806930542,
      "epoch": 3.5220418936373337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2190498799085617,
      "num_tokens": 70038311.0,
      "step": 31360
    },
    {
      "entropy": 5.344569683074951,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003803292677184002,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21984886676073073,
      "num_tokens": 70048613.0,
      "step": 31365
    },
    {
      "entropy": 5.4831862449646,
      "epoch": 3.523165047453249,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038029378824208394,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21814693361520768,
      "num_tokens": 70058686.0,
      "step": 31370
    },
    {
      "entropy": 5.492212915420533,
      "epoch": 3.523726624361206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003802583054131727,
      "loss": 5.0456,
      "mean_token_accuracy": 0.21397042125463486,
      "num_tokens": 70068845.0,
      "step": 31375
    },
    {
      "entropy": 5.378587007522583,
      "epoch": 3.524288201269164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003802228192327961,
      "loss": 4.98,
      "mean_token_accuracy": 0.21384042054414748,
      "num_tokens": 70080223.0,
      "step": 31380
    },
    {
      "entropy": 5.428745174407959,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038018732970208414,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21147552132606506,
      "num_tokens": 70091729.0,
      "step": 31385
    },
    {
      "entropy": 5.3950769901275635,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003801518368221668,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2114290326833725,
      "num_tokens": 70103172.0,
      "step": 31390
    },
    {
      "entropy": 5.463821506500244,
      "epoch": 3.5259729319930364,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003801163405941743,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20802272409200667,
      "num_tokens": 70114303.0,
      "step": 31395
    },
    {
      "entropy": 5.36176323890686,
      "epoch": 3.526534508900994,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038008084101923665,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22090613543987275,
      "num_tokens": 70126802.0,
      "step": 31400
    },
    {
      "entropy": 5.504707670211792,
      "epoch": 3.5270960858089513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038004533809848435,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21766813695430756,
      "num_tokens": 70138768.0,
      "step": 31405
    },
    {
      "entropy": 5.399847507476807,
      "epoch": 3.527657662716909,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038000983183304765,
      "loss": 4.977,
      "mean_token_accuracy": 0.21710626631975175,
      "num_tokens": 70148411.0,
      "step": 31410
    },
    {
      "entropy": 5.349059391021728,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003799743222240572,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2210240676999092,
      "num_tokens": 70159788.0,
      "step": 31415
    },
    {
      "entropy": 5.41290431022644,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003799388092726436,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21792843490839003,
      "num_tokens": 70169707.0,
      "step": 31420
    },
    {
      "entropy": 5.407720279693604,
      "epoch": 3.529342393440782,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037990329297993763,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2200486660003662,
      "num_tokens": 70180958.0,
      "step": 31425
    },
    {
      "entropy": 5.322078084945678,
      "epoch": 3.529903970348739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21585820317268373,
      "num_tokens": 70192848.0,
      "step": 31430
    },
    {
      "entropy": 5.383045482635498,
      "epoch": 3.530465547256697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037983225037517207,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2129748523235321,
      "num_tokens": 70204034.0,
      "step": 31435
    },
    {
      "entropy": 5.458120536804199,
      "epoch": 3.5310271241646545,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037979672406537445,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21414953023195266,
      "num_tokens": 70214274.0,
      "step": 31440
    },
    {
      "entropy": 5.48022871017456,
      "epoch": 3.5315887010726117,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003797611944188086,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2123296543955803,
      "num_tokens": 70223890.0,
      "step": 31445
    },
    {
      "entropy": 5.496121168136597,
      "epoch": 3.5321502779805694,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037972566143660566,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21244687885046004,
      "num_tokens": 70234529.0,
      "step": 31450
    },
    {
      "entropy": 5.328721523284912,
      "epoch": 3.532711854888527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003796901251198969,
      "loss": 4.947,
      "mean_token_accuracy": 0.21929830312728882,
      "num_tokens": 70246900.0,
      "step": 31455
    },
    {
      "entropy": 5.4987688064575195,
      "epoch": 3.5332734317964847,
      "grad_norm": 1.0625,
      "learning_rate": 0.000379654585469814,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21223898380994796,
      "num_tokens": 70257942.0,
      "step": 31460
    },
    {
      "entropy": 5.529710578918457,
      "epoch": 3.5338350087044423,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003796190424874886,
      "loss": 5.0574,
      "mean_token_accuracy": 0.21528240591287612,
      "num_tokens": 70270297.0,
      "step": 31465
    },
    {
      "entropy": 5.47795615196228,
      "epoch": 3.5343965856123996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003795834961740523,
      "loss": 5.0732,
      "mean_token_accuracy": 0.2163182482123375,
      "num_tokens": 70281542.0,
      "step": 31470
    },
    {
      "entropy": 5.385998487472534,
      "epoch": 3.534958162520357,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037954794653063695,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21552914679050444,
      "num_tokens": 70290957.0,
      "step": 31475
    },
    {
      "entropy": 5.4355686664581295,
      "epoch": 3.535519739428315,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003795123935583743,
      "loss": 5.0233,
      "mean_token_accuracy": 0.22180090695619584,
      "num_tokens": 70302796.0,
      "step": 31480
    },
    {
      "entropy": 5.444584941864013,
      "epoch": 3.536081316336272,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003794768372583966,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21688868552446366,
      "num_tokens": 70313709.0,
      "step": 31485
    },
    {
      "entropy": 5.440577507019043,
      "epoch": 3.5366428932442298,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037944127763183587,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2151051551103592,
      "num_tokens": 70324651.0,
      "step": 31490
    },
    {
      "entropy": 5.51100754737854,
      "epoch": 3.5372044701521874,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037940571467982436,
      "loss": 5.135,
      "mean_token_accuracy": 0.20490317344665526,
      "num_tokens": 70335810.0,
      "step": 31495
    },
    {
      "entropy": 5.411005544662475,
      "epoch": 3.5377660470601446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22378795593976974,
      "num_tokens": 70347171.0,
      "step": 31500
    },
    {
      "entropy": 5.405158376693725,
      "epoch": 3.5383276239681023,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037933457880397843,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2243108868598938,
      "num_tokens": 70359122.0,
      "step": 31505
    },
    {
      "entropy": 5.292706060409546,
      "epoch": 3.53888920087606,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2232506662607193,
      "num_tokens": 70369691.0,
      "step": 31510
    },
    {
      "entropy": 5.390719699859619,
      "epoch": 3.5394507777840176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003792634296399187,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21612572371959687,
      "num_tokens": 70379602.0,
      "step": 31515
    },
    {
      "entropy": 5.42822961807251,
      "epoch": 3.5400123546919753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037922785007764043,
      "loss": 5.046,
      "mean_token_accuracy": 0.20529057830572128,
      "num_tokens": 70391819.0,
      "step": 31520
    },
    {
      "entropy": 5.533569240570069,
      "epoch": 3.5405739315999325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000379192267196707,
      "loss": 5.061,
      "mean_token_accuracy": 0.21662346571683883,
      "num_tokens": 70402078.0,
      "step": 31525
    },
    {
      "entropy": 5.378169822692871,
      "epoch": 3.54113550850789,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037915668099825135,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21299780756235123,
      "num_tokens": 70413829.0,
      "step": 31530
    },
    {
      "entropy": 5.295250177383423,
      "epoch": 3.541697085415848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037912109148340663,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21641303598880768,
      "num_tokens": 70426221.0,
      "step": 31535
    },
    {
      "entropy": 5.492299652099609,
      "epoch": 3.542258662323805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037908549865330596,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20914115011692047,
      "num_tokens": 70437215.0,
      "step": 31540
    },
    {
      "entropy": 5.424953746795654,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22384593188762664,
      "num_tokens": 70448918.0,
      "step": 31545
    },
    {
      "entropy": 5.425614738464356,
      "epoch": 3.5433818161397204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037901430305187,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21310271620750426,
      "num_tokens": 70460713.0,
      "step": 31550
    },
    {
      "entropy": 5.396354007720947,
      "epoch": 3.543943393047678,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2137720763683319,
      "num_tokens": 70471460.0,
      "step": 31555
    },
    {
      "entropy": 5.393932437896728,
      "epoch": 3.5445049699556357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003789430942030112,
      "loss": 5.002,
      "mean_token_accuracy": 0.21845425218343734,
      "num_tokens": 70482359.0,
      "step": 31560
    },
    {
      "entropy": 5.463444948196411,
      "epoch": 3.545066546863593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037890748481363216,
      "loss": 5.075,
      "mean_token_accuracy": 0.2097043588757515,
      "num_tokens": 70493272.0,
      "step": 31565
    },
    {
      "entropy": 5.401187086105347,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003788718721157987,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2186749592423439,
      "num_tokens": 70504118.0,
      "step": 31570
    },
    {
      "entropy": 5.503695869445801,
      "epoch": 3.5461897006795082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003788362561106445,
      "loss": 5.1133,
      "mean_token_accuracy": 0.21260030716657638,
      "num_tokens": 70515379.0,
      "step": 31575
    },
    {
      "entropy": 5.481404256820679,
      "epoch": 3.5467512775874654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037880063679930355,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21417918503284455,
      "num_tokens": 70526047.0,
      "step": 31580
    },
    {
      "entropy": 5.457152318954468,
      "epoch": 3.547312854495423,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.9833,
      "mean_token_accuracy": 0.2243720918893814,
      "num_tokens": 70537614.0,
      "step": 31585
    },
    {
      "entropy": 5.440493726730347,
      "epoch": 3.5478744314033808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003787293882625984,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21969517171382905,
      "num_tokens": 70548048.0,
      "step": 31590
    },
    {
      "entropy": 5.388954448699951,
      "epoch": 3.548436008311338,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22555551528930665,
      "num_tokens": 70558668.0,
      "step": 31595
    },
    {
      "entropy": 5.402224779129028,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003786581265147575,
      "loss": 5.083,
      "mean_token_accuracy": 0.20948696583509446,
      "num_tokens": 70570044.0,
      "step": 31600
    },
    {
      "entropy": 5.388058137893677,
      "epoch": 3.5495591621272533,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22623758316040038,
      "num_tokens": 70580857.0,
      "step": 31605
    },
    {
      "entropy": 5.352005481719971,
      "epoch": 3.550120739035211,
      "grad_norm": 1.046875,
      "learning_rate": 0.000378586851564857,
      "loss": 4.8717,
      "mean_token_accuracy": 0.22286927998065947,
      "num_tokens": 70592476.0,
      "step": 31610
    },
    {
      "entropy": 5.376699638366699,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.078125,
      "learning_rate": 0.000378551209141971,
      "loss": 4.9524,
      "mean_token_accuracy": 0.21791828274726868,
      "num_tokens": 70603614.0,
      "step": 31615
    },
    {
      "entropy": 5.417070484161377,
      "epoch": 3.551243892851126,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003785155634219744,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20742738097906113,
      "num_tokens": 70615239.0,
      "step": 31620
    },
    {
      "entropy": 5.459246492385864,
      "epoch": 3.5518054697590835,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003784799144060021,
      "loss": 4.999,
      "mean_token_accuracy": 0.2145635962486267,
      "num_tokens": 70625785.0,
      "step": 31625
    },
    {
      "entropy": 5.4860893249511715,
      "epoch": 3.552367046667041,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037844426209518916,
      "loss": 5.0302,
      "mean_token_accuracy": 0.21509782820940018,
      "num_tokens": 70636942.0,
      "step": 31630
    },
    {
      "entropy": 5.400440788269043,
      "epoch": 3.5529286235749984,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037840860649067084,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22137968987226486,
      "num_tokens": 70648106.0,
      "step": 31635
    },
    {
      "entropy": 5.389050006866455,
      "epoch": 3.553490200482956,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037837294759358235,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21786127239465714,
      "num_tokens": 70658850.0,
      "step": 31640
    },
    {
      "entropy": 5.511782550811768,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037833728540505914,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20641817301511764,
      "num_tokens": 70671060.0,
      "step": 31645
    },
    {
      "entropy": 5.359528017044068,
      "epoch": 3.5546133542988714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003783016199262367,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2217835456132889,
      "num_tokens": 70682186.0,
      "step": 31650
    },
    {
      "entropy": 5.370401811599732,
      "epoch": 3.555174931206829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003782659511582505,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2151901662349701,
      "num_tokens": 70692622.0,
      "step": 31655
    },
    {
      "entropy": 5.3549150943756105,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003782302791022365,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22170427590608596,
      "num_tokens": 70703293.0,
      "step": 31660
    },
    {
      "entropy": 5.33563871383667,
      "epoch": 3.556298085022744,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037819460375933034,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2189187929034233,
      "num_tokens": 70714961.0,
      "step": 31665
    },
    {
      "entropy": 5.504412126541138,
      "epoch": 3.5568596619307016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003781589251306679,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21157828271389006,
      "num_tokens": 70727030.0,
      "step": 31670
    },
    {
      "entropy": 5.466444206237793,
      "epoch": 3.557421238838659,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003781232432173853,
      "loss": 4.962,
      "mean_token_accuracy": 0.21582914441823958,
      "num_tokens": 70738331.0,
      "step": 31675
    },
    {
      "entropy": 5.4191264629364015,
      "epoch": 3.5579828157466165,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037808755802061874,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20953722596168517,
      "num_tokens": 70748347.0,
      "step": 31680
    },
    {
      "entropy": 5.4565976619720455,
      "epoch": 3.558544392654574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003780518695415043,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2195180580019951,
      "num_tokens": 70759905.0,
      "step": 31685
    },
    {
      "entropy": 5.5243126392364506,
      "epoch": 3.5591059695625313,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037801617778117835,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21880548745393752,
      "num_tokens": 70770538.0,
      "step": 31690
    },
    {
      "entropy": 5.442565250396728,
      "epoch": 3.559667546470489,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037798048274077735,
      "loss": 5.057,
      "mean_token_accuracy": 0.21674568951129913,
      "num_tokens": 70781881.0,
      "step": 31695
    },
    {
      "entropy": 5.293679189682007,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22486587315797807,
      "num_tokens": 70792299.0,
      "step": 31700
    },
    {
      "entropy": 5.4038104057312015,
      "epoch": 3.5607907002864043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037790908282429647,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2184235632419586,
      "num_tokens": 70802470.0,
      "step": 31705
    },
    {
      "entropy": 5.4951988697052006,
      "epoch": 3.561352277194362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037787337795049,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21699981540441513,
      "num_tokens": 70813502.0,
      "step": 31710
    },
    {
      "entropy": 5.426673936843872,
      "epoch": 3.561913854102319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037783766980115525,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22016734033823013,
      "num_tokens": 70823917.0,
      "step": 31715
    },
    {
      "entropy": 5.427173042297364,
      "epoch": 3.562475431010277,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037780195837742927,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20950037240982056,
      "num_tokens": 70834503.0,
      "step": 31720
    },
    {
      "entropy": 5.426802349090576,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.9512,
      "mean_token_accuracy": 0.21924301236867905,
      "num_tokens": 70844954.0,
      "step": 31725
    },
    {
      "entropy": 5.514315223693847,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037773052571135157,
      "loss": 5.0959,
      "mean_token_accuracy": 0.2129180058836937,
      "num_tokens": 70856142.0,
      "step": 31730
    },
    {
      "entropy": 5.442246437072754,
      "epoch": 3.5641601617341494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003776948044712745,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22063761800527573,
      "num_tokens": 70866880.0,
      "step": 31735
    },
    {
      "entropy": 5.435084819793701,
      "epoch": 3.564721738642107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000377659079961355,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21995027214288712,
      "num_tokens": 70876836.0,
      "step": 31740
    },
    {
      "entropy": 5.384940576553345,
      "epoch": 3.5652833155500647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037762335218273057,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2113870844244957,
      "num_tokens": 70889430.0,
      "step": 31745
    },
    {
      "entropy": 5.344456386566162,
      "epoch": 3.5658448924580224,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003775876211365388,
      "loss": 4.8868,
      "mean_token_accuracy": 0.2248113602399826,
      "num_tokens": 70899291.0,
      "step": 31750
    },
    {
      "entropy": 5.429926204681396,
      "epoch": 3.5664064693659796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003775518868239174,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21924586594104767,
      "num_tokens": 70910596.0,
      "step": 31755
    },
    {
      "entropy": 5.446308517456055,
      "epoch": 3.5669680462739373,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037751614924600417,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21745488047599792,
      "num_tokens": 70921969.0,
      "step": 31760
    },
    {
      "entropy": 5.392602586746216,
      "epoch": 3.567529623181895,
      "grad_norm": 1.03125,
      "learning_rate": 0.000377480408403937,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21694846749305724,
      "num_tokens": 70933372.0,
      "step": 31765
    },
    {
      "entropy": 5.486953544616699,
      "epoch": 3.568091200089852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.1125,
      "mean_token_accuracy": 0.21130552142858505,
      "num_tokens": 70944989.0,
      "step": 31770
    },
    {
      "entropy": 5.43502516746521,
      "epoch": 3.56865277699781,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037740891693189286,
      "loss": 5.0428,
      "mean_token_accuracy": 0.21377993524074554,
      "num_tokens": 70957516.0,
      "step": 31775
    },
    {
      "entropy": 5.524496364593506,
      "epoch": 3.5692143539057675,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003773731663041922,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20847266763448716,
      "num_tokens": 70969485.0,
      "step": 31780
    },
    {
      "entropy": 5.427566194534302,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037733741241689024,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21719863563776015,
      "num_tokens": 70980335.0,
      "step": 31785
    },
    {
      "entropy": 5.382401752471924,
      "epoch": 3.5703375077216823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003773016552711254,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21710527837276458,
      "num_tokens": 70991870.0,
      "step": 31790
    },
    {
      "entropy": 5.453851270675659,
      "epoch": 3.57089908462964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037726589486803617,
      "loss": 5.1098,
      "mean_token_accuracy": 0.21131422221660615,
      "num_tokens": 71003906.0,
      "step": 31795
    },
    {
      "entropy": 5.347610950469971,
      "epoch": 3.5714606615375977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037723013120876113,
      "loss": 4.961,
      "mean_token_accuracy": 0.21861137598752975,
      "num_tokens": 71015280.0,
      "step": 31800
    },
    {
      "entropy": 5.447812557220459,
      "epoch": 3.5720222384455553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037719436429443915,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21762148588895797,
      "num_tokens": 71026299.0,
      "step": 31805
    },
    {
      "entropy": 5.522577714920044,
      "epoch": 3.5725838153535125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037715859412620895,
      "loss": 5.0938,
      "mean_token_accuracy": 0.2134285181760788,
      "num_tokens": 71037704.0,
      "step": 31810
    },
    {
      "entropy": 5.483775806427002,
      "epoch": 3.57314539226147,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003771228207052095,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21579079180955887,
      "num_tokens": 71048359.0,
      "step": 31815
    },
    {
      "entropy": 5.397462272644043,
      "epoch": 3.573706969169428,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037708704403257974,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22128991782665253,
      "num_tokens": 71059723.0,
      "step": 31820
    },
    {
      "entropy": 5.297647953033447,
      "epoch": 3.574268546077385,
      "grad_norm": 1.0,
      "learning_rate": 0.000377051264109459,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22576578706502914,
      "num_tokens": 71071437.0,
      "step": 31825
    },
    {
      "entropy": 5.364194631576538,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22735091298818588,
      "num_tokens": 71082544.0,
      "step": 31830
    },
    {
      "entropy": 5.433082103729248,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037697969451630135,
      "loss": 4.949,
      "mean_token_accuracy": 0.2172586739063263,
      "num_tokens": 71092429.0,
      "step": 31835
    },
    {
      "entropy": 5.383070659637451,
      "epoch": 3.575953276801258,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037694390484854323,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21397877782583236,
      "num_tokens": 71103154.0,
      "step": 31840
    },
    {
      "entropy": 5.363198947906494,
      "epoch": 3.5765148537092157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22442761063575745,
      "num_tokens": 71112921.0,
      "step": 31845
    },
    {
      "entropy": 5.413502788543701,
      "epoch": 3.577076430617173,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037687231577636626,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2184414193034172,
      "num_tokens": 71123281.0,
      "step": 31850
    },
    {
      "entropy": 5.495950078964233,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.1088,
      "mean_token_accuracy": 0.21119676679372787,
      "num_tokens": 71134358.0,
      "step": 31855
    },
    {
      "entropy": 5.488958311080933,
      "epoch": 3.5781995844330883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037680071372957327,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21348364651203156,
      "num_tokens": 71146805.0,
      "step": 31860
    },
    {
      "entropy": 5.41690845489502,
      "epoch": 3.5787611613410455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037676490784354543,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21579023748636245,
      "num_tokens": 71157388.0,
      "step": 31865
    },
    {
      "entropy": 5.266529655456543,
      "epoch": 3.579322738249003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003767290987172835,
      "loss": 4.8165,
      "mean_token_accuracy": 0.22614523768424988,
      "num_tokens": 71168601.0,
      "step": 31870
    },
    {
      "entropy": 5.396457290649414,
      "epoch": 3.579884315156961,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037669328635192757,
      "loss": 5.007,
      "mean_token_accuracy": 0.21746283918619155,
      "num_tokens": 71180796.0,
      "step": 31875
    },
    {
      "entropy": 5.3838005542755125,
      "epoch": 3.580445892064918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000376657470748618,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22416343092918395,
      "num_tokens": 71191686.0,
      "step": 31880
    },
    {
      "entropy": 5.344283246994019,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.8155,
      "mean_token_accuracy": 0.22537081688642502,
      "num_tokens": 71202454.0,
      "step": 31885
    },
    {
      "entropy": 5.37074556350708,
      "epoch": 3.5815690458808334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003765858298326994,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2254362851381302,
      "num_tokens": 71213729.0,
      "step": 31890
    },
    {
      "entropy": 5.514472436904907,
      "epoch": 3.582130622788791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037655000452237154,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20518959313631058,
      "num_tokens": 71226810.0,
      "step": 31895
    },
    {
      "entropy": 5.365299701690674,
      "epoch": 3.5826921996967487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003765141759786521,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2276046931743622,
      "num_tokens": 71237233.0,
      "step": 31900
    },
    {
      "entropy": 5.435438108444214,
      "epoch": 3.583253776604706,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037647834420268195,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22716102749109268,
      "num_tokens": 71248678.0,
      "step": 31905
    },
    {
      "entropy": 5.297901391983032,
      "epoch": 3.5838153535126636,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22777326554059982,
      "num_tokens": 71259098.0,
      "step": 31910
    },
    {
      "entropy": 5.328636360168457,
      "epoch": 3.584376930420621,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037640667095855307,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22468726336956024,
      "num_tokens": 71270082.0,
      "step": 31915
    },
    {
      "entropy": 5.417966747283936,
      "epoch": 3.5849385073285784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037637082949267654,
      "loss": 5.0047,
      "mean_token_accuracy": 0.22568245977163315,
      "num_tokens": 71280975.0,
      "step": 31920
    },
    {
      "entropy": 5.432625579833984,
      "epoch": 3.585500084236536,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003763349847991134,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2266353264451027,
      "num_tokens": 71292552.0,
      "step": 31925
    },
    {
      "entropy": 5.4968897819519045,
      "epoch": 3.5860616611444938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000376299136879005,
      "loss": 5.0895,
      "mean_token_accuracy": 0.20673315674066545,
      "num_tokens": 71304653.0,
      "step": 31930
    },
    {
      "entropy": 5.42550802230835,
      "epoch": 3.5866232380524514,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037626328573349296,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21750697791576384,
      "num_tokens": 71316734.0,
      "step": 31935
    },
    {
      "entropy": 5.532882452011108,
      "epoch": 3.587184814960409,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003762274313637185,
      "loss": 5.1632,
      "mean_token_accuracy": 0.20454374253749846,
      "num_tokens": 71328228.0,
      "step": 31940
    },
    {
      "entropy": 5.445057725906372,
      "epoch": 3.5877463918683663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003761915737708234,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2172890469431877,
      "num_tokens": 71339921.0,
      "step": 31945
    },
    {
      "entropy": 5.472068452835083,
      "epoch": 3.588307968776324,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037615571295594936,
      "loss": 5.0194,
      "mean_token_accuracy": 0.216056127846241,
      "num_tokens": 71350995.0,
      "step": 31950
    },
    {
      "entropy": 5.361673736572266,
      "epoch": 3.5888695456842816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003761198489202382,
      "loss": 5.01,
      "mean_token_accuracy": 0.2135254770517349,
      "num_tokens": 71362546.0,
      "step": 31955
    },
    {
      "entropy": 5.411526679992676,
      "epoch": 3.589431122592239,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003760839816648317,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2183091789484024,
      "num_tokens": 71373431.0,
      "step": 31960
    },
    {
      "entropy": 5.489752674102784,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037604811119087213,
      "loss": 5.0291,
      "mean_token_accuracy": 0.22437189817428588,
      "num_tokens": 71384488.0,
      "step": 31965
    },
    {
      "entropy": 5.467682218551635,
      "epoch": 3.590554276408154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003760122374995015,
      "loss": 5.0135,
      "mean_token_accuracy": 0.2187003508210182,
      "num_tokens": 71396558.0,
      "step": 31970
    },
    {
      "entropy": 5.360714864730835,
      "epoch": 3.5911158533161114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22387494891881943,
      "num_tokens": 71407107.0,
      "step": 31975
    },
    {
      "entropy": 5.457506608963013,
      "epoch": 3.591677430224069,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037594048046909616,
      "loss": 5.0495,
      "mean_token_accuracy": 0.20773229300975798,
      "num_tokens": 71419801.0,
      "step": 31980
    },
    {
      "entropy": 5.499556398391723,
      "epoch": 3.5922390071320267,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003759045971323462,
      "loss": 5.1188,
      "mean_token_accuracy": 0.21064799129962922,
      "num_tokens": 71431740.0,
      "step": 31985
    },
    {
      "entropy": 5.461505889892578,
      "epoch": 3.5928005840399844,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037586871058275476,
      "loss": 5.0179,
      "mean_token_accuracy": 0.21560925990343094,
      "num_tokens": 71443514.0,
      "step": 31990
    },
    {
      "entropy": 5.400231218338012,
      "epoch": 3.593362160947942,
      "grad_norm": 1.0,
      "learning_rate": 0.00037583282082146434,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21625842452049254,
      "num_tokens": 71454240.0,
      "step": 31995
    },
    {
      "entropy": 5.390089511871338,
      "epoch": 3.5939237378558992,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2199113681912422,
      "num_tokens": 71464707.0,
      "step": 32000
    },
    {
      "entropy": 5.4472901821136475,
      "epoch": 3.594485314763857,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003757610316683581,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20931144207715988,
      "num_tokens": 71475931.0,
      "step": 32005
    },
    {
      "entropy": 5.479285049438476,
      "epoch": 3.5950468916718146,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037572513227882807,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2100420355796814,
      "num_tokens": 71487851.0,
      "step": 32010
    },
    {
      "entropy": 5.486244869232178,
      "epoch": 3.595608468579772,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037568922968217065,
      "loss": 5.0288,
      "mean_token_accuracy": 0.22106293737888336,
      "num_tokens": 71498144.0,
      "step": 32015
    },
    {
      "entropy": 5.422738361358642,
      "epoch": 3.5961700454877294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037565332387952914,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21313035786151885,
      "num_tokens": 71508074.0,
      "step": 32020
    },
    {
      "entropy": 5.394605255126953,
      "epoch": 3.596731622395687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21985551416873933,
      "num_tokens": 71519609.0,
      "step": 32025
    },
    {
      "entropy": 5.327208614349365,
      "epoch": 3.5972931993036448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003755815026608669,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22790366560220718,
      "num_tokens": 71530256.0,
      "step": 32030
    },
    {
      "entropy": 5.47993688583374,
      "epoch": 3.5978547762116024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037554558724713305,
      "loss": 5.005,
      "mean_token_accuracy": 0.21329699456691742,
      "num_tokens": 71540383.0,
      "step": 32035
    },
    {
      "entropy": 5.458826351165771,
      "epoch": 3.5984163531195597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003755096686319886,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2172482803463936,
      "num_tokens": 71551739.0,
      "step": 32040
    },
    {
      "entropy": 5.448180198669434,
      "epoch": 3.5989779300275173,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22240268737077712,
      "num_tokens": 71562429.0,
      "step": 32045
    },
    {
      "entropy": 5.488305997848511,
      "epoch": 3.599539506935475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037543782180204303,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20811505764722824,
      "num_tokens": 71573179.0,
      "step": 32050
    },
    {
      "entropy": 5.4947052001953125,
      "epoch": 3.600101083843432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037540189358952955,
      "loss": 5.0779,
      "mean_token_accuracy": 0.20779660642147063,
      "num_tokens": 71584223.0,
      "step": 32055
    },
    {
      "entropy": 5.411250972747803,
      "epoch": 3.60066266075139,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003753659621801808,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21744544804096222,
      "num_tokens": 71595166.0,
      "step": 32060
    },
    {
      "entropy": 5.418104219436645,
      "epoch": 3.6012242376593475,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037533002757514095,
      "loss": 5.035,
      "mean_token_accuracy": 0.2169846162199974,
      "num_tokens": 71606290.0,
      "step": 32065
    },
    {
      "entropy": 5.407977962493897,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000375294089775554,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21447902917861938,
      "num_tokens": 71617545.0,
      "step": 32070
    },
    {
      "entropy": 5.438481378555298,
      "epoch": 3.6023473914752624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003752581487825644,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2259896069765091,
      "num_tokens": 71627949.0,
      "step": 32075
    },
    {
      "entropy": 5.483710718154907,
      "epoch": 3.60290896838322,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037522220459731644,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21445238143205642,
      "num_tokens": 71639017.0,
      "step": 32080
    },
    {
      "entropy": 5.3851847648620605,
      "epoch": 3.6034705452911777,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22334487438201905,
      "num_tokens": 71649189.0,
      "step": 32085
    },
    {
      "entropy": 5.3427690982818605,
      "epoch": 3.6040321221991354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.9383,
      "mean_token_accuracy": 0.21934900283813477,
      "num_tokens": 71660083.0,
      "step": 32090
    },
    {
      "entropy": 5.346320247650146,
      "epoch": 3.6045936991070926,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037511435289946775,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21984141319990158,
      "num_tokens": 71670874.0,
      "step": 32095
    },
    {
      "entropy": 5.5018798351287845,
      "epoch": 3.6051552760150503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003750783959566322,
      "loss": 5.0387,
      "mean_token_accuracy": 0.2156884953379631,
      "num_tokens": 71681479.0,
      "step": 32100
    },
    {
      "entropy": 5.459052562713623,
      "epoch": 3.605716852923008,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2209658667445183,
      "num_tokens": 71693865.0,
      "step": 32105
    },
    {
      "entropy": 5.367714738845825,
      "epoch": 3.606278429830965,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22172116488218307,
      "num_tokens": 71704513.0,
      "step": 32110
    },
    {
      "entropy": 5.354639196395874,
      "epoch": 3.606840006738923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037497050601349605,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21468107253313065,
      "num_tokens": 71715444.0,
      "step": 32115
    },
    {
      "entropy": 5.3892717361450195,
      "epoch": 3.6074015836468805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22342862337827682,
      "num_tokens": 71725452.0,
      "step": 32120
    },
    {
      "entropy": 5.448363447189331,
      "epoch": 3.607963160554838,
      "grad_norm": 1.0,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.99,
      "mean_token_accuracy": 0.21355825215578078,
      "num_tokens": 71736536.0,
      "step": 32125
    },
    {
      "entropy": 5.389698791503906,
      "epoch": 3.608524737462796,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003748625874224632,
      "loss": 5.001,
      "mean_token_accuracy": 0.21295369565486907,
      "num_tokens": 71748420.0,
      "step": 32130
    },
    {
      "entropy": 5.371073007583618,
      "epoch": 3.609086314370753,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037482660819793134,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22210908830165862,
      "num_tokens": 71758779.0,
      "step": 32135
    },
    {
      "entropy": 5.505191087722778,
      "epoch": 3.6096478912787107,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003747906257948814,
      "loss": 5.1706,
      "mean_token_accuracy": 0.20814643800258636,
      "num_tokens": 71771042.0,
      "step": 32140
    },
    {
      "entropy": 5.45816388130188,
      "epoch": 3.6102094681866683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003747546402144594,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22301269471645355,
      "num_tokens": 71782234.0,
      "step": 32145
    },
    {
      "entropy": 5.4351239681243895,
      "epoch": 3.6107710450946255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.9396,
      "mean_token_accuracy": 0.222194305062294,
      "num_tokens": 71793115.0,
      "step": 32150
    },
    {
      "entropy": 5.364801359176636,
      "epoch": 3.611332622002583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037468265952608196,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22627293020486833,
      "num_tokens": 71804466.0,
      "step": 32155
    },
    {
      "entropy": 5.401900911331177,
      "epoch": 3.611894198910541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003746466644204183,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20401581823825837,
      "num_tokens": 71815248.0,
      "step": 32160
    },
    {
      "entropy": 5.473121166229248,
      "epoch": 3.612455775818498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003746106661419663,
      "loss": 5.034,
      "mean_token_accuracy": 0.21545085310935974,
      "num_tokens": 71825816.0,
      "step": 32165
    },
    {
      "entropy": 5.474753665924072,
      "epoch": 3.6130173527264557,
      "grad_norm": 1.09375,
      "learning_rate": 0.000374574664691872,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21545326262712478,
      "num_tokens": 71837918.0,
      "step": 32170
    },
    {
      "entropy": 5.495144939422607,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037453866007128176,
      "loss": 5.0892,
      "mean_token_accuracy": 0.21281369477510453,
      "num_tokens": 71848736.0,
      "step": 32175
    },
    {
      "entropy": 5.351296854019165,
      "epoch": 3.614140506542371,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037450265228134203,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22188105881214143,
      "num_tokens": 71859903.0,
      "step": 32180
    },
    {
      "entropy": 5.3966148853302,
      "epoch": 3.6147020834503287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037446664132319925,
      "loss": 5.0091,
      "mean_token_accuracy": 0.21634532362222672,
      "num_tokens": 71871656.0,
      "step": 32185
    },
    {
      "entropy": 5.448056602478028,
      "epoch": 3.615263660358286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000374430627198,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21264823973178865,
      "num_tokens": 71883282.0,
      "step": 32190
    },
    {
      "entropy": 5.455802488327026,
      "epoch": 3.6158252372662436,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21939357221126557,
      "num_tokens": 71894473.0,
      "step": 32195
    },
    {
      "entropy": 5.449574995040893,
      "epoch": 3.6163868141742013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003743585894510192,
      "loss": 5.093,
      "mean_token_accuracy": 0.21304668188095094,
      "num_tokens": 71905555.0,
      "step": 32200
    },
    {
      "entropy": 5.392597293853759,
      "epoch": 3.6169483910821585,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037432256583153127,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2096676856279373,
      "num_tokens": 71916599.0,
      "step": 32205
    },
    {
      "entropy": 5.449842882156372,
      "epoch": 3.617509967990116,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037428653904957433,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21696438193321227,
      "num_tokens": 71926173.0,
      "step": 32210
    },
    {
      "entropy": 5.487157773971558,
      "epoch": 3.618071544898074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003742505091062955,
      "loss": 5.0731,
      "mean_token_accuracy": 0.212975911796093,
      "num_tokens": 71937389.0,
      "step": 32215
    },
    {
      "entropy": 5.419522142410278,
      "epoch": 3.6186331218060315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003742144760028419,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21615281701087952,
      "num_tokens": 71949216.0,
      "step": 32220
    },
    {
      "entropy": 5.484242153167725,
      "epoch": 3.619194698713989,
      "grad_norm": 1.0625,
      "learning_rate": 0.000374178439740361,
      "loss": 5.0619,
      "mean_token_accuracy": 0.21490637809038163,
      "num_tokens": 71959861.0,
      "step": 32225
    },
    {
      "entropy": 5.426976346969605,
      "epoch": 3.6197562756219464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003741424003200001,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2166588544845581,
      "num_tokens": 71970043.0,
      "step": 32230
    },
    {
      "entropy": 5.4499061584472654,
      "epoch": 3.620317852529904,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037410635774290667,
      "loss": 4.989,
      "mean_token_accuracy": 0.22220115065574647,
      "num_tokens": 71981096.0,
      "step": 32235
    },
    {
      "entropy": 5.420993375778198,
      "epoch": 3.6208794294378617,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22046905159950256,
      "num_tokens": 71992759.0,
      "step": 32240
    },
    {
      "entropy": 5.42559700012207,
      "epoch": 3.621441006345819,
      "grad_norm": 1.0625,
      "learning_rate": 0.000374034263123113,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21734387725591658,
      "num_tokens": 72003187.0,
      "step": 32245
    },
    {
      "entropy": 5.568624067306518,
      "epoch": 3.6220025832537766,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037399821108270823,
      "loss": 5.1518,
      "mean_token_accuracy": 0.209592866897583,
      "num_tokens": 72013712.0,
      "step": 32250
    },
    {
      "entropy": 5.554506778717041,
      "epoch": 3.622564160161734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000373962155890162,
      "loss": 5.1135,
      "mean_token_accuracy": 0.2081436887383461,
      "num_tokens": 72024693.0,
      "step": 32255
    },
    {
      "entropy": 5.409587144851685,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003739260975466223,
      "loss": 4.956,
      "mean_token_accuracy": 0.22502988427877427,
      "num_tokens": 72036578.0,
      "step": 32260
    },
    {
      "entropy": 5.468867540359497,
      "epoch": 3.623687313977649,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037389003605323737,
      "loss": 5.0758,
      "mean_token_accuracy": 0.2153774842619896,
      "num_tokens": 72047574.0,
      "step": 32265
    },
    {
      "entropy": 5.435374164581299,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21990129649639129,
      "num_tokens": 72059998.0,
      "step": 32270
    },
    {
      "entropy": 5.4741088390350345,
      "epoch": 3.6248104677935644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037381790362152453,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20984285622835158,
      "num_tokens": 72070753.0,
      "step": 32275
    },
    {
      "entropy": 5.3898783206939695,
      "epoch": 3.625372044701522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.926,
      "mean_token_accuracy": 0.21986181735992433,
      "num_tokens": 72081718.0,
      "step": 32280
    },
    {
      "entropy": 5.395722961425781,
      "epoch": 3.6259336216094793,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22227731347084045,
      "num_tokens": 72092403.0,
      "step": 32285
    },
    {
      "entropy": 5.381867122650147,
      "epoch": 3.626495198517437,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003737096813788238,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2212456464767456,
      "num_tokens": 72103711.0,
      "step": 32290
    },
    {
      "entropy": 5.471198415756225,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003736736010104829,
      "loss": 5.0586,
      "mean_token_accuracy": 0.2115407332777977,
      "num_tokens": 72115095.0,
      "step": 32295
    },
    {
      "entropy": 5.462821102142334,
      "epoch": 3.627618352333352,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003736375175003363,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2130939021706581,
      "num_tokens": 72126559.0,
      "step": 32300
    },
    {
      "entropy": 5.443120622634888,
      "epoch": 3.6281799292413095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037360143084953285,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21439512073993683,
      "num_tokens": 72137752.0,
      "step": 32305
    },
    {
      "entropy": 5.3656689643859865,
      "epoch": 3.628741506149267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22616849839687347,
      "num_tokens": 72148888.0,
      "step": 32310
    },
    {
      "entropy": 5.39231424331665,
      "epoch": 3.629303083057225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2218463286757469,
      "num_tokens": 72159607.0,
      "step": 32315
    },
    {
      "entropy": 5.4699952602386475,
      "epoch": 3.6298646599651825,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003734931520646724,
      "loss": 5.045,
      "mean_token_accuracy": 0.21466760337352753,
      "num_tokens": 72171204.0,
      "step": 32320
    },
    {
      "entropy": 5.421824598312378,
      "epoch": 3.6304262368731397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037345705286273285,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21686801165342331,
      "num_tokens": 72182125.0,
      "step": 32325
    },
    {
      "entropy": 5.417802429199218,
      "epoch": 3.6309878137810974,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.994,
      "mean_token_accuracy": 0.21777918934822083,
      "num_tokens": 72193093.0,
      "step": 32330
    },
    {
      "entropy": 5.456452178955078,
      "epoch": 3.631549390689055,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037338484505527103,
      "loss": 5.079,
      "mean_token_accuracy": 0.21039150804281234,
      "num_tokens": 72204235.0,
      "step": 32335
    },
    {
      "entropy": 5.411361169815064,
      "epoch": 3.6321109675970122,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037334873645204785,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21697241067886353,
      "num_tokens": 72215160.0,
      "step": 32340
    },
    {
      "entropy": 5.42682900428772,
      "epoch": 3.63267254450497,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037331262471736285,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2162071332335472,
      "num_tokens": 72227517.0,
      "step": 32345
    },
    {
      "entropy": 5.551150465011597,
      "epoch": 3.6332341214129276,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037327650985236564,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21728498935699464,
      "num_tokens": 72238616.0,
      "step": 32350
    },
    {
      "entropy": 5.517360019683838,
      "epoch": 3.633795698320885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003732403918582063,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21853558272123336,
      "num_tokens": 72250252.0,
      "step": 32355
    },
    {
      "entropy": 5.4928422451019285,
      "epoch": 3.6343572752288424,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037320427073603475,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20591458529233933,
      "num_tokens": 72261197.0,
      "step": 32360
    },
    {
      "entropy": 5.349567317962647,
      "epoch": 3.6349188521368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21939617544412612,
      "num_tokens": 72273838.0,
      "step": 32365
    },
    {
      "entropy": 5.366973829269409,
      "epoch": 3.6354804290447578,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003731320191122557,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22426107972860337,
      "num_tokens": 72285047.0,
      "step": 32370
    },
    {
      "entropy": 5.414035701751709,
      "epoch": 3.6360420059527154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003730958886129486,
      "loss": 5.0505,
      "mean_token_accuracy": 0.2131383016705513,
      "num_tokens": 72295883.0,
      "step": 32375
    },
    {
      "entropy": 5.487714958190918,
      "epoch": 3.6366035828606726,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037305975499023046,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2169301375746727,
      "num_tokens": 72307586.0,
      "step": 32380
    },
    {
      "entropy": 5.440974426269531,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003730236182452517,
      "loss": 4.9537,
      "mean_token_accuracy": 0.21807675212621688,
      "num_tokens": 72318501.0,
      "step": 32385
    },
    {
      "entropy": 5.4146472930908205,
      "epoch": 3.637726736676588,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003729874783791629,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21617843806743622,
      "num_tokens": 72330178.0,
      "step": 32390
    },
    {
      "entropy": 5.4486736297607425,
      "epoch": 3.638288313584545,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003729513353931148,
      "loss": 5.081,
      "mean_token_accuracy": 0.215933895111084,
      "num_tokens": 72340304.0,
      "step": 32395
    },
    {
      "entropy": 5.4787355899810795,
      "epoch": 3.638849890492503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003729151892882581,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21890562325716018,
      "num_tokens": 72350353.0,
      "step": 32400
    },
    {
      "entropy": 5.410263967514038,
      "epoch": 3.6394114674004605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003728790400657439,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2270885154604912,
      "num_tokens": 72361293.0,
      "step": 32405
    },
    {
      "entropy": 5.349084329605103,
      "epoch": 3.639973044308418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.96,
      "mean_token_accuracy": 0.22804917842149736,
      "num_tokens": 72372013.0,
      "step": 32410
    },
    {
      "entropy": 5.3594951152801515,
      "epoch": 3.640534621216376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037280673227234687,
      "loss": 5.0431,
      "mean_token_accuracy": 0.21113687455654145,
      "num_tokens": 72382611.0,
      "step": 32415
    },
    {
      "entropy": 5.4106542587280275,
      "epoch": 3.641096198124333,
      "grad_norm": 1.125,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.9775,
      "mean_token_accuracy": 0.22155033946037292,
      "num_tokens": 72393701.0,
      "step": 32420
    },
    {
      "entropy": 5.60157356262207,
      "epoch": 3.6416577750322907,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003727344120221326,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20395590960979462,
      "num_tokens": 72404168.0,
      "step": 32425
    },
    {
      "entropy": 5.486058044433594,
      "epoch": 3.6422193519402484,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037269824722859743,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21396357864141463,
      "num_tokens": 72415184.0,
      "step": 32430
    },
    {
      "entropy": 5.404782438278199,
      "epoch": 3.6427809288482056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003726620793243121,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21701494455337525,
      "num_tokens": 72426189.0,
      "step": 32435
    },
    {
      "entropy": 5.41213092803955,
      "epoch": 3.6433425057561633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003726259083104284,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21676831245422362,
      "num_tokens": 72436741.0,
      "step": 32440
    },
    {
      "entropy": 5.47843747138977,
      "epoch": 3.643904082664121,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037258973418809784,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21857594549655915,
      "num_tokens": 72448610.0,
      "step": 32445
    },
    {
      "entropy": 5.431500625610352,
      "epoch": 3.644465659572078,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003725535569584723,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21903035938739776,
      "num_tokens": 72460789.0,
      "step": 32450
    },
    {
      "entropy": 5.394928359985352,
      "epoch": 3.645027236480036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21059601306915282,
      "num_tokens": 72471065.0,
      "step": 32455
    },
    {
      "entropy": 5.399752187728882,
      "epoch": 3.6455888133879935,
      "grad_norm": 1.140625,
      "learning_rate": 0.000372481193181944,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22213063836097718,
      "num_tokens": 72481788.0,
      "step": 32460
    },
    {
      "entropy": 5.431295108795166,
      "epoch": 3.646150390295951,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21739132553339005,
      "num_tokens": 72493840.0,
      "step": 32465
    },
    {
      "entropy": 5.37575626373291,
      "epoch": 3.6467119672039088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003724088169900597,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2170462653040886,
      "num_tokens": 72505002.0,
      "step": 32470
    },
    {
      "entropy": 5.325603914260864,
      "epoch": 3.647273544111866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037237262424123975,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2245026409626007,
      "num_tokens": 72515917.0,
      "step": 32475
    },
    {
      "entropy": 5.457018947601318,
      "epoch": 3.6478351210198237,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037233642839203763,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21357165426015853,
      "num_tokens": 72527500.0,
      "step": 32480
    },
    {
      "entropy": 5.351907348632812,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003723002294436059,
      "loss": 4.8986,
      "mean_token_accuracy": 0.2225666418671608,
      "num_tokens": 72538402.0,
      "step": 32485
    },
    {
      "entropy": 5.396405029296875,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.997,
      "mean_token_accuracy": 0.21034591346979142,
      "num_tokens": 72550260.0,
      "step": 32490
    },
    {
      "entropy": 5.489950466156006,
      "epoch": 3.649519851743696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003722278222536639,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21283313482999802,
      "num_tokens": 72562335.0,
      "step": 32495
    },
    {
      "entropy": 5.44412145614624,
      "epoch": 3.650081428651654,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037219161401445914,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2186449721455574,
      "num_tokens": 72572620.0,
      "step": 32500
    },
    {
      "entropy": 5.417562580108642,
      "epoch": 3.6506430055596115,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037215540268063573,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21855491995811463,
      "num_tokens": 72584487.0,
      "step": 32505
    },
    {
      "entropy": 5.401706743240356,
      "epoch": 3.651204582467569,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003721191882533466,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21489957720041275,
      "num_tokens": 72596191.0,
      "step": 32510
    },
    {
      "entropy": 5.4546332359313965,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003720829707337449,
      "loss": 5.0004,
      "mean_token_accuracy": 0.21024405658245088,
      "num_tokens": 72605901.0,
      "step": 32515
    },
    {
      "entropy": 5.444894790649414,
      "epoch": 3.652327736283484,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003720467501229836,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2152911826968193,
      "num_tokens": 72617231.0,
      "step": 32520
    },
    {
      "entropy": 5.494723272323609,
      "epoch": 3.6528893131914417,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003720105264222163,
      "loss": 5.0728,
      "mean_token_accuracy": 0.21193349063396455,
      "num_tokens": 72628777.0,
      "step": 32525
    },
    {
      "entropy": 5.4605554103851315,
      "epoch": 3.653450890099399,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037197429963259615,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21787547469139099,
      "num_tokens": 72639670.0,
      "step": 32530
    },
    {
      "entropy": 5.423086261749267,
      "epoch": 3.6540124670073566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003719380697552767,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22355229109525682,
      "num_tokens": 72651010.0,
      "step": 32535
    },
    {
      "entropy": 5.558578252792358,
      "epoch": 3.6545740439153143,
      "grad_norm": 1.0,
      "learning_rate": 0.0003719018367914115,
      "loss": 5.1078,
      "mean_token_accuracy": 0.21084526628255845,
      "num_tokens": 72662900.0,
      "step": 32540
    },
    {
      "entropy": 5.44339599609375,
      "epoch": 3.6551356208232715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003718656007421542,
      "loss": 5.0879,
      "mean_token_accuracy": 0.21339626759290695,
      "num_tokens": 72674348.0,
      "step": 32545
    },
    {
      "entropy": 5.432695627212524,
      "epoch": 3.655697197731229,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037182936160865866,
      "loss": 4.9129,
      "mean_token_accuracy": 0.21845275908708572,
      "num_tokens": 72685771.0,
      "step": 32550
    },
    {
      "entropy": 5.421091175079345,
      "epoch": 3.656258774639187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22386228144168854,
      "num_tokens": 72697121.0,
      "step": 32555
    },
    {
      "entropy": 5.410254240036011,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22048594802618027,
      "num_tokens": 72707245.0,
      "step": 32560
    },
    {
      "entropy": 5.398268032073974,
      "epoch": 3.657381928455102,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22386341392993928,
      "num_tokens": 72718328.0,
      "step": 32565
    },
    {
      "entropy": 5.476134014129639,
      "epoch": 3.6579435053630593,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003716843742553724,
      "loss": 5.0769,
      "mean_token_accuracy": 0.2130655452609062,
      "num_tokens": 72729192.0,
      "step": 32570
    },
    {
      "entropy": 5.383682203292847,
      "epoch": 3.658505082271017,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003716481197179953,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21030972450971602,
      "num_tokens": 72740213.0,
      "step": 32575
    },
    {
      "entropy": 5.363047409057617,
      "epoch": 3.6590666591789747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037161186210330464,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21937633901834488,
      "num_tokens": 72752477.0,
      "step": 32580
    },
    {
      "entropy": 5.499904155731201,
      "epoch": 3.659628236086932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037157560141245485,
      "loss": 4.9727,
      "mean_token_accuracy": 0.22195008248090745,
      "num_tokens": 72763926.0,
      "step": 32585
    },
    {
      "entropy": 5.44953465461731,
      "epoch": 3.6601898129948895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037153933764660034,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21866535842418672,
      "num_tokens": 72774922.0,
      "step": 32590
    },
    {
      "entropy": 5.386390733718872,
      "epoch": 3.660751389902847,
      "grad_norm": 1.078125,
      "learning_rate": 0.000371503070806896,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21523843556642533,
      "num_tokens": 72786362.0,
      "step": 32595
    },
    {
      "entropy": 5.395166349411011,
      "epoch": 3.661312966810805,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21557731926441193,
      "num_tokens": 72798337.0,
      "step": 32600
    },
    {
      "entropy": 5.457078742980957,
      "epoch": 3.6618745437187625,
      "grad_norm": 1.125,
      "learning_rate": 0.00037143052791055654,
      "loss": 4.9221,
      "mean_token_accuracy": 0.221274234354496,
      "num_tokens": 72808923.0,
      "step": 32605
    },
    {
      "entropy": 5.39433102607727,
      "epoch": 3.6624361206267197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21673766523599625,
      "num_tokens": 72820148.0,
      "step": 32610
    },
    {
      "entropy": 5.4018641948699955,
      "epoch": 3.6629976975346774,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003713579727326754,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21959607154130936,
      "num_tokens": 72830412.0,
      "step": 32615
    },
    {
      "entropy": 5.404740619659424,
      "epoch": 3.663559274442635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2257284551858902,
      "num_tokens": 72840857.0,
      "step": 32620
    },
    {
      "entropy": 5.3995438575744625,
      "epoch": 3.6641208513505923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2158679947257042,
      "num_tokens": 72852269.0,
      "step": 32625
    },
    {
      "entropy": 5.432164049148559,
      "epoch": 3.66468242825855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000371249116958178,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21540742814540864,
      "num_tokens": 72863536.0,
      "step": 32630
    },
    {
      "entropy": 5.342745637893676,
      "epoch": 3.6652440051665076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22599367946386337,
      "num_tokens": 72873835.0,
      "step": 32635
    },
    {
      "entropy": 5.406706619262695,
      "epoch": 3.665805582074465,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003711765311168748,
      "loss": 4.8899,
      "mean_token_accuracy": 0.21812449246644974,
      "num_tokens": 72883916.0,
      "step": 32640
    },
    {
      "entropy": 5.406923818588257,
      "epoch": 3.6663671589824225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037114023360219824,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21148031949996948,
      "num_tokens": 72895928.0,
      "step": 32645
    },
    {
      "entropy": 5.411993312835693,
      "epoch": 3.66692873589038,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037110393302637937,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21257868856191636,
      "num_tokens": 72908138.0,
      "step": 32650
    },
    {
      "entropy": 5.447727108001709,
      "epoch": 3.667490312798338,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037106762939057404,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21587879359722137,
      "num_tokens": 72918326.0,
      "step": 32655
    },
    {
      "entropy": 5.455304193496704,
      "epoch": 3.6680518897062955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003710313226959381,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21518416106700897,
      "num_tokens": 72929140.0,
      "step": 32660
    },
    {
      "entropy": 5.430103445053101,
      "epoch": 3.6686134666142527,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003709950129436276,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22285662293434144,
      "num_tokens": 72941355.0,
      "step": 32665
    },
    {
      "entropy": 5.434720706939697,
      "epoch": 3.6691750435222104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003709587001347987,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2132347121834755,
      "num_tokens": 72951841.0,
      "step": 32670
    },
    {
      "entropy": 5.389640283584595,
      "epoch": 3.669736620430168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037092238427060756,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2191370114684105,
      "num_tokens": 72963843.0,
      "step": 32675
    },
    {
      "entropy": 5.4612068176269535,
      "epoch": 3.6702981973381252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003708860653522104,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22049607187509537,
      "num_tokens": 72973709.0,
      "step": 32680
    },
    {
      "entropy": 5.338049745559692,
      "epoch": 3.670859774246083,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23847329765558242,
      "num_tokens": 72984658.0,
      "step": 32685
    },
    {
      "entropy": 5.327975702285767,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037081341835742433,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21739632040262222,
      "num_tokens": 72995745.0,
      "step": 32690
    },
    {
      "entropy": 5.3787493228912355,
      "epoch": 3.671982928061998,
      "grad_norm": 1.125,
      "learning_rate": 0.0003707770902833483,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21631538420915603,
      "num_tokens": 73007619.0,
      "step": 32695
    },
    {
      "entropy": 5.427275133132935,
      "epoch": 3.672544504969956,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22356079667806625,
      "num_tokens": 73017779.0,
      "step": 32700
    },
    {
      "entropy": 5.458332777023315,
      "epoch": 3.673106081877913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000370704424987614,
      "loss": 4.985,
      "mean_token_accuracy": 0.21958215087652205,
      "num_tokens": 73028246.0,
      "step": 32705
    },
    {
      "entropy": 5.434694766998291,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003706680877682694,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21114393025636674,
      "num_tokens": 73038896.0,
      "step": 32710
    },
    {
      "entropy": 5.414194869995117,
      "epoch": 3.6742292356938284,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037063174750281573,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2101804420351982,
      "num_tokens": 73050960.0,
      "step": 32715
    },
    {
      "entropy": 5.479878044128418,
      "epoch": 3.6747908126017856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22824011892080306,
      "num_tokens": 73061537.0,
      "step": 32720
    },
    {
      "entropy": 5.5057305812835695,
      "epoch": 3.6753523895097433,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037055905783821,
      "loss": 5.0965,
      "mean_token_accuracy": 0.2131681337952614,
      "num_tokens": 73072357.0,
      "step": 32725
    },
    {
      "entropy": 5.398712110519409,
      "epoch": 3.675913966417701,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037052270844137216,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2258327454328537,
      "num_tokens": 73083527.0,
      "step": 32730
    },
    {
      "entropy": 5.376541185379028,
      "epoch": 3.676475543325658,
      "grad_norm": 1.125,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.9084,
      "mean_token_accuracy": 0.21817271858453752,
      "num_tokens": 73095068.0,
      "step": 32735
    },
    {
      "entropy": 5.288842678070068,
      "epoch": 3.677037120233616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003704500005244139,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2258163422346115,
      "num_tokens": 73106318.0,
      "step": 32740
    },
    {
      "entropy": 5.410832023620605,
      "epoch": 3.6775986971415735,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037041364200660835,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2194756180047989,
      "num_tokens": 73118433.0,
      "step": 32745
    },
    {
      "entropy": 5.498656272888184,
      "epoch": 3.678160274049531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037037728045079553,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21309761852025985,
      "num_tokens": 73129887.0,
      "step": 32750
    },
    {
      "entropy": 5.39342041015625,
      "epoch": 3.678721850957489,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000370340915858133,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21882594525814056,
      "num_tokens": 73141529.0,
      "step": 32755
    },
    {
      "entropy": 5.3590904712677006,
      "epoch": 3.679283427865446,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003703045482297788,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22433513253927231,
      "num_tokens": 73151317.0,
      "step": 32760
    },
    {
      "entropy": 5.258422183990478,
      "epoch": 3.6798450047734037,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.856,
      "mean_token_accuracy": 0.22227284759283067,
      "num_tokens": 73162382.0,
      "step": 32765
    },
    {
      "entropy": 5.493445253372192,
      "epoch": 3.6804065816813614,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003702318038706272,
      "loss": 5.0698,
      "mean_token_accuracy": 0.216872675716877,
      "num_tokens": 73173142.0,
      "step": 32770
    },
    {
      "entropy": 5.433839464187622,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21531745046377182,
      "num_tokens": 73185074.0,
      "step": 32775
    },
    {
      "entropy": 5.505591869354248,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037015904738260533,
      "loss": 5.0134,
      "mean_token_accuracy": 0.22001761049032212,
      "num_tokens": 73195806.0,
      "step": 32780
    },
    {
      "entropy": 5.5621216773986815,
      "epoch": 3.682091312405234,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003701226645931637,
      "loss": 5.1361,
      "mean_token_accuracy": 0.2045516401529312,
      "num_tokens": 73207213.0,
      "step": 32785
    },
    {
      "entropy": 5.418919277191162,
      "epoch": 3.6826528893131916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003700862787749796,
      "loss": 5.001,
      "mean_token_accuracy": 0.21523898839950562,
      "num_tokens": 73219715.0,
      "step": 32790
    },
    {
      "entropy": 5.4324378490448,
      "epoch": 3.6832144662211492,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003700498899292115,
      "loss": 4.9411,
      "mean_token_accuracy": 0.21989361047744752,
      "num_tokens": 73229868.0,
      "step": 32795
    },
    {
      "entropy": 5.4289703369140625,
      "epoch": 3.6837760431291064,
      "grad_norm": 1.0,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21428473740816117,
      "num_tokens": 73240347.0,
      "step": 32800
    },
    {
      "entropy": 5.436951160430908,
      "epoch": 3.684337620037064,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003699771031595579,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22343511879444122,
      "num_tokens": 73250165.0,
      "step": 32805
    },
    {
      "entropy": 5.447079610824585,
      "epoch": 3.6848991969450218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036994070523799005,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2147328734397888,
      "num_tokens": 73261320.0,
      "step": 32810
    },
    {
      "entropy": 5.432478713989258,
      "epoch": 3.685460773852979,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036990430429347334,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21538678109645842,
      "num_tokens": 73272342.0,
      "step": 32815
    },
    {
      "entropy": 5.484269905090332,
      "epoch": 3.6860223507609367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036986790032716673,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22273694574832917,
      "num_tokens": 73283790.0,
      "step": 32820
    },
    {
      "entropy": 5.404711008071899,
      "epoch": 3.6865839276688943,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.9847,
      "mean_token_accuracy": 0.2203834131360054,
      "num_tokens": 73295869.0,
      "step": 32825
    },
    {
      "entropy": 5.358232307434082,
      "epoch": 3.6871455045768515,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003697950833338204,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21115920096635818,
      "num_tokens": 73307518.0,
      "step": 32830
    },
    {
      "entropy": 5.432851076126099,
      "epoch": 3.687707081484809,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036975867030909933,
      "loss": 5.0071,
      "mean_token_accuracy": 0.2146642401814461,
      "num_tokens": 73318419.0,
      "step": 32835
    },
    {
      "entropy": 5.426184368133545,
      "epoch": 3.688268658392767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21836827993392943,
      "num_tokens": 73328094.0,
      "step": 32840
    },
    {
      "entropy": 5.402911806106568,
      "epoch": 3.6888302353007245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2245297059416771,
      "num_tokens": 73339156.0,
      "step": 32845
    },
    {
      "entropy": 5.421029329299927,
      "epoch": 3.689391812208682,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036964941313665715,
      "loss": 4.899,
      "mean_token_accuracy": 0.2251039609313011,
      "num_tokens": 73350740.0,
      "step": 32850
    },
    {
      "entropy": 5.457096195220947,
      "epoch": 3.6899533891166394,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003696129880502823,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21794115453958512,
      "num_tokens": 73360745.0,
      "step": 32855
    },
    {
      "entropy": 5.468452882766724,
      "epoch": 3.690514966024597,
      "grad_norm": 1.109375,
      "learning_rate": 0.000369576559951393,
      "loss": 5.0375,
      "mean_token_accuracy": 0.22210952490568162,
      "num_tokens": 73372580.0,
      "step": 32860
    },
    {
      "entropy": 5.435858678817749,
      "epoch": 3.6910765429325547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036954012884114946,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21902744174003602,
      "num_tokens": 73383970.0,
      "step": 32865
    },
    {
      "entropy": 5.405362462997436,
      "epoch": 3.691638119840512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003695036947207115,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22337548285722733,
      "num_tokens": 73394472.0,
      "step": 32870
    },
    {
      "entropy": 5.408407592773438,
      "epoch": 3.6921996967484696,
      "grad_norm": 1.0,
      "learning_rate": 0.00036946725759123936,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2143850713968277,
      "num_tokens": 73405415.0,
      "step": 32875
    },
    {
      "entropy": 5.472079515457153,
      "epoch": 3.6927612736564273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003694308174538929,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21672412306070327,
      "num_tokens": 73416583.0,
      "step": 32880
    },
    {
      "entropy": 5.5194230556488035,
      "epoch": 3.693322850564385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036939437430983274,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21450030207633972,
      "num_tokens": 73427408.0,
      "step": 32885
    },
    {
      "entropy": 5.409603023529053,
      "epoch": 3.6938844274723426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22509371042251586,
      "num_tokens": 73438509.0,
      "step": 32890
    },
    {
      "entropy": 5.38294415473938,
      "epoch": 3.6944460043803,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22144995182752608,
      "num_tokens": 73449197.0,
      "step": 32895
    },
    {
      "entropy": 5.463164806365967,
      "epoch": 3.6950075812882575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036928502684897314,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20889179706573485,
      "num_tokens": 73460777.0,
      "step": 32900
    },
    {
      "entropy": 5.572231245040894,
      "epoch": 3.695569158196215,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036924857168966216,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21190528720617294,
      "num_tokens": 73471747.0,
      "step": 32905
    },
    {
      "entropy": 5.329884672164917,
      "epoch": 3.6961307351041723,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003692121135294401,
      "loss": 4.825,
      "mean_token_accuracy": 0.22388658821582794,
      "num_tokens": 73481258.0,
      "step": 32910
    },
    {
      "entropy": 5.345492649078369,
      "epoch": 3.69669231201213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21604893654584884,
      "num_tokens": 73492073.0,
      "step": 32915
    },
    {
      "entropy": 5.357681560516357,
      "epoch": 3.6972538889200877,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21710047274827957,
      "num_tokens": 73502813.0,
      "step": 32920
    },
    {
      "entropy": 5.509043598175049,
      "epoch": 3.697815465828045,
      "grad_norm": 1.125,
      "learning_rate": 0.00036910272105491625,
      "loss": 5.0406,
      "mean_token_accuracy": 0.22071566730737685,
      "num_tokens": 73514417.0,
      "step": 32925
    },
    {
      "entropy": 5.484415006637573,
      "epoch": 3.6983770427360025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036906625090265925,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22110741585493088,
      "num_tokens": 73525636.0,
      "step": 32930
    },
    {
      "entropy": 5.380090761184692,
      "epoch": 3.69893861964396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21614866107702255,
      "num_tokens": 73536557.0,
      "step": 32935
    },
    {
      "entropy": 5.340582799911499,
      "epoch": 3.699500196551918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003689933016139888,
      "loss": 4.883,
      "mean_token_accuracy": 0.22808073759078978,
      "num_tokens": 73547910.0,
      "step": 32940
    },
    {
      "entropy": 5.331180572509766,
      "epoch": 3.7000617734598755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22415668666362762,
      "num_tokens": 73557583.0,
      "step": 32945
    },
    {
      "entropy": 5.412989807128906,
      "epoch": 3.7006233503678327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003689203403541857,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2158808246254921,
      "num_tokens": 73567300.0,
      "step": 32950
    },
    {
      "entropy": 5.469138717651367,
      "epoch": 3.7011849272757904,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036888385523801325,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21507049202919007,
      "num_tokens": 73577387.0,
      "step": 32955
    },
    {
      "entropy": 5.538281869888306,
      "epoch": 3.701746504183748,
      "grad_norm": 1.125,
      "learning_rate": 0.00036884736713254247,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21075861006975175,
      "num_tokens": 73588340.0,
      "step": 32960
    },
    {
      "entropy": 5.48117504119873,
      "epoch": 3.7023080810917053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036881087603893515,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21877917647361755,
      "num_tokens": 73599650.0,
      "step": 32965
    },
    {
      "entropy": 5.47825345993042,
      "epoch": 3.702869657999663,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036877438195835307,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21953275203704833,
      "num_tokens": 73611279.0,
      "step": 32970
    },
    {
      "entropy": 5.381777048110962,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21917023360729218,
      "num_tokens": 73623279.0,
      "step": 32975
    },
    {
      "entropy": 5.410960102081299,
      "epoch": 3.7039928118155783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036870138484091297,
      "loss": 5.0557,
      "mean_token_accuracy": 0.21168290227651596,
      "num_tokens": 73633756.0,
      "step": 32980
    },
    {
      "entropy": 5.533349609375,
      "epoch": 3.704554388723536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036866488180637926,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20591085851192475,
      "num_tokens": 73646033.0,
      "step": 32985
    },
    {
      "entropy": 5.45375018119812,
      "epoch": 3.705115965631493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003686283757895193,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21237977147102355,
      "num_tokens": 73655833.0,
      "step": 32990
    },
    {
      "entropy": 5.505877447128296,
      "epoch": 3.705677542539451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036859186679149555,
      "loss": 5.0765,
      "mean_token_accuracy": 0.20947602838277818,
      "num_tokens": 73667755.0,
      "step": 32995
    },
    {
      "entropy": 5.504134178161621,
      "epoch": 3.7062391194474085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003685553548134705,
      "loss": 5.0941,
      "mean_token_accuracy": 0.21775619685649872,
      "num_tokens": 73680537.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.220043198853877,
      "eval_loss": 5.150248050689697,
      "eval_mean_token_accuracy": 0.21750620984430777,
      "eval_num_tokens": 73680537.0,
      "eval_runtime": 53.3497,
      "eval_samples_per_second": 581.315,
      "eval_steps_per_second": 72.671,
      "step": 33000
    },
    {
      "entropy": 5.427530431747437,
      "epoch": 3.7068006963553657,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2258942112326622,
      "num_tokens": 73692446.0,
      "step": 33005
    },
    {
      "entropy": 5.417485761642456,
      "epoch": 3.7073622732633233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003684823219220666,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21384463757276534,
      "num_tokens": 73704059.0,
      "step": 33010
    },
    {
      "entropy": 5.389265441894532,
      "epoch": 3.707923850171281,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003684458010110133,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21530304551124574,
      "num_tokens": 73715234.0,
      "step": 33015
    },
    {
      "entropy": 5.461320447921753,
      "epoch": 3.7084854270792382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003684092771246094,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2083536848425865,
      "num_tokens": 73726930.0,
      "step": 33020
    },
    {
      "entropy": 5.459152269363403,
      "epoch": 3.709047003987196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003683727502640178,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21490905284881592,
      "num_tokens": 73738107.0,
      "step": 33025
    },
    {
      "entropy": 5.481708765029907,
      "epoch": 3.7096085808951536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.0992,
      "mean_token_accuracy": 0.21578264683485032,
      "num_tokens": 73749165.0,
      "step": 33030
    },
    {
      "entropy": 5.461365413665772,
      "epoch": 3.710170157803111,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.9371,
      "mean_token_accuracy": 0.2251826912164688,
      "num_tokens": 73760921.0,
      "step": 33035
    },
    {
      "entropy": 5.428691101074219,
      "epoch": 3.710731734711069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003682631518487483,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21630692780017852,
      "num_tokens": 73772045.0,
      "step": 33040
    },
    {
      "entropy": 5.420889616012573,
      "epoch": 3.711293311619026,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003682266131030375,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22212181091308594,
      "num_tokens": 73782698.0,
      "step": 33045
    },
    {
      "entropy": 5.34484281539917,
      "epoch": 3.7118548885269838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036819007138895525,
      "loss": 4.9181,
      "mean_token_accuracy": 0.2193426102399826,
      "num_tokens": 73793758.0,
      "step": 33050
    },
    {
      "entropy": 5.448407602310181,
      "epoch": 3.7124164654349414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000368153526707665,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21002687364816666,
      "num_tokens": 73805479.0,
      "step": 33055
    },
    {
      "entropy": 5.4761302947998045,
      "epoch": 3.7129780423428986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003681169790603303,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21225939691066742,
      "num_tokens": 73817361.0,
      "step": 33060
    },
    {
      "entropy": 5.462630891799927,
      "epoch": 3.7135396192508563,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036808042844811484,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21682698428630828,
      "num_tokens": 73829651.0,
      "step": 33065
    },
    {
      "entropy": 5.367788887023925,
      "epoch": 3.714101196158814,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22377778738737106,
      "num_tokens": 73839608.0,
      "step": 33070
    },
    {
      "entropy": 5.374065923690796,
      "epoch": 3.7146627730667716,
      "grad_norm": 1.015625,
      "learning_rate": 0.000368007318333697,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2199182391166687,
      "num_tokens": 73850663.0,
      "step": 33075
    },
    {
      "entropy": 5.44550199508667,
      "epoch": 3.7152243499747293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003679707588338225,
      "loss": 5.0166,
      "mean_token_accuracy": 0.21642640829086304,
      "num_tokens": 73862441.0,
      "step": 33080
    },
    {
      "entropy": 5.470325326919555,
      "epoch": 3.7157859268826865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000367934196373723,
      "loss": 4.9731,
      "mean_token_accuracy": 0.22307281494140624,
      "num_tokens": 73873549.0,
      "step": 33085
    },
    {
      "entropy": 5.4299825668334964,
      "epoch": 3.716347503790644,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036789763095456264,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22151493430137634,
      "num_tokens": 73884872.0,
      "step": 33090
    },
    {
      "entropy": 5.525178098678589,
      "epoch": 3.716909080698602,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003678610625775056,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20535001903772354,
      "num_tokens": 73896137.0,
      "step": 33095
    },
    {
      "entropy": 5.38016037940979,
      "epoch": 3.717470657606559,
      "grad_norm": 1.0,
      "learning_rate": 0.00036782449124371643,
      "loss": 4.8828,
      "mean_token_accuracy": 0.21694179028272628,
      "num_tokens": 73906135.0,
      "step": 33100
    },
    {
      "entropy": 5.4775200366973875,
      "epoch": 3.7180322345145167,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2196973443031311,
      "num_tokens": 73916856.0,
      "step": 33105
    },
    {
      "entropy": 5.299609231948852,
      "epoch": 3.7185938114224744,
      "grad_norm": 1.0,
      "learning_rate": 0.00036775133971059904,
      "loss": 4.9186,
      "mean_token_accuracy": 0.22716130167245865,
      "num_tokens": 73928268.0,
      "step": 33110
    },
    {
      "entropy": 5.4426452159881595,
      "epoch": 3.7191553883304316,
      "grad_norm": 1.0,
      "learning_rate": 0.0003677147595136001,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21659004986286162,
      "num_tokens": 73940327.0,
      "step": 33115
    },
    {
      "entropy": 5.527439880371094,
      "epoch": 3.7197169652383892,
      "grad_norm": 1.0,
      "learning_rate": 0.00036767817636452723,
      "loss": 5.016,
      "mean_token_accuracy": 0.2245740056037903,
      "num_tokens": 73951488.0,
      "step": 33120
    },
    {
      "entropy": 5.4000715732574465,
      "epoch": 3.720278542146347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036764159026454534,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2219654828310013,
      "num_tokens": 73963479.0,
      "step": 33125
    },
    {
      "entropy": 5.292480325698852,
      "epoch": 3.7208401190543046,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003676050012148193,
      "loss": 4.8653,
      "mean_token_accuracy": 0.21948444098234177,
      "num_tokens": 73974878.0,
      "step": 33130
    },
    {
      "entropy": 5.434208774566651,
      "epoch": 3.7214016959622622,
      "grad_norm": 1.078125,
      "learning_rate": 0.000367568409216514,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21678983718156813,
      "num_tokens": 73985983.0,
      "step": 33135
    },
    {
      "entropy": 5.434379053115845,
      "epoch": 3.7219632728702194,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003675318142707947,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21584286391735077,
      "num_tokens": 73997141.0,
      "step": 33140
    },
    {
      "entropy": 5.432626581192016,
      "epoch": 3.722524849778177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22061228305101394,
      "num_tokens": 74008262.0,
      "step": 33145
    },
    {
      "entropy": 5.431460237503051,
      "epoch": 3.7230864266861348,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036745861554177486,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2165122613310814,
      "num_tokens": 74020465.0,
      "step": 33150
    },
    {
      "entropy": 5.442072534561158,
      "epoch": 3.723648003594092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036742201176080496,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21993062049150466,
      "num_tokens": 74031100.0,
      "step": 33155
    },
    {
      "entropy": 5.47317943572998,
      "epoch": 3.7242095805020496,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003673854050370824,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2098364055156708,
      "num_tokens": 74043313.0,
      "step": 33160
    },
    {
      "entropy": 5.402803802490235,
      "epoch": 3.7247711574100073,
      "grad_norm": 1.0,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2180256024003029,
      "num_tokens": 74054540.0,
      "step": 33165
    },
    {
      "entropy": 5.5105753421783445,
      "epoch": 3.725332734317965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003673121827660415,
      "loss": 5.1436,
      "mean_token_accuracy": 0.21292126774787903,
      "num_tokens": 74065636.0,
      "step": 33170
    },
    {
      "entropy": 5.3694220066070555,
      "epoch": 3.7258943112259226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21962720453739165,
      "num_tokens": 74076516.0,
      "step": 33175
    },
    {
      "entropy": 5.43087363243103,
      "epoch": 3.72645588813388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036723894873797786,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21742255240678787,
      "num_tokens": 74087481.0,
      "step": 33180
    },
    {
      "entropy": 5.504390096664428,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036720232731797726,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2189963147044182,
      "num_tokens": 74098577.0,
      "step": 33185
    },
    {
      "entropy": 5.495723819732666,
      "epoch": 3.727579041949795,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036716570296221873,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20632191002368927,
      "num_tokens": 74108813.0,
      "step": 33190
    },
    {
      "entropy": 5.428347587585449,
      "epoch": 3.7281406188577524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2139742538332939,
      "num_tokens": 74118937.0,
      "step": 33195
    },
    {
      "entropy": 5.44991717338562,
      "epoch": 3.72870219576571,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003670924454480928,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21861491203308106,
      "num_tokens": 74129879.0,
      "step": 33200
    },
    {
      "entropy": 5.399912738800049,
      "epoch": 3.7292637726736677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22036532908678055,
      "num_tokens": 74140455.0,
      "step": 33205
    },
    {
      "entropy": 5.451206684112549,
      "epoch": 3.729825349581625,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036701917620493026,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21249378025531768,
      "num_tokens": 74152508.0,
      "step": 33210
    },
    {
      "entropy": 5.403292942047119,
      "epoch": 3.7303869264895826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003669825371878764,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2236400365829468,
      "num_tokens": 74163700.0,
      "step": 33215
    },
    {
      "entropy": 5.379795837402344,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036694589524206275,
      "loss": 4.9908,
      "mean_token_accuracy": 0.22369305193424224,
      "num_tokens": 74174715.0,
      "step": 33220
    },
    {
      "entropy": 5.475397300720215,
      "epoch": 3.731510080305498,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003669092503686562,
      "loss": 5.0571,
      "mean_token_accuracy": 0.214888134598732,
      "num_tokens": 74186679.0,
      "step": 33225
    },
    {
      "entropy": 5.471183156967163,
      "epoch": 3.7320716572134556,
      "grad_norm": 1.0,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21720597445964812,
      "num_tokens": 74196382.0,
      "step": 33230
    },
    {
      "entropy": 5.487013101577759,
      "epoch": 3.732633234121413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036683595184373147,
      "loss": 5.036,
      "mean_token_accuracy": 0.21445513963699342,
      "num_tokens": 74207587.0,
      "step": 33235
    },
    {
      "entropy": 5.384006357192993,
      "epoch": 3.7331948110293705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21873777508735656,
      "num_tokens": 74218826.0,
      "step": 33240
    },
    {
      "entropy": 5.506264305114746,
      "epoch": 3.733756387937328,
      "grad_norm": 1.125,
      "learning_rate": 0.00036676264162243744,
      "loss": 5.0903,
      "mean_token_accuracy": 0.216038079559803,
      "num_tokens": 74230450.0,
      "step": 33245
    },
    {
      "entropy": 5.4872089385986325,
      "epoch": 3.7343179648452853,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003667259821285698,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2097722813487053,
      "num_tokens": 74241438.0,
      "step": 33250
    },
    {
      "entropy": 5.540822839736938,
      "epoch": 3.734879541753243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003666893197141112,
      "loss": 5.0746,
      "mean_token_accuracy": 0.21250900626182556,
      "num_tokens": 74252460.0,
      "step": 33255
    },
    {
      "entropy": 5.465585374832154,
      "epoch": 3.7354411186612007,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003666526543802291,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22254605442285538,
      "num_tokens": 74262984.0,
      "step": 33260
    },
    {
      "entropy": 5.488776636123657,
      "epoch": 3.7360026955691583,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000366615986128091,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21427225619554519,
      "num_tokens": 74274368.0,
      "step": 33265
    },
    {
      "entropy": 5.367695474624634,
      "epoch": 3.736564272477116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003665793149588644,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22583746165037155,
      "num_tokens": 74286394.0,
      "step": 33270
    },
    {
      "entropy": 5.3609694480896,
      "epoch": 3.737125849385073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.913,
      "mean_token_accuracy": 0.21892166137695312,
      "num_tokens": 74296803.0,
      "step": 33275
    },
    {
      "entropy": 5.423884439468384,
      "epoch": 3.737687426293031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036650596387381615,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2267253130674362,
      "num_tokens": 74306471.0,
      "step": 33280
    },
    {
      "entropy": 5.46139612197876,
      "epoch": 3.7382490032009885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036646928396033,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22172649055719376,
      "num_tokens": 74318875.0,
      "step": 33285
    },
    {
      "entropy": 5.431304979324341,
      "epoch": 3.7388105801089457,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036643260113442634,
      "loss": 4.9963,
      "mean_token_accuracy": 0.21788119226694108,
      "num_tokens": 74328464.0,
      "step": 33290
    },
    {
      "entropy": 5.437408351898194,
      "epoch": 3.7393721570169034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003663959153972732,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22142713218927385,
      "num_tokens": 74339462.0,
      "step": 33295
    },
    {
      "entropy": 5.476829624176025,
      "epoch": 3.739933733924861,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003663592267500388,
      "loss": 5.0437,
      "mean_token_accuracy": 0.2151297941803932,
      "num_tokens": 74350667.0,
      "step": 33300
    },
    {
      "entropy": 5.424339437484742,
      "epoch": 3.7404953108328187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003663225351938911,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22039258778095244,
      "num_tokens": 74361734.0,
      "step": 33305
    },
    {
      "entropy": 5.435579776763916,
      "epoch": 3.741056887740776,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003662858407299984,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20896041244268418,
      "num_tokens": 74373465.0,
      "step": 33310
    },
    {
      "entropy": 5.43818769454956,
      "epoch": 3.7416184646487336,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036624914335952924,
      "loss": 4.9617,
      "mean_token_accuracy": 0.217418272793293,
      "num_tokens": 74384853.0,
      "step": 33315
    },
    {
      "entropy": 5.360261392593384,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036621244308365195,
      "loss": 4.8616,
      "mean_token_accuracy": 0.21941311210393905,
      "num_tokens": 74394356.0,
      "step": 33320
    },
    {
      "entropy": 5.450801038742066,
      "epoch": 3.742741618464649,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036617573990353503,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21255522966384888,
      "num_tokens": 74406721.0,
      "step": 33325
    },
    {
      "entropy": 5.453897285461426,
      "epoch": 3.743303195372606,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003661390338203472,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2195073574781418,
      "num_tokens": 74418217.0,
      "step": 33330
    },
    {
      "entropy": 5.429179286956787,
      "epoch": 3.743864772280564,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003661023248352572,
      "loss": 4.961,
      "mean_token_accuracy": 0.2236451178789139,
      "num_tokens": 74428644.0,
      "step": 33335
    },
    {
      "entropy": 5.392580461502075,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22118284404277802,
      "num_tokens": 74439575.0,
      "step": 33340
    },
    {
      "entropy": 5.416156673431397,
      "epoch": 3.7449879260964787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036602889816404613,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2179145947098732,
      "num_tokens": 74450191.0,
      "step": 33345
    },
    {
      "entropy": 5.435729837417602,
      "epoch": 3.7455495030044363,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036599218048026294,
      "loss": 5.0139,
      "mean_token_accuracy": 0.21928644329309463,
      "num_tokens": 74460816.0,
      "step": 33350
    },
    {
      "entropy": 5.426756525039673,
      "epoch": 3.746111079912394,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036595545989925345,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22695696502923965,
      "num_tokens": 74470999.0,
      "step": 33355
    },
    {
      "entropy": 5.416748523712158,
      "epoch": 3.7466726568203517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.8404,
      "mean_token_accuracy": 0.2325312539935112,
      "num_tokens": 74480884.0,
      "step": 33360
    },
    {
      "entropy": 5.356525516510009,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036588201005023223,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22098326534032822,
      "num_tokens": 74492500.0,
      "step": 33365
    },
    {
      "entropy": 5.3971555709838865,
      "epoch": 3.7477958106362665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003658452807845593,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21919877231121063,
      "num_tokens": 74504635.0,
      "step": 33370
    },
    {
      "entropy": 5.4874554634094235,
      "epoch": 3.748357387544224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036580854862633747,
      "loss": 5.0468,
      "mean_token_accuracy": 0.210519041121006,
      "num_tokens": 74515743.0,
      "step": 33375
    },
    {
      "entropy": 5.40060396194458,
      "epoch": 3.748918964452182,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003657718135767362,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21315452158451081,
      "num_tokens": 74528089.0,
      "step": 33380
    },
    {
      "entropy": 5.467764949798584,
      "epoch": 3.749480541360139,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003657350756369251,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21581330597400666,
      "num_tokens": 74538543.0,
      "step": 33385
    },
    {
      "entropy": 5.4802898406982425,
      "epoch": 3.7500421182680967,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036569833480807403,
      "loss": 5.0698,
      "mean_token_accuracy": 0.21449417471885682,
      "num_tokens": 74549825.0,
      "step": 33390
    },
    {
      "entropy": 5.398254919052124,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003656615910913528,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22279461175203324,
      "num_tokens": 74560304.0,
      "step": 33395
    },
    {
      "entropy": 5.375800704956054,
      "epoch": 3.751165272084012,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036562484448793135,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22508104890584946,
      "num_tokens": 74571549.0,
      "step": 33400
    },
    {
      "entropy": 5.496183109283447,
      "epoch": 3.7517268489919693,
      "grad_norm": 1.125,
      "learning_rate": 0.0003655880949989798,
      "loss": 5.0551,
      "mean_token_accuracy": 0.2147555246949196,
      "num_tokens": 74581865.0,
      "step": 33405
    },
    {
      "entropy": 5.352853012084961,
      "epoch": 3.752288425899927,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036555134262566806,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21847886443138123,
      "num_tokens": 74592079.0,
      "step": 33410
    },
    {
      "entropy": 5.38393874168396,
      "epoch": 3.7528500028078846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003655145873691665,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2174655720591545,
      "num_tokens": 74603058.0,
      "step": 33415
    },
    {
      "entropy": 5.441703128814697,
      "epoch": 3.7534115797158423,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036547782923064537,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21297045946121215,
      "num_tokens": 74614008.0,
      "step": 33420
    },
    {
      "entropy": 5.546002721786499,
      "epoch": 3.7539731566237995,
      "grad_norm": 1.125,
      "learning_rate": 0.00036544106821127505,
      "loss": 5.1017,
      "mean_token_accuracy": 0.21008777171373366,
      "num_tokens": 74625146.0,
      "step": 33425
    },
    {
      "entropy": 5.567825794219971,
      "epoch": 3.754534733531757,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000365404304312226,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2160453751683235,
      "num_tokens": 74636578.0,
      "step": 33430
    },
    {
      "entropy": 5.459314489364624,
      "epoch": 3.755096310439715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036536753753466897,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21661980003118514,
      "num_tokens": 74648551.0,
      "step": 33435
    },
    {
      "entropy": 5.409550142288208,
      "epoch": 3.755657887347672,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003653307678797744,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22512203752994536,
      "num_tokens": 74658712.0,
      "step": 33440
    },
    {
      "entropy": 5.360746288299561,
      "epoch": 3.7562194642556297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003652939953487131,
      "loss": 4.924,
      "mean_token_accuracy": 0.21781559884548188,
      "num_tokens": 74670196.0,
      "step": 33445
    },
    {
      "entropy": 5.444022560119629,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036525721994265606,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2164058580994606,
      "num_tokens": 74680577.0,
      "step": 33450
    },
    {
      "entropy": 5.487272453308106,
      "epoch": 3.757342618071545,
      "grad_norm": 1.0,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22049394100904465,
      "num_tokens": 74692054.0,
      "step": 33455
    },
    {
      "entropy": 5.464502334594727,
      "epoch": 3.7579041949795027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21756777614355088,
      "num_tokens": 74703324.0,
      "step": 33460
    },
    {
      "entropy": 5.389591693878174,
      "epoch": 3.75846577188746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21983040273189544,
      "num_tokens": 74713028.0,
      "step": 33465
    },
    {
      "entropy": 5.376696681976318,
      "epoch": 3.7590273487954176,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036511008959188966,
      "loss": 4.9442,
      "mean_token_accuracy": 0.21717313677072525,
      "num_tokens": 74723964.0,
      "step": 33470
    },
    {
      "entropy": 5.38577127456665,
      "epoch": 3.759588925703375,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2178808942437172,
      "num_tokens": 74734928.0,
      "step": 33475
    },
    {
      "entropy": 5.548198318481445,
      "epoch": 3.7601505026113324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003650365071969803,
      "loss": 5.115,
      "mean_token_accuracy": 0.20926181077957154,
      "num_tokens": 74747198.0,
      "step": 33480
    },
    {
      "entropy": 5.470646858215332,
      "epoch": 3.76071207951929,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003649997116987439,
      "loss": 5.0065,
      "mean_token_accuracy": 0.22390239536762238,
      "num_tokens": 74758180.0,
      "step": 33485
    },
    {
      "entropy": 5.471925163269043,
      "epoch": 3.7612736564272478,
      "grad_norm": 1.0,
      "learning_rate": 0.0003649629133348818,
      "loss": 5.0128,
      "mean_token_accuracy": 0.22022641599178314,
      "num_tokens": 74769227.0,
      "step": 33490
    },
    {
      "entropy": 5.409882497787476,
      "epoch": 3.7618352333352054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21795818507671355,
      "num_tokens": 74780347.0,
      "step": 33495
    },
    {
      "entropy": 5.418921947479248,
      "epoch": 3.7623968102431626,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.9913,
      "mean_token_accuracy": 0.2202054113149643,
      "num_tokens": 74791902.0,
      "step": 33500
    },
    {
      "entropy": 5.418864393234253,
      "epoch": 3.7629583871511203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036485250106125844,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21684525310993194,
      "num_tokens": 74803530.0,
      "step": 33505
    },
    {
      "entropy": 5.4477780818939205,
      "epoch": 3.763519964059078,
      "grad_norm": 1.109375,
      "learning_rate": 0.000364815691246611,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22777428328990937,
      "num_tokens": 74815128.0,
      "step": 33510
    },
    {
      "entropy": 5.5551189422607425,
      "epoch": 3.7640815409670356,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036477887857219726,
      "loss": 5.1561,
      "mean_token_accuracy": 0.21170889884233474,
      "num_tokens": 74827729.0,
      "step": 33515
    },
    {
      "entropy": 5.3820844173431395,
      "epoch": 3.764643117874993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036474206303918914,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21779515594244003,
      "num_tokens": 74837922.0,
      "step": 33520
    },
    {
      "entropy": 5.430291414260864,
      "epoch": 3.7652046947829505,
      "grad_norm": 1.03125,
      "learning_rate": 0.000364705244648759,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2214677482843399,
      "num_tokens": 74848239.0,
      "step": 33525
    },
    {
      "entropy": 5.481971883773804,
      "epoch": 3.765766271690908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003646684234020789,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21280981600284576,
      "num_tokens": 74860454.0,
      "step": 33530
    },
    {
      "entropy": 5.412882280349732,
      "epoch": 3.7663278485988654,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.9371,
      "mean_token_accuracy": 0.2169406771659851,
      "num_tokens": 74871849.0,
      "step": 33535
    },
    {
      "entropy": 5.466569900512695,
      "epoch": 3.766889425506823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036459477234465935,
      "loss": 5.0562,
      "mean_token_accuracy": 0.2105882704257965,
      "num_tokens": 74883877.0,
      "step": 33540
    },
    {
      "entropy": 5.663352727890015,
      "epoch": 3.7674510024147807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036455794253626486,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20494538992643357,
      "num_tokens": 74897218.0,
      "step": 33545
    },
    {
      "entropy": 5.521263885498047,
      "epoch": 3.7680125793227384,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036452110987631066,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2173085480928421,
      "num_tokens": 74908708.0,
      "step": 33550
    },
    {
      "entropy": 5.457334804534912,
      "epoch": 3.768574156230696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22216227054595947,
      "num_tokens": 74919664.0,
      "step": 33555
    },
    {
      "entropy": 5.380177068710327,
      "epoch": 3.7691357331386532,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.982,
      "mean_token_accuracy": 0.22437432557344436,
      "num_tokens": 74930386.0,
      "step": 33560
    },
    {
      "entropy": 5.382160568237305,
      "epoch": 3.769697310046611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.988,
      "mean_token_accuracy": 0.21479461938142777,
      "num_tokens": 74942835.0,
      "step": 33565
    },
    {
      "entropy": 5.4694469451904295,
      "epoch": 3.7702588869545686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036437375074435446,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2134062945842743,
      "num_tokens": 74954332.0,
      "step": 33570
    },
    {
      "entropy": 5.52327823638916,
      "epoch": 3.770820463862526,
      "grad_norm": 1.125,
      "learning_rate": 0.0003643369038441955,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21485630720853804,
      "num_tokens": 74965530.0,
      "step": 33575
    },
    {
      "entropy": 5.435948371887207,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036430005409951476,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22017038017511367,
      "num_tokens": 74976319.0,
      "step": 33580
    },
    {
      "entropy": 5.409151792526245,
      "epoch": 3.771943617678441,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036426320151148575,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21221695989370346,
      "num_tokens": 74987948.0,
      "step": 33585
    },
    {
      "entropy": 5.430173110961914,
      "epoch": 3.7725051945863988,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036422634608128184,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21993712186813355,
      "num_tokens": 74998989.0,
      "step": 33590
    },
    {
      "entropy": 5.4295618534088135,
      "epoch": 3.773066771494356,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036418948781007636,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22002994567155837,
      "num_tokens": 75010510.0,
      "step": 33595
    },
    {
      "entropy": 5.461617803573608,
      "epoch": 3.7736283484023136,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036415262669904305,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2100297912955284,
      "num_tokens": 75020708.0,
      "step": 33600
    },
    {
      "entropy": 5.456891441345215,
      "epoch": 3.7741899253102713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036411576274935546,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22422772347927095,
      "num_tokens": 75031468.0,
      "step": 33605
    },
    {
      "entropy": 5.383851623535156,
      "epoch": 3.774751502218229,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22778162956237794,
      "num_tokens": 75042170.0,
      "step": 33610
    },
    {
      "entropy": 5.377733993530273,
      "epoch": 3.775313079126186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000364042026338713,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22163959294557573,
      "num_tokens": 75053916.0,
      "step": 33615
    },
    {
      "entropy": 5.430508756637574,
      "epoch": 3.775874656034144,
      "grad_norm": 1.0,
      "learning_rate": 0.00036400515388010574,
      "loss": 4.9555,
      "mean_token_accuracy": 0.22411427646875381,
      "num_tokens": 75065026.0,
      "step": 33620
    },
    {
      "entropy": 5.403352212905884,
      "epoch": 3.7764362329421015,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003639682785875399,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2117028594017029,
      "num_tokens": 75075617.0,
      "step": 33625
    },
    {
      "entropy": 5.433413553237915,
      "epoch": 3.7769978098500587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036393140046218957,
      "loss": 5.0,
      "mean_token_accuracy": 0.2102942645549774,
      "num_tokens": 75085931.0,
      "step": 33630
    },
    {
      "entropy": 5.503500652313233,
      "epoch": 3.7775593867580164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036389451950522895,
      "loss": 5.0108,
      "mean_token_accuracy": 0.22571051269769668,
      "num_tokens": 75096404.0,
      "step": 33635
    },
    {
      "entropy": 5.542718744277954,
      "epoch": 3.778120963665974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036385763571783236,
      "loss": 5.0949,
      "mean_token_accuracy": 0.21403420716524124,
      "num_tokens": 75108300.0,
      "step": 33640
    },
    {
      "entropy": 5.454229927062988,
      "epoch": 3.7786825405739317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003638207491011742,
      "loss": 5.0418,
      "mean_token_accuracy": 0.21776841431856156,
      "num_tokens": 75119799.0,
      "step": 33645
    },
    {
      "entropy": 5.385709619522094,
      "epoch": 3.7792441174818894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22519199699163436,
      "num_tokens": 75130803.0,
      "step": 33650
    },
    {
      "entropy": 5.429051303863526,
      "epoch": 3.7798056943898466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036374696738477105,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22574401944875716,
      "num_tokens": 75140621.0,
      "step": 33655
    },
    {
      "entropy": 5.405671119689941,
      "epoch": 3.7803672712978043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003637100722873754,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21484699845314026,
      "num_tokens": 75152446.0,
      "step": 33660
    },
    {
      "entropy": 5.4514782428741455,
      "epoch": 3.780928848205762,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2183723896741867,
      "num_tokens": 75162936.0,
      "step": 33665
    },
    {
      "entropy": 5.3589324951171875,
      "epoch": 3.781490425113719,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22286706268787385,
      "num_tokens": 75174184.0,
      "step": 33670
    },
    {
      "entropy": 5.273860454559326,
      "epoch": 3.782052002021677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.897,
      "mean_token_accuracy": 0.22305538803339003,
      "num_tokens": 75185609.0,
      "step": 33675
    },
    {
      "entropy": 5.503331661224365,
      "epoch": 3.7826135789296345,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036356246366391054,
      "loss": 5.0798,
      "mean_token_accuracy": 0.2154954805970192,
      "num_tokens": 75196098.0,
      "step": 33680
    },
    {
      "entropy": 5.567646598815918,
      "epoch": 3.783175155837592,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036352555445544864,
      "loss": 5.0573,
      "mean_token_accuracy": 0.2149672269821167,
      "num_tokens": 75206411.0,
      "step": 33685
    },
    {
      "entropy": 5.397810792922973,
      "epoch": 3.78373673274555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22084194272756577,
      "num_tokens": 75217029.0,
      "step": 33690
    },
    {
      "entropy": 5.3207813739776615,
      "epoch": 3.784298309653507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2185880273580551,
      "num_tokens": 75228124.0,
      "step": 33695
    },
    {
      "entropy": 5.3315376281738285,
      "epoch": 3.7848598865614647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003634148099226364,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2126242995262146,
      "num_tokens": 75239776.0,
      "step": 33700
    },
    {
      "entropy": 5.3966141700744625,
      "epoch": 3.7854214634694223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21916851699352263,
      "num_tokens": 75251354.0,
      "step": 33705
    },
    {
      "entropy": 5.485003471374512,
      "epoch": 3.7859830403773795,
      "grad_norm": 1.0,
      "learning_rate": 0.0003633409661563267,
      "loss": 5.022,
      "mean_token_accuracy": 0.21041466444730758,
      "num_tokens": 75263167.0,
      "step": 33710
    },
    {
      "entropy": 5.39678258895874,
      "epoch": 3.786544617285337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22262929677963256,
      "num_tokens": 75274200.0,
      "step": 33715
    },
    {
      "entropy": 5.442705345153809,
      "epoch": 3.787106194193295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003632671111387746,
      "loss": 5.0454,
      "mean_token_accuracy": 0.21568792760372163,
      "num_tokens": 75287060.0,
      "step": 33720
    },
    {
      "entropy": 5.537593841552734,
      "epoch": 3.787667771101252,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036323017941372203,
      "loss": 5.0361,
      "mean_token_accuracy": 0.22265829741954804,
      "num_tokens": 75297862.0,
      "step": 33725
    },
    {
      "entropy": 5.5138191223144535,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003631932448793864,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21462939083576202,
      "num_tokens": 75308844.0,
      "step": 33730
    },
    {
      "entropy": 5.517885732650757,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003631563075369436,
      "loss": 5.094,
      "mean_token_accuracy": 0.21445232629776,
      "num_tokens": 75321035.0,
      "step": 33735
    },
    {
      "entropy": 5.361093950271607,
      "epoch": 3.789352501825125,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2222513198852539,
      "num_tokens": 75331407.0,
      "step": 33740
    },
    {
      "entropy": 5.418775653839111,
      "epoch": 3.7899140787330827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22206834852695465,
      "num_tokens": 75343082.0,
      "step": 33745
    },
    {
      "entropy": 5.4102832794189455,
      "epoch": 3.79047565564104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2261229857802391,
      "num_tokens": 75353345.0,
      "step": 33750
    },
    {
      "entropy": 5.521029376983643,
      "epoch": 3.7910372325489976,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003630085301096245,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21755565106868743,
      "num_tokens": 75364039.0,
      "step": 33755
    },
    {
      "entropy": 5.5323279857635494,
      "epoch": 3.7915988094569553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036297157874428933,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21951166689395904,
      "num_tokens": 75374166.0,
      "step": 33760
    },
    {
      "entropy": 5.46776762008667,
      "epoch": 3.7921603863649125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036293462457790505,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21441923081874847,
      "num_tokens": 75384792.0,
      "step": 33765
    },
    {
      "entropy": 5.4583290100097654,
      "epoch": 3.79272196327287,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036289766761164814,
      "loss": 5.0363,
      "mean_token_accuracy": 0.20996575951576232,
      "num_tokens": 75395478.0,
      "step": 33770
    },
    {
      "entropy": 5.444092559814453,
      "epoch": 3.793283540180828,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036286070784669543,
      "loss": 5.0351,
      "mean_token_accuracy": 0.218122398853302,
      "num_tokens": 75406293.0,
      "step": 33775
    },
    {
      "entropy": 5.485165309906006,
      "epoch": 3.7938451170887855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036282374528422373,
      "loss": 5.0394,
      "mean_token_accuracy": 0.22402476519346237,
      "num_tokens": 75417588.0,
      "step": 33780
    },
    {
      "entropy": 5.357405424118042,
      "epoch": 3.794406693996743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22006171196699142,
      "num_tokens": 75428651.0,
      "step": 33785
    },
    {
      "entropy": 5.484849977493286,
      "epoch": 3.7949682709047003,
      "grad_norm": 1.09375,
      "learning_rate": 0.000362749811771431,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21868566125631334,
      "num_tokens": 75438808.0,
      "step": 33790
    },
    {
      "entropy": 5.437881994247436,
      "epoch": 3.795529847812658,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.8727,
      "mean_token_accuracy": 0.22073927521705627,
      "num_tokens": 75449326.0,
      "step": 33795
    },
    {
      "entropy": 5.4278772354125975,
      "epoch": 3.7960914247206157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003626758670826861,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2183094620704651,
      "num_tokens": 75460499.0,
      "step": 33800
    },
    {
      "entropy": 5.3939450740814205,
      "epoch": 3.796653001628573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003626388905502747,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22662028968334197,
      "num_tokens": 75471031.0,
      "step": 33805
    },
    {
      "entropy": 5.454228305816651,
      "epoch": 3.7972145785365305,
      "grad_norm": 1.015625,
      "learning_rate": 0.000362601911227407,
      "loss": 5.0061,
      "mean_token_accuracy": 0.2184986114501953,
      "num_tokens": 75483608.0,
      "step": 33810
    },
    {
      "entropy": 5.515616321563721,
      "epoch": 3.797776155444488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003625649291152604,
      "loss": 5.0392,
      "mean_token_accuracy": 0.22082566618919372,
      "num_tokens": 75495221.0,
      "step": 33815
    },
    {
      "entropy": 5.452381372451782,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036252794421501257,
      "loss": 4.95,
      "mean_token_accuracy": 0.22618527263402938,
      "num_tokens": 75507245.0,
      "step": 33820
    },
    {
      "entropy": 5.496446895599365,
      "epoch": 3.798899309260403,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036249095652784096,
      "loss": 5.0793,
      "mean_token_accuracy": 0.21224284768104554,
      "num_tokens": 75518525.0,
      "step": 33825
    },
    {
      "entropy": 5.424894666671753,
      "epoch": 3.7994608861683608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036245396605492344,
      "loss": 4.974,
      "mean_token_accuracy": 0.2139137864112854,
      "num_tokens": 75529674.0,
      "step": 33830
    },
    {
      "entropy": 5.392486238479615,
      "epoch": 3.8000224630763184,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22481405586004258,
      "num_tokens": 75539596.0,
      "step": 33835
    },
    {
      "entropy": 5.369180727005005,
      "epoch": 3.800584039984276,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036237997675656167,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22570021003484725,
      "num_tokens": 75551164.0,
      "step": 33840
    },
    {
      "entropy": 5.466789054870605,
      "epoch": 3.8011456168922333,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003623429779334732,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21122770607471467,
      "num_tokens": 75563758.0,
      "step": 33845
    },
    {
      "entropy": 5.5225615978240965,
      "epoch": 3.801707193800191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036230597632935053,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2177860513329506,
      "num_tokens": 75574332.0,
      "step": 33850
    },
    {
      "entropy": 5.47590742111206,
      "epoch": 3.8022687707081486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003622689719453717,
      "loss": 5.095,
      "mean_token_accuracy": 0.2096468448638916,
      "num_tokens": 75586229.0,
      "step": 33855
    },
    {
      "entropy": 5.469210433959961,
      "epoch": 3.802830347616106,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036223196478271495,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22487681359052658,
      "num_tokens": 75597396.0,
      "step": 33860
    },
    {
      "entropy": 5.4570338249206545,
      "epoch": 3.8033919245240635,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003621949548425585,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21373271346092224,
      "num_tokens": 75609171.0,
      "step": 33865
    },
    {
      "entropy": 5.516405963897705,
      "epoch": 3.803953501432021,
      "grad_norm": 1.0625,
      "learning_rate": 0.000362157942126081,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21602724492549896,
      "num_tokens": 75619855.0,
      "step": 33870
    },
    {
      "entropy": 5.554366970062256,
      "epoch": 3.804515078339979,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003621209266344608,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2131578207015991,
      "num_tokens": 75631810.0,
      "step": 33875
    },
    {
      "entropy": 5.40648078918457,
      "epoch": 3.8050766552479365,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003620839083688765,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22233926504850388,
      "num_tokens": 75644894.0,
      "step": 33880
    },
    {
      "entropy": 5.471193790435791,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21647387444972993,
      "num_tokens": 75657357.0,
      "step": 33885
    },
    {
      "entropy": 5.4623924732208256,
      "epoch": 3.8061998090638514,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21918362081050874,
      "num_tokens": 75667357.0,
      "step": 33890
    },
    {
      "entropy": 5.390598392486572,
      "epoch": 3.806761385971809,
      "grad_norm": 1.0,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.885,
      "mean_token_accuracy": 0.2233831524848938,
      "num_tokens": 75678758.0,
      "step": 33895
    },
    {
      "entropy": 5.476257801055908,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2218412861227989,
      "num_tokens": 75688392.0,
      "step": 33900
    },
    {
      "entropy": 5.442362451553345,
      "epoch": 3.807884539787724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036189877547275023,
      "loss": 4.959,
      "mean_token_accuracy": 0.22106846868991853,
      "num_tokens": 75699839.0,
      "step": 33905
    },
    {
      "entropy": 5.499070119857788,
      "epoch": 3.8084461166956816,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2181176632642746,
      "num_tokens": 75711877.0,
      "step": 33910
    },
    {
      "entropy": 5.40525712966919,
      "epoch": 3.8090076936036388,
      "grad_norm": 1.0,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.9073,
      "mean_token_accuracy": 0.21960092186927796,
      "num_tokens": 75721646.0,
      "step": 33915
    },
    {
      "entropy": 5.4840850830078125,
      "epoch": 3.8095692705115964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003617876625229543,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20621733963489533,
      "num_tokens": 75733291.0,
      "step": 33920
    },
    {
      "entropy": 5.40609827041626,
      "epoch": 3.810130847419554,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.9913,
      "mean_token_accuracy": 0.22683339267969133,
      "num_tokens": 75744475.0,
      "step": 33925
    },
    {
      "entropy": 5.379101514816284,
      "epoch": 3.8106924243275118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22152695804834366,
      "num_tokens": 75755098.0,
      "step": 33930
    },
    {
      "entropy": 5.485227251052857,
      "epoch": 3.8112540012354694,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.9904,
      "mean_token_accuracy": 0.22521879225969316,
      "num_tokens": 75767405.0,
      "step": 33935
    },
    {
      "entropy": 5.464018678665161,
      "epoch": 3.8118155781434266,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036163947324179186,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2116781145334244,
      "num_tokens": 75778614.0,
      "step": 33940
    },
    {
      "entropy": 5.387935590744019,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22904395312070847,
      "num_tokens": 75789087.0,
      "step": 33945
    },
    {
      "entropy": 5.408736801147461,
      "epoch": 3.812938731959342,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21361952424049377,
      "num_tokens": 75799431.0,
      "step": 33950
    },
    {
      "entropy": 5.421758794784546,
      "epoch": 3.813500308867299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036152830231113013,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22293393909931183,
      "num_tokens": 75810149.0,
      "step": 33955
    },
    {
      "entropy": 5.450659942626953,
      "epoch": 3.814061885775257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003614912398224802,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20948597490787507,
      "num_tokens": 75820072.0,
      "step": 33960
    },
    {
      "entropy": 5.48017520904541,
      "epoch": 3.8146234626832145,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003614541745799156,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21199607104063034,
      "num_tokens": 75831017.0,
      "step": 33965
    },
    {
      "entropy": 5.406284809112549,
      "epoch": 3.815185039591172,
      "grad_norm": 1.0,
      "learning_rate": 0.0003614171065846166,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21443890184164047,
      "num_tokens": 75842606.0,
      "step": 33970
    },
    {
      "entropy": 5.425779533386231,
      "epoch": 3.81574661649913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003613800358377635,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21871304512023926,
      "num_tokens": 75853820.0,
      "step": 33975
    },
    {
      "entropy": 5.401141929626465,
      "epoch": 3.816308193407087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2211194708943367,
      "num_tokens": 75864959.0,
      "step": 33980
    },
    {
      "entropy": 5.4620154857635494,
      "epoch": 3.8168697703150447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036130588609411597,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22302900850772858,
      "num_tokens": 75877247.0,
      "step": 33985
    },
    {
      "entropy": 5.5039143562316895,
      "epoch": 3.8174313472230024,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036126880709968264,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20392042249441147,
      "num_tokens": 75888654.0,
      "step": 33990
    },
    {
      "entropy": 5.433786153793335,
      "epoch": 3.8179929241309596,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21578042656183244,
      "num_tokens": 75899953.0,
      "step": 33995
    },
    {
      "entropy": 5.454100656509399,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003611946408714998,
      "loss": 4.9306,
      "mean_token_accuracy": 0.22434334307909012,
      "num_tokens": 75911174.0,
      "step": 34000
    },
    {
      "entropy": 5.3708292007446286,
      "epoch": 3.819116077946875,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.9877,
      "mean_token_accuracy": 0.22387509644031525,
      "num_tokens": 75922221.0,
      "step": 34005
    },
    {
      "entropy": 5.309935092926025,
      "epoch": 3.819677654854832,
      "grad_norm": 1.0625,
      "learning_rate": 0.000361120463665434,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21740428507328033,
      "num_tokens": 75932630.0,
      "step": 34010
    },
    {
      "entropy": 5.406970691680908,
      "epoch": 3.82023923176279,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003610833709486471,
      "loss": 4.971,
      "mean_token_accuracy": 0.2208850249648094,
      "num_tokens": 75944490.0,
      "step": 34015
    },
    {
      "entropy": 5.427387380599976,
      "epoch": 3.8208008086707475,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22438758313655854,
      "num_tokens": 75956608.0,
      "step": 34020
    },
    {
      "entropy": 5.46316933631897,
      "epoch": 3.821362385578705,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036100917729347073,
      "loss": 5.0395,
      "mean_token_accuracy": 0.2136031597852707,
      "num_tokens": 75967516.0,
      "step": 34025
    },
    {
      "entropy": 5.3868663787841795,
      "epoch": 3.8219239624866628,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036097207635744363,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22432674914598466,
      "num_tokens": 75978329.0,
      "step": 34030
    },
    {
      "entropy": 5.51148042678833,
      "epoch": 3.82248553939462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22754869759082794,
      "num_tokens": 75990170.0,
      "step": 34035
    },
    {
      "entropy": 5.4308576583862305,
      "epoch": 3.8230471163025777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000360897866274418,
      "loss": 4.9325,
      "mean_token_accuracy": 0.220664319396019,
      "num_tokens": 76002368.0,
      "step": 34040
    },
    {
      "entropy": 5.381850099563598,
      "epoch": 3.8236086932105353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22260091900825502,
      "num_tokens": 76012867.0,
      "step": 34045
    },
    {
      "entropy": 5.446662425994873,
      "epoch": 3.8241702701184925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036082364525130684,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2154087871313095,
      "num_tokens": 76023900.0,
      "step": 34050
    },
    {
      "entropy": 5.419698429107666,
      "epoch": 3.82473184702645,
      "grad_norm": 1.125,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2226670950651169,
      "num_tokens": 76034296.0,
      "step": 34055
    },
    {
      "entropy": 5.473008394241333,
      "epoch": 3.825293423934408,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036074941329756327,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2161953940987587,
      "num_tokens": 76045355.0,
      "step": 34060
    },
    {
      "entropy": 5.420407295227051,
      "epoch": 3.8258550008423655,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23162830173969268,
      "num_tokens": 76056158.0,
      "step": 34065
    },
    {
      "entropy": 5.396338844299317,
      "epoch": 3.826416577750323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036067517042264147,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2193853110074997,
      "num_tokens": 76066930.0,
      "step": 34070
    },
    {
      "entropy": 5.4406452655792235,
      "epoch": 3.8269781546582804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22253908663988115,
      "num_tokens": 76078791.0,
      "step": 34075
    },
    {
      "entropy": 5.527487516403198,
      "epoch": 3.827539731566238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003606009166359972,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21527137905359267,
      "num_tokens": 76090356.0,
      "step": 34080
    },
    {
      "entropy": 5.373996162414551,
      "epoch": 3.8281013084741957,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036056378565373434,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2256527692079544,
      "num_tokens": 76101137.0,
      "step": 34085
    },
    {
      "entropy": 5.327095603942871,
      "epoch": 3.828662885382153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22442275434732437,
      "num_tokens": 76111661.0,
      "step": 34090
    },
    {
      "entropy": 5.469878244400024,
      "epoch": 3.8292244622901106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003604895155172386,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21592434644699096,
      "num_tokens": 76122670.0,
      "step": 34095
    },
    {
      "entropy": 5.454479742050171,
      "epoch": 3.8297860391980683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003604523763653706,
      "loss": 5.0799,
      "mean_token_accuracy": 0.21291853487491608,
      "num_tokens": 76133899.0,
      "step": 34100
    },
    {
      "entropy": 5.477661561965943,
      "epoch": 3.8303476161060255,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036041523449266573,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2102467730641365,
      "num_tokens": 76144865.0,
      "step": 34105
    },
    {
      "entropy": 5.507882595062256,
      "epoch": 3.830909193013983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036037808990030667,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21850018501281737,
      "num_tokens": 76157392.0,
      "step": 34110
    },
    {
      "entropy": 5.489041566848755,
      "epoch": 3.831470769921941,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003603409425894761,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21719503849744798,
      "num_tokens": 76168339.0,
      "step": 34115
    },
    {
      "entropy": 5.36748571395874,
      "epoch": 3.8320323468298985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22693848758935928,
      "num_tokens": 76179377.0,
      "step": 34120
    },
    {
      "entropy": 5.428616333007812,
      "epoch": 3.832593923737856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21835720390081406,
      "num_tokens": 76190030.0,
      "step": 34125
    },
    {
      "entropy": 5.344652128219605,
      "epoch": 3.8331555006458133,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22251410484313966,
      "num_tokens": 76201339.0,
      "step": 34130
    },
    {
      "entropy": 5.461196184158325,
      "epoch": 3.833717077553771,
      "grad_norm": 1.0625,
      "learning_rate": 0.000360192326185096,
      "loss": 5.0296,
      "mean_token_accuracy": 0.22054672092199326,
      "num_tokens": 76211760.0,
      "step": 34135
    },
    {
      "entropy": 5.483848333358765,
      "epoch": 3.8342786544617287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003601551652996514,
      "loss": 5.0195,
      "mean_token_accuracy": 0.2158883035182953,
      "num_tokens": 76222573.0,
      "step": 34140
    },
    {
      "entropy": 5.508450984954834,
      "epoch": 3.834840231369686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003601180017028334,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20997832715511322,
      "num_tokens": 76233547.0,
      "step": 34145
    },
    {
      "entropy": 5.424777173995972,
      "epoch": 3.8354018082776435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003600808353958252,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21638840734958648,
      "num_tokens": 76246067.0,
      "step": 34150
    },
    {
      "entropy": 5.34718279838562,
      "epoch": 3.835963385185601,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003600436663798102,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23040239959955217,
      "num_tokens": 76257698.0,
      "step": 34155
    },
    {
      "entropy": 5.3508607864379885,
      "epoch": 3.836524962093559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2235518917441368,
      "num_tokens": 76268533.0,
      "step": 34160
    },
    {
      "entropy": 5.358892965316772,
      "epoch": 3.8370865390015165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22946714460849763,
      "num_tokens": 76278529.0,
      "step": 34165
    },
    {
      "entropy": 5.359953546524048,
      "epoch": 3.8376481159094737,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2127763018012047,
      "num_tokens": 76289461.0,
      "step": 34170
    },
    {
      "entropy": 5.428842449188233,
      "epoch": 3.8382096928174314,
      "grad_norm": 1.03125,
      "learning_rate": 0.000359894963249353,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2193623185157776,
      "num_tokens": 76300641.0,
      "step": 34175
    },
    {
      "entropy": 5.473764944076538,
      "epoch": 3.838771269725389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035985778070605785,
      "loss": 5.0683,
      "mean_token_accuracy": 0.2116386741399765,
      "num_tokens": 76310918.0,
      "step": 34180
    },
    {
      "entropy": 5.429519748687744,
      "epoch": 3.8393328466333463,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003598205954608581,
      "loss": 4.878,
      "mean_token_accuracy": 0.2242760568857193,
      "num_tokens": 76323037.0,
      "step": 34185
    },
    {
      "entropy": 5.4218635082244875,
      "epoch": 3.839894423541304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035978340751493767,
      "loss": 4.9553,
      "mean_token_accuracy": 0.21888386011123656,
      "num_tokens": 76334252.0,
      "step": 34190
    },
    {
      "entropy": 5.438529777526855,
      "epoch": 3.8404560004492616,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035974621686948063,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21493897289037706,
      "num_tokens": 76344928.0,
      "step": 34195
    },
    {
      "entropy": 5.505938482284546,
      "epoch": 3.841017577357219,
      "grad_norm": 1.125,
      "learning_rate": 0.0003597090235256712,
      "loss": 5.0784,
      "mean_token_accuracy": 0.2087486431002617,
      "num_tokens": 76356529.0,
      "step": 34200
    },
    {
      "entropy": 5.4499349117279055,
      "epoch": 3.8415791542651765,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21398191452026366,
      "num_tokens": 76366842.0,
      "step": 34205
    },
    {
      "entropy": 5.469132614135742,
      "epoch": 3.842140731173134,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003596346287477322,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21382647305727004,
      "num_tokens": 76377109.0,
      "step": 34210
    },
    {
      "entropy": 5.43293342590332,
      "epoch": 3.842702308081092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21816105544567108,
      "num_tokens": 76387621.0,
      "step": 34215
    },
    {
      "entropy": 5.490869665145874,
      "epoch": 3.8432638849890495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003595602231905957,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21308456510305404,
      "num_tokens": 76398537.0,
      "step": 34220
    },
    {
      "entropy": 5.463173961639404,
      "epoch": 3.8438254618970067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21746602952480315,
      "num_tokens": 76410192.0,
      "step": 34225
    },
    {
      "entropy": 5.422406196594238,
      "epoch": 3.8443870388049644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035948580686373803,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22337272614240647,
      "num_tokens": 76420024.0,
      "step": 34230
    },
    {
      "entropy": 5.445324230194092,
      "epoch": 3.844948615712922,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035944859466462554,
      "loss": 5.0188,
      "mean_token_accuracy": 0.22217363864183426,
      "num_tokens": 76430833.0,
      "step": 34235
    },
    {
      "entropy": 5.411154556274414,
      "epoch": 3.8455101926208792,
      "grad_norm": 1.0625,
      "learning_rate": 0.000359411379776637,
      "loss": 4.987,
      "mean_token_accuracy": 0.2186681717634201,
      "num_tokens": 76442800.0,
      "step": 34240
    },
    {
      "entropy": 5.461304998397827,
      "epoch": 3.846071769528837,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035937416220095745,
      "loss": 5.0122,
      "mean_token_accuracy": 0.2176061436533928,
      "num_tokens": 76454988.0,
      "step": 34245
    },
    {
      "entropy": 5.572785568237305,
      "epoch": 3.8466333464367946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003593369419387718,
      "loss": 5.1105,
      "mean_token_accuracy": 0.207306145131588,
      "num_tokens": 76466523.0,
      "step": 34250
    },
    {
      "entropy": 5.578875350952148,
      "epoch": 3.847194923344752,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003592997189912652,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20070233792066575,
      "num_tokens": 76478691.0,
      "step": 34255
    },
    {
      "entropy": 5.483144474029541,
      "epoch": 3.84775650025271,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035926249335962296,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21316317468881607,
      "num_tokens": 76490544.0,
      "step": 34260
    },
    {
      "entropy": 5.407374668121338,
      "epoch": 3.848318077160667,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.9015,
      "mean_token_accuracy": 0.22125745713710784,
      "num_tokens": 76500269.0,
      "step": 34265
    },
    {
      "entropy": 5.360837841033936,
      "epoch": 3.8488796540686248,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003591880340486723,
      "loss": 4.9109,
      "mean_token_accuracy": 0.21706343144178392,
      "num_tokens": 76511458.0,
      "step": 34270
    },
    {
      "entropy": 5.4200598239898685,
      "epoch": 3.8494412309765824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003591508003717347,
      "loss": 5.0226,
      "mean_token_accuracy": 0.2169681519269943,
      "num_tokens": 76521722.0,
      "step": 34275
    },
    {
      "entropy": 5.498661136627197,
      "epoch": 3.8500028078845396,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035911356401540306,
      "loss": 5.0844,
      "mean_token_accuracy": 0.22062325477600098,
      "num_tokens": 76532935.0,
      "step": 34280
    },
    {
      "entropy": 5.445328283309936,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2201479196548462,
      "num_tokens": 76543545.0,
      "step": 34285
    },
    {
      "entropy": 5.3701591968536375,
      "epoch": 3.851125961700455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22677624523639678,
      "num_tokens": 76553621.0,
      "step": 34290
    },
    {
      "entropy": 5.465557670593261,
      "epoch": 3.851687538608412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035900183888190005,
      "loss": 4.991,
      "mean_token_accuracy": 0.22425406724214553,
      "num_tokens": 76563154.0,
      "step": 34295
    },
    {
      "entropy": 5.463944768905639,
      "epoch": 3.85224911551637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21515686064958572,
      "num_tokens": 76573930.0,
      "step": 34300
    },
    {
      "entropy": 5.381326913833618,
      "epoch": 3.8528106924243275,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035892734208433264,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22511878609657288,
      "num_tokens": 76585216.0,
      "step": 34305
    },
    {
      "entropy": 5.411333465576172,
      "epoch": 3.853372269332285,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035889008967653725,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22289366871118546,
      "num_tokens": 76596479.0,
      "step": 34310
    },
    {
      "entropy": 5.514521789550781,
      "epoch": 3.853933846240243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003588528345976489,
      "loss": 5.0887,
      "mean_token_accuracy": 0.21017200052738189,
      "num_tokens": 76607123.0,
      "step": 34315
    },
    {
      "entropy": 5.49645414352417,
      "epoch": 3.8544954231482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003588155768488537,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22094994485378266,
      "num_tokens": 76616898.0,
      "step": 34320
    },
    {
      "entropy": 5.377840614318847,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035877831643133803,
      "loss": 4.8353,
      "mean_token_accuracy": 0.22807666659355164,
      "num_tokens": 76627528.0,
      "step": 34325
    },
    {
      "entropy": 5.403672456741333,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003587410533462883,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22213240712881088,
      "num_tokens": 76639173.0,
      "step": 34330
    },
    {
      "entropy": 5.4282958984375,
      "epoch": 3.8561801538720726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035870378759489104,
      "loss": 5.0311,
      "mean_token_accuracy": 0.211903914809227,
      "num_tokens": 76648970.0,
      "step": 34335
    },
    {
      "entropy": 5.381725740432739,
      "epoch": 3.8567417307800302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.9251,
      "mean_token_accuracy": 0.22124284654855728,
      "num_tokens": 76659101.0,
      "step": 34340
    },
    {
      "entropy": 5.4370057582855225,
      "epoch": 3.857303307687988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003586292480977998,
      "loss": 4.9739,
      "mean_token_accuracy": 0.218184195458889,
      "num_tokens": 76670087.0,
      "step": 34345
    },
    {
      "entropy": 5.47442135810852,
      "epoch": 3.8578648845959456,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035859197435447927,
      "loss": 5.0341,
      "mean_token_accuracy": 0.22029916644096376,
      "num_tokens": 76681116.0,
      "step": 34350
    },
    {
      "entropy": 5.42890739440918,
      "epoch": 3.8584264615039032,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035855469794955787,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21057007610797882,
      "num_tokens": 76692050.0,
      "step": 34355
    },
    {
      "entropy": 5.57538857460022,
      "epoch": 3.8589880384118604,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035851741888422254,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21141764372587205,
      "num_tokens": 76704338.0,
      "step": 34360
    },
    {
      "entropy": 5.4070631980896,
      "epoch": 3.859549615319818,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003584801371596601,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2177896648645401,
      "num_tokens": 76715300.0,
      "step": 34365
    },
    {
      "entropy": 5.483849906921387,
      "epoch": 3.8601111922277758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003584428527770578,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21516670137643815,
      "num_tokens": 76726654.0,
      "step": 34370
    },
    {
      "entropy": 5.30983567237854,
      "epoch": 3.860672769135733,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035840556573760266,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2251930132508278,
      "num_tokens": 76737087.0,
      "step": 34375
    },
    {
      "entropy": 5.378384351730347,
      "epoch": 3.8612343460436906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22226417660713196,
      "num_tokens": 76748006.0,
      "step": 34380
    },
    {
      "entropy": 5.339454174041748,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035833098369288303,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23623059839010238,
      "num_tokens": 76758063.0,
      "step": 34385
    },
    {
      "entropy": 5.3873186111450195,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.963,
      "mean_token_accuracy": 0.22257017493247985,
      "num_tokens": 76769080.0,
      "step": 34390
    },
    {
      "entropy": 5.420530605316162,
      "epoch": 3.862919076767563,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035825639103500005,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2185945212841034,
      "num_tokens": 76780743.0,
      "step": 34395
    },
    {
      "entropy": 5.415958309173584,
      "epoch": 3.863480653675521,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035821909072909113,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2230172261595726,
      "num_tokens": 76792277.0,
      "step": 34400
    },
    {
      "entropy": 5.412454557418823,
      "epoch": 3.8640422305834785,
      "grad_norm": 1.125,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22681985050439835,
      "num_tokens": 76802979.0,
      "step": 34405
    },
    {
      "entropy": 5.36803503036499,
      "epoch": 3.864603807491436,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22431184053421022,
      "num_tokens": 76815016.0,
      "step": 34410
    },
    {
      "entropy": 5.3744056224823,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22783102393150328,
      "num_tokens": 76826360.0,
      "step": 34415
    },
    {
      "entropy": 5.428078079223633,
      "epoch": 3.865726961307351,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035806986302005205,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21701684296131135,
      "num_tokens": 76838876.0,
      "step": 34420
    },
    {
      "entropy": 5.337879610061646,
      "epoch": 3.8662885382153087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2264888823032379,
      "num_tokens": 76849156.0,
      "step": 34425
    },
    {
      "entropy": 5.462436628341675,
      "epoch": 3.866850115123266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035799523329092093,
      "loss": 5.0689,
      "mean_token_accuracy": 0.21040559858083724,
      "num_tokens": 76859683.0,
      "step": 34430
    },
    {
      "entropy": 5.374632358551025,
      "epoch": 3.8674116920312236,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035795791446186073,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2256786197423935,
      "num_tokens": 76870509.0,
      "step": 34435
    },
    {
      "entropy": 5.4397424221038815,
      "epoch": 3.8679732689391813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003579205929913882,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21809856593608856,
      "num_tokens": 76882181.0,
      "step": 34440
    },
    {
      "entropy": 5.427903079986573,
      "epoch": 3.868534845847139,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003578832688806919,
      "loss": 5.057,
      "mean_token_accuracy": 0.20871184021234512,
      "num_tokens": 76894709.0,
      "step": 34445
    },
    {
      "entropy": 5.456011390686035,
      "epoch": 3.8690964227550966,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003578459421309602,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22716787904500962,
      "num_tokens": 76905974.0,
      "step": 34450
    },
    {
      "entropy": 5.559581470489502,
      "epoch": 3.869657999663054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003578086127433816,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20920597314834594,
      "num_tokens": 76918009.0,
      "step": 34455
    },
    {
      "entropy": 5.47747769355774,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035777128071914454,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22008588165044785,
      "num_tokens": 76928750.0,
      "step": 34460
    },
    {
      "entropy": 5.479839515686035,
      "epoch": 3.870781153478969,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003577339460594378,
      "loss": 5.0887,
      "mean_token_accuracy": 0.21688045412302018,
      "num_tokens": 76939989.0,
      "step": 34465
    },
    {
      "entropy": 5.331789875030518,
      "epoch": 3.8713427303869263,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003576966087654502,
      "loss": 4.878,
      "mean_token_accuracy": 0.22640405148267745,
      "num_tokens": 76950026.0,
      "step": 34470
    },
    {
      "entropy": 5.47308087348938,
      "epoch": 3.871904307294884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035765926883837054,
      "loss": 5.094,
      "mean_token_accuracy": 0.21217654198408126,
      "num_tokens": 76961198.0,
      "step": 34475
    },
    {
      "entropy": 5.417191123962402,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.125,
      "learning_rate": 0.0003576219262793877,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22268553227186202,
      "num_tokens": 76973012.0,
      "step": 34480
    },
    {
      "entropy": 5.455790615081787,
      "epoch": 3.873027461110799,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2195522144436836,
      "num_tokens": 76983613.0,
      "step": 34485
    },
    {
      "entropy": 5.529917669296265,
      "epoch": 3.8735890380187565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003575472332704684,
      "loss": 5.0991,
      "mean_token_accuracy": 0.21137507110834122,
      "num_tokens": 76995471.0,
      "step": 34490
    },
    {
      "entropy": 5.380887079238891,
      "epoch": 3.874150614926714,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035750988282291037,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2231060028076172,
      "num_tokens": 77005777.0,
      "step": 34495
    },
    {
      "entropy": 5.4737913608551025,
      "epoch": 3.874712191834672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035747252974820565,
      "loss": 5.0647,
      "mean_token_accuracy": 0.21530809998512268,
      "num_tokens": 77017099.0,
      "step": 34500
    },
    {
      "entropy": 5.412206315994263,
      "epoch": 3.8752737687426295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035743517404754354,
      "loss": 5.0431,
      "mean_token_accuracy": 0.21918287724256516,
      "num_tokens": 77028350.0,
      "step": 34505
    },
    {
      "entropy": 5.415634202957153,
      "epoch": 3.8758353456505867,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003573978157221136,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22219243496656418,
      "num_tokens": 77039355.0,
      "step": 34510
    },
    {
      "entropy": 5.522175073623657,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035736045477310517,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20986288040876389,
      "num_tokens": 77049966.0,
      "step": 34515
    },
    {
      "entropy": 5.474979496002197,
      "epoch": 3.876958499466502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035732309120170786,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2189761221408844,
      "num_tokens": 77061102.0,
      "step": 34520
    },
    {
      "entropy": 5.464164876937867,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035728572500911155,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22051455229520797,
      "num_tokens": 77071516.0,
      "step": 34525
    },
    {
      "entropy": 5.46901569366455,
      "epoch": 3.878081653282417,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035724835619650573,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21531451046466826,
      "num_tokens": 77081924.0,
      "step": 34530
    },
    {
      "entropy": 5.434382104873658,
      "epoch": 3.8786432301903746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003572109847650804,
      "loss": 4.9662,
      "mean_token_accuracy": 0.2175379365682602,
      "num_tokens": 77093064.0,
      "step": 34535
    },
    {
      "entropy": 5.456568813323974,
      "epoch": 3.8792048070983323,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035717361071602546,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2229633465409279,
      "num_tokens": 77104428.0,
      "step": 34540
    },
    {
      "entropy": 5.429973840713501,
      "epoch": 3.87976638400629,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22963437736034392,
      "num_tokens": 77114949.0,
      "step": 34545
    },
    {
      "entropy": 5.479398012161255,
      "epoch": 3.880327960914247,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035709885476978666,
      "loss": 5.055,
      "mean_token_accuracy": 0.2151501312851906,
      "num_tokens": 77125178.0,
      "step": 34550
    },
    {
      "entropy": 5.374143123626709,
      "epoch": 3.880889537822205,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035706147287498303,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22760603725910186,
      "num_tokens": 77135967.0,
      "step": 34555
    },
    {
      "entropy": 5.499357461929321,
      "epoch": 3.8814511147301625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035702408836731043,
      "loss": 5.0802,
      "mean_token_accuracy": 0.2169778048992157,
      "num_tokens": 77148052.0,
      "step": 34560
    },
    {
      "entropy": 5.476116132736206,
      "epoch": 3.8820126916381197,
      "grad_norm": 1.015625,
      "learning_rate": 0.000356986701247959,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2194388747215271,
      "num_tokens": 77159008.0,
      "step": 34565
    },
    {
      "entropy": 5.401842069625855,
      "epoch": 3.8825742685460773,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.9482,
      "mean_token_accuracy": 0.216604346036911,
      "num_tokens": 77169846.0,
      "step": 34570
    },
    {
      "entropy": 5.403509569168091,
      "epoch": 3.883135845454035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003569119191789815,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21984901577234267,
      "num_tokens": 77182881.0,
      "step": 34575
    },
    {
      "entropy": 5.432447624206543,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035687452423173655,
      "loss": 4.96,
      "mean_token_accuracy": 0.21972415447235108,
      "num_tokens": 77193852.0,
      "step": 34580
    },
    {
      "entropy": 5.486018562316895,
      "epoch": 3.88425899926995,
      "grad_norm": 1.0,
      "learning_rate": 0.000356837126677575,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2171301931142807,
      "num_tokens": 77205141.0,
      "step": 34585
    },
    {
      "entropy": 5.523805618286133,
      "epoch": 3.8848205761779075,
      "grad_norm": 1.0,
      "learning_rate": 0.0003567997265176876,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20526204705238343,
      "num_tokens": 77217521.0,
      "step": 34590
    },
    {
      "entropy": 5.534891128540039,
      "epoch": 3.885382153085865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035676232375326513,
      "loss": 5.0949,
      "mean_token_accuracy": 0.2123275727033615,
      "num_tokens": 77227468.0,
      "step": 34595
    },
    {
      "entropy": 5.3982603549957275,
      "epoch": 3.885943729993823,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.9593,
      "mean_token_accuracy": 0.21968848407268524,
      "num_tokens": 77237908.0,
      "step": 34600
    },
    {
      "entropy": 5.367333269119262,
      "epoch": 3.88650530690178,
      "grad_norm": 1.0625,
      "learning_rate": 0.000356687510415579,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22525186836719513,
      "num_tokens": 77248621.0,
      "step": 34605
    },
    {
      "entropy": 5.454500341415406,
      "epoch": 3.8870668838097377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003566500998446973,
      "loss": 5.048,
      "mean_token_accuracy": 0.21134770214557647,
      "num_tokens": 77260261.0,
      "step": 34610
    },
    {
      "entropy": 5.420894479751587,
      "epoch": 3.8876284607176954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035661268667404485,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21590457409620284,
      "num_tokens": 77270985.0,
      "step": 34615
    },
    {
      "entropy": 5.458020305633545,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003565752709048127,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21969434916973113,
      "num_tokens": 77281876.0,
      "step": 34620
    },
    {
      "entropy": 5.372719764709473,
      "epoch": 3.8887516145336103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003565378525381924,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22249452918767929,
      "num_tokens": 77293141.0,
      "step": 34625
    },
    {
      "entropy": 5.460032939910889,
      "epoch": 3.889313191441568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2260063797235489,
      "num_tokens": 77305723.0,
      "step": 34630
    },
    {
      "entropy": 5.37106237411499,
      "epoch": 3.8898747683495256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22619739323854446,
      "num_tokens": 77317736.0,
      "step": 34635
    },
    {
      "entropy": 5.407139539718628,
      "epoch": 3.8904363452574833,
      "grad_norm": 1.0,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21480463892221452,
      "num_tokens": 77328293.0,
      "step": 34640
    },
    {
      "entropy": 5.458521604537964,
      "epoch": 3.8909979221654405,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003563881531216579,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21947939693927765,
      "num_tokens": 77339371.0,
      "step": 34645
    },
    {
      "entropy": 5.467627143859863,
      "epoch": 3.891559499073398,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003563507217859691,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21283202320337297,
      "num_tokens": 77350234.0,
      "step": 34650
    },
    {
      "entropy": 5.45972580909729,
      "epoch": 3.892121075981356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003563132878600419,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21943815648555756,
      "num_tokens": 77361144.0,
      "step": 34655
    },
    {
      "entropy": 5.4724077701568605,
      "epoch": 3.892682652889313,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003562758513450678,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22193819433450698,
      "num_tokens": 77372090.0,
      "step": 34660
    },
    {
      "entropy": 5.485793113708496,
      "epoch": 3.8932442297972707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003562384122422393,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21392259895801544,
      "num_tokens": 77382974.0,
      "step": 34665
    },
    {
      "entropy": 5.47131929397583,
      "epoch": 3.8938058067052284,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003562009705527481,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21489569991827012,
      "num_tokens": 77395971.0,
      "step": 34670
    },
    {
      "entropy": 5.4896337509155275,
      "epoch": 3.8943673836131856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035616352627778665,
      "loss": 5.0015,
      "mean_token_accuracy": 0.21083145439624787,
      "num_tokens": 77405776.0,
      "step": 34675
    },
    {
      "entropy": 5.492715644836426,
      "epoch": 3.8949289605211432,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035612607941854695,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2172098383307457,
      "num_tokens": 77417668.0,
      "step": 34680
    },
    {
      "entropy": 5.4678185939788815,
      "epoch": 3.895490537429101,
      "grad_norm": 1.0,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21848851293325425,
      "num_tokens": 77428581.0,
      "step": 34685
    },
    {
      "entropy": 5.518340539932251,
      "epoch": 3.8960521143370586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003560511779520025,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21590711027383805,
      "num_tokens": 77441587.0,
      "step": 34690
    },
    {
      "entropy": 5.454728269577027,
      "epoch": 3.896613691245016,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003560137233470826,
      "loss": 4.9517,
      "mean_token_accuracy": 0.221976375579834,
      "num_tokens": 77452579.0,
      "step": 34695
    },
    {
      "entropy": 5.390017652511597,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21212814897298812,
      "num_tokens": 77464465.0,
      "step": 34700
    },
    {
      "entropy": 5.389188241958618,
      "epoch": 3.897736845060931,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21607416570186616,
      "num_tokens": 77474944.0,
      "step": 34705
    },
    {
      "entropy": 5.391782283782959,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22234292030334474,
      "num_tokens": 77485389.0,
      "step": 34710
    },
    {
      "entropy": 5.431576299667358,
      "epoch": 3.898859998876846,
      "grad_norm": 1.0,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.9389,
      "mean_token_accuracy": 0.2198214665055275,
      "num_tokens": 77496834.0,
      "step": 34715
    },
    {
      "entropy": 5.345710420608521,
      "epoch": 3.8994215757848036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2230816438794136,
      "num_tokens": 77508123.0,
      "step": 34720
    },
    {
      "entropy": 5.382413101196289,
      "epoch": 3.8999831526927613,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003557889415896332,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2216127932071686,
      "num_tokens": 77519599.0,
      "step": 34725
    },
    {
      "entropy": 5.418860721588135,
      "epoch": 3.900544729600719,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035575146895320327,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22027526050806046,
      "num_tokens": 77532241.0,
      "step": 34730
    },
    {
      "entropy": 5.540792179107666,
      "epoch": 3.9011063065086766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003557139937456152,
      "loss": 5.0562,
      "mean_token_accuracy": 0.21725634783506392,
      "num_tokens": 77542702.0,
      "step": 34735
    },
    {
      "entropy": 5.458561182022095,
      "epoch": 3.901667883416634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035567651596806236,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21680158376693726,
      "num_tokens": 77554235.0,
      "step": 34740
    },
    {
      "entropy": 5.400365781784058,
      "epoch": 3.9022294603245915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035563903562173786,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22390355914831161,
      "num_tokens": 77565704.0,
      "step": 34745
    },
    {
      "entropy": 5.400256586074829,
      "epoch": 3.902791037232549,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22008950114250184,
      "num_tokens": 77576524.0,
      "step": 34750
    },
    {
      "entropy": 5.459225177764893,
      "epoch": 3.9033526141405064,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.963,
      "mean_token_accuracy": 0.2243005558848381,
      "num_tokens": 77587292.0,
      "step": 34755
    },
    {
      "entropy": 5.4438012599945065,
      "epoch": 3.903914191048464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003555265791820695,
      "loss": 5.0132,
      "mean_token_accuracy": 0.20928700864315034,
      "num_tokens": 77597958.0,
      "step": 34760
    },
    {
      "entropy": 5.49782977104187,
      "epoch": 3.9044757679564217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003554890885725935,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21439800411462784,
      "num_tokens": 77609264.0,
      "step": 34765
    },
    {
      "entropy": 5.540974950790405,
      "epoch": 3.905037344864379,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003554515954003137,
      "loss": 5.0685,
      "mean_token_accuracy": 0.22002264261245727,
      "num_tokens": 77621224.0,
      "step": 34770
    },
    {
      "entropy": 5.459763860702514,
      "epoch": 3.9055989217723366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003554140996664239,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21318913400173187,
      "num_tokens": 77632009.0,
      "step": 34775
    },
    {
      "entropy": 5.415331602096558,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000355376601372118,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21634882539510727,
      "num_tokens": 77643094.0,
      "step": 34780
    },
    {
      "entropy": 5.430211591720581,
      "epoch": 3.906722075588252,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003553391005185899,
      "loss": 5.023,
      "mean_token_accuracy": 0.22514042854309083,
      "num_tokens": 77654049.0,
      "step": 34785
    },
    {
      "entropy": 5.424267530441284,
      "epoch": 3.9072836524962096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.9426,
      "mean_token_accuracy": 0.22394705712795257,
      "num_tokens": 77664820.0,
      "step": 34790
    },
    {
      "entropy": 5.546099853515625,
      "epoch": 3.907845229404167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003552640911386436,
      "loss": 5.0999,
      "mean_token_accuracy": 0.2110679864883423,
      "num_tokens": 77675107.0,
      "step": 34795
    },
    {
      "entropy": 5.447711420059204,
      "epoch": 3.9084068063121244,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035522658261461366,
      "loss": 4.9542,
      "mean_token_accuracy": 0.21461385488510132,
      "num_tokens": 77685487.0,
      "step": 34800
    },
    {
      "entropy": 5.437941551208496,
      "epoch": 3.908968383220082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21757581532001496,
      "num_tokens": 77697391.0,
      "step": 34805
    },
    {
      "entropy": 5.4682526111602785,
      "epoch": 3.9095299601280393,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.9572,
      "mean_token_accuracy": 0.22340391129255294,
      "num_tokens": 77707270.0,
      "step": 34810
    },
    {
      "entropy": 5.3404991149902346,
      "epoch": 3.910091537035997,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22623810023069382,
      "num_tokens": 77717164.0,
      "step": 34815
    },
    {
      "entropy": 5.4963030338287355,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003550765229859828,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2147359013557434,
      "num_tokens": 77728142.0,
      "step": 34820
    },
    {
      "entropy": 5.4969322204589846,
      "epoch": 3.9112146908519123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003550390017016698,
      "loss": 5.0559,
      "mean_token_accuracy": 0.22220600098371507,
      "num_tokens": 77740162.0,
      "step": 34825
    },
    {
      "entropy": 5.453863716125488,
      "epoch": 3.91177626775987,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035500147786888387,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2285173863172531,
      "num_tokens": 77751574.0,
      "step": 34830
    },
    {
      "entropy": 5.496347188949585,
      "epoch": 3.912337844667827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035496395148881994,
      "loss": 5.0532,
      "mean_token_accuracy": 0.2166771709918976,
      "num_tokens": 77762919.0,
      "step": 34835
    },
    {
      "entropy": 5.417273044586182,
      "epoch": 3.912899421575785,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.9866,
      "mean_token_accuracy": 0.221032577753067,
      "num_tokens": 77774578.0,
      "step": 34840
    },
    {
      "entropy": 5.483873605728149,
      "epoch": 3.9134609984837425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035488889109163725,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21870765835046768,
      "num_tokens": 77784872.0,
      "step": 34845
    },
    {
      "entropy": 5.426029253005981,
      "epoch": 3.9140225753916997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21990254670381545,
      "num_tokens": 77796159.0,
      "step": 34850
    },
    {
      "entropy": 5.365622186660767,
      "epoch": 3.9145841522996574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22596972286701203,
      "num_tokens": 77805779.0,
      "step": 34855
    },
    {
      "entropy": 5.480260753631592,
      "epoch": 3.915145729207615,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003547762814211517,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21907289922237397,
      "num_tokens": 77818102.0,
      "step": 34860
    },
    {
      "entropy": 5.45202898979187,
      "epoch": 3.9157073061155723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000354738739782514,
      "loss": 5.036,
      "mean_token_accuracy": 0.21392963081598282,
      "num_tokens": 77830107.0,
      "step": 34865
    },
    {
      "entropy": 5.467381572723388,
      "epoch": 3.91626888302353,
      "grad_norm": 1.109375,
      "learning_rate": 0.000354701195604964,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21332841664552687,
      "num_tokens": 77841388.0,
      "step": 34870
    },
    {
      "entropy": 5.458281373977661,
      "epoch": 3.9168304599314876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003546636488896969,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21399015933275223,
      "num_tokens": 77853418.0,
      "step": 34875
    },
    {
      "entropy": 5.505796957015991,
      "epoch": 3.9173920368394453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003546260996379083,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21461955159902574,
      "num_tokens": 77865019.0,
      "step": 34880
    },
    {
      "entropy": 5.466100645065308,
      "epoch": 3.917953613747403,
      "grad_norm": 1.03125,
      "learning_rate": 0.000354588547850794,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2204435482621193,
      "num_tokens": 77876546.0,
      "step": 34885
    },
    {
      "entropy": 5.523967981338501,
      "epoch": 3.91851519065536,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003545509935295493,
      "loss": 5.0805,
      "mean_token_accuracy": 0.21517561376094818,
      "num_tokens": 77888081.0,
      "step": 34890
    },
    {
      "entropy": 5.354397201538086,
      "epoch": 3.919076767563318,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003545134366753701,
      "loss": 4.8823,
      "mean_token_accuracy": 0.225845268368721,
      "num_tokens": 77900106.0,
      "step": 34895
    },
    {
      "entropy": 5.428485774993897,
      "epoch": 3.9196383444712755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22681449651718139,
      "num_tokens": 77910452.0,
      "step": 34900
    },
    {
      "entropy": 5.466759157180786,
      "epoch": 3.9201999213792327,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035443831537299177,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21461546868085862,
      "num_tokens": 77921826.0,
      "step": 34905
    },
    {
      "entropy": 5.572246026992798,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003544007509271844,
      "loss": 5.1089,
      "mean_token_accuracy": 0.2108939617872238,
      "num_tokens": 77933909.0,
      "step": 34910
    },
    {
      "entropy": 5.554079914093018,
      "epoch": 3.921323075195148,
      "grad_norm": 1.0,
      "learning_rate": 0.00035436318395322636,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21244825869798661,
      "num_tokens": 77945232.0,
      "step": 34915
    },
    {
      "entropy": 5.441824674606323,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003543256144523137,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21628457307815552,
      "num_tokens": 77957305.0,
      "step": 34920
    },
    {
      "entropy": 5.378567790985107,
      "epoch": 3.9224462290110633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2281341165304184,
      "num_tokens": 77968036.0,
      "step": 34925
    },
    {
      "entropy": 5.404850578308105,
      "epoch": 3.9230078059190205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003542504678744099,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22325969040393828,
      "num_tokens": 77978826.0,
      "step": 34930
    },
    {
      "entropy": 5.404387807846069,
      "epoch": 3.923569382826978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21726534515619278,
      "num_tokens": 77989820.0,
      "step": 34935
    },
    {
      "entropy": 5.444500541687011,
      "epoch": 3.924130959734936,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2129973739385605,
      "num_tokens": 78000114.0,
      "step": 34940
    },
    {
      "entropy": 5.397372770309448,
      "epoch": 3.924692536642893,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22431577742099762,
      "num_tokens": 78010680.0,
      "step": 34945
    },
    {
      "entropy": 5.4540637016296385,
      "epoch": 3.9252541135508507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.9591,
      "mean_token_accuracy": 0.2154833823442459,
      "num_tokens": 78022172.0,
      "step": 34950
    },
    {
      "entropy": 5.450309896469117,
      "epoch": 3.9258156904588084,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21303560584783554,
      "num_tokens": 78033535.0,
      "step": 34955
    },
    {
      "entropy": 5.439764738082886,
      "epoch": 3.9263772673667656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003540249676182131,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21759533286094665,
      "num_tokens": 78044555.0,
      "step": 34960
    },
    {
      "entropy": 5.414005899429322,
      "epoch": 3.9269388442747233,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035398737542854936,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22370222359895706,
      "num_tokens": 78056078.0,
      "step": 34965
    },
    {
      "entropy": 5.457328510284424,
      "epoch": 3.927500421182681,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.997,
      "mean_token_accuracy": 0.21994853019714355,
      "num_tokens": 78068324.0,
      "step": 34970
    },
    {
      "entropy": 5.406261539459228,
      "epoch": 3.9280619980906386,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035391218350545287,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21537727862596512,
      "num_tokens": 78080013.0,
      "step": 34975
    },
    {
      "entropy": 5.447155857086182,
      "epoch": 3.9286235749985963,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035387458377441434,
      "loss": 5.0037,
      "mean_token_accuracy": 0.2161048471927643,
      "num_tokens": 78091494.0,
      "step": 34980
    },
    {
      "entropy": 5.474175596237183,
      "epoch": 3.9291851519065535,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22314296364784242,
      "num_tokens": 78102037.0,
      "step": 34985
    },
    {
      "entropy": 5.43297929763794,
      "epoch": 3.929746728814511,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035379937677934245,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21497827172279357,
      "num_tokens": 78112651.0,
      "step": 34990
    },
    {
      "entropy": 5.456510591506958,
      "epoch": 3.930308305722469,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003537617695177037,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2180771842598915,
      "num_tokens": 78123273.0,
      "step": 34995
    },
    {
      "entropy": 5.491047525405884,
      "epoch": 3.930869882630426,
      "grad_norm": 1.125,
      "learning_rate": 0.00035372415974825974,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2123734325170517,
      "num_tokens": 78133868.0,
      "step": 35000
    },
    {
      "entropy": 5.528883600234986,
      "epoch": 3.9314314595383837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035368654747220795,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21136365979909896,
      "num_tokens": 78145001.0,
      "step": 35005
    },
    {
      "entropy": 5.501938056945801,
      "epoch": 3.9319930364463413,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035364893269074606,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21308618485927583,
      "num_tokens": 78155699.0,
      "step": 35010
    },
    {
      "entropy": 5.388974714279175,
      "epoch": 3.932554613354299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035361131540507157,
      "loss": 4.8294,
      "mean_token_accuracy": 0.224798721075058,
      "num_tokens": 78165941.0,
      "step": 35015
    },
    {
      "entropy": 5.403348970413208,
      "epoch": 3.9331161902622567,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003535736956163823,
      "loss": 5.007,
      "mean_token_accuracy": 0.22347320914268493,
      "num_tokens": 78177049.0,
      "step": 35020
    },
    {
      "entropy": 5.471497058868408,
      "epoch": 3.933677767170214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035353607332587613,
      "loss": 5.09,
      "mean_token_accuracy": 0.21281858831644057,
      "num_tokens": 78188834.0,
      "step": 35025
    },
    {
      "entropy": 5.543138599395752,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003534984485347509,
      "loss": 5.0189,
      "mean_token_accuracy": 0.22663862705230714,
      "num_tokens": 78199847.0,
      "step": 35030
    },
    {
      "entropy": 5.369149875640869,
      "epoch": 3.934800920986129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003534608212442046,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22344624102115632,
      "num_tokens": 78210747.0,
      "step": 35035
    },
    {
      "entropy": 5.407032060623169,
      "epoch": 3.9353624978940864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035342319145543533,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2232777014374733,
      "num_tokens": 78221711.0,
      "step": 35040
    },
    {
      "entropy": 5.570589685440064,
      "epoch": 3.935924074802044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035338555916964126,
      "loss": 5.1562,
      "mean_token_accuracy": 0.2054454505443573,
      "num_tokens": 78233957.0,
      "step": 35045
    },
    {
      "entropy": 5.430166435241699,
      "epoch": 3.9364856517100018,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035334792438802056,
      "loss": 5.0025,
      "mean_token_accuracy": 0.21588361710309983,
      "num_tokens": 78245403.0,
      "step": 35050
    },
    {
      "entropy": 5.470647811889648,
      "epoch": 3.937047228617959,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21761251837015153,
      "num_tokens": 78256062.0,
      "step": 35055
    },
    {
      "entropy": 5.4520172595977785,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003532726473420927,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22097249776124955,
      "num_tokens": 78265895.0,
      "step": 35060
    },
    {
      "entropy": 5.440299558639526,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21425224840641022,
      "num_tokens": 78276752.0,
      "step": 35065
    },
    {
      "entropy": 5.4226884841918945,
      "epoch": 3.938731959341832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035319736032723905,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22373635172843934,
      "num_tokens": 78288357.0,
      "step": 35070
    },
    {
      "entropy": 5.444322347640991,
      "epoch": 3.9392935362497896,
      "grad_norm": 1.0,
      "learning_rate": 0.0003531597130844616,
      "loss": 4.946,
      "mean_token_accuracy": 0.22660247087478638,
      "num_tokens": 78299558.0,
      "step": 35075
    },
    {
      "entropy": 5.542957830429077,
      "epoch": 3.939855113157747,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035312206335304843,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21739863753318786,
      "num_tokens": 78311443.0,
      "step": 35080
    },
    {
      "entropy": 5.519734239578247,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035308441113419855,
      "loss": 5.0783,
      "mean_token_accuracy": 0.218059179186821,
      "num_tokens": 78322575.0,
      "step": 35085
    },
    {
      "entropy": 5.463766193389892,
      "epoch": 3.940978266973662,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035304675642911074,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21701926589012147,
      "num_tokens": 78333583.0,
      "step": 35090
    },
    {
      "entropy": 5.447640705108642,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003530090992389838,
      "loss": 4.9721,
      "mean_token_accuracy": 0.21478767693042755,
      "num_tokens": 78343339.0,
      "step": 35095
    },
    {
      "entropy": 5.462275075912475,
      "epoch": 3.942101420789577,
      "grad_norm": 1.078125,
      "learning_rate": 0.000352971439565017,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2211902469396591,
      "num_tokens": 78354613.0,
      "step": 35100
    },
    {
      "entropy": 5.417634439468384,
      "epoch": 3.9426629976975347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2178771272301674,
      "num_tokens": 78365105.0,
      "step": 35105
    },
    {
      "entropy": 5.391190433502198,
      "epoch": 3.9432245746054924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035289611277036,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22754145562648773,
      "num_tokens": 78375311.0,
      "step": 35110
    },
    {
      "entropy": 5.3941864490509035,
      "epoch": 3.94378615151345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22397405952215194,
      "num_tokens": 78386653.0,
      "step": 35115
    },
    {
      "entropy": 5.289754056930542,
      "epoch": 3.9443477284214072,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035282077605473317,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22362440824508667,
      "num_tokens": 78397296.0,
      "step": 35120
    },
    {
      "entropy": 5.427908611297608,
      "epoch": 3.944909305329365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003527831039795545,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22305500209331514,
      "num_tokens": 78408825.0,
      "step": 35125
    },
    {
      "entropy": 5.41396656036377,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.93,
      "mean_token_accuracy": 0.22582117319107056,
      "num_tokens": 78419895.0,
      "step": 35130
    },
    {
      "entropy": 5.443940591812134,
      "epoch": 3.94603245914528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.845,
      "mean_token_accuracy": 0.22588656097650528,
      "num_tokens": 78430111.0,
      "step": 35135
    },
    {
      "entropy": 5.398477268218994,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035267007289895176,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21530262529850006,
      "num_tokens": 78441778.0,
      "step": 35140
    },
    {
      "entropy": 5.385983037948608,
      "epoch": 3.947155612961195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22483974546194077,
      "num_tokens": 78451930.0,
      "step": 35145
    },
    {
      "entropy": 5.374666166305542,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.9225,
      "mean_token_accuracy": 0.2257457286119461,
      "num_tokens": 78463002.0,
      "step": 35150
    },
    {
      "entropy": 5.384811305999756,
      "epoch": 3.94827876677711,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21961035430431367,
      "num_tokens": 78473573.0,
      "step": 35155
    },
    {
      "entropy": 5.453182888031006,
      "epoch": 3.9488403436850676,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035251933017444814,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21174660623073577,
      "num_tokens": 78485336.0,
      "step": 35160
    },
    {
      "entropy": 5.476636075973511,
      "epoch": 3.9494019205930253,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21983083784580232,
      "num_tokens": 78497801.0,
      "step": 35165
    },
    {
      "entropy": 5.447101211547851,
      "epoch": 3.949963497500983,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21985769718885423,
      "num_tokens": 78509288.0,
      "step": 35170
    },
    {
      "entropy": 5.434791374206543,
      "epoch": 3.95052507440894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003524062472102927,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22261300384998323,
      "num_tokens": 78519871.0,
      "step": 35175
    },
    {
      "entropy": 5.383049011230469,
      "epoch": 3.951086651316898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003523685479580174,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21813647747039794,
      "num_tokens": 78530287.0,
      "step": 35180
    },
    {
      "entropy": 5.453476524353027,
      "epoch": 3.9516482282248555,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21386591494083404,
      "num_tokens": 78542341.0,
      "step": 35185
    },
    {
      "entropy": 5.3831947326660154,
      "epoch": 3.9522098051328127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22958169877529144,
      "num_tokens": 78552533.0,
      "step": 35190
    },
    {
      "entropy": 5.447121334075928,
      "epoch": 3.9527713820407704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21881452351808547,
      "num_tokens": 78564007.0,
      "step": 35195
    },
    {
      "entropy": 5.403392362594604,
      "epoch": 3.953332958948728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035221772632647503,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21391237080097197,
      "num_tokens": 78574015.0,
      "step": 35200
    },
    {
      "entropy": 5.521846628189087,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035218001476898194,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21474697887897493,
      "num_tokens": 78586230.0,
      "step": 35205
    },
    {
      "entropy": 5.4549071311950685,
      "epoch": 3.9544561127646434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.9899,
      "mean_token_accuracy": 0.22162221670150756,
      "num_tokens": 78596819.0,
      "step": 35210
    },
    {
      "entropy": 5.451625108718872,
      "epoch": 3.9550176896726006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003521045842828718,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21803141683340072,
      "num_tokens": 78608788.0,
      "step": 35215
    },
    {
      "entropy": 5.438762569427491,
      "epoch": 3.9555792665805583,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2143615186214447,
      "num_tokens": 78619124.0,
      "step": 35220
    },
    {
      "entropy": 5.5130208969116214,
      "epoch": 3.956140843488516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003520291439766024,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21848011910915374,
      "num_tokens": 78630799.0,
      "step": 35225
    },
    {
      "entropy": 5.32649450302124,
      "epoch": 3.956702420396473,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22492473125457763,
      "num_tokens": 78641664.0,
      "step": 35230
    },
    {
      "entropy": 5.449229717254639,
      "epoch": 3.957263997304431,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035195369385978165,
      "loss": 4.96,
      "mean_token_accuracy": 0.22289867252111434,
      "num_tokens": 78652025.0,
      "step": 35235
    },
    {
      "entropy": 5.471516561508179,
      "epoch": 3.9578255742123885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035191596512541747,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22599594593048095,
      "num_tokens": 78663539.0,
      "step": 35240
    },
    {
      "entropy": 5.400595569610596,
      "epoch": 3.9583871511203457,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003518782339420192,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21465740203857422,
      "num_tokens": 78674398.0,
      "step": 35245
    },
    {
      "entropy": 5.434558343887329,
      "epoch": 3.9589487280283033,
      "grad_norm": 1.0,
      "learning_rate": 0.00035184050031078815,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21438826620578766,
      "num_tokens": 78685945.0,
      "step": 35250
    },
    {
      "entropy": 5.39529070854187,
      "epoch": 3.959510304936261,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21447912603616714,
      "num_tokens": 78696664.0,
      "step": 35255
    },
    {
      "entropy": 5.395404577255249,
      "epoch": 3.9600718818442187,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23242322057485582,
      "num_tokens": 78706812.0,
      "step": 35260
    },
    {
      "entropy": 5.36605134010315,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22586668878793717,
      "num_tokens": 78716657.0,
      "step": 35265
    },
    {
      "entropy": 5.434974431991577,
      "epoch": 3.9611950356601335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.998,
      "mean_token_accuracy": 0.2181496202945709,
      "num_tokens": 78728313.0,
      "step": 35270
    },
    {
      "entropy": 5.464006853103638,
      "epoch": 3.961756612568091,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003516517954791944,
      "loss": 4.996,
      "mean_token_accuracy": 0.21821155697107314,
      "num_tokens": 78739888.0,
      "step": 35275
    },
    {
      "entropy": 5.458188390731811,
      "epoch": 3.962318189476049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22009536474943162,
      "num_tokens": 78750624.0,
      "step": 35280
    },
    {
      "entropy": 5.437712097167969,
      "epoch": 3.962879766384006,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22613925635814666,
      "num_tokens": 78762506.0,
      "step": 35285
    },
    {
      "entropy": 5.455118417739868,
      "epoch": 3.9634413432919637,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003515385432831499,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21512505263090134,
      "num_tokens": 78772565.0,
      "step": 35290
    },
    {
      "entropy": 5.405574321746826,
      "epoch": 3.9640029201999214,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21888073682785034,
      "num_tokens": 78784107.0,
      "step": 35295
    },
    {
      "entropy": 5.469899129867554,
      "epoch": 3.964564497107879,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003514630296320323,
      "loss": 5.0403,
      "mean_token_accuracy": 0.2134338602423668,
      "num_tokens": 78796182.0,
      "step": 35300
    },
    {
      "entropy": 5.480256223678589,
      "epoch": 3.9651260740158367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003514252691539536,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2199527695775032,
      "num_tokens": 78806620.0,
      "step": 35305
    },
    {
      "entropy": 5.420833492279053,
      "epoch": 3.965687650923794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23045612126588821,
      "num_tokens": 78817034.0,
      "step": 35310
    },
    {
      "entropy": 5.44416618347168,
      "epoch": 3.9662492278317516,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000351349740898768,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22053406089544297,
      "num_tokens": 78828974.0,
      "step": 35315
    },
    {
      "entropy": 5.472843217849731,
      "epoch": 3.9668108047397093,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003513119731240659,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21181270480155945,
      "num_tokens": 78840514.0,
      "step": 35320
    },
    {
      "entropy": 5.459736633300781,
      "epoch": 3.9673723816476665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035127420291956104,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22286323457956314,
      "num_tokens": 78850839.0,
      "step": 35325
    },
    {
      "entropy": 5.44039216041565,
      "epoch": 3.967933958555624,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003512364302864559,
      "loss": 5.0092,
      "mean_token_accuracy": 0.22171708643436433,
      "num_tokens": 78862931.0,
      "step": 35330
    },
    {
      "entropy": 5.449551916122436,
      "epoch": 3.968495535463582,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.943,
      "mean_token_accuracy": 0.21648455560207366,
      "num_tokens": 78872324.0,
      "step": 35335
    },
    {
      "entropy": 5.4072509765625,
      "epoch": 3.969057112371539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.9299,
      "mean_token_accuracy": 0.21803183108568192,
      "num_tokens": 78884075.0,
      "step": 35340
    },
    {
      "entropy": 5.3324939727783205,
      "epoch": 3.9696186892794967,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22278828769922257,
      "num_tokens": 78894804.0,
      "step": 35345
    },
    {
      "entropy": 5.495571327209473,
      "epoch": 3.9701802661874543,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035108531549208846,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21304801255464553,
      "num_tokens": 78906069.0,
      "step": 35350
    },
    {
      "entropy": 5.597161674499512,
      "epoch": 3.970741843095412,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035104753073402433,
      "loss": 5.1357,
      "mean_token_accuracy": 0.21182256639003755,
      "num_tokens": 78918490.0,
      "step": 35355
    },
    {
      "entropy": 5.509946346282959,
      "epoch": 3.9713034200033697,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003510097435545774,
      "loss": 5.0044,
      "mean_token_accuracy": 0.21467436105012894,
      "num_tokens": 78929790.0,
      "step": 35360
    },
    {
      "entropy": 5.446705293655396,
      "epoch": 3.971864996911327,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22864970117807387,
      "num_tokens": 78941600.0,
      "step": 35365
    },
    {
      "entropy": 5.5030371189117435,
      "epoch": 3.9724265738192845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003509341619363474,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21319745779037474,
      "num_tokens": 78953764.0,
      "step": 35370
    },
    {
      "entropy": 5.463508224487304,
      "epoch": 3.972988150727242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035089636749997095,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21991973966360093,
      "num_tokens": 78964781.0,
      "step": 35375
    },
    {
      "entropy": 5.543183612823486,
      "epoch": 3.9735497276351994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035085857064702487,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21385376155376434,
      "num_tokens": 78976325.0,
      "step": 35380
    },
    {
      "entropy": 5.435974645614624,
      "epoch": 3.974111304543157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003508207713787123,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2229997217655182,
      "num_tokens": 78987174.0,
      "step": 35385
    },
    {
      "entropy": 5.483600282669068,
      "epoch": 3.9746728814511147,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.9806,
      "mean_token_accuracy": 0.21907915472984313,
      "num_tokens": 78999195.0,
      "step": 35390
    },
    {
      "entropy": 5.330482864379883,
      "epoch": 3.9752344583590724,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22684643566608428,
      "num_tokens": 79010981.0,
      "step": 35395
    },
    {
      "entropy": 5.405303335189819,
      "epoch": 3.97579603526703,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22464494854211808,
      "num_tokens": 79022059.0,
      "step": 35400
    },
    {
      "entropy": 5.617932367324829,
      "epoch": 3.9763576121749873,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035066955017587135,
      "loss": 5.1693,
      "mean_token_accuracy": 0.2110062301158905,
      "num_tokens": 79033261.0,
      "step": 35405
    },
    {
      "entropy": 5.436311101913452,
      "epoch": 3.976919189082945,
      "grad_norm": 1.09375,
      "learning_rate": 0.000350631738848782,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2146225690841675,
      "num_tokens": 79044518.0,
      "step": 35410
    },
    {
      "entropy": 5.43720235824585,
      "epoch": 3.9774807659909026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003505939251135488,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21652488708496093,
      "num_tokens": 79055202.0,
      "step": 35415
    },
    {
      "entropy": 5.48226842880249,
      "epoch": 3.97804234289886,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003505561089713756,
      "loss": 5.025,
      "mean_token_accuracy": 0.2145179733633995,
      "num_tokens": 79066448.0,
      "step": 35420
    },
    {
      "entropy": 5.468845367431641,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003505182904234667,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21629902869462966,
      "num_tokens": 79077745.0,
      "step": 35425
    },
    {
      "entropy": 5.528349733352661,
      "epoch": 3.979165496714775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003504804694710261,
      "loss": 5.059,
      "mean_token_accuracy": 0.21022931337356568,
      "num_tokens": 79088011.0,
      "step": 35430
    },
    {
      "entropy": 5.5749105453491214,
      "epoch": 3.9797270736227324,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003504426461152581,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20975074917078018,
      "num_tokens": 79098718.0,
      "step": 35435
    },
    {
      "entropy": 5.41304235458374,
      "epoch": 3.98028865053069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21851980835199356,
      "num_tokens": 79108216.0,
      "step": 35440
    },
    {
      "entropy": 5.413367366790771,
      "epoch": 3.9808502274386477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21556641310453414,
      "num_tokens": 79119709.0,
      "step": 35445
    },
    {
      "entropy": 5.5161521434783936,
      "epoch": 3.9814118043466054,
      "grad_norm": 1.015625,
      "learning_rate": 0.000350329161640033,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21756214499473572,
      "num_tokens": 79131824.0,
      "step": 35450
    },
    {
      "entropy": 5.507373952865601,
      "epoch": 3.981973381254563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035029132868299914,
      "loss": 5.0013,
      "mean_token_accuracy": 0.2083567276597023,
      "num_tokens": 79142855.0,
      "step": 35455
    },
    {
      "entropy": 5.557955121994018,
      "epoch": 3.9825349581625202,
      "grad_norm": 1.125,
      "learning_rate": 0.00035025349332866017,
      "loss": 5.112,
      "mean_token_accuracy": 0.20994910448789597,
      "num_tokens": 79153759.0,
      "step": 35460
    },
    {
      "entropy": 5.473484849929809,
      "epoch": 3.983096535070478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035021565557822085,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21612292826175689,
      "num_tokens": 79163586.0,
      "step": 35465
    },
    {
      "entropy": 5.504828071594238,
      "epoch": 3.9836581119784356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003501778154328858,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21020968556404113,
      "num_tokens": 79175067.0,
      "step": 35470
    },
    {
      "entropy": 5.349946212768555,
      "epoch": 3.9842196888863928,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22980794608592986,
      "num_tokens": 79186039.0,
      "step": 35475
    },
    {
      "entropy": 5.481285095214844,
      "epoch": 3.9847812657943504,
      "grad_norm": 1.0,
      "learning_rate": 0.00035010212796234827,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21181695014238358,
      "num_tokens": 79198571.0,
      "step": 35480
    },
    {
      "entropy": 5.3585028648376465,
      "epoch": 3.985342842702308,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2240922674536705,
      "num_tokens": 79210253.0,
      "step": 35485
    },
    {
      "entropy": 5.403286361694336,
      "epoch": 3.9859044196102658,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.936,
      "mean_token_accuracy": 0.22786005437374116,
      "num_tokens": 79221611.0,
      "step": 35490
    },
    {
      "entropy": 5.471501445770263,
      "epoch": 3.9864659965182234,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034998857882494823,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21436578333377837,
      "num_tokens": 79233270.0,
      "step": 35495
    },
    {
      "entropy": 5.464900207519531,
      "epoch": 3.9870275734261806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034995072433554377,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21007405668497087,
      "num_tokens": 79245338.0,
      "step": 35500
    },
    {
      "entropy": 5.518237924575805,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003499128674596792,
      "loss": 5.059,
      "mean_token_accuracy": 0.21506733149290086,
      "num_tokens": 79255558.0,
      "step": 35505
    },
    {
      "entropy": 5.531598424911499,
      "epoch": 3.988150727242096,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034987500819855976,
      "loss": 5.0595,
      "mean_token_accuracy": 0.22153622806072235,
      "num_tokens": 79266075.0,
      "step": 35510
    },
    {
      "entropy": 5.521352958679199,
      "epoch": 3.988712304150053,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034983714655339115,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21957447826862336,
      "num_tokens": 79277667.0,
      "step": 35515
    },
    {
      "entropy": 5.398689794540405,
      "epoch": 3.989273881058011,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003497992825253787,
      "loss": 4.931,
      "mean_token_accuracy": 0.22350920587778092,
      "num_tokens": 79288755.0,
      "step": 35520
    },
    {
      "entropy": 5.4374244689941404,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.125,
      "learning_rate": 0.00034976141611572817,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21831277161836624,
      "num_tokens": 79299048.0,
      "step": 35525
    },
    {
      "entropy": 5.455310440063476,
      "epoch": 3.9903970348739257,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21868229657411575,
      "num_tokens": 79310847.0,
      "step": 35530
    },
    {
      "entropy": 5.487273168563843,
      "epoch": 3.9909586117818834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034968567615633516,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21360173672437668,
      "num_tokens": 79322496.0,
      "step": 35535
    },
    {
      "entropy": 5.471784782409668,
      "epoch": 3.991520188689841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034964780260900445,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21848510950803757,
      "num_tokens": 79334216.0,
      "step": 35540
    },
    {
      "entropy": 5.410571813583374,
      "epoch": 3.9920817655977987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22785455286502837,
      "num_tokens": 79344320.0,
      "step": 35545
    },
    {
      "entropy": 5.504690408706665,
      "epoch": 3.9926433425057564,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034957204838510406,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20801470726728438,
      "num_tokens": 79355775.0,
      "step": 35550
    },
    {
      "entropy": 5.381856679916382,
      "epoch": 3.9932049194137136,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22438577562570572,
      "num_tokens": 79366866.0,
      "step": 35555
    },
    {
      "entropy": 5.445256280899048,
      "epoch": 3.9937664963216712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000349496284663592,
      "loss": 5.0029,
      "mean_token_accuracy": 0.2116694226861,
      "num_tokens": 79378599.0,
      "step": 35560
    },
    {
      "entropy": 5.383032131195068,
      "epoch": 3.994328073229629,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.902,
      "mean_token_accuracy": 0.22327440977096558,
      "num_tokens": 79389283.0,
      "step": 35565
    },
    {
      "entropy": 5.460935497283936,
      "epoch": 3.994889650137586,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003494205114541176,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2122328609228134,
      "num_tokens": 79400858.0,
      "step": 35570
    },
    {
      "entropy": 5.510056066513061,
      "epoch": 3.995451227045544,
      "grad_norm": 1.125,
      "learning_rate": 0.00034938262129441026,
      "loss": 5.0421,
      "mean_token_accuracy": 0.216183640062809,
      "num_tokens": 79411122.0,
      "step": 35575
    },
    {
      "entropy": 5.40553388595581,
      "epoch": 3.9960128039535014,
      "grad_norm": 1.0,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23313819020986556,
      "num_tokens": 79420153.0,
      "step": 35580
    },
    {
      "entropy": 5.546561193466187,
      "epoch": 3.996574380861459,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003493068338710877,
      "loss": 5.1181,
      "mean_token_accuracy": 0.21331992894411086,
      "num_tokens": 79432964.0,
      "step": 35585
    },
    {
      "entropy": 5.412816524505615,
      "epoch": 3.9971359577694168,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003492689366098855,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22359354943037033,
      "num_tokens": 79445191.0,
      "step": 35590
    },
    {
      "entropy": 5.471116256713867,
      "epoch": 3.997697534677374,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22190708220005034,
      "num_tokens": 79455553.0,
      "step": 35595
    },
    {
      "entropy": 5.498412179946899,
      "epoch": 3.9982591115853316,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003491931349944325,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21384828239679338,
      "num_tokens": 79466781.0,
      "step": 35600
    },
    {
      "entropy": 5.395805978775025,
      "epoch": 3.9988206884932893,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034915523064259533,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21700815707445145,
      "num_tokens": 79478656.0,
      "step": 35605
    },
    {
      "entropy": 5.399086761474609,
      "epoch": 3.9993822654012465,
      "grad_norm": 1.0,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.953,
      "mean_token_accuracy": 0.22050281316041948,
      "num_tokens": 79489763.0,
      "step": 35610
    },
    {
      "entropy": 5.473388433456421,
      "epoch": 3.999943842309204,
      "grad_norm": 1.0,
      "learning_rate": 0.000349079414856734,
      "loss": 5.0806,
      "mean_token_accuracy": 0.20976900309324265,
      "num_tokens": 79501828.0,
      "step": 35615
    },
    {
      "entropy": 5.471547815534803,
      "epoch": 4.000449261526366,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.931,
      "mean_token_accuracy": 0.21515231993463305,
      "num_tokens": 79510316.0,
      "step": 35620
    },
    {
      "entropy": 5.395777845382691,
      "epoch": 4.001010838434324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2215465098619461,
      "num_tokens": 79521555.0,
      "step": 35625
    },
    {
      "entropy": 5.544995737075806,
      "epoch": 4.001572415342281,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034896567349058027,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21871211379766464,
      "num_tokens": 79533147.0,
      "step": 35630
    },
    {
      "entropy": 5.5175745487213135,
      "epoch": 4.002133992250239,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034892775499006126,
      "loss": 5.0339,
      "mean_token_accuracy": 0.21613783687353133,
      "num_tokens": 79544695.0,
      "step": 35635
    },
    {
      "entropy": 5.359199047088623,
      "epoch": 4.002695569158196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003488898341356538,
      "loss": 4.8018,
      "mean_token_accuracy": 0.23133434802293779,
      "num_tokens": 79555740.0,
      "step": 35640
    },
    {
      "entropy": 5.427336072921753,
      "epoch": 4.0032571460661535,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22098218500614167,
      "num_tokens": 79565648.0,
      "step": 35645
    },
    {
      "entropy": 5.494509601593018,
      "epoch": 4.003818722974112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22143245339393616,
      "num_tokens": 79577107.0,
      "step": 35650
    },
    {
      "entropy": 5.443895483016968,
      "epoch": 4.004380299882069,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22181719690561294,
      "num_tokens": 79586689.0,
      "step": 35655
    },
    {
      "entropy": 5.477811574935913,
      "epoch": 4.004941876790026,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003487381272032897,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21919209212064744,
      "num_tokens": 79597517.0,
      "step": 35660
    },
    {
      "entropy": 5.470027685165405,
      "epoch": 4.005503453697984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000348700194597553,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2183989852666855,
      "num_tokens": 79609516.0,
      "step": 35665
    },
    {
      "entropy": 5.439858770370483,
      "epoch": 4.006065030605941,
      "grad_norm": 1.125,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22276511192321777,
      "num_tokens": 79620706.0,
      "step": 35670
    },
    {
      "entropy": 5.420510721206665,
      "epoch": 4.006626607513899,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003486243223473592,
      "loss": 4.8293,
      "mean_token_accuracy": 0.22794631421566008,
      "num_tokens": 79630897.0,
      "step": 35675
    },
    {
      "entropy": 5.373473358154297,
      "epoch": 4.007188184421857,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22421146631240846,
      "num_tokens": 79642426.0,
      "step": 35680
    },
    {
      "entropy": 5.421261739730835,
      "epoch": 4.007749761329814,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22518074065446853,
      "num_tokens": 79653491.0,
      "step": 35685
    },
    {
      "entropy": 5.381777429580689,
      "epoch": 4.008311338237772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.863,
      "mean_token_accuracy": 0.22482653856277465,
      "num_tokens": 79665329.0,
      "step": 35690
    },
    {
      "entropy": 5.419335269927979,
      "epoch": 4.008872915145729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22075097262859344,
      "num_tokens": 79676552.0,
      "step": 35695
    },
    {
      "entropy": 5.378837919235229,
      "epoch": 4.009434492053686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034843460071904407,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22171896696090698,
      "num_tokens": 79687359.0,
      "step": 35700
    },
    {
      "entropy": 5.395798015594482,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.832,
      "mean_token_accuracy": 0.22733590751886368,
      "num_tokens": 79698728.0,
      "step": 35705
    },
    {
      "entropy": 5.4735329151153564,
      "epoch": 4.010557645869602,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003483586956919572,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2156719222664833,
      "num_tokens": 79709018.0,
      "step": 35710
    },
    {
      "entropy": 5.482189464569092,
      "epoch": 4.011119222777559,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21609508991241455,
      "num_tokens": 79720166.0,
      "step": 35715
    },
    {
      "entropy": 5.413097238540649,
      "epoch": 4.011680799685517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003482827813217932,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23402056097984314,
      "num_tokens": 79730467.0,
      "step": 35720
    },
    {
      "entropy": 5.462182521820068,
      "epoch": 4.012242376593474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22300498336553573,
      "num_tokens": 79740570.0,
      "step": 35725
    },
    {
      "entropy": 5.4214338779449465,
      "epoch": 4.012803953501432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21967718601226807,
      "num_tokens": 79751856.0,
      "step": 35730
    },
    {
      "entropy": 5.429662990570068,
      "epoch": 4.01336553040939,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22983449697494507,
      "num_tokens": 79763267.0,
      "step": 35735
    },
    {
      "entropy": 5.501342821121216,
      "epoch": 4.013927107317347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003481309245909088,
      "loss": 5.0424,
      "mean_token_accuracy": 0.22176155149936677,
      "num_tokens": 79775202.0,
      "step": 35740
    },
    {
      "entropy": 5.426898193359375,
      "epoch": 4.014488684225305,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21946090012788771,
      "num_tokens": 79787087.0,
      "step": 35745
    },
    {
      "entropy": 5.499755144119263,
      "epoch": 4.015050261133262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003480549822495293,
      "loss": 5.0214,
      "mean_token_accuracy": 0.22048043310642243,
      "num_tokens": 79799014.0,
      "step": 35750
    },
    {
      "entropy": 5.415890312194824,
      "epoch": 4.015611838041219,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23274248242378234,
      "num_tokens": 79810455.0,
      "step": 35755
    },
    {
      "entropy": 5.48612699508667,
      "epoch": 4.0161734149491775,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22217288464307786,
      "num_tokens": 79821614.0,
      "step": 35760
    },
    {
      "entropy": 5.476651287078857,
      "epoch": 4.016734991857135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003479410512947407,
      "loss": 5.0504,
      "mean_token_accuracy": 0.21738536655902863,
      "num_tokens": 79832418.0,
      "step": 35765
    },
    {
      "entropy": 5.524921417236328,
      "epoch": 4.017296568765093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034790306966325606,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21669320315122603,
      "num_tokens": 79844234.0,
      "step": 35770
    },
    {
      "entropy": 5.42331223487854,
      "epoch": 4.01785814567305,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22446131557226182,
      "num_tokens": 79855284.0,
      "step": 35775
    },
    {
      "entropy": 5.402645683288574,
      "epoch": 4.018419722581007,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.8555,
      "mean_token_accuracy": 0.22327370941638947,
      "num_tokens": 79866169.0,
      "step": 35780
    },
    {
      "entropy": 5.488822078704834,
      "epoch": 4.018981299488965,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034778911084606793,
      "loss": 5.0709,
      "mean_token_accuracy": 0.20761737525463103,
      "num_tokens": 79878044.0,
      "step": 35785
    },
    {
      "entropy": 5.368129396438599,
      "epoch": 4.0195428763969225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23022831976413727,
      "num_tokens": 79888859.0,
      "step": 35790
    },
    {
      "entropy": 5.495619869232177,
      "epoch": 4.02010445330488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21911725401878357,
      "num_tokens": 79899886.0,
      "step": 35795
    },
    {
      "entropy": 5.4980857372283936,
      "epoch": 4.020666030212838,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.981,
      "mean_token_accuracy": 0.22767989933490754,
      "num_tokens": 79911752.0,
      "step": 35800
    },
    {
      "entropy": 5.41571273803711,
      "epoch": 4.021227607120795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2167004570364952,
      "num_tokens": 79923076.0,
      "step": 35805
    },
    {
      "entropy": 5.430156278610229,
      "epoch": 4.021789184028753,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2224578857421875,
      "num_tokens": 79933866.0,
      "step": 35810
    },
    {
      "entropy": 5.466361331939697,
      "epoch": 4.02235076093671,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22610590904951094,
      "num_tokens": 79946046.0,
      "step": 35815
    },
    {
      "entropy": 5.416965436935425,
      "epoch": 4.022912337844668,
      "grad_norm": 1.0,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22775917053222655,
      "num_tokens": 79957887.0,
      "step": 35820
    },
    {
      "entropy": 5.365596294403076,
      "epoch": 4.023473914752626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.8268,
      "mean_token_accuracy": 0.22717266529798508,
      "num_tokens": 79968101.0,
      "step": 35825
    },
    {
      "entropy": 5.403276252746582,
      "epoch": 4.024035491660583,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22275644540786743,
      "num_tokens": 79978176.0,
      "step": 35830
    },
    {
      "entropy": 5.443884992599488,
      "epoch": 4.02459706856854,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2175081878900528,
      "num_tokens": 79990324.0,
      "step": 35835
    },
    {
      "entropy": 5.498500537872315,
      "epoch": 4.025158645476498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22409330159425736,
      "num_tokens": 80001898.0,
      "step": 35840
    },
    {
      "entropy": 5.375147867202759,
      "epoch": 4.0257202223844555,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003473330673169572,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21721529066562653,
      "num_tokens": 80012072.0,
      "step": 35845
    },
    {
      "entropy": 5.362394618988037,
      "epoch": 4.026281799292413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22004484981298447,
      "num_tokens": 80022718.0,
      "step": 35850
    },
    {
      "entropy": 5.5100830554962155,
      "epoch": 4.026843376200371,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21969416588544846,
      "num_tokens": 80033899.0,
      "step": 35855
    },
    {
      "entropy": 5.554616546630859,
      "epoch": 4.027404953108328,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22219123393297197,
      "num_tokens": 80045398.0,
      "step": 35860
    },
    {
      "entropy": 5.519767618179321,
      "epoch": 4.027966530016286,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003471809790043667,
      "loss": 5.0048,
      "mean_token_accuracy": 0.22453017085790633,
      "num_tokens": 80055186.0,
      "step": 35865
    },
    {
      "entropy": 5.456296586990357,
      "epoch": 4.028528106924243,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.9673,
      "mean_token_accuracy": 0.22856031358242035,
      "num_tokens": 80065545.0,
      "step": 35870
    },
    {
      "entropy": 5.464864301681518,
      "epoch": 4.029089683832201,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034710492105360653,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22364983409643174,
      "num_tokens": 80077451.0,
      "step": 35875
    },
    {
      "entropy": 5.390609169006348,
      "epoch": 4.029651260740159,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22432430386543273,
      "num_tokens": 80088215.0,
      "step": 35880
    },
    {
      "entropy": 5.347185897827148,
      "epoch": 4.030212837648116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.7808,
      "mean_token_accuracy": 0.22760084122419358,
      "num_tokens": 80098393.0,
      "step": 35885
    },
    {
      "entropy": 5.41992220878601,
      "epoch": 4.030774414556073,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000346990816911629,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21682439148426055,
      "num_tokens": 80110183.0,
      "step": 35890
    },
    {
      "entropy": 5.392170190811157,
      "epoch": 4.031335991464031,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22471424341201782,
      "num_tokens": 80120954.0,
      "step": 35895
    },
    {
      "entropy": 5.399175930023193,
      "epoch": 4.031897568371988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.8242,
      "mean_token_accuracy": 0.23077375143766404,
      "num_tokens": 80131607.0,
      "step": 35900
    },
    {
      "entropy": 5.373232030868531,
      "epoch": 4.032459145279946,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22548570930957795,
      "num_tokens": 80141700.0,
      "step": 35905
    },
    {
      "entropy": 5.392943954467773,
      "epoch": 4.033020722187904,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.936,
      "mean_token_accuracy": 0.22190795987844467,
      "num_tokens": 80151776.0,
      "step": 35910
    },
    {
      "entropy": 5.473532247543335,
      "epoch": 4.033582299095861,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21788766980171204,
      "num_tokens": 80162838.0,
      "step": 35915
    },
    {
      "entropy": 5.488190698623657,
      "epoch": 4.034143876003819,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22241785675287246,
      "num_tokens": 80174104.0,
      "step": 35920
    },
    {
      "entropy": 5.457074308395386,
      "epoch": 4.034705452911776,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034672449374266575,
      "loss": 5.0062,
      "mean_token_accuracy": 0.220038940012455,
      "num_tokens": 80185015.0,
      "step": 35925
    },
    {
      "entropy": 5.461394929885865,
      "epoch": 4.0352670298197335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22483104169368745,
      "num_tokens": 80196893.0,
      "step": 35930
    },
    {
      "entropy": 5.5345744609832765,
      "epoch": 4.035828606727692,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22547780126333236,
      "num_tokens": 80209910.0,
      "step": 35935
    },
    {
      "entropy": 5.41735954284668,
      "epoch": 4.036390183635649,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.929,
      "mean_token_accuracy": 0.22247291654348372,
      "num_tokens": 80219780.0,
      "step": 35940
    },
    {
      "entropy": 5.46521224975586,
      "epoch": 4.036951760543606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003465722588058497,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22036375254392623,
      "num_tokens": 80231316.0,
      "step": 35945
    },
    {
      "entropy": 5.50279016494751,
      "epoch": 4.037513337451564,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21695162653923034,
      "num_tokens": 80242622.0,
      "step": 35950
    },
    {
      "entropy": 5.498411321640015,
      "epoch": 4.038074914359521,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22724716663360595,
      "num_tokens": 80255357.0,
      "step": 35955
    },
    {
      "entropy": 5.504025793075561,
      "epoch": 4.0386364912674795,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.9684,
      "mean_token_accuracy": 0.22005678415298463,
      "num_tokens": 80267430.0,
      "step": 35960
    },
    {
      "entropy": 5.391363573074341,
      "epoch": 4.039198068175437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22690337151288986,
      "num_tokens": 80277820.0,
      "step": 35965
    },
    {
      "entropy": 5.520750713348389,
      "epoch": 4.039759645083394,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034638191386893115,
      "loss": 5.0578,
      "mean_token_accuracy": 0.20828215777873993,
      "num_tokens": 80289559.0,
      "step": 35970
    },
    {
      "entropy": 5.4724479675292965,
      "epoch": 4.040321221991352,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034634383805821504,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21591122895479203,
      "num_tokens": 80300452.0,
      "step": 35975
    },
    {
      "entropy": 5.480446481704712,
      "epoch": 4.040882798899309,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.9594,
      "mean_token_accuracy": 0.22420481592416763,
      "num_tokens": 80311403.0,
      "step": 35980
    },
    {
      "entropy": 5.440109014511108,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000346267679623149,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2250768333673477,
      "num_tokens": 80323245.0,
      "step": 35985
    },
    {
      "entropy": 5.402356386184692,
      "epoch": 4.042005952715225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.922,
      "mean_token_accuracy": 0.21486283242702484,
      "num_tokens": 80333666.0,
      "step": 35990
    },
    {
      "entropy": 5.473085498809814,
      "epoch": 4.042567529623182,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23221387714147568,
      "num_tokens": 80345291.0,
      "step": 35995
    },
    {
      "entropy": 5.400797700881958,
      "epoch": 4.04312910653114,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22822534590959548,
      "num_tokens": 80355612.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.228513087936059,
      "eval_loss": 5.131678104400635,
      "eval_mean_token_accuracy": 0.21992043310645193,
      "eval_num_tokens": 80355612.0,
      "eval_runtime": 53.2478,
      "eval_samples_per_second": 582.428,
      "eval_steps_per_second": 72.811,
      "step": 36000
    },
    {
      "entropy": 5.3761132717132565,
      "epoch": 4.043690683439097,
      "grad_norm": 1.046875,
      "learning_rate": 0.000346115335532431,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22393494844436646,
      "num_tokens": 80365952.0,
      "step": 36005
    },
    {
      "entropy": 5.431370687484741,
      "epoch": 4.044252260347054,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034607724384587006,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22233061045408248,
      "num_tokens": 80376962.0,
      "step": 36010
    },
    {
      "entropy": 5.546143484115601,
      "epoch": 4.044813837255012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034603914989618197,
      "loss": 5.024,
      "mean_token_accuracy": 0.2159732311964035,
      "num_tokens": 80388671.0,
      "step": 36015
    },
    {
      "entropy": 5.490001392364502,
      "epoch": 4.04537541416297,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003460010536845799,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22236818373203276,
      "num_tokens": 80399215.0,
      "step": 36020
    },
    {
      "entropy": 5.417406606674194,
      "epoch": 4.045936991070927,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003459629552122767,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2199813336133957,
      "num_tokens": 80410590.0,
      "step": 36025
    },
    {
      "entropy": 5.357647371292114,
      "epoch": 4.046498567978885,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003459248544804855,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22946387231349946,
      "num_tokens": 80422085.0,
      "step": 36030
    },
    {
      "entropy": 5.388637685775757,
      "epoch": 4.047060144886842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003458867514904194,
      "loss": 4.8398,
      "mean_token_accuracy": 0.22906061708927156,
      "num_tokens": 80434389.0,
      "step": 36035
    },
    {
      "entropy": 5.444587564468383,
      "epoch": 4.047621721794799,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003458486462432916,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22620894610881806,
      "num_tokens": 80445073.0,
      "step": 36040
    },
    {
      "entropy": 5.417082643508911,
      "epoch": 4.0481832987027575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034581053874031554,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22530607283115386,
      "num_tokens": 80455206.0,
      "step": 36045
    },
    {
      "entropy": 5.443615484237671,
      "epoch": 4.048744875610715,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003457724289827044,
      "loss": 4.931,
      "mean_token_accuracy": 0.22621400654315948,
      "num_tokens": 80467707.0,
      "step": 36050
    },
    {
      "entropy": 5.426360607147217,
      "epoch": 4.049306452518673,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003457343169716716,
      "loss": 4.9404,
      "mean_token_accuracy": 0.2232485070824623,
      "num_tokens": 80478493.0,
      "step": 36055
    },
    {
      "entropy": 5.403174686431885,
      "epoch": 4.04986802942663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034569620270843087,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22181425243616104,
      "num_tokens": 80490057.0,
      "step": 36060
    },
    {
      "entropy": 5.44171028137207,
      "epoch": 4.050429606334587,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034565808619419554,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21364964693784713,
      "num_tokens": 80500590.0,
      "step": 36065
    },
    {
      "entropy": 5.486949014663696,
      "epoch": 4.050991183242545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003456199674301793,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2220090761780739,
      "num_tokens": 80510836.0,
      "step": 36070
    },
    {
      "entropy": 5.44735050201416,
      "epoch": 4.051552760150503,
      "grad_norm": 1.0625,
      "learning_rate": 0.000345581846417596,
      "loss": 4.9815,
      "mean_token_accuracy": 0.21667204350233077,
      "num_tokens": 80521770.0,
      "step": 36075
    },
    {
      "entropy": 5.500047731399536,
      "epoch": 4.05211433705846,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034554372315765927,
      "loss": 4.9489,
      "mean_token_accuracy": 0.2268861085176468,
      "num_tokens": 80532331.0,
      "step": 36080
    },
    {
      "entropy": 5.490350961685181,
      "epoch": 4.052675913966418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003455055976515831,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22157858312129974,
      "num_tokens": 80544858.0,
      "step": 36085
    },
    {
      "entropy": 5.470765733718872,
      "epoch": 4.053237490874375,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034546746990058125,
      "loss": 5.0269,
      "mean_token_accuracy": 0.22071103900671005,
      "num_tokens": 80556493.0,
      "step": 36090
    },
    {
      "entropy": 5.450572967529297,
      "epoch": 4.053799067782332,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034542933990586795,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22407526075839995,
      "num_tokens": 80567416.0,
      "step": 36095
    },
    {
      "entropy": 5.4614784717559814,
      "epoch": 4.0543606446902904,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034539120766865713,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22070671617984772,
      "num_tokens": 80578175.0,
      "step": 36100
    },
    {
      "entropy": 5.4004474639892575,
      "epoch": 4.054922221598248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003453530731901628,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23230642229318618,
      "num_tokens": 80588755.0,
      "step": 36105
    },
    {
      "entropy": 5.415850019454956,
      "epoch": 4.055483798506206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003453149364715995,
      "loss": 4.977,
      "mean_token_accuracy": 0.2219657689332962,
      "num_tokens": 80599734.0,
      "step": 36110
    },
    {
      "entropy": 5.365566539764404,
      "epoch": 4.056045375414163,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034527679751418126,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23710959404706955,
      "num_tokens": 80610765.0,
      "step": 36115
    },
    {
      "entropy": 5.490283012390137,
      "epoch": 4.05660695232212,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003452386563191226,
      "loss": 4.985,
      "mean_token_accuracy": 0.22549961358308793,
      "num_tokens": 80622680.0,
      "step": 36120
    },
    {
      "entropy": 5.470968627929688,
      "epoch": 4.057168529230078,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034520051288763784,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2176246538758278,
      "num_tokens": 80633115.0,
      "step": 36125
    },
    {
      "entropy": 5.465948057174683,
      "epoch": 4.0577301061380355,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003451623672209415,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22344064712524414,
      "num_tokens": 80643750.0,
      "step": 36130
    },
    {
      "entropy": 5.507699489593506,
      "epoch": 4.058291683045993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034512421932024825,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21805242747068404,
      "num_tokens": 80654392.0,
      "step": 36135
    },
    {
      "entropy": 5.397718858718872,
      "epoch": 4.058853259953951,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003450860691867726,
      "loss": 4.937,
      "mean_token_accuracy": 0.21703851819038392,
      "num_tokens": 80665509.0,
      "step": 36140
    },
    {
      "entropy": 5.403131723403931,
      "epoch": 4.059414836861908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034504791682172934,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2236579567193985,
      "num_tokens": 80676895.0,
      "step": 36145
    },
    {
      "entropy": 5.45961012840271,
      "epoch": 4.059976413769866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003450097622263333,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22592142671346666,
      "num_tokens": 80688321.0,
      "step": 36150
    },
    {
      "entropy": 5.439686298370361,
      "epoch": 4.060537990677823,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034497160540179924,
      "loss": 4.8329,
      "mean_token_accuracy": 0.24019111543893815,
      "num_tokens": 80699534.0,
      "step": 36155
    },
    {
      "entropy": 5.339393472671508,
      "epoch": 4.061099567585781,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034493344634934205,
      "loss": 4.8374,
      "mean_token_accuracy": 0.23828351944684983,
      "num_tokens": 80710999.0,
      "step": 36160
    },
    {
      "entropy": 5.362044095993042,
      "epoch": 4.061661144493739,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003448952850701769,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22030559778213502,
      "num_tokens": 80722943.0,
      "step": 36165
    },
    {
      "entropy": 5.3928242206573485,
      "epoch": 4.062222721401696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034485712156551873,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21641165316104888,
      "num_tokens": 80733142.0,
      "step": 36170
    },
    {
      "entropy": 5.411113023757935,
      "epoch": 4.062784298309653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034481895583658267,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22154947072267533,
      "num_tokens": 80745228.0,
      "step": 36175
    },
    {
      "entropy": 5.455417203903198,
      "epoch": 4.063345875217611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034478078788458406,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22280148416757584,
      "num_tokens": 80756866.0,
      "step": 36180
    },
    {
      "entropy": 5.388067054748535,
      "epoch": 4.0639074521255685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000344742617710738,
      "loss": 4.8438,
      "mean_token_accuracy": 0.23137331753969193,
      "num_tokens": 80767541.0,
      "step": 36185
    },
    {
      "entropy": 5.44798469543457,
      "epoch": 4.064469029033527,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034470444531626003,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2243697836995125,
      "num_tokens": 80779577.0,
      "step": 36190
    },
    {
      "entropy": 5.471805381774902,
      "epoch": 4.065030605941484,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034466627070236545,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22290571331977843,
      "num_tokens": 80790569.0,
      "step": 36195
    },
    {
      "entropy": 5.524180555343628,
      "epoch": 4.065592182849441,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003446280938702698,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21925498098134993,
      "num_tokens": 80800887.0,
      "step": 36200
    },
    {
      "entropy": 5.4990074157714846,
      "epoch": 4.066153759757399,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034458991482118865,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22187986522912978,
      "num_tokens": 80811967.0,
      "step": 36205
    },
    {
      "entropy": 5.413905477523803,
      "epoch": 4.066715336665356,
      "grad_norm": 1.125,
      "learning_rate": 0.0003445517335563376,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23049579858779906,
      "num_tokens": 80824770.0,
      "step": 36210
    },
    {
      "entropy": 5.380900096893311,
      "epoch": 4.067276913573314,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034451355007693237,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22539349943399428,
      "num_tokens": 80834913.0,
      "step": 36215
    },
    {
      "entropy": 5.424925708770752,
      "epoch": 4.067838490481272,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003444753643841888,
      "loss": 4.9104,
      "mean_token_accuracy": 0.2180577665567398,
      "num_tokens": 80846555.0,
      "step": 36220
    },
    {
      "entropy": 5.559576177597046,
      "epoch": 4.068400067389229,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034443717647932265,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2180319145321846,
      "num_tokens": 80857774.0,
      "step": 36225
    },
    {
      "entropy": 5.5771996021270756,
      "epoch": 4.068961644297186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003443989863635498,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21593445986509324,
      "num_tokens": 80868618.0,
      "step": 36230
    },
    {
      "entropy": 5.499241065979004,
      "epoch": 4.069523221205144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034436079403808647,
      "loss": 5.033,
      "mean_token_accuracy": 0.20998706966638564,
      "num_tokens": 80880319.0,
      "step": 36235
    },
    {
      "entropy": 5.426686382293701,
      "epoch": 4.070084798113101,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034432259950414846,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22057519108057022,
      "num_tokens": 80891194.0,
      "step": 36240
    },
    {
      "entropy": 5.346396827697754,
      "epoch": 4.0706463750210595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000344284402762952,
      "loss": 4.8055,
      "mean_token_accuracy": 0.22565880715847014,
      "num_tokens": 80902322.0,
      "step": 36245
    },
    {
      "entropy": 5.48210506439209,
      "epoch": 4.071207951929017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003442462038157132,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2182726487517357,
      "num_tokens": 80913324.0,
      "step": 36250
    },
    {
      "entropy": 5.389068603515625,
      "epoch": 4.071769528836974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003442080026636485,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23049925565719603,
      "num_tokens": 80923360.0,
      "step": 36255
    },
    {
      "entropy": 5.477072477340698,
      "epoch": 4.072331105744932,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003441697993079742,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2164838805794716,
      "num_tokens": 80935359.0,
      "step": 36260
    },
    {
      "entropy": 5.369146013259888,
      "epoch": 4.072892682652889,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003441315937499066,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22713578194379808,
      "num_tokens": 80945885.0,
      "step": 36265
    },
    {
      "entropy": 5.467874813079834,
      "epoch": 4.0734542595608465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034409338599066224,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2263147085905075,
      "num_tokens": 80955700.0,
      "step": 36270
    },
    {
      "entropy": 5.423817443847656,
      "epoch": 4.074015836468805,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003440551760314577,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2271623879671097,
      "num_tokens": 80967130.0,
      "step": 36275
    },
    {
      "entropy": 5.471257305145263,
      "epoch": 4.074577413376762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034401696387350954,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22129244357347488,
      "num_tokens": 80979677.0,
      "step": 36280
    },
    {
      "entropy": 5.462257385253906,
      "epoch": 4.075138990284719,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003439787495180345,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21525031477212905,
      "num_tokens": 80990563.0,
      "step": 36285
    },
    {
      "entropy": 5.404213571548462,
      "epoch": 4.075700567192677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034394053296624936,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22815835922956468,
      "num_tokens": 81001731.0,
      "step": 36290
    },
    {
      "entropy": 5.495741081237793,
      "epoch": 4.076262144100634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003439023142193709,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22327785640954972,
      "num_tokens": 81013072.0,
      "step": 36295
    },
    {
      "entropy": 5.425687646865844,
      "epoch": 4.0768237210085925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000343864093278616,
      "loss": 4.9434,
      "mean_token_accuracy": 0.230279840528965,
      "num_tokens": 81023628.0,
      "step": 36300
    },
    {
      "entropy": 5.396432828903198,
      "epoch": 4.07738529791655,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003438258701452017,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23041560649871826,
      "num_tokens": 81034856.0,
      "step": 36305
    },
    {
      "entropy": 5.489410781860352,
      "epoch": 4.077946874824507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003437876448203449,
      "loss": 4.9804,
      "mean_token_accuracy": 0.22601220905780792,
      "num_tokens": 81046998.0,
      "step": 36310
    },
    {
      "entropy": 5.4090217590332035,
      "epoch": 4.078508451732465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003437494173052629,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22620512992143632,
      "num_tokens": 81058292.0,
      "step": 36315
    },
    {
      "entropy": 5.428522682189941,
      "epoch": 4.079070028640422,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003437111876011728,
      "loss": 4.9053,
      "mean_token_accuracy": 0.22751108855009078,
      "num_tokens": 81069907.0,
      "step": 36320
    },
    {
      "entropy": 5.44788293838501,
      "epoch": 4.0796316055483794,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034367295570929176,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21710942685604095,
      "num_tokens": 81080700.0,
      "step": 36325
    },
    {
      "entropy": 5.430348205566406,
      "epoch": 4.0801931824563376,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003436347216308373,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2212770774960518,
      "num_tokens": 81092484.0,
      "step": 36330
    },
    {
      "entropy": 5.434270000457763,
      "epoch": 4.080754759364295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003435964853670265,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22424254864454268,
      "num_tokens": 81103992.0,
      "step": 36335
    },
    {
      "entropy": 5.4609935760498045,
      "epoch": 4.081316336272253,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034355824691907714,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21537371575832367,
      "num_tokens": 81114995.0,
      "step": 36340
    },
    {
      "entropy": 5.318227672576905,
      "epoch": 4.08187791318021,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034352000628820655,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23897641152143478,
      "num_tokens": 81125627.0,
      "step": 36345
    },
    {
      "entropy": 5.467759418487549,
      "epoch": 4.082439490088167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003434817634756324,
      "loss": 5.0484,
      "mean_token_accuracy": 0.2184484302997589,
      "num_tokens": 81136957.0,
      "step": 36350
    },
    {
      "entropy": 5.432562589645386,
      "epoch": 4.083001066996125,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003434435184825724,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22530073076486587,
      "num_tokens": 81148840.0,
      "step": 36355
    },
    {
      "entropy": 5.527706480026245,
      "epoch": 4.083562643904083,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034340527131024415,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22304429858922958,
      "num_tokens": 81160385.0,
      "step": 36360
    },
    {
      "entropy": 5.421856880187988,
      "epoch": 4.08412422081204,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003433670219598656,
      "loss": 4.913,
      "mean_token_accuracy": 0.22120100855827332,
      "num_tokens": 81171334.0,
      "step": 36365
    },
    {
      "entropy": 5.40715446472168,
      "epoch": 4.084685797719998,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003433287704326545,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23071181327104567,
      "num_tokens": 81182547.0,
      "step": 36370
    },
    {
      "entropy": 5.406431245803833,
      "epoch": 4.085247374627955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003432905167298288,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2237130492925644,
      "num_tokens": 81193427.0,
      "step": 36375
    },
    {
      "entropy": 5.418706035614013,
      "epoch": 4.085808951535913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034325226085260666,
      "loss": 4.8661,
      "mean_token_accuracy": 0.22958991080522537,
      "num_tokens": 81204522.0,
      "step": 36380
    },
    {
      "entropy": 5.440243530273437,
      "epoch": 4.0863705284438705,
      "grad_norm": 1.125,
      "learning_rate": 0.000343214002802206,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21950859874486922,
      "num_tokens": 81215209.0,
      "step": 36385
    },
    {
      "entropy": 5.409033966064453,
      "epoch": 4.086932105351828,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034317574257984514,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2203845500946045,
      "num_tokens": 81226083.0,
      "step": 36390
    },
    {
      "entropy": 5.407149934768677,
      "epoch": 4.087493682259786,
      "grad_norm": 1.0625,
      "learning_rate": 0.000343137480186742,
      "loss": 4.8557,
      "mean_token_accuracy": 0.22648276388645172,
      "num_tokens": 81236531.0,
      "step": 36395
    },
    {
      "entropy": 5.440310335159301,
      "epoch": 4.088055259167743,
      "grad_norm": 1.0,
      "learning_rate": 0.00034309921562411527,
      "loss": 4.9392,
      "mean_token_accuracy": 0.21713693737983703,
      "num_tokens": 81246856.0,
      "step": 36400
    },
    {
      "entropy": 5.485646915435791,
      "epoch": 4.0886168360757,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034306094889318295,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21904641389846802,
      "num_tokens": 81259352.0,
      "step": 36405
    },
    {
      "entropy": 5.481346368789673,
      "epoch": 4.089178412983658,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003430226799951637,
      "loss": 5.035,
      "mean_token_accuracy": 0.21562340259552001,
      "num_tokens": 81270020.0,
      "step": 36410
    },
    {
      "entropy": 5.429272031784057,
      "epoch": 4.089739989891616,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034298440893127584,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21878219693899154,
      "num_tokens": 81281625.0,
      "step": 36415
    },
    {
      "entropy": 5.395605993270874,
      "epoch": 4.090301566799573,
      "grad_norm": 1.125,
      "learning_rate": 0.00034294613570273817,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23670536279678345,
      "num_tokens": 81292340.0,
      "step": 36420
    },
    {
      "entropy": 5.415265130996704,
      "epoch": 4.090863143707531,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034290786031076917,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22203906774520873,
      "num_tokens": 81304494.0,
      "step": 36425
    },
    {
      "entropy": 5.427768230438232,
      "epoch": 4.091424720615488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003428695827565875,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22704363465309144,
      "num_tokens": 81315365.0,
      "step": 36430
    },
    {
      "entropy": 5.464547920227051,
      "epoch": 4.091986297523446,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034283130304141204,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21996108442544937,
      "num_tokens": 81326970.0,
      "step": 36435
    },
    {
      "entropy": 5.415773200988769,
      "epoch": 4.092547874431403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003427930211664615,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22785191237926483,
      "num_tokens": 81337117.0,
      "step": 36440
    },
    {
      "entropy": 5.4475147247314455,
      "epoch": 4.093109451339361,
      "grad_norm": 1.078125,
      "learning_rate": 0.000342754737132955,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2211699217557907,
      "num_tokens": 81347984.0,
      "step": 36445
    },
    {
      "entropy": 5.481760931015015,
      "epoch": 4.093671028247319,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034271645094211125,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22810928523540497,
      "num_tokens": 81358685.0,
      "step": 36450
    },
    {
      "entropy": 5.444785785675049,
      "epoch": 4.094232605155276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003426781625951495,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21794984489679337,
      "num_tokens": 81368724.0,
      "step": 36455
    },
    {
      "entropy": 5.5131677150726315,
      "epoch": 4.094794182063233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034263987209328886,
      "loss": 5.0725,
      "mean_token_accuracy": 0.2162942573428154,
      "num_tokens": 81380570.0,
      "step": 36460
    },
    {
      "entropy": 5.454878473281861,
      "epoch": 4.095355758971191,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003426015794377483,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2224457949399948,
      "num_tokens": 81391606.0,
      "step": 36465
    },
    {
      "entropy": 5.519129896163941,
      "epoch": 4.0959173358791485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003425632846297474,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22118670493364334,
      "num_tokens": 81402439.0,
      "step": 36470
    },
    {
      "entropy": 5.426743602752685,
      "epoch": 4.096478912787107,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034252498767050525,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21968540400266648,
      "num_tokens": 81413500.0,
      "step": 36475
    },
    {
      "entropy": 5.3637086868286135,
      "epoch": 4.097040489695064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003424866885612413,
      "loss": 4.8472,
      "mean_token_accuracy": 0.22630822360515596,
      "num_tokens": 81424656.0,
      "step": 36480
    },
    {
      "entropy": 5.556217050552368,
      "epoch": 4.097602066603021,
      "grad_norm": 1.078125,
      "learning_rate": 0.000342448387303175,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21847063452005386,
      "num_tokens": 81436597.0,
      "step": 36485
    },
    {
      "entropy": 5.526525974273682,
      "epoch": 4.098163643510979,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034241008389752594,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21743009239435196,
      "num_tokens": 81448955.0,
      "step": 36490
    },
    {
      "entropy": 5.473721647262574,
      "epoch": 4.098725220418936,
      "grad_norm": 1.796875,
      "learning_rate": 0.0003423717783455136,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22408452332019807,
      "num_tokens": 81460445.0,
      "step": 36495
    },
    {
      "entropy": 5.456004428863525,
      "epoch": 4.099286797326894,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034233347064835775,
      "loss": 4.9283,
      "mean_token_accuracy": 0.21802510768175126,
      "num_tokens": 81472103.0,
      "step": 36500
    },
    {
      "entropy": 5.40439772605896,
      "epoch": 4.099848374234852,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003422951608072781,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22538803219795228,
      "num_tokens": 81483051.0,
      "step": 36505
    },
    {
      "entropy": 5.3894140243530275,
      "epoch": 4.100409951142809,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003422568488234944,
      "loss": 4.863,
      "mean_token_accuracy": 0.22680689245462418,
      "num_tokens": 81493755.0,
      "step": 36510
    },
    {
      "entropy": 5.328900527954102,
      "epoch": 4.100971528050766,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034221853469822654,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22631326764822007,
      "num_tokens": 81505738.0,
      "step": 36515
    },
    {
      "entropy": 5.458355140686035,
      "epoch": 4.101533104958724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034218021843269444,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2181945264339447,
      "num_tokens": 81516410.0,
      "step": 36520
    },
    {
      "entropy": 5.469938850402832,
      "epoch": 4.1020946818666815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003421419000281182,
      "loss": 4.9718,
      "mean_token_accuracy": 0.21943264156579972,
      "num_tokens": 81527516.0,
      "step": 36525
    },
    {
      "entropy": 5.505574321746826,
      "epoch": 4.10265625877464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003421035794857178,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22195605039596558,
      "num_tokens": 81538226.0,
      "step": 36530
    },
    {
      "entropy": 5.494089269638062,
      "epoch": 4.103217835682597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003420652568067134,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2161216229200363,
      "num_tokens": 81549699.0,
      "step": 36535
    },
    {
      "entropy": 5.535510540008545,
      "epoch": 4.103779412590554,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034202693199232516,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21109698712825775,
      "num_tokens": 81560404.0,
      "step": 36540
    },
    {
      "entropy": 5.516623067855835,
      "epoch": 4.104340989498512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034198860504377343,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21969870179891587,
      "num_tokens": 81571758.0,
      "step": 36545
    },
    {
      "entropy": 5.435519170761109,
      "epoch": 4.104902566406469,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034195027596227857,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23014482408761977,
      "num_tokens": 81583555.0,
      "step": 36550
    },
    {
      "entropy": 5.3485359191894535,
      "epoch": 4.1054641433144266,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003419119447490609,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23024367839097976,
      "num_tokens": 81594901.0,
      "step": 36555
    },
    {
      "entropy": 5.428869819641113,
      "epoch": 4.106025720222385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000341873611405341,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22897766679525375,
      "num_tokens": 81606229.0,
      "step": 36560
    },
    {
      "entropy": 5.609891176223755,
      "epoch": 4.106587297130342,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003418352759323394,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21797014623880387,
      "num_tokens": 81616543.0,
      "step": 36565
    },
    {
      "entropy": 5.456070041656494,
      "epoch": 4.1071488740383,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034179693833127667,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22834302335977555,
      "num_tokens": 81627517.0,
      "step": 36570
    },
    {
      "entropy": 5.399865388870239,
      "epoch": 4.107710450946257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034175859860337347,
      "loss": 4.926,
      "mean_token_accuracy": 0.22650080919265747,
      "num_tokens": 81637235.0,
      "step": 36575
    },
    {
      "entropy": 5.4107582569122314,
      "epoch": 4.108272027854214,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034172025674985076,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22121364027261733,
      "num_tokens": 81648539.0,
      "step": 36580
    },
    {
      "entropy": 5.42774600982666,
      "epoch": 4.1088336047621725,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034168191277192904,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22568051517009735,
      "num_tokens": 81659981.0,
      "step": 36585
    },
    {
      "entropy": 5.443951177597046,
      "epoch": 4.10939518167013,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034164356667082933,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22648842483758927,
      "num_tokens": 81669602.0,
      "step": 36590
    },
    {
      "entropy": 5.438381290435791,
      "epoch": 4.109956758578087,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003416052184477727,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2210528314113617,
      "num_tokens": 81679995.0,
      "step": 36595
    },
    {
      "entropy": 5.398089456558227,
      "epoch": 4.110518335486045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003415668681039801,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22185308784246444,
      "num_tokens": 81691295.0,
      "step": 36600
    },
    {
      "entropy": 5.449099349975586,
      "epoch": 4.111079912394002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034152851564067254,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22342463731765747,
      "num_tokens": 81700966.0,
      "step": 36605
    },
    {
      "entropy": 5.549625110626221,
      "epoch": 4.1116414893019595,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003414901610590712,
      "loss": 5.041,
      "mean_token_accuracy": 0.21454848796129228,
      "num_tokens": 81713469.0,
      "step": 36610
    },
    {
      "entropy": 5.482658243179321,
      "epoch": 4.112203066209918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003414518043603974,
      "loss": 4.928,
      "mean_token_accuracy": 0.22331971377134324,
      "num_tokens": 81724748.0,
      "step": 36615
    },
    {
      "entropy": 5.452486562728882,
      "epoch": 4.112764643117875,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034141344554587243,
      "loss": 5.0438,
      "mean_token_accuracy": 0.2162758231163025,
      "num_tokens": 81735694.0,
      "step": 36620
    },
    {
      "entropy": 5.514401054382324,
      "epoch": 4.113326220025833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003413750846167175,
      "loss": 4.9581,
      "mean_token_accuracy": 0.21518886536359788,
      "num_tokens": 81748251.0,
      "step": 36625
    },
    {
      "entropy": 5.4402806758880615,
      "epoch": 4.11388779693379,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003413367215741542,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22733558863401412,
      "num_tokens": 81760750.0,
      "step": 36630
    },
    {
      "entropy": 5.486662912368774,
      "epoch": 4.114449373841747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034129835641940394,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21325063705444336,
      "num_tokens": 81772100.0,
      "step": 36635
    },
    {
      "entropy": 5.455642795562744,
      "epoch": 4.1150109507497055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034125998915368827,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21600612998008728,
      "num_tokens": 81783795.0,
      "step": 36640
    },
    {
      "entropy": 5.4123035907745365,
      "epoch": 4.115572527657663,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034122161977822885,
      "loss": 4.96,
      "mean_token_accuracy": 0.22377222329378127,
      "num_tokens": 81794713.0,
      "step": 36645
    },
    {
      "entropy": 5.439058113098144,
      "epoch": 4.11613410456562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003411832482942474,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2236935168504715,
      "num_tokens": 81806493.0,
      "step": 36650
    },
    {
      "entropy": 5.393202400207519,
      "epoch": 4.116695681473578,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034114487470296575,
      "loss": 4.8258,
      "mean_token_accuracy": 0.23110629618167877,
      "num_tokens": 81818232.0,
      "step": 36655
    },
    {
      "entropy": 5.359373283386231,
      "epoch": 4.117257258381535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034110649900560554,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22689494639635086,
      "num_tokens": 81829439.0,
      "step": 36660
    },
    {
      "entropy": 5.47729458808899,
      "epoch": 4.117818835289493,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003410681212033887,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2267334908246994,
      "num_tokens": 81839585.0,
      "step": 36665
    },
    {
      "entropy": 5.470051097869873,
      "epoch": 4.1183804121974505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034102974129753726,
      "loss": 5.0014,
      "mean_token_accuracy": 0.22091586291790008,
      "num_tokens": 81850648.0,
      "step": 36670
    },
    {
      "entropy": 5.446447563171387,
      "epoch": 4.118941989105408,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003409913592892733,
      "loss": 4.9316,
      "mean_token_accuracy": 0.21857986599206924,
      "num_tokens": 81862597.0,
      "step": 36675
    },
    {
      "entropy": 5.499032783508301,
      "epoch": 4.119503566013366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003409529751798189,
      "loss": 4.979,
      "mean_token_accuracy": 0.2249290704727173,
      "num_tokens": 81874540.0,
      "step": 36680
    },
    {
      "entropy": 5.428309154510498,
      "epoch": 4.120065142921323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003409145889703961,
      "loss": 4.8612,
      "mean_token_accuracy": 0.228695909678936,
      "num_tokens": 81884205.0,
      "step": 36685
    },
    {
      "entropy": 5.3693845748901365,
      "epoch": 4.12062671982928,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034087620066222733,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22267425507307054,
      "num_tokens": 81895552.0,
      "step": 36690
    },
    {
      "entropy": 5.4880446434021,
      "epoch": 4.121188296737238,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003408378102565347,
      "loss": 5.0621,
      "mean_token_accuracy": 0.20870553106069564,
      "num_tokens": 81908169.0,
      "step": 36695
    },
    {
      "entropy": 5.439088153839111,
      "epoch": 4.121749873645196,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003407994177545407,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22023088932037355,
      "num_tokens": 81919465.0,
      "step": 36700
    },
    {
      "entropy": 5.495008182525635,
      "epoch": 4.122311450553153,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003407610231574677,
      "loss": 5.027,
      "mean_token_accuracy": 0.21799664795398713,
      "num_tokens": 81930558.0,
      "step": 36705
    },
    {
      "entropy": 5.48842487335205,
      "epoch": 4.122873027461111,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003407226264665382,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22635471522808076,
      "num_tokens": 81941325.0,
      "step": 36710
    },
    {
      "entropy": 5.426229047775268,
      "epoch": 4.123434604369068,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034068422768297486,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2222608745098114,
      "num_tokens": 81953225.0,
      "step": 36715
    },
    {
      "entropy": 5.489975643157959,
      "epoch": 4.123996181277026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003406458268080001,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22292101830244065,
      "num_tokens": 81964365.0,
      "step": 36720
    },
    {
      "entropy": 5.3724778175354,
      "epoch": 4.1245577581849835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034060742384283695,
      "loss": 4.944,
      "mean_token_accuracy": 0.21955933570861816,
      "num_tokens": 81974648.0,
      "step": 36725
    },
    {
      "entropy": 5.362296056747437,
      "epoch": 4.125119335092941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003405690187887078,
      "loss": 4.8667,
      "mean_token_accuracy": 0.23144982904195785,
      "num_tokens": 81985502.0,
      "step": 36730
    },
    {
      "entropy": 5.555491256713867,
      "epoch": 4.125680912000899,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034053061164683574,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21988180726766587,
      "num_tokens": 81998279.0,
      "step": 36735
    },
    {
      "entropy": 5.4810707569122314,
      "epoch": 4.126242488908856,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034049220241844357,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22617652118206025,
      "num_tokens": 82009911.0,
      "step": 36740
    },
    {
      "entropy": 5.344607925415039,
      "epoch": 4.126804065816813,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003404537911047543,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23450527638196944,
      "num_tokens": 82020219.0,
      "step": 36745
    },
    {
      "entropy": 5.353504371643067,
      "epoch": 4.127365642724771,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034041537770699095,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23204591870307922,
      "num_tokens": 82030997.0,
      "step": 36750
    },
    {
      "entropy": 5.485170030593872,
      "epoch": 4.127927219632729,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003403769622263765,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22609869241714478,
      "num_tokens": 82043384.0,
      "step": 36755
    },
    {
      "entropy": 5.431669521331787,
      "epoch": 4.128488796540687,
      "grad_norm": 1.125,
      "learning_rate": 0.00034033854466413425,
      "loss": 4.8612,
      "mean_token_accuracy": 0.22571199834346772,
      "num_tokens": 82053346.0,
      "step": 36760
    },
    {
      "entropy": 5.4369720935821535,
      "epoch": 4.129050373448644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003403001250214875,
      "loss": 4.8637,
      "mean_token_accuracy": 0.22336917072534562,
      "num_tokens": 82063628.0,
      "step": 36765
    },
    {
      "entropy": 5.398493862152099,
      "epoch": 4.129611950356601,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003402617032996593,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23357643336057662,
      "num_tokens": 82074097.0,
      "step": 36770
    },
    {
      "entropy": 5.4311059474945065,
      "epoch": 4.130173527264559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003402232794998733,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2125575214624405,
      "num_tokens": 82085159.0,
      "step": 36775
    },
    {
      "entropy": 5.455510759353638,
      "epoch": 4.130735104172516,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034018485362335266,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22419775724411012,
      "num_tokens": 82096299.0,
      "step": 36780
    },
    {
      "entropy": 5.393710613250732,
      "epoch": 4.131296681080474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003401464256713211,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22388868331909179,
      "num_tokens": 82107111.0,
      "step": 36785
    },
    {
      "entropy": 5.418604135513306,
      "epoch": 4.131858257988432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034010799564500197,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22545289397239685,
      "num_tokens": 82117889.0,
      "step": 36790
    },
    {
      "entropy": 5.409048700332642,
      "epoch": 4.132419834896389,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003400695635456191,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22808441668748855,
      "num_tokens": 82128380.0,
      "step": 36795
    },
    {
      "entropy": 5.4062902450561525,
      "epoch": 4.132981411804346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003400311293743961,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23228008449077606,
      "num_tokens": 82138368.0,
      "step": 36800
    },
    {
      "entropy": 5.508398246765137,
      "epoch": 4.133542988712304,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003399926931325567,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21781384199857712,
      "num_tokens": 82148449.0,
      "step": 36805
    },
    {
      "entropy": 5.347343826293946,
      "epoch": 4.1341045656202615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033995425482132464,
      "loss": 4.8351,
      "mean_token_accuracy": 0.2297959491610527,
      "num_tokens": 82160937.0,
      "step": 36810
    },
    {
      "entropy": 5.385529661178589,
      "epoch": 4.13466614252822,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033991581444192405,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23060031682252885,
      "num_tokens": 82171546.0,
      "step": 36815
    },
    {
      "entropy": 5.422704362869263,
      "epoch": 4.135227719436177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003398773719955787,
      "loss": 4.9068,
      "mean_token_accuracy": 0.230114084482193,
      "num_tokens": 82182613.0,
      "step": 36820
    },
    {
      "entropy": 5.340125656127929,
      "epoch": 4.135789296344134,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003398389274835126,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22936685830354692,
      "num_tokens": 82193757.0,
      "step": 36825
    },
    {
      "entropy": 5.424433183670044,
      "epoch": 4.136350873252092,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033980048090694997,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2197564959526062,
      "num_tokens": 82205601.0,
      "step": 36830
    },
    {
      "entropy": 5.465984344482422,
      "epoch": 4.136912450160049,
      "grad_norm": 1.125,
      "learning_rate": 0.0003397620322671149,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22245920151472093,
      "num_tokens": 82217022.0,
      "step": 36835
    },
    {
      "entropy": 5.399036502838134,
      "epoch": 4.137474027068007,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033972358156523163,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22809913903474807,
      "num_tokens": 82227682.0,
      "step": 36840
    },
    {
      "entropy": 5.374110555648803,
      "epoch": 4.138035603975965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003396851288025243,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22322924435138702,
      "num_tokens": 82238754.0,
      "step": 36845
    },
    {
      "entropy": 5.528666543960571,
      "epoch": 4.138597180883922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003396466739802175,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2150941848754883,
      "num_tokens": 82249536.0,
      "step": 36850
    },
    {
      "entropy": 5.455479669570923,
      "epoch": 4.13915875779188,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003396082170995355,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2253570333123207,
      "num_tokens": 82260374.0,
      "step": 36855
    },
    {
      "entropy": 5.419986867904663,
      "epoch": 4.139720334699837,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003395697581617028,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22664140164852142,
      "num_tokens": 82271801.0,
      "step": 36860
    },
    {
      "entropy": 5.4699193954467775,
      "epoch": 4.1402819116077945,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033953129716794396,
      "loss": 5.0568,
      "mean_token_accuracy": 0.21430994272232057,
      "num_tokens": 82283149.0,
      "step": 36865
    },
    {
      "entropy": 5.466754674911499,
      "epoch": 4.140843488515753,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003394928341194836,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22777072340250015,
      "num_tokens": 82293003.0,
      "step": 36870
    },
    {
      "entropy": 5.48637113571167,
      "epoch": 4.14140506542371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003394543690175464,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22452080696821214,
      "num_tokens": 82303734.0,
      "step": 36875
    },
    {
      "entropy": 5.372871160507202,
      "epoch": 4.141966642331667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033941590186335703,
      "loss": 4.904,
      "mean_token_accuracy": 0.22427805066108703,
      "num_tokens": 82314569.0,
      "step": 36880
    },
    {
      "entropy": 5.445491647720337,
      "epoch": 4.142528219239625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003393774326581404,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2239125296473503,
      "num_tokens": 82325174.0,
      "step": 36885
    },
    {
      "entropy": 5.4745180130004885,
      "epoch": 4.143089796147582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033933896140312135,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22396914660930634,
      "num_tokens": 82337448.0,
      "step": 36890
    },
    {
      "entropy": 5.493487119674683,
      "epoch": 4.1436513730555395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033930048809952484,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22056402117013932,
      "num_tokens": 82348705.0,
      "step": 36895
    },
    {
      "entropy": 5.51518063545227,
      "epoch": 4.144212949963498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033926201274857575,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22217679470777513,
      "num_tokens": 82360439.0,
      "step": 36900
    },
    {
      "entropy": 5.440210962295533,
      "epoch": 4.144774526871455,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033922353535149934,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22659475803375245,
      "num_tokens": 82371346.0,
      "step": 36905
    },
    {
      "entropy": 5.363571119308472,
      "epoch": 4.145336103779413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003391850559095207,
      "loss": 4.8404,
      "mean_token_accuracy": 0.23183912485837938,
      "num_tokens": 82382221.0,
      "step": 36910
    },
    {
      "entropy": 5.452501440048218,
      "epoch": 4.14589768068737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003391465744238649,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21603289544582366,
      "num_tokens": 82392290.0,
      "step": 36915
    },
    {
      "entropy": 5.497771883010865,
      "epoch": 4.146459257595327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003391080908957573,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21558503657579423,
      "num_tokens": 82403513.0,
      "step": 36920
    },
    {
      "entropy": 5.415082597732544,
      "epoch": 4.1470208345032855,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033906960532642327,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23785481005907058,
      "num_tokens": 82414537.0,
      "step": 36925
    },
    {
      "entropy": 5.412952041625976,
      "epoch": 4.147582411411243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003390311177170881,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22820528149604796,
      "num_tokens": 82425330.0,
      "step": 36930
    },
    {
      "entropy": 5.309172964096069,
      "epoch": 4.1481439883192,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003389926280689773,
      "loss": 4.8608,
      "mean_token_accuracy": 0.2256098210811615,
      "num_tokens": 82435703.0,
      "step": 36935
    },
    {
      "entropy": 5.439262008666992,
      "epoch": 4.148705565227158,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033895413638331646,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22197455912828445,
      "num_tokens": 82447302.0,
      "step": 36940
    },
    {
      "entropy": 5.43472375869751,
      "epoch": 4.149267142135115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003389156426613311,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21679594963788987,
      "num_tokens": 82458379.0,
      "step": 36945
    },
    {
      "entropy": 5.463861703872681,
      "epoch": 4.149828719043073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033887714690424687,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22778795808553695,
      "num_tokens": 82469728.0,
      "step": 36950
    },
    {
      "entropy": 5.517714929580689,
      "epoch": 4.150390295951031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033883864911328954,
      "loss": 5.014,
      "mean_token_accuracy": 0.21839865148067475,
      "num_tokens": 82482216.0,
      "step": 36955
    },
    {
      "entropy": 5.462868070602417,
      "epoch": 4.150951872858988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033880014928968483,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2244784951210022,
      "num_tokens": 82494511.0,
      "step": 36960
    },
    {
      "entropy": 5.4502537727355955,
      "epoch": 4.151513449766946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033876164743465865,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22427585422992707,
      "num_tokens": 82505982.0,
      "step": 36965
    },
    {
      "entropy": 5.392003870010376,
      "epoch": 4.152075026674903,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003387231435494368,
      "loss": 4.8885,
      "mean_token_accuracy": 0.21868932545185088,
      "num_tokens": 82516682.0,
      "step": 36970
    },
    {
      "entropy": 5.426611137390137,
      "epoch": 4.15263660358286,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033868463763524545,
      "loss": 4.945,
      "mean_token_accuracy": 0.21553872674703597,
      "num_tokens": 82527742.0,
      "step": 36975
    },
    {
      "entropy": 5.423478460311889,
      "epoch": 4.1531981804908185,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033864612969331046,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22626941949129104,
      "num_tokens": 82538130.0,
      "step": 36980
    },
    {
      "entropy": 5.414081287384033,
      "epoch": 4.153759757398776,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003386076197248581,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2240848496556282,
      "num_tokens": 82549695.0,
      "step": 36985
    },
    {
      "entropy": 5.474125099182129,
      "epoch": 4.154321334306733,
      "grad_norm": 1.125,
      "learning_rate": 0.0003385691077311144,
      "loss": 4.9788,
      "mean_token_accuracy": 0.217303329706192,
      "num_tokens": 82560778.0,
      "step": 36990
    },
    {
      "entropy": 5.389772701263428,
      "epoch": 4.154882911214691,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033853059371330565,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22314440906047822,
      "num_tokens": 82572021.0,
      "step": 36995
    },
    {
      "entropy": 5.510976886749267,
      "epoch": 4.155444488122648,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003384920776726582,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22648510038852693,
      "num_tokens": 82583173.0,
      "step": 37000
    },
    {
      "entropy": 5.430298280715943,
      "epoch": 4.156006065030606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033845355961039823,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22898722887039186,
      "num_tokens": 82594279.0,
      "step": 37005
    },
    {
      "entropy": 5.4888568878173825,
      "epoch": 4.1565676419385635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033841503952775255,
      "loss": 5.0236,
      "mean_token_accuracy": 0.22348776012659072,
      "num_tokens": 82605038.0,
      "step": 37010
    },
    {
      "entropy": 5.547130393981933,
      "epoch": 4.157129218846521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033837651742594734,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21970628201961517,
      "num_tokens": 82616627.0,
      "step": 37015
    },
    {
      "entropy": 5.391753101348877,
      "epoch": 4.157690795754479,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033833799330620917,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22154066413640977,
      "num_tokens": 82628057.0,
      "step": 37020
    },
    {
      "entropy": 5.425816440582276,
      "epoch": 4.158252372662436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003382994671697647,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22288416624069213,
      "num_tokens": 82639890.0,
      "step": 37025
    },
    {
      "entropy": 5.3966724395751955,
      "epoch": 4.158813949570393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033826093901784065,
      "loss": 4.8915,
      "mean_token_accuracy": 0.21775314658880235,
      "num_tokens": 82651770.0,
      "step": 37030
    },
    {
      "entropy": 5.432868528366089,
      "epoch": 4.159375526478351,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033822240885166384,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22817357033491134,
      "num_tokens": 82662159.0,
      "step": 37035
    },
    {
      "entropy": 5.438778257369995,
      "epoch": 4.159937103386309,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033818387667246095,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22597167193889617,
      "num_tokens": 82674348.0,
      "step": 37040
    },
    {
      "entropy": 5.437805414199829,
      "epoch": 4.160498680294267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000338145342481459,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2296017125248909,
      "num_tokens": 82685674.0,
      "step": 37045
    },
    {
      "entropy": 5.536522340774536,
      "epoch": 4.161060257202224,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003381068062798848,
      "loss": 5.035,
      "mean_token_accuracy": 0.21014996618032455,
      "num_tokens": 82696992.0,
      "step": 37050
    },
    {
      "entropy": 5.552812385559082,
      "epoch": 4.161621834110181,
      "grad_norm": 1.0,
      "learning_rate": 0.0003380682680689654,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21754339188337327,
      "num_tokens": 82708741.0,
      "step": 37055
    },
    {
      "entropy": 5.440478563308716,
      "epoch": 4.162183411018139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033802972784992784,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22159737050533296,
      "num_tokens": 82720687.0,
      "step": 37060
    },
    {
      "entropy": 5.378065872192383,
      "epoch": 4.1627449879260965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003379911856239994,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22264955937862396,
      "num_tokens": 82731863.0,
      "step": 37065
    },
    {
      "entropy": 5.402432537078857,
      "epoch": 4.163306564834054,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003379526413924071,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22937424927949907,
      "num_tokens": 82742931.0,
      "step": 37070
    },
    {
      "entropy": 5.438032817840576,
      "epoch": 4.163868141742012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003379140951563784,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2204241096973419,
      "num_tokens": 82754244.0,
      "step": 37075
    },
    {
      "entropy": 5.366430044174194,
      "epoch": 4.164429718649969,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033787554691714033,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22152822315692902,
      "num_tokens": 82764929.0,
      "step": 37080
    },
    {
      "entropy": 5.411551380157471,
      "epoch": 4.164991295557926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033783699667592066,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2257144942879677,
      "num_tokens": 82775767.0,
      "step": 37085
    },
    {
      "entropy": 5.387930917739868,
      "epoch": 4.165552872465884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033779844443394656,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22321588546037674,
      "num_tokens": 82786902.0,
      "step": 37090
    },
    {
      "entropy": 5.522421646118164,
      "epoch": 4.166114449373842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033775989019244556,
      "loss": 4.9837,
      "mean_token_accuracy": 0.22013596892356874,
      "num_tokens": 82797677.0,
      "step": 37095
    },
    {
      "entropy": 5.5003267288208,
      "epoch": 4.1666760262818,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033772133395264535,
      "loss": 4.9195,
      "mean_token_accuracy": 0.21941313296556472,
      "num_tokens": 82807246.0,
      "step": 37100
    },
    {
      "entropy": 5.492604207992554,
      "epoch": 4.167237603189757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033768277571577364,
      "loss": 4.954,
      "mean_token_accuracy": 0.2220607429742813,
      "num_tokens": 82818983.0,
      "step": 37105
    },
    {
      "entropy": 5.387652635574341,
      "epoch": 4.167799180097714,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033764421548305795,
      "loss": 4.8887,
      "mean_token_accuracy": 0.23176541328430175,
      "num_tokens": 82829340.0,
      "step": 37110
    },
    {
      "entropy": 5.425860023498535,
      "epoch": 4.168360757005672,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033760565325572614,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21684991121292113,
      "num_tokens": 82840741.0,
      "step": 37115
    },
    {
      "entropy": 5.488474416732788,
      "epoch": 4.168922333913629,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003375670890350061,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2202910378575325,
      "num_tokens": 82851284.0,
      "step": 37120
    },
    {
      "entropy": 5.302756881713867,
      "epoch": 4.169483910821587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033752852282212557,
      "loss": 4.7482,
      "mean_token_accuracy": 0.22917028665542602,
      "num_tokens": 82861856.0,
      "step": 37125
    },
    {
      "entropy": 5.462600660324097,
      "epoch": 4.170045487729545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003374899546183128,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21523133367300035,
      "num_tokens": 82873354.0,
      "step": 37130
    },
    {
      "entropy": 5.432726001739502,
      "epoch": 4.170607064637502,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033745138442479544,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22606950253248215,
      "num_tokens": 82883891.0,
      "step": 37135
    },
    {
      "entropy": 5.480425882339477,
      "epoch": 4.17116864154546,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003374128122428019,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23132419735193252,
      "num_tokens": 82893685.0,
      "step": 37140
    },
    {
      "entropy": 5.429477882385254,
      "epoch": 4.171730218453417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003373742380735602,
      "loss": 4.9357,
      "mean_token_accuracy": 0.23456268906593322,
      "num_tokens": 82904947.0,
      "step": 37145
    },
    {
      "entropy": 5.384034633636475,
      "epoch": 4.1722917953613745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003373356619182985,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21553608477115632,
      "num_tokens": 82915368.0,
      "step": 37150
    },
    {
      "entropy": 5.449951601028443,
      "epoch": 4.172853372269333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003372970837782453,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22097211927175522,
      "num_tokens": 82926594.0,
      "step": 37155
    },
    {
      "entropy": 5.43921217918396,
      "epoch": 4.17341494917729,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003372585036546287,
      "loss": 4.8869,
      "mean_token_accuracy": 0.2281961590051651,
      "num_tokens": 82937699.0,
      "step": 37160
    },
    {
      "entropy": 5.463379669189453,
      "epoch": 4.173976526085247,
      "grad_norm": 1.0,
      "learning_rate": 0.0003372199215486772,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22139305174350737,
      "num_tokens": 82948822.0,
      "step": 37165
    },
    {
      "entropy": 5.418450784683228,
      "epoch": 4.174538102993205,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003371813374616192,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21898352354764938,
      "num_tokens": 82959597.0,
      "step": 37170
    },
    {
      "entropy": 5.434878492355347,
      "epoch": 4.175099679901162,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033714275139468336,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22032560855150224,
      "num_tokens": 82970740.0,
      "step": 37175
    },
    {
      "entropy": 5.515315818786621,
      "epoch": 4.17566125680912,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003371041633490982,
      "loss": 5.032,
      "mean_token_accuracy": 0.2121607780456543,
      "num_tokens": 82982923.0,
      "step": 37180
    },
    {
      "entropy": 5.396599912643433,
      "epoch": 4.176222833717078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003370655733260924,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22347286641597747,
      "num_tokens": 82993271.0,
      "step": 37185
    },
    {
      "entropy": 5.468991422653199,
      "epoch": 4.176784410625035,
      "grad_norm": 1.0,
      "learning_rate": 0.00033702698132689467,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22572373747825622,
      "num_tokens": 83004688.0,
      "step": 37190
    },
    {
      "entropy": 5.409436798095703,
      "epoch": 4.177345987532993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003369883873527338,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21252969056367874,
      "num_tokens": 83016310.0,
      "step": 37195
    },
    {
      "entropy": 5.4954156398773195,
      "epoch": 4.17790756444095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033694979140483865,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2174718350172043,
      "num_tokens": 83027327.0,
      "step": 37200
    },
    {
      "entropy": 5.472800064086914,
      "epoch": 4.1784691413489075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003369111934844381,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21592868417501448,
      "num_tokens": 83038493.0,
      "step": 37205
    },
    {
      "entropy": 5.423410940170288,
      "epoch": 4.179030718256866,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033687259359276115,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2188231974840164,
      "num_tokens": 83050082.0,
      "step": 37210
    },
    {
      "entropy": 5.422526502609253,
      "epoch": 4.179592295164823,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003368339917310368,
      "loss": 4.8608,
      "mean_token_accuracy": 0.230422680079937,
      "num_tokens": 83061234.0,
      "step": 37215
    },
    {
      "entropy": 5.417610454559326,
      "epoch": 4.18015387207278,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003367953879004942,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21833456307649612,
      "num_tokens": 83071454.0,
      "step": 37220
    },
    {
      "entropy": 5.416381692886352,
      "epoch": 4.180715448980738,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003367567821023625,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22102924734354018,
      "num_tokens": 83083204.0,
      "step": 37225
    },
    {
      "entropy": 5.4379395008087155,
      "epoch": 4.181277025888695,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003367181743378708,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22561691254377364,
      "num_tokens": 83094654.0,
      "step": 37230
    },
    {
      "entropy": 5.515936422348022,
      "epoch": 4.181838602796653,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033667956460824866,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21570766121149063,
      "num_tokens": 83105539.0,
      "step": 37235
    },
    {
      "entropy": 5.457009553909302,
      "epoch": 4.182400179704611,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003366409529147251,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2273418724536896,
      "num_tokens": 83115741.0,
      "step": 37240
    },
    {
      "entropy": 5.421052598953247,
      "epoch": 4.182961756612568,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033660233925852977,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2261724665760994,
      "num_tokens": 83126739.0,
      "step": 37245
    },
    {
      "entropy": 5.353790187835694,
      "epoch": 4.183523333520526,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033656372364089206,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22916687428951263,
      "num_tokens": 83138821.0,
      "step": 37250
    },
    {
      "entropy": 5.478930330276489,
      "epoch": 4.184084910428483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033652510606304155,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21627798825502395,
      "num_tokens": 83149757.0,
      "step": 37255
    },
    {
      "entropy": 5.437226295471191,
      "epoch": 4.18464648733644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033648648652620766,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21406979709863663,
      "num_tokens": 83161627.0,
      "step": 37260
    },
    {
      "entropy": 5.4171020030975345,
      "epoch": 4.1852080642443985,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033644786503162036,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22870920449495316,
      "num_tokens": 83172903.0,
      "step": 37265
    },
    {
      "entropy": 5.337483596801758,
      "epoch": 4.185769641152356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003364092415805091,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2290056198835373,
      "num_tokens": 83184142.0,
      "step": 37270
    },
    {
      "entropy": 5.474261522293091,
      "epoch": 4.186331218060313,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033637061617410377,
      "loss": 4.9713,
      "mean_token_accuracy": 0.2182529538869858,
      "num_tokens": 83195206.0,
      "step": 37275
    },
    {
      "entropy": 5.441357231140136,
      "epoch": 4.186892794968271,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003363319888136343,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22907545864582063,
      "num_tokens": 83207579.0,
      "step": 37280
    },
    {
      "entropy": 5.493651962280273,
      "epoch": 4.187454371876228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003362933595003305,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22045973241329192,
      "num_tokens": 83219964.0,
      "step": 37285
    },
    {
      "entropy": 5.414311313629151,
      "epoch": 4.188015948784186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003362547282354223,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22435457706451417,
      "num_tokens": 83231241.0,
      "step": 37290
    },
    {
      "entropy": 5.378537225723266,
      "epoch": 4.188577525692144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003362160950201398,
      "loss": 4.848,
      "mean_token_accuracy": 0.2366323485970497,
      "num_tokens": 83241784.0,
      "step": 37295
    },
    {
      "entropy": 5.369288921356201,
      "epoch": 4.189139102600101,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003361774598557132,
      "loss": 4.8754,
      "mean_token_accuracy": 0.22410111576318742,
      "num_tokens": 83252815.0,
      "step": 37300
    },
    {
      "entropy": 5.400293397903442,
      "epoch": 4.189700679508059,
      "grad_norm": 1.0,
      "learning_rate": 0.00033613882274337244,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22958636283874512,
      "num_tokens": 83263938.0,
      "step": 37305
    },
    {
      "entropy": 5.504263925552368,
      "epoch": 4.190262256416016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003361001836843479,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2190396085381508,
      "num_tokens": 83274746.0,
      "step": 37310
    },
    {
      "entropy": 5.519247198104859,
      "epoch": 4.190823833323973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003360615426798697,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21649404019117355,
      "num_tokens": 83285707.0,
      "step": 37315
    },
    {
      "entropy": 5.500043487548828,
      "epoch": 4.1913854102319315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003360228997311685,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22190604358911514,
      "num_tokens": 83297175.0,
      "step": 37320
    },
    {
      "entropy": 5.421435832977295,
      "epoch": 4.191946987139889,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003359842548394744,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23323349803686141,
      "num_tokens": 83307731.0,
      "step": 37325
    },
    {
      "entropy": 5.4769714832305905,
      "epoch": 4.192508564047847,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003359456080060179,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22383499294519424,
      "num_tokens": 83318698.0,
      "step": 37330
    },
    {
      "entropy": 5.362858390808105,
      "epoch": 4.193070140955804,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003359069592320297,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22599163949489592,
      "num_tokens": 83330085.0,
      "step": 37335
    },
    {
      "entropy": 5.484649562835694,
      "epoch": 4.193631717863761,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003358683085187403,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2220660775899887,
      "num_tokens": 83340606.0,
      "step": 37340
    },
    {
      "entropy": 5.406899642944336,
      "epoch": 4.194193294771719,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033582965586738037,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22515281289815903,
      "num_tokens": 83352177.0,
      "step": 37345
    },
    {
      "entropy": 5.508094453811646,
      "epoch": 4.1947548716796765,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003357910012791804,
      "loss": 5.0446,
      "mean_token_accuracy": 0.214927576482296,
      "num_tokens": 83362474.0,
      "step": 37350
    },
    {
      "entropy": 5.489491701126099,
      "epoch": 4.195316448587634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003357523447553716,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21961501985788345,
      "num_tokens": 83373187.0,
      "step": 37355
    },
    {
      "entropy": 5.441519546508789,
      "epoch": 4.195878025495592,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033571368629718457,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22688770592212676,
      "num_tokens": 83385110.0,
      "step": 37360
    },
    {
      "entropy": 5.446689128875732,
      "epoch": 4.196439602403549,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033567502590585007,
      "loss": 4.9606,
      "mean_token_accuracy": 0.22534715682268142,
      "num_tokens": 83396692.0,
      "step": 37365
    },
    {
      "entropy": 5.426073598861694,
      "epoch": 4.197001179311506,
      "grad_norm": 1.0,
      "learning_rate": 0.0003356363635825994,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2273035779595375,
      "num_tokens": 83408433.0,
      "step": 37370
    },
    {
      "entropy": 5.471146440505981,
      "epoch": 4.197562756219464,
      "grad_norm": 1.125,
      "learning_rate": 0.0003355976993286633,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22567424029111863,
      "num_tokens": 83419169.0,
      "step": 37375
    },
    {
      "entropy": 5.450080442428589,
      "epoch": 4.198124333127422,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003355590331452729,
      "loss": 4.9311,
      "mean_token_accuracy": 0.23245418518781663,
      "num_tokens": 83430247.0,
      "step": 37380
    },
    {
      "entropy": 5.420957136154175,
      "epoch": 4.19868591003538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033552036503365945,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23059021532535554,
      "num_tokens": 83442294.0,
      "step": 37385
    },
    {
      "entropy": 5.393273830413818,
      "epoch": 4.199247486943337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003354816949950541,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22803356051445006,
      "num_tokens": 83453388.0,
      "step": 37390
    },
    {
      "entropy": 5.325131797790528,
      "epoch": 4.199809063851294,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033544302303068814,
      "loss": 4.8908,
      "mean_token_accuracy": 0.23138397336006164,
      "num_tokens": 83465151.0,
      "step": 37395
    },
    {
      "entropy": 5.403776550292969,
      "epoch": 4.200370640759252,
      "grad_norm": 1.0,
      "learning_rate": 0.0003354043491417929,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22193790078163148,
      "num_tokens": 83474731.0,
      "step": 37400
    },
    {
      "entropy": 5.394750261306763,
      "epoch": 4.2009322176672095,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033536567332959966,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22294367253780364,
      "num_tokens": 83486456.0,
      "step": 37405
    },
    {
      "entropy": 5.443454742431641,
      "epoch": 4.201493794575167,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033532699559534,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2281699687242508,
      "num_tokens": 83496503.0,
      "step": 37410
    },
    {
      "entropy": 5.546840524673462,
      "epoch": 4.202055371483125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033528831594024553,
      "loss": 5.037,
      "mean_token_accuracy": 0.21815346628427507,
      "num_tokens": 83508198.0,
      "step": 37415
    },
    {
      "entropy": 5.466953945159912,
      "epoch": 4.202616948391082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033524963436554753,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22318112552165986,
      "num_tokens": 83520255.0,
      "step": 37420
    },
    {
      "entropy": 5.469125986099243,
      "epoch": 4.20317852529904,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033521095087247794,
      "loss": 4.9785,
      "mean_token_accuracy": 0.222655288875103,
      "num_tokens": 83532083.0,
      "step": 37425
    },
    {
      "entropy": 5.498526334762573,
      "epoch": 4.203740102206997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003351722654622683,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2192293092608452,
      "num_tokens": 83542891.0,
      "step": 37430
    },
    {
      "entropy": 5.415678310394287,
      "epoch": 4.204301679114955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003351335781361503,
      "loss": 4.9107,
      "mean_token_accuracy": 0.23000191599130632,
      "num_tokens": 83554238.0,
      "step": 37435
    },
    {
      "entropy": 5.4847665309906,
      "epoch": 4.204863256022913,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033509488889535595,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2173205152153969,
      "num_tokens": 83565681.0,
      "step": 37440
    },
    {
      "entropy": 5.36694450378418,
      "epoch": 4.20542483293087,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033505619774111696,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23285714834928511,
      "num_tokens": 83576844.0,
      "step": 37445
    },
    {
      "entropy": 5.442114973068238,
      "epoch": 4.205986409838827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033501750467466535,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22929039746522903,
      "num_tokens": 83589501.0,
      "step": 37450
    },
    {
      "entropy": 5.45819354057312,
      "epoch": 4.206547986746785,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003349788096972331,
      "loss": 4.969,
      "mean_token_accuracy": 0.22938144207000732,
      "num_tokens": 83600305.0,
      "step": 37455
    },
    {
      "entropy": 5.470835876464844,
      "epoch": 4.207109563654742,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003349401128100523,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21386773586273194,
      "num_tokens": 83611908.0,
      "step": 37460
    },
    {
      "entropy": 5.407688999176026,
      "epoch": 4.2076711405627,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033490141401435516,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22175033241510392,
      "num_tokens": 83624156.0,
      "step": 37465
    },
    {
      "entropy": 5.385529661178589,
      "epoch": 4.208232717470658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003348627133113736,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23159741014242172,
      "num_tokens": 83634655.0,
      "step": 37470
    },
    {
      "entropy": 5.364322376251221,
      "epoch": 4.208794294378615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033482401070234006,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22805216461420058,
      "num_tokens": 83645015.0,
      "step": 37475
    },
    {
      "entropy": 5.415023231506348,
      "epoch": 4.209355871286573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000334785306188487,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22951314896345137,
      "num_tokens": 83656540.0,
      "step": 37480
    },
    {
      "entropy": 5.449527597427368,
      "epoch": 4.20991744819453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003347465997710464,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2307795524597168,
      "num_tokens": 83666798.0,
      "step": 37485
    },
    {
      "entropy": 5.413037538528442,
      "epoch": 4.2104790251024875,
      "grad_norm": 1.0625,
      "learning_rate": 0.000334707891451251,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2313299983739853,
      "num_tokens": 83677058.0,
      "step": 37490
    },
    {
      "entropy": 5.3280517578125,
      "epoch": 4.211040602010446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003346691812303331,
      "loss": 4.8291,
      "mean_token_accuracy": 0.22917085886001587,
      "num_tokens": 83688126.0,
      "step": 37495
    },
    {
      "entropy": 5.460555219650269,
      "epoch": 4.211602178918403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003346304691095254,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2166024327278137,
      "num_tokens": 83700298.0,
      "step": 37500
    },
    {
      "entropy": 5.387555122375488,
      "epoch": 4.21216375582636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003345917550900604,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2303667590022087,
      "num_tokens": 83711037.0,
      "step": 37505
    },
    {
      "entropy": 5.465621376037598,
      "epoch": 4.212725332734318,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033455303917317075,
      "loss": 4.902,
      "mean_token_accuracy": 0.2260332554578781,
      "num_tokens": 83722306.0,
      "step": 37510
    },
    {
      "entropy": 5.471441602706909,
      "epoch": 4.213286909642275,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033451432136008937,
      "loss": 4.953,
      "mean_token_accuracy": 0.22284950613975524,
      "num_tokens": 83733306.0,
      "step": 37515
    },
    {
      "entropy": 5.373544406890869,
      "epoch": 4.2138484865502335,
      "grad_norm": 1.125,
      "learning_rate": 0.0003344756016520488,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22529838234186172,
      "num_tokens": 83745475.0,
      "step": 37520
    },
    {
      "entropy": 5.414590692520141,
      "epoch": 4.214410063458191,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033443688005028213,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2247951790690422,
      "num_tokens": 83756992.0,
      "step": 37525
    },
    {
      "entropy": 5.391875553131103,
      "epoch": 4.214971640366148,
      "grad_norm": 1.03125,
      "learning_rate": 0.000334398156556022,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22451052218675613,
      "num_tokens": 83768881.0,
      "step": 37530
    },
    {
      "entropy": 5.593256425857544,
      "epoch": 4.215533217274106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003343594311705016,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22411651611328126,
      "num_tokens": 83781144.0,
      "step": 37535
    },
    {
      "entropy": 5.5223907947540285,
      "epoch": 4.216094794182063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033432070389495386,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21981520503759383,
      "num_tokens": 83792449.0,
      "step": 37540
    },
    {
      "entropy": 5.381565809249878,
      "epoch": 4.2166563710900204,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003342819747306119,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22999639213085174,
      "num_tokens": 83802889.0,
      "step": 37545
    },
    {
      "entropy": 5.437777471542359,
      "epoch": 4.2172179479979786,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033424324367870896,
      "loss": 4.9871,
      "mean_token_accuracy": 0.21959020346403121,
      "num_tokens": 83813527.0,
      "step": 37550
    },
    {
      "entropy": 5.336101865768432,
      "epoch": 4.217779524905936,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033420451074047807,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2300384074449539,
      "num_tokens": 83824963.0,
      "step": 37555
    },
    {
      "entropy": 5.491553592681885,
      "epoch": 4.218341101813893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003341657759171526,
      "loss": 5.0073,
      "mean_token_accuracy": 0.2162635311484337,
      "num_tokens": 83836188.0,
      "step": 37560
    },
    {
      "entropy": 5.483624696731567,
      "epoch": 4.218902678721851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003341270392099659,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22907852083444596,
      "num_tokens": 83848297.0,
      "step": 37565
    },
    {
      "entropy": 5.401229047775269,
      "epoch": 4.219464255629808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033408830062015135,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22119218707084656,
      "num_tokens": 83859074.0,
      "step": 37570
    },
    {
      "entropy": 5.342350196838379,
      "epoch": 4.220025832537766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033404956014894237,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2255649521946907,
      "num_tokens": 83869921.0,
      "step": 37575
    },
    {
      "entropy": 5.413160276412964,
      "epoch": 4.220587409445724,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033401081779757254,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21876488178968428,
      "num_tokens": 83881994.0,
      "step": 37580
    },
    {
      "entropy": 5.4653526782989506,
      "epoch": 4.221148986353681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003339720735672753,
      "loss": 4.939,
      "mean_token_accuracy": 0.2175651505589485,
      "num_tokens": 83893530.0,
      "step": 37585
    },
    {
      "entropy": 5.43386754989624,
      "epoch": 4.221710563261639,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033393332745928453,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22778006196022033,
      "num_tokens": 83904929.0,
      "step": 37590
    },
    {
      "entropy": 5.4019529819488525,
      "epoch": 4.222272140169596,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033389457947483357,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22496678978204726,
      "num_tokens": 83915888.0,
      "step": 37595
    },
    {
      "entropy": 5.366207885742187,
      "epoch": 4.222833717077553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033385582961515645,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22313133627176285,
      "num_tokens": 83927481.0,
      "step": 37600
    },
    {
      "entropy": 5.474659872055054,
      "epoch": 4.2233952939855115,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033381707788148683,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2229377031326294,
      "num_tokens": 83938644.0,
      "step": 37605
    },
    {
      "entropy": 5.388237047195434,
      "epoch": 4.223956870893469,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033377832427505864,
      "loss": 4.8486,
      "mean_token_accuracy": 0.23696217834949493,
      "num_tokens": 83949165.0,
      "step": 37610
    },
    {
      "entropy": 5.4504640102386475,
      "epoch": 4.224518447801427,
      "grad_norm": 1.0,
      "learning_rate": 0.00033373956879710593,
      "loss": 4.93,
      "mean_token_accuracy": 0.22194320112466812,
      "num_tokens": 83961112.0,
      "step": 37615
    },
    {
      "entropy": 5.455247783660889,
      "epoch": 4.225080024709384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033370081144886246,
      "loss": 4.9393,
      "mean_token_accuracy": 0.23292788714170456,
      "num_tokens": 83972451.0,
      "step": 37620
    },
    {
      "entropy": 5.3283271312713625,
      "epoch": 4.225641601617341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033366205223156237,
      "loss": 4.8314,
      "mean_token_accuracy": 0.2279622286558151,
      "num_tokens": 83984082.0,
      "step": 37625
    },
    {
      "entropy": 5.512401866912842,
      "epoch": 4.226203178525299,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033362329114643987,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21119129806756973,
      "num_tokens": 83995853.0,
      "step": 37630
    },
    {
      "entropy": 5.487409782409668,
      "epoch": 4.226764755433257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033358452819472895,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22000881284475327,
      "num_tokens": 84006440.0,
      "step": 37635
    },
    {
      "entropy": 5.537296962738037,
      "epoch": 4.227326332341214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033354576337766396,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21711939573287964,
      "num_tokens": 84018968.0,
      "step": 37640
    },
    {
      "entropy": 5.446925497055053,
      "epoch": 4.227887909249172,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003335069966964792,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22981393039226533,
      "num_tokens": 84030607.0,
      "step": 37645
    },
    {
      "entropy": 5.416278553009033,
      "epoch": 4.228449486157129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000333468228152409,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22657398879528046,
      "num_tokens": 84042598.0,
      "step": 37650
    },
    {
      "entropy": 5.32530460357666,
      "epoch": 4.229011063065086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033342945774668766,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22861914932727814,
      "num_tokens": 84052708.0,
      "step": 37655
    },
    {
      "entropy": 5.464290285110474,
      "epoch": 4.2295726399730444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033339068548054974,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21959422677755355,
      "num_tokens": 84063573.0,
      "step": 37660
    },
    {
      "entropy": 5.532953500747681,
      "epoch": 4.230134216881002,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033335191135522974,
      "loss": 4.9635,
      "mean_token_accuracy": 0.21979272365570068,
      "num_tokens": 84074132.0,
      "step": 37665
    },
    {
      "entropy": 5.366282987594604,
      "epoch": 4.23069579378896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033331313537196235,
      "loss": 4.8521,
      "mean_token_accuracy": 0.24656212627887725,
      "num_tokens": 84086424.0,
      "step": 37670
    },
    {
      "entropy": 5.467771577835083,
      "epoch": 4.231257370696917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000333274357531982,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22409769296646118,
      "num_tokens": 84098413.0,
      "step": 37675
    },
    {
      "entropy": 5.332815885543823,
      "epoch": 4.231818947604874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003332355778365235,
      "loss": 4.9004,
      "mean_token_accuracy": 0.2231147274374962,
      "num_tokens": 84109156.0,
      "step": 37680
    },
    {
      "entropy": 5.416134643554687,
      "epoch": 4.232380524512832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003331967962868217,
      "loss": 4.8296,
      "mean_token_accuracy": 0.22592823058366776,
      "num_tokens": 84120677.0,
      "step": 37685
    },
    {
      "entropy": 5.43780837059021,
      "epoch": 4.2329421014207895,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003331580128841112,
      "loss": 4.9355,
      "mean_token_accuracy": 0.22210324555635452,
      "num_tokens": 84132607.0,
      "step": 37690
    },
    {
      "entropy": 5.463098526000977,
      "epoch": 4.233503678328747,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033311922762962716,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22402994632720946,
      "num_tokens": 84143958.0,
      "step": 37695
    },
    {
      "entropy": 5.548026514053345,
      "epoch": 4.234065255236705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003330804405246043,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21999145150184632,
      "num_tokens": 84155154.0,
      "step": 37700
    },
    {
      "entropy": 5.4399713516235355,
      "epoch": 4.234626832144662,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003330416515702777,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2179627910256386,
      "num_tokens": 84166100.0,
      "step": 37705
    },
    {
      "entropy": 5.470032215118408,
      "epoch": 4.23518840905262,
      "grad_norm": 1.125,
      "learning_rate": 0.00033300286076788237,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22607191503047944,
      "num_tokens": 84177249.0,
      "step": 37710
    },
    {
      "entropy": 5.509691047668457,
      "epoch": 4.235749985960577,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003329640681186535,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2205980345606804,
      "num_tokens": 84187922.0,
      "step": 37715
    },
    {
      "entropy": 5.487896823883057,
      "epoch": 4.236311562868535,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033292527362382626,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2288982629776001,
      "num_tokens": 84198929.0,
      "step": 37720
    },
    {
      "entropy": 5.380988121032715,
      "epoch": 4.236873139776493,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033288647728463576,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22099193483591079,
      "num_tokens": 84209503.0,
      "step": 37725
    },
    {
      "entropy": 5.437314414978028,
      "epoch": 4.23743471668445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033284767910231745,
      "loss": 4.9982,
      "mean_token_accuracy": 0.21846625953912735,
      "num_tokens": 84220667.0,
      "step": 37730
    },
    {
      "entropy": 5.474794054031372,
      "epoch": 4.237996293592407,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003328088790781065,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22226817011833191,
      "num_tokens": 84232072.0,
      "step": 37735
    },
    {
      "entropy": 5.52205605506897,
      "epoch": 4.238557870500365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033277007721323854,
      "loss": 4.9929,
      "mean_token_accuracy": 0.22187653332948684,
      "num_tokens": 84243765.0,
      "step": 37740
    },
    {
      "entropy": 5.500497198104858,
      "epoch": 4.2391194474083225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033273127350894884,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21909938305616378,
      "num_tokens": 84254869.0,
      "step": 37745
    },
    {
      "entropy": 5.495359992980957,
      "epoch": 4.23968102431628,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033269246796647296,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21224257797002793,
      "num_tokens": 84264992.0,
      "step": 37750
    },
    {
      "entropy": 5.3908332824707035,
      "epoch": 4.240242601224238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003326536605870466,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23232346922159194,
      "num_tokens": 84275648.0,
      "step": 37755
    },
    {
      "entropy": 5.5205153465271,
      "epoch": 4.240804178132195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033261485137190533,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22095723003149031,
      "num_tokens": 84287217.0,
      "step": 37760
    },
    {
      "entropy": 5.461372041702271,
      "epoch": 4.241365755040153,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033257604032228485,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21603074669837952,
      "num_tokens": 84298764.0,
      "step": 37765
    },
    {
      "entropy": 5.508271598815918,
      "epoch": 4.24192733194811,
      "grad_norm": 1.0,
      "learning_rate": 0.00033253722743942086,
      "loss": 4.996,
      "mean_token_accuracy": 0.22075465619564055,
      "num_tokens": 84310016.0,
      "step": 37770
    },
    {
      "entropy": 5.432967805862427,
      "epoch": 4.2424889088560676,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003324984127245492,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22106491327285765,
      "num_tokens": 84321836.0,
      "step": 37775
    },
    {
      "entropy": 5.408997201919556,
      "epoch": 4.243050485764026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003324595961789058,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23002008199691773,
      "num_tokens": 84332434.0,
      "step": 37780
    },
    {
      "entropy": 5.35033483505249,
      "epoch": 4.243612062671983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033242077780372664,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23451683074235916,
      "num_tokens": 84343997.0,
      "step": 37785
    },
    {
      "entropy": 5.388645505905151,
      "epoch": 4.24417363957994,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033238195760024766,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23132121413946152,
      "num_tokens": 84354157.0,
      "step": 37790
    },
    {
      "entropy": 5.364294099807739,
      "epoch": 4.244735216487898,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033234313556970485,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2319982349872589,
      "num_tokens": 84365888.0,
      "step": 37795
    },
    {
      "entropy": 5.46644811630249,
      "epoch": 4.245296793395855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033230431171333435,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22495216876268387,
      "num_tokens": 84376121.0,
      "step": 37800
    },
    {
      "entropy": 5.360619306564331,
      "epoch": 4.2458583703038135,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003322654860323723,
      "loss": 4.7838,
      "mean_token_accuracy": 0.2338026598095894,
      "num_tokens": 84387318.0,
      "step": 37805
    },
    {
      "entropy": 5.4292162418365475,
      "epoch": 4.246419947211771,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033222665852805514,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21384691596031188,
      "num_tokens": 84398952.0,
      "step": 37810
    },
    {
      "entropy": 5.433514595031738,
      "epoch": 4.246981524119728,
      "grad_norm": 1.125,
      "learning_rate": 0.00033218782920161884,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21700209230184556,
      "num_tokens": 84410010.0,
      "step": 37815
    },
    {
      "entropy": 5.460071086883545,
      "epoch": 4.247543101027686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033214899805429984,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2259935572743416,
      "num_tokens": 84421879.0,
      "step": 37820
    },
    {
      "entropy": 5.531859922409057,
      "epoch": 4.248104677935643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033211016508733463,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21367260217666625,
      "num_tokens": 84432811.0,
      "step": 37825
    },
    {
      "entropy": 5.495768737792969,
      "epoch": 4.2486662548436005,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033207133030195964,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2239308699965477,
      "num_tokens": 84444330.0,
      "step": 37830
    },
    {
      "entropy": 5.446014785766602,
      "epoch": 4.249227831751559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003320324936994113,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22818640917539595,
      "num_tokens": 84455827.0,
      "step": 37835
    },
    {
      "entropy": 5.400070238113403,
      "epoch": 4.249789408659516,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033199365528092625,
      "loss": 4.85,
      "mean_token_accuracy": 0.22813124358654022,
      "num_tokens": 84465608.0,
      "step": 37840
    },
    {
      "entropy": 5.415810537338257,
      "epoch": 4.250350985567474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033195481504774113,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22512352466583252,
      "num_tokens": 84476142.0,
      "step": 37845
    },
    {
      "entropy": 5.49345908164978,
      "epoch": 4.250912562475431,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033191597300109263,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22226145416498183,
      "num_tokens": 84486496.0,
      "step": 37850
    },
    {
      "entropy": 5.396511077880859,
      "epoch": 4.251474139383388,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033187712914221744,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22774270325899124,
      "num_tokens": 84497529.0,
      "step": 37855
    },
    {
      "entropy": 5.347085905075073,
      "epoch": 4.2520357162913465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003318382834723524,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22773233354091643,
      "num_tokens": 84508016.0,
      "step": 37860
    },
    {
      "entropy": 5.5221634864807125,
      "epoch": 4.252597293199304,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033179943599273445,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21782655417919158,
      "num_tokens": 84519787.0,
      "step": 37865
    },
    {
      "entropy": 5.421408081054688,
      "epoch": 4.253158870107261,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003317605867046003,
      "loss": 4.8486,
      "mean_token_accuracy": 0.21908371895551682,
      "num_tokens": 84530908.0,
      "step": 37870
    },
    {
      "entropy": 5.577149200439453,
      "epoch": 4.253720447015219,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033172173560918716,
      "loss": 5.0634,
      "mean_token_accuracy": 0.21370524019002915,
      "num_tokens": 84543502.0,
      "step": 37875
    },
    {
      "entropy": 5.392487382888794,
      "epoch": 4.254282023923176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033168288270773197,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22749721705913545,
      "num_tokens": 84554500.0,
      "step": 37880
    },
    {
      "entropy": 5.503260517120362,
      "epoch": 4.254843600831133,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033164402800147177,
      "loss": 5.0852,
      "mean_token_accuracy": 0.214706389605999,
      "num_tokens": 84566704.0,
      "step": 37885
    },
    {
      "entropy": 5.416437005996704,
      "epoch": 4.2554051777390915,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003316051714916436,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2281710371375084,
      "num_tokens": 84578264.0,
      "step": 37890
    },
    {
      "entropy": 5.368616104125977,
      "epoch": 4.255966754647049,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033156631317948505,
      "loss": 4.7182,
      "mean_token_accuracy": 0.2415535643696785,
      "num_tokens": 84588513.0,
      "step": 37895
    },
    {
      "entropy": 5.480933952331543,
      "epoch": 4.256528331555007,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000331527453066233,
      "loss": 4.9157,
      "mean_token_accuracy": 0.23084976971149446,
      "num_tokens": 84600384.0,
      "step": 37900
    },
    {
      "entropy": 5.500741243362427,
      "epoch": 4.257089908462964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000331488591153125,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22318375408649443,
      "num_tokens": 84610827.0,
      "step": 37905
    },
    {
      "entropy": 5.397649049758911,
      "epoch": 4.257651485370921,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003314497274413983,
      "loss": 4.843,
      "mean_token_accuracy": 0.23738518953323365,
      "num_tokens": 84621130.0,
      "step": 37910
    },
    {
      "entropy": 5.454805755615235,
      "epoch": 4.258213062278879,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033141086193229043,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22096305787563325,
      "num_tokens": 84633428.0,
      "step": 37915
    },
    {
      "entropy": 5.4665281772613525,
      "epoch": 4.258774639186837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003313719946270388,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2264236479997635,
      "num_tokens": 84644519.0,
      "step": 37920
    },
    {
      "entropy": 5.490434646606445,
      "epoch": 4.259336216094794,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033133312552688094,
      "loss": 5.0183,
      "mean_token_accuracy": 0.218091781437397,
      "num_tokens": 84654752.0,
      "step": 37925
    },
    {
      "entropy": 5.444280385971069,
      "epoch": 4.259897793002752,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033129425463305465,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22793712615966796,
      "num_tokens": 84666384.0,
      "step": 37930
    },
    {
      "entropy": 5.545771884918213,
      "epoch": 4.260459369910709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003312553819467974,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21969504058361053,
      "num_tokens": 84676630.0,
      "step": 37935
    },
    {
      "entropy": 5.352607107162475,
      "epoch": 4.261020946818666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033121650746934693,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2265059918165207,
      "num_tokens": 84689407.0,
      "step": 37940
    },
    {
      "entropy": 5.408863973617554,
      "epoch": 4.2615825237266245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003311776312019412,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22247940748929979,
      "num_tokens": 84699237.0,
      "step": 37945
    },
    {
      "entropy": 5.464523077011108,
      "epoch": 4.262144100634582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003311387531458177,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2266036257147789,
      "num_tokens": 84710074.0,
      "step": 37950
    },
    {
      "entropy": 5.475925350189209,
      "epoch": 4.26270567754254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003310998733022146,
      "loss": 5.0711,
      "mean_token_accuracy": 0.2087671473622322,
      "num_tokens": 84721315.0,
      "step": 37955
    },
    {
      "entropy": 5.510786247253418,
      "epoch": 4.263267254450497,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003310609916723698,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22837463170289993,
      "num_tokens": 84732654.0,
      "step": 37960
    },
    {
      "entropy": 5.548636245727539,
      "epoch": 4.263828831358454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003310221082575213,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2252579629421234,
      "num_tokens": 84745568.0,
      "step": 37965
    },
    {
      "entropy": 5.512429094314575,
      "epoch": 4.264390408266412,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033098322305890715,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21666004806756972,
      "num_tokens": 84756443.0,
      "step": 37970
    },
    {
      "entropy": 5.485815763473511,
      "epoch": 4.26495198517437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033094433607776545,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22790285646915437,
      "num_tokens": 84766789.0,
      "step": 37975
    },
    {
      "entropy": 5.49920129776001,
      "epoch": 4.265513562082327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033090544731533434,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21634891182184218,
      "num_tokens": 84778186.0,
      "step": 37980
    },
    {
      "entropy": 5.446406126022339,
      "epoch": 4.266075138990285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003308665567728522,
      "loss": 4.9015,
      "mean_token_accuracy": 0.217153163254261,
      "num_tokens": 84788431.0,
      "step": 37985
    },
    {
      "entropy": 5.494599628448486,
      "epoch": 4.266636715898242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033082766445155715,
      "loss": 5.0128,
      "mean_token_accuracy": 0.22146278321743013,
      "num_tokens": 84799881.0,
      "step": 37990
    },
    {
      "entropy": 5.472569561004638,
      "epoch": 4.2671982928062,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033078877035268766,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21977291405200958,
      "num_tokens": 84810710.0,
      "step": 37995
    },
    {
      "entropy": 5.497346973419189,
      "epoch": 4.267759869714157,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033074987447748205,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2296091288328171,
      "num_tokens": 84821932.0,
      "step": 38000
    },
    {
      "entropy": 5.369991779327393,
      "epoch": 4.268321446622115,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033071097682717883,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22482355833053588,
      "num_tokens": 84832662.0,
      "step": 38005
    },
    {
      "entropy": 5.4392030239105225,
      "epoch": 4.268883023530073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003306720774030165,
      "loss": 4.9388,
      "mean_token_accuracy": 0.226839317381382,
      "num_tokens": 84843199.0,
      "step": 38010
    },
    {
      "entropy": 5.406073570251465,
      "epoch": 4.26944460043803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003306331762062336,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22542897313833238,
      "num_tokens": 84853367.0,
      "step": 38015
    },
    {
      "entropy": 5.433145046234131,
      "epoch": 4.270006177345987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003305942732380689,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2288215085864067,
      "num_tokens": 84863527.0,
      "step": 38020
    },
    {
      "entropy": 5.355737543106079,
      "epoch": 4.270567754253945,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003305553684997609,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2251868098974228,
      "num_tokens": 84874929.0,
      "step": 38025
    },
    {
      "entropy": 5.414251327514648,
      "epoch": 4.2711293311619025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033051646199254843,
      "loss": 4.881,
      "mean_token_accuracy": 0.22541238814592363,
      "num_tokens": 84885182.0,
      "step": 38030
    },
    {
      "entropy": 5.485614013671875,
      "epoch": 4.271690908069861,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003304775537176702,
      "loss": 4.9481,
      "mean_token_accuracy": 0.21802280396223067,
      "num_tokens": 84896115.0,
      "step": 38035
    },
    {
      "entropy": 5.4502626895904545,
      "epoch": 4.272252484977818,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003304386436763652,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21900599747896193,
      "num_tokens": 84908174.0,
      "step": 38040
    },
    {
      "entropy": 5.4910022735595705,
      "epoch": 4.272814061885775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003303997318698723,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21615429520606994,
      "num_tokens": 84920893.0,
      "step": 38045
    },
    {
      "entropy": 5.47519211769104,
      "epoch": 4.273375638793733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003303608182994304,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22305091470479965,
      "num_tokens": 84931860.0,
      "step": 38050
    },
    {
      "entropy": 5.378074359893799,
      "epoch": 4.27393721570169,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003303219029662787,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2376406118273735,
      "num_tokens": 84942533.0,
      "step": 38055
    },
    {
      "entropy": 5.322369050979614,
      "epoch": 4.274498792609648,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033028298587165606,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23040280044078826,
      "num_tokens": 84954018.0,
      "step": 38060
    },
    {
      "entropy": 5.501421785354614,
      "epoch": 4.275060369517606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003302440670168017,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21728321462869643,
      "num_tokens": 84966150.0,
      "step": 38065
    },
    {
      "entropy": 5.498354196548462,
      "epoch": 4.275621946425563,
      "grad_norm": 1.0,
      "learning_rate": 0.0003302051464029548,
      "loss": 4.9287,
      "mean_token_accuracy": 0.21917082816362382,
      "num_tokens": 84977668.0,
      "step": 38070
    },
    {
      "entropy": 5.409559869766236,
      "epoch": 4.27618352333352,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033016622403135473,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22508633136749268,
      "num_tokens": 84988735.0,
      "step": 38075
    },
    {
      "entropy": 5.407833814620972,
      "epoch": 4.276745100241478,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033012729990324063,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2246340721845627,
      "num_tokens": 84999573.0,
      "step": 38080
    },
    {
      "entropy": 5.366381406784058,
      "epoch": 4.2773066771494355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003300883740198519,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22258910685777664,
      "num_tokens": 85010582.0,
      "step": 38085
    },
    {
      "entropy": 5.4105157375335695,
      "epoch": 4.277868254057394,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033004944638242805,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23491198867559432,
      "num_tokens": 85021563.0,
      "step": 38090
    },
    {
      "entropy": 5.550600576400757,
      "epoch": 4.278429830965351,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003300105169922084,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20780457854270934,
      "num_tokens": 85033231.0,
      "step": 38095
    },
    {
      "entropy": 5.38087010383606,
      "epoch": 4.278991407873308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003299715858504326,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2309536650776863,
      "num_tokens": 85044110.0,
      "step": 38100
    },
    {
      "entropy": 5.405970525741577,
      "epoch": 4.279552984781266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032993265295834013,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22551467567682265,
      "num_tokens": 85055425.0,
      "step": 38105
    },
    {
      "entropy": 5.496796178817749,
      "epoch": 4.280114561689223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003298937183171707,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22552022486925125,
      "num_tokens": 85067358.0,
      "step": 38110
    },
    {
      "entropy": 5.519613409042359,
      "epoch": 4.2806761385971805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000329854781928164,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22503392845392228,
      "num_tokens": 85078967.0,
      "step": 38115
    },
    {
      "entropy": 5.41651291847229,
      "epoch": 4.281237715505139,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003298158437925598,
      "loss": 4.8482,
      "mean_token_accuracy": 0.2263336256146431,
      "num_tokens": 85088515.0,
      "step": 38120
    },
    {
      "entropy": 5.372963237762451,
      "epoch": 4.281799292413096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003297769039115978,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22736939042806625,
      "num_tokens": 85099535.0,
      "step": 38125
    },
    {
      "entropy": 5.559076118469238,
      "epoch": 4.282360869321053,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032973796228651797,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20559803098440171,
      "num_tokens": 85110818.0,
      "step": 38130
    },
    {
      "entropy": 5.469377326965332,
      "epoch": 4.282922446229011,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032969901891856027,
      "loss": 4.9436,
      "mean_token_accuracy": 0.21664611846208573,
      "num_tokens": 85120362.0,
      "step": 38135
    },
    {
      "entropy": 5.4390068531036375,
      "epoch": 4.283484023136968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032966007380896443,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22675210535526275,
      "num_tokens": 85131358.0,
      "step": 38140
    },
    {
      "entropy": 5.478372383117676,
      "epoch": 4.2840456000449265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003296211269589708,
      "loss": 4.933,
      "mean_token_accuracy": 0.2261539801955223,
      "num_tokens": 85143395.0,
      "step": 38145
    },
    {
      "entropy": 5.40168867111206,
      "epoch": 4.284607176952884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003295821783698192,
      "loss": 4.8813,
      "mean_token_accuracy": 0.23238845616579057,
      "num_tokens": 85155654.0,
      "step": 38150
    },
    {
      "entropy": 5.419315338134766,
      "epoch": 4.285168753860841,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032954322804274995,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21349204927682877,
      "num_tokens": 85168285.0,
      "step": 38155
    },
    {
      "entropy": 5.489283752441406,
      "epoch": 4.285730330768799,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032950427597900315,
      "loss": 5.011,
      "mean_token_accuracy": 0.2174928978085518,
      "num_tokens": 85180058.0,
      "step": 38160
    },
    {
      "entropy": 5.449603796005249,
      "epoch": 4.286291907676756,
      "grad_norm": 1.03125,
      "learning_rate": 0.000329465322179819,
      "loss": 4.9861,
      "mean_token_accuracy": 0.22005026638507844,
      "num_tokens": 85190990.0,
      "step": 38165
    },
    {
      "entropy": 5.553635215759277,
      "epoch": 4.2868534845847135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032942636664643794,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2179936036467552,
      "num_tokens": 85201136.0,
      "step": 38170
    },
    {
      "entropy": 5.4730278015136715,
      "epoch": 4.287415061492672,
      "grad_norm": 1.125,
      "learning_rate": 0.00032938740938010024,
      "loss": 4.992,
      "mean_token_accuracy": 0.22200053334236144,
      "num_tokens": 85212424.0,
      "step": 38175
    },
    {
      "entropy": 5.517501783370972,
      "epoch": 4.287976638400629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003293484503820463,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22064849436283113,
      "num_tokens": 85223767.0,
      "step": 38180
    },
    {
      "entropy": 5.541213035583496,
      "epoch": 4.288538215308587,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032930948965351664,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22231877595186234,
      "num_tokens": 85234374.0,
      "step": 38185
    },
    {
      "entropy": 5.426975727081299,
      "epoch": 4.289099792216544,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003292705271957518,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22683661431074142,
      "num_tokens": 85245322.0,
      "step": 38190
    },
    {
      "entropy": 5.559979248046875,
      "epoch": 4.289661369124501,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032923156300999223,
      "loss": 5.0372,
      "mean_token_accuracy": 0.2157924011349678,
      "num_tokens": 85257114.0,
      "step": 38195
    },
    {
      "entropy": 5.464731979370117,
      "epoch": 4.2902229460324595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032919259709747874,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2267296642065048,
      "num_tokens": 85267564.0,
      "step": 38200
    },
    {
      "entropy": 5.426394271850586,
      "epoch": 4.290784522940417,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032915362945945195,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22811576426029206,
      "num_tokens": 85277763.0,
      "step": 38205
    },
    {
      "entropy": 5.522031545639038,
      "epoch": 4.291346099848374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032911466009715254,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21940020173788072,
      "num_tokens": 85288504.0,
      "step": 38210
    },
    {
      "entropy": 5.4985973834991455,
      "epoch": 4.291907676756332,
      "grad_norm": 1.0,
      "learning_rate": 0.00032907568901182136,
      "loss": 4.9349,
      "mean_token_accuracy": 0.230858913064003,
      "num_tokens": 85299928.0,
      "step": 38215
    },
    {
      "entropy": 5.426014232635498,
      "epoch": 4.292469253664289,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003290367162046993,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22778550833463668,
      "num_tokens": 85311052.0,
      "step": 38220
    },
    {
      "entropy": 5.4412778377532955,
      "epoch": 4.293030830572247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032899774167702723,
      "loss": 4.983,
      "mean_token_accuracy": 0.2215246304869652,
      "num_tokens": 85322023.0,
      "step": 38225
    },
    {
      "entropy": 5.404849672317505,
      "epoch": 4.2935924074802045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003289587654300461,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22352275401353836,
      "num_tokens": 85334160.0,
      "step": 38230
    },
    {
      "entropy": 5.534461307525635,
      "epoch": 4.294153984388162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000328919787464997,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21469973772764206,
      "num_tokens": 85345378.0,
      "step": 38235
    },
    {
      "entropy": 5.494585037231445,
      "epoch": 4.29471556129612,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000328880807783121,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22287463992834092,
      "num_tokens": 85357610.0,
      "step": 38240
    },
    {
      "entropy": 5.497124242782593,
      "epoch": 4.295277138204077,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032884182638565914,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22046888023614883,
      "num_tokens": 85368749.0,
      "step": 38245
    },
    {
      "entropy": 5.4447413921356205,
      "epoch": 4.295838715112034,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032880284327385255,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2216956779360771,
      "num_tokens": 85379488.0,
      "step": 38250
    },
    {
      "entropy": 5.418927192687988,
      "epoch": 4.296400292019992,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003287638584489426,
      "loss": 4.914,
      "mean_token_accuracy": 0.22616717517375945,
      "num_tokens": 85390581.0,
      "step": 38255
    },
    {
      "entropy": 5.412367630004883,
      "epoch": 4.29696186892795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003287248719121706,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22052263468503952,
      "num_tokens": 85402430.0,
      "step": 38260
    },
    {
      "entropy": 5.419243478775025,
      "epoch": 4.297523445835907,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032868588366477785,
      "loss": 4.8187,
      "mean_token_accuracy": 0.2290537416934967,
      "num_tokens": 85414932.0,
      "step": 38265
    },
    {
      "entropy": 5.436607313156128,
      "epoch": 4.298085022743865,
      "grad_norm": 1.125,
      "learning_rate": 0.00032864689370800564,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23583237528800965,
      "num_tokens": 85426484.0,
      "step": 38270
    },
    {
      "entropy": 5.381043004989624,
      "epoch": 4.298646599651822,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003286079020430957,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22095342874526977,
      "num_tokens": 85436864.0,
      "step": 38275
    },
    {
      "entropy": 5.372885990142822,
      "epoch": 4.29920817655978,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032856890867128926,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23463254421949387,
      "num_tokens": 85446303.0,
      "step": 38280
    },
    {
      "entropy": 5.437014961242676,
      "epoch": 4.2997697534677375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003285299135938279,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22899353504180908,
      "num_tokens": 85457218.0,
      "step": 38285
    },
    {
      "entropy": 5.4680122375488285,
      "epoch": 4.300331330375695,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032849091681195347,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22404097467660905,
      "num_tokens": 85469085.0,
      "step": 38290
    },
    {
      "entropy": 5.411242580413818,
      "epoch": 4.300892907283653,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032845191832690747,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22600938826799394,
      "num_tokens": 85478966.0,
      "step": 38295
    },
    {
      "entropy": 5.40232892036438,
      "epoch": 4.30145448419161,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003284129181399317,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23662393391132355,
      "num_tokens": 85489742.0,
      "step": 38300
    },
    {
      "entropy": 5.385412931442261,
      "epoch": 4.302016061099567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003283739162522678,
      "loss": 4.8144,
      "mean_token_accuracy": 0.24473751932382584,
      "num_tokens": 85500766.0,
      "step": 38305
    },
    {
      "entropy": 5.42709436416626,
      "epoch": 4.302577638007525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003283349126651578,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22144956290721893,
      "num_tokens": 85512206.0,
      "step": 38310
    },
    {
      "entropy": 5.417858648300171,
      "epoch": 4.303139214915483,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032829590737984345,
      "loss": 4.9291,
      "mean_token_accuracy": 0.226166433095932,
      "num_tokens": 85523881.0,
      "step": 38315
    },
    {
      "entropy": 5.348527479171753,
      "epoch": 4.30370079182344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032825690039756675,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2247912496328354,
      "num_tokens": 85535130.0,
      "step": 38320
    },
    {
      "entropy": 5.485222339630127,
      "epoch": 4.304262368731398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032821789171956966,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22234477549791337,
      "num_tokens": 85547261.0,
      "step": 38325
    },
    {
      "entropy": 5.518471765518188,
      "epoch": 4.304823945639355,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003281788813470943,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22177961319684983,
      "num_tokens": 85558300.0,
      "step": 38330
    },
    {
      "entropy": 5.424336004257202,
      "epoch": 4.305385522547313,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003281398692813827,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2255878120660782,
      "num_tokens": 85569370.0,
      "step": 38335
    },
    {
      "entropy": 5.399345588684082,
      "epoch": 4.30594709945527,
      "grad_norm": 1.0625,
      "learning_rate": 0.000328100855523677,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2334865853190422,
      "num_tokens": 85579527.0,
      "step": 38340
    },
    {
      "entropy": 5.440371704101563,
      "epoch": 4.306508676363228,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032806184007521955,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22170567661523818,
      "num_tokens": 85590791.0,
      "step": 38345
    },
    {
      "entropy": 5.364723968505859,
      "epoch": 4.307070253271186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003280228229372525,
      "loss": 4.837,
      "mean_token_accuracy": 0.22879211157560347,
      "num_tokens": 85601210.0,
      "step": 38350
    },
    {
      "entropy": 5.461073350906372,
      "epoch": 4.307631830179143,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003279838041110182,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22486176043748857,
      "num_tokens": 85611811.0,
      "step": 38355
    },
    {
      "entropy": 5.485902929306031,
      "epoch": 4.3081934070871,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032794478359775904,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21965544521808625,
      "num_tokens": 85622046.0,
      "step": 38360
    },
    {
      "entropy": 5.473137378692627,
      "epoch": 4.308754983995058,
      "grad_norm": 1.0,
      "learning_rate": 0.00032790576139871743,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22786286622285842,
      "num_tokens": 85633094.0,
      "step": 38365
    },
    {
      "entropy": 5.383824300765991,
      "epoch": 4.3093165609030155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003278667375151359,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22005852460861205,
      "num_tokens": 85644224.0,
      "step": 38370
    },
    {
      "entropy": 5.451056861877442,
      "epoch": 4.309878137810974,
      "grad_norm": 1.125,
      "learning_rate": 0.00032782771194825685,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2197220355272293,
      "num_tokens": 85654950.0,
      "step": 38375
    },
    {
      "entropy": 5.466577005386353,
      "epoch": 4.310439714718931,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032778868469932304,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2253563031554222,
      "num_tokens": 85665250.0,
      "step": 38380
    },
    {
      "entropy": 5.538761186599731,
      "epoch": 4.311001291626888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032774965576957705,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22455768138170243,
      "num_tokens": 85676856.0,
      "step": 38385
    },
    {
      "entropy": 5.645647382736206,
      "epoch": 4.311562868534846,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003277106251602615,
      "loss": 5.1177,
      "mean_token_accuracy": 0.21346690207719804,
      "num_tokens": 85689733.0,
      "step": 38390
    },
    {
      "entropy": 5.455279111862183,
      "epoch": 4.312124445442803,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032767159287261914,
      "loss": 4.8733,
      "mean_token_accuracy": 0.229646198451519,
      "num_tokens": 85699678.0,
      "step": 38395
    },
    {
      "entropy": 5.4969861030578615,
      "epoch": 4.312686022350761,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000327632558907893,
      "loss": 5.067,
      "mean_token_accuracy": 0.21240358650684357,
      "num_tokens": 85712130.0,
      "step": 38400
    },
    {
      "entropy": 5.42769570350647,
      "epoch": 4.313247599258719,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032759352326732573,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23107061684131622,
      "num_tokens": 85723025.0,
      "step": 38405
    },
    {
      "entropy": 5.401553964614868,
      "epoch": 4.313809176166676,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032755448595216027,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2277603879570961,
      "num_tokens": 85733551.0,
      "step": 38410
    },
    {
      "entropy": 5.492598295211792,
      "epoch": 4.314370753074634,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032751544696363945,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21583308577537536,
      "num_tokens": 85745633.0,
      "step": 38415
    },
    {
      "entropy": 5.447570323944092,
      "epoch": 4.314932329982591,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032747640630300664,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22652867287397385,
      "num_tokens": 85757522.0,
      "step": 38420
    },
    {
      "entropy": 5.521631908416748,
      "epoch": 4.3154939068905485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032743736397150455,
      "loss": 4.934,
      "mean_token_accuracy": 0.2276088312268257,
      "num_tokens": 85768666.0,
      "step": 38425
    },
    {
      "entropy": 5.402419137954712,
      "epoch": 4.316055483798507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003273983199703765,
      "loss": 4.8711,
      "mean_token_accuracy": 0.22878836542367936,
      "num_tokens": 85780527.0,
      "step": 38430
    },
    {
      "entropy": 5.421187448501587,
      "epoch": 4.316617060706464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003273592743008656,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21977301836013793,
      "num_tokens": 85792166.0,
      "step": 38435
    },
    {
      "entropy": 5.523520469665527,
      "epoch": 4.317178637614421,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032732022696421514,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21545934826135635,
      "num_tokens": 85803706.0,
      "step": 38440
    },
    {
      "entropy": 5.494450092315674,
      "epoch": 4.317740214522379,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003272811779616684,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2249071031808853,
      "num_tokens": 85815515.0,
      "step": 38445
    },
    {
      "entropy": 5.507548379898071,
      "epoch": 4.318301791430336,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032724212729446864,
      "loss": 4.8795,
      "mean_token_accuracy": 0.230177041888237,
      "num_tokens": 85826359.0,
      "step": 38450
    },
    {
      "entropy": 5.464695072174072,
      "epoch": 4.3188633683382935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003272030749638593,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22342449724674224,
      "num_tokens": 85837977.0,
      "step": 38455
    },
    {
      "entropy": 5.484880876541138,
      "epoch": 4.319424945246252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003271640209710837,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21861580312252044,
      "num_tokens": 85850356.0,
      "step": 38460
    },
    {
      "entropy": 5.437144565582275,
      "epoch": 4.319986522154209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003271249653173855,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22120324969291688,
      "num_tokens": 85863101.0,
      "step": 38465
    },
    {
      "entropy": 5.448912906646728,
      "epoch": 4.320548099062167,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032708590800400826,
      "loss": 4.921,
      "mean_token_accuracy": 0.2203420490026474,
      "num_tokens": 85873374.0,
      "step": 38470
    },
    {
      "entropy": 5.5455645561218265,
      "epoch": 4.321109675970124,
      "grad_norm": 1.0,
      "learning_rate": 0.0003270468490321955,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2155260145664215,
      "num_tokens": 85886167.0,
      "step": 38475
    },
    {
      "entropy": 5.480037546157837,
      "epoch": 4.321671252878081,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032700778840319086,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22502726763486863,
      "num_tokens": 85897040.0,
      "step": 38480
    },
    {
      "entropy": 5.425618839263916,
      "epoch": 4.3222328297860395,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000326968726118238,
      "loss": 4.898,
      "mean_token_accuracy": 0.22807476818561553,
      "num_tokens": 85908386.0,
      "step": 38485
    },
    {
      "entropy": 5.370502376556397,
      "epoch": 4.322794406693997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032692966217858074,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23682193011045455,
      "num_tokens": 85920313.0,
      "step": 38490
    },
    {
      "entropy": 5.431816816329956,
      "epoch": 4.323355983601954,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003268905965854629,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2213150829076767,
      "num_tokens": 85931273.0,
      "step": 38495
    },
    {
      "entropy": 5.5107818126678465,
      "epoch": 4.323917560509912,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032685152934012844,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21729506105184554,
      "num_tokens": 85942952.0,
      "step": 38500
    },
    {
      "entropy": 5.417878293991089,
      "epoch": 4.324479137417869,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003268124604438211,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22418098151683807,
      "num_tokens": 85953331.0,
      "step": 38505
    },
    {
      "entropy": 5.455777835845947,
      "epoch": 4.3250407143258265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032677338989778494,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22682186514139174,
      "num_tokens": 85965142.0,
      "step": 38510
    },
    {
      "entropy": 5.460382699966431,
      "epoch": 4.325602291233785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000326734317703264,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22598697245121002,
      "num_tokens": 85976258.0,
      "step": 38515
    },
    {
      "entropy": 5.462629270553589,
      "epoch": 4.326163868141742,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003266952438615023,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2273354321718216,
      "num_tokens": 85987976.0,
      "step": 38520
    },
    {
      "entropy": 5.390769577026367,
      "epoch": 4.3267254450497,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032665616837374397,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23044266402721406,
      "num_tokens": 85999283.0,
      "step": 38525
    },
    {
      "entropy": 5.370438575744629,
      "epoch": 4.327287021957657,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003266170912412332,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23099258691072463,
      "num_tokens": 86010484.0,
      "step": 38530
    },
    {
      "entropy": 5.494019317626953,
      "epoch": 4.327848598865614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003265780124652143,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22222049832344054,
      "num_tokens": 86021582.0,
      "step": 38535
    },
    {
      "entropy": 5.423544311523438,
      "epoch": 4.3284101757735725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032653893204693136,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21784015893936157,
      "num_tokens": 86033847.0,
      "step": 38540
    },
    {
      "entropy": 5.456597995758057,
      "epoch": 4.32897175268153,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003264998499876289,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2224305287003517,
      "num_tokens": 86044870.0,
      "step": 38545
    },
    {
      "entropy": 5.385165405273438,
      "epoch": 4.329533329589487,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003264607662885512,
      "loss": 4.8031,
      "mean_token_accuracy": 0.23015362173318862,
      "num_tokens": 86055807.0,
      "step": 38550
    },
    {
      "entropy": 5.4418017864227295,
      "epoch": 4.330094906497445,
      "grad_norm": 0.90625,
      "learning_rate": 0.00032642168095094274,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2189194768667221,
      "num_tokens": 86066869.0,
      "step": 38555
    },
    {
      "entropy": 5.535571813583374,
      "epoch": 4.330656483405402,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003263825939760481,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21412504017353057,
      "num_tokens": 86078870.0,
      "step": 38560
    },
    {
      "entropy": 5.42948055267334,
      "epoch": 4.33121806031336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032634350536511167,
      "loss": 4.9638,
      "mean_token_accuracy": 0.218853859603405,
      "num_tokens": 86090940.0,
      "step": 38565
    },
    {
      "entropy": 5.392476749420166,
      "epoch": 4.3317796372213175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032630441511937813,
      "loss": 4.85,
      "mean_token_accuracy": 0.22865450829267503,
      "num_tokens": 86101171.0,
      "step": 38570
    },
    {
      "entropy": 5.407442617416382,
      "epoch": 4.332341214129275,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032626532324009204,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21769392788410186,
      "num_tokens": 86111614.0,
      "step": 38575
    },
    {
      "entropy": 5.421198320388794,
      "epoch": 4.332902791037233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003262262297284983,
      "loss": 4.9369,
      "mean_token_accuracy": 0.23260684609413146,
      "num_tokens": 86124131.0,
      "step": 38580
    },
    {
      "entropy": 5.4651830196380615,
      "epoch": 4.33346436794519,
      "grad_norm": 1.125,
      "learning_rate": 0.00032618713458584147,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22437238395214082,
      "num_tokens": 86135584.0,
      "step": 38585
    },
    {
      "entropy": 5.432363414764405,
      "epoch": 4.334025944853147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003261480378133665,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23605918884277344,
      "num_tokens": 86146163.0,
      "step": 38590
    },
    {
      "entropy": 5.4625650405883786,
      "epoch": 4.334587521761105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003261089394123181,
      "loss": 5.002,
      "mean_token_accuracy": 0.21865949183702468,
      "num_tokens": 86157199.0,
      "step": 38595
    },
    {
      "entropy": 5.425432395935059,
      "epoch": 4.335149098669063,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003260698393839413,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22506657391786575,
      "num_tokens": 86167523.0,
      "step": 38600
    },
    {
      "entropy": 5.440423107147216,
      "epoch": 4.335710675577021,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032603073772948096,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2285637378692627,
      "num_tokens": 86179341.0,
      "step": 38605
    },
    {
      "entropy": 5.439254951477051,
      "epoch": 4.336272252484978,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032599163445018204,
      "loss": 4.93,
      "mean_token_accuracy": 0.22704899907112122,
      "num_tokens": 86190154.0,
      "step": 38610
    },
    {
      "entropy": 5.403077888488769,
      "epoch": 4.336833829392935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003259525295472899,
      "loss": 4.9182,
      "mean_token_accuracy": 0.2250693693757057,
      "num_tokens": 86201598.0,
      "step": 38615
    },
    {
      "entropy": 5.447387409210205,
      "epoch": 4.337395406300893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032591342302204935,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22369637936353684,
      "num_tokens": 86212033.0,
      "step": 38620
    },
    {
      "entropy": 5.400721406936645,
      "epoch": 4.3379569832088505,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003258743148757057,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2194783091545105,
      "num_tokens": 86222931.0,
      "step": 38625
    },
    {
      "entropy": 5.412833642959595,
      "epoch": 4.338518560116808,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032583520510950405,
      "loss": 4.8841,
      "mean_token_accuracy": 0.222505384683609,
      "num_tokens": 86234022.0,
      "step": 38630
    },
    {
      "entropy": 5.388411617279052,
      "epoch": 4.339080137024766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032579609372468984,
      "loss": 4.8645,
      "mean_token_accuracy": 0.22592876553535463,
      "num_tokens": 86244267.0,
      "step": 38635
    },
    {
      "entropy": 5.431416368484497,
      "epoch": 4.339641713932723,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032575698072250827,
      "loss": 4.879,
      "mean_token_accuracy": 0.23610179275274276,
      "num_tokens": 86254200.0,
      "step": 38640
    },
    {
      "entropy": 5.456893539428711,
      "epoch": 4.34020329084068,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003257178661042047,
      "loss": 4.9868,
      "mean_token_accuracy": 0.21576672345399855,
      "num_tokens": 86266523.0,
      "step": 38645
    },
    {
      "entropy": 5.503312969207764,
      "epoch": 4.340764867748638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032567874987102463,
      "loss": 4.967,
      "mean_token_accuracy": 0.21966600567102432,
      "num_tokens": 86276949.0,
      "step": 38650
    },
    {
      "entropy": 5.473315572738647,
      "epoch": 4.341326444656596,
      "grad_norm": 1.0,
      "learning_rate": 0.0003256396320242136,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22202335596084594,
      "num_tokens": 86287412.0,
      "step": 38655
    },
    {
      "entropy": 5.461539363861084,
      "epoch": 4.341888021564554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000325600512565017,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22781429141759874,
      "num_tokens": 86298540.0,
      "step": 38660
    },
    {
      "entropy": 5.358958435058594,
      "epoch": 4.342449598472511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032556139149468046,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23377001285552979,
      "num_tokens": 86309351.0,
      "step": 38665
    },
    {
      "entropy": 5.526310539245605,
      "epoch": 4.343011175380468,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032552226881444956,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2212403580546379,
      "num_tokens": 86320873.0,
      "step": 38670
    },
    {
      "entropy": 5.5431581974029545,
      "epoch": 4.343572752288426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032548314452557006,
      "loss": 5.0176,
      "mean_token_accuracy": 0.2143500417470932,
      "num_tokens": 86332889.0,
      "step": 38675
    },
    {
      "entropy": 5.393198919296265,
      "epoch": 4.344134329196383,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032544401862928766,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23059802502393723,
      "num_tokens": 86344155.0,
      "step": 38680
    },
    {
      "entropy": 5.380664396286011,
      "epoch": 4.344695906104341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032540489112684804,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22890018969774245,
      "num_tokens": 86355718.0,
      "step": 38685
    },
    {
      "entropy": 5.500594758987427,
      "epoch": 4.345257483012299,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032536576201949724,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2129070997238159,
      "num_tokens": 86367174.0,
      "step": 38690
    },
    {
      "entropy": 5.487231922149658,
      "epoch": 4.345819059920256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003253266313084811,
      "loss": 5.0,
      "mean_token_accuracy": 0.21662745177745818,
      "num_tokens": 86379086.0,
      "step": 38695
    },
    {
      "entropy": 5.5506911277771,
      "epoch": 4.346380636828213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003252874989950454,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22297999113798142,
      "num_tokens": 86390113.0,
      "step": 38700
    },
    {
      "entropy": 5.537986612319946,
      "epoch": 4.346942213736171,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032524836508043625,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21612437814474106,
      "num_tokens": 86400824.0,
      "step": 38705
    },
    {
      "entropy": 5.438606071472168,
      "epoch": 4.3475037906441285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032520922956589973,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22785817235708236,
      "num_tokens": 86412972.0,
      "step": 38710
    },
    {
      "entropy": 5.4909200191497805,
      "epoch": 4.348065367552087,
      "grad_norm": 1.0,
      "learning_rate": 0.00032517009245268175,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22334469109773636,
      "num_tokens": 86425332.0,
      "step": 38715
    },
    {
      "entropy": 5.43266544342041,
      "epoch": 4.348626944460044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032513095374202863,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23254610747098922,
      "num_tokens": 86435615.0,
      "step": 38720
    },
    {
      "entropy": 5.523071336746216,
      "epoch": 4.349188521368001,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032509181343518646,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2266810342669487,
      "num_tokens": 86446875.0,
      "step": 38725
    },
    {
      "entropy": 5.454200506210327,
      "epoch": 4.349750098275959,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032505267153340163,
      "loss": 4.9174,
      "mean_token_accuracy": 0.23178026378154754,
      "num_tokens": 86458284.0,
      "step": 38730
    },
    {
      "entropy": 5.535055160522461,
      "epoch": 4.350311675183916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003250135280379202,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20970666259527207,
      "num_tokens": 86470977.0,
      "step": 38735
    },
    {
      "entropy": 5.492198896408081,
      "epoch": 4.350873252091874,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003249743829499888,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21859696060419082,
      "num_tokens": 86482578.0,
      "step": 38740
    },
    {
      "entropy": 5.456911277770996,
      "epoch": 4.351434828999832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003249352362708535,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2263573169708252,
      "num_tokens": 86493123.0,
      "step": 38745
    },
    {
      "entropy": 5.579812669754029,
      "epoch": 4.351996405907789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000324896088001761,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21822014302015305,
      "num_tokens": 86503865.0,
      "step": 38750
    },
    {
      "entropy": 5.43505892753601,
      "epoch": 4.352557982815747,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003248569381439577,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2250353753566742,
      "num_tokens": 86515006.0,
      "step": 38755
    },
    {
      "entropy": 5.409599733352661,
      "epoch": 4.353119559723704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003248177866986901,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2220209851861,
      "num_tokens": 86526256.0,
      "step": 38760
    },
    {
      "entropy": 5.428135967254638,
      "epoch": 4.3536811366316615,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003247786336672049,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2312880665063858,
      "num_tokens": 86537050.0,
      "step": 38765
    },
    {
      "entropy": 5.415249300003052,
      "epoch": 4.35424271353962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032473947905074866,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23092418313026428,
      "num_tokens": 86547678.0,
      "step": 38770
    },
    {
      "entropy": 5.436520051956177,
      "epoch": 4.354804290447577,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032470032285056814,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2342531353235245,
      "num_tokens": 86558747.0,
      "step": 38775
    },
    {
      "entropy": 5.43987717628479,
      "epoch": 4.355365867355534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032466116506791004,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22307591289281845,
      "num_tokens": 86569500.0,
      "step": 38780
    },
    {
      "entropy": 5.351621913909912,
      "epoch": 4.355927444263492,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032462200570402124,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22135955691337586,
      "num_tokens": 86581312.0,
      "step": 38785
    },
    {
      "entropy": 5.417032766342163,
      "epoch": 4.356489021171449,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003245828447601484,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2229658767580986,
      "num_tokens": 86592526.0,
      "step": 38790
    },
    {
      "entropy": 5.4891602993011475,
      "epoch": 4.357050598079407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003245436822375387,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22551260143518448,
      "num_tokens": 86604284.0,
      "step": 38795
    },
    {
      "entropy": 5.405604839324951,
      "epoch": 4.357612174987365,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003245045181374389,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22381707578897475,
      "num_tokens": 86615567.0,
      "step": 38800
    },
    {
      "entropy": 5.366706895828247,
      "epoch": 4.358173751895322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000324465352461096,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2367199406027794,
      "num_tokens": 86626188.0,
      "step": 38805
    },
    {
      "entropy": 5.508548212051392,
      "epoch": 4.35873532880328,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003244261852097572,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21988991200923919,
      "num_tokens": 86638468.0,
      "step": 38810
    },
    {
      "entropy": 5.524399709701538,
      "epoch": 4.359296905711237,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032438701638466943,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22473135888576506,
      "num_tokens": 86649119.0,
      "step": 38815
    },
    {
      "entropy": 5.486215019226075,
      "epoch": 4.359858482619194,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032434784598707987,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22345405220985412,
      "num_tokens": 86660924.0,
      "step": 38820
    },
    {
      "entropy": 5.441120433807373,
      "epoch": 4.3604200595271525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003243086740182358,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22381024211645126,
      "num_tokens": 86672655.0,
      "step": 38825
    },
    {
      "entropy": 5.388194894790649,
      "epoch": 4.36098163643511,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032426950047938436,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23226060718297958,
      "num_tokens": 86684146.0,
      "step": 38830
    },
    {
      "entropy": 5.513617277145386,
      "epoch": 4.361543213343067,
      "grad_norm": 1.03125,
      "learning_rate": 0.000324230325371773,
      "loss": 5.0982,
      "mean_token_accuracy": 0.2155735284090042,
      "num_tokens": 86695435.0,
      "step": 38835
    },
    {
      "entropy": 5.6521138668060305,
      "epoch": 4.362104790251025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003241911486966489,
      "loss": 5.125,
      "mean_token_accuracy": 0.2070575922727585,
      "num_tokens": 86706387.0,
      "step": 38840
    },
    {
      "entropy": 5.490468502044678,
      "epoch": 4.362666367158982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003241519704552596,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2272973269224167,
      "num_tokens": 86717661.0,
      "step": 38845
    },
    {
      "entropy": 5.428877210617065,
      "epoch": 4.36322794406694,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032411279064885253,
      "loss": 4.9132,
      "mean_token_accuracy": 0.2245629221200943,
      "num_tokens": 86729045.0,
      "step": 38850
    },
    {
      "entropy": 5.481683969497681,
      "epoch": 4.363789520974898,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032407360927867506,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21480597108602523,
      "num_tokens": 86740763.0,
      "step": 38855
    },
    {
      "entropy": 5.488077974319458,
      "epoch": 4.364351097882855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003240344263459749,
      "loss": 4.9911,
      "mean_token_accuracy": 0.22475745230913163,
      "num_tokens": 86752369.0,
      "step": 38860
    },
    {
      "entropy": 5.469762945175171,
      "epoch": 4.364912674790813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032399524185199957,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21748938411474228,
      "num_tokens": 86765947.0,
      "step": 38865
    },
    {
      "entropy": 5.3865001678466795,
      "epoch": 4.36547425169877,
      "grad_norm": 1.0,
      "learning_rate": 0.00032395605579799673,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2294256016612053,
      "num_tokens": 86777202.0,
      "step": 38870
    },
    {
      "entropy": 5.454809284210205,
      "epoch": 4.366035828606727,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032391686818521414,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22364619821310044,
      "num_tokens": 86787838.0,
      "step": 38875
    },
    {
      "entropy": 5.515267038345337,
      "epoch": 4.3665974055146854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032387767901489946,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21970974653959274,
      "num_tokens": 86799875.0,
      "step": 38880
    },
    {
      "entropy": 5.470282173156738,
      "epoch": 4.367158982422643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032383848828830054,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22256029099226,
      "num_tokens": 86811018.0,
      "step": 38885
    },
    {
      "entropy": 5.436767673492431,
      "epoch": 4.3677205593306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032379929600666525,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22696240544319152,
      "num_tokens": 86821556.0,
      "step": 38890
    },
    {
      "entropy": 5.336217784881592,
      "epoch": 4.368282136238558,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032376010217124133,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2416122302412987,
      "num_tokens": 86832041.0,
      "step": 38895
    },
    {
      "entropy": 5.459322595596314,
      "epoch": 4.368843713146515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032372090678327694,
      "loss": 4.9281,
      "mean_token_accuracy": 0.2212374061346054,
      "num_tokens": 86844723.0,
      "step": 38900
    },
    {
      "entropy": 5.513222360610962,
      "epoch": 4.369405290054473,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032368170984401997,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22516104578971863,
      "num_tokens": 86856211.0,
      "step": 38905
    },
    {
      "entropy": 5.467984962463379,
      "epoch": 4.3699668669624305,
      "grad_norm": 1.125,
      "learning_rate": 0.00032364251135471846,
      "loss": 4.9704,
      "mean_token_accuracy": 0.2185332491993904,
      "num_tokens": 86867701.0,
      "step": 38910
    },
    {
      "entropy": 5.504336547851563,
      "epoch": 4.370528443870388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003236033113166205,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21785043478012084,
      "num_tokens": 86879838.0,
      "step": 38915
    },
    {
      "entropy": 5.495995998382568,
      "epoch": 4.371090020778346,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032356410973097423,
      "loss": 5.0111,
      "mean_token_accuracy": 0.22295461744070053,
      "num_tokens": 86891101.0,
      "step": 38920
    },
    {
      "entropy": 5.52413854598999,
      "epoch": 4.371651597686303,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032352490659902795,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2291983038187027,
      "num_tokens": 86902796.0,
      "step": 38925
    },
    {
      "entropy": 5.498848915100098,
      "epoch": 4.37221317459426,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003234857019220298,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21585609763860703,
      "num_tokens": 86914381.0,
      "step": 38930
    },
    {
      "entropy": 5.509270095825196,
      "epoch": 4.372774751502218,
      "grad_norm": 1.03125,
      "learning_rate": 0.000323446495701228,
      "loss": 5.1025,
      "mean_token_accuracy": 0.21833983957767486,
      "num_tokens": 86926082.0,
      "step": 38935
    },
    {
      "entropy": 5.450918245315552,
      "epoch": 4.373336328410176,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003234072879378711,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22858246862888337,
      "num_tokens": 86936644.0,
      "step": 38940
    },
    {
      "entropy": 5.450481033325195,
      "epoch": 4.373897905318134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003233680786332074,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22225441336631774,
      "num_tokens": 86947586.0,
      "step": 38945
    },
    {
      "entropy": 5.488393783569336,
      "epoch": 4.374459482226091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003233288677884853,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2238599345088005,
      "num_tokens": 86958265.0,
      "step": 38950
    },
    {
      "entropy": 5.46598629951477,
      "epoch": 4.375021059134048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003232896554049533,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2253301903605461,
      "num_tokens": 86969119.0,
      "step": 38955
    },
    {
      "entropy": 5.438471078872681,
      "epoch": 4.375582636042006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032325044148385986,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21784359514713286,
      "num_tokens": 86980232.0,
      "step": 38960
    },
    {
      "entropy": 5.41444034576416,
      "epoch": 4.3761442129499635,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032321122602645374,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2244138464331627,
      "num_tokens": 86993590.0,
      "step": 38965
    },
    {
      "entropy": 5.491230630874634,
      "epoch": 4.376705789857921,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003231720090339834,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21607110798358917,
      "num_tokens": 87004905.0,
      "step": 38970
    },
    {
      "entropy": 5.462052822113037,
      "epoch": 4.377267366765879,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032313279050769767,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2294619545340538,
      "num_tokens": 87016920.0,
      "step": 38975
    },
    {
      "entropy": 5.53444128036499,
      "epoch": 4.377828943673836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003230935704488452,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21357343196868897,
      "num_tokens": 87027740.0,
      "step": 38980
    },
    {
      "entropy": 5.5301838397979735,
      "epoch": 4.378390520581794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032305434885867483,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21756920516490935,
      "num_tokens": 87039742.0,
      "step": 38985
    },
    {
      "entropy": 5.423443126678467,
      "epoch": 4.378952097489751,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032301512573843535,
      "loss": 4.8723,
      "mean_token_accuracy": 0.21939280033111572,
      "num_tokens": 87049750.0,
      "step": 38990
    },
    {
      "entropy": 5.503546333312988,
      "epoch": 4.3795136743977086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003229759010893756,
      "loss": 4.9835,
      "mean_token_accuracy": 0.23007114827632905,
      "num_tokens": 87062890.0,
      "step": 38995
    },
    {
      "entropy": 5.476947212219239,
      "epoch": 4.380075251305667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003229366749127446,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21839510202407836,
      "num_tokens": 87074450.0,
      "step": 39000
    },
    {
      "epoch": 4.380075251305667,
      "eval_entropy": 5.285658815825204,
      "eval_loss": 5.123254299163818,
      "eval_mean_token_accuracy": 0.2217517003968554,
      "eval_num_tokens": 87074450.0,
      "eval_runtime": 53.5363,
      "eval_samples_per_second": 579.289,
      "eval_steps_per_second": 72.418,
      "step": 39000
    },
    {
      "entropy": 5.607362222671509,
      "epoch": 4.380636828213624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003228974472097913,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21709448397159575,
      "num_tokens": 87085931.0,
      "step": 39005
    },
    {
      "entropy": 5.482158088684082,
      "epoch": 4.381198405121581,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032285821798176456,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22456406354904174,
      "num_tokens": 87096655.0,
      "step": 39010
    },
    {
      "entropy": 5.45820107460022,
      "epoch": 4.381759982029539,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032281898722991376,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21492307484149933,
      "num_tokens": 87108359.0,
      "step": 39015
    },
    {
      "entropy": 5.471878623962402,
      "epoch": 4.382321558937496,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032277975495548775,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21689616292715072,
      "num_tokens": 87119028.0,
      "step": 39020
    },
    {
      "entropy": 5.429191589355469,
      "epoch": 4.382883135845454,
      "grad_norm": 1.0,
      "learning_rate": 0.0003227405211597359,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22615356743335724,
      "num_tokens": 87130574.0,
      "step": 39025
    },
    {
      "entropy": 5.504112958908081,
      "epoch": 4.383444712753412,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032270128584390735,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22350075393915175,
      "num_tokens": 87141052.0,
      "step": 39030
    },
    {
      "entropy": 5.492665576934814,
      "epoch": 4.384006289661369,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003226620490092513,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21525583267211915,
      "num_tokens": 87151323.0,
      "step": 39035
    },
    {
      "entropy": 5.413222074508667,
      "epoch": 4.384567866569327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032262281065701716,
      "loss": 4.8666,
      "mean_token_accuracy": 0.21706949025392533,
      "num_tokens": 87162614.0,
      "step": 39040
    },
    {
      "entropy": 5.412001752853394,
      "epoch": 4.385129443477284,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003225835707884542,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22351963222026824,
      "num_tokens": 87172712.0,
      "step": 39045
    },
    {
      "entropy": 5.4799079418182375,
      "epoch": 4.3856910203852415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003225443294048119,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22887052744626998,
      "num_tokens": 87183612.0,
      "step": 39050
    },
    {
      "entropy": 5.439529228210449,
      "epoch": 4.3862525972932,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032250508650733976,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22343820333480835,
      "num_tokens": 87194707.0,
      "step": 39055
    },
    {
      "entropy": 5.387938690185547,
      "epoch": 4.386814174201157,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003224658420972871,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23036515861749648,
      "num_tokens": 87206451.0,
      "step": 39060
    },
    {
      "entropy": 5.472826337814331,
      "epoch": 4.387375751109114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003224265961759038,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21814199388027192,
      "num_tokens": 87217542.0,
      "step": 39065
    },
    {
      "entropy": 5.406626224517822,
      "epoch": 4.387937328017072,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003223873487444392,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23286348283290864,
      "num_tokens": 87228399.0,
      "step": 39070
    },
    {
      "entropy": 5.467068815231324,
      "epoch": 4.388498904925029,
      "grad_norm": 1.0,
      "learning_rate": 0.0003223480998041431,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22586329132318497,
      "num_tokens": 87239875.0,
      "step": 39075
    },
    {
      "entropy": 5.499501657485962,
      "epoch": 4.389060481832987,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032230884935626505,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21679504811763764,
      "num_tokens": 87252919.0,
      "step": 39080
    },
    {
      "entropy": 5.392881727218628,
      "epoch": 4.389622058740945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000322269597402055,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2208663046360016,
      "num_tokens": 87264458.0,
      "step": 39085
    },
    {
      "entropy": 5.485640192031861,
      "epoch": 4.390183635648902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003222303439427625,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22694895565509796,
      "num_tokens": 87274935.0,
      "step": 39090
    },
    {
      "entropy": 5.421903610229492,
      "epoch": 4.39074521255686,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032219108897963767,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22886288315057754,
      "num_tokens": 87285870.0,
      "step": 39095
    },
    {
      "entropy": 5.432970571517944,
      "epoch": 4.391306789464817,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003221518325139302,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22442394644021987,
      "num_tokens": 87297374.0,
      "step": 39100
    },
    {
      "entropy": 5.486106729507446,
      "epoch": 4.3918683663727744,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003221125745468901,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22268571704626083,
      "num_tokens": 87307597.0,
      "step": 39105
    },
    {
      "entropy": 5.4063084602355955,
      "epoch": 4.3924299432807326,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003220733150797673,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22755901217460633,
      "num_tokens": 87318459.0,
      "step": 39110
    },
    {
      "entropy": 5.553591060638428,
      "epoch": 4.39299152018869,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003220340541138119,
      "loss": 5.0968,
      "mean_token_accuracy": 0.2092335566878319,
      "num_tokens": 87330613.0,
      "step": 39115
    },
    {
      "entropy": 5.511180353164673,
      "epoch": 4.393553097096647,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032199479165027407,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2159762516617775,
      "num_tokens": 87341988.0,
      "step": 39120
    },
    {
      "entropy": 5.430227041244507,
      "epoch": 4.394114674004605,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032195552769040376,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22389326095581055,
      "num_tokens": 87353532.0,
      "step": 39125
    },
    {
      "entropy": 5.479941940307617,
      "epoch": 4.394676250912562,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003219162622354513,
      "loss": 4.9986,
      "mean_token_accuracy": 0.22558475732803346,
      "num_tokens": 87364754.0,
      "step": 39130
    },
    {
      "entropy": 5.41253023147583,
      "epoch": 4.39523782782052,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032187699528666674,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21854687184095384,
      "num_tokens": 87375983.0,
      "step": 39135
    },
    {
      "entropy": 5.415592432022095,
      "epoch": 4.395799404728478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003218377268453006,
      "loss": 4.9213,
      "mean_token_accuracy": 0.23373583257198333,
      "num_tokens": 87388169.0,
      "step": 39140
    },
    {
      "entropy": 5.513821411132812,
      "epoch": 4.396360981636435,
      "grad_norm": 1.078125,
      "learning_rate": 0.000321798456912603,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2264143407344818,
      "num_tokens": 87399341.0,
      "step": 39145
    },
    {
      "entropy": 5.449951219558716,
      "epoch": 4.396922558544393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032175918548982437,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22582374811172484,
      "num_tokens": 87410941.0,
      "step": 39150
    },
    {
      "entropy": 5.432147598266601,
      "epoch": 4.39748413545235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032171991257821513,
      "loss": 4.8953,
      "mean_token_accuracy": 0.2241746112704277,
      "num_tokens": 87421833.0,
      "step": 39155
    },
    {
      "entropy": 5.371978235244751,
      "epoch": 4.398045712360307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003216806381790258,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2332782730460167,
      "num_tokens": 87431648.0,
      "step": 39160
    },
    {
      "entropy": 5.493801879882812,
      "epoch": 4.3986072892682655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003216413622935069,
      "loss": 5.024,
      "mean_token_accuracy": 0.2237239509820938,
      "num_tokens": 87444207.0,
      "step": 39165
    },
    {
      "entropy": 5.39956636428833,
      "epoch": 4.399168866176223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032160208492290886,
      "loss": 4.8603,
      "mean_token_accuracy": 0.23240079134702682,
      "num_tokens": 87455919.0,
      "step": 39170
    },
    {
      "entropy": 5.44054012298584,
      "epoch": 4.399730443084181,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032156280606848237,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22502132058143615,
      "num_tokens": 87466979.0,
      "step": 39175
    },
    {
      "entropy": 5.502376699447632,
      "epoch": 4.400292019992138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003215235257314781,
      "loss": 4.968,
      "mean_token_accuracy": 0.2177415445446968,
      "num_tokens": 87478193.0,
      "step": 39180
    },
    {
      "entropy": 5.446694421768188,
      "epoch": 4.400853596900095,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032148424391314665,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2242877796292305,
      "num_tokens": 87488961.0,
      "step": 39185
    },
    {
      "entropy": 5.497700500488281,
      "epoch": 4.401415173808053,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032144496061473895,
      "loss": 5.0067,
      "mean_token_accuracy": 0.22452557682991028,
      "num_tokens": 87500508.0,
      "step": 39190
    },
    {
      "entropy": 5.4520103454589846,
      "epoch": 4.401976750716011,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003214056758375056,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2258489102125168,
      "num_tokens": 87510755.0,
      "step": 39195
    },
    {
      "entropy": 5.478241300582885,
      "epoch": 4.402538327623968,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003213663895826976,
      "loss": 4.95,
      "mean_token_accuracy": 0.21955355852842331,
      "num_tokens": 87522425.0,
      "step": 39200
    },
    {
      "entropy": 5.503452920913697,
      "epoch": 4.403099904531926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003213271018515657,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2260384351015091,
      "num_tokens": 87534209.0,
      "step": 39205
    },
    {
      "entropy": 5.386435508728027,
      "epoch": 4.403661481439883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000321287812645361,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22532330453395844,
      "num_tokens": 87544647.0,
      "step": 39210
    },
    {
      "entropy": 5.3859001159667965,
      "epoch": 4.40422305834784,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032124852196533434,
      "loss": 4.8387,
      "mean_token_accuracy": 0.23190588653087615,
      "num_tokens": 87555924.0,
      "step": 39215
    },
    {
      "entropy": 5.3889007568359375,
      "epoch": 4.404784635255798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003212092298127368,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22716501653194426,
      "num_tokens": 87566998.0,
      "step": 39220
    },
    {
      "entropy": 5.39280686378479,
      "epoch": 4.405346212163756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032116993618881945,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22347783744335176,
      "num_tokens": 87579516.0,
      "step": 39225
    },
    {
      "entropy": 5.535956382751465,
      "epoch": 4.405907789071714,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032113064109483345,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2207686185836792,
      "num_tokens": 87591512.0,
      "step": 39230
    },
    {
      "entropy": 5.4990699768066404,
      "epoch": 4.406469365979671,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032109134453202993,
      "loss": 4.909,
      "mean_token_accuracy": 0.217495858669281,
      "num_tokens": 87602783.0,
      "step": 39235
    },
    {
      "entropy": 5.424880647659302,
      "epoch": 4.407030942887628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032105204650166016,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2227158486843109,
      "num_tokens": 87613585.0,
      "step": 39240
    },
    {
      "entropy": 5.505851221084595,
      "epoch": 4.407592519795586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032101274700497537,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22053801119327546,
      "num_tokens": 87624003.0,
      "step": 39245
    },
    {
      "entropy": 5.429603719711304,
      "epoch": 4.4081540967035435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003209734460432269,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2284000739455223,
      "num_tokens": 87634429.0,
      "step": 39250
    },
    {
      "entropy": 5.427091455459594,
      "epoch": 4.408715673611501,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000320934143617666,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22650336474180222,
      "num_tokens": 87645887.0,
      "step": 39255
    },
    {
      "entropy": 5.484008169174194,
      "epoch": 4.409277250519459,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003208948397295441,
      "loss": 4.942,
      "mean_token_accuracy": 0.22250875532627107,
      "num_tokens": 87657439.0,
      "step": 39260
    },
    {
      "entropy": 5.409818029403686,
      "epoch": 4.409838827427416,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032085553438011285,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22470861077308654,
      "num_tokens": 87668461.0,
      "step": 39265
    },
    {
      "entropy": 5.5745322704315186,
      "epoch": 4.410400404335373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003208162275706235,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21706801205873488,
      "num_tokens": 87680595.0,
      "step": 39270
    },
    {
      "entropy": 5.4258605480194095,
      "epoch": 4.410961981243331,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003207769193023277,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22573997974395751,
      "num_tokens": 87691563.0,
      "step": 39275
    },
    {
      "entropy": 5.4989790439605715,
      "epoch": 4.411523558151289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032073760957647705,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21748031973838805,
      "num_tokens": 87702052.0,
      "step": 39280
    },
    {
      "entropy": 5.423162317276001,
      "epoch": 4.412085135059247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003206982983943232,
      "loss": 4.9421,
      "mean_token_accuracy": 0.21700615286827088,
      "num_tokens": 87713123.0,
      "step": 39285
    },
    {
      "entropy": 5.49748969078064,
      "epoch": 4.412646711967204,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003206589857571178,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2186909317970276,
      "num_tokens": 87723294.0,
      "step": 39290
    },
    {
      "entropy": 5.4749397277832035,
      "epoch": 4.413208288875161,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032061967166611255,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22911046892404557,
      "num_tokens": 87733751.0,
      "step": 39295
    },
    {
      "entropy": 5.4317535877227785,
      "epoch": 4.413769865783119,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032058035612255934,
      "loss": 4.8967,
      "mean_token_accuracy": 0.23303252458572388,
      "num_tokens": 87744584.0,
      "step": 39300
    },
    {
      "entropy": 5.474879312515259,
      "epoch": 4.4143314426910765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003205410391277098,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22651025801897048,
      "num_tokens": 87755420.0,
      "step": 39305
    },
    {
      "entropy": 5.520950031280518,
      "epoch": 4.414893019599034,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032050172068281594,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22209255695343016,
      "num_tokens": 87767213.0,
      "step": 39310
    },
    {
      "entropy": 5.4937355518341064,
      "epoch": 4.415454596506992,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003204624007891297,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22287026643753052,
      "num_tokens": 87778058.0,
      "step": 39315
    },
    {
      "entropy": 5.470255422592163,
      "epoch": 4.416016173414949,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000320423079447903,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21940568089485168,
      "num_tokens": 87788279.0,
      "step": 39320
    },
    {
      "entropy": 5.420727729797363,
      "epoch": 4.416577750322907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032038375666038776,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22200984209775926,
      "num_tokens": 87799643.0,
      "step": 39325
    },
    {
      "entropy": 5.388582324981689,
      "epoch": 4.417139327230864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003203444324278361,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22743337899446486,
      "num_tokens": 87810493.0,
      "step": 39330
    },
    {
      "entropy": 5.4632173538208,
      "epoch": 4.4177009041388215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003203051067515002,
      "loss": 4.9577,
      "mean_token_accuracy": 0.222740475833416,
      "num_tokens": 87821876.0,
      "step": 39335
    },
    {
      "entropy": 5.457635498046875,
      "epoch": 4.41826248104678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003202657796326321,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23467588275671006,
      "num_tokens": 87832058.0,
      "step": 39340
    },
    {
      "entropy": 5.50344614982605,
      "epoch": 4.418824057954737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000320226451072484,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22254351079463958,
      "num_tokens": 87842549.0,
      "step": 39345
    },
    {
      "entropy": 5.426849603652954,
      "epoch": 4.419385634862694,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032018712107230815,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22485638409852982,
      "num_tokens": 87853507.0,
      "step": 39350
    },
    {
      "entropy": 5.510846662521362,
      "epoch": 4.419947211770652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003201477896333568,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22579678595066072,
      "num_tokens": 87864546.0,
      "step": 39355
    },
    {
      "entropy": 5.477618885040283,
      "epoch": 4.420508788678609,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003201084567568824,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2328917846083641,
      "num_tokens": 87876553.0,
      "step": 39360
    },
    {
      "entropy": 5.4230975151062015,
      "epoch": 4.4210703655865675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032006912244413727,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22229408621788024,
      "num_tokens": 87886257.0,
      "step": 39365
    },
    {
      "entropy": 5.4300939559936525,
      "epoch": 4.421631942494525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003200297866963737,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23210005611181259,
      "num_tokens": 87897704.0,
      "step": 39370
    },
    {
      "entropy": 5.429366731643677,
      "epoch": 4.422193519402482,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003199904495148443,
      "loss": 4.943,
      "mean_token_accuracy": 0.2313595488667488,
      "num_tokens": 87908892.0,
      "step": 39375
    },
    {
      "entropy": 5.463156032562256,
      "epoch": 4.42275509631044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003199511109008016,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22885927557945251,
      "num_tokens": 87921502.0,
      "step": 39380
    },
    {
      "entropy": 5.491264867782593,
      "epoch": 4.423316673218397,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031991177085549807,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2281908795237541,
      "num_tokens": 87933250.0,
      "step": 39385
    },
    {
      "entropy": 5.4236914157867435,
      "epoch": 4.4238782501263545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031987242938018626,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23987323194742202,
      "num_tokens": 87945715.0,
      "step": 39390
    },
    {
      "entropy": 5.377259111404419,
      "epoch": 4.424439827034313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031983308647611896,
      "loss": 4.866,
      "mean_token_accuracy": 0.2336372524499893,
      "num_tokens": 87956973.0,
      "step": 39395
    },
    {
      "entropy": 5.496961212158203,
      "epoch": 4.42500140394227,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003197937421445488,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22017743587493896,
      "num_tokens": 87968341.0,
      "step": 39400
    },
    {
      "entropy": 5.416796636581421,
      "epoch": 4.425562980850227,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003197543963867285,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22212274372577667,
      "num_tokens": 87978980.0,
      "step": 39405
    },
    {
      "entropy": 5.510531377792359,
      "epoch": 4.426124557758185,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003197150492039108,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22574230283498764,
      "num_tokens": 87990612.0,
      "step": 39410
    },
    {
      "entropy": 5.523446989059448,
      "epoch": 4.426686134666142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003196757005973487,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22486769706010817,
      "num_tokens": 88001052.0,
      "step": 39415
    },
    {
      "entropy": 5.474443626403809,
      "epoch": 4.4272477115741005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031963635056829487,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2279496192932129,
      "num_tokens": 88013102.0,
      "step": 39420
    },
    {
      "entropy": 5.496119165420533,
      "epoch": 4.427809288482058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031959699911800237,
      "loss": 5.0145,
      "mean_token_accuracy": 0.22585564851760864,
      "num_tokens": 88023952.0,
      "step": 39425
    },
    {
      "entropy": 5.44144868850708,
      "epoch": 4.428370865390015,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003195576462477241,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22196145802736283,
      "num_tokens": 88035148.0,
      "step": 39430
    },
    {
      "entropy": 5.441307449340821,
      "epoch": 4.428932442297973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031951829195871315,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21757904291152955,
      "num_tokens": 88046294.0,
      "step": 39435
    },
    {
      "entropy": 5.447620439529419,
      "epoch": 4.42949401920593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031947893625222244,
      "loss": 4.922,
      "mean_token_accuracy": 0.22397839426994323,
      "num_tokens": 88058225.0,
      "step": 39440
    },
    {
      "entropy": 5.505614566802978,
      "epoch": 4.430055596113887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031943957912950523,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21688448786735534,
      "num_tokens": 88069994.0,
      "step": 39445
    },
    {
      "entropy": 5.520434761047364,
      "epoch": 4.4306171730218455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003194002205918145,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22396422177553177,
      "num_tokens": 88081912.0,
      "step": 39450
    },
    {
      "entropy": 5.453877115249634,
      "epoch": 4.431178749929803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003193608606404036,
      "loss": 4.923,
      "mean_token_accuracy": 0.22667381465435027,
      "num_tokens": 88092457.0,
      "step": 39455
    },
    {
      "entropy": 5.543041324615478,
      "epoch": 4.43174032683776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003193214992765257,
      "loss": 5.0848,
      "mean_token_accuracy": 0.21107200384140015,
      "num_tokens": 88104407.0,
      "step": 39460
    },
    {
      "entropy": 5.413370847702026,
      "epoch": 4.432301903745718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031928213650143407,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22913040965795517,
      "num_tokens": 88114653.0,
      "step": 39465
    },
    {
      "entropy": 5.5764322757720945,
      "epoch": 4.432863480653675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000319242772316382,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22635554075241088,
      "num_tokens": 88125477.0,
      "step": 39470
    },
    {
      "entropy": 5.439753723144531,
      "epoch": 4.433425057561633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003192034067226228,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23228394389152526,
      "num_tokens": 88136419.0,
      "step": 39475
    },
    {
      "entropy": 5.423063087463379,
      "epoch": 4.433986634469591,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031916403972141017,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2302507072687149,
      "num_tokens": 88146578.0,
      "step": 39480
    },
    {
      "entropy": 5.280356597900391,
      "epoch": 4.434548211377548,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003191246713139973,
      "loss": 4.7871,
      "mean_token_accuracy": 0.2371435597538948,
      "num_tokens": 88157351.0,
      "step": 39485
    },
    {
      "entropy": 5.454060173034668,
      "epoch": 4.435109788285506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031908530150163777,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22122818529605864,
      "num_tokens": 88168460.0,
      "step": 39490
    },
    {
      "entropy": 5.54032301902771,
      "epoch": 4.435671365193463,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031904593028558517,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22368544787168504,
      "num_tokens": 88178994.0,
      "step": 39495
    },
    {
      "entropy": 5.486897659301758,
      "epoch": 4.43623294210142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031900655766709305,
      "loss": 4.8713,
      "mean_token_accuracy": 0.2258126065135002,
      "num_tokens": 88188916.0,
      "step": 39500
    },
    {
      "entropy": 5.380896043777466,
      "epoch": 4.4367945190093785,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003189671836474151,
      "loss": 4.924,
      "mean_token_accuracy": 0.22603742331266402,
      "num_tokens": 88200833.0,
      "step": 39505
    },
    {
      "entropy": 5.47540979385376,
      "epoch": 4.437356095917336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003189278082278049,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2161484733223915,
      "num_tokens": 88212022.0,
      "step": 39510
    },
    {
      "entropy": 5.459175252914429,
      "epoch": 4.437917672825294,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031888843140951627,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23255739659070968,
      "num_tokens": 88222327.0,
      "step": 39515
    },
    {
      "entropy": 5.348937892913819,
      "epoch": 4.438479249733251,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003188490531938031,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2380232632160187,
      "num_tokens": 88233141.0,
      "step": 39520
    },
    {
      "entropy": 5.5083517074584964,
      "epoch": 4.439040826641208,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031880967358191884,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21251277327537538,
      "num_tokens": 88245663.0,
      "step": 39525
    },
    {
      "entropy": 5.475397729873658,
      "epoch": 4.439602403549166,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031877029257511774,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22351008802652358,
      "num_tokens": 88256291.0,
      "step": 39530
    },
    {
      "entropy": 5.530876684188843,
      "epoch": 4.440163980457124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031873091017465356,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22142276167869568,
      "num_tokens": 88268382.0,
      "step": 39535
    },
    {
      "entropy": 5.398684549331665,
      "epoch": 4.440725557365081,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003186915263817802,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2317397952079773,
      "num_tokens": 88279638.0,
      "step": 39540
    },
    {
      "entropy": 5.469324827194214,
      "epoch": 4.441287134273039,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031865214119775173,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22296528071165084,
      "num_tokens": 88290599.0,
      "step": 39545
    },
    {
      "entropy": 5.465048694610596,
      "epoch": 4.441848711180996,
      "grad_norm": 1.078125,
      "learning_rate": 0.000318612754623822,
      "loss": 4.9301,
      "mean_token_accuracy": 0.2288094013929367,
      "num_tokens": 88301572.0,
      "step": 39550
    },
    {
      "entropy": 5.534448194503784,
      "epoch": 4.442410288088954,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031857336666124537,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2204577624797821,
      "num_tokens": 88312978.0,
      "step": 39555
    },
    {
      "entropy": 5.472358179092407,
      "epoch": 4.442971864996911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003185339773112758,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2156166687607765,
      "num_tokens": 88324023.0,
      "step": 39560
    },
    {
      "entropy": 5.410524892807007,
      "epoch": 4.443533441904869,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003184945865751675,
      "loss": 4.8851,
      "mean_token_accuracy": 0.21819587498903276,
      "num_tokens": 88335366.0,
      "step": 39565
    },
    {
      "entropy": 5.462276411056519,
      "epoch": 4.444095018812827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031845519445417475,
      "loss": 4.9423,
      "mean_token_accuracy": 0.22152890712022782,
      "num_tokens": 88347786.0,
      "step": 39570
    },
    {
      "entropy": 5.5407476902008055,
      "epoch": 4.444656595720784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003184158009495518,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21760598570108414,
      "num_tokens": 88358720.0,
      "step": 39575
    },
    {
      "entropy": 5.547049570083618,
      "epoch": 4.445218172628741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003183764060625528,
      "loss": 4.9926,
      "mean_token_accuracy": 0.22072043120861054,
      "num_tokens": 88370152.0,
      "step": 39580
    },
    {
      "entropy": 5.448483800888061,
      "epoch": 4.445779749536699,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031833700979443227,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22526223808526993,
      "num_tokens": 88381385.0,
      "step": 39585
    },
    {
      "entropy": 5.44879674911499,
      "epoch": 4.4463413264446565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031829761214644453,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22428101897239686,
      "num_tokens": 88391450.0,
      "step": 39590
    },
    {
      "entropy": 5.348173046112061,
      "epoch": 4.446902903352614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000318258213119844,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23957473337650298,
      "num_tokens": 88402310.0,
      "step": 39595
    },
    {
      "entropy": 5.43366584777832,
      "epoch": 4.447464480260572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003182188127158852,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2156695231795311,
      "num_tokens": 88412286.0,
      "step": 39600
    },
    {
      "entropy": 5.5346479415893555,
      "epoch": 4.448026057168529,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003181794109358227,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2227863445878029,
      "num_tokens": 88423335.0,
      "step": 39605
    },
    {
      "entropy": 5.3994426250457765,
      "epoch": 4.448587634076487,
      "grad_norm": 1.0,
      "learning_rate": 0.0003181400077809109,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23297772258520127,
      "num_tokens": 88434328.0,
      "step": 39610
    },
    {
      "entropy": 5.455141305923462,
      "epoch": 4.449149210984444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031810060325240466,
      "loss": 4.9306,
      "mean_token_accuracy": 0.22355677783489228,
      "num_tokens": 88445611.0,
      "step": 39615
    },
    {
      "entropy": 5.485100269317627,
      "epoch": 4.449710787892402,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003180611973515583,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2212423101067543,
      "num_tokens": 88456207.0,
      "step": 39620
    },
    {
      "entropy": 5.574441576004029,
      "epoch": 4.45027236480036,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031802179007962687,
      "loss": 5.093,
      "mean_token_accuracy": 0.21439885199069977,
      "num_tokens": 88467322.0,
      "step": 39625
    },
    {
      "entropy": 5.482133483886718,
      "epoch": 4.450833941708317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003179823814378649,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22316208481788635,
      "num_tokens": 88479703.0,
      "step": 39630
    },
    {
      "entropy": 5.3728655815124515,
      "epoch": 4.451395518616274,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031794297142752727,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22211243212223053,
      "num_tokens": 88490284.0,
      "step": 39635
    },
    {
      "entropy": 5.432567930221557,
      "epoch": 4.451957095524232,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003179035600498687,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22097376734018326,
      "num_tokens": 88502232.0,
      "step": 39640
    },
    {
      "entropy": 5.4118931770324705,
      "epoch": 4.4525186724321895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031786414730614423,
      "loss": 4.91,
      "mean_token_accuracy": 0.2252911776304245,
      "num_tokens": 88513788.0,
      "step": 39645
    },
    {
      "entropy": 5.436362266540527,
      "epoch": 4.453080249340147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003178247331976086,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21956576108932496,
      "num_tokens": 88524942.0,
      "step": 39650
    },
    {
      "entropy": 5.3970095157623295,
      "epoch": 4.453641826248105,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031778531772551696,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23154211789369583,
      "num_tokens": 88535692.0,
      "step": 39655
    },
    {
      "entropy": 5.448426246643066,
      "epoch": 4.454203403156062,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003177459008911241,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2261297434568405,
      "num_tokens": 88547007.0,
      "step": 39660
    },
    {
      "entropy": 5.517078256607055,
      "epoch": 4.45476498006402,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031770648269568524,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22878472954034806,
      "num_tokens": 88558197.0,
      "step": 39665
    },
    {
      "entropy": 5.520717477798462,
      "epoch": 4.455326556971977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003176670631404554,
      "loss": 4.9958,
      "mean_token_accuracy": 0.2206620991230011,
      "num_tokens": 88569488.0,
      "step": 39670
    },
    {
      "entropy": 5.452864789962769,
      "epoch": 4.4558881338799345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031762764222668976,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22079122215509414,
      "num_tokens": 88582054.0,
      "step": 39675
    },
    {
      "entropy": 5.4288801670074465,
      "epoch": 4.456449710787893,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031758821995564343,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22733242064714432,
      "num_tokens": 88592528.0,
      "step": 39680
    },
    {
      "entropy": 5.49313530921936,
      "epoch": 4.45701128769585,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003175487963285716,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22786980569362641,
      "num_tokens": 88604136.0,
      "step": 39685
    },
    {
      "entropy": 5.544993543624878,
      "epoch": 4.457572864603807,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003175093713467297,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22295190542936325,
      "num_tokens": 88615629.0,
      "step": 39690
    },
    {
      "entropy": 5.391041421890259,
      "epoch": 4.458134441511765,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031746994501137283,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23288982659578322,
      "num_tokens": 88626100.0,
      "step": 39695
    },
    {
      "entropy": 5.425187397003174,
      "epoch": 4.458696018419722,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003174305173237565,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2184377685189247,
      "num_tokens": 88638948.0,
      "step": 39700
    },
    {
      "entropy": 5.4720478534698485,
      "epoch": 4.4592575953276805,
      "grad_norm": 1.125,
      "learning_rate": 0.000317391088285136,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2267664134502411,
      "num_tokens": 88650488.0,
      "step": 39705
    },
    {
      "entropy": 5.5436676979064945,
      "epoch": 4.459819172235638,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003173516578967668,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2157947063446045,
      "num_tokens": 88661584.0,
      "step": 39710
    },
    {
      "entropy": 5.420422887802124,
      "epoch": 4.460380749143595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031731222615990443,
      "loss": 4.8927,
      "mean_token_accuracy": 0.2266027718782425,
      "num_tokens": 88673051.0,
      "step": 39715
    },
    {
      "entropy": 5.452140808105469,
      "epoch": 4.460942326051553,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031727279307580436,
      "loss": 5.021,
      "mean_token_accuracy": 0.21752583384513854,
      "num_tokens": 88684122.0,
      "step": 39720
    },
    {
      "entropy": 5.517731618881226,
      "epoch": 4.46150390295951,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031723335864572216,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21879401803016663,
      "num_tokens": 88694620.0,
      "step": 39725
    },
    {
      "entropy": 5.586868667602539,
      "epoch": 4.4620654798674675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031719392287091343,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2232214868068695,
      "num_tokens": 88706137.0,
      "step": 39730
    },
    {
      "entropy": 5.478714513778686,
      "epoch": 4.462627056775426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003171544857526338,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22429933845996858,
      "num_tokens": 88717326.0,
      "step": 39735
    },
    {
      "entropy": 5.4524843215942385,
      "epoch": 4.463188633683383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003171150472921391,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22681159526109695,
      "num_tokens": 88727125.0,
      "step": 39740
    },
    {
      "entropy": 5.421259021759033,
      "epoch": 4.463750210591341,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031707560749068487,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22187957167625427,
      "num_tokens": 88737098.0,
      "step": 39745
    },
    {
      "entropy": 5.482561683654785,
      "epoch": 4.464311787499298,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000317036166349527,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2225729390978813,
      "num_tokens": 88748454.0,
      "step": 39750
    },
    {
      "entropy": 5.4548882961273195,
      "epoch": 4.464873364407255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003169967238699213,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2200855553150177,
      "num_tokens": 88760165.0,
      "step": 39755
    },
    {
      "entropy": 5.4404644012451175,
      "epoch": 4.4654349413152135,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031695728005312363,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22141189724206925,
      "num_tokens": 88772124.0,
      "step": 39760
    },
    {
      "entropy": 5.384089422225952,
      "epoch": 4.465996518223171,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003169178349003898,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22390239387750627,
      "num_tokens": 88782429.0,
      "step": 39765
    },
    {
      "entropy": 5.446290969848633,
      "epoch": 4.466558095131128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000316878388412976,
      "loss": 5.005,
      "mean_token_accuracy": 0.21791324019432068,
      "num_tokens": 88794060.0,
      "step": 39770
    },
    {
      "entropy": 5.522270584106446,
      "epoch": 4.467119672039086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000316838940592138,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22695681303739548,
      "num_tokens": 88806174.0,
      "step": 39775
    },
    {
      "entropy": 5.415129137039185,
      "epoch": 4.467681248947043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031679949143913183,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22688726633787154,
      "num_tokens": 88817587.0,
      "step": 39780
    },
    {
      "entropy": 5.516943359375,
      "epoch": 4.468242825855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003167600409552137,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2295068919658661,
      "num_tokens": 88829435.0,
      "step": 39785
    },
    {
      "entropy": 5.438367509841919,
      "epoch": 4.4688044027629585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003167205891416397,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21947738975286485,
      "num_tokens": 88840244.0,
      "step": 39790
    },
    {
      "entropy": 5.53250527381897,
      "epoch": 4.469365979670916,
      "grad_norm": 1.078125,
      "learning_rate": 0.000316681135999666,
      "loss": 5.0507,
      "mean_token_accuracy": 0.21977671533823012,
      "num_tokens": 88852085.0,
      "step": 39795
    },
    {
      "entropy": 5.5074316501617435,
      "epoch": 4.469927556578874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003166416815305486,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21804299354553222,
      "num_tokens": 88864605.0,
      "step": 39800
    },
    {
      "entropy": 5.50226936340332,
      "epoch": 4.470489133486831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031660222573554406,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22579880952835082,
      "num_tokens": 88876209.0,
      "step": 39805
    },
    {
      "entropy": 5.429501867294311,
      "epoch": 4.471050710394788,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003165627686159085,
      "loss": 4.9281,
      "mean_token_accuracy": 0.23044562339782715,
      "num_tokens": 88886346.0,
      "step": 39810
    },
    {
      "entropy": 5.382834768295288,
      "epoch": 4.471612287302746,
      "grad_norm": 1.125,
      "learning_rate": 0.0003165233101728982,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22495102882385254,
      "num_tokens": 88898190.0,
      "step": 39815
    },
    {
      "entropy": 5.472168874740601,
      "epoch": 4.472173864210704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031648385040776965,
      "loss": 5.0122,
      "mean_token_accuracy": 0.20940009653568267,
      "num_tokens": 88909957.0,
      "step": 39820
    },
    {
      "entropy": 5.6455107688903805,
      "epoch": 4.472735441118661,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003164443893217793,
      "loss": 5.1059,
      "mean_token_accuracy": 0.21384526789188385,
      "num_tokens": 88922993.0,
      "step": 39825
    },
    {
      "entropy": 5.417866039276123,
      "epoch": 4.473297018026619,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003164049269161834,
      "loss": 4.8386,
      "mean_token_accuracy": 0.22999694049358368,
      "num_tokens": 88934506.0,
      "step": 39830
    },
    {
      "entropy": 5.455194187164307,
      "epoch": 4.473858594934576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031636546319223845,
      "loss": 4.9369,
      "mean_token_accuracy": 0.2235606476664543,
      "num_tokens": 88946963.0,
      "step": 39835
    },
    {
      "entropy": 5.465081834793091,
      "epoch": 4.474420171842533,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003163259981512013,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22110593914985657,
      "num_tokens": 88956995.0,
      "step": 39840
    },
    {
      "entropy": 5.495630598068237,
      "epoch": 4.4749817487504915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003162865317943283,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22669213861227036,
      "num_tokens": 88967507.0,
      "step": 39845
    },
    {
      "entropy": 5.575601530075073,
      "epoch": 4.475543325658449,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031624706412287614,
      "loss": 5.0274,
      "mean_token_accuracy": 0.22169420421123504,
      "num_tokens": 88979943.0,
      "step": 39850
    },
    {
      "entropy": 5.4888323783874515,
      "epoch": 4.476104902566407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003162075951381014,
      "loss": 5.0382,
      "mean_token_accuracy": 0.21488074511289595,
      "num_tokens": 88990840.0,
      "step": 39855
    },
    {
      "entropy": 5.482742977142334,
      "epoch": 4.476666479474364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031616812484126084,
      "loss": 5.034,
      "mean_token_accuracy": 0.21025368869304656,
      "num_tokens": 89002245.0,
      "step": 39860
    },
    {
      "entropy": 5.555473041534424,
      "epoch": 4.477228056382321,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003161286532336113,
      "loss": 5.0289,
      "mean_token_accuracy": 0.2255301758646965,
      "num_tokens": 89013520.0,
      "step": 39865
    },
    {
      "entropy": 5.527439975738526,
      "epoch": 4.477789633290279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031608918031640946,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22401530742645265,
      "num_tokens": 89025377.0,
      "step": 39870
    },
    {
      "entropy": 5.452583169937133,
      "epoch": 4.478351210198237,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003160497060909121,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22116939425468446,
      "num_tokens": 89035299.0,
      "step": 39875
    },
    {
      "entropy": 5.483480215072632,
      "epoch": 4.478912787106194,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031601023055837624,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22540455907583237,
      "num_tokens": 89046255.0,
      "step": 39880
    },
    {
      "entropy": 5.504537439346313,
      "epoch": 4.479474364014152,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031597075372005875,
      "loss": 5.0459,
      "mean_token_accuracy": 0.2156398668885231,
      "num_tokens": 89057873.0,
      "step": 39885
    },
    {
      "entropy": 5.43861403465271,
      "epoch": 4.480035940922109,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031593127557721657,
      "loss": 4.8407,
      "mean_token_accuracy": 0.2258278027176857,
      "num_tokens": 89068830.0,
      "step": 39890
    },
    {
      "entropy": 5.414667320251465,
      "epoch": 4.480597517830067,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003158917961311067,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2270885154604912,
      "num_tokens": 89080328.0,
      "step": 39895
    },
    {
      "entropy": 5.471400928497315,
      "epoch": 4.481159094738024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003158523153829862,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22080646753311156,
      "num_tokens": 89090118.0,
      "step": 39900
    },
    {
      "entropy": 5.526761054992676,
      "epoch": 4.481720671645982,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003158128333341121,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22288597375154495,
      "num_tokens": 89103338.0,
      "step": 39905
    },
    {
      "entropy": 5.496445178985596,
      "epoch": 4.48228224855394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031577334998574155,
      "loss": 4.914,
      "mean_token_accuracy": 0.226732736825943,
      "num_tokens": 89113721.0,
      "step": 39910
    },
    {
      "entropy": 5.6198082447052,
      "epoch": 4.482843825461897,
      "grad_norm": 1.125,
      "learning_rate": 0.0003157338653391317,
      "loss": 5.018,
      "mean_token_accuracy": 0.21876502335071563,
      "num_tokens": 89125425.0,
      "step": 39915
    },
    {
      "entropy": 5.494821500778198,
      "epoch": 4.483405402369854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003156943793955398,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2194610357284546,
      "num_tokens": 89135407.0,
      "step": 39920
    },
    {
      "entropy": 5.481952381134033,
      "epoch": 4.483966979277812,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031565489215622304,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2221390947699547,
      "num_tokens": 89147563.0,
      "step": 39925
    },
    {
      "entropy": 5.508932447433471,
      "epoch": 4.4845285561857695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003156154036224388,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22341302037239075,
      "num_tokens": 89158336.0,
      "step": 39930
    },
    {
      "entropy": 5.422506618499756,
      "epoch": 4.485090133093728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003155759137954444,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22869114428758622,
      "num_tokens": 89168894.0,
      "step": 39935
    },
    {
      "entropy": 5.541752767562866,
      "epoch": 4.485651710001685,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031553642267649715,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22736286967992783,
      "num_tokens": 89180701.0,
      "step": 39940
    },
    {
      "entropy": 5.487986612319946,
      "epoch": 4.486213286909642,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003154969302668544,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22744088619947433,
      "num_tokens": 89191192.0,
      "step": 39945
    },
    {
      "entropy": 5.451310491561889,
      "epoch": 4.4867748638176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003154574365677737,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22468456774950027,
      "num_tokens": 89202377.0,
      "step": 39950
    },
    {
      "entropy": 5.388812017440796,
      "epoch": 4.487336440725557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003154179415805126,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23155782967805863,
      "num_tokens": 89212869.0,
      "step": 39955
    },
    {
      "entropy": 5.427642822265625,
      "epoch": 4.487898017633515,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003153784453063285,
      "loss": 4.947,
      "mean_token_accuracy": 0.22262288630008698,
      "num_tokens": 89224558.0,
      "step": 39960
    },
    {
      "entropy": 5.392936325073242,
      "epoch": 4.488459594541473,
      "grad_norm": 1.0,
      "learning_rate": 0.0003153389477464791,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2279853641986847,
      "num_tokens": 89235645.0,
      "step": 39965
    },
    {
      "entropy": 5.4235576629638675,
      "epoch": 4.48902117144943,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003152994489022219,
      "loss": 4.874,
      "mean_token_accuracy": 0.2276056483387947,
      "num_tokens": 89247063.0,
      "step": 39970
    },
    {
      "entropy": 5.461188316345215,
      "epoch": 4.489582748357387,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031525994877481465,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2191716030240059,
      "num_tokens": 89258297.0,
      "step": 39975
    },
    {
      "entropy": 5.436674737930298,
      "epoch": 4.490144325265345,
      "grad_norm": 1.078125,
      "learning_rate": 0.000315220447365515,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22500652372837066,
      "num_tokens": 89269245.0,
      "step": 39980
    },
    {
      "entropy": 5.507085180282592,
      "epoch": 4.4907059021733025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031518094467558074,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2170154258608818,
      "num_tokens": 89281959.0,
      "step": 39985
    },
    {
      "entropy": 5.441592264175415,
      "epoch": 4.491267479081261,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031514144070626956,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23263459652662277,
      "num_tokens": 89292656.0,
      "step": 39990
    },
    {
      "entropy": 5.477483654022217,
      "epoch": 4.491829055989218,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031510193545883944,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22280487567186355,
      "num_tokens": 89304198.0,
      "step": 39995
    },
    {
      "entropy": 5.485585498809814,
      "epoch": 4.492390632897175,
      "grad_norm": 1.140625,
      "learning_rate": 0.000315062428934548,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22988380789756774,
      "num_tokens": 89315443.0,
      "step": 40000
    },
    {
      "entropy": 5.423489379882812,
      "epoch": 4.492952209805133,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031502292113465334,
      "loss": 4.969,
      "mean_token_accuracy": 0.22458784133195878,
      "num_tokens": 89326597.0,
      "step": 40005
    },
    {
      "entropy": 5.507819890975952,
      "epoch": 4.49351378671309,
      "grad_norm": 1.125,
      "learning_rate": 0.00031498341206041337,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21966915279626847,
      "num_tokens": 89338385.0,
      "step": 40010
    },
    {
      "entropy": 5.454037380218506,
      "epoch": 4.4940753636210475,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000314943901713086,
      "loss": 4.8999,
      "mean_token_accuracy": 0.2298089772462845,
      "num_tokens": 89350950.0,
      "step": 40015
    },
    {
      "entropy": 5.4400533676147464,
      "epoch": 4.494636940529006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003149043900939294,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22299591898918153,
      "num_tokens": 89361154.0,
      "step": 40020
    },
    {
      "entropy": 5.472492980957031,
      "epoch": 4.495198517436963,
      "grad_norm": 1.0,
      "learning_rate": 0.00031486487720420146,
      "loss": 4.9762,
      "mean_token_accuracy": 0.22396794706583023,
      "num_tokens": 89373188.0,
      "step": 40025
    },
    {
      "entropy": 5.482180404663086,
      "epoch": 4.49576009434492,
      "grad_norm": 1.125,
      "learning_rate": 0.00031482536304516047,
      "loss": 4.909,
      "mean_token_accuracy": 0.22902706265449524,
      "num_tokens": 89383651.0,
      "step": 40030
    },
    {
      "entropy": 5.39109320640564,
      "epoch": 4.496321671252878,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003147858476180644,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22213201224803925,
      "num_tokens": 89395385.0,
      "step": 40035
    },
    {
      "entropy": 5.415160179138184,
      "epoch": 4.496883248160835,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031474633092417155,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22234738171100615,
      "num_tokens": 89406071.0,
      "step": 40040
    },
    {
      "entropy": 5.5439516544342045,
      "epoch": 4.4974448250687935,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031470681296474006,
      "loss": 5.0634,
      "mean_token_accuracy": 0.21998815834522248,
      "num_tokens": 89417044.0,
      "step": 40045
    },
    {
      "entropy": 5.456154870986938,
      "epoch": 4.498006401976751,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003146672937410283,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2218027114868164,
      "num_tokens": 89428253.0,
      "step": 40050
    },
    {
      "entropy": 5.471809339523316,
      "epoch": 4.498567978884708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031462777325429447,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21967702656984328,
      "num_tokens": 89438760.0,
      "step": 40055
    },
    {
      "entropy": 5.39780478477478,
      "epoch": 4.499129555792666,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000314588251505797,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23626451790332795,
      "num_tokens": 89450335.0,
      "step": 40060
    },
    {
      "entropy": 5.5226410865783695,
      "epoch": 4.499691132700623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031454872849679425,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22229253500699997,
      "num_tokens": 89461077.0,
      "step": 40065
    },
    {
      "entropy": 5.540978240966797,
      "epoch": 4.500252709608581,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003145092042285446,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21893227398395537,
      "num_tokens": 89472281.0,
      "step": 40070
    },
    {
      "entropy": 5.483084964752197,
      "epoch": 4.500814286516539,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003144696787023066,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22305919826030732,
      "num_tokens": 89483030.0,
      "step": 40075
    },
    {
      "entropy": 5.404547691345215,
      "epoch": 4.501375863424496,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031443015191933864,
      "loss": 4.8285,
      "mean_token_accuracy": 0.22835124284029007,
      "num_tokens": 89493355.0,
      "step": 40080
    },
    {
      "entropy": 5.491171789169312,
      "epoch": 4.501937440332454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031439062388089953,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2223987713456154,
      "num_tokens": 89504253.0,
      "step": 40085
    },
    {
      "entropy": 5.393430185317993,
      "epoch": 4.502499017240411,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003143510945882475,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22483769953250884,
      "num_tokens": 89514720.0,
      "step": 40090
    },
    {
      "entropy": 5.440873146057129,
      "epoch": 4.503060594148368,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031431156404264143,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2226941019296646,
      "num_tokens": 89526583.0,
      "step": 40095
    },
    {
      "entropy": 5.46413221359253,
      "epoch": 4.5036221710563265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031427203224533996,
      "loss": 4.962,
      "mean_token_accuracy": 0.2221837416291237,
      "num_tokens": 89537791.0,
      "step": 40100
    },
    {
      "entropy": 5.547964668273925,
      "epoch": 4.504183747964284,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031423249919760167,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2200085461139679,
      "num_tokens": 89549480.0,
      "step": 40105
    },
    {
      "entropy": 5.4913959980010985,
      "epoch": 4.504745324872241,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003141929649006854,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2236751511693001,
      "num_tokens": 89559531.0,
      "step": 40110
    },
    {
      "entropy": 5.456775379180908,
      "epoch": 4.505306901780199,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031415342935585,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21901922374963761,
      "num_tokens": 89570238.0,
      "step": 40115
    },
    {
      "entropy": 5.431304502487182,
      "epoch": 4.505868478688156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003141138925643542,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22565024793148042,
      "num_tokens": 89581300.0,
      "step": 40120
    },
    {
      "entropy": 5.508476972579956,
      "epoch": 4.506430055596114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003140743545274569,
      "loss": 5.012,
      "mean_token_accuracy": 0.2200285479426384,
      "num_tokens": 89592461.0,
      "step": 40125
    },
    {
      "entropy": 5.579326009750366,
      "epoch": 4.5069916325040715,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031403481524641696,
      "loss": 5.0854,
      "mean_token_accuracy": 0.21196956187486649,
      "num_tokens": 89603380.0,
      "step": 40130
    },
    {
      "entropy": 5.5200258731842045,
      "epoch": 4.507553209412029,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031399527472249336,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22292020320892333,
      "num_tokens": 89614268.0,
      "step": 40135
    },
    {
      "entropy": 5.508366298675537,
      "epoch": 4.508114786319987,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031395573295694507,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2206068143248558,
      "num_tokens": 89625480.0,
      "step": 40140
    },
    {
      "entropy": 5.478950834274292,
      "epoch": 4.508676363227944,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031391618995103116,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21827653497457505,
      "num_tokens": 89637437.0,
      "step": 40145
    },
    {
      "entropy": 5.461371326446534,
      "epoch": 4.509237940135901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003138766457060107,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22896285653114318,
      "num_tokens": 89649196.0,
      "step": 40150
    },
    {
      "entropy": 5.384342622756958,
      "epoch": 4.509799517043859,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003138371002231427,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22693414688110353,
      "num_tokens": 89659536.0,
      "step": 40155
    },
    {
      "entropy": 5.4844386100769045,
      "epoch": 4.510361093951817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003137975535036864,
      "loss": 4.999,
      "mean_token_accuracy": 0.22144571393728257,
      "num_tokens": 89671021.0,
      "step": 40160
    },
    {
      "entropy": 5.379784631729126,
      "epoch": 4.510922670859774,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031375800554890094,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23385951966047286,
      "num_tokens": 89681689.0,
      "step": 40165
    },
    {
      "entropy": 5.457186937332153,
      "epoch": 4.511484247767732,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031371845636004556,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22237533777952195,
      "num_tokens": 89692353.0,
      "step": 40170
    },
    {
      "entropy": 5.43677659034729,
      "epoch": 4.512045824675689,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003136789059383795,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22580281645059586,
      "num_tokens": 89702852.0,
      "step": 40175
    },
    {
      "entropy": 5.521664619445801,
      "epoch": 4.512607401583647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003136393542851622,
      "loss": 5.0106,
      "mean_token_accuracy": 0.2170362502336502,
      "num_tokens": 89716125.0,
      "step": 40180
    },
    {
      "entropy": 5.453791570663452,
      "epoch": 4.5131689784916045,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031359980140165274,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22214728146791457,
      "num_tokens": 89726611.0,
      "step": 40185
    },
    {
      "entropy": 5.480514430999756,
      "epoch": 4.513730555399562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003135602472891106,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2248659461736679,
      "num_tokens": 89738563.0,
      "step": 40190
    },
    {
      "entropy": 5.431926107406616,
      "epoch": 4.51429213230752,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031352069194879527,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22407346665859224,
      "num_tokens": 89748842.0,
      "step": 40195
    },
    {
      "entropy": 5.418686723709106,
      "epoch": 4.514853709215477,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003134811353819662,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22626644223928452,
      "num_tokens": 89760747.0,
      "step": 40200
    },
    {
      "entropy": 5.428221416473389,
      "epoch": 4.515415286123434,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003134415775898828,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22835561037063598,
      "num_tokens": 89773059.0,
      "step": 40205
    },
    {
      "entropy": 5.459879684448242,
      "epoch": 4.515976863031392,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003134020185738047,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23220747709274292,
      "num_tokens": 89783347.0,
      "step": 40210
    },
    {
      "entropy": 5.462179803848267,
      "epoch": 4.51653843993935,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031336245833499146,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2294219046831131,
      "num_tokens": 89793094.0,
      "step": 40215
    },
    {
      "entropy": 5.459250020980835,
      "epoch": 4.517100016847307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003133228968747026,
      "loss": 4.9526,
      "mean_token_accuracy": 0.2206479474902153,
      "num_tokens": 89804591.0,
      "step": 40220
    },
    {
      "entropy": 5.49304461479187,
      "epoch": 4.517661593755265,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003132833341941979,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2201897144317627,
      "num_tokens": 89815749.0,
      "step": 40225
    },
    {
      "entropy": 5.530298089981079,
      "epoch": 4.518223170663222,
      "grad_norm": 0.953125,
      "learning_rate": 0.000313243770294737,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2201068416237831,
      "num_tokens": 89826940.0,
      "step": 40230
    },
    {
      "entropy": 5.559852266311646,
      "epoch": 4.51878474757118,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031320420517757955,
      "loss": 5.033,
      "mean_token_accuracy": 0.21716344952583314,
      "num_tokens": 89838955.0,
      "step": 40235
    },
    {
      "entropy": 5.434154891967774,
      "epoch": 4.519346324479137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003131646388439854,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21849194467067717,
      "num_tokens": 89850177.0,
      "step": 40240
    },
    {
      "entropy": 5.420894432067871,
      "epoch": 4.519907901387095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003131250712952144,
      "loss": 4.949,
      "mean_token_accuracy": 0.22039094269275666,
      "num_tokens": 89861185.0,
      "step": 40245
    },
    {
      "entropy": 5.533088731765747,
      "epoch": 4.520469478295053,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031308550253252634,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2246254190802574,
      "num_tokens": 89873000.0,
      "step": 40250
    },
    {
      "entropy": 5.470024538040161,
      "epoch": 4.52103105520301,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00031304593255718113,
      "loss": 4.9189,
      "mean_token_accuracy": 0.2279578536748886,
      "num_tokens": 89884223.0,
      "step": 40255
    },
    {
      "entropy": 5.558011484146118,
      "epoch": 4.521592632110968,
      "grad_norm": 1.125,
      "learning_rate": 0.00031300636137043857,
      "loss": 5.0397,
      "mean_token_accuracy": 0.22098381519317628,
      "num_tokens": 89895655.0,
      "step": 40260
    },
    {
      "entropy": 5.450252199172974,
      "epoch": 4.522154209018925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031296678897355877,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22019732594490052,
      "num_tokens": 89906475.0,
      "step": 40265
    },
    {
      "entropy": 5.427516508102417,
      "epoch": 4.5227157859268825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003129272153678017,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2250470131635666,
      "num_tokens": 89917523.0,
      "step": 40270
    },
    {
      "entropy": 5.356239271163941,
      "epoch": 4.523277362834841,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003128876405544274,
      "loss": 4.7865,
      "mean_token_accuracy": 0.23471252620220184,
      "num_tokens": 89928106.0,
      "step": 40275
    },
    {
      "entropy": 5.421920537948608,
      "epoch": 4.523838939742798,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031284806453469596,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23388457000255586,
      "num_tokens": 89938638.0,
      "step": 40280
    },
    {
      "entropy": 5.434865188598633,
      "epoch": 4.524400516650755,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031280848730986735,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22891588807106017,
      "num_tokens": 89949760.0,
      "step": 40285
    },
    {
      "entropy": 5.538129377365112,
      "epoch": 4.524962093558713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003127689088812019,
      "loss": 5.0569,
      "mean_token_accuracy": 0.22157867699861528,
      "num_tokens": 89960578.0,
      "step": 40290
    },
    {
      "entropy": 5.403190565109253,
      "epoch": 4.52552367046667,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003127293292499598,
      "loss": 4.8944,
      "mean_token_accuracy": 0.23044501692056657,
      "num_tokens": 89971370.0,
      "step": 40295
    },
    {
      "entropy": 5.420849323272705,
      "epoch": 4.526085247374628,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003126897484174012,
      "loss": 4.8596,
      "mean_token_accuracy": 0.227194744348526,
      "num_tokens": 89982068.0,
      "step": 40300
    },
    {
      "entropy": 5.45878415107727,
      "epoch": 4.526646824282586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003126501663847863,
      "loss": 4.8466,
      "mean_token_accuracy": 0.2325056254863739,
      "num_tokens": 89993108.0,
      "step": 40305
    },
    {
      "entropy": 5.457764434814453,
      "epoch": 4.527208401190543,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003126105831533756,
      "loss": 4.9337,
      "mean_token_accuracy": 0.2177932545542717,
      "num_tokens": 90004604.0,
      "step": 40310
    },
    {
      "entropy": 5.505823945999145,
      "epoch": 4.527769978098501,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003125709987244293,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21996448040008545,
      "num_tokens": 90015241.0,
      "step": 40315
    },
    {
      "entropy": 5.359634447097778,
      "epoch": 4.528331555006458,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031253141309920784,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23214155286550522,
      "num_tokens": 90026511.0,
      "step": 40320
    },
    {
      "entropy": 5.357134532928467,
      "epoch": 4.5288931319144154,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031249182627897165,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23342389911413192,
      "num_tokens": 90038994.0,
      "step": 40325
    },
    {
      "entropy": 5.476423072814941,
      "epoch": 4.5294547088223736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031245223826498116,
      "loss": 4.9685,
      "mean_token_accuracy": 0.215977743268013,
      "num_tokens": 90049183.0,
      "step": 40330
    },
    {
      "entropy": 5.4687083721160885,
      "epoch": 4.530016285730331,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031241264905849687,
      "loss": 4.895,
      "mean_token_accuracy": 0.22717846930027008,
      "num_tokens": 90060642.0,
      "step": 40335
    },
    {
      "entropy": 5.473373985290527,
      "epoch": 4.530577862638288,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031237305866077926,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21769315153360366,
      "num_tokens": 90069709.0,
      "step": 40340
    },
    {
      "entropy": 5.444840669631958,
      "epoch": 4.531139439546246,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031233346707308906,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22170692831277847,
      "num_tokens": 90080219.0,
      "step": 40345
    },
    {
      "entropy": 5.484726095199585,
      "epoch": 4.531701016454203,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031229387429668675,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21833734661340715,
      "num_tokens": 90091769.0,
      "step": 40350
    },
    {
      "entropy": 5.511235761642456,
      "epoch": 4.5322625933621605,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031225428033283305,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22703082263469695,
      "num_tokens": 90102273.0,
      "step": 40355
    },
    {
      "entropy": 5.4857275009155275,
      "epoch": 4.532824170270119,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003122146851827886,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2170167624950409,
      "num_tokens": 90112311.0,
      "step": 40360
    },
    {
      "entropy": 5.359253692626953,
      "epoch": 4.533385747178076,
      "grad_norm": 1.0,
      "learning_rate": 0.00031217508884781413,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22847431600093843,
      "num_tokens": 90123361.0,
      "step": 40365
    },
    {
      "entropy": 5.387179136276245,
      "epoch": 4.533947324086034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003121354913291704,
      "loss": 4.8454,
      "mean_token_accuracy": 0.2276599586009979,
      "num_tokens": 90134811.0,
      "step": 40370
    },
    {
      "entropy": 5.471611070632934,
      "epoch": 4.534508900993991,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003120958926281182,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22347267866134643,
      "num_tokens": 90145218.0,
      "step": 40375
    },
    {
      "entropy": 5.5336127281188965,
      "epoch": 4.535070477901948,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031205629274591844,
      "loss": 5.018,
      "mean_token_accuracy": 0.22260277718305588,
      "num_tokens": 90157217.0,
      "step": 40380
    },
    {
      "entropy": 5.506981229782104,
      "epoch": 4.5356320548099065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031201669168383193,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22575457394123077,
      "num_tokens": 90167752.0,
      "step": 40385
    },
    {
      "entropy": 5.43422155380249,
      "epoch": 4.536193631717864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031197708944311964,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23223264813423156,
      "num_tokens": 90178309.0,
      "step": 40390
    },
    {
      "entropy": 5.371665000915527,
      "epoch": 4.536755208625821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003119374860250425,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23473186194896697,
      "num_tokens": 90189158.0,
      "step": 40395
    },
    {
      "entropy": 5.474858379364013,
      "epoch": 4.537316785533779,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003118978814308614,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22570581883192062,
      "num_tokens": 90199611.0,
      "step": 40400
    },
    {
      "entropy": 5.516820764541626,
      "epoch": 4.537878362441736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003118582756618375,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22325025051832198,
      "num_tokens": 90210852.0,
      "step": 40405
    },
    {
      "entropy": 5.489544677734375,
      "epoch": 4.5384399393496935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003118186687192318,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21984528601169587,
      "num_tokens": 90221193.0,
      "step": 40410
    },
    {
      "entropy": 5.469363975524902,
      "epoch": 4.539001516257652,
      "grad_norm": 1.125,
      "learning_rate": 0.00031177906060430536,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21606320589780809,
      "num_tokens": 90232811.0,
      "step": 40415
    },
    {
      "entropy": 5.473674869537353,
      "epoch": 4.539563093165609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031173945131831936,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22421042770147323,
      "num_tokens": 90243566.0,
      "step": 40420
    },
    {
      "entropy": 5.503416347503662,
      "epoch": 4.540124670073567,
      "grad_norm": 1.125,
      "learning_rate": 0.000311699840862535,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22609136253595352,
      "num_tokens": 90255292.0,
      "step": 40425
    },
    {
      "entropy": 5.472876930236817,
      "epoch": 4.540686246981524,
      "grad_norm": 1.0,
      "learning_rate": 0.0003116602292382135,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22223843783140182,
      "num_tokens": 90266691.0,
      "step": 40430
    },
    {
      "entropy": 5.467611646652221,
      "epoch": 4.541247823889481,
      "grad_norm": 1.0,
      "learning_rate": 0.00031162061644661606,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22920531183481216,
      "num_tokens": 90278240.0,
      "step": 40435
    },
    {
      "entropy": 5.531290197372437,
      "epoch": 4.5418094007974394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000311581002489004,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22457275390625,
      "num_tokens": 90288203.0,
      "step": 40440
    },
    {
      "entropy": 5.453151226043701,
      "epoch": 4.542370977705397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003115413873666387,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2213039755821228,
      "num_tokens": 90299326.0,
      "step": 40445
    },
    {
      "entropy": 5.436215400695801,
      "epoch": 4.542932554613355,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003115017710807813,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22550439983606338,
      "num_tokens": 90309504.0,
      "step": 40450
    },
    {
      "entropy": 5.43272442817688,
      "epoch": 4.543494131521312,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003114621536326934,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22715132385492326,
      "num_tokens": 90320919.0,
      "step": 40455
    },
    {
      "entropy": 5.413691520690918,
      "epoch": 4.544055708429269,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003114225350236365,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2368725433945656,
      "num_tokens": 90331864.0,
      "step": 40460
    },
    {
      "entropy": 5.4379761695861815,
      "epoch": 4.544617285337227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003113829152548719,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2217637449502945,
      "num_tokens": 90342563.0,
      "step": 40465
    },
    {
      "entropy": 5.406955718994141,
      "epoch": 4.5451788622451845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003113432943276611,
      "loss": 4.8302,
      "mean_token_accuracy": 0.2282145008444786,
      "num_tokens": 90353464.0,
      "step": 40470
    },
    {
      "entropy": 5.432149076461792,
      "epoch": 4.545740439153142,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031130367224326585,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22739073038101196,
      "num_tokens": 90363255.0,
      "step": 40475
    },
    {
      "entropy": 5.505935192108154,
      "epoch": 4.5463020160611,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003112640490029475,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2226555034518242,
      "num_tokens": 90374478.0,
      "step": 40480
    },
    {
      "entropy": 5.50180606842041,
      "epoch": 4.546863592969057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031122442460796774,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22842514514923096,
      "num_tokens": 90386916.0,
      "step": 40485
    },
    {
      "entropy": 5.37881064414978,
      "epoch": 4.547425169877014,
      "grad_norm": 1.125,
      "learning_rate": 0.0003111847990595883,
      "loss": 4.8376,
      "mean_token_accuracy": 0.2284505620598793,
      "num_tokens": 90397407.0,
      "step": 40490
    },
    {
      "entropy": 5.49995789527893,
      "epoch": 4.547986746784972,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003111451723590708,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21681496053934096,
      "num_tokens": 90408762.0,
      "step": 40495
    },
    {
      "entropy": 5.47885308265686,
      "epoch": 4.54854832369293,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000311105544507677,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21695224642753602,
      "num_tokens": 90418942.0,
      "step": 40500
    },
    {
      "entropy": 5.599104642868042,
      "epoch": 4.549109900600888,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031106591550666865,
      "loss": 5.1055,
      "mean_token_accuracy": 0.21053329855203629,
      "num_tokens": 90430052.0,
      "step": 40505
    },
    {
      "entropy": 5.567095041275024,
      "epoch": 4.549671477508845,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003110262853573076,
      "loss": 5.0433,
      "mean_token_accuracy": 0.22364627569913864,
      "num_tokens": 90442631.0,
      "step": 40510
    },
    {
      "entropy": 5.4917521476745605,
      "epoch": 4.550233054416802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003109866540608556,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22580652683973312,
      "num_tokens": 90454429.0,
      "step": 40515
    },
    {
      "entropy": 5.479446077346802,
      "epoch": 4.55079463132476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031094702161857455,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21480597257614137,
      "num_tokens": 90466367.0,
      "step": 40520
    },
    {
      "entropy": 5.457577323913574,
      "epoch": 4.5513562082327175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031090738803172646,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22836467623710632,
      "num_tokens": 90478480.0,
      "step": 40525
    },
    {
      "entropy": 5.448171901702881,
      "epoch": 4.551917785140675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031086775330157324,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21764908134937286,
      "num_tokens": 90490501.0,
      "step": 40530
    },
    {
      "entropy": 5.436358880996704,
      "epoch": 4.552479362048633,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003108281174293768,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22313677072525023,
      "num_tokens": 90500647.0,
      "step": 40535
    },
    {
      "entropy": 5.394740009307862,
      "epoch": 4.55304093895659,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031078848041639926,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23098958730697633,
      "num_tokens": 90510963.0,
      "step": 40540
    },
    {
      "entropy": 5.496988821029663,
      "epoch": 4.553602515864547,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031074884226390246,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22360682785511016,
      "num_tokens": 90523475.0,
      "step": 40545
    },
    {
      "entropy": 5.466540145874023,
      "epoch": 4.554164092772505,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003107092029731488,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22420813590288163,
      "num_tokens": 90535348.0,
      "step": 40550
    },
    {
      "entropy": 5.405141115188599,
      "epoch": 4.5547256696804626,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031066956254540025,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22538721263408662,
      "num_tokens": 90545841.0,
      "step": 40555
    },
    {
      "entropy": 5.47594518661499,
      "epoch": 4.555287246588421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031062992098191897,
      "loss": 4.9799,
      "mean_token_accuracy": 0.22009463906288146,
      "num_tokens": 90557221.0,
      "step": 40560
    },
    {
      "entropy": 5.394782018661499,
      "epoch": 4.555848823496378,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003105902782839672,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2286055028438568,
      "num_tokens": 90568080.0,
      "step": 40565
    },
    {
      "entropy": 5.519039011001587,
      "epoch": 4.556410400404335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003105506344528071,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2223922371864319,
      "num_tokens": 90579075.0,
      "step": 40570
    },
    {
      "entropy": 5.490881872177124,
      "epoch": 4.556971977312293,
      "grad_norm": 1.125,
      "learning_rate": 0.000310510989489701,
      "loss": 5.0073,
      "mean_token_accuracy": 0.22263161838054657,
      "num_tokens": 90589937.0,
      "step": 40575
    },
    {
      "entropy": 5.412973546981812,
      "epoch": 4.55753355422025,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031047134339591136,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23307521045207977,
      "num_tokens": 90600262.0,
      "step": 40580
    },
    {
      "entropy": 5.52299919128418,
      "epoch": 4.558095131128208,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031043169617270033,
      "loss": 5.0315,
      "mean_token_accuracy": 0.22245253026485443,
      "num_tokens": 90612161.0,
      "step": 40585
    },
    {
      "entropy": 5.519643640518188,
      "epoch": 4.558656708036166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031039204782133033,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22115375399589537,
      "num_tokens": 90624304.0,
      "step": 40590
    },
    {
      "entropy": 5.46839017868042,
      "epoch": 4.559218284944123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003103523983430638,
      "loss": 4.923,
      "mean_token_accuracy": 0.2160951554775238,
      "num_tokens": 90636196.0,
      "step": 40595
    },
    {
      "entropy": 5.444388580322266,
      "epoch": 4.55977986185208,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031031274773916326,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22800838351249694,
      "num_tokens": 90649024.0,
      "step": 40600
    },
    {
      "entropy": 5.439424657821656,
      "epoch": 4.560341438760038,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031027309601089104,
      "loss": 4.9176,
      "mean_token_accuracy": 0.21568093448877335,
      "num_tokens": 90661110.0,
      "step": 40605
    },
    {
      "entropy": 5.4203362464904785,
      "epoch": 4.5609030156679955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003102334431595098,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22487906217575074,
      "num_tokens": 90671691.0,
      "step": 40610
    },
    {
      "entropy": 5.461644029617309,
      "epoch": 4.561464592575954,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031019378918628203,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23179313391447068,
      "num_tokens": 90683147.0,
      "step": 40615
    },
    {
      "entropy": 5.512072563171387,
      "epoch": 4.562026169483911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031015413409247046,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22659795582294465,
      "num_tokens": 90694614.0,
      "step": 40620
    },
    {
      "entropy": 5.480062198638916,
      "epoch": 4.562587746391868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031011447787933756,
      "loss": 4.9973,
      "mean_token_accuracy": 0.22257253229618074,
      "num_tokens": 90705326.0,
      "step": 40625
    },
    {
      "entropy": 5.441899061203003,
      "epoch": 4.563149323299826,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003100748205481461,
      "loss": 4.9401,
      "mean_token_accuracy": 0.2257190078496933,
      "num_tokens": 90716140.0,
      "step": 40630
    },
    {
      "entropy": 5.346983242034912,
      "epoch": 4.563710900207783,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003100351621001587,
      "loss": 4.7455,
      "mean_token_accuracy": 0.24131762534379958,
      "num_tokens": 90727181.0,
      "step": 40635
    },
    {
      "entropy": 5.396084833145141,
      "epoch": 4.5642724771157415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003099955025366382,
      "loss": 4.942,
      "mean_token_accuracy": 0.22593189626932145,
      "num_tokens": 90737931.0,
      "step": 40640
    },
    {
      "entropy": 5.40970401763916,
      "epoch": 4.564834054023699,
      "grad_norm": 1.125,
      "learning_rate": 0.00030995584185884733,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22465057969093322,
      "num_tokens": 90749011.0,
      "step": 40645
    },
    {
      "entropy": 5.419178867340088,
      "epoch": 4.565395630931656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003099161800680488,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22316871583461761,
      "num_tokens": 90760413.0,
      "step": 40650
    },
    {
      "entropy": 5.457279205322266,
      "epoch": 4.565957207839614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003098765171655056,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21810862571001052,
      "num_tokens": 90770973.0,
      "step": 40655
    },
    {
      "entropy": 5.498063707351685,
      "epoch": 4.566518784747571,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003098368531524806,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2289290651679039,
      "num_tokens": 90782204.0,
      "step": 40660
    },
    {
      "entropy": 5.484995269775391,
      "epoch": 4.567080361655528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003097971880302367,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23143701553344725,
      "num_tokens": 90791906.0,
      "step": 40665
    },
    {
      "entropy": 5.439677953720093,
      "epoch": 4.5676419385634865,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030975752180003674,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23034979552030563,
      "num_tokens": 90802153.0,
      "step": 40670
    },
    {
      "entropy": 5.39012770652771,
      "epoch": 4.568203515471444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000309717854463144,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22366369217634202,
      "num_tokens": 90813129.0,
      "step": 40675
    },
    {
      "entropy": 5.491723918914795,
      "epoch": 4.568765092379401,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003096781860208211,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2207631215453148,
      "num_tokens": 90823445.0,
      "step": 40680
    },
    {
      "entropy": 5.463934326171875,
      "epoch": 4.569326669287359,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003096385164743314,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22553213983774184,
      "num_tokens": 90834823.0,
      "step": 40685
    },
    {
      "entropy": 5.3581061363220215,
      "epoch": 4.569888246195316,
      "grad_norm": 1.0,
      "learning_rate": 0.00030959884582493793,
      "loss": 4.795,
      "mean_token_accuracy": 0.23424050062894822,
      "num_tokens": 90845503.0,
      "step": 40690
    },
    {
      "entropy": 5.51166934967041,
      "epoch": 4.570449823103274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003095591740739038,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2201641783118248,
      "num_tokens": 90857562.0,
      "step": 40695
    },
    {
      "entropy": 5.510186815261841,
      "epoch": 4.571011400011232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003095195012224921,
      "loss": 4.9955,
      "mean_token_accuracy": 0.22411537170410156,
      "num_tokens": 90868207.0,
      "step": 40700
    },
    {
      "entropy": 5.410302019119262,
      "epoch": 4.571572976919189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003094798272719661,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2245311215519905,
      "num_tokens": 90879710.0,
      "step": 40705
    },
    {
      "entropy": 5.476915550231934,
      "epoch": 4.572134553827147,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030944015222358906,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22623330950737,
      "num_tokens": 90890702.0,
      "step": 40710
    },
    {
      "entropy": 5.446897792816162,
      "epoch": 4.572696130735104,
      "grad_norm": 1.0,
      "learning_rate": 0.00030940047607862415,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22814477235078812,
      "num_tokens": 90902171.0,
      "step": 40715
    },
    {
      "entropy": 5.460239601135254,
      "epoch": 4.573257707643061,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003093607988383348,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22337245643138887,
      "num_tokens": 90913735.0,
      "step": 40720
    },
    {
      "entropy": 5.44628267288208,
      "epoch": 4.5738192845510195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003093211205039842,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22480099201202391,
      "num_tokens": 90924137.0,
      "step": 40725
    },
    {
      "entropy": 5.413126754760742,
      "epoch": 4.574380861458977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003092814410768359,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23101178854703902,
      "num_tokens": 90935214.0,
      "step": 40730
    },
    {
      "entropy": 5.4115783214569095,
      "epoch": 4.574942438366934,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030924176055815327,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23168229311704636,
      "num_tokens": 90946797.0,
      "step": 40735
    },
    {
      "entropy": 5.429314088821411,
      "epoch": 4.575504015274892,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030920207894919957,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2256943702697754,
      "num_tokens": 90958055.0,
      "step": 40740
    },
    {
      "entropy": 5.391261434555053,
      "epoch": 4.576065592182849,
      "grad_norm": 1.0,
      "learning_rate": 0.0003091623962512385,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22994834780693055,
      "num_tokens": 90969516.0,
      "step": 40745
    },
    {
      "entropy": 5.339202928543091,
      "epoch": 4.576627169090807,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030912271246553346,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22264113277196884,
      "num_tokens": 90980536.0,
      "step": 40750
    },
    {
      "entropy": 5.462917613983154,
      "epoch": 4.577188745998765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000309083027593348,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22792242765426635,
      "num_tokens": 90992150.0,
      "step": 40755
    },
    {
      "entropy": 5.478546190261841,
      "epoch": 4.577750322906722,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030904334163594577,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22149543464183807,
      "num_tokens": 91003323.0,
      "step": 40760
    },
    {
      "entropy": 5.429043388366699,
      "epoch": 4.57831189981468,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003090036545945904,
      "loss": 4.8292,
      "mean_token_accuracy": 0.2284099817276001,
      "num_tokens": 91014238.0,
      "step": 40765
    },
    {
      "entropy": 5.445560264587402,
      "epoch": 4.578873476722637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030896396647054535,
      "loss": 4.9391,
      "mean_token_accuracy": 0.224538354575634,
      "num_tokens": 91025375.0,
      "step": 40770
    },
    {
      "entropy": 5.39719524383545,
      "epoch": 4.579435053630594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003089242772650744,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2208469271659851,
      "num_tokens": 91036498.0,
      "step": 40775
    },
    {
      "entropy": 5.417607641220092,
      "epoch": 4.579996630538552,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030888458697944144,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22948740273714066,
      "num_tokens": 91046909.0,
      "step": 40780
    },
    {
      "entropy": 5.525335836410522,
      "epoch": 4.58055820744651,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00030884489561491005,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21786029189825057,
      "num_tokens": 91056052.0,
      "step": 40785
    },
    {
      "entropy": 5.657703828811646,
      "epoch": 4.581119784354467,
      "grad_norm": 1.046875,
      "learning_rate": 0.000308805203172744,
      "loss": 5.1239,
      "mean_token_accuracy": 0.21118472665548324,
      "num_tokens": 91067614.0,
      "step": 40790
    },
    {
      "entropy": 5.405837106704712,
      "epoch": 4.581681361262425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030876550965420725,
      "loss": 4.8665,
      "mean_token_accuracy": 0.23103329986333848,
      "num_tokens": 91078539.0,
      "step": 40795
    },
    {
      "entropy": 5.43439474105835,
      "epoch": 4.582242938170382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003087258150605635,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2244063764810562,
      "num_tokens": 91088909.0,
      "step": 40800
    },
    {
      "entropy": 5.425942993164062,
      "epoch": 4.58280451507834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003086861193930767,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2236924946308136,
      "num_tokens": 91101073.0,
      "step": 40805
    },
    {
      "entropy": 5.444519662857056,
      "epoch": 4.5833660919862975,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003086464226530109,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22395261973142624,
      "num_tokens": 91113518.0,
      "step": 40810
    },
    {
      "entropy": 5.491447496414184,
      "epoch": 4.583927668894255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030860672484162987,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22434968054294585,
      "num_tokens": 91124287.0,
      "step": 40815
    },
    {
      "entropy": 5.533636903762817,
      "epoch": 4.584489245802213,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003085670259601977,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22094009071588516,
      "num_tokens": 91134828.0,
      "step": 40820
    },
    {
      "entropy": 5.502436542510987,
      "epoch": 4.58505082271017,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030852732600997846,
      "loss": 4.9903,
      "mean_token_accuracy": 0.2164643034338951,
      "num_tokens": 91146647.0,
      "step": 40825
    },
    {
      "entropy": 5.430188417434692,
      "epoch": 4.585612399618128,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030848762499223616,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2206990510225296,
      "num_tokens": 91156769.0,
      "step": 40830
    },
    {
      "entropy": 5.466915655136108,
      "epoch": 4.586173976526085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003084479229082348,
      "loss": 4.89,
      "mean_token_accuracy": 0.22599827498197556,
      "num_tokens": 91168876.0,
      "step": 40835
    },
    {
      "entropy": 5.558495330810547,
      "epoch": 4.586735553434043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003084082197592387,
      "loss": 5.0445,
      "mean_token_accuracy": 0.22136587053537368,
      "num_tokens": 91178587.0,
      "step": 40840
    },
    {
      "entropy": 5.421411561965942,
      "epoch": 4.587297130342001,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030836851554651184,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2368401348590851,
      "num_tokens": 91190210.0,
      "step": 40845
    },
    {
      "entropy": 5.459425449371338,
      "epoch": 4.587858707249958,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030832881027131854,
      "loss": 4.9546,
      "mean_token_accuracy": 0.22767448723316192,
      "num_tokens": 91200921.0,
      "step": 40850
    },
    {
      "entropy": 5.406929683685303,
      "epoch": 4.588420284157915,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030828910393492306,
      "loss": 4.8836,
      "mean_token_accuracy": 0.23053905963897706,
      "num_tokens": 91211553.0,
      "step": 40855
    },
    {
      "entropy": 5.496520328521728,
      "epoch": 4.588981861065873,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030824939653858953,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22470089197158813,
      "num_tokens": 91222224.0,
      "step": 40860
    },
    {
      "entropy": 5.50420823097229,
      "epoch": 4.5895434379738305,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030820968808358223,
      "loss": 4.9087,
      "mean_token_accuracy": 0.21877786368131638,
      "num_tokens": 91233645.0,
      "step": 40865
    },
    {
      "entropy": 5.503484582901001,
      "epoch": 4.590105014881788,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030816997857116573,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22498093098402022,
      "num_tokens": 91245194.0,
      "step": 40870
    },
    {
      "entropy": 5.421435308456421,
      "epoch": 4.590666591789746,
      "grad_norm": 1.0,
      "learning_rate": 0.0003081302680026042,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2229336455464363,
      "num_tokens": 91255124.0,
      "step": 40875
    },
    {
      "entropy": 5.503187561035157,
      "epoch": 4.591228168697703,
      "grad_norm": 1.140625,
      "learning_rate": 0.000308090556379162,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21645701080560684,
      "num_tokens": 91267129.0,
      "step": 40880
    },
    {
      "entropy": 5.453730297088623,
      "epoch": 4.591789745605661,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030805084370210377,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22386665195226668,
      "num_tokens": 91277614.0,
      "step": 40885
    },
    {
      "entropy": 5.512607002258301,
      "epoch": 4.592351322513618,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030801112997269386,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21659952849149705,
      "num_tokens": 91289856.0,
      "step": 40890
    },
    {
      "entropy": 5.398402357101441,
      "epoch": 4.5929128994215755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003079714151921967,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2218829110264778,
      "num_tokens": 91300593.0,
      "step": 40895
    },
    {
      "entropy": 5.448042535781861,
      "epoch": 4.593474476329534,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030793169936187695,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23587338626384735,
      "num_tokens": 91312934.0,
      "step": 40900
    },
    {
      "entropy": 5.4160514831542965,
      "epoch": 4.594036053237491,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030789198248299906,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2274225890636444,
      "num_tokens": 91323856.0,
      "step": 40905
    },
    {
      "entropy": 5.51603627204895,
      "epoch": 4.594597630145448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030785226455682773,
      "loss": 4.945,
      "mean_token_accuracy": 0.2270421102643013,
      "num_tokens": 91334545.0,
      "step": 40910
    },
    {
      "entropy": 5.508879137039185,
      "epoch": 4.595159207053406,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003078125455846276,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22625827491283418,
      "num_tokens": 91346006.0,
      "step": 40915
    },
    {
      "entropy": 5.481260824203491,
      "epoch": 4.595720783961363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030777282556766316,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21328455954790115,
      "num_tokens": 91356710.0,
      "step": 40920
    },
    {
      "entropy": 5.4455667495727536,
      "epoch": 4.596282360869321,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030773310450719937,
      "loss": 4.8854,
      "mean_token_accuracy": 0.22193866670131684,
      "num_tokens": 91367747.0,
      "step": 40925
    },
    {
      "entropy": 5.515042448043824,
      "epoch": 4.596843937777279,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003076933824045008,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2176276281476021,
      "num_tokens": 91379079.0,
      "step": 40930
    },
    {
      "entropy": 5.473390674591064,
      "epoch": 4.597405514685236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030765365926083214,
      "loss": 4.9991,
      "mean_token_accuracy": 0.22306723147630692,
      "num_tokens": 91390264.0,
      "step": 40935
    },
    {
      "entropy": 5.444041156768799,
      "epoch": 4.597967091593194,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003076139350774584,
      "loss": 4.8845,
      "mean_token_accuracy": 0.23096632659435273,
      "num_tokens": 91401322.0,
      "step": 40940
    },
    {
      "entropy": 5.425783014297485,
      "epoch": 4.598528668501151,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003075742098556444,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22536594271659852,
      "num_tokens": 91412277.0,
      "step": 40945
    },
    {
      "entropy": 5.386979722976685,
      "epoch": 4.5990902454091085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003075344835966548,
      "loss": 4.861,
      "mean_token_accuracy": 0.22910823822021484,
      "num_tokens": 91422941.0,
      "step": 40950
    },
    {
      "entropy": 5.429729557037353,
      "epoch": 4.599651822317067,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003074947563017547,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2299702301621437,
      "num_tokens": 91434194.0,
      "step": 40955
    },
    {
      "entropy": 5.430826187133789,
      "epoch": 4.600213399225024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003074550279722089,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2258673131465912,
      "num_tokens": 91445330.0,
      "step": 40960
    },
    {
      "entropy": 5.506629276275635,
      "epoch": 4.600774976132981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003074152986092825,
      "loss": 4.9365,
      "mean_token_accuracy": 0.2229177325963974,
      "num_tokens": 91456762.0,
      "step": 40965
    },
    {
      "entropy": 5.418737506866455,
      "epoch": 4.601336553040939,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003073755682142404,
      "loss": 4.8191,
      "mean_token_accuracy": 0.22971097528934478,
      "num_tokens": 91467272.0,
      "step": 40970
    },
    {
      "entropy": 5.512608528137207,
      "epoch": 4.601898129948896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030733583678834765,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2174053192138672,
      "num_tokens": 91478188.0,
      "step": 40975
    },
    {
      "entropy": 5.431144571304321,
      "epoch": 4.602459706856854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030729610433286926,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22509733736515045,
      "num_tokens": 91489759.0,
      "step": 40980
    },
    {
      "entropy": 5.505617380142212,
      "epoch": 4.603021283764812,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030725637084907037,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22806721180677414,
      "num_tokens": 91500705.0,
      "step": 40985
    },
    {
      "entropy": 5.413644075393677,
      "epoch": 4.603582860672769,
      "grad_norm": 1.0,
      "learning_rate": 0.0003072166363382162,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22141510993242264,
      "num_tokens": 91511832.0,
      "step": 40990
    },
    {
      "entropy": 5.49561882019043,
      "epoch": 4.604144437580727,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003071769008015717,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22124620527029037,
      "num_tokens": 91522454.0,
      "step": 40995
    },
    {
      "entropy": 5.50156946182251,
      "epoch": 4.604706014488684,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003071371642404024,
      "loss": 5.0102,
      "mean_token_accuracy": 0.22194377779960633,
      "num_tokens": 91533987.0,
      "step": 41000
    },
    {
      "entropy": 5.449102592468262,
      "epoch": 4.605267591396641,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003070974266559732,
      "loss": 4.927,
      "mean_token_accuracy": 0.2268700897693634,
      "num_tokens": 91545000.0,
      "step": 41005
    },
    {
      "entropy": 5.5901635646820065,
      "epoch": 4.6058291683045995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003070576880495495,
      "loss": 4.9874,
      "mean_token_accuracy": 0.222335384786129,
      "num_tokens": 91556481.0,
      "step": 41010
    },
    {
      "entropy": 5.464625453948974,
      "epoch": 4.606390745212557,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003070179484223966,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23451564759016036,
      "num_tokens": 91568668.0,
      "step": 41015
    },
    {
      "entropy": 5.43808331489563,
      "epoch": 4.606952322120515,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003069782077757797,
      "loss": 4.8858,
      "mean_token_accuracy": 0.23207415789365768,
      "num_tokens": 91579884.0,
      "step": 41020
    },
    {
      "entropy": 5.509147214889526,
      "epoch": 4.607513899028472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030693846611096437,
      "loss": 5.0536,
      "mean_token_accuracy": 0.22799602150917053,
      "num_tokens": 91590488.0,
      "step": 41025
    },
    {
      "entropy": 5.462219095230102,
      "epoch": 4.608075475936429,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030689872342921595,
      "loss": 4.934,
      "mean_token_accuracy": 0.22629468590021135,
      "num_tokens": 91602163.0,
      "step": 41030
    },
    {
      "entropy": 5.442301225662232,
      "epoch": 4.608637052844387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030685897973179967,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22634771317243577,
      "num_tokens": 91614251.0,
      "step": 41035
    },
    {
      "entropy": 5.447401332855224,
      "epoch": 4.609198629752345,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030681923501998124,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22879401743412017,
      "num_tokens": 91625102.0,
      "step": 41040
    },
    {
      "entropy": 5.500440216064453,
      "epoch": 4.609760206660302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030677948929502596,
      "loss": 4.994,
      "mean_token_accuracy": 0.22525483816862107,
      "num_tokens": 91636613.0,
      "step": 41045
    },
    {
      "entropy": 5.422806024551392,
      "epoch": 4.61032178356826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030673974255819944,
      "loss": 4.899,
      "mean_token_accuracy": 0.221329365670681,
      "num_tokens": 91647648.0,
      "step": 41050
    },
    {
      "entropy": 5.593122339248657,
      "epoch": 4.610883360476217,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003066999948107672,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21848564147949218,
      "num_tokens": 91659160.0,
      "step": 41055
    },
    {
      "entropy": 5.483921194076538,
      "epoch": 4.611444937384174,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003066602460539948,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23251988291740416,
      "num_tokens": 91670973.0,
      "step": 41060
    },
    {
      "entropy": 5.508328580856324,
      "epoch": 4.6120065142921325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030662049628914787,
      "loss": 5.0271,
      "mean_token_accuracy": 0.22527264952659606,
      "num_tokens": 91682710.0,
      "step": 41065
    },
    {
      "entropy": 5.494505167007446,
      "epoch": 4.61256809120009,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003065807455174921,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21900745630264282,
      "num_tokens": 91694008.0,
      "step": 41070
    },
    {
      "entropy": 5.401672792434693,
      "epoch": 4.613129668108048,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003065409937402932,
      "loss": 4.889,
      "mean_token_accuracy": 0.2185859888792038,
      "num_tokens": 91705289.0,
      "step": 41075
    },
    {
      "entropy": 5.4460498809814455,
      "epoch": 4.613691245016005,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030650124095881674,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22112310975790023,
      "num_tokens": 91717078.0,
      "step": 41080
    },
    {
      "entropy": 5.424749565124512,
      "epoch": 4.614252821923962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030646148717432855,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22415444254875183,
      "num_tokens": 91728971.0,
      "step": 41085
    },
    {
      "entropy": 5.433434724807739,
      "epoch": 4.61481439883192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030642173238809453,
      "loss": 4.8715,
      "mean_token_accuracy": 0.232856822013855,
      "num_tokens": 91739496.0,
      "step": 41090
    },
    {
      "entropy": 5.492396974563599,
      "epoch": 4.615375975739878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003063819766013802,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21512610614299774,
      "num_tokens": 91749732.0,
      "step": 41095
    },
    {
      "entropy": 5.435218000411988,
      "epoch": 4.615937552647835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003063422198154516,
      "loss": 4.8648,
      "mean_token_accuracy": 0.22006469815969468,
      "num_tokens": 91758772.0,
      "step": 41100
    },
    {
      "entropy": 5.426222944259644,
      "epoch": 4.616499129555793,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003063024620315746,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22828404158353804,
      "num_tokens": 91770012.0,
      "step": 41105
    },
    {
      "entropy": 5.473956203460693,
      "epoch": 4.61706070646375,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003062627032510151,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21847863495349884,
      "num_tokens": 91781926.0,
      "step": 41110
    },
    {
      "entropy": 5.514020824432373,
      "epoch": 4.617622283371707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000306222943475039,
      "loss": 5.0372,
      "mean_token_accuracy": 0.2211952582001686,
      "num_tokens": 91794598.0,
      "step": 41115
    },
    {
      "entropy": 5.462335634231567,
      "epoch": 4.618183860279665,
      "grad_norm": 1.0,
      "learning_rate": 0.0003061831827049123,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22102042883634568,
      "num_tokens": 91805898.0,
      "step": 41120
    },
    {
      "entropy": 5.48613977432251,
      "epoch": 4.618745437187623,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030614342094190095,
      "loss": 4.949,
      "mean_token_accuracy": 0.2196344405412674,
      "num_tokens": 91816369.0,
      "step": 41125
    },
    {
      "entropy": 5.369981908798218,
      "epoch": 4.619307014095581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030610365818727097,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2248953774571419,
      "num_tokens": 91827223.0,
      "step": 41130
    },
    {
      "entropy": 5.479789924621582,
      "epoch": 4.619868591003538,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003060638944422884,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21993738412857056,
      "num_tokens": 91838898.0,
      "step": 41135
    },
    {
      "entropy": 5.5014571189880375,
      "epoch": 4.620430167911495,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030602412970821944,
      "loss": 4.932,
      "mean_token_accuracy": 0.22159878611564637,
      "num_tokens": 91850660.0,
      "step": 41140
    },
    {
      "entropy": 5.488295698165894,
      "epoch": 4.620991744819453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003059843639863301,
      "loss": 4.958,
      "mean_token_accuracy": 0.22173966467380524,
      "num_tokens": 91860594.0,
      "step": 41145
    },
    {
      "entropy": 5.419105005264282,
      "epoch": 4.6215533217274105,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030594459727788665,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22427579909563064,
      "num_tokens": 91872209.0,
      "step": 41150
    },
    {
      "entropy": 5.491024160385132,
      "epoch": 4.622114898635368,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030590482958415524,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2219543382525444,
      "num_tokens": 91885241.0,
      "step": 41155
    },
    {
      "entropy": 5.587158250808716,
      "epoch": 4.622676475543326,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00030586506090640205,
      "loss": 5.0152,
      "mean_token_accuracy": 0.23169474452733993,
      "num_tokens": 91896999.0,
      "step": 41160
    },
    {
      "entropy": 5.441745281219482,
      "epoch": 4.623238052451283,
      "grad_norm": 1.0,
      "learning_rate": 0.0003058252912458933,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22360673695802688,
      "num_tokens": 91908554.0,
      "step": 41165
    },
    {
      "entropy": 5.397407817840576,
      "epoch": 4.62379962935924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003057855206038954,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2286001980304718,
      "num_tokens": 91920013.0,
      "step": 41170
    },
    {
      "entropy": 5.463139629364013,
      "epoch": 4.624361206267198,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003057457489816745,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22828772515058518,
      "num_tokens": 91930951.0,
      "step": 41175
    },
    {
      "entropy": 5.490615653991699,
      "epoch": 4.624922783175156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003057059763804971,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22791792899370195,
      "num_tokens": 91941718.0,
      "step": 41180
    },
    {
      "entropy": 5.40921835899353,
      "epoch": 4.625484360083114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003056662028016295,
      "loss": 4.8402,
      "mean_token_accuracy": 0.2289062112569809,
      "num_tokens": 91952756.0,
      "step": 41185
    },
    {
      "entropy": 5.433768224716187,
      "epoch": 4.626045936991071,
      "grad_norm": 1.0,
      "learning_rate": 0.00030562642824633805,
      "loss": 4.9338,
      "mean_token_accuracy": 0.22176509499549865,
      "num_tokens": 91964200.0,
      "step": 41190
    },
    {
      "entropy": 5.39778413772583,
      "epoch": 4.626607513899028,
      "grad_norm": 1.0,
      "learning_rate": 0.0003055866527158893,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2166033700108528,
      "num_tokens": 91975032.0,
      "step": 41195
    },
    {
      "entropy": 5.434349918365479,
      "epoch": 4.627169090806986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003055468762115497,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2239617645740509,
      "num_tokens": 91986566.0,
      "step": 41200
    },
    {
      "entropy": 5.4334112167358395,
      "epoch": 4.6277306677149435,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030550709873458567,
      "loss": 4.8652,
      "mean_token_accuracy": 0.2352556422352791,
      "num_tokens": 91996944.0,
      "step": 41205
    },
    {
      "entropy": 5.461637830734253,
      "epoch": 4.628292244622902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030546732028626373,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2324964240193367,
      "num_tokens": 92007692.0,
      "step": 41210
    },
    {
      "entropy": 5.440851831436158,
      "epoch": 4.628853821530859,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030542754086785055,
      "loss": 4.899,
      "mean_token_accuracy": 0.23231306672096252,
      "num_tokens": 92017514.0,
      "step": 41215
    },
    {
      "entropy": 5.451291227340699,
      "epoch": 4.629415398438816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003053877604806127,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2287612736225128,
      "num_tokens": 92028311.0,
      "step": 41220
    },
    {
      "entropy": 5.373444890975952,
      "epoch": 4.629976975346774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030534797912581664,
      "loss": 4.834,
      "mean_token_accuracy": 0.23543382287025452,
      "num_tokens": 92038795.0,
      "step": 41225
    },
    {
      "entropy": 5.358739805221558,
      "epoch": 4.630538552254731,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030530819680472926,
      "loss": 4.9,
      "mean_token_accuracy": 0.22328694611787797,
      "num_tokens": 92049578.0,
      "step": 41230
    },
    {
      "entropy": 5.5613843441009525,
      "epoch": 4.6311001291626885,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030526841351861717,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2138779103755951,
      "num_tokens": 92062868.0,
      "step": 41235
    },
    {
      "entropy": 5.529632663726806,
      "epoch": 4.631661706070647,
      "grad_norm": 1.0,
      "learning_rate": 0.00030522862926874697,
      "loss": 4.9456,
      "mean_token_accuracy": 0.23154184371232986,
      "num_tokens": 92073819.0,
      "step": 41240
    },
    {
      "entropy": 5.497158622741699,
      "epoch": 4.632223282978604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003051888440563855,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21472576260566711,
      "num_tokens": 92085212.0,
      "step": 41245
    },
    {
      "entropy": 5.474449634552002,
      "epoch": 4.632784859886561,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003051490578827996,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2254885643720627,
      "num_tokens": 92096807.0,
      "step": 41250
    },
    {
      "entropy": 5.436577177047729,
      "epoch": 4.633346436794519,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030510927074925594,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2297043666243553,
      "num_tokens": 92108619.0,
      "step": 41255
    },
    {
      "entropy": 5.501138496398926,
      "epoch": 4.633908013702476,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030506948265702144,
      "loss": 5.0348,
      "mean_token_accuracy": 0.2187516823410988,
      "num_tokens": 92120815.0,
      "step": 41260
    },
    {
      "entropy": 5.423004388809204,
      "epoch": 4.6344695906104345,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000305029693607363,
      "loss": 4.91,
      "mean_token_accuracy": 0.22992446571588515,
      "num_tokens": 92131957.0,
      "step": 41265
    },
    {
      "entropy": 5.416411638259888,
      "epoch": 4.635031167518392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030498990360154746,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22991663515567778,
      "num_tokens": 92143936.0,
      "step": 41270
    },
    {
      "entropy": 5.4788267612457275,
      "epoch": 4.635592744426349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030495011264084165,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22511032819747925,
      "num_tokens": 92155320.0,
      "step": 41275
    },
    {
      "entropy": 5.496634769439697,
      "epoch": 4.636154321334307,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030491032072651266,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22723082900047303,
      "num_tokens": 92166831.0,
      "step": 41280
    },
    {
      "entropy": 5.504222393035889,
      "epoch": 4.636715898242264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030487052785982757,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22023556083440782,
      "num_tokens": 92177927.0,
      "step": 41285
    },
    {
      "entropy": 5.474713850021362,
      "epoch": 4.6372774751502215,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003048307340420532,
      "loss": 4.971,
      "mean_token_accuracy": 0.22185453921556472,
      "num_tokens": 92188755.0,
      "step": 41290
    },
    {
      "entropy": 5.477785587310791,
      "epoch": 4.63783905205818,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003047909392744566,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22582642883062362,
      "num_tokens": 92199156.0,
      "step": 41295
    },
    {
      "entropy": 5.4435224533081055,
      "epoch": 4.638400628966137,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030475114355830513,
      "loss": 4.895,
      "mean_token_accuracy": 0.22558722794055938,
      "num_tokens": 92211080.0,
      "step": 41300
    },
    {
      "entropy": 5.50134768486023,
      "epoch": 4.638962205874094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030471134689486563,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2236796110868454,
      "num_tokens": 92220734.0,
      "step": 41305
    },
    {
      "entropy": 5.528839635848999,
      "epoch": 4.639523782782052,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003046715492854053,
      "loss": 5.046,
      "mean_token_accuracy": 0.21678799390792847,
      "num_tokens": 92231166.0,
      "step": 41310
    },
    {
      "entropy": 5.514361810684204,
      "epoch": 4.640085359690009,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030463175073119133,
      "loss": 4.969,
      "mean_token_accuracy": 0.21745579242706298,
      "num_tokens": 92241920.0,
      "step": 41315
    },
    {
      "entropy": 5.436991214752197,
      "epoch": 4.6406469365979675,
      "grad_norm": 1.0,
      "learning_rate": 0.00030459195123349095,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22368635535240172,
      "num_tokens": 92253745.0,
      "step": 41320
    },
    {
      "entropy": 5.448040914535523,
      "epoch": 4.641208513505925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003045521507935714,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22612234055995942,
      "num_tokens": 92263672.0,
      "step": 41325
    },
    {
      "entropy": 5.4804564952850345,
      "epoch": 4.641770090413882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003045123494126999,
      "loss": 4.9939,
      "mean_token_accuracy": 0.22234402894973754,
      "num_tokens": 92275124.0,
      "step": 41330
    },
    {
      "entropy": 5.589838600158691,
      "epoch": 4.64233166732184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003044725470921437,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21864518523216248,
      "num_tokens": 92287077.0,
      "step": 41335
    },
    {
      "entropy": 5.566555738449097,
      "epoch": 4.642893244229797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003044327438331702,
      "loss": 5.0206,
      "mean_token_accuracy": 0.22343625724315644,
      "num_tokens": 92298649.0,
      "step": 41340
    },
    {
      "entropy": 5.469246435165405,
      "epoch": 4.643454821137754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003043929396370467,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22515467405319214,
      "num_tokens": 92309589.0,
      "step": 41345
    },
    {
      "entropy": 5.346590852737426,
      "epoch": 4.6440163980457125,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030435313450504056,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22766272872686386,
      "num_tokens": 92319927.0,
      "step": 41350
    },
    {
      "entropy": 5.466430187225342,
      "epoch": 4.64457797495367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003043133284384193,
      "loss": 5.007,
      "mean_token_accuracy": 0.22052559107542039,
      "num_tokens": 92333156.0,
      "step": 41355
    },
    {
      "entropy": 5.354211711883545,
      "epoch": 4.645139551861627,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003042735214384503,
      "loss": 4.7902,
      "mean_token_accuracy": 0.23237168043851852,
      "num_tokens": 92345228.0,
      "step": 41360
    },
    {
      "entropy": 5.508993816375733,
      "epoch": 4.645701128769585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000304233713506401,
      "loss": 4.9954,
      "mean_token_accuracy": 0.22812074571847915,
      "num_tokens": 92355922.0,
      "step": 41365
    },
    {
      "entropy": 5.480924654006958,
      "epoch": 4.646262705677542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003041939046435389,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22587586790323258,
      "num_tokens": 92367255.0,
      "step": 41370
    },
    {
      "entropy": 5.397941827774048,
      "epoch": 4.6468242825855,
      "grad_norm": 1.125,
      "learning_rate": 0.00030415409485113156,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22951108664274217,
      "num_tokens": 92378577.0,
      "step": 41375
    },
    {
      "entropy": 5.459941577911377,
      "epoch": 4.647385859493458,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030411428413044657,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22166165411472322,
      "num_tokens": 92389822.0,
      "step": 41380
    },
    {
      "entropy": 5.569951677322388,
      "epoch": 4.647947436401415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030407447248275137,
      "loss": 5.0464,
      "mean_token_accuracy": 0.22016583383083344,
      "num_tokens": 92400682.0,
      "step": 41385
    },
    {
      "entropy": 5.534773921966552,
      "epoch": 4.648509013309373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003040346599093138,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2305434077978134,
      "num_tokens": 92412194.0,
      "step": 41390
    },
    {
      "entropy": 5.5042320728302006,
      "epoch": 4.64907059021733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003039948464114014,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22524748742580414,
      "num_tokens": 92423311.0,
      "step": 41395
    },
    {
      "entropy": 5.446036720275879,
      "epoch": 4.649632167125288,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030395503199028186,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2171839565038681,
      "num_tokens": 92434008.0,
      "step": 41400
    },
    {
      "entropy": 5.4064445972442625,
      "epoch": 4.6501937440332455,
      "grad_norm": 1.0,
      "learning_rate": 0.0003039152166472228,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23166422694921493,
      "num_tokens": 92445887.0,
      "step": 41405
    },
    {
      "entropy": 5.4839414119720455,
      "epoch": 4.650755320941203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003038754003834921,
      "loss": 5.0067,
      "mean_token_accuracy": 0.22992882281541824,
      "num_tokens": 92457856.0,
      "step": 41410
    },
    {
      "entropy": 5.527867078781128,
      "epoch": 4.651316897849161,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030383558320035744,
      "loss": 5.0062,
      "mean_token_accuracy": 0.22542784959077836,
      "num_tokens": 92469976.0,
      "step": 41415
    },
    {
      "entropy": 5.437243795394897,
      "epoch": 4.651878474757118,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030379576509908656,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22599224746227264,
      "num_tokens": 92481035.0,
      "step": 41420
    },
    {
      "entropy": 5.457373857498169,
      "epoch": 4.652440051665075,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030375594608094737,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2243888035416603,
      "num_tokens": 92492145.0,
      "step": 41425
    },
    {
      "entropy": 5.474706697463989,
      "epoch": 4.653001628573033,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003037161261472078,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21830472350120544,
      "num_tokens": 92504541.0,
      "step": 41430
    },
    {
      "entropy": 5.531746292114258,
      "epoch": 4.653563205480991,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030367630529913554,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21907250285148622,
      "num_tokens": 92516911.0,
      "step": 41435
    },
    {
      "entropy": 5.522646045684814,
      "epoch": 4.654124782388948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030363648353799857,
      "loss": 4.974,
      "mean_token_accuracy": 0.2257963702082634,
      "num_tokens": 92526619.0,
      "step": 41440
    },
    {
      "entropy": 5.5586542129516605,
      "epoch": 4.654686359296906,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030359666086506493,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2216309666633606,
      "num_tokens": 92536579.0,
      "step": 41445
    },
    {
      "entropy": 5.4694969177246096,
      "epoch": 4.655247936204863,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030355683728160247,
      "loss": 4.8978,
      "mean_token_accuracy": 0.23194750696420668,
      "num_tokens": 92548147.0,
      "step": 41450
    },
    {
      "entropy": 5.455047988891602,
      "epoch": 4.655809513112821,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003035170127888793,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2317287564277649,
      "num_tokens": 92559335.0,
      "step": 41455
    },
    {
      "entropy": 5.4144515037536625,
      "epoch": 4.656371090020778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030347718738816333,
      "loss": 4.8698,
      "mean_token_accuracy": 0.21929376125335692,
      "num_tokens": 92571527.0,
      "step": 41460
    },
    {
      "entropy": 5.487369680404663,
      "epoch": 4.656932666928736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030343736108072266,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2305314302444458,
      "num_tokens": 92581996.0,
      "step": 41465
    },
    {
      "entropy": 5.457761478424072,
      "epoch": 4.657494243836694,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003033975338678254,
      "loss": 4.8942,
      "mean_token_accuracy": 0.23184183090925217,
      "num_tokens": 92593134.0,
      "step": 41470
    },
    {
      "entropy": 5.449653196334839,
      "epoch": 4.658055820744651,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00030335770575073967,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22109778970479965,
      "num_tokens": 92605196.0,
      "step": 41475
    },
    {
      "entropy": 5.411291122436523,
      "epoch": 4.658617397652608,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003033178767307336,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2232678547501564,
      "num_tokens": 92615151.0,
      "step": 41480
    },
    {
      "entropy": 5.45254111289978,
      "epoch": 4.659178974560566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003032780468090753,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21937613040208817,
      "num_tokens": 92625723.0,
      "step": 41485
    },
    {
      "entropy": 5.476891899108887,
      "epoch": 4.6597405514685235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003032382159870331,
      "loss": 5.0158,
      "mean_token_accuracy": 0.22984239906072618,
      "num_tokens": 92636699.0,
      "step": 41490
    },
    {
      "entropy": 5.457426309585571,
      "epoch": 4.660302128376481,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030319838426587505,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2205788090825081,
      "num_tokens": 92646977.0,
      "step": 41495
    },
    {
      "entropy": 5.423472452163696,
      "epoch": 4.660863705284439,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030315855164686956,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2311264008283615,
      "num_tokens": 92658021.0,
      "step": 41500
    },
    {
      "entropy": 5.415776157379151,
      "epoch": 4.661425282192396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003031187181312848,
      "loss": 4.8382,
      "mean_token_accuracy": 0.23514327108860017,
      "num_tokens": 92668573.0,
      "step": 41505
    },
    {
      "entropy": 5.441014575958252,
      "epoch": 4.661986859100354,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030307888372038927,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2340692847967148,
      "num_tokens": 92680286.0,
      "step": 41510
    },
    {
      "entropy": 5.428313779830932,
      "epoch": 4.662548436008311,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030303904841545113,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22055338621139525,
      "num_tokens": 92692309.0,
      "step": 41515
    },
    {
      "entropy": 5.388029146194458,
      "epoch": 4.663110012916269,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003029992122177388,
      "loss": 4.882,
      "mean_token_accuracy": 0.2321081802248955,
      "num_tokens": 92703194.0,
      "step": 41520
    },
    {
      "entropy": 5.589586305618286,
      "epoch": 4.663671589824227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030295937512852075,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21646709740161896,
      "num_tokens": 92714218.0,
      "step": 41525
    },
    {
      "entropy": 5.481814813613892,
      "epoch": 4.664233166732184,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003029195371490652,
      "loss": 4.9035,
      "mean_token_accuracy": 0.23250290602445603,
      "num_tokens": 92726422.0,
      "step": 41530
    },
    {
      "entropy": 5.4706755638122555,
      "epoch": 4.664794743640142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030287969828064087,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22286178767681122,
      "num_tokens": 92736835.0,
      "step": 41535
    },
    {
      "entropy": 5.456659317016602,
      "epoch": 4.665356320548099,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030283985852451624,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23604504764080048,
      "num_tokens": 92746665.0,
      "step": 41540
    },
    {
      "entropy": 5.4743509769439695,
      "epoch": 4.6659178974560565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003028000178819596,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22648150771856307,
      "num_tokens": 92758059.0,
      "step": 41545
    },
    {
      "entropy": 5.397711277008057,
      "epoch": 4.666479474364014,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030276017635423965,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23087209314107895,
      "num_tokens": 92768982.0,
      "step": 41550
    },
    {
      "entropy": 5.38611364364624,
      "epoch": 4.667041051271972,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030272033394262483,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23151155412197114,
      "num_tokens": 92779117.0,
      "step": 41555
    },
    {
      "entropy": 5.444413852691651,
      "epoch": 4.667602628179929,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030268049064838395,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22439152449369432,
      "num_tokens": 92790595.0,
      "step": 41560
    },
    {
      "entropy": 5.509378290176391,
      "epoch": 4.668164205087887,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030264064647278537,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22884678989648818,
      "num_tokens": 92801661.0,
      "step": 41565
    },
    {
      "entropy": 5.369493722915649,
      "epoch": 4.668725781995844,
      "grad_norm": 1.09375,
      "learning_rate": 0.000302600801417098,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23358503580093384,
      "num_tokens": 92813223.0,
      "step": 41570
    },
    {
      "entropy": 5.403044366836548,
      "epoch": 4.6692873589038015,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030256095548259035,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2263038858771324,
      "num_tokens": 92823839.0,
      "step": 41575
    },
    {
      "entropy": 5.326963615417481,
      "epoch": 4.66984893581176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003025211086705312,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23345523774623872,
      "num_tokens": 92834739.0,
      "step": 41580
    },
    {
      "entropy": 5.4520470142364506,
      "epoch": 4.670410512719717,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003024812609821893,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2298668086528778,
      "num_tokens": 92846369.0,
      "step": 41585
    },
    {
      "entropy": 5.429536008834839,
      "epoch": 4.670972089627675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003024414124188333,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22957131415605544,
      "num_tokens": 92857963.0,
      "step": 41590
    },
    {
      "entropy": 5.465985584259033,
      "epoch": 4.671533666535632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003024015629817322,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22549217641353608,
      "num_tokens": 92868922.0,
      "step": 41595
    },
    {
      "entropy": 5.439185762405396,
      "epoch": 4.672095243443589,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003023617126721547,
      "loss": 4.921,
      "mean_token_accuracy": 0.2235422044992447,
      "num_tokens": 92881107.0,
      "step": 41600
    },
    {
      "entropy": 5.491609764099121,
      "epoch": 4.6726568203515475,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003023218614913696,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2253548488020897,
      "num_tokens": 92891538.0,
      "step": 41605
    },
    {
      "entropy": 5.4591772079467775,
      "epoch": 4.673218397259505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003022820094406458,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21640805453062056,
      "num_tokens": 92903015.0,
      "step": 41610
    },
    {
      "entropy": 5.414551639556885,
      "epoch": 4.673779974167462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003022421565212524,
      "loss": 4.8666,
      "mean_token_accuracy": 0.22028053998947145,
      "num_tokens": 92914237.0,
      "step": 41615
    },
    {
      "entropy": 5.426530456542968,
      "epoch": 4.67434155107542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030220230273445806,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22820313423871993,
      "num_tokens": 92925351.0,
      "step": 41620
    },
    {
      "entropy": 5.461531114578247,
      "epoch": 4.674903127983377,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030216244808153186,
      "loss": 4.9369,
      "mean_token_accuracy": 0.23124035596847534,
      "num_tokens": 92935472.0,
      "step": 41625
    },
    {
      "entropy": 5.526190662384034,
      "epoch": 4.6754647048913345,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003021225925637428,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21501239985227585,
      "num_tokens": 92948328.0,
      "step": 41630
    },
    {
      "entropy": 5.421201753616333,
      "epoch": 4.676026281799293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003020827361823598,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22875114679336547,
      "num_tokens": 92958458.0,
      "step": 41635
    },
    {
      "entropy": 5.486659669876099,
      "epoch": 4.67658785870725,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030204287893865204,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21780622750520706,
      "num_tokens": 92969870.0,
      "step": 41640
    },
    {
      "entropy": 5.419363641738892,
      "epoch": 4.677149435615208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030200302083388856,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22507369220256807,
      "num_tokens": 92981597.0,
      "step": 41645
    },
    {
      "entropy": 5.52564263343811,
      "epoch": 4.677711012523165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030196316186933834,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22978515028953553,
      "num_tokens": 92992525.0,
      "step": 41650
    },
    {
      "entropy": 5.466914176940918,
      "epoch": 4.678272589431122,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030192330204627054,
      "loss": 4.939,
      "mean_token_accuracy": 0.22339371591806412,
      "num_tokens": 93003652.0,
      "step": 41655
    },
    {
      "entropy": 5.544053459167481,
      "epoch": 4.6788341663390804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003018834413659545,
      "loss": 5.0386,
      "mean_token_accuracy": 0.22018277645111084,
      "num_tokens": 93015794.0,
      "step": 41660
    },
    {
      "entropy": 5.408869743347168,
      "epoch": 4.679395743247038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003018435798296592,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22861088067293167,
      "num_tokens": 93026454.0,
      "step": 41665
    },
    {
      "entropy": 5.4577024459838865,
      "epoch": 4.679957320154995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003018037174386539,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2199436604976654,
      "num_tokens": 93039263.0,
      "step": 41670
    },
    {
      "entropy": 5.51140079498291,
      "epoch": 4.680518897062953,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003017638541942078,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22627294510602952,
      "num_tokens": 93050437.0,
      "step": 41675
    },
    {
      "entropy": 5.522812938690185,
      "epoch": 4.68108047397091,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030172399009759034,
      "loss": 4.9826,
      "mean_token_accuracy": 0.22089919894933702,
      "num_tokens": 93060602.0,
      "step": 41680
    },
    {
      "entropy": 5.527175045013427,
      "epoch": 4.681642050878867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003016841251500705,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21913072019815444,
      "num_tokens": 93073062.0,
      "step": 41685
    },
    {
      "entropy": 5.45185866355896,
      "epoch": 4.6822036277868255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030164425935291794,
      "loss": 4.9698,
      "mean_token_accuracy": 0.22936414778232575,
      "num_tokens": 93083150.0,
      "step": 41690
    },
    {
      "entropy": 5.506683492660523,
      "epoch": 4.682765204694783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003016043927074018,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21836733371019362,
      "num_tokens": 93094506.0,
      "step": 41695
    },
    {
      "entropy": 5.380016040802002,
      "epoch": 4.683326781602741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003015645252147914,
      "loss": 4.8059,
      "mean_token_accuracy": 0.2277416467666626,
      "num_tokens": 93105148.0,
      "step": 41700
    },
    {
      "entropy": 5.439466190338135,
      "epoch": 4.683888358510698,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003015246568763562,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22188717424869536,
      "num_tokens": 93116470.0,
      "step": 41705
    },
    {
      "entropy": 5.407010793685913,
      "epoch": 4.684449935418655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030148478769336564,
      "loss": 4.9111,
      "mean_token_accuracy": 0.21974487006664276,
      "num_tokens": 93126662.0,
      "step": 41710
    },
    {
      "entropy": 5.468749618530273,
      "epoch": 4.685011512326613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003014449176670892,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22059730738401412,
      "num_tokens": 93137690.0,
      "step": 41715
    },
    {
      "entropy": 5.409414148330688,
      "epoch": 4.685573089234571,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030140504679879637,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22726468592882157,
      "num_tokens": 93149760.0,
      "step": 41720
    },
    {
      "entropy": 5.501344299316406,
      "epoch": 4.686134666142529,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003013651750897566,
      "loss": 4.9819,
      "mean_token_accuracy": 0.23308862298727034,
      "num_tokens": 93162046.0,
      "step": 41725
    },
    {
      "entropy": 5.469012832641601,
      "epoch": 4.686696243050486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003013253025412393,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23302853405475615,
      "num_tokens": 93172912.0,
      "step": 41730
    },
    {
      "entropy": 5.493797302246094,
      "epoch": 4.687257819958443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030128542915451427,
      "loss": 4.9926,
      "mean_token_accuracy": 0.22095739245414733,
      "num_tokens": 93184595.0,
      "step": 41735
    },
    {
      "entropy": 5.441645717620849,
      "epoch": 4.6878193968664,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030124555493085095,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22496240884065627,
      "num_tokens": 93194753.0,
      "step": 41740
    },
    {
      "entropy": 5.450566911697388,
      "epoch": 4.6883809737743585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003012056798715189,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22689427584409713,
      "num_tokens": 93204259.0,
      "step": 41745
    },
    {
      "entropy": 5.4326395988464355,
      "epoch": 4.688942550682316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000301165803977788,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23596798181533812,
      "num_tokens": 93214984.0,
      "step": 41750
    },
    {
      "entropy": 5.442271709442139,
      "epoch": 4.689504127590274,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003011259272509277,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22311572432518006,
      "num_tokens": 93227004.0,
      "step": 41755
    },
    {
      "entropy": 5.551017189025879,
      "epoch": 4.690065704498231,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003010860496922077,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2173895999789238,
      "num_tokens": 93238485.0,
      "step": 41760
    },
    {
      "entropy": 5.519052791595459,
      "epoch": 4.690627281406188,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030104617130289783,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22812331467866898,
      "num_tokens": 93249318.0,
      "step": 41765
    },
    {
      "entropy": 5.540460777282715,
      "epoch": 4.691188858314146,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003010062920842677,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22961132675409318,
      "num_tokens": 93262040.0,
      "step": 41770
    },
    {
      "entropy": 5.467882823944092,
      "epoch": 4.6917504352221036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030096641203758716,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22289322912693024,
      "num_tokens": 93272766.0,
      "step": 41775
    },
    {
      "entropy": 5.5660011768341064,
      "epoch": 4.692312012130062,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030092653116412605,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22955501526594163,
      "num_tokens": 93283179.0,
      "step": 41780
    },
    {
      "entropy": 5.470059442520141,
      "epoch": 4.692873589038019,
      "grad_norm": 1.0625,
      "learning_rate": 0.000300886649465154,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22026987373828888,
      "num_tokens": 93293484.0,
      "step": 41785
    },
    {
      "entropy": 5.4354651927947994,
      "epoch": 4.693435165945976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003008467669419411,
      "loss": 4.905,
      "mean_token_accuracy": 0.22103402614593506,
      "num_tokens": 93305611.0,
      "step": 41790
    },
    {
      "entropy": 5.481216526031494,
      "epoch": 4.693996742853934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003008068835957571,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22715714424848557,
      "num_tokens": 93316495.0,
      "step": 41795
    },
    {
      "entropy": 5.3790501117706295,
      "epoch": 4.694558319761891,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003007669994278719,
      "loss": 4.853,
      "mean_token_accuracy": 0.233861044049263,
      "num_tokens": 93328484.0,
      "step": 41800
    },
    {
      "entropy": 5.467354583740234,
      "epoch": 4.695119896669849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003007271144395554,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22233079820871354,
      "num_tokens": 93339826.0,
      "step": 41805
    },
    {
      "entropy": 5.4765983581542965,
      "epoch": 4.695681473577807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003006872286320777,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22931426167488098,
      "num_tokens": 93351923.0,
      "step": 41810
    },
    {
      "entropy": 5.434081315994263,
      "epoch": 4.696243050485764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003006473420067086,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2296685218811035,
      "num_tokens": 93363813.0,
      "step": 41815
    },
    {
      "entropy": 5.506272268295288,
      "epoch": 4.696804627393721,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003006074545647182,
      "loss": 5.0043,
      "mean_token_accuracy": 0.23070923537015914,
      "num_tokens": 93374905.0,
      "step": 41820
    },
    {
      "entropy": 5.516914415359497,
      "epoch": 4.697366204301679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030056756630737655,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22296809405088425,
      "num_tokens": 93386473.0,
      "step": 41825
    },
    {
      "entropy": 5.470607423782349,
      "epoch": 4.6979277812096365,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030052767723595365,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22126565873622894,
      "num_tokens": 93398010.0,
      "step": 41830
    },
    {
      "entropy": 5.468153762817383,
      "epoch": 4.698489358117595,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003004877873517196,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2202627182006836,
      "num_tokens": 93410692.0,
      "step": 41835
    },
    {
      "entropy": 5.434916877746582,
      "epoch": 4.699050935025552,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030044789665594444,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22780765891075133,
      "num_tokens": 93421830.0,
      "step": 41840
    },
    {
      "entropy": 5.5645357131958,
      "epoch": 4.699612511933509,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003004080051498985,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21729116886854172,
      "num_tokens": 93434136.0,
      "step": 41845
    },
    {
      "entropy": 5.431099796295166,
      "epoch": 4.700174088841467,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030036811283485184,
      "loss": 4.8562,
      "mean_token_accuracy": 0.22976442277431489,
      "num_tokens": 93444559.0,
      "step": 41850
    },
    {
      "entropy": 5.489794540405273,
      "epoch": 4.700735665749424,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030032821971207453,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21954151690006257,
      "num_tokens": 93456258.0,
      "step": 41855
    },
    {
      "entropy": 5.475648498535156,
      "epoch": 4.701297242657382,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003002883257828369,
      "loss": 4.911,
      "mean_token_accuracy": 0.22844208329916,
      "num_tokens": 93467402.0,
      "step": 41860
    },
    {
      "entropy": 5.492329740524292,
      "epoch": 4.70185881956534,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030024843104840924,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2178042784333229,
      "num_tokens": 93477515.0,
      "step": 41865
    },
    {
      "entropy": 5.492426633834839,
      "epoch": 4.702420396473297,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030020853551006164,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2192101001739502,
      "num_tokens": 93488173.0,
      "step": 41870
    },
    {
      "entropy": 5.400456428527832,
      "epoch": 4.702981973381254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003001686391690645,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23221338242292405,
      "num_tokens": 93498236.0,
      "step": 41875
    },
    {
      "entropy": 5.408120632171631,
      "epoch": 4.703543550289212,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030012874202668823,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22954028844833374,
      "num_tokens": 93510162.0,
      "step": 41880
    },
    {
      "entropy": 5.466610908508301,
      "epoch": 4.7041051271971694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030008884408420293,
      "loss": 4.9598,
      "mean_token_accuracy": 0.23276390135288239,
      "num_tokens": 93521344.0,
      "step": 41885
    },
    {
      "entropy": 5.489440011978149,
      "epoch": 4.7046667041051276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003000489453428792,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22554476410150529,
      "num_tokens": 93533116.0,
      "step": 41890
    },
    {
      "entropy": 5.4957098960876465,
      "epoch": 4.705228281013085,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003000090458039873,
      "loss": 4.9911,
      "mean_token_accuracy": 0.222378733754158,
      "num_tokens": 93544612.0,
      "step": 41895
    },
    {
      "entropy": 5.526552057266235,
      "epoch": 4.705789857921042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029996914546879773,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22389042228460312,
      "num_tokens": 93556033.0,
      "step": 41900
    },
    {
      "entropy": 5.561739826202393,
      "epoch": 4.706351434829,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002999292443385809,
      "loss": 5.0746,
      "mean_token_accuracy": 0.216574028134346,
      "num_tokens": 93567250.0,
      "step": 41905
    },
    {
      "entropy": 5.520518684387207,
      "epoch": 4.706913011736957,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002998893424146072,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2195129707455635,
      "num_tokens": 93578521.0,
      "step": 41910
    },
    {
      "entropy": 5.50751428604126,
      "epoch": 4.707474588644915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002998494396981472,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2256704717874527,
      "num_tokens": 93589809.0,
      "step": 41915
    },
    {
      "entropy": 5.407452869415283,
      "epoch": 4.708036165552873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002998095361904714,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23557835817337036,
      "num_tokens": 93599688.0,
      "step": 41920
    },
    {
      "entropy": 5.403969383239746,
      "epoch": 4.70859774246083,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002997696318928503,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22941371351480483,
      "num_tokens": 93610575.0,
      "step": 41925
    },
    {
      "entropy": 5.40067195892334,
      "epoch": 4.709159319368787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002997297268065546,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23649199008941652,
      "num_tokens": 93622329.0,
      "step": 41930
    },
    {
      "entropy": 5.41402850151062,
      "epoch": 4.709720896276745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029968982093285475,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22533971965312957,
      "num_tokens": 93632686.0,
      "step": 41935
    },
    {
      "entropy": 5.521656274795532,
      "epoch": 4.710282473184702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029964991427302147,
      "loss": 4.9438,
      "mean_token_accuracy": 0.23242482990026475,
      "num_tokens": 93643086.0,
      "step": 41940
    },
    {
      "entropy": 5.551564407348633,
      "epoch": 4.7108440500926605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002996100068283253,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2199747696518898,
      "num_tokens": 93654457.0,
      "step": 41945
    },
    {
      "entropy": 5.513864135742187,
      "epoch": 4.711405627000618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029957009860003697,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21762975454330444,
      "num_tokens": 93665806.0,
      "step": 41950
    },
    {
      "entropy": 5.431041622161866,
      "epoch": 4.711967203908575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029953018958942717,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22233254909515382,
      "num_tokens": 93676685.0,
      "step": 41955
    },
    {
      "entropy": 5.531437397003174,
      "epoch": 4.712528780816533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002994902797977667,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21472801715135575,
      "num_tokens": 93688111.0,
      "step": 41960
    },
    {
      "entropy": 5.469184684753418,
      "epoch": 4.71309035772449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029945036922632615,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22181306332349776,
      "num_tokens": 93698919.0,
      "step": 41965
    },
    {
      "entropy": 5.506165075302124,
      "epoch": 4.713651934632448,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029941045787637636,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2239682510495186,
      "num_tokens": 93711367.0,
      "step": 41970
    },
    {
      "entropy": 5.465181589126587,
      "epoch": 4.714213511540406,
      "grad_norm": 1.0,
      "learning_rate": 0.0002993705457491882,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2324758440256119,
      "num_tokens": 93722649.0,
      "step": 41975
    },
    {
      "entropy": 5.49464168548584,
      "epoch": 4.714775088448363,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029933063284603235,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21841963529586791,
      "num_tokens": 93733873.0,
      "step": 41980
    },
    {
      "entropy": 5.387407016754151,
      "epoch": 4.715336665356321,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002992907191681797,
      "loss": 4.7776,
      "mean_token_accuracy": 0.24182587414979934,
      "num_tokens": 93744954.0,
      "step": 41985
    },
    {
      "entropy": 5.498066663742065,
      "epoch": 4.715898242264278,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029925080471690117,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22577090859413146,
      "num_tokens": 93756607.0,
      "step": 41990
    },
    {
      "entropy": 5.418418645858765,
      "epoch": 4.716459819172235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029921088949346765,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22616310119628907,
      "num_tokens": 93767496.0,
      "step": 41995
    },
    {
      "entropy": 5.381891441345215,
      "epoch": 4.717021396080193,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002991709734991499,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22560623288154602,
      "num_tokens": 93778432.0,
      "step": 42000
    },
    {
      "epoch": 4.717021396080193,
      "eval_entropy": 5.224169265495184,
      "eval_loss": 5.114164352416992,
      "eval_mean_token_accuracy": 0.22326922189345333,
      "eval_num_tokens": 93778432.0,
      "eval_runtime": 53.5523,
      "eval_samples_per_second": 579.116,
      "eval_steps_per_second": 72.397,
      "step": 42000
    },
    {
      "entropy": 5.407142877578735,
      "epoch": 4.717582972988151,
      "grad_norm": 1.0,
      "learning_rate": 0.00029913105673521907,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22892675250768663,
      "num_tokens": 93789652.0,
      "step": 42005
    },
    {
      "entropy": 5.453160953521729,
      "epoch": 4.718144549896108,
      "grad_norm": 1.015625,
      "learning_rate": 0.000299091139202946,
      "loss": 4.8946,
      "mean_token_accuracy": 0.22600020170211793,
      "num_tokens": 93800386.0,
      "step": 42010
    },
    {
      "entropy": 5.51993145942688,
      "epoch": 4.718706126804066,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029905122090360167,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22577731013298036,
      "num_tokens": 93810184.0,
      "step": 42015
    },
    {
      "entropy": 5.442842292785644,
      "epoch": 4.719267703712023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002990113018384572,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22931873202323913,
      "num_tokens": 93820139.0,
      "step": 42020
    },
    {
      "entropy": 5.4438871383667,
      "epoch": 4.719829280619981,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002989713820087836,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23378548622131348,
      "num_tokens": 93831323.0,
      "step": 42025
    },
    {
      "entropy": 5.483273458480835,
      "epoch": 4.7203908575279385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029893146141585187,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22401926368474961,
      "num_tokens": 93841160.0,
      "step": 42030
    },
    {
      "entropy": 5.514167308807373,
      "epoch": 4.720952434435896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029889154006093306,
      "loss": 5.0548,
      "mean_token_accuracy": 0.21568068414926528,
      "num_tokens": 93851530.0,
      "step": 42035
    },
    {
      "entropy": 5.533733320236206,
      "epoch": 4.721514011343854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002988516179452984,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22648120075464248,
      "num_tokens": 93862466.0,
      "step": 42040
    },
    {
      "entropy": 5.508970117568969,
      "epoch": 4.722075588251811,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029881169507021895,
      "loss": 4.9201,
      "mean_token_accuracy": 0.23749003559350967,
      "num_tokens": 93874399.0,
      "step": 42045
    },
    {
      "entropy": 5.459762096405029,
      "epoch": 4.722637165159768,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000298771771436966,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22586722671985626,
      "num_tokens": 93885054.0,
      "step": 42050
    },
    {
      "entropy": 5.447548580169678,
      "epoch": 4.723198742067726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002987318470468105,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2370377317070961,
      "num_tokens": 93894813.0,
      "step": 42055
    },
    {
      "entropy": 5.374237203598023,
      "epoch": 4.723760318975684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029869192190102385,
      "loss": 4.9109,
      "mean_token_accuracy": 0.23235447704792023,
      "num_tokens": 93905625.0,
      "step": 42060
    },
    {
      "entropy": 5.436096048355102,
      "epoch": 4.724321895883641,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029865199600087724,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23088692724704743,
      "num_tokens": 93917539.0,
      "step": 42065
    },
    {
      "entropy": 5.459396409988403,
      "epoch": 4.724883472791599,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002986120693476418,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2230369731783867,
      "num_tokens": 93927307.0,
      "step": 42070
    },
    {
      "entropy": 5.454146766662598,
      "epoch": 4.725445049699556,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029857214194258896,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2307046741247177,
      "num_tokens": 93938252.0,
      "step": 42075
    },
    {
      "entropy": 5.481504106521607,
      "epoch": 4.726006626607514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029853221378699,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23050923198461531,
      "num_tokens": 93947762.0,
      "step": 42080
    },
    {
      "entropy": 5.488630723953247,
      "epoch": 4.7265682035154715,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002984922848821162,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21914098262786866,
      "num_tokens": 93958410.0,
      "step": 42085
    },
    {
      "entropy": 5.429141998291016,
      "epoch": 4.727129780423429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029845235522923903,
      "loss": 4.9197,
      "mean_token_accuracy": 0.23202846944332123,
      "num_tokens": 93969610.0,
      "step": 42090
    },
    {
      "entropy": 5.486133670806884,
      "epoch": 4.727691357331387,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029841242482962966,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22760361880064012,
      "num_tokens": 93978404.0,
      "step": 42095
    },
    {
      "entropy": 5.409827184677124,
      "epoch": 4.728252934239344,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029837249368455963,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22256179749965668,
      "num_tokens": 93989578.0,
      "step": 42100
    },
    {
      "entropy": 5.520192861557007,
      "epoch": 4.728814511147302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002983325617953004,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2110009968280792,
      "num_tokens": 94000010.0,
      "step": 42105
    },
    {
      "entropy": 5.471327257156372,
      "epoch": 4.729376088055259,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029829262916312326,
      "loss": 4.8801,
      "mean_token_accuracy": 0.240518781542778,
      "num_tokens": 94009386.0,
      "step": 42110
    },
    {
      "entropy": 5.374818468093872,
      "epoch": 4.7299376649632165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029825269578929986,
      "loss": 4.8094,
      "mean_token_accuracy": 0.2347433239221573,
      "num_tokens": 94021594.0,
      "step": 42115
    },
    {
      "entropy": 5.480156469345093,
      "epoch": 4.730499241871174,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029821276167510153,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22177234739065171,
      "num_tokens": 94033518.0,
      "step": 42120
    },
    {
      "entropy": 5.4185600757598875,
      "epoch": 4.731060818779132,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002981728268217999,
      "loss": 4.9216,
      "mean_token_accuracy": 0.2217911198735237,
      "num_tokens": 94044174.0,
      "step": 42125
    },
    {
      "entropy": 5.42535982131958,
      "epoch": 4.731622395687089,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002981328912306665,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22606829106807708,
      "num_tokens": 94056000.0,
      "step": 42130
    },
    {
      "entropy": 5.543865728378296,
      "epoch": 4.732183972595047,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029809295490297285,
      "loss": 5.0261,
      "mean_token_accuracy": 0.22615868002176284,
      "num_tokens": 94067274.0,
      "step": 42135
    },
    {
      "entropy": 5.485551023483277,
      "epoch": 4.732745549503004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002980530178399906,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22560227960348128,
      "num_tokens": 94079113.0,
      "step": 42140
    },
    {
      "entropy": 5.4589982509613035,
      "epoch": 4.733307126410962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002980130800429912,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22688924074172973,
      "num_tokens": 94090438.0,
      "step": 42145
    },
    {
      "entropy": 5.434643745422363,
      "epoch": 4.73386870331892,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029797314151324653,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2219032198190689,
      "num_tokens": 94101153.0,
      "step": 42150
    },
    {
      "entropy": 5.564031744003296,
      "epoch": 4.734430280226877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029793320225202807,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2227512076497078,
      "num_tokens": 94112366.0,
      "step": 42155
    },
    {
      "entropy": 5.451723432540893,
      "epoch": 4.734991857134835,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002978932622606075,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22643409073352813,
      "num_tokens": 94122919.0,
      "step": 42160
    },
    {
      "entropy": 5.522156286239624,
      "epoch": 4.735553434042792,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029785332154025676,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21867509186267853,
      "num_tokens": 94135158.0,
      "step": 42165
    },
    {
      "entropy": 5.498574829101562,
      "epoch": 4.7361150109507495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002978133800922473,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2161959558725357,
      "num_tokens": 94145773.0,
      "step": 42170
    },
    {
      "entropy": 5.413529682159424,
      "epoch": 4.736676587858708,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002977734379178509,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22796997725963591,
      "num_tokens": 94157260.0,
      "step": 42175
    },
    {
      "entropy": 5.416730785369873,
      "epoch": 4.737238164766665,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002977334950183394,
      "loss": 4.902,
      "mean_token_accuracy": 0.22819317877292633,
      "num_tokens": 94166513.0,
      "step": 42180
    },
    {
      "entropy": 5.548943996429443,
      "epoch": 4.737799741674622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029769355139498465,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20435759276151658,
      "num_tokens": 94178565.0,
      "step": 42185
    },
    {
      "entropy": 5.461018705368042,
      "epoch": 4.73836131858258,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002976536070490585,
      "loss": 4.894,
      "mean_token_accuracy": 0.2273544639348984,
      "num_tokens": 94189319.0,
      "step": 42190
    },
    {
      "entropy": 5.514475917816162,
      "epoch": 4.738922895490537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029761366198183264,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2177510753273964,
      "num_tokens": 94203030.0,
      "step": 42195
    },
    {
      "entropy": 5.5035864353179935,
      "epoch": 4.739484472398495,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029757371619457904,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22170187383890153,
      "num_tokens": 94214901.0,
      "step": 42200
    },
    {
      "entropy": 5.430035018920899,
      "epoch": 4.740046049306453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029753376968856955,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22584094107151031,
      "num_tokens": 94226833.0,
      "step": 42205
    },
    {
      "entropy": 5.5036920547485355,
      "epoch": 4.74060762621441,
      "grad_norm": 1.125,
      "learning_rate": 0.00029749382246507604,
      "loss": 4.9248,
      "mean_token_accuracy": 0.23071799874305726,
      "num_tokens": 94239464.0,
      "step": 42210
    },
    {
      "entropy": 5.459576654434204,
      "epoch": 4.741169203122368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029745387452537056,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2360416457056999,
      "num_tokens": 94250692.0,
      "step": 42215
    },
    {
      "entropy": 5.430537843704224,
      "epoch": 4.741730780030325,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029741392587072504,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22246290743350983,
      "num_tokens": 94261130.0,
      "step": 42220
    },
    {
      "entropy": 5.446351146697998,
      "epoch": 4.742292356938282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029737397650241135,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22226280122995376,
      "num_tokens": 94271581.0,
      "step": 42225
    },
    {
      "entropy": 5.47103362083435,
      "epoch": 4.7428539338462405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002973340264217016,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22324565201997756,
      "num_tokens": 94283353.0,
      "step": 42230
    },
    {
      "entropy": 5.491291236877442,
      "epoch": 4.743415510754198,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029729407562986774,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21773881167173387,
      "num_tokens": 94294669.0,
      "step": 42235
    },
    {
      "entropy": 5.493033313751221,
      "epoch": 4.743977087662155,
      "grad_norm": 1.0,
      "learning_rate": 0.00029725412412818196,
      "loss": 4.9203,
      "mean_token_accuracy": 0.23192385584115982,
      "num_tokens": 94306127.0,
      "step": 42240
    },
    {
      "entropy": 5.429143238067627,
      "epoch": 4.744538664570113,
      "grad_norm": 1.125,
      "learning_rate": 0.00029721417191791617,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2325584828853607,
      "num_tokens": 94316621.0,
      "step": 42245
    },
    {
      "entropy": 5.466883611679077,
      "epoch": 4.74510024147807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029717421900034253,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22151844650506974,
      "num_tokens": 94327715.0,
      "step": 42250
    },
    {
      "entropy": 5.494129228591919,
      "epoch": 4.7456618183860275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002971342653767332,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22776928991079332,
      "num_tokens": 94339452.0,
      "step": 42255
    },
    {
      "entropy": 5.48996639251709,
      "epoch": 4.746223395293986,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029709431104836027,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22751633524894715,
      "num_tokens": 94350309.0,
      "step": 42260
    },
    {
      "entropy": 5.5284199714660645,
      "epoch": 4.746784972201943,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029705435601649584,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2174219086766243,
      "num_tokens": 94361665.0,
      "step": 42265
    },
    {
      "entropy": 5.593537473678589,
      "epoch": 4.747346549109901,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002970144002824122,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21710591912269592,
      "num_tokens": 94374141.0,
      "step": 42270
    },
    {
      "entropy": 5.461695909500122,
      "epoch": 4.747908126017858,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002969744438473815,
      "loss": 4.8678,
      "mean_token_accuracy": 0.21778475493192673,
      "num_tokens": 94384379.0,
      "step": 42275
    },
    {
      "entropy": 5.411889600753784,
      "epoch": 4.748469702925815,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002969344867126761,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23297653794288636,
      "num_tokens": 94394747.0,
      "step": 42280
    },
    {
      "entropy": 5.473565673828125,
      "epoch": 4.7490312798337735,
      "grad_norm": 1.125,
      "learning_rate": 0.00029689452887956804,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2194908306002617,
      "num_tokens": 94407530.0,
      "step": 42285
    },
    {
      "entropy": 5.486377954483032,
      "epoch": 4.749592856741731,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029685457034932966,
      "loss": 5.0046,
      "mean_token_accuracy": 0.2164237380027771,
      "num_tokens": 94420031.0,
      "step": 42290
    },
    {
      "entropy": 5.497297668457032,
      "epoch": 4.750154433649689,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002968146111232333,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22431379109621047,
      "num_tokens": 94429634.0,
      "step": 42295
    },
    {
      "entropy": 5.501425695419312,
      "epoch": 4.750716010557646,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029677465120255133,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22106616348028182,
      "num_tokens": 94440310.0,
      "step": 42300
    },
    {
      "entropy": 5.479865550994873,
      "epoch": 4.751277587465603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002967346905885559,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21894653439521788,
      "num_tokens": 94451050.0,
      "step": 42305
    },
    {
      "entropy": 5.511062908172607,
      "epoch": 4.7518391643735605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029669472928251954,
      "loss": 4.9982,
      "mean_token_accuracy": 0.23162644654512404,
      "num_tokens": 94462210.0,
      "step": 42310
    },
    {
      "entropy": 5.40460958480835,
      "epoch": 4.752400741281519,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029665476728571466,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2240043669939041,
      "num_tokens": 94472716.0,
      "step": 42315
    },
    {
      "entropy": 5.41677041053772,
      "epoch": 4.752962318189476,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002966148045994135,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23357120305299758,
      "num_tokens": 94482446.0,
      "step": 42320
    },
    {
      "entropy": 5.44665150642395,
      "epoch": 4.753523895097434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002965748412248886,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22943754196166993,
      "num_tokens": 94492476.0,
      "step": 42325
    },
    {
      "entropy": 5.393251037597656,
      "epoch": 4.754085472005391,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002965348771634124,
      "loss": 4.8519,
      "mean_token_accuracy": 0.23238711953163146,
      "num_tokens": 94502607.0,
      "step": 42330
    },
    {
      "entropy": 5.408693408966064,
      "epoch": 4.754647048913348,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029649491241625736,
      "loss": 4.8035,
      "mean_token_accuracy": 0.22917423099279405,
      "num_tokens": 94512528.0,
      "step": 42335
    },
    {
      "entropy": 5.643634796142578,
      "epoch": 4.755208625821306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029645494698469604,
      "loss": 5.1297,
      "mean_token_accuracy": 0.2164226546883583,
      "num_tokens": 94524890.0,
      "step": 42340
    },
    {
      "entropy": 5.458085203170777,
      "epoch": 4.755770202729264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029641498087000083,
      "loss": 4.8222,
      "mean_token_accuracy": 0.23505277186632156,
      "num_tokens": 94536483.0,
      "step": 42345
    },
    {
      "entropy": 5.537858343124389,
      "epoch": 4.756331779637222,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002963750140734443,
      "loss": 5.0788,
      "mean_token_accuracy": 0.21687097549438478,
      "num_tokens": 94547235.0,
      "step": 42350
    },
    {
      "entropy": 5.529531669616699,
      "epoch": 4.756893356545179,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002963350465962991,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2280014395713806,
      "num_tokens": 94558511.0,
      "step": 42355
    },
    {
      "entropy": 5.443650722503662,
      "epoch": 4.757454933453136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002962950784398376,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22323132455348968,
      "num_tokens": 94569949.0,
      "step": 42360
    },
    {
      "entropy": 5.451601934432984,
      "epoch": 4.758016510361094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002962551096053327,
      "loss": 4.918,
      "mean_token_accuracy": 0.22254585921764375,
      "num_tokens": 94581353.0,
      "step": 42365
    },
    {
      "entropy": 5.414772081375122,
      "epoch": 4.7585780872690515,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029621514009405675,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23076555579900743,
      "num_tokens": 94591653.0,
      "step": 42370
    },
    {
      "entropy": 5.46034197807312,
      "epoch": 4.759139664177009,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002961751699072826,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22408088445663452,
      "num_tokens": 94603615.0,
      "step": 42375
    },
    {
      "entropy": 5.501304292678833,
      "epoch": 4.759701241084967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002961351990462827,
      "loss": 5.0336,
      "mean_token_accuracy": 0.22138273417949678,
      "num_tokens": 94614813.0,
      "step": 42380
    },
    {
      "entropy": 5.463575792312622,
      "epoch": 4.760262817992924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029609522751232996,
      "loss": 4.8165,
      "mean_token_accuracy": 0.22936659306287766,
      "num_tokens": 94624483.0,
      "step": 42385
    },
    {
      "entropy": 5.494738006591797,
      "epoch": 4.760824394900881,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029605525530669705,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2265668421983719,
      "num_tokens": 94634332.0,
      "step": 42390
    },
    {
      "entropy": 5.429692983627319,
      "epoch": 4.761385971808839,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029601528243065656,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23120235353708268,
      "num_tokens": 94645019.0,
      "step": 42395
    },
    {
      "entropy": 5.516799736022949,
      "epoch": 4.761947548716797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029597530888548135,
      "loss": 5.0376,
      "mean_token_accuracy": 0.2221710577607155,
      "num_tokens": 94656393.0,
      "step": 42400
    },
    {
      "entropy": 5.451876878738403,
      "epoch": 4.762509125624755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029593533467244415,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2322724238038063,
      "num_tokens": 94666730.0,
      "step": 42405
    },
    {
      "entropy": 5.4544233798980715,
      "epoch": 4.763070702532712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002958953597928178,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2229684591293335,
      "num_tokens": 94677937.0,
      "step": 42410
    },
    {
      "entropy": 5.500162887573242,
      "epoch": 4.763632279440669,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002958553842478751,
      "loss": 4.948,
      "mean_token_accuracy": 0.2247547686100006,
      "num_tokens": 94688141.0,
      "step": 42415
    },
    {
      "entropy": 5.4917339324951175,
      "epoch": 4.764193856348627,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029581540803888886,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22849182784557343,
      "num_tokens": 94700055.0,
      "step": 42420
    },
    {
      "entropy": 5.502260017395019,
      "epoch": 4.7647554332565845,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029577543116713197,
      "loss": 4.9992,
      "mean_token_accuracy": 0.22376129776239395,
      "num_tokens": 94711858.0,
      "step": 42425
    },
    {
      "entropy": 5.468829536437989,
      "epoch": 4.765317010164542,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029573545363387725,
      "loss": 4.9841,
      "mean_token_accuracy": 0.2229260802268982,
      "num_tokens": 94722691.0,
      "step": 42430
    },
    {
      "entropy": 5.4648223400115965,
      "epoch": 4.7658785870725,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002956954754403977,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22710380852222442,
      "num_tokens": 94733570.0,
      "step": 42435
    },
    {
      "entropy": 5.506772661209107,
      "epoch": 4.766440163980457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002956554965879662,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22942209541797637,
      "num_tokens": 94744719.0,
      "step": 42440
    },
    {
      "entropy": 5.547110509872437,
      "epoch": 4.767001740888414,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002956155170778556,
      "loss": 5.0207,
      "mean_token_accuracy": 0.215211983025074,
      "num_tokens": 94755272.0,
      "step": 42445
    },
    {
      "entropy": 5.557039833068847,
      "epoch": 4.767563317796372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002955755369113389,
      "loss": 5.0688,
      "mean_token_accuracy": 0.21831802725791932,
      "num_tokens": 94767178.0,
      "step": 42450
    },
    {
      "entropy": 5.4473754405975345,
      "epoch": 4.7681248947043295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029553555608968925,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22326680570840834,
      "num_tokens": 94777088.0,
      "step": 42455
    },
    {
      "entropy": 5.558001708984375,
      "epoch": 4.768686471612288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002954955746141795,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2219248041510582,
      "num_tokens": 94787995.0,
      "step": 42460
    },
    {
      "entropy": 5.508326530456543,
      "epoch": 4.769248048520245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002954555924860827,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22730502486228943,
      "num_tokens": 94797810.0,
      "step": 42465
    },
    {
      "entropy": 5.528858280181884,
      "epoch": 4.769809625428202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002954156097066718,
      "loss": 4.954,
      "mean_token_accuracy": 0.22685584425926208,
      "num_tokens": 94809725.0,
      "step": 42470
    },
    {
      "entropy": 5.54455976486206,
      "epoch": 4.77037120233616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002953756262772201,
      "loss": 4.9273,
      "mean_token_accuracy": 0.23178496062755585,
      "num_tokens": 94819740.0,
      "step": 42475
    },
    {
      "entropy": 5.58225383758545,
      "epoch": 4.770932779244117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002953356421990005,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21943700313568115,
      "num_tokens": 94830419.0,
      "step": 42480
    },
    {
      "entropy": 5.532826280593872,
      "epoch": 4.7714943561520755,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002952956574732862,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2147657960653305,
      "num_tokens": 94842796.0,
      "step": 42485
    },
    {
      "entropy": 5.5227803707122805,
      "epoch": 4.772055933060033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002952556721013503,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2211341604590416,
      "num_tokens": 94854065.0,
      "step": 42490
    },
    {
      "entropy": 5.517033815383911,
      "epoch": 4.77261750996799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002952156860844659,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2294651001691818,
      "num_tokens": 94866264.0,
      "step": 42495
    },
    {
      "entropy": 5.521196746826172,
      "epoch": 4.773179086875947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002951756994239063,
      "loss": 4.9838,
      "mean_token_accuracy": 0.223395836353302,
      "num_tokens": 94878549.0,
      "step": 42500
    },
    {
      "entropy": 5.5188240051269535,
      "epoch": 4.773740663783905,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029513571212094443,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22471031099557875,
      "num_tokens": 94890527.0,
      "step": 42505
    },
    {
      "entropy": 5.434290599822998,
      "epoch": 4.7743022406918625,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002950957241768539,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22012683153152465,
      "num_tokens": 94901424.0,
      "step": 42510
    },
    {
      "entropy": 5.407250452041626,
      "epoch": 4.774863817599821,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029505573559290756,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22808320671319962,
      "num_tokens": 94912153.0,
      "step": 42515
    },
    {
      "entropy": 5.473645448684692,
      "epoch": 4.775425394507778,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029501574637037884,
      "loss": 4.9429,
      "mean_token_accuracy": 0.2269621193408966,
      "num_tokens": 94923912.0,
      "step": 42520
    },
    {
      "entropy": 5.425438356399536,
      "epoch": 4.775986971415735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000294975756510541,
      "loss": 4.824,
      "mean_token_accuracy": 0.23368587344884872,
      "num_tokens": 94934768.0,
      "step": 42525
    },
    {
      "entropy": 5.473843383789062,
      "epoch": 4.776548548323693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002949357660146673,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21738053858280182,
      "num_tokens": 94946617.0,
      "step": 42530
    },
    {
      "entropy": 5.5746564865112305,
      "epoch": 4.77711012523165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002948957748840311,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21390352994203568,
      "num_tokens": 94958863.0,
      "step": 42535
    },
    {
      "entropy": 5.5108147144317625,
      "epoch": 4.7776717021396085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029485578311990574,
      "loss": 4.948,
      "mean_token_accuracy": 0.22647946923971177,
      "num_tokens": 94969930.0,
      "step": 42540
    },
    {
      "entropy": 5.488322305679321,
      "epoch": 4.778233279047566,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029481579072356456,
      "loss": 4.975,
      "mean_token_accuracy": 0.2214006796479225,
      "num_tokens": 94980858.0,
      "step": 42545
    },
    {
      "entropy": 5.617687320709228,
      "epoch": 4.778794855955523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002947757976962809,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21715106219053268,
      "num_tokens": 94993075.0,
      "step": 42550
    },
    {
      "entropy": 5.431242179870606,
      "epoch": 4.779356432863481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029473580403932814,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22258812189102173,
      "num_tokens": 95003849.0,
      "step": 42555
    },
    {
      "entropy": 5.472601985931396,
      "epoch": 4.779918009771438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002946958097539798,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22712336331605912,
      "num_tokens": 95015766.0,
      "step": 42560
    },
    {
      "entropy": 5.475570297241211,
      "epoch": 4.780479586679395,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002946558148415092,
      "loss": 4.9964,
      "mean_token_accuracy": 0.2266492500901222,
      "num_tokens": 95026690.0,
      "step": 42565
    },
    {
      "entropy": 5.444657325744629,
      "epoch": 4.7810411635873535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002946158193031898,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2262527361512184,
      "num_tokens": 95038591.0,
      "step": 42570
    },
    {
      "entropy": 5.589418745040893,
      "epoch": 4.781602740495311,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029457582314029507,
      "loss": 5.0377,
      "mean_token_accuracy": 0.22142512649297713,
      "num_tokens": 95050268.0,
      "step": 42575
    },
    {
      "entropy": 5.573477125167846,
      "epoch": 4.782164317403268,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029453582635409866,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22230153232812883,
      "num_tokens": 95061448.0,
      "step": 42580
    },
    {
      "entropy": 5.515412139892578,
      "epoch": 4.782725894311226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002944958289458739,
      "loss": 4.931,
      "mean_token_accuracy": 0.22074800580739976,
      "num_tokens": 95072331.0,
      "step": 42585
    },
    {
      "entropy": 5.483947038650513,
      "epoch": 4.783287471219183,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029445583091689437,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2219451203942299,
      "num_tokens": 95082536.0,
      "step": 42590
    },
    {
      "entropy": 5.373032855987549,
      "epoch": 4.783849048127141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002944158322684336,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23899985998868942,
      "num_tokens": 95093877.0,
      "step": 42595
    },
    {
      "entropy": 5.541982173919678,
      "epoch": 4.784410625035099,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029437583300176526,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22473303973674774,
      "num_tokens": 95104672.0,
      "step": 42600
    },
    {
      "entropy": 5.497195100784301,
      "epoch": 4.784972201943056,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029433583311816284,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21903145611286162,
      "num_tokens": 95115206.0,
      "step": 42605
    },
    {
      "entropy": 5.451729869842529,
      "epoch": 4.785533778851014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002942958326189,
      "loss": 4.877,
      "mean_token_accuracy": 0.22466294169425965,
      "num_tokens": 95125142.0,
      "step": 42610
    },
    {
      "entropy": 5.4865477085113525,
      "epoch": 4.786095355758971,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029425583150525044,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22213102430105208,
      "num_tokens": 95136108.0,
      "step": 42615
    },
    {
      "entropy": 5.467717123031616,
      "epoch": 4.786656932666928,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029421582977848763,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22030642181634902,
      "num_tokens": 95146452.0,
      "step": 42620
    },
    {
      "entropy": 5.359899234771729,
      "epoch": 4.7872185095748865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029417582743988547,
      "loss": 4.7241,
      "mean_token_accuracy": 0.23501054793596268,
      "num_tokens": 95156665.0,
      "step": 42625
    },
    {
      "entropy": 5.4582263946533205,
      "epoch": 4.787780086482844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029413582449071745,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22532953172922135,
      "num_tokens": 95167092.0,
      "step": 42630
    },
    {
      "entropy": 5.475995016098023,
      "epoch": 4.788341663390801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002940958209322574,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23016841411590577,
      "num_tokens": 95178158.0,
      "step": 42635
    },
    {
      "entropy": 5.418591928482056,
      "epoch": 4.788903240298759,
      "grad_norm": 1.0,
      "learning_rate": 0.00029405581676577897,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2239123538136482,
      "num_tokens": 95189432.0,
      "step": 42640
    },
    {
      "entropy": 5.433315706253052,
      "epoch": 4.789464817206716,
      "grad_norm": 1.140625,
      "learning_rate": 0.000294015811992556,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22950695008039473,
      "num_tokens": 95200682.0,
      "step": 42645
    },
    {
      "entropy": 5.4284257888793945,
      "epoch": 4.790026394114674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002939758066138621,
      "loss": 4.8751,
      "mean_token_accuracy": 0.2272473007440567,
      "num_tokens": 95212643.0,
      "step": 42650
    },
    {
      "entropy": 5.4751077651977536,
      "epoch": 4.790587971022632,
      "grad_norm": 1.125,
      "learning_rate": 0.0002939358006309712,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2273278832435608,
      "num_tokens": 95223098.0,
      "step": 42655
    },
    {
      "entropy": 5.439370346069336,
      "epoch": 4.791149547930589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002938957940451571,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2253586918115616,
      "num_tokens": 95232946.0,
      "step": 42660
    },
    {
      "entropy": 5.493184375762939,
      "epoch": 4.791711124838547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029385578685769346,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22881893068552017,
      "num_tokens": 95243966.0,
      "step": 42665
    },
    {
      "entropy": 5.456683111190796,
      "epoch": 4.792272701746504,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002938157790698544,
      "loss": 4.9486,
      "mean_token_accuracy": 0.226042503118515,
      "num_tokens": 95255251.0,
      "step": 42670
    },
    {
      "entropy": 5.407070207595825,
      "epoch": 4.792834278654462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002937757706829136,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23212108463048936,
      "num_tokens": 95267517.0,
      "step": 42675
    },
    {
      "entropy": 5.476585483551025,
      "epoch": 4.793395855562419,
      "grad_norm": 1.109375,
      "learning_rate": 0.000293735761698145,
      "loss": 4.975,
      "mean_token_accuracy": 0.22424352914094925,
      "num_tokens": 95277508.0,
      "step": 42680
    },
    {
      "entropy": 5.436635160446167,
      "epoch": 4.793957432470377,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002936957521168224,
      "loss": 4.9182,
      "mean_token_accuracy": 0.2269693583250046,
      "num_tokens": 95289757.0,
      "step": 42685
    },
    {
      "entropy": 5.524445390701294,
      "epoch": 4.794519009378335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002936557419402198,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22770176380872725,
      "num_tokens": 95301541.0,
      "step": 42690
    },
    {
      "entropy": 5.452566146850586,
      "epoch": 4.795080586286292,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002936157311696112,
      "loss": 4.9081,
      "mean_token_accuracy": 0.2241094246506691,
      "num_tokens": 95312840.0,
      "step": 42695
    },
    {
      "entropy": 5.413245964050293,
      "epoch": 4.795642163194249,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002935757198062705,
      "loss": 4.8603,
      "mean_token_accuracy": 0.23341536670923232,
      "num_tokens": 95322909.0,
      "step": 42700
    },
    {
      "entropy": 5.297267246246338,
      "epoch": 4.796203740102207,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002935357078514717,
      "loss": 4.7728,
      "mean_token_accuracy": 0.2296014606952667,
      "num_tokens": 95333437.0,
      "step": 42705
    },
    {
      "entropy": 5.415021705627441,
      "epoch": 4.7967653170101645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029349569530648877,
      "loss": 4.9049,
      "mean_token_accuracy": 0.228548064827919,
      "num_tokens": 95344477.0,
      "step": 42710
    },
    {
      "entropy": 5.521950101852417,
      "epoch": 4.797326893918122,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002934556821725957,
      "loss": 4.968,
      "mean_token_accuracy": 0.21660072058439256,
      "num_tokens": 95354354.0,
      "step": 42715
    },
    {
      "entropy": 5.474449110031128,
      "epoch": 4.79788847082608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029341566845106647,
      "loss": 4.9148,
      "mean_token_accuracy": 0.2206682562828064,
      "num_tokens": 95365026.0,
      "step": 42720
    },
    {
      "entropy": 5.530383825302124,
      "epoch": 4.798450047734037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002933756541431753,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2292229101061821,
      "num_tokens": 95376625.0,
      "step": 42725
    },
    {
      "entropy": 5.450825262069702,
      "epoch": 4.799011624641995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029333563925019616,
      "loss": 4.957,
      "mean_token_accuracy": 0.21800345927476883,
      "num_tokens": 95387066.0,
      "step": 42730
    },
    {
      "entropy": 5.399088621139526,
      "epoch": 4.799573201549952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029329562377340327,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22426166534423828,
      "num_tokens": 95398623.0,
      "step": 42735
    },
    {
      "entropy": 5.5611457347869875,
      "epoch": 4.80013477845791,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002932556077140704,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22189248204231263,
      "num_tokens": 95412514.0,
      "step": 42740
    },
    {
      "entropy": 5.544858407974243,
      "epoch": 4.800696355365868,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029321559107347206,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21560375541448593,
      "num_tokens": 95422885.0,
      "step": 42745
    },
    {
      "entropy": 5.524244594573974,
      "epoch": 4.801257932273825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029317557385288216,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22407928258180618,
      "num_tokens": 95433659.0,
      "step": 42750
    },
    {
      "entropy": 5.517619609832764,
      "epoch": 4.801819509181782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000293135556053575,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2226235195994377,
      "num_tokens": 95444161.0,
      "step": 42755
    },
    {
      "entropy": 5.453628253936768,
      "epoch": 4.80238108608974,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029309553767682467,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22781706005334854,
      "num_tokens": 95454840.0,
      "step": 42760
    },
    {
      "entropy": 5.507527875900268,
      "epoch": 4.8029426629976975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002930555187239054,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22417127788066865,
      "num_tokens": 95466766.0,
      "step": 42765
    },
    {
      "entropy": 5.4760902404785154,
      "epoch": 4.803504239905655,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029301549919609146,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22831126302480698,
      "num_tokens": 95477767.0,
      "step": 42770
    },
    {
      "entropy": 5.446821880340576,
      "epoch": 4.804065816813613,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029297547909465697,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22068424075841903,
      "num_tokens": 95489428.0,
      "step": 42775
    },
    {
      "entropy": 5.512834882736206,
      "epoch": 4.80462739372157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002929354584208763,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2198573023080826,
      "num_tokens": 95500844.0,
      "step": 42780
    },
    {
      "entropy": 5.452223062515259,
      "epoch": 4.805188970629528,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002928954371760237,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23800189793109894,
      "num_tokens": 95511829.0,
      "step": 42785
    },
    {
      "entropy": 5.5740502834320065,
      "epoch": 4.805750547537485,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002928554153613734,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21971351355314256,
      "num_tokens": 95523366.0,
      "step": 42790
    },
    {
      "entropy": 5.397696113586425,
      "epoch": 4.8063121244454425,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002928153929781997,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23321444243192674,
      "num_tokens": 95533495.0,
      "step": 42795
    },
    {
      "entropy": 5.409657049179077,
      "epoch": 4.806873701353401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029277537002777703,
      "loss": 4.922,
      "mean_token_accuracy": 0.22160994708538057,
      "num_tokens": 95545375.0,
      "step": 42800
    },
    {
      "entropy": 5.383014059066772,
      "epoch": 4.807435278261358,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029273534651137967,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22839583158493043,
      "num_tokens": 95557212.0,
      "step": 42805
    },
    {
      "entropy": 5.58746075630188,
      "epoch": 4.807996855169315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029269532243028195,
      "loss": 5.0873,
      "mean_token_accuracy": 0.221536985039711,
      "num_tokens": 95568608.0,
      "step": 42810
    },
    {
      "entropy": 5.3534462451934814,
      "epoch": 4.808558432077273,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002926552977857584,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22866297513246536,
      "num_tokens": 95579069.0,
      "step": 42815
    },
    {
      "entropy": 5.477660131454468,
      "epoch": 4.80912000898523,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029261527257908325,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22757295966148378,
      "num_tokens": 95589244.0,
      "step": 42820
    },
    {
      "entropy": 5.415616512298584,
      "epoch": 4.809681585893188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029257524681153106,
      "loss": 4.9623,
      "mean_token_accuracy": 0.23163859993219377,
      "num_tokens": 95600277.0,
      "step": 42825
    },
    {
      "entropy": 5.519608402252198,
      "epoch": 4.810243162801146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002925352204843761,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22521614879369736,
      "num_tokens": 95609913.0,
      "step": 42830
    },
    {
      "entropy": 5.634596061706543,
      "epoch": 4.810804739709103,
      "grad_norm": 1.03125,
      "learning_rate": 0.000292495193598893,
      "loss": 5.0704,
      "mean_token_accuracy": 0.21617207676172256,
      "num_tokens": 95623142.0,
      "step": 42835
    },
    {
      "entropy": 5.475644588470459,
      "epoch": 4.811366316617061,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029245516615635615,
      "loss": 4.982,
      "mean_token_accuracy": 0.21747396290302276,
      "num_tokens": 95633711.0,
      "step": 42840
    },
    {
      "entropy": 5.450428819656372,
      "epoch": 4.811927893525018,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029241513815804,
      "loss": 4.911,
      "mean_token_accuracy": 0.22779988944530488,
      "num_tokens": 95646956.0,
      "step": 42845
    },
    {
      "entropy": 5.532745599746704,
      "epoch": 4.8124894704329755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029237510960521904,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21933685541152953,
      "num_tokens": 95659260.0,
      "step": 42850
    },
    {
      "entropy": 5.549359083175659,
      "epoch": 4.813051047340934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000292335080499168,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2205290749669075,
      "num_tokens": 95671122.0,
      "step": 42855
    },
    {
      "entropy": 5.470278310775757,
      "epoch": 4.813612624248891,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029229505084116123,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22526505589485168,
      "num_tokens": 95681513.0,
      "step": 42860
    },
    {
      "entropy": 5.4932385921478275,
      "epoch": 4.814174201156849,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002922550206324732,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2243470087647438,
      "num_tokens": 95691670.0,
      "step": 42865
    },
    {
      "entropy": 5.4617818832397464,
      "epoch": 4.814735778064806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002922149898743787,
      "loss": 4.9209,
      "mean_token_accuracy": 0.23055792599916458,
      "num_tokens": 95702359.0,
      "step": 42870
    },
    {
      "entropy": 5.538556051254273,
      "epoch": 4.815297354972763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002921749585681523,
      "loss": 4.9684,
      "mean_token_accuracy": 0.23466706722974778,
      "num_tokens": 95714069.0,
      "step": 42875
    },
    {
      "entropy": 5.474097013473511,
      "epoch": 4.8158589318807214,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029213492671506855,
      "loss": 4.9277,
      "mean_token_accuracy": 0.21899548023939133,
      "num_tokens": 95724420.0,
      "step": 42880
    },
    {
      "entropy": 5.472040033340454,
      "epoch": 4.816420508788679,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029209489431640203,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22670433819293975,
      "num_tokens": 95735229.0,
      "step": 42885
    },
    {
      "entropy": 5.377327537536621,
      "epoch": 4.816982085696636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002920548613734275,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2301700621843338,
      "num_tokens": 95745246.0,
      "step": 42890
    },
    {
      "entropy": 5.412742328643799,
      "epoch": 4.817543662604594,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029201482788741953,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22148043662309647,
      "num_tokens": 95755752.0,
      "step": 42895
    },
    {
      "entropy": 5.507479476928711,
      "epoch": 4.818105239512551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002919747938596528,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2202646940946579,
      "num_tokens": 95766476.0,
      "step": 42900
    },
    {
      "entropy": 5.56518235206604,
      "epoch": 4.818666816420508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002919347592914021,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21483041793107988,
      "num_tokens": 95776862.0,
      "step": 42905
    },
    {
      "entropy": 5.50231442451477,
      "epoch": 4.8192283933284665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029189472418394215,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2246096283197403,
      "num_tokens": 95788216.0,
      "step": 42910
    },
    {
      "entropy": 5.474801969528198,
      "epoch": 4.819789970236424,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029185468853854755,
      "loss": 4.9437,
      "mean_token_accuracy": 0.22559484392404555,
      "num_tokens": 95799050.0,
      "step": 42915
    },
    {
      "entropy": 5.490537691116333,
      "epoch": 4.820351547144382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002918146523564932,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22519136667251588,
      "num_tokens": 95810763.0,
      "step": 42920
    },
    {
      "entropy": 5.518945407867432,
      "epoch": 4.820913124052339,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029177461563905375,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2278888165950775,
      "num_tokens": 95821730.0,
      "step": 42925
    },
    {
      "entropy": 5.580334568023682,
      "epoch": 4.821474700960296,
      "grad_norm": 1.0,
      "learning_rate": 0.00029173457838750405,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21703006327152252,
      "num_tokens": 95834037.0,
      "step": 42930
    },
    {
      "entropy": 5.436626768112182,
      "epoch": 4.822036277868254,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029169454060311886,
      "loss": 4.897,
      "mean_token_accuracy": 0.22954510599374772,
      "num_tokens": 95845074.0,
      "step": 42935
    },
    {
      "entropy": 5.544756650924683,
      "epoch": 4.822597854776212,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000291654502287173,
      "loss": 5.0267,
      "mean_token_accuracy": 0.2237790748476982,
      "num_tokens": 95855439.0,
      "step": 42940
    },
    {
      "entropy": 5.585258674621582,
      "epoch": 4.823159431684169,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002916144634409414,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22296859920024872,
      "num_tokens": 95866511.0,
      "step": 42945
    },
    {
      "entropy": 5.528468227386474,
      "epoch": 4.823721008592127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029157442406569885,
      "loss": 4.946,
      "mean_token_accuracy": 0.2257267192006111,
      "num_tokens": 95877795.0,
      "step": 42950
    },
    {
      "entropy": 5.467168188095092,
      "epoch": 4.824282585500084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002915343841627201,
      "loss": 4.9504,
      "mean_token_accuracy": 0.22310132831335067,
      "num_tokens": 95889039.0,
      "step": 42955
    },
    {
      "entropy": 5.446141195297241,
      "epoch": 4.824844162408041,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002914943437332802,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2304059624671936,
      "num_tokens": 95900101.0,
      "step": 42960
    },
    {
      "entropy": 5.4430849075317385,
      "epoch": 4.8254057393159995,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029145430277865406,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23036620318889617,
      "num_tokens": 95910023.0,
      "step": 42965
    },
    {
      "entropy": 5.454978370666504,
      "epoch": 4.825967316223957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002914142613001165,
      "loss": 4.8511,
      "mean_token_accuracy": 0.2359600380063057,
      "num_tokens": 95920610.0,
      "step": 42970
    },
    {
      "entropy": 5.397142028808593,
      "epoch": 4.826528893131915,
      "grad_norm": 1.0,
      "learning_rate": 0.00029137421929894235,
      "loss": 4.8459,
      "mean_token_accuracy": 0.23296699225902556,
      "num_tokens": 95931521.0,
      "step": 42975
    },
    {
      "entropy": 5.5700677871704105,
      "epoch": 4.827090470039872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002913341767764069,
      "loss": 5.0287,
      "mean_token_accuracy": 0.22543496936559676,
      "num_tokens": 95943179.0,
      "step": 42980
    },
    {
      "entropy": 5.535800552368164,
      "epoch": 4.827652046947829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029129413373378484,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22684888839721679,
      "num_tokens": 95955282.0,
      "step": 42985
    },
    {
      "entropy": 5.466316413879395,
      "epoch": 4.828213623855787,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029125409017235126,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22621762305498122,
      "num_tokens": 95966296.0,
      "step": 42990
    },
    {
      "entropy": 5.522579956054687,
      "epoch": 4.828775200763745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002912140460933812,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22446183264255523,
      "num_tokens": 95977039.0,
      "step": 42995
    },
    {
      "entropy": 5.586356353759766,
      "epoch": 4.829336777671702,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029117400149814953,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21627237200736998,
      "num_tokens": 95988041.0,
      "step": 43000
    },
    {
      "entropy": 5.426169490814209,
      "epoch": 4.82989835457966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002911339563879315,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23552426248788833,
      "num_tokens": 95998950.0,
      "step": 43005
    },
    {
      "entropy": 5.4661928653717045,
      "epoch": 4.830459931487617,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002910939107640019,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22449470907449723,
      "num_tokens": 96010380.0,
      "step": 43010
    },
    {
      "entropy": 5.456477975845337,
      "epoch": 4.831021508395574,
      "grad_norm": 1.046875,
      "learning_rate": 0.000291053864627636,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22461861968040467,
      "num_tokens": 96021860.0,
      "step": 43015
    },
    {
      "entropy": 5.551606893539429,
      "epoch": 4.831583085303532,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029101381798010885,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22579393535852432,
      "num_tokens": 96034127.0,
      "step": 43020
    },
    {
      "entropy": 5.483212423324585,
      "epoch": 4.83214466221149,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002909737708226954,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22763255536556243,
      "num_tokens": 96045029.0,
      "step": 43025
    },
    {
      "entropy": 5.4739926815032955,
      "epoch": 4.832706239119448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029093372315667105,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22885517179965972,
      "num_tokens": 96055409.0,
      "step": 43030
    },
    {
      "entropy": 5.4830348014831545,
      "epoch": 4.833267816027405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029089367498331075,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22858651727437973,
      "num_tokens": 96064928.0,
      "step": 43035
    },
    {
      "entropy": 5.51757845878601,
      "epoch": 4.833829392935362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029085362630388965,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2286778748035431,
      "num_tokens": 96077111.0,
      "step": 43040
    },
    {
      "entropy": 5.459997367858887,
      "epoch": 4.83439096984332,
      "grad_norm": 1.0,
      "learning_rate": 0.0002908135771196829,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2425326317548752,
      "num_tokens": 96088850.0,
      "step": 43045
    },
    {
      "entropy": 5.506393003463745,
      "epoch": 4.8349525467512775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002907735274319657,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22427674233913422,
      "num_tokens": 96099273.0,
      "step": 43050
    },
    {
      "entropy": 5.506422805786133,
      "epoch": 4.835514123659236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002907334772420134,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21593190282583236,
      "num_tokens": 96111332.0,
      "step": 43055
    },
    {
      "entropy": 5.503973770141601,
      "epoch": 4.836075700567193,
      "grad_norm": 1.03125,
      "learning_rate": 0.000290693426551101,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22386019825935363,
      "num_tokens": 96122225.0,
      "step": 43060
    },
    {
      "entropy": 5.486712312698364,
      "epoch": 4.83663727747515,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029065337536050384,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23582800030708312,
      "num_tokens": 96133465.0,
      "step": 43065
    },
    {
      "entropy": 5.417854309082031,
      "epoch": 4.837198854383108,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029061332367149705,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23494486659765243,
      "num_tokens": 96145036.0,
      "step": 43070
    },
    {
      "entropy": 5.350524187088013,
      "epoch": 4.837760431291065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029057327148535614,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23787256330251694,
      "num_tokens": 96154986.0,
      "step": 43075
    },
    {
      "entropy": 5.49015851020813,
      "epoch": 4.838322008199023,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000290533218803356,
      "loss": 5.019,
      "mean_token_accuracy": 0.22035479843616484,
      "num_tokens": 96166271.0,
      "step": 43080
    },
    {
      "entropy": 5.50107889175415,
      "epoch": 4.838883585106981,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029049316562677235,
      "loss": 5.0157,
      "mean_token_accuracy": 0.22897062748670577,
      "num_tokens": 96177203.0,
      "step": 43085
    },
    {
      "entropy": 5.562479639053345,
      "epoch": 4.839445162014938,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029045311195688016,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21930212676525115,
      "num_tokens": 96188749.0,
      "step": 43090
    },
    {
      "entropy": 5.525820398330689,
      "epoch": 4.840006738922895,
      "grad_norm": 1.09375,
      "learning_rate": 0.000290413057794955,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22649697661399842,
      "num_tokens": 96198701.0,
      "step": 43095
    },
    {
      "entropy": 5.6243125915527346,
      "epoch": 4.840568315830853,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029037300314227205,
      "loss": 5.0989,
      "mean_token_accuracy": 0.21776521652936937,
      "num_tokens": 96211016.0,
      "step": 43100
    },
    {
      "entropy": 5.5418964385986325,
      "epoch": 4.8411298927388104,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029033294800010666,
      "loss": 4.9881,
      "mean_token_accuracy": 0.21574331670999528,
      "num_tokens": 96222449.0,
      "step": 43105
    },
    {
      "entropy": 5.43619122505188,
      "epoch": 4.8416914696467686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029029289236973436,
      "loss": 4.9238,
      "mean_token_accuracy": 0.2272673323750496,
      "num_tokens": 96234434.0,
      "step": 43110
    },
    {
      "entropy": 5.447023534774781,
      "epoch": 4.842253046554726,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029025283625243043,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22875184416770936,
      "num_tokens": 96245612.0,
      "step": 43115
    },
    {
      "entropy": 5.486077213287354,
      "epoch": 4.842814623462683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029021277964947023,
      "loss": 4.8408,
      "mean_token_accuracy": 0.22796191275119781,
      "num_tokens": 96256790.0,
      "step": 43120
    },
    {
      "entropy": 5.554425048828125,
      "epoch": 4.843376200370641,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029017272256212927,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22286737114191055,
      "num_tokens": 96267562.0,
      "step": 43125
    },
    {
      "entropy": 5.487060546875,
      "epoch": 4.843937777278598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002901326649916829,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2283039927482605,
      "num_tokens": 96278469.0,
      "step": 43130
    },
    {
      "entropy": 5.534913349151611,
      "epoch": 4.8444993541865555,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029009260693940665,
      "loss": 5.0545,
      "mean_token_accuracy": 0.22176521718502046,
      "num_tokens": 96289460.0,
      "step": 43135
    },
    {
      "entropy": 5.509739923477173,
      "epoch": 4.845060931094514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002900525484065759,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2217169150710106,
      "num_tokens": 96300963.0,
      "step": 43140
    },
    {
      "entropy": 5.464438533782959,
      "epoch": 4.845622508002471,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029001248939446617,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22964289784431458,
      "num_tokens": 96311521.0,
      "step": 43145
    },
    {
      "entropy": 5.490410184860229,
      "epoch": 4.846184084910428,
      "grad_norm": 1.046875,
      "learning_rate": 0.000289972429904353,
      "loss": 4.9796,
      "mean_token_accuracy": 0.22369348853826523,
      "num_tokens": 96321959.0,
      "step": 43150
    },
    {
      "entropy": 5.615848350524902,
      "epoch": 4.846745661818386,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028993236993751174,
      "loss": 5.1513,
      "mean_token_accuracy": 0.21927058398723603,
      "num_tokens": 96333916.0,
      "step": 43155
    },
    {
      "entropy": 5.542170143127441,
      "epoch": 4.847307238726343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028989230949521817,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22720308154821395,
      "num_tokens": 96345081.0,
      "step": 43160
    },
    {
      "entropy": 5.458407020568847,
      "epoch": 4.8478688156343015,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028985224857874765,
      "loss": 4.8835,
      "mean_token_accuracy": 0.21910833865404128,
      "num_tokens": 96356394.0,
      "step": 43165
    },
    {
      "entropy": 5.365903425216675,
      "epoch": 4.848430392542259,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028981218718937573,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22394414842128754,
      "num_tokens": 96366952.0,
      "step": 43170
    },
    {
      "entropy": 5.465608882904053,
      "epoch": 4.848991969450216,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028977212532837804,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21827811598777772,
      "num_tokens": 96378165.0,
      "step": 43175
    },
    {
      "entropy": 5.584663343429566,
      "epoch": 4.849553546358174,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002897320629970302,
      "loss": 4.9745,
      "mean_token_accuracy": 0.23343257009983062,
      "num_tokens": 96390606.0,
      "step": 43180
    },
    {
      "entropy": 5.475037574768066,
      "epoch": 4.850115123266131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028969200019660774,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23044926077127456,
      "num_tokens": 96401089.0,
      "step": 43185
    },
    {
      "entropy": 5.513542795181275,
      "epoch": 4.8506767001740885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002896519369283862,
      "loss": 5.0271,
      "mean_token_accuracy": 0.22343237698078156,
      "num_tokens": 96411908.0,
      "step": 43190
    },
    {
      "entropy": 5.4668869972229,
      "epoch": 4.851238277082047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002896118731936414,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22565155029296874,
      "num_tokens": 96422859.0,
      "step": 43195
    },
    {
      "entropy": 5.484217739105224,
      "epoch": 4.851799853990004,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002895718089936489,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2281451478600502,
      "num_tokens": 96435122.0,
      "step": 43200
    },
    {
      "entropy": 5.500491094589234,
      "epoch": 4.852361430897961,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002895317443296843,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21310558915138245,
      "num_tokens": 96446916.0,
      "step": 43205
    },
    {
      "entropy": 5.478422403335571,
      "epoch": 4.852923007805919,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028949167920302337,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22337192595005034,
      "num_tokens": 96458241.0,
      "step": 43210
    },
    {
      "entropy": 5.517196607589722,
      "epoch": 4.853484584713876,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028945161361494177,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22371796518564224,
      "num_tokens": 96468581.0,
      "step": 43215
    },
    {
      "entropy": 5.467471933364868,
      "epoch": 4.854046161621834,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002894115475667151,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23045792132616044,
      "num_tokens": 96479119.0,
      "step": 43220
    },
    {
      "entropy": 5.524138116836548,
      "epoch": 4.854607738529792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028937148105961924,
      "loss": 5.0207,
      "mean_token_accuracy": 0.22098875194787979,
      "num_tokens": 96490306.0,
      "step": 43225
    },
    {
      "entropy": 5.388018655776977,
      "epoch": 4.855169315437749,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002893314140949298,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22937648743391037,
      "num_tokens": 96501222.0,
      "step": 43230
    },
    {
      "entropy": 5.510744619369507,
      "epoch": 4.855730892345707,
      "grad_norm": 1.0,
      "learning_rate": 0.00028929134667392265,
      "loss": 5.017,
      "mean_token_accuracy": 0.22227704524993896,
      "num_tokens": 96513283.0,
      "step": 43235
    },
    {
      "entropy": 5.48338098526001,
      "epoch": 4.856292469253664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028925127879787343,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22310985028743743,
      "num_tokens": 96523546.0,
      "step": 43240
    },
    {
      "entropy": 5.462495517730713,
      "epoch": 4.856854046161622,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002892112104680579,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22647399753332137,
      "num_tokens": 96534906.0,
      "step": 43245
    },
    {
      "entropy": 5.426185941696167,
      "epoch": 4.8574156230695795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028917114168575196,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22676144242286683,
      "num_tokens": 96547029.0,
      "step": 43250
    },
    {
      "entropy": 5.521560621261597,
      "epoch": 4.857977199977537,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002891310724522314,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2304091766476631,
      "num_tokens": 96558632.0,
      "step": 43255
    },
    {
      "entropy": 5.511607551574707,
      "epoch": 4.858538776885495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028909100276877203,
      "loss": 4.9579,
      "mean_token_accuracy": 0.2305277794599533,
      "num_tokens": 96568644.0,
      "step": 43260
    },
    {
      "entropy": 5.443189001083374,
      "epoch": 4.859100353793452,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028905093263664966,
      "loss": 4.889,
      "mean_token_accuracy": 0.22792809307575226,
      "num_tokens": 96578844.0,
      "step": 43265
    },
    {
      "entropy": 5.481982326507568,
      "epoch": 4.859661930701409,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028901086205714016,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22194791138172149,
      "num_tokens": 96590715.0,
      "step": 43270
    },
    {
      "entropy": 5.458102893829346,
      "epoch": 4.860223507609367,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028897079103151936,
      "loss": 4.915,
      "mean_token_accuracy": 0.22501542121171952,
      "num_tokens": 96601001.0,
      "step": 43275
    },
    {
      "entropy": 5.421552133560181,
      "epoch": 4.860785084517325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028893071956106315,
      "loss": 4.863,
      "mean_token_accuracy": 0.23103848248720169,
      "num_tokens": 96612591.0,
      "step": 43280
    },
    {
      "entropy": 5.383371782302857,
      "epoch": 4.861346661425282,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028889064764704735,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2272133231163025,
      "num_tokens": 96624353.0,
      "step": 43285
    },
    {
      "entropy": 5.485684251785278,
      "epoch": 4.86190823833324,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028885057529074804,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22158446609973909,
      "num_tokens": 96634293.0,
      "step": 43290
    },
    {
      "entropy": 5.571002006530762,
      "epoch": 4.862469815241197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028881050249344103,
      "loss": 5.004,
      "mean_token_accuracy": 0.21872071474790572,
      "num_tokens": 96645308.0,
      "step": 43295
    },
    {
      "entropy": 5.515964555740356,
      "epoch": 4.863031392149155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002887704292564022,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2245854392647743,
      "num_tokens": 96655198.0,
      "step": 43300
    },
    {
      "entropy": 5.416020393371582,
      "epoch": 4.8635929690571125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028873035558090766,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2234005257487297,
      "num_tokens": 96665753.0,
      "step": 43305
    },
    {
      "entropy": 5.4561761856079105,
      "epoch": 4.86415454596507,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002886902814682332,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22531859576702118,
      "num_tokens": 96676038.0,
      "step": 43310
    },
    {
      "entropy": 5.569386100769043,
      "epoch": 4.864716122873028,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028865020691965485,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22381328791379929,
      "num_tokens": 96687893.0,
      "step": 43315
    },
    {
      "entropy": 5.55728816986084,
      "epoch": 4.865277699780985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028861013193644865,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21830248087644577,
      "num_tokens": 96698937.0,
      "step": 43320
    },
    {
      "entropy": 5.464916563034057,
      "epoch": 4.865839276688942,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002885700565198906,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22659693658351898,
      "num_tokens": 96710117.0,
      "step": 43325
    },
    {
      "entropy": 5.424099779129028,
      "epoch": 4.8664008535969,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028852998067125675,
      "loss": 4.872,
      "mean_token_accuracy": 0.22963308840990065,
      "num_tokens": 96719831.0,
      "step": 43330
    },
    {
      "entropy": 5.471237754821777,
      "epoch": 4.8669624305048576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002884899043918229,
      "loss": 5.0403,
      "mean_token_accuracy": 0.2222439631819725,
      "num_tokens": 96732127.0,
      "step": 43335
    },
    {
      "entropy": 5.5598986625671385,
      "epoch": 4.867524007412815,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002884498276828653,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21828823685646057,
      "num_tokens": 96744406.0,
      "step": 43340
    },
    {
      "entropy": 5.515527296066284,
      "epoch": 4.868085584320773,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028840975054566007,
      "loss": 4.9368,
      "mean_token_accuracy": 0.23326347917318344,
      "num_tokens": 96754810.0,
      "step": 43345
    },
    {
      "entropy": 5.437874460220337,
      "epoch": 4.86864716122873,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028836967298148317,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22845949679613115,
      "num_tokens": 96765520.0,
      "step": 43350
    },
    {
      "entropy": 5.552954816818238,
      "epoch": 4.869208738136688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028832959499161065,
      "loss": 5.0711,
      "mean_token_accuracy": 0.2223411411046982,
      "num_tokens": 96777262.0,
      "step": 43355
    },
    {
      "entropy": 5.485742998123169,
      "epoch": 4.869770315044645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002882895165773187,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22478465884923934,
      "num_tokens": 96788452.0,
      "step": 43360
    },
    {
      "entropy": 5.440671443939209,
      "epoch": 4.870331891952603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002882494377398833,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2303902193903923,
      "num_tokens": 96799402.0,
      "step": 43365
    },
    {
      "entropy": 5.5242105484008786,
      "epoch": 4.870893468860561,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028820935848058066,
      "loss": 5.0282,
      "mean_token_accuracy": 0.22675897926092148,
      "num_tokens": 96810467.0,
      "step": 43370
    },
    {
      "entropy": 5.479265356063843,
      "epoch": 4.871455045768518,
      "grad_norm": 1.015625,
      "learning_rate": 0.000288169278800687,
      "loss": 4.8639,
      "mean_token_accuracy": 0.23196078985929489,
      "num_tokens": 96821480.0,
      "step": 43375
    },
    {
      "entropy": 5.43449764251709,
      "epoch": 4.872016622676475,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028812919870147837,
      "loss": 4.8681,
      "mean_token_accuracy": 0.2365085870027542,
      "num_tokens": 96832473.0,
      "step": 43380
    },
    {
      "entropy": 5.494804716110229,
      "epoch": 4.872578199584433,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028808911818423096,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22210337817668915,
      "num_tokens": 96844019.0,
      "step": 43385
    },
    {
      "entropy": 5.411420392990112,
      "epoch": 4.8731397764923905,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028804903725022094,
      "loss": 4.9007,
      "mean_token_accuracy": 0.225067076086998,
      "num_tokens": 96853902.0,
      "step": 43390
    },
    {
      "entropy": 5.498079872131347,
      "epoch": 4.873701353400348,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002880089559007245,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22783800661563874,
      "num_tokens": 96865066.0,
      "step": 43395
    },
    {
      "entropy": 5.522425746917724,
      "epoch": 4.874262930308306,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002879688741370179,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22417174726724626,
      "num_tokens": 96876329.0,
      "step": 43400
    },
    {
      "entropy": 5.4789206981658936,
      "epoch": 4.874824507216263,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002879287919603773,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22377232015132903,
      "num_tokens": 96887108.0,
      "step": 43405
    },
    {
      "entropy": 5.465694713592529,
      "epoch": 4.875386084124221,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028788870937207896,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22226077318191528,
      "num_tokens": 96898956.0,
      "step": 43410
    },
    {
      "entropy": 5.451403379440308,
      "epoch": 4.875947661032178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002878486263733991,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22732333987951278,
      "num_tokens": 96910889.0,
      "step": 43415
    },
    {
      "entropy": 5.447277402877807,
      "epoch": 4.876509237940136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000287808542965614,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23131901770830154,
      "num_tokens": 96921682.0,
      "step": 43420
    },
    {
      "entropy": 5.497397136688233,
      "epoch": 4.877070814848094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028776845914999996,
      "loss": 4.9055,
      "mean_token_accuracy": 0.23285999596118928,
      "num_tokens": 96933502.0,
      "step": 43425
    },
    {
      "entropy": 5.402181816101074,
      "epoch": 4.877632391756051,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002877283749278332,
      "loss": 4.8323,
      "mean_token_accuracy": 0.23040764033794403,
      "num_tokens": 96945621.0,
      "step": 43430
    },
    {
      "entropy": 5.575005388259887,
      "epoch": 4.878193968664009,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028768829030039,
      "loss": 5.0676,
      "mean_token_accuracy": 0.21556686162948607,
      "num_tokens": 96957164.0,
      "step": 43435
    },
    {
      "entropy": 5.551993036270142,
      "epoch": 4.878755545571966,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028764820526894675,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2212304264307022,
      "num_tokens": 96968646.0,
      "step": 43440
    },
    {
      "entropy": 5.522346162796021,
      "epoch": 4.879317122479923,
      "grad_norm": 1.0,
      "learning_rate": 0.00028760811983477976,
      "loss": 4.8988,
      "mean_token_accuracy": 0.23169928789138794,
      "num_tokens": 96979855.0,
      "step": 43445
    },
    {
      "entropy": 5.453479623794555,
      "epoch": 4.8798786993878815,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028756803399916534,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23626819252967834,
      "num_tokens": 96989603.0,
      "step": 43450
    },
    {
      "entropy": 5.459464120864868,
      "epoch": 4.880440276295839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002875279477633798,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2319524884223938,
      "num_tokens": 96999511.0,
      "step": 43455
    },
    {
      "entropy": 5.51837477684021,
      "epoch": 4.881001853203796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002874878611286995,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22385403364896775,
      "num_tokens": 97011210.0,
      "step": 43460
    },
    {
      "entropy": 5.509891510009766,
      "epoch": 4.881563430111754,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000287447774096401,
      "loss": 4.9791,
      "mean_token_accuracy": 0.23082538247108458,
      "num_tokens": 97022351.0,
      "step": 43465
    },
    {
      "entropy": 5.506597995758057,
      "epoch": 4.882125007019711,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028740768666776046,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22108504474163054,
      "num_tokens": 97033269.0,
      "step": 43470
    },
    {
      "entropy": 5.504105281829834,
      "epoch": 4.8826865839276685,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028736759884405433,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22101922929286957,
      "num_tokens": 97045278.0,
      "step": 43475
    },
    {
      "entropy": 5.492180681228637,
      "epoch": 4.883248160835627,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002873275106265591,
      "loss": 4.8736,
      "mean_token_accuracy": 0.23586749881505967,
      "num_tokens": 97055617.0,
      "step": 43480
    },
    {
      "entropy": 5.557194566726684,
      "epoch": 4.883809737743584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028728742201655116,
      "loss": 5.0597,
      "mean_token_accuracy": 0.22560517191886903,
      "num_tokens": 97068815.0,
      "step": 43485
    },
    {
      "entropy": 5.503067255020142,
      "epoch": 4.884371314651542,
      "grad_norm": 1.078125,
      "learning_rate": 0.000287247333015307,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21703422367572783,
      "num_tokens": 97078812.0,
      "step": 43490
    },
    {
      "entropy": 5.484073114395142,
      "epoch": 4.884932891559499,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002872072436241029,
      "loss": 4.8596,
      "mean_token_accuracy": 0.24206791520118714,
      "num_tokens": 97090423.0,
      "step": 43495
    },
    {
      "entropy": 5.445108366012573,
      "epoch": 4.885494468467456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002871671538442155,
      "loss": 4.7944,
      "mean_token_accuracy": 0.2338974729180336,
      "num_tokens": 97100756.0,
      "step": 43500
    },
    {
      "entropy": 5.383710861206055,
      "epoch": 4.8860560453754145,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002871270636769212,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2295299306511879,
      "num_tokens": 97111016.0,
      "step": 43505
    },
    {
      "entropy": 5.492556619644165,
      "epoch": 4.886617622283372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002870869731234965,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21879408806562423,
      "num_tokens": 97122148.0,
      "step": 43510
    },
    {
      "entropy": 5.493264436721802,
      "epoch": 4.887179199191329,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002870468821852178,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22374285459518434,
      "num_tokens": 97133301.0,
      "step": 43515
    },
    {
      "entropy": 5.484845876693726,
      "epoch": 4.887740776099287,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028700679086336175,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22179259955883027,
      "num_tokens": 97144637.0,
      "step": 43520
    },
    {
      "entropy": 5.445652151107788,
      "epoch": 4.888302353007244,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002869666991592048,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23567636758089067,
      "num_tokens": 97156047.0,
      "step": 43525
    },
    {
      "entropy": 5.464514970779419,
      "epoch": 4.8888639299152015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002869266070740235,
      "loss": 4.919,
      "mean_token_accuracy": 0.22343428581953048,
      "num_tokens": 97167767.0,
      "step": 43530
    },
    {
      "entropy": 5.380908632278443,
      "epoch": 4.88942550682316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002868865146090944,
      "loss": 4.888,
      "mean_token_accuracy": 0.22416869848966597,
      "num_tokens": 97179326.0,
      "step": 43535
    },
    {
      "entropy": 5.449964809417724,
      "epoch": 4.889987083731117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028684642176569414,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22886897772550582,
      "num_tokens": 97190437.0,
      "step": 43540
    },
    {
      "entropy": 5.4278027534484865,
      "epoch": 4.890548660639075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002868063285450992,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23216653764247894,
      "num_tokens": 97200913.0,
      "step": 43545
    },
    {
      "entropy": 5.406724166870117,
      "epoch": 4.891110237547032,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002867662349485861,
      "loss": 4.848,
      "mean_token_accuracy": 0.22910453528165817,
      "num_tokens": 97210935.0,
      "step": 43550
    },
    {
      "entropy": 5.462351560592651,
      "epoch": 4.891671814454989,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028672614097743156,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23270974457263946,
      "num_tokens": 97221568.0,
      "step": 43555
    },
    {
      "entropy": 5.460868787765503,
      "epoch": 4.892233391362947,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028668604663291213,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22260284125804902,
      "num_tokens": 97232149.0,
      "step": 43560
    },
    {
      "entropy": 5.490008592605591,
      "epoch": 4.892794968270905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028664595191630455,
      "loss": 4.9845,
      "mean_token_accuracy": 0.22320212423801422,
      "num_tokens": 97242800.0,
      "step": 43565
    },
    {
      "entropy": 5.497044944763184,
      "epoch": 4.893356545178862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028660585682888524,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2296442434191704,
      "num_tokens": 97253681.0,
      "step": 43570
    },
    {
      "entropy": 5.551821708679199,
      "epoch": 4.89391812208682,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028656576137193096,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22383990436792373,
      "num_tokens": 97264949.0,
      "step": 43575
    },
    {
      "entropy": 5.529165506362915,
      "epoch": 4.894479698994777,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002865256655467184,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2355342835187912,
      "num_tokens": 97276670.0,
      "step": 43580
    },
    {
      "entropy": 5.421679306030273,
      "epoch": 4.895041275902734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002864855693545241,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23442933708429337,
      "num_tokens": 97287134.0,
      "step": 43585
    },
    {
      "entropy": 5.4618288516998295,
      "epoch": 4.8956028528106925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028644547279662485,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2260987564921379,
      "num_tokens": 97298250.0,
      "step": 43590
    },
    {
      "entropy": 5.414427995681763,
      "epoch": 4.89616442971865,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028640537587429734,
      "loss": 4.868,
      "mean_token_accuracy": 0.2258552610874176,
      "num_tokens": 97309144.0,
      "step": 43595
    },
    {
      "entropy": 5.511423587799072,
      "epoch": 4.896726006626608,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002863652785888182,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22494738548994064,
      "num_tokens": 97320586.0,
      "step": 43600
    },
    {
      "entropy": 5.371093273162842,
      "epoch": 4.897287583534565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002863251809414642,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23583719432353972,
      "num_tokens": 97330924.0,
      "step": 43605
    },
    {
      "entropy": 5.544574975967407,
      "epoch": 4.897849160442522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000286285082933512,
      "loss": 5.0183,
      "mean_token_accuracy": 0.22143056988716125,
      "num_tokens": 97341926.0,
      "step": 43610
    },
    {
      "entropy": 5.54961895942688,
      "epoch": 4.89841073735048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002862449845662385,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2183466449379921,
      "num_tokens": 97353809.0,
      "step": 43615
    },
    {
      "entropy": 5.51994481086731,
      "epoch": 4.898972314258438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002862048858409201,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2276705950498581,
      "num_tokens": 97365317.0,
      "step": 43620
    },
    {
      "entropy": 5.496413850784302,
      "epoch": 4.899533891166396,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000286164786758834,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22595206797122955,
      "num_tokens": 97375581.0,
      "step": 43625
    },
    {
      "entropy": 5.637809753417969,
      "epoch": 4.900095468074353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028612468732125667,
      "loss": 5.079,
      "mean_token_accuracy": 0.2206890270113945,
      "num_tokens": 97387371.0,
      "step": 43630
    },
    {
      "entropy": 5.454562950134277,
      "epoch": 4.90065704498231,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002860845875294651,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23474393635988236,
      "num_tokens": 97398929.0,
      "step": 43635
    },
    {
      "entropy": 5.613053798675537,
      "epoch": 4.901218621890268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002860444873847358,
      "loss": 5.1208,
      "mean_token_accuracy": 0.21043747365474702,
      "num_tokens": 97410469.0,
      "step": 43640
    },
    {
      "entropy": 5.478086805343628,
      "epoch": 4.9017801987982255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028600438688834587,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22578085958957672,
      "num_tokens": 97422253.0,
      "step": 43645
    },
    {
      "entropy": 5.5141050815582275,
      "epoch": 4.902341775706183,
      "grad_norm": 1.03125,
      "learning_rate": 0.000285964286041572,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21645319312810898,
      "num_tokens": 97433384.0,
      "step": 43650
    },
    {
      "entropy": 5.537349271774292,
      "epoch": 4.902903352614141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028592418484569095,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22643389403820038,
      "num_tokens": 97444755.0,
      "step": 43655
    },
    {
      "entropy": 5.401272439956665,
      "epoch": 4.903464929522098,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002858840833019795,
      "loss": 4.7776,
      "mean_token_accuracy": 0.23039863705635072,
      "num_tokens": 97456156.0,
      "step": 43660
    },
    {
      "entropy": 5.452666282653809,
      "epoch": 4.904026506430055,
      "grad_norm": 1.125,
      "learning_rate": 0.0002858439814117148,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23238523453474044,
      "num_tokens": 97467494.0,
      "step": 43665
    },
    {
      "entropy": 5.402915334701538,
      "epoch": 4.904588083338013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028580387917617343,
      "loss": 4.8592,
      "mean_token_accuracy": 0.2310606062412262,
      "num_tokens": 97478317.0,
      "step": 43670
    },
    {
      "entropy": 5.449511241912842,
      "epoch": 4.9051496602459705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002857637765966324,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2300128698348999,
      "num_tokens": 97488530.0,
      "step": 43675
    },
    {
      "entropy": 5.451708030700684,
      "epoch": 4.905711237153929,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002857236736743685,
      "loss": 4.879,
      "mean_token_accuracy": 0.23411962240934373,
      "num_tokens": 97501148.0,
      "step": 43680
    },
    {
      "entropy": 5.477384996414185,
      "epoch": 4.906272814061886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002856835704106587,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22490883618593216,
      "num_tokens": 97512136.0,
      "step": 43685
    },
    {
      "entropy": 5.482175016403199,
      "epoch": 4.906834390969843,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028564346680677995,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2293742999434471,
      "num_tokens": 97523787.0,
      "step": 43690
    },
    {
      "entropy": 5.537923097610474,
      "epoch": 4.907395967877801,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002856033628640089,
      "loss": 4.963,
      "mean_token_accuracy": 0.22516103088855743,
      "num_tokens": 97534851.0,
      "step": 43695
    },
    {
      "entropy": 5.47380428314209,
      "epoch": 4.907957544785758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002855632585836228,
      "loss": 4.8665,
      "mean_token_accuracy": 0.22429047971963884,
      "num_tokens": 97545160.0,
      "step": 43700
    },
    {
      "entropy": 5.489389705657959,
      "epoch": 4.908519121693716,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028552315396689846,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22395735681056977,
      "num_tokens": 97556321.0,
      "step": 43705
    },
    {
      "entropy": 5.48088698387146,
      "epoch": 4.909080698601674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002854830490151128,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2215038850903511,
      "num_tokens": 97567406.0,
      "step": 43710
    },
    {
      "entropy": 5.4700844287872314,
      "epoch": 4.909642275509631,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028544294372954276,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2286268636584282,
      "num_tokens": 97577998.0,
      "step": 43715
    },
    {
      "entropy": 5.453070497512817,
      "epoch": 4.910203852417588,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002854028381114654,
      "loss": 4.897,
      "mean_token_accuracy": 0.22750664204359056,
      "num_tokens": 97588262.0,
      "step": 43720
    },
    {
      "entropy": 5.405086994171143,
      "epoch": 4.910765429325546,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028536273216215756,
      "loss": 4.8337,
      "mean_token_accuracy": 0.231433866918087,
      "num_tokens": 97600011.0,
      "step": 43725
    },
    {
      "entropy": 5.566233587265015,
      "epoch": 4.9113270062335035,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028532262588289636,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2260650783777237,
      "num_tokens": 97612037.0,
      "step": 43730
    },
    {
      "entropy": 5.468508529663086,
      "epoch": 4.911888583141462,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028528251927495874,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23383875489234923,
      "num_tokens": 97622643.0,
      "step": 43735
    },
    {
      "entropy": 5.519642305374146,
      "epoch": 4.912450160049419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028524241233962176,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22195694744586944,
      "num_tokens": 97634720.0,
      "step": 43740
    },
    {
      "entropy": 5.62167763710022,
      "epoch": 4.913011736957376,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028520230507816236,
      "loss": 5.1189,
      "mean_token_accuracy": 0.20986909419298172,
      "num_tokens": 97646314.0,
      "step": 43745
    },
    {
      "entropy": 5.542956161499023,
      "epoch": 4.913573313865334,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028516219749185755,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22890951484441757,
      "num_tokens": 97657478.0,
      "step": 43750
    },
    {
      "entropy": 5.487882089614868,
      "epoch": 4.914134890773291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002851220895819845,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22739728093147277,
      "num_tokens": 97668084.0,
      "step": 43755
    },
    {
      "entropy": 5.470142841339111,
      "epoch": 4.914696467681249,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028508198134982016,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2229931280016899,
      "num_tokens": 97678769.0,
      "step": 43760
    },
    {
      "entropy": 5.473022937774658,
      "epoch": 4.915258044589207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028504187279664164,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22215166836977004,
      "num_tokens": 97689587.0,
      "step": 43765
    },
    {
      "entropy": 5.471416330337524,
      "epoch": 4.915819621497164,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000285001763923726,
      "loss": 4.9589,
      "mean_token_accuracy": 0.23041018843650818,
      "num_tokens": 97700334.0,
      "step": 43770
    },
    {
      "entropy": 5.470821762084961,
      "epoch": 4.916381198405121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002849616547323503,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22976914048194885,
      "num_tokens": 97713038.0,
      "step": 43775
    },
    {
      "entropy": 5.512460470199585,
      "epoch": 4.916942775313079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028492154522379167,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2234805256128311,
      "num_tokens": 97724651.0,
      "step": 43780
    },
    {
      "entropy": 5.489507436752319,
      "epoch": 4.917504352221036,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002848814353993273,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22095756977796555,
      "num_tokens": 97735862.0,
      "step": 43785
    },
    {
      "entropy": 5.482861423492432,
      "epoch": 4.9180659291289945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028484132526023404,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22495709657669066,
      "num_tokens": 97747356.0,
      "step": 43790
    },
    {
      "entropy": 5.477084302902222,
      "epoch": 4.918627506036952,
      "grad_norm": 1.0,
      "learning_rate": 0.0002848012148077892,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22266436964273453,
      "num_tokens": 97758670.0,
      "step": 43795
    },
    {
      "entropy": 5.542899703979492,
      "epoch": 4.919189082944909,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00028476110404326984,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22702872902154922,
      "num_tokens": 97770291.0,
      "step": 43800
    },
    {
      "entropy": 5.478105211257935,
      "epoch": 4.919750659852867,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002847209929679532,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22788389027118683,
      "num_tokens": 97781567.0,
      "step": 43805
    },
    {
      "entropy": 5.408502626419067,
      "epoch": 4.920312236760824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002846808815831164,
      "loss": 4.86,
      "mean_token_accuracy": 0.2323884218931198,
      "num_tokens": 97792281.0,
      "step": 43810
    },
    {
      "entropy": 5.360383701324463,
      "epoch": 4.920873813668782,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002846407698900365,
      "loss": 4.8192,
      "mean_token_accuracy": 0.23352174162864686,
      "num_tokens": 97803606.0,
      "step": 43815
    },
    {
      "entropy": 5.444654941558838,
      "epoch": 4.92143539057674,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028460065788999073,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22703173160552978,
      "num_tokens": 97816024.0,
      "step": 43820
    },
    {
      "entropy": 5.498255920410156,
      "epoch": 4.921996967484697,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002845605455842564,
      "loss": 4.9047,
      "mean_token_accuracy": 0.2329721122980118,
      "num_tokens": 97827121.0,
      "step": 43825
    },
    {
      "entropy": 5.532002735137939,
      "epoch": 4.922558544392655,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002845204329741106,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22677357494831085,
      "num_tokens": 97838758.0,
      "step": 43830
    },
    {
      "entropy": 5.364478683471679,
      "epoch": 4.923120121300612,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002844803200608304,
      "loss": 4.7857,
      "mean_token_accuracy": 0.24071400761604309,
      "num_tokens": 97850208.0,
      "step": 43835
    },
    {
      "entropy": 5.335463762283325,
      "epoch": 4.923681698208569,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002844402068456931,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2278517797589302,
      "num_tokens": 97860210.0,
      "step": 43840
    },
    {
      "entropy": 5.443119955062866,
      "epoch": 4.9242432751165275,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002844000933299761,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22583219408988953,
      "num_tokens": 97871348.0,
      "step": 43845
    },
    {
      "entropy": 5.471803283691406,
      "epoch": 4.924804852024485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028435997951495643,
      "loss": 4.946,
      "mean_token_accuracy": 0.22777656614780425,
      "num_tokens": 97882767.0,
      "step": 43850
    },
    {
      "entropy": 5.507179927825928,
      "epoch": 4.925366428932442,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002843198654019114,
      "loss": 4.979,
      "mean_token_accuracy": 0.22312871366739273,
      "num_tokens": 97895690.0,
      "step": 43855
    },
    {
      "entropy": 5.497745227813721,
      "epoch": 4.9259280058404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002842797509921182,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22744575291872024,
      "num_tokens": 97906951.0,
      "step": 43860
    },
    {
      "entropy": 5.51332278251648,
      "epoch": 4.926489582748357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028423963628685414,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2212610736489296,
      "num_tokens": 97918538.0,
      "step": 43865
    },
    {
      "entropy": 5.456428718566895,
      "epoch": 4.927051159656315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002841995212873965,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23356722593307494,
      "num_tokens": 97930175.0,
      "step": 43870
    },
    {
      "entropy": 5.484073066711426,
      "epoch": 4.927612736564273,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002841594059950225,
      "loss": 4.981,
      "mean_token_accuracy": 0.2267455652356148,
      "num_tokens": 97940870.0,
      "step": 43875
    },
    {
      "entropy": 5.453384733200073,
      "epoch": 4.92817431347223,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028411929041100947,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22565243691205977,
      "num_tokens": 97951492.0,
      "step": 43880
    },
    {
      "entropy": 5.45724835395813,
      "epoch": 4.928735890380188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002840791745366347,
      "loss": 4.8658,
      "mean_token_accuracy": 0.23786222487688063,
      "num_tokens": 97962243.0,
      "step": 43885
    },
    {
      "entropy": 5.430580425262451,
      "epoch": 4.929297467288145,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028403905837317547,
      "loss": 4.8653,
      "mean_token_accuracy": 0.233624367415905,
      "num_tokens": 97973150.0,
      "step": 43890
    },
    {
      "entropy": 5.428069686889648,
      "epoch": 4.929859044196102,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028399894192190916,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22354853451251983,
      "num_tokens": 97985314.0,
      "step": 43895
    },
    {
      "entropy": 5.464573907852173,
      "epoch": 4.93042062110406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028395882518411305,
      "loss": 4.931,
      "mean_token_accuracy": 0.22880469262599945,
      "num_tokens": 97997065.0,
      "step": 43900
    },
    {
      "entropy": 5.455758190155029,
      "epoch": 4.930982198012018,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002839187081610645,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22286141961812972,
      "num_tokens": 98007576.0,
      "step": 43905
    },
    {
      "entropy": 5.448855400085449,
      "epoch": 4.931543774919975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002838785908540407,
      "loss": 4.903,
      "mean_token_accuracy": 0.22856711596250534,
      "num_tokens": 98017870.0,
      "step": 43910
    },
    {
      "entropy": 5.489424753189087,
      "epoch": 4.932105351827933,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028383847326431924,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21831856817007064,
      "num_tokens": 98029076.0,
      "step": 43915
    },
    {
      "entropy": 5.524903726577759,
      "epoch": 4.93266692873589,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028379835539317727,
      "loss": 4.888,
      "mean_token_accuracy": 0.2297412186861038,
      "num_tokens": 98042099.0,
      "step": 43920
    },
    {
      "entropy": 5.439872074127197,
      "epoch": 4.933228505643848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002837582372418923,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23236961662769318,
      "num_tokens": 98053335.0,
      "step": 43925
    },
    {
      "entropy": 5.52184419631958,
      "epoch": 4.9337900825518055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002837181188117417,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22211210429668427,
      "num_tokens": 98063723.0,
      "step": 43930
    },
    {
      "entropy": 5.469312238693237,
      "epoch": 4.934351659459763,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002836780001040028,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23140270113945008,
      "num_tokens": 98074756.0,
      "step": 43935
    },
    {
      "entropy": 5.438142967224121,
      "epoch": 4.934913236367721,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000283637881119953,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22842407822608948,
      "num_tokens": 98086584.0,
      "step": 43940
    },
    {
      "entropy": 5.515545415878296,
      "epoch": 4.935474813275678,
      "grad_norm": 1.0,
      "learning_rate": 0.00028359776186086965,
      "loss": 5.1409,
      "mean_token_accuracy": 0.2150951698422432,
      "num_tokens": 98099996.0,
      "step": 43945
    },
    {
      "entropy": 5.436253595352173,
      "epoch": 4.936036390183635,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002835576423280303,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2246584638953209,
      "num_tokens": 98111740.0,
      "step": 43950
    },
    {
      "entropy": 5.535261011123657,
      "epoch": 4.936597967091593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002835175225227123,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22615235447883605,
      "num_tokens": 98123946.0,
      "step": 43955
    },
    {
      "entropy": 5.492046594619751,
      "epoch": 4.937159543999551,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002834774024461931,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22415673583745957,
      "num_tokens": 98134980.0,
      "step": 43960
    },
    {
      "entropy": 5.50024094581604,
      "epoch": 4.937721120907508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028343728209975,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22673149555921554,
      "num_tokens": 98145639.0,
      "step": 43965
    },
    {
      "entropy": 5.487412738800049,
      "epoch": 4.938282697815466,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028339716148466067,
      "loss": 4.9712,
      "mean_token_accuracy": 0.23011333793401717,
      "num_tokens": 98156078.0,
      "step": 43970
    },
    {
      "entropy": 5.5019269466400145,
      "epoch": 4.938844274723423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028335704060220246,
      "loss": 5.0459,
      "mean_token_accuracy": 0.21944239288568496,
      "num_tokens": 98168486.0,
      "step": 43975
    },
    {
      "entropy": 5.581345653533935,
      "epoch": 4.939405851631381,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028331691945365274,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22500444054603577,
      "num_tokens": 98178196.0,
      "step": 43980
    },
    {
      "entropy": 5.51515645980835,
      "epoch": 4.9399674285393385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002832767980402892,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22391725480556487,
      "num_tokens": 98188912.0,
      "step": 43985
    },
    {
      "entropy": 5.500329208374024,
      "epoch": 4.940529005447296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002832366763633892,
      "loss": 5.03,
      "mean_token_accuracy": 0.22130554914474487,
      "num_tokens": 98200207.0,
      "step": 43990
    },
    {
      "entropy": 5.3709522724151615,
      "epoch": 4.941090582355254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028319655442423015,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23325294405221939,
      "num_tokens": 98210846.0,
      "step": 43995
    },
    {
      "entropy": 5.5325939655303955,
      "epoch": 4.941652159263211,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028315643222408965,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21521669179201125,
      "num_tokens": 98222303.0,
      "step": 44000
    },
    {
      "entropy": 5.409994316101074,
      "epoch": 4.942213736171169,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028311630976424526,
      "loss": 4.8695,
      "mean_token_accuracy": 0.22886465042829512,
      "num_tokens": 98233374.0,
      "step": 44005
    },
    {
      "entropy": 5.581387376785278,
      "epoch": 4.942775313079126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002830761870459743,
      "loss": 5.0558,
      "mean_token_accuracy": 0.22472795397043227,
      "num_tokens": 98244840.0,
      "step": 44010
    },
    {
      "entropy": 5.429254150390625,
      "epoch": 4.9433368899870835,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002830360640705545,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22237589061260224,
      "num_tokens": 98254829.0,
      "step": 44015
    },
    {
      "entropy": 5.523613548278808,
      "epoch": 4.943898466895042,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028299594083926324,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2274458423256874,
      "num_tokens": 98266296.0,
      "step": 44020
    },
    {
      "entropy": 5.4587208271026615,
      "epoch": 4.944460043802999,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028295581735337817,
      "loss": 4.9136,
      "mean_token_accuracy": 0.23545611947774886,
      "num_tokens": 98277723.0,
      "step": 44025
    },
    {
      "entropy": 5.50061445236206,
      "epoch": 4.945021620710956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028291569361417674,
      "loss": 4.9634,
      "mean_token_accuracy": 0.23360185325145721,
      "num_tokens": 98288799.0,
      "step": 44030
    },
    {
      "entropy": 5.466094207763672,
      "epoch": 4.945583197618914,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002828755696229366,
      "loss": 4.964,
      "mean_token_accuracy": 0.22313659340143205,
      "num_tokens": 98299731.0,
      "step": 44035
    },
    {
      "entropy": 5.493797683715821,
      "epoch": 4.946144774526871,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002828354453809352,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22603518664836883,
      "num_tokens": 98310607.0,
      "step": 44040
    },
    {
      "entropy": 5.458073282241822,
      "epoch": 4.946706351434829,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002827953208894503,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2266363248229027,
      "num_tokens": 98320149.0,
      "step": 44045
    },
    {
      "entropy": 5.44878077507019,
      "epoch": 4.947267928342787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002827551961497593,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22677122056484222,
      "num_tokens": 98330816.0,
      "step": 44050
    },
    {
      "entropy": 5.383896064758301,
      "epoch": 4.947829505250744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002827150711631398,
      "loss": 4.787,
      "mean_token_accuracy": 0.23135144710540773,
      "num_tokens": 98342041.0,
      "step": 44055
    },
    {
      "entropy": 5.450576543807983,
      "epoch": 4.948391082158702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028267494593086945,
      "loss": 5.0212,
      "mean_token_accuracy": 0.22442660331726075,
      "num_tokens": 98353891.0,
      "step": 44060
    },
    {
      "entropy": 5.5171675205230715,
      "epoch": 4.948952659066659,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028263482045422597,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2282612517476082,
      "num_tokens": 98364931.0,
      "step": 44065
    },
    {
      "entropy": 5.582327079772949,
      "epoch": 4.9495142359746165,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002825946947344867,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22581800669431687,
      "num_tokens": 98376787.0,
      "step": 44070
    },
    {
      "entropy": 5.543013286590576,
      "epoch": 4.950075812882575,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002825545687729295,
      "loss": 4.9432,
      "mean_token_accuracy": 0.2168271943926811,
      "num_tokens": 98387750.0,
      "step": 44075
    },
    {
      "entropy": 5.453578996658325,
      "epoch": 4.950637389790532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028251444257083186,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23152336925268174,
      "num_tokens": 98399309.0,
      "step": 44080
    },
    {
      "entropy": 5.420279550552368,
      "epoch": 4.951198966698489,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028247431612947146,
      "loss": 4.843,
      "mean_token_accuracy": 0.22479046881198883,
      "num_tokens": 98410850.0,
      "step": 44085
    },
    {
      "entropy": 5.521691131591797,
      "epoch": 4.951760543606447,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002824341894501259,
      "loss": 4.9813,
      "mean_token_accuracy": 0.22611639201641082,
      "num_tokens": 98422166.0,
      "step": 44090
    },
    {
      "entropy": 5.501344347000122,
      "epoch": 4.952322120514404,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028239406253407286,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22467085123062133,
      "num_tokens": 98432885.0,
      "step": 44095
    },
    {
      "entropy": 5.536303520202637,
      "epoch": 4.952883697422362,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028235393538259007,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22571698129177092,
      "num_tokens": 98444513.0,
      "step": 44100
    },
    {
      "entropy": 5.493028354644776,
      "epoch": 4.95344527433032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002823138079969551,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22864649146795274,
      "num_tokens": 98456009.0,
      "step": 44105
    },
    {
      "entropy": 5.543236780166626,
      "epoch": 4.954006851238277,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028227368037844557,
      "loss": 5.0214,
      "mean_token_accuracy": 0.22533411979675294,
      "num_tokens": 98466282.0,
      "step": 44110
    },
    {
      "entropy": 5.514357328414917,
      "epoch": 4.954568428146235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028223355252833933,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22610896974802017,
      "num_tokens": 98478334.0,
      "step": 44115
    },
    {
      "entropy": 5.509020900726318,
      "epoch": 4.955130005054192,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028219342444791396,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22951127737760543,
      "num_tokens": 98488446.0,
      "step": 44120
    },
    {
      "entropy": 5.459474802017212,
      "epoch": 4.955691581962149,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028215329613844715,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23202061802148818,
      "num_tokens": 98499698.0,
      "step": 44125
    },
    {
      "entropy": 5.474583292007447,
      "epoch": 4.9562531588701075,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028211316760121656,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2336056187748909,
      "num_tokens": 98509980.0,
      "step": 44130
    },
    {
      "entropy": 5.576911211013794,
      "epoch": 4.956814735778065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002820730388375001,
      "loss": 5.0259,
      "mean_token_accuracy": 0.22568201571702956,
      "num_tokens": 98521406.0,
      "step": 44135
    },
    {
      "entropy": 5.622286176681518,
      "epoch": 4.957376312686022,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002820329098485752,
      "loss": 5.1019,
      "mean_token_accuracy": 0.22162072956562043,
      "num_tokens": 98532845.0,
      "step": 44140
    },
    {
      "entropy": 5.426628541946411,
      "epoch": 4.95793788959398,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028199278063571976,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23653644174337388,
      "num_tokens": 98543639.0,
      "step": 44145
    },
    {
      "entropy": 5.521172904968262,
      "epoch": 4.958499466501937,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028195265120021157,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22431274652481079,
      "num_tokens": 98554604.0,
      "step": 44150
    },
    {
      "entropy": 5.513560819625854,
      "epoch": 4.9590610434098945,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028191252154332816,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22151417583227156,
      "num_tokens": 98566388.0,
      "step": 44155
    },
    {
      "entropy": 5.518068408966064,
      "epoch": 4.959622620317853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028187239166634737,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2250673845410347,
      "num_tokens": 98577215.0,
      "step": 44160
    },
    {
      "entropy": 5.592435216903686,
      "epoch": 4.96018419722581,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028183226157054705,
      "loss": 5.1274,
      "mean_token_accuracy": 0.2150576591491699,
      "num_tokens": 98588640.0,
      "step": 44165
    },
    {
      "entropy": 5.505161619186401,
      "epoch": 4.960745774133768,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002817921312572048,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2286597430706024,
      "num_tokens": 98599710.0,
      "step": 44170
    },
    {
      "entropy": 5.486812114715576,
      "epoch": 4.961307351041725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002817520007275985,
      "loss": 5.0085,
      "mean_token_accuracy": 0.22273856550455093,
      "num_tokens": 98611510.0,
      "step": 44175
    },
    {
      "entropy": 5.465967226028442,
      "epoch": 4.961868927949682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002817118699830058,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2239406242966652,
      "num_tokens": 98622982.0,
      "step": 44180
    },
    {
      "entropy": 5.590875434875488,
      "epoch": 4.9624305048576405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002816717390247046,
      "loss": 5.0748,
      "mean_token_accuracy": 0.22316273152828217,
      "num_tokens": 98634216.0,
      "step": 44185
    },
    {
      "entropy": 5.561990737915039,
      "epoch": 4.962992081765598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002816316078539727,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21978497356176377,
      "num_tokens": 98646469.0,
      "step": 44190
    },
    {
      "entropy": 5.590316295623779,
      "epoch": 4.963553658673556,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00028159147647208784,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21300797164440155,
      "num_tokens": 98659249.0,
      "step": 44195
    },
    {
      "entropy": 5.490548753738404,
      "epoch": 4.964115235581513,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002815513448803277,
      "loss": 4.8714,
      "mean_token_accuracy": 0.23387259542942046,
      "num_tokens": 98671322.0,
      "step": 44200
    },
    {
      "entropy": 5.53375473022461,
      "epoch": 4.96467681248947,
      "grad_norm": 1.0,
      "learning_rate": 0.0002815112130799703,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2266753524541855,
      "num_tokens": 98684236.0,
      "step": 44205
    },
    {
      "entropy": 5.407411670684814,
      "epoch": 4.965238389397428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028147108107229324,
      "loss": 4.842,
      "mean_token_accuracy": 0.223918317258358,
      "num_tokens": 98694581.0,
      "step": 44210
    },
    {
      "entropy": 5.518760681152344,
      "epoch": 4.965799966305386,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028143094885857455,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22144197225570678,
      "num_tokens": 98705002.0,
      "step": 44215
    },
    {
      "entropy": 5.499848413467407,
      "epoch": 4.966361543213343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000281390816440092,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22700729668140412,
      "num_tokens": 98715131.0,
      "step": 44220
    },
    {
      "entropy": 5.406033611297607,
      "epoch": 4.966923120121301,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002813506838181233,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22894641011953354,
      "num_tokens": 98724992.0,
      "step": 44225
    },
    {
      "entropy": 5.422825288772583,
      "epoch": 4.967484697029258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002813105509939464,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2249909445643425,
      "num_tokens": 98736651.0,
      "step": 44230
    },
    {
      "entropy": 5.503113174438477,
      "epoch": 4.968046273937215,
      "grad_norm": 1.046875,
      "learning_rate": 0.000281270417968839,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22891122996807098,
      "num_tokens": 98747365.0,
      "step": 44235
    },
    {
      "entropy": 5.5109223365783695,
      "epoch": 4.968607850845173,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002812302847440791,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21391773223876953,
      "num_tokens": 98759610.0,
      "step": 44240
    },
    {
      "entropy": 5.509530067443848,
      "epoch": 4.969169427753131,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028119015132094455,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22565089017152787,
      "num_tokens": 98770166.0,
      "step": 44245
    },
    {
      "entropy": 5.500471735000611,
      "epoch": 4.969731004661089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002811500177007132,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2239633321762085,
      "num_tokens": 98782416.0,
      "step": 44250
    },
    {
      "entropy": 5.551662731170654,
      "epoch": 4.970292581569046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028110988388466287,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2243271917104721,
      "num_tokens": 98793341.0,
      "step": 44255
    },
    {
      "entropy": 5.410412836074829,
      "epoch": 4.970854158477003,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002810697498740715,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23068729490041734,
      "num_tokens": 98803642.0,
      "step": 44260
    },
    {
      "entropy": 5.44674859046936,
      "epoch": 4.971415735384961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002810296156702169,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22624965161085128,
      "num_tokens": 98814650.0,
      "step": 44265
    },
    {
      "entropy": 5.495485734939575,
      "epoch": 4.9719773122929185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028098948127437695,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2241385668516159,
      "num_tokens": 98825928.0,
      "step": 44270
    },
    {
      "entropy": 5.49595308303833,
      "epoch": 4.972538889200876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002809493466878297,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22707158029079438,
      "num_tokens": 98837447.0,
      "step": 44275
    },
    {
      "entropy": 5.4692432403564455,
      "epoch": 4.973100466108834,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028090921191185295,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22881969362497329,
      "num_tokens": 98847720.0,
      "step": 44280
    },
    {
      "entropy": 5.421547079086304,
      "epoch": 4.973662043016791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002808690769477245,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2301239088177681,
      "num_tokens": 98859315.0,
      "step": 44285
    },
    {
      "entropy": 5.417299461364746,
      "epoch": 4.974223619924748,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002808289417967225,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22366369962692262,
      "num_tokens": 98869921.0,
      "step": 44290
    },
    {
      "entropy": 5.454826021194458,
      "epoch": 4.974785196832706,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028078880646012467,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23173736780881882,
      "num_tokens": 98881473.0,
      "step": 44295
    },
    {
      "entropy": 5.485437202453613,
      "epoch": 4.975346773740664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028074867093920897,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22900353372097015,
      "num_tokens": 98892601.0,
      "step": 44300
    },
    {
      "entropy": 5.563317823410034,
      "epoch": 4.975908350648622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028070853523525337,
      "loss": 4.9926,
      "mean_token_accuracy": 0.22579190731048585,
      "num_tokens": 98903823.0,
      "step": 44305
    },
    {
      "entropy": 5.4655718326568605,
      "epoch": 4.976469927556579,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028066839934953573,
      "loss": 4.959,
      "mean_token_accuracy": 0.22352727502584457,
      "num_tokens": 98914266.0,
      "step": 44310
    },
    {
      "entropy": 5.453049993515014,
      "epoch": 4.977031504464536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028062826328333415,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2262068435549736,
      "num_tokens": 98924693.0,
      "step": 44315
    },
    {
      "entropy": 5.414642572402954,
      "epoch": 4.977593081372494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002805881270379265,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22405864000320436,
      "num_tokens": 98936081.0,
      "step": 44320
    },
    {
      "entropy": 5.46228175163269,
      "epoch": 4.9781546582804515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002805479906145906,
      "loss": 4.91,
      "mean_token_accuracy": 0.2274413600564003,
      "num_tokens": 98946495.0,
      "step": 44325
    },
    {
      "entropy": 5.495104074478149,
      "epoch": 4.978716235188409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002805078540146046,
      "loss": 4.898,
      "mean_token_accuracy": 0.22490564733743668,
      "num_tokens": 98956873.0,
      "step": 44330
    },
    {
      "entropy": 5.447467994689942,
      "epoch": 4.979277812096367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002804677172392464,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22581837922334672,
      "num_tokens": 98968349.0,
      "step": 44335
    },
    {
      "entropy": 5.451490354537964,
      "epoch": 4.979839389004324,
      "grad_norm": 1.0,
      "learning_rate": 0.000280427580289794,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23503319174051285,
      "num_tokens": 98978558.0,
      "step": 44340
    },
    {
      "entropy": 5.566263151168823,
      "epoch": 4.980400965912281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028038744316752524,
      "loss": 5.0586,
      "mean_token_accuracy": 0.22287654429674147,
      "num_tokens": 98990435.0,
      "step": 44345
    },
    {
      "entropy": 5.534523820877075,
      "epoch": 4.980962542820239,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028034730587371824,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22296777367591858,
      "num_tokens": 99001516.0,
      "step": 44350
    },
    {
      "entropy": 5.521559238433838,
      "epoch": 4.9815241197281965,
      "grad_norm": 1.125,
      "learning_rate": 0.00028030716840965096,
      "loss": 5.003,
      "mean_token_accuracy": 0.22257322520017625,
      "num_tokens": 99012810.0,
      "step": 44355
    },
    {
      "entropy": 5.526064729690551,
      "epoch": 4.982085696636155,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028026703077660143,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21559089571237564,
      "num_tokens": 99023342.0,
      "step": 44360
    },
    {
      "entropy": 5.39044451713562,
      "epoch": 4.982647273544112,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002802268929758475,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2306942641735077,
      "num_tokens": 99034047.0,
      "step": 44365
    },
    {
      "entropy": 5.525187730789185,
      "epoch": 4.983208850452069,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002801867550086674,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22426154911518098,
      "num_tokens": 99045333.0,
      "step": 44370
    },
    {
      "entropy": 5.539105606079102,
      "epoch": 4.983770427360027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002801466168763389,
      "loss": 4.9901,
      "mean_token_accuracy": 0.22736834585666657,
      "num_tokens": 99057582.0,
      "step": 44375
    },
    {
      "entropy": 5.505902814865112,
      "epoch": 4.984332004267984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002801064785801401,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21981275230646133,
      "num_tokens": 99067940.0,
      "step": 44380
    },
    {
      "entropy": 5.501267433166504,
      "epoch": 4.9848935811759425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028006634012134915,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21996189951896666,
      "num_tokens": 99079408.0,
      "step": 44385
    },
    {
      "entropy": 5.445398902893066,
      "epoch": 4.9854551580839,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028002620150124386,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23353392630815506,
      "num_tokens": 99090320.0,
      "step": 44390
    },
    {
      "entropy": 5.514816188812256,
      "epoch": 4.986016734991857,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00027998606272110237,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22112341225147247,
      "num_tokens": 99101503.0,
      "step": 44395
    },
    {
      "entropy": 5.584442520141602,
      "epoch": 4.986578311899815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002799459237822028,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2176099792122841,
      "num_tokens": 99112768.0,
      "step": 44400
    },
    {
      "entropy": 5.4655508518219,
      "epoch": 4.987139888807772,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027990578468582304,
      "loss": 4.9111,
      "mean_token_accuracy": 0.23119820058345794,
      "num_tokens": 99123473.0,
      "step": 44405
    },
    {
      "entropy": 5.523442792892456,
      "epoch": 4.9877014657157295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002798656454332412,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2250715211033821,
      "num_tokens": 99136005.0,
      "step": 44410
    },
    {
      "entropy": 5.561041355133057,
      "epoch": 4.988263042623688,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027982550602573524,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2289936736226082,
      "num_tokens": 99147494.0,
      "step": 44415
    },
    {
      "entropy": 5.568865203857422,
      "epoch": 4.988824619531645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002797853664645834,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22099230736494063,
      "num_tokens": 99159255.0,
      "step": 44420
    },
    {
      "entropy": 5.521811056137085,
      "epoch": 4.989386196439602,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002797452267510637,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22710692137479782,
      "num_tokens": 99169406.0,
      "step": 44425
    },
    {
      "entropy": 5.467849922180176,
      "epoch": 4.98994777334756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000279705086886454,
      "loss": 4.9203,
      "mean_token_accuracy": 0.229062257707119,
      "num_tokens": 99180392.0,
      "step": 44430
    },
    {
      "entropy": 5.441602659225464,
      "epoch": 4.990509350255517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002796649468720326,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2295740246772766,
      "num_tokens": 99190387.0,
      "step": 44435
    },
    {
      "entropy": 5.467052555084228,
      "epoch": 4.9910709271634754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027962480670907743,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22811322063207626,
      "num_tokens": 99202070.0,
      "step": 44440
    },
    {
      "entropy": 5.539659070968628,
      "epoch": 4.991632504071433,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027958466639886663,
      "loss": 4.9957,
      "mean_token_accuracy": 0.2218868464231491,
      "num_tokens": 99213323.0,
      "step": 44445
    },
    {
      "entropy": 5.496681118011475,
      "epoch": 4.99219408097939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027954452594267825,
      "loss": 4.92,
      "mean_token_accuracy": 0.2263687625527382,
      "num_tokens": 99224178.0,
      "step": 44450
    },
    {
      "entropy": 5.589364242553711,
      "epoch": 4.992755657887348,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027950438534179046,
      "loss": 5.024,
      "mean_token_accuracy": 0.21928937286138533,
      "num_tokens": 99234538.0,
      "step": 44455
    },
    {
      "entropy": 5.5392852306365965,
      "epoch": 4.993317234795305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027946424459748126,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22076205015182496,
      "num_tokens": 99245737.0,
      "step": 44460
    },
    {
      "entropy": 5.47588701248169,
      "epoch": 4.993878811703262,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027942410371102877,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22394698858261108,
      "num_tokens": 99257047.0,
      "step": 44465
    },
    {
      "entropy": 5.538810348510742,
      "epoch": 4.9944403886112205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002793839626837111,
      "loss": 5.0492,
      "mean_token_accuracy": 0.220108500123024,
      "num_tokens": 99267664.0,
      "step": 44470
    },
    {
      "entropy": 5.4640552520751955,
      "epoch": 4.995001965519178,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027934382151680644,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22177913039922714,
      "num_tokens": 99279270.0,
      "step": 44475
    },
    {
      "entropy": 5.4871570587158205,
      "epoch": 4.995563542427135,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027930368021159273,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22807272523641586,
      "num_tokens": 99289724.0,
      "step": 44480
    },
    {
      "entropy": 5.6028053760528564,
      "epoch": 4.996125119335093,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002792635387693481,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22815712839365004,
      "num_tokens": 99300326.0,
      "step": 44485
    },
    {
      "entropy": 5.520056915283203,
      "epoch": 4.99668669624305,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027922339719135093,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22832891792058946,
      "num_tokens": 99312073.0,
      "step": 44490
    },
    {
      "entropy": 5.536628150939942,
      "epoch": 4.997248273151008,
      "grad_norm": 1.03125,
      "learning_rate": 0.000279183255478879,
      "loss": 5.0521,
      "mean_token_accuracy": 0.22065576314926147,
      "num_tokens": 99323938.0,
      "step": 44495
    },
    {
      "entropy": 5.4916524410247805,
      "epoch": 4.997809850058966,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027914311363321073,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22896270006895064,
      "num_tokens": 99334773.0,
      "step": 44500
    },
    {
      "entropy": 5.492211580276489,
      "epoch": 4.998371426966923,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027910297165562406,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23792830407619475,
      "num_tokens": 99345688.0,
      "step": 44505
    },
    {
      "entropy": 5.505130052566528,
      "epoch": 4.998933003874881,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027906282954739714,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2247784659266472,
      "num_tokens": 99356307.0,
      "step": 44510
    },
    {
      "entropy": 5.418028497695923,
      "epoch": 4.999494580782838,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027902268730980816,
      "loss": 4.849,
      "mean_token_accuracy": 0.23138141334056855,
      "num_tokens": 99367348.0,
      "step": 44515
    },
    {
      "entropy": 5.531285815768772,
      "epoch": 5.0,
      "grad_norm": 5.625,
      "learning_rate": 0.0002789825449441353,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21747081809573704,
      "num_tokens": 99377525.0,
      "step": 44520
    },
    {
      "entropy": 5.477304935455322,
      "epoch": 5.000561576907957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002789424024516566,
      "loss": 4.8214,
      "mean_token_accuracy": 0.2316776230931282,
      "num_tokens": 99388342.0,
      "step": 44525
    },
    {
      "entropy": 5.520528745651245,
      "epoch": 5.001123153815915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027890225983365034,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22725383639335633,
      "num_tokens": 99400387.0,
      "step": 44530
    },
    {
      "entropy": 5.374107408523559,
      "epoch": 5.0016847307238725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002788621170913946,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23304048627614976,
      "num_tokens": 99411232.0,
      "step": 44535
    },
    {
      "entropy": 5.4188963890075685,
      "epoch": 5.00224630763183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027882197422616746,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22984031289815904,
      "num_tokens": 99422839.0,
      "step": 44540
    },
    {
      "entropy": 5.491669416427612,
      "epoch": 5.002807884539788,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002787818312392473,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22840026021003723,
      "num_tokens": 99434835.0,
      "step": 44545
    },
    {
      "entropy": 5.457922315597534,
      "epoch": 5.003369461447745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027874168813191207,
      "loss": 4.886,
      "mean_token_accuracy": 0.2291823610663414,
      "num_tokens": 99445705.0,
      "step": 44550
    },
    {
      "entropy": 5.451347160339355,
      "epoch": 5.003931038355703,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002787015449054401,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23324509114027023,
      "num_tokens": 99457127.0,
      "step": 44555
    },
    {
      "entropy": 5.408340263366699,
      "epoch": 5.00449261526366,
      "grad_norm": 1.125,
      "learning_rate": 0.0002786614015611095,
      "loss": 4.824,
      "mean_token_accuracy": 0.23245389759540558,
      "num_tokens": 99468653.0,
      "step": 44560
    },
    {
      "entropy": 5.603671741485596,
      "epoch": 5.005054192171618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027862125810019844,
      "loss": 5.0722,
      "mean_token_accuracy": 0.21252766251564026,
      "num_tokens": 99481238.0,
      "step": 44565
    },
    {
      "entropy": 5.486661100387574,
      "epoch": 5.005615769079576,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027858111452398515,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23610934764146804,
      "num_tokens": 99493025.0,
      "step": 44570
    },
    {
      "entropy": 5.448735475540161,
      "epoch": 5.006177345987533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027854097083374773,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23531297147274016,
      "num_tokens": 99504675.0,
      "step": 44575
    },
    {
      "entropy": 5.513508129119873,
      "epoch": 5.00673892289549,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002785008270307645,
      "loss": 4.8886,
      "mean_token_accuracy": 0.23305243402719497,
      "num_tokens": 99516005.0,
      "step": 44580
    },
    {
      "entropy": 5.566770410537719,
      "epoch": 5.007300499803448,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002784606831163136,
      "loss": 5.0312,
      "mean_token_accuracy": 0.21470609605312346,
      "num_tokens": 99526917.0,
      "step": 44585
    },
    {
      "entropy": 5.4467216491699215,
      "epoch": 5.0078620767114055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027842053909167314,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23020866364240647,
      "num_tokens": 99538799.0,
      "step": 44590
    },
    {
      "entropy": 5.485759210586548,
      "epoch": 5.008423653619364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027838039495812133,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2368365079164505,
      "num_tokens": 99550431.0,
      "step": 44595
    },
    {
      "entropy": 5.437341117858887,
      "epoch": 5.008985230527321,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00027834025071693655,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23270939886569977,
      "num_tokens": 99561562.0,
      "step": 44600
    },
    {
      "entropy": 5.424635744094848,
      "epoch": 5.009546807435278,
      "grad_norm": 1.0,
      "learning_rate": 0.00027830010636939684,
      "loss": 4.814,
      "mean_token_accuracy": 0.23347311317920685,
      "num_tokens": 99572101.0,
      "step": 44605
    },
    {
      "entropy": 5.440806341171265,
      "epoch": 5.010108384343236,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027825996191678043,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22618837356567384,
      "num_tokens": 99583186.0,
      "step": 44610
    },
    {
      "entropy": 5.458729267120361,
      "epoch": 5.010669961251193,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027821981736036555,
      "loss": 4.8689,
      "mean_token_accuracy": 0.23714529871940612,
      "num_tokens": 99593716.0,
      "step": 44615
    },
    {
      "entropy": 5.554162979125977,
      "epoch": 5.011231538159151,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002781796727014305,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22327640950679778,
      "num_tokens": 99605798.0,
      "step": 44620
    },
    {
      "entropy": 5.475872564315796,
      "epoch": 5.011793115067109,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002781395279412535,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22816588431596757,
      "num_tokens": 99617938.0,
      "step": 44625
    },
    {
      "entropy": 5.488706922531128,
      "epoch": 5.012354691975066,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027809938308111254,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2318232238292694,
      "num_tokens": 99628761.0,
      "step": 44630
    },
    {
      "entropy": 5.483943748474121,
      "epoch": 5.012916268883023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027805923812228615,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22466466277837754,
      "num_tokens": 99640072.0,
      "step": 44635
    },
    {
      "entropy": 5.483944320678711,
      "epoch": 5.013477845790981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027801909306605243,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22774286866188048,
      "num_tokens": 99651280.0,
      "step": 44640
    },
    {
      "entropy": 5.444033861160278,
      "epoch": 5.014039422698938,
      "grad_norm": 1.125,
      "learning_rate": 0.0002779789479136896,
      "loss": 4.8834,
      "mean_token_accuracy": 0.23166434317827225,
      "num_tokens": 99661574.0,
      "step": 44645
    },
    {
      "entropy": 5.475733804702759,
      "epoch": 5.0146009996068965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002779388026664759,
      "loss": 4.8432,
      "mean_token_accuracy": 0.23567759543657302,
      "num_tokens": 99671092.0,
      "step": 44650
    },
    {
      "entropy": 5.6396770000457765,
      "epoch": 5.015162576514854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002778986573256896,
      "loss": 5.013,
      "mean_token_accuracy": 0.22555311918258666,
      "num_tokens": 99682104.0,
      "step": 44655
    },
    {
      "entropy": 5.572243928909302,
      "epoch": 5.015724153422811,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002778585118926089,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22298432737588883,
      "num_tokens": 99692152.0,
      "step": 44660
    },
    {
      "entropy": 5.445234060287476,
      "epoch": 5.016285730330769,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027781836636851207,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2313980981707573,
      "num_tokens": 99703807.0,
      "step": 44665
    },
    {
      "entropy": 5.543313646316529,
      "epoch": 5.016847307238726,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027777822075467735,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2282588467001915,
      "num_tokens": 99714039.0,
      "step": 44670
    },
    {
      "entropy": 5.427083349227905,
      "epoch": 5.0174088841466835,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000277738075052383,
      "loss": 4.864,
      "mean_token_accuracy": 0.23167277723550797,
      "num_tokens": 99725215.0,
      "step": 44675
    },
    {
      "entropy": 5.487498712539673,
      "epoch": 5.017970461054642,
      "grad_norm": 1.0,
      "learning_rate": 0.0002776979292629074,
      "loss": 4.905,
      "mean_token_accuracy": 0.23197535276412964,
      "num_tokens": 99736281.0,
      "step": 44680
    },
    {
      "entropy": 5.4977792263031,
      "epoch": 5.018532037962599,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002776577833875285,
      "loss": 4.8926,
      "mean_token_accuracy": 0.23179379403591155,
      "num_tokens": 99747544.0,
      "step": 44685
    },
    {
      "entropy": 5.426753282546997,
      "epoch": 5.019093614870557,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027761763742752485,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23115358054637908,
      "num_tokens": 99758806.0,
      "step": 44690
    },
    {
      "entropy": 5.497817182540894,
      "epoch": 5.019655191778514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002775774913841746,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22942211329936982,
      "num_tokens": 99769477.0,
      "step": 44695
    },
    {
      "entropy": 5.448174953460693,
      "epoch": 5.020216768686471,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000277537345258756,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23229531347751617,
      "num_tokens": 99780283.0,
      "step": 44700
    },
    {
      "entropy": 5.5535157203674315,
      "epoch": 5.0207783455944295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027749719905254737,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2293313905596733,
      "num_tokens": 99790119.0,
      "step": 44705
    },
    {
      "entropy": 5.532082843780517,
      "epoch": 5.021339922502387,
      "grad_norm": 1.015625,
      "learning_rate": 0.000277457052766827,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21846643537282945,
      "num_tokens": 99801619.0,
      "step": 44710
    },
    {
      "entropy": 5.466307926177978,
      "epoch": 5.021901499410344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027741690640287306,
      "loss": 4.849,
      "mean_token_accuracy": 0.23532640635967256,
      "num_tokens": 99812972.0,
      "step": 44715
    },
    {
      "entropy": 5.466490316390991,
      "epoch": 5.022463076318302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002773767599619639,
      "loss": 4.8338,
      "mean_token_accuracy": 0.23186297714710236,
      "num_tokens": 99823667.0,
      "step": 44720
    },
    {
      "entropy": 5.388620758056641,
      "epoch": 5.023024653226259,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027733661344537776,
      "loss": 4.7512,
      "mean_token_accuracy": 0.24410629719495774,
      "num_tokens": 99834654.0,
      "step": 44725
    },
    {
      "entropy": 5.48935832977295,
      "epoch": 5.0235862301342165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000277296466854393,
      "loss": 4.9748,
      "mean_token_accuracy": 0.23433843553066253,
      "num_tokens": 99846393.0,
      "step": 44730
    },
    {
      "entropy": 5.419358777999878,
      "epoch": 5.024147807042175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002772563201902877,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23970647603273393,
      "num_tokens": 99857206.0,
      "step": 44735
    },
    {
      "entropy": 5.612226819992065,
      "epoch": 5.024709383950132,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027721617345434035,
      "loss": 5.0463,
      "mean_token_accuracy": 0.22763677835464477,
      "num_tokens": 99871497.0,
      "step": 44740
    },
    {
      "entropy": 5.468939352035522,
      "epoch": 5.02527096085809,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027717602664782925,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22174068987369538,
      "num_tokens": 99882337.0,
      "step": 44745
    },
    {
      "entropy": 5.418942785263061,
      "epoch": 5.025832537766047,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002771358797720326,
      "loss": 4.812,
      "mean_token_accuracy": 0.2380195677280426,
      "num_tokens": 99892863.0,
      "step": 44750
    },
    {
      "entropy": 5.4293317794799805,
      "epoch": 5.026394114674004,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027709573282822864,
      "loss": 4.803,
      "mean_token_accuracy": 0.2367755353450775,
      "num_tokens": 99903566.0,
      "step": 44755
    },
    {
      "entropy": 5.517363119125366,
      "epoch": 5.026955691581962,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002770555858176958,
      "loss": 4.8547,
      "mean_token_accuracy": 0.23724362403154373,
      "num_tokens": 99913466.0,
      "step": 44760
    },
    {
      "entropy": 5.4832642555236815,
      "epoch": 5.02751726848992,
      "grad_norm": 1.125,
      "learning_rate": 0.0002770154387417123,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22862743437290192,
      "num_tokens": 99923579.0,
      "step": 44765
    },
    {
      "entropy": 5.544750642776489,
      "epoch": 5.028078845397877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002769752916015565,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22032272070646286,
      "num_tokens": 99935471.0,
      "step": 44770
    },
    {
      "entropy": 5.569924259185791,
      "epoch": 5.028640422305835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027693514439850646,
      "loss": 4.9633,
      "mean_token_accuracy": 0.23117049038410187,
      "num_tokens": 99947082.0,
      "step": 44775
    },
    {
      "entropy": 5.54443392753601,
      "epoch": 5.029201999213792,
      "grad_norm": 1.125,
      "learning_rate": 0.00027689499713384085,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22578049898147584,
      "num_tokens": 99958186.0,
      "step": 44780
    },
    {
      "entropy": 5.544092082977295,
      "epoch": 5.02976357612175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002768548498088378,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2256103053689003,
      "num_tokens": 99970928.0,
      "step": 44785
    },
    {
      "entropy": 5.388844633102417,
      "epoch": 5.0303251530297075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002768147024247756,
      "loss": 4.8485,
      "mean_token_accuracy": 0.23256482183933258,
      "num_tokens": 99982549.0,
      "step": 44790
    },
    {
      "entropy": 5.378140878677368,
      "epoch": 5.030886729937665,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002767745549829325,
      "loss": 4.7709,
      "mean_token_accuracy": 0.23346623182296752,
      "num_tokens": 99993835.0,
      "step": 44795
    },
    {
      "entropy": 5.523657894134521,
      "epoch": 5.031448306845623,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027673440748458687,
      "loss": 4.916,
      "mean_token_accuracy": 0.22583357244729996,
      "num_tokens": 100004336.0,
      "step": 44800
    },
    {
      "entropy": 5.480328702926636,
      "epoch": 5.03200988375358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027669425993101716,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23256247192621232,
      "num_tokens": 100015334.0,
      "step": 44805
    },
    {
      "entropy": 5.517094612121582,
      "epoch": 5.032571460661537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002766541123235014,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22413976043462752,
      "num_tokens": 100026665.0,
      "step": 44810
    },
    {
      "entropy": 5.421613264083862,
      "epoch": 5.033133037569495,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002766139646633182,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2355615496635437,
      "num_tokens": 100037263.0,
      "step": 44815
    },
    {
      "entropy": 5.444287109375,
      "epoch": 5.033694614477453,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027657381695174563,
      "loss": 4.826,
      "mean_token_accuracy": 0.22665845602750778,
      "num_tokens": 100048482.0,
      "step": 44820
    },
    {
      "entropy": 5.448045635223389,
      "epoch": 5.03425619138541,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027653366919006217,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22785186022520065,
      "num_tokens": 100059351.0,
      "step": 44825
    },
    {
      "entropy": 5.440320920944214,
      "epoch": 5.034817768293368,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002764935213795461,
      "loss": 4.7507,
      "mean_token_accuracy": 0.24024568796157836,
      "num_tokens": 100069924.0,
      "step": 44830
    },
    {
      "entropy": 5.4145032405853275,
      "epoch": 5.035379345201325,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002764533735214757,
      "loss": 4.797,
      "mean_token_accuracy": 0.2411456361413002,
      "num_tokens": 100080733.0,
      "step": 44835
    },
    {
      "entropy": 5.446532583236694,
      "epoch": 5.035940922109283,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027641322561712935,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22791907787322999,
      "num_tokens": 100091834.0,
      "step": 44840
    },
    {
      "entropy": 5.5948686599731445,
      "epoch": 5.0365024990172405,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00027637307766778524,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21982720196247102,
      "num_tokens": 100103872.0,
      "step": 44845
    },
    {
      "entropy": 5.576116466522217,
      "epoch": 5.037064075925198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000276332929674722,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22797320932149887,
      "num_tokens": 100114391.0,
      "step": 44850
    },
    {
      "entropy": 5.549773120880127,
      "epoch": 5.037625652833156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027629278163921763,
      "loss": 4.865,
      "mean_token_accuracy": 0.23460142463445663,
      "num_tokens": 100125000.0,
      "step": 44855
    },
    {
      "entropy": 5.524862432479859,
      "epoch": 5.038187229741113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027625263356255067,
      "loss": 4.8951,
      "mean_token_accuracy": 0.23412115275859832,
      "num_tokens": 100135394.0,
      "step": 44860
    },
    {
      "entropy": 5.3969159603118895,
      "epoch": 5.03874880664907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002762124854459992,
      "loss": 4.7989,
      "mean_token_accuracy": 0.24094276130199432,
      "num_tokens": 100146610.0,
      "step": 44865
    },
    {
      "entropy": 5.434390115737915,
      "epoch": 5.039310383557028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027617233729084194,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23276289105415343,
      "num_tokens": 100156936.0,
      "step": 44870
    },
    {
      "entropy": 5.440294408798218,
      "epoch": 5.0398719604649855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027613218909835693,
      "loss": 4.866,
      "mean_token_accuracy": 0.23841193169355393,
      "num_tokens": 100168406.0,
      "step": 44875
    },
    {
      "entropy": 5.413275527954101,
      "epoch": 5.040433537372944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002760920408698225,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22684036791324616,
      "num_tokens": 100179317.0,
      "step": 44880
    },
    {
      "entropy": 5.405751705169678,
      "epoch": 5.040995114280901,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027605189260651716,
      "loss": 4.7772,
      "mean_token_accuracy": 0.2375454217195511,
      "num_tokens": 100190916.0,
      "step": 44885
    },
    {
      "entropy": 5.412471771240234,
      "epoch": 5.041556691188858,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002760117443097191,
      "loss": 4.7708,
      "mean_token_accuracy": 0.24563691020011902,
      "num_tokens": 100201437.0,
      "step": 44890
    },
    {
      "entropy": 5.427309036254883,
      "epoch": 5.042118268096816,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027597159598070675,
      "loss": 4.8143,
      "mean_token_accuracy": 0.23853641152381896,
      "num_tokens": 100212405.0,
      "step": 44895
    },
    {
      "entropy": 5.415934276580811,
      "epoch": 5.042679845004773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027593144762075837,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23565881997346877,
      "num_tokens": 100223428.0,
      "step": 44900
    },
    {
      "entropy": 5.355186462402344,
      "epoch": 5.043241421912731,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027589129923115234,
      "loss": 4.7656,
      "mean_token_accuracy": 0.23753758817911147,
      "num_tokens": 100234128.0,
      "step": 44905
    },
    {
      "entropy": 5.483733892440796,
      "epoch": 5.043802998820689,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000275851150813167,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22383883595466614,
      "num_tokens": 100244920.0,
      "step": 44910
    },
    {
      "entropy": 5.443733739852905,
      "epoch": 5.044364575728646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002758110023680807,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2314569517970085,
      "num_tokens": 100256518.0,
      "step": 44915
    },
    {
      "entropy": 5.540986251831055,
      "epoch": 5.044926152636603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002757708538971717,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2288044199347496,
      "num_tokens": 100268203.0,
      "step": 44920
    },
    {
      "entropy": 5.454263210296631,
      "epoch": 5.045487729544561,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002757307054017185,
      "loss": 4.7839,
      "mean_token_accuracy": 0.2329866662621498,
      "num_tokens": 100279440.0,
      "step": 44925
    },
    {
      "entropy": 5.502652502059936,
      "epoch": 5.0460493064525185,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002756905568829994,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2319154903292656,
      "num_tokens": 100290616.0,
      "step": 44930
    },
    {
      "entropy": 5.477941274642944,
      "epoch": 5.046610883360477,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027565040834229266,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22626966536045073,
      "num_tokens": 100301602.0,
      "step": 44935
    },
    {
      "entropy": 5.428617191314697,
      "epoch": 5.047172460268434,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027561025978087663,
      "loss": 4.8365,
      "mean_token_accuracy": 0.23294224441051484,
      "num_tokens": 100313311.0,
      "step": 44940
    },
    {
      "entropy": 5.5026857376098635,
      "epoch": 5.047734037176391,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027557011120002975,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22296959310770034,
      "num_tokens": 100325394.0,
      "step": 44945
    },
    {
      "entropy": 5.5178553581237795,
      "epoch": 5.048295614084349,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002755299626010302,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2301964282989502,
      "num_tokens": 100336643.0,
      "step": 44950
    },
    {
      "entropy": 5.507905340194702,
      "epoch": 5.048857190992306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002754898139851565,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2288292109966278,
      "num_tokens": 100348409.0,
      "step": 44955
    },
    {
      "entropy": 5.4947350025177,
      "epoch": 5.049418767900264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027544966535368695,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2368205890059471,
      "num_tokens": 100360115.0,
      "step": 44960
    },
    {
      "entropy": 5.432924079895019,
      "epoch": 5.049980344808222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002754095167078999,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23017006665468215,
      "num_tokens": 100369618.0,
      "step": 44965
    },
    {
      "entropy": 5.414655017852783,
      "epoch": 5.050541921716179,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002753693680490736,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23445862233638765,
      "num_tokens": 100379470.0,
      "step": 44970
    },
    {
      "entropy": 5.397604465484619,
      "epoch": 5.051103498624137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002753292193784865,
      "loss": 4.8588,
      "mean_token_accuracy": 0.2318849727511406,
      "num_tokens": 100390490.0,
      "step": 44975
    },
    {
      "entropy": 5.433482646942139,
      "epoch": 5.051665075532094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027528907069741696,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23750723898410797,
      "num_tokens": 100402817.0,
      "step": 44980
    },
    {
      "entropy": 5.567031621932983,
      "epoch": 5.052226652440051,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002752489220071433,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22237911820411682,
      "num_tokens": 100413411.0,
      "step": 44985
    },
    {
      "entropy": 5.427756214141846,
      "epoch": 5.0527882293480095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002752087733089438,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23630745708942413,
      "num_tokens": 100425118.0,
      "step": 44990
    },
    {
      "entropy": 5.560967445373535,
      "epoch": 5.053349806255967,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002751686246040969,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2231506198644638,
      "num_tokens": 100436544.0,
      "step": 44995
    },
    {
      "entropy": 5.533036947250366,
      "epoch": 5.053911383163924,
      "grad_norm": 1.0625,
      "learning_rate": 0.000275128475893881,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22702505588531494,
      "num_tokens": 100448315.0,
      "step": 45000
    },
    {
      "epoch": 5.053911383163924,
      "eval_entropy": 5.256718953737501,
      "eval_loss": 5.107872009277344,
      "eval_mean_token_accuracy": 0.22466659463455851,
      "eval_num_tokens": 100448315.0,
      "eval_runtime": 52.994,
      "eval_samples_per_second": 585.217,
      "eval_steps_per_second": 73.159,
      "step": 45000
    },
    {
      "entropy": 5.4173136234283445,
      "epoch": 5.054472960071882,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027508832717957433,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2340691566467285,
      "num_tokens": 100459548.0,
      "step": 45005
    },
    {
      "entropy": 5.525400114059448,
      "epoch": 5.055034536979839,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002750481784624553,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22838965356349944,
      "num_tokens": 100468687.0,
      "step": 45010
    },
    {
      "entropy": 5.4471265316009525,
      "epoch": 5.0555961138877965,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027500802974380224,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22989449203014373,
      "num_tokens": 100479607.0,
      "step": 45015
    },
    {
      "entropy": 5.473666477203369,
      "epoch": 5.056157690795755,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002749678810248935,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2292293131351471,
      "num_tokens": 100492245.0,
      "step": 45020
    },
    {
      "entropy": 5.463179111480713,
      "epoch": 5.056719267703712,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002749277323070075,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23225232511758803,
      "num_tokens": 100503973.0,
      "step": 45025
    },
    {
      "entropy": 5.492497539520263,
      "epoch": 5.05728084461167,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027488758359142245,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23472894132137298,
      "num_tokens": 100516003.0,
      "step": 45030
    },
    {
      "entropy": 5.481388807296753,
      "epoch": 5.057842421519627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002748474348794167,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22866024374961852,
      "num_tokens": 100528130.0,
      "step": 45035
    },
    {
      "entropy": 5.449507188796997,
      "epoch": 5.058403998427584,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027480728617226884,
      "loss": 4.8861,
      "mean_token_accuracy": 0.23283398598432542,
      "num_tokens": 100539532.0,
      "step": 45040
    },
    {
      "entropy": 5.521431255340576,
      "epoch": 5.0589655753355425,
      "grad_norm": 1.078125,
      "learning_rate": 0.000274767137471257,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22547551095485688,
      "num_tokens": 100550520.0,
      "step": 45045
    },
    {
      "entropy": 5.4244121551513675,
      "epoch": 5.0595271522435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027472698877765967,
      "loss": 4.7732,
      "mean_token_accuracy": 0.23832533359527588,
      "num_tokens": 100561596.0,
      "step": 45050
    },
    {
      "entropy": 5.463217878341675,
      "epoch": 5.060088729151457,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002746868400927551,
      "loss": 4.8697,
      "mean_token_accuracy": 0.232737198472023,
      "num_tokens": 100571238.0,
      "step": 45055
    },
    {
      "entropy": 5.511422872543335,
      "epoch": 5.060650306059415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002746466914178216,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22705770581960677,
      "num_tokens": 100582687.0,
      "step": 45060
    },
    {
      "entropy": 5.492000722885132,
      "epoch": 5.061211882967372,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002746065427541376,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22587278187274934,
      "num_tokens": 100594270.0,
      "step": 45065
    },
    {
      "entropy": 5.476876592636108,
      "epoch": 5.06177345987533,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00027456639410298144,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22768248468637467,
      "num_tokens": 100605449.0,
      "step": 45070
    },
    {
      "entropy": 5.3803283214569095,
      "epoch": 5.062335036783288,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002745262454656314,
      "loss": 4.783,
      "mean_token_accuracy": 0.23578767031431197,
      "num_tokens": 100616985.0,
      "step": 45075
    },
    {
      "entropy": 5.411811780929566,
      "epoch": 5.062896613691245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000274486096843366,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23978390395641327,
      "num_tokens": 100628088.0,
      "step": 45080
    },
    {
      "entropy": 5.56168041229248,
      "epoch": 5.063458190599203,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027444594823746344,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2257420003414154,
      "num_tokens": 100639805.0,
      "step": 45085
    },
    {
      "entropy": 5.650282192230224,
      "epoch": 5.06401976750716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002744057996492021,
      "loss": 5.0719,
      "mean_token_accuracy": 0.21834689676761626,
      "num_tokens": 100650437.0,
      "step": 45090
    },
    {
      "entropy": 5.437141418457031,
      "epoch": 5.064581344415117,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002743656510798603,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23204113095998763,
      "num_tokens": 100661466.0,
      "step": 45095
    },
    {
      "entropy": 5.505884838104248,
      "epoch": 5.065142921323075,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002743255025307165,
      "loss": 4.919,
      "mean_token_accuracy": 0.23068351447582244,
      "num_tokens": 100671706.0,
      "step": 45100
    },
    {
      "entropy": 5.446944952011108,
      "epoch": 5.065704498231033,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027428535400304886,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2304103747010231,
      "num_tokens": 100682592.0,
      "step": 45105
    },
    {
      "entropy": 5.485529184341431,
      "epoch": 5.06626607513899,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002742452054981358,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2235440269112587,
      "num_tokens": 100693893.0,
      "step": 45110
    },
    {
      "entropy": 5.365427827835083,
      "epoch": 5.066827652046948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002742050570172558,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23714247196912766,
      "num_tokens": 100705455.0,
      "step": 45115
    },
    {
      "entropy": 5.439892768859863,
      "epoch": 5.067389228954905,
      "grad_norm": 1.125,
      "learning_rate": 0.000274164908561687,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22938322126865388,
      "num_tokens": 100714945.0,
      "step": 45120
    },
    {
      "entropy": 5.508907508850098,
      "epoch": 5.067950805862863,
      "grad_norm": 1.125,
      "learning_rate": 0.00027412476013270786,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22504537403583527,
      "num_tokens": 100726299.0,
      "step": 45125
    },
    {
      "entropy": 5.52562198638916,
      "epoch": 5.0685123827708205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027408461173159676,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22381579279899597,
      "num_tokens": 100739415.0,
      "step": 45130
    },
    {
      "entropy": 5.553932952880859,
      "epoch": 5.069073959678778,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027404446335963195,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2204652890563011,
      "num_tokens": 100751018.0,
      "step": 45135
    },
    {
      "entropy": 5.432482004165649,
      "epoch": 5.069635536586736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002740043150180918,
      "loss": 4.8288,
      "mean_token_accuracy": 0.23520049303770066,
      "num_tokens": 100761425.0,
      "step": 45140
    },
    {
      "entropy": 5.523691844940186,
      "epoch": 5.070197113494693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002739641667082546,
      "loss": 4.925,
      "mean_token_accuracy": 0.2249401852488518,
      "num_tokens": 100772906.0,
      "step": 45145
    },
    {
      "entropy": 5.52758469581604,
      "epoch": 5.07075869040265,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027392401843139876,
      "loss": 4.9356,
      "mean_token_accuracy": 0.23113394975662233,
      "num_tokens": 100784106.0,
      "step": 45150
    },
    {
      "entropy": 5.525169467926025,
      "epoch": 5.071320267310608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002738838701888026,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21539560109376907,
      "num_tokens": 100795066.0,
      "step": 45155
    },
    {
      "entropy": 5.428374147415161,
      "epoch": 5.071881844218566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002738437219817444,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2290758326649666,
      "num_tokens": 100806936.0,
      "step": 45160
    },
    {
      "entropy": 5.318535661697387,
      "epoch": 5.072443421126524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027380357381150263,
      "loss": 4.7557,
      "mean_token_accuracy": 0.23462794721126556,
      "num_tokens": 100817483.0,
      "step": 45165
    },
    {
      "entropy": 5.572995948791504,
      "epoch": 5.073004998034481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002737634256793555,
      "loss": 5.0079,
      "mean_token_accuracy": 0.22498151808977127,
      "num_tokens": 100828482.0,
      "step": 45170
    },
    {
      "entropy": 5.506433391571045,
      "epoch": 5.073566574942438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027372327758658135,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22867220789194107,
      "num_tokens": 100839745.0,
      "step": 45175
    },
    {
      "entropy": 5.4458802223205565,
      "epoch": 5.074128151850396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027368312953445845,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23431621342897416,
      "num_tokens": 100850441.0,
      "step": 45180
    },
    {
      "entropy": 5.374617958068848,
      "epoch": 5.0746897287583534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002736429815242653,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2319557100534439,
      "num_tokens": 100861428.0,
      "step": 45185
    },
    {
      "entropy": 5.408046293258667,
      "epoch": 5.075251305666311,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002736028335572802,
      "loss": 4.8186,
      "mean_token_accuracy": 0.2315792351961136,
      "num_tokens": 100872874.0,
      "step": 45190
    },
    {
      "entropy": 5.516990232467651,
      "epoch": 5.075812882574269,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027356268563478135,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21665584146976472,
      "num_tokens": 100884150.0,
      "step": 45195
    },
    {
      "entropy": 5.500853872299194,
      "epoch": 5.076374459482226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002735225377580472,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23192399144172668,
      "num_tokens": 100896440.0,
      "step": 45200
    },
    {
      "entropy": 5.464272308349609,
      "epoch": 5.076936036390183,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000273482389928356,
      "loss": 4.8666,
      "mean_token_accuracy": 0.23984038531780244,
      "num_tokens": 100906854.0,
      "step": 45205
    },
    {
      "entropy": 5.455808162689209,
      "epoch": 5.077497613298141,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002734422421469861,
      "loss": 4.8712,
      "mean_token_accuracy": 0.24243172705173494,
      "num_tokens": 100918644.0,
      "step": 45210
    },
    {
      "entropy": 5.488126611709594,
      "epoch": 5.0780591902060985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002734020944152157,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2292349800467491,
      "num_tokens": 100929313.0,
      "step": 45215
    },
    {
      "entropy": 5.524488019943237,
      "epoch": 5.078620767114057,
      "grad_norm": 1.0,
      "learning_rate": 0.0002733619467343234,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2296699270606041,
      "num_tokens": 100940406.0,
      "step": 45220
    },
    {
      "entropy": 5.429388475418091,
      "epoch": 5.079182344022014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002733217991055873,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23226650506258012,
      "num_tokens": 100951086.0,
      "step": 45225
    },
    {
      "entropy": 5.467359209060669,
      "epoch": 5.079743920929971,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002732816515302858,
      "loss": 4.943,
      "mean_token_accuracy": 0.22968518882989883,
      "num_tokens": 100961417.0,
      "step": 45230
    },
    {
      "entropy": 5.477453708648682,
      "epoch": 5.080305497837929,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002732415040096971,
      "loss": 4.9385,
      "mean_token_accuracy": 0.23316119760274887,
      "num_tokens": 100973667.0,
      "step": 45235
    },
    {
      "entropy": 5.543975925445556,
      "epoch": 5.080867074745886,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027320135654509964,
      "loss": 4.951,
      "mean_token_accuracy": 0.23017015755176545,
      "num_tokens": 100984891.0,
      "step": 45240
    },
    {
      "entropy": 5.492526721954346,
      "epoch": 5.081428651653844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002731612091377717,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22610901445150375,
      "num_tokens": 100997273.0,
      "step": 45245
    },
    {
      "entropy": 5.363886117935181,
      "epoch": 5.081990228561802,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027312106178899154,
      "loss": 4.7315,
      "mean_token_accuracy": 0.24743276238441467,
      "num_tokens": 101007341.0,
      "step": 45250
    },
    {
      "entropy": 5.5131653308868405,
      "epoch": 5.082551805469759,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002730809145000374,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2288408488035202,
      "num_tokens": 101017471.0,
      "step": 45255
    },
    {
      "entropy": 5.3962303638458256,
      "epoch": 5.083113382377717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002730407672721878,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23744911700487137,
      "num_tokens": 101028726.0,
      "step": 45260
    },
    {
      "entropy": 5.636175870895386,
      "epoch": 5.083674959285674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000273000620106721,
      "loss": 5.0395,
      "mean_token_accuracy": 0.22020127773284912,
      "num_tokens": 101040824.0,
      "step": 45265
    },
    {
      "entropy": 5.5257645606994625,
      "epoch": 5.0842365361936315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002729604730049151,
      "loss": 4.9431,
      "mean_token_accuracy": 0.2268101915717125,
      "num_tokens": 101051849.0,
      "step": 45270
    },
    {
      "entropy": 5.456682825088501,
      "epoch": 5.08479811310159,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002729203259680486,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2268223538994789,
      "num_tokens": 101062667.0,
      "step": 45275
    },
    {
      "entropy": 5.474559783935547,
      "epoch": 5.085359690009547,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002728801789973998,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22678239196538924,
      "num_tokens": 101073679.0,
      "step": 45280
    },
    {
      "entropy": 5.557666444778443,
      "epoch": 5.085921266917504,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002728400320942468,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2260642796754837,
      "num_tokens": 101085394.0,
      "step": 45285
    },
    {
      "entropy": 5.5498768329620365,
      "epoch": 5.086482843825462,
      "grad_norm": 1.078125,
      "learning_rate": 0.000272799885259868,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22776736319065094,
      "num_tokens": 101096748.0,
      "step": 45290
    },
    {
      "entropy": 5.391570997238159,
      "epoch": 5.087044420733419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002727597384955418,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24810547530651092,
      "num_tokens": 101107704.0,
      "step": 45295
    },
    {
      "entropy": 5.464524507522583,
      "epoch": 5.0876059976413766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027271959180254636,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2311531811952591,
      "num_tokens": 101119890.0,
      "step": 45300
    },
    {
      "entropy": 5.469335460662842,
      "epoch": 5.088167574549335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027267944518215997,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22886326611042024,
      "num_tokens": 101130172.0,
      "step": 45305
    },
    {
      "entropy": 5.439871454238892,
      "epoch": 5.088729151457292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027263929863566104,
      "loss": 4.8812,
      "mean_token_accuracy": 0.23770459592342377,
      "num_tokens": 101141374.0,
      "step": 45310
    },
    {
      "entropy": 5.492646741867065,
      "epoch": 5.08929072836525,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002725991521643277,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22421987801790239,
      "num_tokens": 101155192.0,
      "step": 45315
    },
    {
      "entropy": 5.53441276550293,
      "epoch": 5.089852305273207,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002725590057694383,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22673005014657974,
      "num_tokens": 101168438.0,
      "step": 45320
    },
    {
      "entropy": 5.506564331054688,
      "epoch": 5.090413882181164,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027251885945227093,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22875392884016038,
      "num_tokens": 101180609.0,
      "step": 45325
    },
    {
      "entropy": 5.511861562728882,
      "epoch": 5.0909754590891225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027247871321410425,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23051070421934128,
      "num_tokens": 101191211.0,
      "step": 45330
    },
    {
      "entropy": 5.443926239013672,
      "epoch": 5.09153703599708,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027243856705621623,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22983463257551193,
      "num_tokens": 101202706.0,
      "step": 45335
    },
    {
      "entropy": 5.441290664672851,
      "epoch": 5.092098612905037,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027239842097988526,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22985194474458695,
      "num_tokens": 101213890.0,
      "step": 45340
    },
    {
      "entropy": 5.563672256469727,
      "epoch": 5.092660189812995,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002723582749863896,
      "loss": 5.006,
      "mean_token_accuracy": 0.22286151647567748,
      "num_tokens": 101225946.0,
      "step": 45345
    },
    {
      "entropy": 5.553834438323975,
      "epoch": 5.093221766720952,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027231812907700746,
      "loss": 4.9602,
      "mean_token_accuracy": 0.219678857922554,
      "num_tokens": 101236269.0,
      "step": 45350
    },
    {
      "entropy": 5.516775417327881,
      "epoch": 5.09378334362891,
      "grad_norm": 1.125,
      "learning_rate": 0.0002722779832530171,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23529492765665055,
      "num_tokens": 101247641.0,
      "step": 45355
    },
    {
      "entropy": 5.475974464416504,
      "epoch": 5.094344920536868,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027223783751569687,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22438330352306365,
      "num_tokens": 101259271.0,
      "step": 45360
    },
    {
      "entropy": 5.392143583297729,
      "epoch": 5.094906497444825,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027219769186632493,
      "loss": 4.7738,
      "mean_token_accuracy": 0.23940883129835128,
      "num_tokens": 101269709.0,
      "step": 45365
    },
    {
      "entropy": 5.466506862640381,
      "epoch": 5.095468074352783,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002721575463061796,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21700910478830338,
      "num_tokens": 101280559.0,
      "step": 45370
    },
    {
      "entropy": 5.509766340255737,
      "epoch": 5.09602965126074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027211740083653903,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23226659446954728,
      "num_tokens": 101292356.0,
      "step": 45375
    },
    {
      "entropy": 5.494313287734985,
      "epoch": 5.096591228168697,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002720772554586816,
      "loss": 4.8838,
      "mean_token_accuracy": 0.23069509118795395,
      "num_tokens": 101303881.0,
      "step": 45380
    },
    {
      "entropy": 5.463854837417602,
      "epoch": 5.0971528050766555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002720371101738855,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23687127828598023,
      "num_tokens": 101315310.0,
      "step": 45385
    },
    {
      "entropy": 5.394137620925903,
      "epoch": 5.097714381984613,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027199696498342893,
      "loss": 4.8397,
      "mean_token_accuracy": 0.2391986295580864,
      "num_tokens": 101327085.0,
      "step": 45390
    },
    {
      "entropy": 5.4343773365020756,
      "epoch": 5.09827595889257,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002719568198885901,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22926938235759736,
      "num_tokens": 101340473.0,
      "step": 45395
    },
    {
      "entropy": 5.500201225280762,
      "epoch": 5.098837535800528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002719166748906473,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2386144742369652,
      "num_tokens": 101353434.0,
      "step": 45400
    },
    {
      "entropy": 5.435666036605835,
      "epoch": 5.099399112708485,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027187652999087887,
      "loss": 4.8436,
      "mean_token_accuracy": 0.2308952033519745,
      "num_tokens": 101363782.0,
      "step": 45405
    },
    {
      "entropy": 5.492531824111938,
      "epoch": 5.099960689616443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027183638519056284,
      "loss": 4.9377,
      "mean_token_accuracy": 0.225948566198349,
      "num_tokens": 101374985.0,
      "step": 45410
    },
    {
      "entropy": 5.443627262115479,
      "epoch": 5.1005222665244005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002717962404909776,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23832945078611373,
      "num_tokens": 101386916.0,
      "step": 45415
    },
    {
      "entropy": 5.435193634033203,
      "epoch": 5.101083843432358,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027175609589340135,
      "loss": 4.7946,
      "mean_token_accuracy": 0.2355433687567711,
      "num_tokens": 101397656.0,
      "step": 45420
    },
    {
      "entropy": 5.445901107788086,
      "epoch": 5.101645420340316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002717159513991122,
      "loss": 4.8337,
      "mean_token_accuracy": 0.23730040788650514,
      "num_tokens": 101408725.0,
      "step": 45425
    },
    {
      "entropy": 5.384482955932617,
      "epoch": 5.102206997248273,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027167580700938844,
      "loss": 4.7778,
      "mean_token_accuracy": 0.24391164034605026,
      "num_tokens": 101419580.0,
      "step": 45430
    },
    {
      "entropy": 5.466874599456787,
      "epoch": 5.10276857415623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027163566272550826,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22958824336528777,
      "num_tokens": 101430735.0,
      "step": 45435
    },
    {
      "entropy": 5.4402025699615475,
      "epoch": 5.103330151064188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027159551854874986,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23232457041740417,
      "num_tokens": 101440544.0,
      "step": 45440
    },
    {
      "entropy": 5.506302452087402,
      "epoch": 5.103891727972146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002715553744803915,
      "loss": 4.9346,
      "mean_token_accuracy": 0.2241054505109787,
      "num_tokens": 101450793.0,
      "step": 45445
    },
    {
      "entropy": 5.503649044036865,
      "epoch": 5.104453304880104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002715152305217114,
      "loss": 5.033,
      "mean_token_accuracy": 0.22178096771240235,
      "num_tokens": 101463593.0,
      "step": 45450
    },
    {
      "entropy": 5.41388258934021,
      "epoch": 5.105014881788061,
      "grad_norm": 1.0,
      "learning_rate": 0.0002714750866739877,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22960714101791382,
      "num_tokens": 101474334.0,
      "step": 45455
    },
    {
      "entropy": 5.469512557983398,
      "epoch": 5.105576458696018,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002714349429384986,
      "loss": 4.9315,
      "mean_token_accuracy": 0.23194750994443894,
      "num_tokens": 101485356.0,
      "step": 45460
    },
    {
      "entropy": 5.447671937942505,
      "epoch": 5.106138035603976,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027139479931652226,
      "loss": 4.8183,
      "mean_token_accuracy": 0.2372212141752243,
      "num_tokens": 101495939.0,
      "step": 45465
    },
    {
      "entropy": 5.4805036067962645,
      "epoch": 5.1066996125119335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002713546558093368,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22828693836927413,
      "num_tokens": 101505563.0,
      "step": 45470
    },
    {
      "entropy": 5.351919412612915,
      "epoch": 5.107261189419891,
      "grad_norm": 1.125,
      "learning_rate": 0.0002713145124182207,
      "loss": 4.7731,
      "mean_token_accuracy": 0.24333920180797577,
      "num_tokens": 101516829.0,
      "step": 45475
    },
    {
      "entropy": 5.492102241516113,
      "epoch": 5.107822766327849,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027127436914445194,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22709906995296478,
      "num_tokens": 101527427.0,
      "step": 45480
    },
    {
      "entropy": 5.486274003982544,
      "epoch": 5.108384343235806,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002712342259893086,
      "loss": 4.8292,
      "mean_token_accuracy": 0.23833133578300475,
      "num_tokens": 101537991.0,
      "step": 45485
    },
    {
      "entropy": 5.466387224197388,
      "epoch": 5.108945920143763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027119408295406906,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22331980168819426,
      "num_tokens": 101549485.0,
      "step": 45490
    },
    {
      "entropy": 5.5258513450622555,
      "epoch": 5.109507497051721,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002711539400400114,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22183857709169388,
      "num_tokens": 101560719.0,
      "step": 45495
    },
    {
      "entropy": 5.51230525970459,
      "epoch": 5.110069073959679,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027111379724841373,
      "loss": 5.0024,
      "mean_token_accuracy": 0.2182866483926773,
      "num_tokens": 101572118.0,
      "step": 45500
    },
    {
      "entropy": 5.545909786224366,
      "epoch": 5.110630650867637,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027107365458055424,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2296516329050064,
      "num_tokens": 101582050.0,
      "step": 45505
    },
    {
      "entropy": 5.414169454574585,
      "epoch": 5.111192227775594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002710335120377112,
      "loss": 4.7329,
      "mean_token_accuracy": 0.24135648608207702,
      "num_tokens": 101592529.0,
      "step": 45510
    },
    {
      "entropy": 5.436296367645264,
      "epoch": 5.111753804683551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027099336962116253,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22834448963403703,
      "num_tokens": 101603106.0,
      "step": 45515
    },
    {
      "entropy": 5.505915594100952,
      "epoch": 5.112315381591509,
      "grad_norm": 1.125,
      "learning_rate": 0.00027095322733218665,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22363995015621185,
      "num_tokens": 101614171.0,
      "step": 45520
    },
    {
      "entropy": 5.541736698150634,
      "epoch": 5.112876958499466,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002709130851720616,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22581171989440918,
      "num_tokens": 101626311.0,
      "step": 45525
    },
    {
      "entropy": 5.422439813613892,
      "epoch": 5.113438535407424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027087294314206544,
      "loss": 4.839,
      "mean_token_accuracy": 0.23613592833280564,
      "num_tokens": 101637923.0,
      "step": 45530
    },
    {
      "entropy": 5.589333724975586,
      "epoch": 5.114000112315382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002708328012434764,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21952267587184907,
      "num_tokens": 101647943.0,
      "step": 45535
    },
    {
      "entropy": 5.494402551651001,
      "epoch": 5.114561689223339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002707926594775725,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22963372617959976,
      "num_tokens": 101660095.0,
      "step": 45540
    },
    {
      "entropy": 5.458812427520752,
      "epoch": 5.115123266131297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027075251784563196,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23435254395008087,
      "num_tokens": 101670092.0,
      "step": 45545
    },
    {
      "entropy": 5.471007013320923,
      "epoch": 5.115684843039254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000270712376348933,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22508775889873506,
      "num_tokens": 101681692.0,
      "step": 45550
    },
    {
      "entropy": 5.4641820907592775,
      "epoch": 5.1162464199472115,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002706722349887536,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2290792405605316,
      "num_tokens": 101692569.0,
      "step": 45555
    },
    {
      "entropy": 5.566471433639526,
      "epoch": 5.11680799685517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002706320937663719,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22443944215774536,
      "num_tokens": 101704185.0,
      "step": 45560
    },
    {
      "entropy": 5.498971891403198,
      "epoch": 5.117369573763127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002705919526830661,
      "loss": 4.958,
      "mean_token_accuracy": 0.22598046213388442,
      "num_tokens": 101715500.0,
      "step": 45565
    },
    {
      "entropy": 5.559986162185669,
      "epoch": 5.117931150671084,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027055181174011415,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2330833539366722,
      "num_tokens": 101727562.0,
      "step": 45570
    },
    {
      "entropy": 5.481236267089844,
      "epoch": 5.118492727579042,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027051167093879424,
      "loss": 4.7991,
      "mean_token_accuracy": 0.235282202064991,
      "num_tokens": 101739269.0,
      "step": 45575
    },
    {
      "entropy": 5.490000581741333,
      "epoch": 5.119054304486999,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002704715302803846,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22758667171001434,
      "num_tokens": 101750189.0,
      "step": 45580
    },
    {
      "entropy": 5.504766130447388,
      "epoch": 5.119615881394957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027043138976616304,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22267509251832962,
      "num_tokens": 101760824.0,
      "step": 45585
    },
    {
      "entropy": 5.419432640075684,
      "epoch": 5.120177458302915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027039124939740787,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24323923587799073,
      "num_tokens": 101772647.0,
      "step": 45590
    },
    {
      "entropy": 5.459737920761109,
      "epoch": 5.120739035210872,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027035110917539713,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2282433792948723,
      "num_tokens": 101783020.0,
      "step": 45595
    },
    {
      "entropy": 5.487755584716797,
      "epoch": 5.12130061211883,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027031096910140877,
      "loss": 4.8694,
      "mean_token_accuracy": 0.23529072105884552,
      "num_tokens": 101793372.0,
      "step": 45600
    },
    {
      "entropy": 5.458370065689087,
      "epoch": 5.121862189026787,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027027082917672107,
      "loss": 4.9045,
      "mean_token_accuracy": 0.23624385744333268,
      "num_tokens": 101805077.0,
      "step": 45605
    },
    {
      "entropy": 5.458372783660889,
      "epoch": 5.1224237659347445,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027023068940261203,
      "loss": 4.908,
      "mean_token_accuracy": 0.2342908948659897,
      "num_tokens": 101816334.0,
      "step": 45610
    },
    {
      "entropy": 5.544493103027344,
      "epoch": 5.122985342842703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002701905497803595,
      "loss": 5.0516,
      "mean_token_accuracy": 0.22163873612880708,
      "num_tokens": 101827812.0,
      "step": 45615
    },
    {
      "entropy": 5.424270677566528,
      "epoch": 5.12354691975066,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002701504103112419,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23602136969566345,
      "num_tokens": 101839400.0,
      "step": 45620
    },
    {
      "entropy": 5.539113998413086,
      "epoch": 5.124108496658617,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027011027099653706,
      "loss": 4.9812,
      "mean_token_accuracy": 0.23038126528263092,
      "num_tokens": 101851948.0,
      "step": 45625
    },
    {
      "entropy": 5.49214997291565,
      "epoch": 5.124670073566575,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002700701318375232,
      "loss": 4.964,
      "mean_token_accuracy": 0.22232070416212082,
      "num_tokens": 101862985.0,
      "step": 45630
    },
    {
      "entropy": 5.461416721343994,
      "epoch": 5.125231650474532,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002700299928354781,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2275281861424446,
      "num_tokens": 101873247.0,
      "step": 45635
    },
    {
      "entropy": 5.5090944290161135,
      "epoch": 5.12579322738249,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026998985399168005,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2324616864323616,
      "num_tokens": 101883686.0,
      "step": 45640
    },
    {
      "entropy": 5.533977317810058,
      "epoch": 5.126354804290448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026994971530740696,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22731431424617768,
      "num_tokens": 101894905.0,
      "step": 45645
    },
    {
      "entropy": 5.487450361251831,
      "epoch": 5.126916381198405,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026990957678393684,
      "loss": 4.8953,
      "mean_token_accuracy": 0.2301494523882866,
      "num_tokens": 101906671.0,
      "step": 45650
    },
    {
      "entropy": 5.400568914413452,
      "epoch": 5.127477958106363,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026986943842254783,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23347915709018707,
      "num_tokens": 101917440.0,
      "step": 45655
    },
    {
      "entropy": 5.476686334609985,
      "epoch": 5.12803953501432,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002698293002245179,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22416940331459045,
      "num_tokens": 101929177.0,
      "step": 45660
    },
    {
      "entropy": 5.505867528915405,
      "epoch": 5.128601111922277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026978916219112505,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2270990192890167,
      "num_tokens": 101940034.0,
      "step": 45665
    },
    {
      "entropy": 5.585483741760254,
      "epoch": 5.1291626888302355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026974902432364727,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22170396000146866,
      "num_tokens": 101952108.0,
      "step": 45670
    },
    {
      "entropy": 5.577408981323242,
      "epoch": 5.129724265738193,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026970888662336263,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22466563433408737,
      "num_tokens": 101963617.0,
      "step": 45675
    },
    {
      "entropy": 5.50240421295166,
      "epoch": 5.130285842646151,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002696687490915491,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23172969073057176,
      "num_tokens": 101975565.0,
      "step": 45680
    },
    {
      "entropy": 5.420003509521484,
      "epoch": 5.130847419554108,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026962861172948456,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23502437621355057,
      "num_tokens": 101988048.0,
      "step": 45685
    },
    {
      "entropy": 5.390117454528808,
      "epoch": 5.131408996462065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026958847453844713,
      "loss": 4.851,
      "mean_token_accuracy": 0.22679630368947984,
      "num_tokens": 101999045.0,
      "step": 45690
    },
    {
      "entropy": 5.48456563949585,
      "epoch": 5.131970573370023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026954833751971486,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22106243371963502,
      "num_tokens": 102009679.0,
      "step": 45695
    },
    {
      "entropy": 5.461053657531738,
      "epoch": 5.132532150277981,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002695082006745656,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22585563659667968,
      "num_tokens": 102021378.0,
      "step": 45700
    },
    {
      "entropy": 5.525784969329834,
      "epoch": 5.133093727185938,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002694680640042774,
      "loss": 4.9092,
      "mean_token_accuracy": 0.23190022110939026,
      "num_tokens": 102031037.0,
      "step": 45705
    },
    {
      "entropy": 5.510881519317627,
      "epoch": 5.133655304093896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026942792751012814,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2394351452589035,
      "num_tokens": 102041767.0,
      "step": 45710
    },
    {
      "entropy": 5.451894474029541,
      "epoch": 5.134216881001853,
      "grad_norm": 1.125,
      "learning_rate": 0.00026938779119339585,
      "loss": 4.8793,
      "mean_token_accuracy": 0.228719961643219,
      "num_tokens": 102052792.0,
      "step": 45715
    },
    {
      "entropy": 5.469319820404053,
      "epoch": 5.13477845790981,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002693476550553584,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2405119091272354,
      "num_tokens": 102064238.0,
      "step": 45720
    },
    {
      "entropy": 5.482106447219849,
      "epoch": 5.1353400348177685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026930751909729384,
      "loss": 4.9124,
      "mean_token_accuracy": 0.23022722750902175,
      "num_tokens": 102075557.0,
      "step": 45725
    },
    {
      "entropy": 5.428512239456177,
      "epoch": 5.135901611725726,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026926738332048013,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2377573937177658,
      "num_tokens": 102086691.0,
      "step": 45730
    },
    {
      "entropy": 5.491974353790283,
      "epoch": 5.136463188633684,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026922724772619516,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2315070375800133,
      "num_tokens": 102097215.0,
      "step": 45735
    },
    {
      "entropy": 5.506746625900268,
      "epoch": 5.137024765541641,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002691871123157168,
      "loss": 4.9194,
      "mean_token_accuracy": 0.22899875789880753,
      "num_tokens": 102108408.0,
      "step": 45740
    },
    {
      "entropy": 5.5105218410491945,
      "epoch": 5.137586342449598,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002691469770903231,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23155590295791625,
      "num_tokens": 102120594.0,
      "step": 45745
    },
    {
      "entropy": 5.476330995559692,
      "epoch": 5.138147919357556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002691068420512919,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2340434178709984,
      "num_tokens": 102132306.0,
      "step": 45750
    },
    {
      "entropy": 5.520437335968017,
      "epoch": 5.1387094962655135,
      "grad_norm": 1.078125,
      "learning_rate": 0.000269066707199901,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22708712816238402,
      "num_tokens": 102143271.0,
      "step": 45755
    },
    {
      "entropy": 5.482902717590332,
      "epoch": 5.139271073173471,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026902657253742855,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22782009541988374,
      "num_tokens": 102154657.0,
      "step": 45760
    },
    {
      "entropy": 5.465263414382934,
      "epoch": 5.139832650081429,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002689864380651524,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2348073035478592,
      "num_tokens": 102167135.0,
      "step": 45765
    },
    {
      "entropy": 5.392302560806274,
      "epoch": 5.140394226989386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002689463037843504,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23751562237739562,
      "num_tokens": 102178485.0,
      "step": 45770
    },
    {
      "entropy": 5.469488668441772,
      "epoch": 5.140955803897343,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002689061696963004,
      "loss": 4.9351,
      "mean_token_accuracy": 0.231259685754776,
      "num_tokens": 102189378.0,
      "step": 45775
    },
    {
      "entropy": 5.465953636169433,
      "epoch": 5.141517380805301,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002688660358022803,
      "loss": 4.9121,
      "mean_token_accuracy": 0.23028521835803986,
      "num_tokens": 102199864.0,
      "step": 45780
    },
    {
      "entropy": 5.401256418228149,
      "epoch": 5.142078957713259,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000268825902103568,
      "loss": 4.7943,
      "mean_token_accuracy": 0.23135091364383698,
      "num_tokens": 102210699.0,
      "step": 45785
    },
    {
      "entropy": 5.492848062515259,
      "epoch": 5.142640534621217,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002687857686014414,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2286272391676903,
      "num_tokens": 102221724.0,
      "step": 45790
    },
    {
      "entropy": 5.408935260772705,
      "epoch": 5.143202111529174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002687456352971783,
      "loss": 4.6942,
      "mean_token_accuracy": 0.24891749024391174,
      "num_tokens": 102232023.0,
      "step": 45795
    },
    {
      "entropy": 5.428686571121216,
      "epoch": 5.143763688437131,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002687055021920566,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2352169543504715,
      "num_tokens": 102242526.0,
      "step": 45800
    },
    {
      "entropy": 5.378143453598023,
      "epoch": 5.144325265345089,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002686653692873542,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23135044276714326,
      "num_tokens": 102253271.0,
      "step": 45805
    },
    {
      "entropy": 5.486434602737427,
      "epoch": 5.1448868422530465,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026862523658434883,
      "loss": 4.8999,
      "mean_token_accuracy": 0.228525809943676,
      "num_tokens": 102263089.0,
      "step": 45810
    },
    {
      "entropy": 5.467685079574585,
      "epoch": 5.145448419161004,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002685851040843185,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23434340059757233,
      "num_tokens": 102273312.0,
      "step": 45815
    },
    {
      "entropy": 5.519162130355835,
      "epoch": 5.146009996068962,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026854497178854087,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22512232065200805,
      "num_tokens": 102283940.0,
      "step": 45820
    },
    {
      "entropy": 5.504747772216797,
      "epoch": 5.146571572976919,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002685048396982939,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2243882328271866,
      "num_tokens": 102295666.0,
      "step": 45825
    },
    {
      "entropy": 5.448159313201904,
      "epoch": 5.147133149884877,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00026846470781485517,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22914567291736604,
      "num_tokens": 102306463.0,
      "step": 45830
    },
    {
      "entropy": 5.422773933410644,
      "epoch": 5.147694726792834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002684245761395028,
      "loss": 4.7897,
      "mean_token_accuracy": 0.23686749786138533,
      "num_tokens": 102317443.0,
      "step": 45835
    },
    {
      "entropy": 5.46085638999939,
      "epoch": 5.148256303700792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026838444467351447,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2282401517033577,
      "num_tokens": 102327924.0,
      "step": 45840
    },
    {
      "entropy": 5.624072027206421,
      "epoch": 5.14881788060875,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026834431341816805,
      "loss": 5.0293,
      "mean_token_accuracy": 0.22284373939037322,
      "num_tokens": 102339364.0,
      "step": 45845
    },
    {
      "entropy": 5.4984358787536625,
      "epoch": 5.149379457516707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026830418237474126,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22565895169973374,
      "num_tokens": 102350947.0,
      "step": 45850
    },
    {
      "entropy": 5.547277450561523,
      "epoch": 5.149941034424664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002682640515445118,
      "loss": 4.9952,
      "mean_token_accuracy": 0.22333341389894484,
      "num_tokens": 102362412.0,
      "step": 45855
    },
    {
      "entropy": 5.494638299942016,
      "epoch": 5.150502611332622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002682239209287577,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23205736726522447,
      "num_tokens": 102373630.0,
      "step": 45860
    },
    {
      "entropy": 5.47002067565918,
      "epoch": 5.151064188240579,
      "grad_norm": 1.125,
      "learning_rate": 0.00026818379052875646,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22329700440168382,
      "num_tokens": 102385081.0,
      "step": 45865
    },
    {
      "entropy": 5.451447439193726,
      "epoch": 5.1516257651485375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000268143660345786,
      "loss": 4.8134,
      "mean_token_accuracy": 0.22976609021425248,
      "num_tokens": 102396519.0,
      "step": 45870
    },
    {
      "entropy": 5.482119750976563,
      "epoch": 5.152187342056495,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002681035303811241,
      "loss": 4.9334,
      "mean_token_accuracy": 0.2262008011341095,
      "num_tokens": 102406326.0,
      "step": 45875
    },
    {
      "entropy": 5.58284969329834,
      "epoch": 5.152748918964452,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026806340063604845,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22293318063020706,
      "num_tokens": 102416965.0,
      "step": 45880
    },
    {
      "entropy": 5.494816398620605,
      "epoch": 5.15331049587241,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026802327111183686,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2292417675256729,
      "num_tokens": 102428414.0,
      "step": 45885
    },
    {
      "entropy": 5.467991590499878,
      "epoch": 5.153872072780367,
      "grad_norm": 1.125,
      "learning_rate": 0.00026798314180976696,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2357127383351326,
      "num_tokens": 102438665.0,
      "step": 45890
    },
    {
      "entropy": 5.50474443435669,
      "epoch": 5.1544336496883245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026794301273111654,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2234726533293724,
      "num_tokens": 102449828.0,
      "step": 45895
    },
    {
      "entropy": 5.519992351531982,
      "epoch": 5.154995226596283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002679028838771633,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2292328342795372,
      "num_tokens": 102461007.0,
      "step": 45900
    },
    {
      "entropy": 5.4586756229400635,
      "epoch": 5.15555680350424,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026786275524918503,
      "loss": 4.8151,
      "mean_token_accuracy": 0.2393457919359207,
      "num_tokens": 102472372.0,
      "step": 45905
    },
    {
      "entropy": 5.447400617599487,
      "epoch": 5.156118380412197,
      "grad_norm": 1.125,
      "learning_rate": 0.0002678226268484594,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23341257125139236,
      "num_tokens": 102482361.0,
      "step": 45910
    },
    {
      "entropy": 5.347509574890137,
      "epoch": 5.156679957320155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002677824986762641,
      "loss": 4.7392,
      "mean_token_accuracy": 0.23225993514060975,
      "num_tokens": 102493058.0,
      "step": 45915
    },
    {
      "entropy": 5.548799228668213,
      "epoch": 5.157241534228112,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002677423707338769,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22717125713825226,
      "num_tokens": 102505242.0,
      "step": 45920
    },
    {
      "entropy": 5.500287866592407,
      "epoch": 5.1578031111360705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002677022430225754,
      "loss": 4.8971,
      "mean_token_accuracy": 0.23288116306066514,
      "num_tokens": 102517580.0,
      "step": 45925
    },
    {
      "entropy": 5.517814874649048,
      "epoch": 5.158364688044028,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026766211554363727,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22851693332195283,
      "num_tokens": 102527336.0,
      "step": 45930
    },
    {
      "entropy": 5.513534212112427,
      "epoch": 5.158926264951985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002676219882983403,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22133236825466157,
      "num_tokens": 102538511.0,
      "step": 45935
    },
    {
      "entropy": 5.542808628082275,
      "epoch": 5.159487841859943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026758186128796204,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23253501802682877,
      "num_tokens": 102549981.0,
      "step": 45940
    },
    {
      "entropy": 5.594483423233032,
      "epoch": 5.1600494187679,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026754173451378014,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2272316813468933,
      "num_tokens": 102560504.0,
      "step": 45945
    },
    {
      "entropy": 5.508536052703858,
      "epoch": 5.1606109956758575,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026750160797707245,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2328792169690132,
      "num_tokens": 102571136.0,
      "step": 45950
    },
    {
      "entropy": 5.469540786743164,
      "epoch": 5.161172572583816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002674614816791164,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23996069729328157,
      "num_tokens": 102582879.0,
      "step": 45955
    },
    {
      "entropy": 5.4965691566467285,
      "epoch": 5.161734149491773,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026742135562118967,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23177324384450912,
      "num_tokens": 102593171.0,
      "step": 45960
    },
    {
      "entropy": 5.544973707199096,
      "epoch": 5.16229572639973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002673812298045699,
      "loss": 5.0103,
      "mean_token_accuracy": 0.22415003925561905,
      "num_tokens": 102603816.0,
      "step": 45965
    },
    {
      "entropy": 5.47808084487915,
      "epoch": 5.162857303307688,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026734110423053483,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23296729773283004,
      "num_tokens": 102614360.0,
      "step": 45970
    },
    {
      "entropy": 5.555528545379639,
      "epoch": 5.163418880215645,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026730097890036175,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2214311346411705,
      "num_tokens": 102625320.0,
      "step": 45975
    },
    {
      "entropy": 5.481560230255127,
      "epoch": 5.163980457123603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026726085381532867,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22857046276330947,
      "num_tokens": 102635728.0,
      "step": 45980
    },
    {
      "entropy": 5.414718866348267,
      "epoch": 5.164542034031561,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000267220728976713,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23007957488298417,
      "num_tokens": 102645880.0,
      "step": 45985
    },
    {
      "entropy": 5.4534811019897464,
      "epoch": 5.165103610939518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002671806043857923,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22689806073904037,
      "num_tokens": 102657854.0,
      "step": 45990
    },
    {
      "entropy": 5.481391382217407,
      "epoch": 5.165665187847476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002671404800438442,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2322985604405403,
      "num_tokens": 102669637.0,
      "step": 45995
    },
    {
      "entropy": 5.482212877273559,
      "epoch": 5.166226764755433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002671003559521463,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2360638052225113,
      "num_tokens": 102680520.0,
      "step": 46000
    },
    {
      "entropy": 5.44390606880188,
      "epoch": 5.16678834166339,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026706023211197613,
      "loss": 4.8149,
      "mean_token_accuracy": 0.22979138791561127,
      "num_tokens": 102691633.0,
      "step": 46005
    },
    {
      "entropy": 5.510936832427978,
      "epoch": 5.1673499185713485,
      "grad_norm": 1.125,
      "learning_rate": 0.0002670201085246111,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22662893682718277,
      "num_tokens": 102703171.0,
      "step": 46010
    },
    {
      "entropy": 5.5542405128479,
      "epoch": 5.167911495479306,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026697998519132905,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22670727223157883,
      "num_tokens": 102714636.0,
      "step": 46015
    },
    {
      "entropy": 5.501375150680542,
      "epoch": 5.168473072387264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026693986211340736,
      "loss": 5.0466,
      "mean_token_accuracy": 0.22851192355155944,
      "num_tokens": 102725554.0,
      "step": 46020
    },
    {
      "entropy": 5.4967015266418455,
      "epoch": 5.169034649295221,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026689973929212363,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22717914283275603,
      "num_tokens": 102736705.0,
      "step": 46025
    },
    {
      "entropy": 5.597528982162475,
      "epoch": 5.169596226203178,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026685961672875524,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2188520535826683,
      "num_tokens": 102749187.0,
      "step": 46030
    },
    {
      "entropy": 5.491814661026001,
      "epoch": 5.170157803111136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002668194944245799,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22465777546167373,
      "num_tokens": 102760095.0,
      "step": 46035
    },
    {
      "entropy": 5.472312498092651,
      "epoch": 5.170719380019094,
      "grad_norm": 1.09375,
      "learning_rate": 0.000266779372380875,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2256108120083809,
      "num_tokens": 102771086.0,
      "step": 46040
    },
    {
      "entropy": 5.521039009094238,
      "epoch": 5.171280956927051,
      "grad_norm": 1.09375,
      "learning_rate": 0.000266739250598918,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22076838314533234,
      "num_tokens": 102782024.0,
      "step": 46045
    },
    {
      "entropy": 5.633496332168579,
      "epoch": 5.171842533835009,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026669912907998657,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22493979781866075,
      "num_tokens": 102793654.0,
      "step": 46050
    },
    {
      "entropy": 5.606273984909057,
      "epoch": 5.172404110742966,
      "grad_norm": 1.109375,
      "learning_rate": 0.000266659007825358,
      "loss": 5.074,
      "mean_token_accuracy": 0.22115335017442703,
      "num_tokens": 102804908.0,
      "step": 46055
    },
    {
      "entropy": 5.38487138748169,
      "epoch": 5.172965687650924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026661888683631,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23697859644889832,
      "num_tokens": 102815399.0,
      "step": 46060
    },
    {
      "entropy": 5.410790872573853,
      "epoch": 5.1735272645588815,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026657876611411974,
      "loss": 4.8286,
      "mean_token_accuracy": 0.23193645477294922,
      "num_tokens": 102825734.0,
      "step": 46065
    },
    {
      "entropy": 5.442007160186767,
      "epoch": 5.174088841466839,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026653864566006494,
      "loss": 4.7304,
      "mean_token_accuracy": 0.23923491835594177,
      "num_tokens": 102836251.0,
      "step": 46070
    },
    {
      "entropy": 5.534513282775879,
      "epoch": 5.174650418374797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002664985254754229,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22153810113668443,
      "num_tokens": 102847626.0,
      "step": 46075
    },
    {
      "entropy": 5.4645754337310795,
      "epoch": 5.175211995282754,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026645840556147103,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23611160069704057,
      "num_tokens": 102859424.0,
      "step": 46080
    },
    {
      "entropy": 5.466070556640625,
      "epoch": 5.175773572190711,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002664182859194869,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23740164190530777,
      "num_tokens": 102870466.0,
      "step": 46085
    },
    {
      "entropy": 5.451819372177124,
      "epoch": 5.176335149098669,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002663781665507479,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2292115420103073,
      "num_tokens": 102882181.0,
      "step": 46090
    },
    {
      "entropy": 5.490746927261353,
      "epoch": 5.1768967260066265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026633804745653146,
      "loss": 4.9222,
      "mean_token_accuracy": 0.22980857491493226,
      "num_tokens": 102892562.0,
      "step": 46095
    },
    {
      "entropy": 5.430376672744751,
      "epoch": 5.177458302914584,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026629792863811484,
      "loss": 4.8211,
      "mean_token_accuracy": 0.24317560642957686,
      "num_tokens": 102904380.0,
      "step": 46100
    },
    {
      "entropy": 5.427767944335938,
      "epoch": 5.178019879822542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002662578100967756,
      "loss": 4.8522,
      "mean_token_accuracy": 0.22583554238080977,
      "num_tokens": 102916018.0,
      "step": 46105
    },
    {
      "entropy": 5.470332860946655,
      "epoch": 5.178581456730499,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002662176918337911,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21751319617033005,
      "num_tokens": 102927885.0,
      "step": 46110
    },
    {
      "entropy": 5.42456021308899,
      "epoch": 5.179143033638457,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002661775738504387,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2347772851586342,
      "num_tokens": 102938343.0,
      "step": 46115
    },
    {
      "entropy": 5.421642160415649,
      "epoch": 5.179704610546414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00026613745614799574,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2315191864967346,
      "num_tokens": 102949341.0,
      "step": 46120
    },
    {
      "entropy": 5.546340036392212,
      "epoch": 5.180266187454372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002660973387277396,
      "loss": 4.9613,
      "mean_token_accuracy": 0.23733300864696502,
      "num_tokens": 102959722.0,
      "step": 46125
    },
    {
      "entropy": 5.531007766723633,
      "epoch": 5.18082776436233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002660572215909477,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23100270628929137,
      "num_tokens": 102971423.0,
      "step": 46130
    },
    {
      "entropy": 5.483974409103394,
      "epoch": 5.181389341270287,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002660171047388973,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23724421709775925,
      "num_tokens": 102983218.0,
      "step": 46135
    },
    {
      "entropy": 5.500847434997558,
      "epoch": 5.181950918178244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026597698817286576,
      "loss": 4.9257,
      "mean_token_accuracy": 0.231524321436882,
      "num_tokens": 102995287.0,
      "step": 46140
    },
    {
      "entropy": 5.482327604293824,
      "epoch": 5.182512495086202,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026593687189413044,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23138115406036378,
      "num_tokens": 103006578.0,
      "step": 46145
    },
    {
      "entropy": 5.54092001914978,
      "epoch": 5.1830740719941595,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026589675590396865,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2258661776781082,
      "num_tokens": 103017959.0,
      "step": 46150
    },
    {
      "entropy": 5.464177465438842,
      "epoch": 5.183635648902117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002658566402036576,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22490028589963912,
      "num_tokens": 103029931.0,
      "step": 46155
    },
    {
      "entropy": 5.4839826107025145,
      "epoch": 5.184197225810075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002658165247944747,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2356893703341484,
      "num_tokens": 103041603.0,
      "step": 46160
    },
    {
      "entropy": 5.453537511825561,
      "epoch": 5.184758802718032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026577640967769727,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2303358405828476,
      "num_tokens": 103053353.0,
      "step": 46165
    },
    {
      "entropy": 5.404115438461304,
      "epoch": 5.18532037962599,
      "grad_norm": 0.953125,
      "learning_rate": 0.00026573629485460257,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23797769397497176,
      "num_tokens": 103064295.0,
      "step": 46170
    },
    {
      "entropy": 5.526855993270874,
      "epoch": 5.185881956533947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026569618032646773,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2311050370335579,
      "num_tokens": 103075582.0,
      "step": 46175
    },
    {
      "entropy": 5.482394695281982,
      "epoch": 5.186443533441905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026565606609457016,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23118651062250137,
      "num_tokens": 103087969.0,
      "step": 46180
    },
    {
      "entropy": 5.51949634552002,
      "epoch": 5.187005110349863,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002656159521601871,
      "loss": 4.907,
      "mean_token_accuracy": 0.2267500415444374,
      "num_tokens": 103099505.0,
      "step": 46185
    },
    {
      "entropy": 5.548426580429077,
      "epoch": 5.18756668725782,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002655758385245956,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22202078253030777,
      "num_tokens": 103110706.0,
      "step": 46190
    },
    {
      "entropy": 5.4418271541595455,
      "epoch": 5.188128264165777,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002655357251890732,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2416694089770317,
      "num_tokens": 103121129.0,
      "step": 46195
    },
    {
      "entropy": 5.425090312957764,
      "epoch": 5.188689841073735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026549561215489695,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22899033278226852,
      "num_tokens": 103130706.0,
      "step": 46200
    },
    {
      "entropy": 5.504629039764405,
      "epoch": 5.189251417981692,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002654554994233442,
      "loss": 4.9522,
      "mean_token_accuracy": 0.2262539952993393,
      "num_tokens": 103140441.0,
      "step": 46205
    },
    {
      "entropy": 5.40563325881958,
      "epoch": 5.1898129948896505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002654153869956919,
      "loss": 4.832,
      "mean_token_accuracy": 0.231583933532238,
      "num_tokens": 103151245.0,
      "step": 46210
    },
    {
      "entropy": 5.500794458389282,
      "epoch": 5.190374571797608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002653752748732175,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23184361159801484,
      "num_tokens": 103161985.0,
      "step": 46215
    },
    {
      "entropy": 5.491947269439697,
      "epoch": 5.190936148705565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000265335163057198,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22954720705747605,
      "num_tokens": 103174471.0,
      "step": 46220
    },
    {
      "entropy": 5.432328748703003,
      "epoch": 5.191497725613523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002652950515489107,
      "loss": 4.8739,
      "mean_token_accuracy": 0.23716988712549208,
      "num_tokens": 103184853.0,
      "step": 46225
    },
    {
      "entropy": 5.5254052639007565,
      "epoch": 5.19205930252148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002652549403496327,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22064388245344163,
      "num_tokens": 103195909.0,
      "step": 46230
    },
    {
      "entropy": 5.570833921432495,
      "epoch": 5.1926208794294375,
      "grad_norm": 1.125,
      "learning_rate": 0.00026521482946064123,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22648655772209167,
      "num_tokens": 103206557.0,
      "step": 46235
    },
    {
      "entropy": 5.414399671554565,
      "epoch": 5.193182456337396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002651747188832134,
      "loss": 4.8361,
      "mean_token_accuracy": 0.23680318295955657,
      "num_tokens": 103216887.0,
      "step": 46240
    },
    {
      "entropy": 5.485964250564575,
      "epoch": 5.193744033245353,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002651346086186262,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22607296854257583,
      "num_tokens": 103228879.0,
      "step": 46245
    },
    {
      "entropy": 5.46967716217041,
      "epoch": 5.194305610153311,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000265094498668157,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2230827808380127,
      "num_tokens": 103240564.0,
      "step": 46250
    },
    {
      "entropy": 5.468884658813477,
      "epoch": 5.194867187061268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026505438903308285,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22979423105716706,
      "num_tokens": 103251297.0,
      "step": 46255
    },
    {
      "entropy": 5.515857172012329,
      "epoch": 5.195428763969225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026501427971468056,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23330413401126862,
      "num_tokens": 103262041.0,
      "step": 46260
    },
    {
      "entropy": 5.470881128311158,
      "epoch": 5.1959903408771835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002649741707142277,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2338673397898674,
      "num_tokens": 103273024.0,
      "step": 46265
    },
    {
      "entropy": 5.46405291557312,
      "epoch": 5.196551917785141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026493406203300106,
      "loss": 4.898,
      "mean_token_accuracy": 0.22713986337184905,
      "num_tokens": 103284511.0,
      "step": 46270
    },
    {
      "entropy": 5.503048610687256,
      "epoch": 5.197113494693098,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026489395367227785,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2297889068722725,
      "num_tokens": 103296787.0,
      "step": 46275
    },
    {
      "entropy": 5.50200629234314,
      "epoch": 5.197675071601056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000264853845633335,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2319417491555214,
      "num_tokens": 103308422.0,
      "step": 46280
    },
    {
      "entropy": 5.416675472259522,
      "epoch": 5.198236648509013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002648137379174497,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23166537433862686,
      "num_tokens": 103319592.0,
      "step": 46285
    },
    {
      "entropy": 5.458620548248291,
      "epoch": 5.1987982254169705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002647736305258989,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23449993878602982,
      "num_tokens": 103330702.0,
      "step": 46290
    },
    {
      "entropy": 5.393842697143555,
      "epoch": 5.199359802324929,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002647335234599597,
      "loss": 4.8199,
      "mean_token_accuracy": 0.22958943545818328,
      "num_tokens": 103342339.0,
      "step": 46295
    },
    {
      "entropy": 5.398126840591431,
      "epoch": 5.199921379232886,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026469341672090906,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2324619024991989,
      "num_tokens": 103353604.0,
      "step": 46300
    },
    {
      "entropy": 5.447867393493652,
      "epoch": 5.200482956140844,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026465331031002404,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23298390507698058,
      "num_tokens": 103363713.0,
      "step": 46305
    },
    {
      "entropy": 5.539913034439087,
      "epoch": 5.201044533048801,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026461320422858166,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22113827764987945,
      "num_tokens": 103374966.0,
      "step": 46310
    },
    {
      "entropy": 5.520695686340332,
      "epoch": 5.201606109956758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002645730984778589,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2236853078007698,
      "num_tokens": 103387025.0,
      "step": 46315
    },
    {
      "entropy": 5.558062410354614,
      "epoch": 5.202167686864716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002645329930591327,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22464126199483872,
      "num_tokens": 103397947.0,
      "step": 46320
    },
    {
      "entropy": 5.520836210250854,
      "epoch": 5.202729263772674,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026449288797368006,
      "loss": 4.9499,
      "mean_token_accuracy": 0.23059084862470627,
      "num_tokens": 103409874.0,
      "step": 46325
    },
    {
      "entropy": 5.473841953277588,
      "epoch": 5.203290840680631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026445278322277796,
      "loss": 4.9413,
      "mean_token_accuracy": 0.22731707841157914,
      "num_tokens": 103423484.0,
      "step": 46330
    },
    {
      "entropy": 5.471132469177246,
      "epoch": 5.203852417588589,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002644126788077032,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22919536530971527,
      "num_tokens": 103434933.0,
      "step": 46335
    },
    {
      "entropy": 5.504837894439698,
      "epoch": 5.204413994496546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026437257472973296,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22275933921337127,
      "num_tokens": 103447111.0,
      "step": 46340
    },
    {
      "entropy": 5.468713617324829,
      "epoch": 5.204975571404503,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026433247099014406,
      "loss": 4.8322,
      "mean_token_accuracy": 0.24390691220760347,
      "num_tokens": 103457627.0,
      "step": 46345
    },
    {
      "entropy": 5.478974056243897,
      "epoch": 5.2055371483124615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002642923675902135,
      "loss": 4.8977,
      "mean_token_accuracy": 0.23063876032829284,
      "num_tokens": 103468047.0,
      "step": 46350
    },
    {
      "entropy": 5.409407663345337,
      "epoch": 5.206098725220419,
      "grad_norm": 1.015625,
      "learning_rate": 0.000264252264531218,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23332178592681885,
      "num_tokens": 103479671.0,
      "step": 46355
    },
    {
      "entropy": 5.4242401123046875,
      "epoch": 5.206660302128377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026421216181443457,
      "loss": 4.8358,
      "mean_token_accuracy": 0.22665320783853532,
      "num_tokens": 103490565.0,
      "step": 46360
    },
    {
      "entropy": 5.470352506637573,
      "epoch": 5.207221879036334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002641720594411401,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23110245019197465,
      "num_tokens": 103500915.0,
      "step": 46365
    },
    {
      "entropy": 5.496555423736572,
      "epoch": 5.207783455944291,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026413195741261145,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23020098805427552,
      "num_tokens": 103511739.0,
      "step": 46370
    },
    {
      "entropy": 5.511297988891601,
      "epoch": 5.208345032852249,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002640918557301256,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22127144932746887,
      "num_tokens": 103521957.0,
      "step": 46375
    },
    {
      "entropy": 5.491480302810669,
      "epoch": 5.208906609760207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002640517543949592,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2276636242866516,
      "num_tokens": 103532511.0,
      "step": 46380
    },
    {
      "entropy": 5.517809772491455,
      "epoch": 5.209468186668164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002640116534083892,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22811852693557738,
      "num_tokens": 103544484.0,
      "step": 46385
    },
    {
      "entropy": 5.492289876937866,
      "epoch": 5.210029763576122,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026397155277169246,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23226603269577026,
      "num_tokens": 103555020.0,
      "step": 46390
    },
    {
      "entropy": 5.485903072357178,
      "epoch": 5.210591340484079,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002639314524861457,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23386629074811935,
      "num_tokens": 103565331.0,
      "step": 46395
    },
    {
      "entropy": 5.503296899795532,
      "epoch": 5.211152917392037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002638913525530258,
      "loss": 4.9402,
      "mean_token_accuracy": 0.23266783058643342,
      "num_tokens": 103576970.0,
      "step": 46400
    },
    {
      "entropy": 5.479026365280151,
      "epoch": 5.2117144942999944,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026385125297360946,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2213706597685814,
      "num_tokens": 103587981.0,
      "step": 46405
    },
    {
      "entropy": 5.476946687698364,
      "epoch": 5.212276071207952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026381115374917363,
      "loss": 4.8703,
      "mean_token_accuracy": 0.23217329978942872,
      "num_tokens": 103599508.0,
      "step": 46410
    },
    {
      "entropy": 5.462182426452637,
      "epoch": 5.21283764811591,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026377105488099503,
      "loss": 4.8243,
      "mean_token_accuracy": 0.2349432021379471,
      "num_tokens": 103609982.0,
      "step": 46415
    },
    {
      "entropy": 5.434537315368653,
      "epoch": 5.213399225023867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026373095637035037,
      "loss": 4.8535,
      "mean_token_accuracy": 0.23662357032299042,
      "num_tokens": 103620419.0,
      "step": 46420
    },
    {
      "entropy": 5.404030084609985,
      "epoch": 5.213960801931824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026369085821851645,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23976810425519943,
      "num_tokens": 103630512.0,
      "step": 46425
    },
    {
      "entropy": 5.514559412002564,
      "epoch": 5.214522378839782,
      "grad_norm": 1.125,
      "learning_rate": 0.00026365076042677,
      "loss": 4.9626,
      "mean_token_accuracy": 0.23102798014879228,
      "num_tokens": 103641596.0,
      "step": 46430
    },
    {
      "entropy": 5.49004430770874,
      "epoch": 5.2150839557477395,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002636106629963877,
      "loss": 4.8989,
      "mean_token_accuracy": 0.23138244301080704,
      "num_tokens": 103654291.0,
      "step": 46435
    },
    {
      "entropy": 5.3542516231536865,
      "epoch": 5.215645532655698,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002635705659286463,
      "loss": 4.7461,
      "mean_token_accuracy": 0.2398676887154579,
      "num_tokens": 103665843.0,
      "step": 46440
    },
    {
      "entropy": 5.583477210998535,
      "epoch": 5.216207109563655,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026353046922482256,
      "loss": 5.0364,
      "mean_token_accuracy": 0.22368732392787932,
      "num_tokens": 103676682.0,
      "step": 46445
    },
    {
      "entropy": 5.504293298721313,
      "epoch": 5.216768686471612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002634903728861931,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22198515087366105,
      "num_tokens": 103687189.0,
      "step": 46450
    },
    {
      "entropy": 5.482424736022949,
      "epoch": 5.21733026337957,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002634502769140346,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2258310079574585,
      "num_tokens": 103698332.0,
      "step": 46455
    },
    {
      "entropy": 5.468286228179932,
      "epoch": 5.217891840287527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026341018130962384,
      "loss": 4.8702,
      "mean_token_accuracy": 0.22989479452371597,
      "num_tokens": 103708652.0,
      "step": 46460
    },
    {
      "entropy": 5.509663438796997,
      "epoch": 5.218453417195485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026337008607423733,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22210776060819626,
      "num_tokens": 103719275.0,
      "step": 46465
    },
    {
      "entropy": 5.446264791488647,
      "epoch": 5.219014994103443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026332999120915184,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2348087474703789,
      "num_tokens": 103730865.0,
      "step": 46470
    },
    {
      "entropy": 5.54120659828186,
      "epoch": 5.2195765710114,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026328989671564395,
      "loss": 4.971,
      "mean_token_accuracy": 0.21766522377729416,
      "num_tokens": 103743784.0,
      "step": 46475
    },
    {
      "entropy": 5.576703214645386,
      "epoch": 5.220138147919357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002632498025949902,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22465194016695023,
      "num_tokens": 103755580.0,
      "step": 46480
    },
    {
      "entropy": 5.537931299209594,
      "epoch": 5.220699724827315,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002632097088484674,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2236671507358551,
      "num_tokens": 103767296.0,
      "step": 46485
    },
    {
      "entropy": 5.470188760757447,
      "epoch": 5.2212613017352725,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026316961547735203,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23469316363334655,
      "num_tokens": 103778176.0,
      "step": 46490
    },
    {
      "entropy": 5.58207597732544,
      "epoch": 5.221822878643231,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026312952248292067,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22552027106285094,
      "num_tokens": 103791499.0,
      "step": 46495
    },
    {
      "entropy": 5.528898668289185,
      "epoch": 5.222384455551188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002630894298664499,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2294543519616127,
      "num_tokens": 103803102.0,
      "step": 46500
    },
    {
      "entropy": 5.487461137771606,
      "epoch": 5.222946032459145,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026304933762921637,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23537017703056334,
      "num_tokens": 103814273.0,
      "step": 46505
    },
    {
      "entropy": 5.400848293304444,
      "epoch": 5.223507609367103,
      "grad_norm": 1.125,
      "learning_rate": 0.0002630092457724965,
      "loss": 4.7263,
      "mean_token_accuracy": 0.2442261889576912,
      "num_tokens": 103823713.0,
      "step": 46510
    },
    {
      "entropy": 5.430395078659058,
      "epoch": 5.22406918627506,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002629691542975669,
      "loss": 4.7898,
      "mean_token_accuracy": 0.2377841055393219,
      "num_tokens": 103834244.0,
      "step": 46515
    },
    {
      "entropy": 5.354468297958374,
      "epoch": 5.224630763183018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002629290632057041,
      "loss": 4.7348,
      "mean_token_accuracy": 0.24283353090286255,
      "num_tokens": 103844561.0,
      "step": 46520
    },
    {
      "entropy": 5.518036556243897,
      "epoch": 5.225192340090976,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002628889724981846,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2266793355345726,
      "num_tokens": 103856812.0,
      "step": 46525
    },
    {
      "entropy": 5.550623941421509,
      "epoch": 5.225753916998933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002628488821762849,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2180622085928917,
      "num_tokens": 103870062.0,
      "step": 46530
    },
    {
      "entropy": 5.541503190994263,
      "epoch": 5.22631549390689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026280879224128165,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22562605887651443,
      "num_tokens": 103880763.0,
      "step": 46535
    },
    {
      "entropy": 5.522700691223145,
      "epoch": 5.226877070814848,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000262768702694451,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22398133724927902,
      "num_tokens": 103892528.0,
      "step": 46540
    },
    {
      "entropy": 5.455557537078858,
      "epoch": 5.227438647722805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002627286135370697,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22852995991706848,
      "num_tokens": 103903589.0,
      "step": 46545
    },
    {
      "entropy": 5.406871747970581,
      "epoch": 5.2280002246307635,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000262688524770414,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23025539219379426,
      "num_tokens": 103914192.0,
      "step": 46550
    },
    {
      "entropy": 5.457571506500244,
      "epoch": 5.228561801538721,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026264843639576055,
      "loss": 4.814,
      "mean_token_accuracy": 0.23362844586372375,
      "num_tokens": 103924649.0,
      "step": 46555
    },
    {
      "entropy": 5.414918756484985,
      "epoch": 5.229123378446678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002626083484143856,
      "loss": 4.8115,
      "mean_token_accuracy": 0.23245368003845215,
      "num_tokens": 103935834.0,
      "step": 46560
    },
    {
      "entropy": 5.398805427551269,
      "epoch": 5.229684955354636,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002625682608275657,
      "loss": 4.82,
      "mean_token_accuracy": 0.23627463430166246,
      "num_tokens": 103947007.0,
      "step": 46565
    },
    {
      "entropy": 5.4752813339233395,
      "epoch": 5.230246532262593,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002625281736365772,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22789255231618882,
      "num_tokens": 103958424.0,
      "step": 46570
    },
    {
      "entropy": 5.488842630386353,
      "epoch": 5.2308081091705505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026248808684269646,
      "loss": 4.9264,
      "mean_token_accuracy": 0.23529693186283113,
      "num_tokens": 103969289.0,
      "step": 46575
    },
    {
      "entropy": 5.598863124847412,
      "epoch": 5.231369686078509,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002624480004472,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2275763913989067,
      "num_tokens": 103981081.0,
      "step": 46580
    },
    {
      "entropy": 5.542724895477295,
      "epoch": 5.231931262986466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026240791445136384,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22570017129182815,
      "num_tokens": 103990904.0,
      "step": 46585
    },
    {
      "entropy": 5.471207332611084,
      "epoch": 5.232492839894424,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002623678288564647,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23407834768295288,
      "num_tokens": 104001535.0,
      "step": 46590
    },
    {
      "entropy": 5.4588840961456295,
      "epoch": 5.233054416802381,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002623277436637788,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2319916680455208,
      "num_tokens": 104012840.0,
      "step": 46595
    },
    {
      "entropy": 5.453291559219361,
      "epoch": 5.233615993710338,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002622876588745825,
      "loss": 4.8038,
      "mean_token_accuracy": 0.23526890873908995,
      "num_tokens": 104024035.0,
      "step": 46600
    },
    {
      "entropy": 5.452269697189331,
      "epoch": 5.2341775706182965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000262247574490152,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23293790221214294,
      "num_tokens": 104035391.0,
      "step": 46605
    },
    {
      "entropy": 5.478780078887939,
      "epoch": 5.234739147526254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002622074905117637,
      "loss": 4.8818,
      "mean_token_accuracy": 0.23556840121746064,
      "num_tokens": 104046645.0,
      "step": 46610
    },
    {
      "entropy": 5.434322881698608,
      "epoch": 5.235300724434211,
      "grad_norm": 1.0,
      "learning_rate": 0.00026216740694069387,
      "loss": 4.8287,
      "mean_token_accuracy": 0.233448326587677,
      "num_tokens": 104057706.0,
      "step": 46615
    },
    {
      "entropy": 5.472760152816773,
      "epoch": 5.235862301342169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002621273237782188,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22963133454322815,
      "num_tokens": 104068076.0,
      "step": 46620
    },
    {
      "entropy": 5.456567764282227,
      "epoch": 5.236423878250126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002620872410256146,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2303246095776558,
      "num_tokens": 104078116.0,
      "step": 46625
    },
    {
      "entropy": 5.415603399276733,
      "epoch": 5.236985455158084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002620471586841577,
      "loss": 4.7241,
      "mean_token_accuracy": 0.2425907462835312,
      "num_tokens": 104089543.0,
      "step": 46630
    },
    {
      "entropy": 5.431361150741577,
      "epoch": 5.2375470320660416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002620070767551244,
      "loss": 4.866,
      "mean_token_accuracy": 0.2327197328209877,
      "num_tokens": 104100824.0,
      "step": 46635
    },
    {
      "entropy": 5.458374166488648,
      "epoch": 5.238108608973999,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002619669952397908,
      "loss": 4.9016,
      "mean_token_accuracy": 0.23159324377775192,
      "num_tokens": 104112410.0,
      "step": 46640
    },
    {
      "entropy": 5.42400631904602,
      "epoch": 5.238670185881957,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002619269141394331,
      "loss": 4.91,
      "mean_token_accuracy": 0.23049300014972687,
      "num_tokens": 104124299.0,
      "step": 46645
    },
    {
      "entropy": 5.573159646987915,
      "epoch": 5.239231762789914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002618868334553275,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22278565168380737,
      "num_tokens": 104136266.0,
      "step": 46650
    },
    {
      "entropy": 5.549076843261719,
      "epoch": 5.239793339697871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026184675318875023,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2263721600174904,
      "num_tokens": 104146449.0,
      "step": 46655
    },
    {
      "entropy": 5.452953290939331,
      "epoch": 5.240354916605829,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026180667334097733,
      "loss": 4.7783,
      "mean_token_accuracy": 0.2325626641511917,
      "num_tokens": 104156959.0,
      "step": 46660
    },
    {
      "entropy": 5.405904483795166,
      "epoch": 5.240916493513787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002617665939132852,
      "loss": 4.8339,
      "mean_token_accuracy": 0.2280412271618843,
      "num_tokens": 104167834.0,
      "step": 46665
    },
    {
      "entropy": 5.5289695262908936,
      "epoch": 5.241478070421744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002617265149069498,
      "loss": 4.972,
      "mean_token_accuracy": 0.22490683645009996,
      "num_tokens": 104180035.0,
      "step": 46670
    },
    {
      "entropy": 5.5932775974273685,
      "epoch": 5.242039647329702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002616864363232474,
      "loss": 5.1123,
      "mean_token_accuracy": 0.22470228523015975,
      "num_tokens": 104192960.0,
      "step": 46675
    },
    {
      "entropy": 5.545761060714722,
      "epoch": 5.242601224237659,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026164635816345395,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22346824407577515,
      "num_tokens": 104204120.0,
      "step": 46680
    },
    {
      "entropy": 5.4762208461761475,
      "epoch": 5.243162801145617,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002616062804288457,
      "loss": 4.869,
      "mean_token_accuracy": 0.23322182446718215,
      "num_tokens": 104215222.0,
      "step": 46685
    },
    {
      "entropy": 5.48479380607605,
      "epoch": 5.2437243780535745,
      "grad_norm": 1.0,
      "learning_rate": 0.0002615662031206986,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22929975241422654,
      "num_tokens": 104227208.0,
      "step": 46690
    },
    {
      "entropy": 5.526920795440674,
      "epoch": 5.244285954961532,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002615261262402887,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2182338058948517,
      "num_tokens": 104236836.0,
      "step": 46695
    },
    {
      "entropy": 5.502941036224366,
      "epoch": 5.24484753186949,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026148604978889237,
      "loss": 4.9041,
      "mean_token_accuracy": 0.2339513286948204,
      "num_tokens": 104248950.0,
      "step": 46700
    },
    {
      "entropy": 5.5197367668151855,
      "epoch": 5.245409108777447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026144597376778533,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2281977117061615,
      "num_tokens": 104260179.0,
      "step": 46705
    },
    {
      "entropy": 5.519990539550781,
      "epoch": 5.245970685685404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002614058981782438,
      "loss": 4.991,
      "mean_token_accuracy": 0.2238247275352478,
      "num_tokens": 104273704.0,
      "step": 46710
    },
    {
      "entropy": 5.458600282669067,
      "epoch": 5.246532262593362,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026136582302154377,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2328959360718727,
      "num_tokens": 104283851.0,
      "step": 46715
    },
    {
      "entropy": 5.474749040603638,
      "epoch": 5.24709383950132,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026132574829896123,
      "loss": 4.8828,
      "mean_token_accuracy": 0.2389626160264015,
      "num_tokens": 104296064.0,
      "step": 46720
    },
    {
      "entropy": 5.543599367141724,
      "epoch": 5.247655416409277,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002612856740117721,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21965974122285842,
      "num_tokens": 104309016.0,
      "step": 46725
    },
    {
      "entropy": 5.47886552810669,
      "epoch": 5.248216993317235,
      "grad_norm": 0.96875,
      "learning_rate": 0.00026124560016125236,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2283435732126236,
      "num_tokens": 104320598.0,
      "step": 46730
    },
    {
      "entropy": 5.565599250793457,
      "epoch": 5.248778570225192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002612055267486782,
      "loss": 4.9293,
      "mean_token_accuracy": 0.222578227519989,
      "num_tokens": 104332055.0,
      "step": 46735
    },
    {
      "entropy": 5.528148365020752,
      "epoch": 5.24934014713315,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026116545377532536,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2330057442188263,
      "num_tokens": 104343879.0,
      "step": 46740
    },
    {
      "entropy": 5.510451412200927,
      "epoch": 5.249901724041107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002611253812424698,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22486024498939514,
      "num_tokens": 104354653.0,
      "step": 46745
    },
    {
      "entropy": 5.52443528175354,
      "epoch": 5.250463300949065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002610853091513876,
      "loss": 4.9541,
      "mean_token_accuracy": 0.2289663404226303,
      "num_tokens": 104365273.0,
      "step": 46750
    },
    {
      "entropy": 5.544688606262207,
      "epoch": 5.251024877857023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026104523750335445,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2229477733373642,
      "num_tokens": 104375990.0,
      "step": 46755
    },
    {
      "entropy": 5.464516067504883,
      "epoch": 5.25158645476498,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002610051662996464,
      "loss": 4.7891,
      "mean_token_accuracy": 0.2419636219739914,
      "num_tokens": 104385975.0,
      "step": 46760
    },
    {
      "entropy": 5.515393686294556,
      "epoch": 5.252148031672937,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026096509554153914,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22842023968696595,
      "num_tokens": 104396053.0,
      "step": 46765
    },
    {
      "entropy": 5.432059955596924,
      "epoch": 5.252709608580895,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002609250252303088,
      "loss": 4.8093,
      "mean_token_accuracy": 0.23116195052862168,
      "num_tokens": 104407874.0,
      "step": 46770
    },
    {
      "entropy": 5.431617546081543,
      "epoch": 5.2532711854888525,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002608849553672311,
      "loss": 4.8858,
      "mean_token_accuracy": 0.23428474217653275,
      "num_tokens": 104418527.0,
      "step": 46775
    },
    {
      "entropy": 5.4260032176971436,
      "epoch": 5.253832762396811,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026084488595358197,
      "loss": 4.913,
      "mean_token_accuracy": 0.23078569620847703,
      "num_tokens": 104430490.0,
      "step": 46780
    },
    {
      "entropy": 5.463120985031128,
      "epoch": 5.254394339304768,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002608048169906371,
      "loss": 4.8283,
      "mean_token_accuracy": 0.23413912951946259,
      "num_tokens": 104441767.0,
      "step": 46785
    },
    {
      "entropy": 5.4229625225067135,
      "epoch": 5.254955916212725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026076474847967236,
      "loss": 4.89,
      "mean_token_accuracy": 0.23068717420101165,
      "num_tokens": 104452720.0,
      "step": 46790
    },
    {
      "entropy": 5.504771709442139,
      "epoch": 5.255517493120683,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026072468042196356,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2287175640463829,
      "num_tokens": 104463829.0,
      "step": 46795
    },
    {
      "entropy": 5.497013187408447,
      "epoch": 5.25607907002864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026068461281878645,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23125652521848677,
      "num_tokens": 104474830.0,
      "step": 46800
    },
    {
      "entropy": 5.461471033096314,
      "epoch": 5.256640646936598,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026064454567141683,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23174595534801484,
      "num_tokens": 104485518.0,
      "step": 46805
    },
    {
      "entropy": 5.565434980392456,
      "epoch": 5.257202223844556,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002606044789811304,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2289423242211342,
      "num_tokens": 104495562.0,
      "step": 46810
    },
    {
      "entropy": 5.503579425811767,
      "epoch": 5.257763800752513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000260564412749203,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2314058870077133,
      "num_tokens": 104506057.0,
      "step": 46815
    },
    {
      "entropy": 5.451401567459106,
      "epoch": 5.258325377660471,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026052434697691035,
      "loss": 4.8704,
      "mean_token_accuracy": 0.24010505974292756,
      "num_tokens": 104518269.0,
      "step": 46820
    },
    {
      "entropy": 5.412418365478516,
      "epoch": 5.258886954568428,
      "grad_norm": 1.015625,
      "learning_rate": 0.000260484281665528,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2341032937169075,
      "num_tokens": 104528469.0,
      "step": 46825
    },
    {
      "entropy": 5.483194875717163,
      "epoch": 5.2594485314763855,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002604442168163318,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22798044085502625,
      "num_tokens": 104539079.0,
      "step": 46830
    },
    {
      "entropy": 5.480548429489136,
      "epoch": 5.260010108384344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002604041524305974,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2244734913110733,
      "num_tokens": 104551235.0,
      "step": 46835
    },
    {
      "entropy": 5.637797737121582,
      "epoch": 5.260571685292301,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002603640885096004,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22572603523731233,
      "num_tokens": 104563911.0,
      "step": 46840
    },
    {
      "entropy": 5.521130609512329,
      "epoch": 5.261133262200258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002603240250546165,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2258487954735756,
      "num_tokens": 104574877.0,
      "step": 46845
    },
    {
      "entropy": 5.543003416061401,
      "epoch": 5.261694839108216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026028396206692134,
      "loss": 4.9457,
      "mean_token_accuracy": 0.23016353696584702,
      "num_tokens": 104585209.0,
      "step": 46850
    },
    {
      "entropy": 5.421526145935059,
      "epoch": 5.262256416016173,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026024389954779054,
      "loss": 4.8693,
      "mean_token_accuracy": 0.2370264321565628,
      "num_tokens": 104596213.0,
      "step": 46855
    },
    {
      "entropy": 5.466647672653198,
      "epoch": 5.2628179929241306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002602038374984997,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22939542531967164,
      "num_tokens": 104606813.0,
      "step": 46860
    },
    {
      "entropy": 5.504680919647217,
      "epoch": 5.263379569832089,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026016377592032446,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2334458649158478,
      "num_tokens": 104619040.0,
      "step": 46865
    },
    {
      "entropy": 5.355341196060181,
      "epoch": 5.263941146740046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002601237148145403,
      "loss": 4.7447,
      "mean_token_accuracy": 0.23960089981555938,
      "num_tokens": 104629879.0,
      "step": 46870
    },
    {
      "entropy": 5.420273399353027,
      "epoch": 5.264502723648004,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026008365418242284,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23667743057012558,
      "num_tokens": 104641408.0,
      "step": 46875
    },
    {
      "entropy": 5.563666963577271,
      "epoch": 5.265064300555961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002600435940252476,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22532201260328294,
      "num_tokens": 104653537.0,
      "step": 46880
    },
    {
      "entropy": 5.396293783187867,
      "epoch": 5.265625877463918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002600035343442902,
      "loss": 4.819,
      "mean_token_accuracy": 0.23363107740879058,
      "num_tokens": 104665499.0,
      "step": 46885
    },
    {
      "entropy": 5.503332662582397,
      "epoch": 5.2661874543718765,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00025996347514082603,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2276037186384201,
      "num_tokens": 104676964.0,
      "step": 46890
    },
    {
      "entropy": 5.5844666957855225,
      "epoch": 5.266749031279834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00025992341641613065,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22505789250135422,
      "num_tokens": 104688853.0,
      "step": 46895
    },
    {
      "entropy": 5.453549337387085,
      "epoch": 5.267310608187791,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002598833581714795,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2355702444911003,
      "num_tokens": 104699723.0,
      "step": 46900
    },
    {
      "entropy": 5.48667540550232,
      "epoch": 5.267872185095749,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025984330040814817,
      "loss": 4.8974,
      "mean_token_accuracy": 0.23401861190795897,
      "num_tokens": 104711370.0,
      "step": 46905
    },
    {
      "entropy": 5.566730690002442,
      "epoch": 5.268433762003706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002598032431274119,
      "loss": 4.993,
      "mean_token_accuracy": 0.22286490350961685,
      "num_tokens": 104722939.0,
      "step": 46910
    },
    {
      "entropy": 5.543216133117676,
      "epoch": 5.2689953389116635,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025976318633054636,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22493660002946853,
      "num_tokens": 104735433.0,
      "step": 46915
    },
    {
      "entropy": 5.513320016860962,
      "epoch": 5.269556915819622,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002597231300188269,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23168213665485382,
      "num_tokens": 104748325.0,
      "step": 46920
    },
    {
      "entropy": 5.507792329788208,
      "epoch": 5.270118492727579,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025968307419352884,
      "loss": 4.923,
      "mean_token_accuracy": 0.22768891006708145,
      "num_tokens": 104758745.0,
      "step": 46925
    },
    {
      "entropy": 5.430860090255737,
      "epoch": 5.270680069635537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025964301885592766,
      "loss": 4.796,
      "mean_token_accuracy": 0.23355162441730498,
      "num_tokens": 104769424.0,
      "step": 46930
    },
    {
      "entropy": 5.42628846168518,
      "epoch": 5.271241646543494,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002596029640072988,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2387458473443985,
      "num_tokens": 104779913.0,
      "step": 46935
    },
    {
      "entropy": 5.452473402023315,
      "epoch": 5.271803223451451,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002595629096489174,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22325150519609452,
      "num_tokens": 104793066.0,
      "step": 46940
    },
    {
      "entropy": 5.443156909942627,
      "epoch": 5.2723648003594095,
      "grad_norm": 1.046875,
      "learning_rate": 0.000259522855782059,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2278371497988701,
      "num_tokens": 104803641.0,
      "step": 46945
    },
    {
      "entropy": 5.597671842575073,
      "epoch": 5.272926377267367,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002594828024079989,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21625016778707504,
      "num_tokens": 104816025.0,
      "step": 46950
    },
    {
      "entropy": 5.547744512557983,
      "epoch": 5.273487954175324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025944274952801233,
      "loss": 4.9866,
      "mean_token_accuracy": 0.22987845540046692,
      "num_tokens": 104828188.0,
      "step": 46955
    },
    {
      "entropy": 5.5158043384552,
      "epoch": 5.274049531083282,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002594026971433747,
      "loss": 4.932,
      "mean_token_accuracy": 0.2289382115006447,
      "num_tokens": 104838276.0,
      "step": 46960
    },
    {
      "entropy": 5.530891275405883,
      "epoch": 5.274611107991239,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025936264525536126,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22979342937469482,
      "num_tokens": 104848295.0,
      "step": 46965
    },
    {
      "entropy": 5.574198198318482,
      "epoch": 5.275172684899197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002593225938652472,
      "loss": 5.0137,
      "mean_token_accuracy": 0.22519916146993638,
      "num_tokens": 104857857.0,
      "step": 46970
    },
    {
      "entropy": 5.469282817840576,
      "epoch": 5.2757342618071545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002592825429743079,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22826243340969085,
      "num_tokens": 104867819.0,
      "step": 46975
    },
    {
      "entropy": 5.559512758255005,
      "epoch": 5.276295838715112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002592424925838185,
      "loss": 4.931,
      "mean_token_accuracy": 0.22707390636205674,
      "num_tokens": 104879034.0,
      "step": 46980
    },
    {
      "entropy": 5.549154996871948,
      "epoch": 5.27685741562307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002592024426950541,
      "loss": 4.999,
      "mean_token_accuracy": 0.22419430911540986,
      "num_tokens": 104890245.0,
      "step": 46985
    },
    {
      "entropy": 5.504646444320679,
      "epoch": 5.277418992531027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025916239330929026,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22741807252168655,
      "num_tokens": 104900805.0,
      "step": 46990
    },
    {
      "entropy": 5.447807359695434,
      "epoch": 5.277980569438984,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025912234442780185,
      "loss": 4.8149,
      "mean_token_accuracy": 0.2376958504319191,
      "num_tokens": 104911636.0,
      "step": 46995
    },
    {
      "entropy": 5.508055734634399,
      "epoch": 5.278542146346942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025908229605186423,
      "loss": 4.8837,
      "mean_token_accuracy": 0.23219896256923675,
      "num_tokens": 104922253.0,
      "step": 47000
    },
    {
      "entropy": 5.4464414596557615,
      "epoch": 5.2791037232549,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025904224818275247,
      "loss": 4.867,
      "mean_token_accuracy": 0.23286062777042388,
      "num_tokens": 104933274.0,
      "step": 47005
    },
    {
      "entropy": 5.538744211196899,
      "epoch": 5.279665300162858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025900220082174165,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22070562541484834,
      "num_tokens": 104943931.0,
      "step": 47010
    },
    {
      "entropy": 5.413900518417359,
      "epoch": 5.280226877070815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025896215397010696,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2268196389079094,
      "num_tokens": 104955080.0,
      "step": 47015
    },
    {
      "entropy": 5.445582866668701,
      "epoch": 5.280788453978772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002589221076291236,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2355947658419609,
      "num_tokens": 104967232.0,
      "step": 47020
    },
    {
      "entropy": 5.515359115600586,
      "epoch": 5.28135003088673,
      "grad_norm": 1.0,
      "learning_rate": 0.00025888206180006647,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22250103950500488,
      "num_tokens": 104979462.0,
      "step": 47025
    },
    {
      "entropy": 5.5138133525848385,
      "epoch": 5.2819116077946875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002588420164842108,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22548136115074158,
      "num_tokens": 104990567.0,
      "step": 47030
    },
    {
      "entropy": 5.389851713180542,
      "epoch": 5.282473184702645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002588019716828316,
      "loss": 4.7556,
      "mean_token_accuracy": 0.2388958603143692,
      "num_tokens": 105001965.0,
      "step": 47035
    },
    {
      "entropy": 5.44419002532959,
      "epoch": 5.283034761610603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025876192739720383,
      "loss": 4.7676,
      "mean_token_accuracy": 0.24015588760375978,
      "num_tokens": 105013349.0,
      "step": 47040
    },
    {
      "entropy": 5.462402725219727,
      "epoch": 5.28359633851856,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025872188362860264,
      "loss": 4.904,
      "mean_token_accuracy": 0.22398434579372406,
      "num_tokens": 105025009.0,
      "step": 47045
    },
    {
      "entropy": 5.4033092021942135,
      "epoch": 5.284157915426517,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025868184037830305,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23369160741567613,
      "num_tokens": 105035474.0,
      "step": 47050
    },
    {
      "entropy": 5.5333436012268065,
      "epoch": 5.284719492334475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002586417976475798,
      "loss": 4.952,
      "mean_token_accuracy": 0.2269732803106308,
      "num_tokens": 105047310.0,
      "step": 47055
    },
    {
      "entropy": 5.386792182922363,
      "epoch": 5.285281069242433,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025860175543770817,
      "loss": 4.7759,
      "mean_token_accuracy": 0.2332301452755928,
      "num_tokens": 105057198.0,
      "step": 47060
    },
    {
      "entropy": 5.554803276062012,
      "epoch": 5.285842646150391,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002585617137499631,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22878552675247193,
      "num_tokens": 105067885.0,
      "step": 47065
    },
    {
      "entropy": 5.523142099380493,
      "epoch": 5.286404223058348,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002585216725856193,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22765052318572998,
      "num_tokens": 105078917.0,
      "step": 47070
    },
    {
      "entropy": 5.546887969970703,
      "epoch": 5.286965799966305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002584816319459519,
      "loss": 4.996,
      "mean_token_accuracy": 0.22355802357196808,
      "num_tokens": 105089425.0,
      "step": 47075
    },
    {
      "entropy": 5.541159534454346,
      "epoch": 5.287527376874263,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002584415918322358,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23129201978445052,
      "num_tokens": 105100043.0,
      "step": 47080
    },
    {
      "entropy": 5.415031099319458,
      "epoch": 5.28808895378222,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002584015522457458,
      "loss": 4.8072,
      "mean_token_accuracy": 0.2328536331653595,
      "num_tokens": 105110317.0,
      "step": 47085
    },
    {
      "entropy": 5.487929439544677,
      "epoch": 5.288650530690178,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025836151318775675,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22357714027166367,
      "num_tokens": 105120688.0,
      "step": 47090
    },
    {
      "entropy": 5.619646120071411,
      "epoch": 5.289212107598136,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025832147465954365,
      "loss": 5.0258,
      "mean_token_accuracy": 0.22272807955741883,
      "num_tokens": 105130883.0,
      "step": 47095
    },
    {
      "entropy": 5.5816563129425045,
      "epoch": 5.289773684506093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025828143666238124,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22541789710521698,
      "num_tokens": 105141203.0,
      "step": 47100
    },
    {
      "entropy": 5.492149496078492,
      "epoch": 5.29033526141405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025824139919754436,
      "loss": 4.8901,
      "mean_token_accuracy": 0.23536172807216643,
      "num_tokens": 105152419.0,
      "step": 47105
    },
    {
      "entropy": 5.583808183670044,
      "epoch": 5.290896838322008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002582013622663079,
      "loss": 5.041,
      "mean_token_accuracy": 0.22361351996660234,
      "num_tokens": 105164242.0,
      "step": 47110
    },
    {
      "entropy": 5.5338358879089355,
      "epoch": 5.2914584152299655,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002581613258699465,
      "loss": 4.9349,
      "mean_token_accuracy": 0.23363456577062608,
      "num_tokens": 105174008.0,
      "step": 47115
    },
    {
      "entropy": 5.4236218452453615,
      "epoch": 5.292019992137924,
      "grad_norm": 1.15625,
      "learning_rate": 0.000258121290009735,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23153066486120225,
      "num_tokens": 105184880.0,
      "step": 47120
    },
    {
      "entropy": 5.520227479934692,
      "epoch": 5.292581569045881,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002580812546869481,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22457044422626496,
      "num_tokens": 105196492.0,
      "step": 47125
    },
    {
      "entropy": 5.435938358306885,
      "epoch": 5.293143145953838,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002580412199028607,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2298796847462654,
      "num_tokens": 105206227.0,
      "step": 47130
    },
    {
      "entropy": 5.612427568435669,
      "epoch": 5.293704722861796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025800118565874754,
      "loss": 5.0808,
      "mean_token_accuracy": 0.216640405356884,
      "num_tokens": 105217877.0,
      "step": 47135
    },
    {
      "entropy": 5.493530321121216,
      "epoch": 5.294266299769753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002579611519558831,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23266010284423827,
      "num_tokens": 105229124.0,
      "step": 47140
    },
    {
      "entropy": 5.419517850875854,
      "epoch": 5.294827876677711,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025792111879554224,
      "loss": 4.7801,
      "mean_token_accuracy": 0.240191088616848,
      "num_tokens": 105240952.0,
      "step": 47145
    },
    {
      "entropy": 5.462467861175537,
      "epoch": 5.295389453585669,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002578810861789996,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23170026391744614,
      "num_tokens": 105252594.0,
      "step": 47150
    },
    {
      "entropy": 5.531341648101806,
      "epoch": 5.295951030493626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002578410541075298,
      "loss": 4.9843,
      "mean_token_accuracy": 0.22480778843164445,
      "num_tokens": 105263457.0,
      "step": 47155
    },
    {
      "entropy": 5.502911138534546,
      "epoch": 5.296512607401584,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025780102258240745,
      "loss": 4.8838,
      "mean_token_accuracy": 0.23336536586284637,
      "num_tokens": 105275455.0,
      "step": 47160
    },
    {
      "entropy": 5.529229021072387,
      "epoch": 5.297074184309541,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002577609916049072,
      "loss": 4.934,
      "mean_token_accuracy": 0.23119357079267502,
      "num_tokens": 105285643.0,
      "step": 47165
    },
    {
      "entropy": 5.495049667358399,
      "epoch": 5.2976357612174985,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002577209611763037,
      "loss": 4.9208,
      "mean_token_accuracy": 0.23065212219953538,
      "num_tokens": 105297322.0,
      "step": 47170
    },
    {
      "entropy": 5.431743955612182,
      "epoch": 5.298197338125457,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002576809312978715,
      "loss": 4.8002,
      "mean_token_accuracy": 0.23061246424913406,
      "num_tokens": 105307977.0,
      "step": 47175
    },
    {
      "entropy": 5.46057653427124,
      "epoch": 5.298758915033414,
      "grad_norm": 1.125,
      "learning_rate": 0.0002576409019708852,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23107492327690124,
      "num_tokens": 105318461.0,
      "step": 47180
    },
    {
      "entropy": 5.593760251998901,
      "epoch": 5.299320491941371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002576008731966193,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21937705129384993,
      "num_tokens": 105329494.0,
      "step": 47185
    },
    {
      "entropy": 5.353734016418457,
      "epoch": 5.299882068849329,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025756084497634827,
      "loss": 4.7288,
      "mean_token_accuracy": 0.240184922516346,
      "num_tokens": 105341475.0,
      "step": 47190
    },
    {
      "entropy": 5.480043125152588,
      "epoch": 5.300443645757286,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002575208173113468,
      "loss": 4.8061,
      "mean_token_accuracy": 0.24251188933849335,
      "num_tokens": 105352075.0,
      "step": 47195
    },
    {
      "entropy": 5.536924505233765,
      "epoch": 5.301005222665244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002574807902028891,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22492753118276596,
      "num_tokens": 105363801.0,
      "step": 47200
    },
    {
      "entropy": 5.529438781738281,
      "epoch": 5.301566799573202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025744076365225,
      "loss": 4.921,
      "mean_token_accuracy": 0.22975416332483292,
      "num_tokens": 105374821.0,
      "step": 47205
    },
    {
      "entropy": 5.448573064804077,
      "epoch": 5.302128376481159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025740073766070376,
      "loss": 4.8681,
      "mean_token_accuracy": 0.2354705199599266,
      "num_tokens": 105385890.0,
      "step": 47210
    },
    {
      "entropy": 5.504185628890991,
      "epoch": 5.302689953389117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002573607122295249,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2211478739976883,
      "num_tokens": 105396943.0,
      "step": 47215
    },
    {
      "entropy": 5.542298746109009,
      "epoch": 5.303251530297074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002573206873599878,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22903476506471634,
      "num_tokens": 105407223.0,
      "step": 47220
    },
    {
      "entropy": 5.545210981369019,
      "epoch": 5.303813107205031,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025728066305336685,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22682155966758727,
      "num_tokens": 105417714.0,
      "step": 47225
    },
    {
      "entropy": 5.635655498504638,
      "epoch": 5.3043746841129895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002572406393109364,
      "loss": 5.0499,
      "mean_token_accuracy": 0.227703095972538,
      "num_tokens": 105429035.0,
      "step": 47230
    },
    {
      "entropy": 5.516148614883423,
      "epoch": 5.304936261020947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002572006161339709,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22556700855493544,
      "num_tokens": 105440347.0,
      "step": 47235
    },
    {
      "entropy": 5.568880176544189,
      "epoch": 5.305497837928904,
      "grad_norm": 1.125,
      "learning_rate": 0.0002571605935237448,
      "loss": 4.9766,
      "mean_token_accuracy": 0.23009970784187317,
      "num_tokens": 105451265.0,
      "step": 47240
    },
    {
      "entropy": 5.577010583877564,
      "epoch": 5.306059414836862,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002571205714815322,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2195965066552162,
      "num_tokens": 105462034.0,
      "step": 47245
    },
    {
      "entropy": 5.505531263351441,
      "epoch": 5.306620991744819,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002570805500086076,
      "loss": 4.8639,
      "mean_token_accuracy": 0.23436724543571472,
      "num_tokens": 105472733.0,
      "step": 47250
    },
    {
      "entropy": 5.528012323379516,
      "epoch": 5.307182568652777,
      "grad_norm": 1.125,
      "learning_rate": 0.0002570405291062452,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22361893951892853,
      "num_tokens": 105483152.0,
      "step": 47255
    },
    {
      "entropy": 5.405060195922852,
      "epoch": 5.307744145560735,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002570005087757193,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23654100596904754,
      "num_tokens": 105493541.0,
      "step": 47260
    },
    {
      "entropy": 5.4702184200286865,
      "epoch": 5.308305722468692,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002569604890183043,
      "loss": 4.9045,
      "mean_token_accuracy": 0.23051210939884187,
      "num_tokens": 105505550.0,
      "step": 47265
    },
    {
      "entropy": 5.435751676559448,
      "epoch": 5.30886729937665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002569204698352741,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22896383106708526,
      "num_tokens": 105518460.0,
      "step": 47270
    },
    {
      "entropy": 5.456562852859497,
      "epoch": 5.309428876284607,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002568804512279033,
      "loss": 4.8733,
      "mean_token_accuracy": 0.23586559742689134,
      "num_tokens": 105529074.0,
      "step": 47275
    },
    {
      "entropy": 5.496738243103027,
      "epoch": 5.309990453192564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025684043319746595,
      "loss": 4.8989,
      "mean_token_accuracy": 0.23132887184619905,
      "num_tokens": 105540501.0,
      "step": 47280
    },
    {
      "entropy": 5.477481317520142,
      "epoch": 5.3105520301005225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025680041574523626,
      "loss": 4.8584,
      "mean_token_accuracy": 0.23028894811868666,
      "num_tokens": 105551158.0,
      "step": 47285
    },
    {
      "entropy": 5.486521530151367,
      "epoch": 5.31111360700848,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002567603988724884,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23795565515756606,
      "num_tokens": 105563219.0,
      "step": 47290
    },
    {
      "entropy": 5.510918998718262,
      "epoch": 5.311675183916437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002567203825804966,
      "loss": 4.9186,
      "mean_token_accuracy": 0.23125733733177184,
      "num_tokens": 105574846.0,
      "step": 47295
    },
    {
      "entropy": 5.4328478336334225,
      "epoch": 5.312236760824395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002566803668705347,
      "loss": 4.8563,
      "mean_token_accuracy": 0.23633792996406555,
      "num_tokens": 105585848.0,
      "step": 47300
    },
    {
      "entropy": 5.493729829788208,
      "epoch": 5.312798337732352,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002566403517438772,
      "loss": 4.9314,
      "mean_token_accuracy": 0.23041922450065613,
      "num_tokens": 105596757.0,
      "step": 47305
    },
    {
      "entropy": 5.508331966400147,
      "epoch": 5.31335991464031,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025660033720179795,
      "loss": 5.0486,
      "mean_token_accuracy": 0.2205788552761078,
      "num_tokens": 105609363.0,
      "step": 47310
    },
    {
      "entropy": 5.469094848632812,
      "epoch": 5.3139214915482675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025656032324557107,
      "loss": 4.915,
      "mean_token_accuracy": 0.23233708292245864,
      "num_tokens": 105620909.0,
      "step": 47315
    },
    {
      "entropy": 5.426692390441895,
      "epoch": 5.314483068456225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025652030987647066,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23626600354909896,
      "num_tokens": 105631134.0,
      "step": 47320
    },
    {
      "entropy": 5.416753578186035,
      "epoch": 5.315044645364183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002564802970957708,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23664402663707734,
      "num_tokens": 105641979.0,
      "step": 47325
    },
    {
      "entropy": 5.470412302017212,
      "epoch": 5.31560622227214,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025644028490474544,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22902540415525435,
      "num_tokens": 105652320.0,
      "step": 47330
    },
    {
      "entropy": 5.479973220825196,
      "epoch": 5.316167799180098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025640027330466866,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2232399895787239,
      "num_tokens": 105663278.0,
      "step": 47335
    },
    {
      "entropy": 5.495741081237793,
      "epoch": 5.316729376088055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025636026229681433,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23296345323324202,
      "num_tokens": 105673930.0,
      "step": 47340
    },
    {
      "entropy": 5.441064023971558,
      "epoch": 5.317290952996013,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002563202518824564,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2357548713684082,
      "num_tokens": 105684617.0,
      "step": 47345
    },
    {
      "entropy": 5.445990514755249,
      "epoch": 5.317852529903971,
      "grad_norm": 1.09375,
      "learning_rate": 0.000256280242062869,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2302331358194351,
      "num_tokens": 105695278.0,
      "step": 47350
    },
    {
      "entropy": 5.508570909500122,
      "epoch": 5.318414106811928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002562402328393259,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22394426465034484,
      "num_tokens": 105706026.0,
      "step": 47355
    },
    {
      "entropy": 5.47455735206604,
      "epoch": 5.318975683719885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025620022421310115,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2361288011074066,
      "num_tokens": 105717141.0,
      "step": 47360
    },
    {
      "entropy": 5.507042264938354,
      "epoch": 5.319537260627843,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025616021618546847,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2328542023897171,
      "num_tokens": 105729240.0,
      "step": 47365
    },
    {
      "entropy": 5.505671405792237,
      "epoch": 5.3200988375358005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025612020875770186,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22664554864168168,
      "num_tokens": 105741344.0,
      "step": 47370
    },
    {
      "entropy": 5.492160892486572,
      "epoch": 5.320660414443758,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002560802019310751,
      "loss": 4.9042,
      "mean_token_accuracy": 0.23073910772800446,
      "num_tokens": 105752750.0,
      "step": 47375
    },
    {
      "entropy": 5.495358848571778,
      "epoch": 5.321221991351716,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025604019570686203,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22557903528213502,
      "num_tokens": 105764677.0,
      "step": 47380
    },
    {
      "entropy": 5.498144149780273,
      "epoch": 5.321783568259673,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002560001900863365,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2260882630944252,
      "num_tokens": 105775649.0,
      "step": 47385
    },
    {
      "entropy": 5.541204977035522,
      "epoch": 5.322345145167631,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002559601850707723,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23077133744955064,
      "num_tokens": 105785605.0,
      "step": 47390
    },
    {
      "entropy": 5.512674188613891,
      "epoch": 5.322906722075588,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002559201806614432,
      "loss": 4.9071,
      "mean_token_accuracy": 0.232770200073719,
      "num_tokens": 105795887.0,
      "step": 47395
    },
    {
      "entropy": 5.4183652877807615,
      "epoch": 5.323468298983546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025588017685962297,
      "loss": 4.8124,
      "mean_token_accuracy": 0.2356306403875351,
      "num_tokens": 105806129.0,
      "step": 47400
    },
    {
      "entropy": 5.525734615325928,
      "epoch": 5.324029875891504,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025584017366658525,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2283807873725891,
      "num_tokens": 105817048.0,
      "step": 47405
    },
    {
      "entropy": 5.501811218261719,
      "epoch": 5.324591452799461,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025580017108360386,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2281448170542717,
      "num_tokens": 105829847.0,
      "step": 47410
    },
    {
      "entropy": 5.4985826969146725,
      "epoch": 5.325153029707418,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002557601691119524,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23134850859642028,
      "num_tokens": 105839797.0,
      "step": 47415
    },
    {
      "entropy": 5.529952955245972,
      "epoch": 5.325714606615376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025572016775290475,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22361051738262178,
      "num_tokens": 105851004.0,
      "step": 47420
    },
    {
      "entropy": 5.566834115982056,
      "epoch": 5.326276183523333,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025568016700773435,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22217485904693604,
      "num_tokens": 105861558.0,
      "step": 47425
    },
    {
      "entropy": 5.484556055068969,
      "epoch": 5.326837760431291,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025564016687771495,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23947307616472244,
      "num_tokens": 105872705.0,
      "step": 47430
    },
    {
      "entropy": 5.447583389282227,
      "epoch": 5.327399337339249,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025560016736412015,
      "loss": 4.941,
      "mean_token_accuracy": 0.22464296817779542,
      "num_tokens": 105884426.0,
      "step": 47435
    },
    {
      "entropy": 5.458125448226928,
      "epoch": 5.327960914247206,
      "grad_norm": 1.125,
      "learning_rate": 0.00025556016846822353,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22849782407283784,
      "num_tokens": 105896141.0,
      "step": 47440
    },
    {
      "entropy": 5.472576761245728,
      "epoch": 5.328522491155164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025552017019129874,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23690968453884126,
      "num_tokens": 105907216.0,
      "step": 47445
    },
    {
      "entropy": 5.571914768218994,
      "epoch": 5.329084068063121,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025548017253461923,
      "loss": 4.9795,
      "mean_token_accuracy": 0.23501780033111572,
      "num_tokens": 105918011.0,
      "step": 47450
    },
    {
      "entropy": 5.59455041885376,
      "epoch": 5.3296456449710785,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002554401754994586,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22343018054962158,
      "num_tokens": 105929173.0,
      "step": 47455
    },
    {
      "entropy": 5.503238153457642,
      "epoch": 5.330207221879037,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002554001790870904,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22969183325767517,
      "num_tokens": 105940825.0,
      "step": 47460
    },
    {
      "entropy": 5.453517484664917,
      "epoch": 5.330768798786994,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025536018329878814,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23374899476766586,
      "num_tokens": 105951335.0,
      "step": 47465
    },
    {
      "entropy": 5.508804607391357,
      "epoch": 5.331330375694951,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002553201881358252,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21906238943338394,
      "num_tokens": 105963529.0,
      "step": 47470
    },
    {
      "entropy": 5.613299179077148,
      "epoch": 5.331891952602909,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025528019359947514,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21983711868524553,
      "num_tokens": 105975386.0,
      "step": 47475
    },
    {
      "entropy": 5.463528633117676,
      "epoch": 5.332453529510866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002552401996910113,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23058575987815857,
      "num_tokens": 105985710.0,
      "step": 47480
    },
    {
      "entropy": 5.406627368927002,
      "epoch": 5.333015106418824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002552002064117072,
      "loss": 4.8085,
      "mean_token_accuracy": 0.2373457968235016,
      "num_tokens": 105997216.0,
      "step": 47485
    },
    {
      "entropy": 5.498259973526001,
      "epoch": 5.333576683326782,
      "grad_norm": 1.125,
      "learning_rate": 0.0002551602137628361,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23405273556709288,
      "num_tokens": 106008111.0,
      "step": 47490
    },
    {
      "entropy": 5.494059276580811,
      "epoch": 5.334138260234739,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025512022174567164,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2380556970834732,
      "num_tokens": 106018732.0,
      "step": 47495
    },
    {
      "entropy": 5.448333501815796,
      "epoch": 5.334699837142697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025508023036148697,
      "loss": 4.834,
      "mean_token_accuracy": 0.23093049079179764,
      "num_tokens": 106029740.0,
      "step": 47500
    },
    {
      "entropy": 5.473051643371582,
      "epoch": 5.335261414050654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025504023961155553,
      "loss": 4.8293,
      "mean_token_accuracy": 0.24207490384578706,
      "num_tokens": 106040553.0,
      "step": 47505
    },
    {
      "entropy": 5.504683876037598,
      "epoch": 5.3358229909586115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025500024949715055,
      "loss": 4.931,
      "mean_token_accuracy": 0.23054544925689696,
      "num_tokens": 106051667.0,
      "step": 47510
    },
    {
      "entropy": 5.5408001899719235,
      "epoch": 5.33638456786657,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025496026001954543,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2304633617401123,
      "num_tokens": 106062492.0,
      "step": 47515
    },
    {
      "entropy": 5.555714511871338,
      "epoch": 5.336946144774527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025492027118001345,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22832075506448746,
      "num_tokens": 106072645.0,
      "step": 47520
    },
    {
      "entropy": 5.466731071472168,
      "epoch": 5.337507721682485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002548802829798277,
      "loss": 4.8262,
      "mean_token_accuracy": 0.24047624170780182,
      "num_tokens": 106084115.0,
      "step": 47525
    },
    {
      "entropy": 5.461771869659424,
      "epoch": 5.338069298590442,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002548402954202616,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22597669661045075,
      "num_tokens": 106094093.0,
      "step": 47530
    },
    {
      "entropy": 5.465782690048218,
      "epoch": 5.338630875498399,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002548003085025883,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23610792756080629,
      "num_tokens": 106104452.0,
      "step": 47535
    },
    {
      "entropy": 5.480342769622803,
      "epoch": 5.339192452406357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025476032222808106,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23226362019777297,
      "num_tokens": 106114540.0,
      "step": 47540
    },
    {
      "entropy": 5.516150760650635,
      "epoch": 5.339754029314315,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025472033659801297,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22871264666318894,
      "num_tokens": 106124981.0,
      "step": 47545
    },
    {
      "entropy": 5.477128505706787,
      "epoch": 5.340315606222272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002546803516136572,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22226879745721817,
      "num_tokens": 106135037.0,
      "step": 47550
    },
    {
      "entropy": 5.533163452148438,
      "epoch": 5.34087718313023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025464036727628695,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22818878293037415,
      "num_tokens": 106145515.0,
      "step": 47555
    },
    {
      "entropy": 5.523593997955322,
      "epoch": 5.341438760038187,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025460038358717527,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22693444937467575,
      "num_tokens": 106157246.0,
      "step": 47560
    },
    {
      "entropy": 5.501403760910034,
      "epoch": 5.342000336946144,
      "grad_norm": 1.125,
      "learning_rate": 0.0002545604005475953,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22550505846738816,
      "num_tokens": 106167889.0,
      "step": 47565
    },
    {
      "entropy": 5.465127277374267,
      "epoch": 5.3425619138541025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002545204181588201,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2375788927078247,
      "num_tokens": 106178330.0,
      "step": 47570
    },
    {
      "entropy": 5.468624258041382,
      "epoch": 5.34312349076206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025448043642212274,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22792986035346985,
      "num_tokens": 106188995.0,
      "step": 47575
    },
    {
      "entropy": 5.485120868682861,
      "epoch": 5.343685067670018,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002544404553387763,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2308419942855835,
      "num_tokens": 106199271.0,
      "step": 47580
    },
    {
      "entropy": 5.455610847473144,
      "epoch": 5.344246644577975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025440047491005367,
      "loss": 4.904,
      "mean_token_accuracy": 0.23004486709833144,
      "num_tokens": 106211495.0,
      "step": 47585
    },
    {
      "entropy": 5.4397053718566895,
      "epoch": 5.344808221485932,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025436049513722795,
      "loss": 4.8598,
      "mean_token_accuracy": 0.23079854547977446,
      "num_tokens": 106223437.0,
      "step": 47590
    },
    {
      "entropy": 5.500260591506958,
      "epoch": 5.34536979839389,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025432051602157203,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22249676734209062,
      "num_tokens": 106233456.0,
      "step": 47595
    },
    {
      "entropy": 5.484407138824463,
      "epoch": 5.345931375301848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002542805375643589,
      "loss": 4.8125,
      "mean_token_accuracy": 0.2350120037794113,
      "num_tokens": 106244559.0,
      "step": 47600
    },
    {
      "entropy": 5.426497983932495,
      "epoch": 5.346492952209805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002542405597668615,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23428538590669631,
      "num_tokens": 106254427.0,
      "step": 47605
    },
    {
      "entropy": 5.4553139209747314,
      "epoch": 5.347054529117763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025420058263035276,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23177750259637833,
      "num_tokens": 106264616.0,
      "step": 47610
    },
    {
      "entropy": 5.470527267456054,
      "epoch": 5.34761610602572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002541606061561055,
      "loss": 4.861,
      "mean_token_accuracy": 0.23223581314086914,
      "num_tokens": 106275103.0,
      "step": 47615
    },
    {
      "entropy": 5.527747297286988,
      "epoch": 5.348177682933677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002541206303453927,
      "loss": 4.993,
      "mean_token_accuracy": 0.22961623221635818,
      "num_tokens": 106286004.0,
      "step": 47620
    },
    {
      "entropy": 5.5298285484313965,
      "epoch": 5.3487392598416355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000254080655199487,
      "loss": 4.9165,
      "mean_token_accuracy": 0.23070854395627977,
      "num_tokens": 106297161.0,
      "step": 47625
    },
    {
      "entropy": 5.549318456649781,
      "epoch": 5.349300836749593,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025404068071966133,
      "loss": 4.9916,
      "mean_token_accuracy": 0.22259076982736586,
      "num_tokens": 106309330.0,
      "step": 47630
    },
    {
      "entropy": 5.522893810272217,
      "epoch": 5.349862413657551,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025400070690718865,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22463311702013017,
      "num_tokens": 106319801.0,
      "step": 47635
    },
    {
      "entropy": 5.491478395462036,
      "epoch": 5.350423990565508,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025396073376334155,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22113758474588394,
      "num_tokens": 106331328.0,
      "step": 47640
    },
    {
      "entropy": 5.506081676483154,
      "epoch": 5.350985567473465,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025392076128939283,
      "loss": 4.9001,
      "mean_token_accuracy": 0.2320844739675522,
      "num_tokens": 106343740.0,
      "step": 47645
    },
    {
      "entropy": 5.4536723613739015,
      "epoch": 5.351547144381423,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002538807894866153,
      "loss": 4.8534,
      "mean_token_accuracy": 0.24348312765359878,
      "num_tokens": 106354620.0,
      "step": 47650
    },
    {
      "entropy": 5.430413770675659,
      "epoch": 5.3521087212893805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025384081835628163,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23244879245758057,
      "num_tokens": 106366247.0,
      "step": 47655
    },
    {
      "entropy": 5.488042640686035,
      "epoch": 5.352670298197338,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002538008478996645,
      "loss": 4.925,
      "mean_token_accuracy": 0.22918449640274047,
      "num_tokens": 106377045.0,
      "step": 47660
    },
    {
      "entropy": 5.447242975234985,
      "epoch": 5.353231875105296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002537608781180366,
      "loss": 4.8629,
      "mean_token_accuracy": 0.23150699883699416,
      "num_tokens": 106389070.0,
      "step": 47665
    },
    {
      "entropy": 5.4500627517700195,
      "epoch": 5.353793452013253,
      "grad_norm": 1.0,
      "learning_rate": 0.0002537209090126704,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23206169456243514,
      "num_tokens": 106401292.0,
      "step": 47670
    },
    {
      "entropy": 5.500263357162476,
      "epoch": 5.35435502892121,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00025368094058483885,
      "loss": 4.8786,
      "mean_token_accuracy": 0.23017141073942185,
      "num_tokens": 106413045.0,
      "step": 47675
    },
    {
      "entropy": 5.487235355377197,
      "epoch": 5.354916605829168,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002536409728358145,
      "loss": 4.8546,
      "mean_token_accuracy": 0.23145971745252608,
      "num_tokens": 106424655.0,
      "step": 47680
    },
    {
      "entropy": 5.536852645874023,
      "epoch": 5.355478182737126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002536010057668698,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2174648180603981,
      "num_tokens": 106436544.0,
      "step": 47685
    },
    {
      "entropy": 5.4961169242858885,
      "epoch": 5.356039759645084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002535610393792773,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2281959354877472,
      "num_tokens": 106446936.0,
      "step": 47690
    },
    {
      "entropy": 5.532784175872803,
      "epoch": 5.356601336553041,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025352107367430975,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22819947600364685,
      "num_tokens": 106457328.0,
      "step": 47695
    },
    {
      "entropy": 5.463089323043823,
      "epoch": 5.357162913460998,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025348110865323946,
      "loss": 4.9133,
      "mean_token_accuracy": 0.23361240178346634,
      "num_tokens": 106468037.0,
      "step": 47700
    },
    {
      "entropy": 5.503396987915039,
      "epoch": 5.357724490368956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002534411443173389,
      "loss": 4.9011,
      "mean_token_accuracy": 0.23293057531118394,
      "num_tokens": 106479406.0,
      "step": 47705
    },
    {
      "entropy": 5.37979211807251,
      "epoch": 5.3582860672769135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002534011806678809,
      "loss": 4.7912,
      "mean_token_accuracy": 0.2278803750872612,
      "num_tokens": 106491142.0,
      "step": 47710
    },
    {
      "entropy": 5.463598823547363,
      "epoch": 5.358847644184872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002533612177061376,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22881633639335633,
      "num_tokens": 106502743.0,
      "step": 47715
    },
    {
      "entropy": 5.496118879318237,
      "epoch": 5.359409221092829,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00025332125543338146,
      "loss": 4.9121,
      "mean_token_accuracy": 0.23209248483181,
      "num_tokens": 106513765.0,
      "step": 47720
    },
    {
      "entropy": 5.417875623703003,
      "epoch": 5.359970798000786,
      "grad_norm": 1.140625,
      "learning_rate": 0.000253281293850885,
      "loss": 4.7611,
      "mean_token_accuracy": 0.24136034846305848,
      "num_tokens": 106524154.0,
      "step": 47725
    },
    {
      "entropy": 5.465169286727905,
      "epoch": 5.360532374908744,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00025324133295992056,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2291560336947441,
      "num_tokens": 106535946.0,
      "step": 47730
    },
    {
      "entropy": 5.576036596298218,
      "epoch": 5.361093951816701,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002532013727617605,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2178375631570816,
      "num_tokens": 106546352.0,
      "step": 47735
    },
    {
      "entropy": 5.479981374740601,
      "epoch": 5.361655528724659,
      "grad_norm": 1.125,
      "learning_rate": 0.0002531614132576772,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23627692610025405,
      "num_tokens": 106557640.0,
      "step": 47740
    },
    {
      "entropy": 5.491077947616577,
      "epoch": 5.362217105632617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000253121454448943,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22352838814258574,
      "num_tokens": 106567762.0,
      "step": 47745
    },
    {
      "entropy": 5.476313352584839,
      "epoch": 5.362778682540574,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025308149633683014,
      "loss": 4.8807,
      "mean_token_accuracy": 0.24005515724420548,
      "num_tokens": 106579589.0,
      "step": 47750
    },
    {
      "entropy": 5.462822580337525,
      "epoch": 5.363340259448531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000253041538922611,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22768848836421968,
      "num_tokens": 106590122.0,
      "step": 47755
    },
    {
      "entropy": 5.524800872802734,
      "epoch": 5.363901836356489,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025300158220755776,
      "loss": 4.976,
      "mean_token_accuracy": 0.2277942642569542,
      "num_tokens": 106601269.0,
      "step": 47760
    },
    {
      "entropy": 5.577778387069702,
      "epoch": 5.364463413264446,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025296162619294267,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21923573464155197,
      "num_tokens": 106612462.0,
      "step": 47765
    },
    {
      "entropy": 5.534153270721435,
      "epoch": 5.3650249901724045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002529216708800379,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23313615024089812,
      "num_tokens": 106623333.0,
      "step": 47770
    },
    {
      "entropy": 5.4849793910980225,
      "epoch": 5.365586567080362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025288171627011564,
      "loss": 4.8419,
      "mean_token_accuracy": 0.2353183001279831,
      "num_tokens": 106634505.0,
      "step": 47775
    },
    {
      "entropy": 5.565237998962402,
      "epoch": 5.366148143988319,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025284176236444817,
      "loss": 5.0027,
      "mean_token_accuracy": 0.22506024241447448,
      "num_tokens": 106645764.0,
      "step": 47780
    },
    {
      "entropy": 5.411687994003296,
      "epoch": 5.366709720896277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025280180916430764,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23510556817054748,
      "num_tokens": 106657484.0,
      "step": 47785
    },
    {
      "entropy": 5.423128414154053,
      "epoch": 5.367271297804234,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025276185667096605,
      "loss": 4.8518,
      "mean_token_accuracy": 0.23208666890859603,
      "num_tokens": 106669038.0,
      "step": 47790
    },
    {
      "entropy": 5.532661008834839,
      "epoch": 5.3678328747121915,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025272190488569564,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23534328639507293,
      "num_tokens": 106679447.0,
      "step": 47795
    },
    {
      "entropy": 5.492126178741455,
      "epoch": 5.36839445162015,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025268195380976835,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22988718301057814,
      "num_tokens": 106689596.0,
      "step": 47800
    },
    {
      "entropy": 5.495362091064453,
      "epoch": 5.368956028528107,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002526420034444564,
      "loss": 4.8823,
      "mean_token_accuracy": 0.24032578766345977,
      "num_tokens": 106701825.0,
      "step": 47805
    },
    {
      "entropy": 5.585808420181275,
      "epoch": 5.369517605436064,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025260205379103157,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2156889706850052,
      "num_tokens": 106713451.0,
      "step": 47810
    },
    {
      "entropy": 5.580071973800659,
      "epoch": 5.370079182344022,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025256210485076616,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2217390328645706,
      "num_tokens": 106724732.0,
      "step": 47815
    },
    {
      "entropy": 5.573336410522461,
      "epoch": 5.370640759251979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000252522156624932,
      "loss": 4.9486,
      "mean_token_accuracy": 0.23318471908569335,
      "num_tokens": 106736533.0,
      "step": 47820
    },
    {
      "entropy": 5.453620195388794,
      "epoch": 5.3712023361599375,
      "grad_norm": 0.96875,
      "learning_rate": 0.00025248220911480105,
      "loss": 4.841,
      "mean_token_accuracy": 0.23660558313131333,
      "num_tokens": 106747454.0,
      "step": 47825
    },
    {
      "entropy": 5.404984426498413,
      "epoch": 5.371763913067895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025244226232164533,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23693080693483354,
      "num_tokens": 106757833.0,
      "step": 47830
    },
    {
      "entropy": 5.377712345123291,
      "epoch": 5.372325489975852,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025240231624673674,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22868526428937913,
      "num_tokens": 106768646.0,
      "step": 47835
    },
    {
      "entropy": 5.463355302810669,
      "epoch": 5.37288706688381,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002523623708913472,
      "loss": 4.9061,
      "mean_token_accuracy": 0.23181371688842772,
      "num_tokens": 106780434.0,
      "step": 47840
    },
    {
      "entropy": 5.493573427200317,
      "epoch": 5.373448643791767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002523224262567485,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23653798252344133,
      "num_tokens": 106792372.0,
      "step": 47845
    },
    {
      "entropy": 5.531528520584106,
      "epoch": 5.3740102206997244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025228248234421247,
      "loss": 4.8824,
      "mean_token_accuracy": 0.21821400225162507,
      "num_tokens": 106802251.0,
      "step": 47850
    },
    {
      "entropy": 5.489517879486084,
      "epoch": 5.3745717976076826,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002522425391550111,
      "loss": 4.9399,
      "mean_token_accuracy": 0.23089901506900787,
      "num_tokens": 106813818.0,
      "step": 47855
    },
    {
      "entropy": 5.549230480194092,
      "epoch": 5.37513337451564,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025220259669041613,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21911114007234572,
      "num_tokens": 106824096.0,
      "step": 47860
    },
    {
      "entropy": 5.588487672805786,
      "epoch": 5.375694951423597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002521626549516993,
      "loss": 5.0546,
      "mean_token_accuracy": 0.22138909101486207,
      "num_tokens": 106836037.0,
      "step": 47865
    },
    {
      "entropy": 5.569146156311035,
      "epoch": 5.376256528331555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002521227139401324,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2248838409781456,
      "num_tokens": 106846370.0,
      "step": 47870
    },
    {
      "entropy": 5.536303997039795,
      "epoch": 5.376818105239512,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00025208277365698715,
      "loss": 4.9959,
      "mean_token_accuracy": 0.22317198365926744,
      "num_tokens": 106857945.0,
      "step": 47875
    },
    {
      "entropy": 5.4957784652709964,
      "epoch": 5.37737968214747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002520428341035353,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23581562340259551,
      "num_tokens": 106867899.0,
      "step": 47880
    },
    {
      "entropy": 5.476692819595337,
      "epoch": 5.377941259055428,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002520028952810485,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22517258822917938,
      "num_tokens": 106879305.0,
      "step": 47885
    },
    {
      "entropy": 5.546954488754272,
      "epoch": 5.378502835963385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025196295719079835,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22664524018764495,
      "num_tokens": 106891944.0,
      "step": 47890
    },
    {
      "entropy": 5.486609554290771,
      "epoch": 5.379064412871343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002519230198340567,
      "loss": 4.8973,
      "mean_token_accuracy": 0.23221572488546371,
      "num_tokens": 106903420.0,
      "step": 47895
    },
    {
      "entropy": 5.514862728118897,
      "epoch": 5.3796259897793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025188308321209504,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23120735287666322,
      "num_tokens": 106914838.0,
      "step": 47900
    },
    {
      "entropy": 5.386690378189087,
      "epoch": 5.380187566687258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025184314732618496,
      "loss": 4.8132,
      "mean_token_accuracy": 0.23580189943313598,
      "num_tokens": 106926519.0,
      "step": 47905
    },
    {
      "entropy": 5.49598650932312,
      "epoch": 5.3807491435952155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000251803212177598,
      "loss": 4.9455,
      "mean_token_accuracy": 0.23219493180513381,
      "num_tokens": 106938303.0,
      "step": 47910
    },
    {
      "entropy": 5.512967681884765,
      "epoch": 5.381310720503173,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002517632777676058,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2322253555059433,
      "num_tokens": 106950120.0,
      "step": 47915
    },
    {
      "entropy": 5.511247539520264,
      "epoch": 5.381872297411131,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025172334409747976,
      "loss": 4.942,
      "mean_token_accuracy": 0.2232983738183975,
      "num_tokens": 106960998.0,
      "step": 47920
    },
    {
      "entropy": 5.483645105361939,
      "epoch": 5.382433874319088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002516834111684915,
      "loss": 4.892,
      "mean_token_accuracy": 0.22890195846557618,
      "num_tokens": 106972640.0,
      "step": 47925
    },
    {
      "entropy": 5.46949577331543,
      "epoch": 5.382995451227045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002516434789819126,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22962397933006287,
      "num_tokens": 106985300.0,
      "step": 47930
    },
    {
      "entropy": 5.527017641067505,
      "epoch": 5.383557028135003,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002516035475390143,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2197425588965416,
      "num_tokens": 106996074.0,
      "step": 47935
    },
    {
      "entropy": 5.5518945217132565,
      "epoch": 5.384118605042961,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025156361684106814,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2336969032883644,
      "num_tokens": 107006493.0,
      "step": 47940
    },
    {
      "entropy": 5.477571439743042,
      "epoch": 5.384680181950918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002515236868893455,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22731538712978364,
      "num_tokens": 107018739.0,
      "step": 47945
    },
    {
      "entropy": 5.425822401046753,
      "epoch": 5.385241758858876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025148375768511776,
      "loss": 4.7801,
      "mean_token_accuracy": 0.24172314256429672,
      "num_tokens": 107028762.0,
      "step": 47950
    },
    {
      "entropy": 5.5032164573669435,
      "epoch": 5.385803335766833,
      "grad_norm": 1.125,
      "learning_rate": 0.0002514438292296563,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22583913654088975,
      "num_tokens": 107039609.0,
      "step": 47955
    },
    {
      "entropy": 5.4079827785491945,
      "epoch": 5.386364912674791,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025140390152423247,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23872132301330568,
      "num_tokens": 107051871.0,
      "step": 47960
    },
    {
      "entropy": 5.46141791343689,
      "epoch": 5.3869264895827484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002513639745701176,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22724697291851043,
      "num_tokens": 107063560.0,
      "step": 47965
    },
    {
      "entropy": 5.520855188369751,
      "epoch": 5.387488066490706,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025132404836858294,
      "loss": 4.9292,
      "mean_token_accuracy": 0.23207831978797913,
      "num_tokens": 107074466.0,
      "step": 47970
    },
    {
      "entropy": 5.504336595535278,
      "epoch": 5.388049643398664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002512841229208997,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23885941803455352,
      "num_tokens": 107085470.0,
      "step": 47975
    },
    {
      "entropy": 5.513006067276001,
      "epoch": 5.388611220306621,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002512441982283392,
      "loss": 4.9495,
      "mean_token_accuracy": 0.23669832795858384,
      "num_tokens": 107098179.0,
      "step": 47980
    },
    {
      "entropy": 5.456262540817261,
      "epoch": 5.389172797214578,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002512042742921727,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2322048157453537,
      "num_tokens": 107108941.0,
      "step": 47985
    },
    {
      "entropy": 5.515744972229004,
      "epoch": 5.389734374122536,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002511643511136712,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22577069848775863,
      "num_tokens": 107120019.0,
      "step": 47990
    },
    {
      "entropy": 5.5167584896087645,
      "epoch": 5.3902959510304935,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025112442869410607,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2261642560362816,
      "num_tokens": 107130254.0,
      "step": 47995
    },
    {
      "entropy": 5.453172397613526,
      "epoch": 5.390857527938451,
      "grad_norm": 1.0,
      "learning_rate": 0.00025108450703474844,
      "loss": 4.8111,
      "mean_token_accuracy": 0.24059183299541473,
      "num_tokens": 107140625.0,
      "step": 48000
    },
    {
      "epoch": 5.390857527938451,
      "eval_entropy": 5.249294281005859,
      "eval_loss": 5.105429649353027,
      "eval_mean_token_accuracy": 0.2256936084153359,
      "eval_num_tokens": 107140625.0,
      "eval_runtime": 53.5917,
      "eval_samples_per_second": 578.691,
      "eval_steps_per_second": 72.343,
      "step": 48000
    },
    {
      "entropy": 5.480610990524292,
      "epoch": 5.391419104846409,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00025104458613686936,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22482458204030992,
      "num_tokens": 107150318.0,
      "step": 48005
    },
    {
      "entropy": 5.409901142120361,
      "epoch": 5.391980681754366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00025100466600174,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22616744488477708,
      "num_tokens": 107162831.0,
      "step": 48010
    },
    {
      "entropy": 5.558383369445801,
      "epoch": 5.392542258662324,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002509647466306313,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22813290357589722,
      "num_tokens": 107174775.0,
      "step": 48015
    },
    {
      "entropy": 5.5373907566070555,
      "epoch": 5.393103835570281,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025092482802481443,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22275751680135727,
      "num_tokens": 107185644.0,
      "step": 48020
    },
    {
      "entropy": 5.494850540161133,
      "epoch": 5.393665412478239,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002508849101855604,
      "loss": 4.9052,
      "mean_token_accuracy": 0.23253665268421173,
      "num_tokens": 107196669.0,
      "step": 48025
    },
    {
      "entropy": 5.518252897262573,
      "epoch": 5.394226989386197,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002508449931141401,
      "loss": 4.8995,
      "mean_token_accuracy": 0.24154942184686662,
      "num_tokens": 107207625.0,
      "step": 48030
    },
    {
      "entropy": 5.556953287124633,
      "epoch": 5.394788566294154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025080507681182464,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22663915008306504,
      "num_tokens": 107218925.0,
      "step": 48035
    },
    {
      "entropy": 5.510553646087646,
      "epoch": 5.395350143202111,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002507651612798849,
      "loss": 4.9243,
      "mean_token_accuracy": 0.23432310670614243,
      "num_tokens": 107231669.0,
      "step": 48040
    },
    {
      "entropy": 5.4442047595977785,
      "epoch": 5.395911720110069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025072524651959183,
      "loss": 4.8932,
      "mean_token_accuracy": 0.22792393267154692,
      "num_tokens": 107242922.0,
      "step": 48045
    },
    {
      "entropy": 5.508061122894287,
      "epoch": 5.3964732970180265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002506853325322164,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22403692901134492,
      "num_tokens": 107254713.0,
      "step": 48050
    },
    {
      "entropy": 5.550611686706543,
      "epoch": 5.397034873925985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002506454193190293,
      "loss": 4.9315,
      "mean_token_accuracy": 0.2291585922241211,
      "num_tokens": 107266202.0,
      "step": 48055
    },
    {
      "entropy": 5.506698036193848,
      "epoch": 5.397596450833942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025060550688130157,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22555219531059265,
      "num_tokens": 107277914.0,
      "step": 48060
    },
    {
      "entropy": 5.530933856964111,
      "epoch": 5.398158027741899,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025056559522030383,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2244277909398079,
      "num_tokens": 107289592.0,
      "step": 48065
    },
    {
      "entropy": 5.407431173324585,
      "epoch": 5.398719604649857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002505256843373071,
      "loss": 4.8168,
      "mean_token_accuracy": 0.2363855391740799,
      "num_tokens": 107301428.0,
      "step": 48070
    },
    {
      "entropy": 5.448621559143066,
      "epoch": 5.399281181557814,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002504857742335821,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22859592884778976,
      "num_tokens": 107313651.0,
      "step": 48075
    },
    {
      "entropy": 5.5377795696258545,
      "epoch": 5.3998427584657716,
      "grad_norm": 1.125,
      "learning_rate": 0.00025044586491039956,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22161124795675277,
      "num_tokens": 107324375.0,
      "step": 48080
    },
    {
      "entropy": 5.48480634689331,
      "epoch": 5.40040433537373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002504059563690302,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23232621550559998,
      "num_tokens": 107335445.0,
      "step": 48085
    },
    {
      "entropy": 5.475756216049194,
      "epoch": 5.400965912281687,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002503660486107447,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2317176416516304,
      "num_tokens": 107346384.0,
      "step": 48090
    },
    {
      "entropy": 5.4503637790679935,
      "epoch": 5.401527489189645,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002503261416368138,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22991699874401092,
      "num_tokens": 107356188.0,
      "step": 48095
    },
    {
      "entropy": 5.450072479248047,
      "epoch": 5.402089066097602,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002502862354485081,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23822680711746216,
      "num_tokens": 107366370.0,
      "step": 48100
    },
    {
      "entropy": 5.4943774223327635,
      "epoch": 5.402650643005559,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025024633004709824,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22749283462762832,
      "num_tokens": 107376975.0,
      "step": 48105
    },
    {
      "entropy": 5.502799606323242,
      "epoch": 5.4032122199135175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025020642543385486,
      "loss": 5.0003,
      "mean_token_accuracy": 0.22007739990949632,
      "num_tokens": 107388980.0,
      "step": 48110
    },
    {
      "entropy": 5.53503007888794,
      "epoch": 5.403773796821475,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025016652161004847,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2256879910826683,
      "num_tokens": 107399299.0,
      "step": 48115
    },
    {
      "entropy": 5.457827615737915,
      "epoch": 5.404335373729432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002501266185769497,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22901976853609085,
      "num_tokens": 107410438.0,
      "step": 48120
    },
    {
      "entropy": 5.495449066162109,
      "epoch": 5.40489695063739,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025008671633582904,
      "loss": 4.9335,
      "mean_token_accuracy": 0.236519518494606,
      "num_tokens": 107422597.0,
      "step": 48125
    },
    {
      "entropy": 5.5092453956604,
      "epoch": 5.405458527545347,
      "grad_norm": 1.046875,
      "learning_rate": 0.000250046814887957,
      "loss": 4.906,
      "mean_token_accuracy": 0.23640007823705672,
      "num_tokens": 107432762.0,
      "step": 48130
    },
    {
      "entropy": 5.473772621154785,
      "epoch": 5.4060201044533045,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002500069142346039,
      "loss": 4.8734,
      "mean_token_accuracy": 0.22860792875289918,
      "num_tokens": 107443644.0,
      "step": 48135
    },
    {
      "entropy": 5.535027980804443,
      "epoch": 5.406581681361263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002499670143770405,
      "loss": 4.9761,
      "mean_token_accuracy": 0.21882829666137696,
      "num_tokens": 107454752.0,
      "step": 48140
    },
    {
      "entropy": 5.44027624130249,
      "epoch": 5.40714325826922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002499271153165371,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22827683985233307,
      "num_tokens": 107464829.0,
      "step": 48145
    },
    {
      "entropy": 5.592629384994507,
      "epoch": 5.407704835177178,
      "grad_norm": 1.09375,
      "learning_rate": 0.000249887217054364,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22592387050390245,
      "num_tokens": 107475214.0,
      "step": 48150
    },
    {
      "entropy": 5.613207674026489,
      "epoch": 5.408266412085135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002498473195917917,
      "loss": 5.0421,
      "mean_token_accuracy": 0.2223970502614975,
      "num_tokens": 107486693.0,
      "step": 48155
    },
    {
      "entropy": 5.579286813735962,
      "epoch": 5.408827988993092,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002498074229300905,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22609905600547792,
      "num_tokens": 107498347.0,
      "step": 48160
    },
    {
      "entropy": 5.551806449890137,
      "epoch": 5.4093895659010505,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002497675270705307,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22102952152490615,
      "num_tokens": 107507525.0,
      "step": 48165
    },
    {
      "entropy": 5.460461664199829,
      "epoch": 5.409951142809008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002497276320143826,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23126355856657027,
      "num_tokens": 107519196.0,
      "step": 48170
    },
    {
      "entropy": 5.52610821723938,
      "epoch": 5.410512719716965,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024968773776291647,
      "loss": 4.981,
      "mean_token_accuracy": 0.22774393558502198,
      "num_tokens": 107532740.0,
      "step": 48175
    },
    {
      "entropy": 5.45966362953186,
      "epoch": 5.411074296624923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002496478443174026,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23265248388051987,
      "num_tokens": 107544098.0,
      "step": 48180
    },
    {
      "entropy": 5.612999439239502,
      "epoch": 5.41163587353288,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002496079516791113,
      "loss": 5.027,
      "mean_token_accuracy": 0.2279532939195633,
      "num_tokens": 107555560.0,
      "step": 48185
    },
    {
      "entropy": 5.439470529556274,
      "epoch": 5.412197450440837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002495680598493126,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23526695519685745,
      "num_tokens": 107565761.0,
      "step": 48190
    },
    {
      "entropy": 5.469681644439698,
      "epoch": 5.4127590273487955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002495281688292768,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22828537076711655,
      "num_tokens": 107576742.0,
      "step": 48195
    },
    {
      "entropy": 5.524663019180298,
      "epoch": 5.413320604256753,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024948827862027385,
      "loss": 4.8938,
      "mean_token_accuracy": 0.23379677385091782,
      "num_tokens": 107588810.0,
      "step": 48200
    },
    {
      "entropy": 5.52324914932251,
      "epoch": 5.413882181164711,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002494483892235741,
      "loss": 4.9108,
      "mean_token_accuracy": 0.23634403198957443,
      "num_tokens": 107599392.0,
      "step": 48205
    },
    {
      "entropy": 5.548250293731689,
      "epoch": 5.414443758072668,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002494085006404474,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2378223106265068,
      "num_tokens": 107608288.0,
      "step": 48210
    },
    {
      "entropy": 5.522441434860229,
      "epoch": 5.415005334980625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002493686128721641,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22748434990644456,
      "num_tokens": 107620747.0,
      "step": 48215
    },
    {
      "entropy": 5.626024389266968,
      "epoch": 5.415566911888583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024932872591999414,
      "loss": 4.959,
      "mean_token_accuracy": 0.22892976552248,
      "num_tokens": 107632754.0,
      "step": 48220
    },
    {
      "entropy": 5.439723157882691,
      "epoch": 5.416128488796541,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024928883978520743,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23290538489818574,
      "num_tokens": 107643326.0,
      "step": 48225
    },
    {
      "entropy": 5.4669571876525875,
      "epoch": 5.416690065704498,
      "grad_norm": 1.140625,
      "learning_rate": 0.00024924895446907407,
      "loss": 4.8811,
      "mean_token_accuracy": 0.22507030218839646,
      "num_tokens": 107654474.0,
      "step": 48230
    },
    {
      "entropy": 5.395746278762817,
      "epoch": 5.417251642612456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000249209069972864,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23720510303974152,
      "num_tokens": 107665110.0,
      "step": 48235
    },
    {
      "entropy": 5.5247022151947025,
      "epoch": 5.417813219520413,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024916918629784714,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22229215651750564,
      "num_tokens": 107675290.0,
      "step": 48240
    },
    {
      "entropy": 5.4802697658538815,
      "epoch": 5.418374796428371,
      "grad_norm": 1.0,
      "learning_rate": 0.0002491293034452933,
      "loss": 4.8376,
      "mean_token_accuracy": 0.23323350846767427,
      "num_tokens": 107685202.0,
      "step": 48245
    },
    {
      "entropy": 5.49799747467041,
      "epoch": 5.4189363733363285,
      "grad_norm": 1.125,
      "learning_rate": 0.0002490894214164726,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22126352936029434,
      "num_tokens": 107695459.0,
      "step": 48250
    },
    {
      "entropy": 5.458436679840088,
      "epoch": 5.419497950244286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002490495402126548,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23365006893873214,
      "num_tokens": 107706776.0,
      "step": 48255
    },
    {
      "entropy": 5.534139251708984,
      "epoch": 5.420059527152244,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002490096598351096,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23958513587713243,
      "num_tokens": 107717969.0,
      "step": 48260
    },
    {
      "entropy": 5.538957977294922,
      "epoch": 5.420621104060201,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024896978028510696,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2229563593864441,
      "num_tokens": 107727812.0,
      "step": 48265
    },
    {
      "entropy": 5.536340188980103,
      "epoch": 5.421182680968158,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002489299015639167,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23498108685016633,
      "num_tokens": 107737823.0,
      "step": 48270
    },
    {
      "entropy": 5.531022739410401,
      "epoch": 5.421744257876116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002488900236728083,
      "loss": 4.9456,
      "mean_token_accuracy": 0.23478585928678514,
      "num_tokens": 107747835.0,
      "step": 48275
    },
    {
      "entropy": 5.433973503112793,
      "epoch": 5.422305834784074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002488501466130517,
      "loss": 4.8451,
      "mean_token_accuracy": 0.2375443011522293,
      "num_tokens": 107759167.0,
      "step": 48280
    },
    {
      "entropy": 5.4623195171356205,
      "epoch": 5.422867411692032,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024881027038591664,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23801836371421814,
      "num_tokens": 107770130.0,
      "step": 48285
    },
    {
      "entropy": 5.530704164505005,
      "epoch": 5.423428988599989,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024877039499267273,
      "loss": 4.9157,
      "mean_token_accuracy": 0.22722937762737275,
      "num_tokens": 107780923.0,
      "step": 48290
    },
    {
      "entropy": 5.429027509689331,
      "epoch": 5.423990565507946,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002487305204345896,
      "loss": 4.887,
      "mean_token_accuracy": 0.23358372300863267,
      "num_tokens": 107792417.0,
      "step": 48295
    },
    {
      "entropy": 5.540813255310058,
      "epoch": 5.424552142415904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024869064671293687,
      "loss": 4.9614,
      "mean_token_accuracy": 0.22485365867614746,
      "num_tokens": 107802124.0,
      "step": 48300
    },
    {
      "entropy": 5.5201239585876465,
      "epoch": 5.425113719323861,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024865077382898423,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2318449169397354,
      "num_tokens": 107814025.0,
      "step": 48305
    },
    {
      "entropy": 5.519247770309448,
      "epoch": 5.425675296231819,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024861090178400106,
      "loss": 4.988,
      "mean_token_accuracy": 0.22338881045579911,
      "num_tokens": 107825271.0,
      "step": 48310
    },
    {
      "entropy": 5.46104474067688,
      "epoch": 5.426236873139777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024857103057925707,
      "loss": 4.8249,
      "mean_token_accuracy": 0.22794119119644166,
      "num_tokens": 107836419.0,
      "step": 48315
    },
    {
      "entropy": 5.476489496231079,
      "epoch": 5.426798450047734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024853116021602173,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22874287515878677,
      "num_tokens": 107847700.0,
      "step": 48320
    },
    {
      "entropy": 5.435369825363159,
      "epoch": 5.427360026955691,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002484912906955645,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23274308145046235,
      "num_tokens": 107858870.0,
      "step": 48325
    },
    {
      "entropy": 5.618700695037842,
      "epoch": 5.427921603863649,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002484514220191549,
      "loss": 5.1784,
      "mean_token_accuracy": 0.21332009136676788,
      "num_tokens": 107870414.0,
      "step": 48330
    },
    {
      "entropy": 5.545728731155395,
      "epoch": 5.4284831807716065,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002484115541880623,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21828954219818114,
      "num_tokens": 107880798.0,
      "step": 48335
    },
    {
      "entropy": 5.446378946304321,
      "epoch": 5.429044757679565,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024837168720355605,
      "loss": 4.774,
      "mean_token_accuracy": 0.2435053691267967,
      "num_tokens": 107891332.0,
      "step": 48340
    },
    {
      "entropy": 5.485541486740113,
      "epoch": 5.429606334587522,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002483318210669057,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23341669291257858,
      "num_tokens": 107902579.0,
      "step": 48345
    },
    {
      "entropy": 5.474950885772705,
      "epoch": 5.430167911495479,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002482919557793805,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22258930057287216,
      "num_tokens": 107912921.0,
      "step": 48350
    },
    {
      "entropy": 5.4513331890106205,
      "epoch": 5.430729488403437,
      "grad_norm": 1.15625,
      "learning_rate": 0.00024825209134224965,
      "loss": 4.7898,
      "mean_token_accuracy": 0.23764537125825883,
      "num_tokens": 107923767.0,
      "step": 48355
    },
    {
      "entropy": 5.493067741394043,
      "epoch": 5.431291065311394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002482122277567827,
      "loss": 4.989,
      "mean_token_accuracy": 0.22160961031913756,
      "num_tokens": 107935905.0,
      "step": 48360
    },
    {
      "entropy": 5.32234263420105,
      "epoch": 5.431852642219352,
      "grad_norm": 1.25,
      "learning_rate": 0.00024817236502424885,
      "loss": 4.7552,
      "mean_token_accuracy": 0.2552019476890564,
      "num_tokens": 107947235.0,
      "step": 48365
    },
    {
      "entropy": 5.511327075958252,
      "epoch": 5.43241421912731,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002481325031459172,
      "loss": 4.8489,
      "mean_token_accuracy": 0.23512881398200988,
      "num_tokens": 107959323.0,
      "step": 48370
    },
    {
      "entropy": 5.535067510604859,
      "epoch": 5.432975796035267,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024809264212305714,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22362808138132095,
      "num_tokens": 107970790.0,
      "step": 48375
    },
    {
      "entropy": 5.522066736221314,
      "epoch": 5.433537372943224,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002480527819569378,
      "loss": 4.928,
      "mean_token_accuracy": 0.2268450975418091,
      "num_tokens": 107982122.0,
      "step": 48380
    },
    {
      "entropy": 5.4828023433685305,
      "epoch": 5.434098949851182,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002480129226488283,
      "loss": 4.884,
      "mean_token_accuracy": 0.23869821578264236,
      "num_tokens": 107993093.0,
      "step": 48385
    },
    {
      "entropy": 5.457908248901367,
      "epoch": 5.4346605267591395,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024797306419999777,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22752158343791962,
      "num_tokens": 108004805.0,
      "step": 48390
    },
    {
      "entropy": 5.551687288284302,
      "epoch": 5.435222103667098,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002479332066117154,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22301529496908187,
      "num_tokens": 108016515.0,
      "step": 48395
    },
    {
      "entropy": 5.452988481521606,
      "epoch": 5.435783680575055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002478933498852503,
      "loss": 4.865,
      "mean_token_accuracy": 0.23313224464654922,
      "num_tokens": 108027439.0,
      "step": 48400
    },
    {
      "entropy": 5.573643207550049,
      "epoch": 5.436345257483012,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002478534940218714,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23131155222654343,
      "num_tokens": 108038373.0,
      "step": 48405
    },
    {
      "entropy": 5.536352157592773,
      "epoch": 5.43690683439097,
      "grad_norm": 1.15625,
      "learning_rate": 0.00024781363902284776,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2290602758526802,
      "num_tokens": 108048336.0,
      "step": 48410
    },
    {
      "entropy": 5.525331115722656,
      "epoch": 5.437468411298927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002477737848894484,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2283505842089653,
      "num_tokens": 108059189.0,
      "step": 48415
    },
    {
      "entropy": 5.499941396713257,
      "epoch": 5.4380299882068845,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002477339316229423,
      "loss": 4.954,
      "mean_token_accuracy": 0.22781021445989608,
      "num_tokens": 108071579.0,
      "step": 48420
    },
    {
      "entropy": 5.534363079071045,
      "epoch": 5.438591565114843,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002476940792245983,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21832381933927536,
      "num_tokens": 108082709.0,
      "step": 48425
    },
    {
      "entropy": 5.446618556976318,
      "epoch": 5.4391531420228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024765422769568553,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23263368308544158,
      "num_tokens": 108094139.0,
      "step": 48430
    },
    {
      "entropy": 5.474534273147583,
      "epoch": 5.439714718930758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002476143770374727,
      "loss": 4.8746,
      "mean_token_accuracy": 0.23155325055122375,
      "num_tokens": 108105352.0,
      "step": 48435
    },
    {
      "entropy": 5.411159610748291,
      "epoch": 5.440276295838715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024757452725122876,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23011818826198577,
      "num_tokens": 108118313.0,
      "step": 48440
    },
    {
      "entropy": 5.52334566116333,
      "epoch": 5.440837872746672,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002475346783382225,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22389149963855742,
      "num_tokens": 108129002.0,
      "step": 48445
    },
    {
      "entropy": 5.553470230102539,
      "epoch": 5.4413994496546305,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024749483029972276,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22442157566547394,
      "num_tokens": 108141071.0,
      "step": 48450
    },
    {
      "entropy": 5.565377044677734,
      "epoch": 5.441961026562588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002474549831369983,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2289014920592308,
      "num_tokens": 108152923.0,
      "step": 48455
    },
    {
      "entropy": 5.512358808517456,
      "epoch": 5.442522603470545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024741513685131773,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22946113497018814,
      "num_tokens": 108164645.0,
      "step": 48460
    },
    {
      "entropy": 5.45045952796936,
      "epoch": 5.443084180378503,
      "grad_norm": 1.125,
      "learning_rate": 0.00024737529144395,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23924597948789597,
      "num_tokens": 108174547.0,
      "step": 48465
    },
    {
      "entropy": 5.512485694885254,
      "epoch": 5.44364575728646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002473354469161637,
      "loss": 4.9244,
      "mean_token_accuracy": 0.23160437643527984,
      "num_tokens": 108186322.0,
      "step": 48470
    },
    {
      "entropy": 5.477369117736816,
      "epoch": 5.444207334194418,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002472956032692275,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23787690848112106,
      "num_tokens": 108196771.0,
      "step": 48475
    },
    {
      "entropy": 5.519158172607422,
      "epoch": 5.444768911102376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002472557605044101,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22431601732969284,
      "num_tokens": 108208222.0,
      "step": 48480
    },
    {
      "entropy": 5.4008739471435545,
      "epoch": 5.445330488010333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024721591862297994,
      "loss": 4.7885,
      "mean_token_accuracy": 0.24877588003873824,
      "num_tokens": 108218572.0,
      "step": 48485
    },
    {
      "entropy": 5.477782487869263,
      "epoch": 5.445892064918291,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024717607762620575,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2323409855365753,
      "num_tokens": 108230458.0,
      "step": 48490
    },
    {
      "entropy": 5.566490316390992,
      "epoch": 5.446453641826248,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002471362375153559,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2190639927983284,
      "num_tokens": 108243840.0,
      "step": 48495
    },
    {
      "entropy": 5.573229598999023,
      "epoch": 5.447015218734205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024709639829169924,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21587361693382262,
      "num_tokens": 108255390.0,
      "step": 48500
    },
    {
      "entropy": 5.429841423034668,
      "epoch": 5.4475767956421635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024705655995650407,
      "loss": 4.82,
      "mean_token_accuracy": 0.2387775033712387,
      "num_tokens": 108267166.0,
      "step": 48505
    },
    {
      "entropy": 5.521072435379028,
      "epoch": 5.448138372550121,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002470167225110388,
      "loss": 4.9476,
      "mean_token_accuracy": 0.22815096080303193,
      "num_tokens": 108278353.0,
      "step": 48510
    },
    {
      "entropy": 5.544791221618652,
      "epoch": 5.448699949458078,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000246976885956572,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2311861425638199,
      "num_tokens": 108290239.0,
      "step": 48515
    },
    {
      "entropy": 5.434009218215943,
      "epoch": 5.449261526366036,
      "grad_norm": 1.0625,
      "learning_rate": 0.000246937050294372,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23341335654258727,
      "num_tokens": 108301866.0,
      "step": 48520
    },
    {
      "entropy": 5.447401428222657,
      "epoch": 5.449823103273993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024689721552570714,
      "loss": 4.8948,
      "mean_token_accuracy": 0.2303045317530632,
      "num_tokens": 108314514.0,
      "step": 48525
    },
    {
      "entropy": 5.553496932983398,
      "epoch": 5.450384680181951,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002468573816518459,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22832864373922349,
      "num_tokens": 108325018.0,
      "step": 48530
    },
    {
      "entropy": 5.490911197662354,
      "epoch": 5.4509462570899085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002468175486740565,
      "loss": 4.8571,
      "mean_token_accuracy": 0.2365182176232338,
      "num_tokens": 108336195.0,
      "step": 48535
    },
    {
      "entropy": 5.445481777191162,
      "epoch": 5.451507833997866,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002467777165936073,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23215649127960206,
      "num_tokens": 108348169.0,
      "step": 48540
    },
    {
      "entropy": 5.464890432357788,
      "epoch": 5.452069410905824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002467378854117666,
      "loss": 4.906,
      "mean_token_accuracy": 0.23527680039405824,
      "num_tokens": 108359168.0,
      "step": 48545
    },
    {
      "entropy": 5.49435019493103,
      "epoch": 5.452630987813781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002466980551298026,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23027172386646272,
      "num_tokens": 108370615.0,
      "step": 48550
    },
    {
      "entropy": 5.503913164138794,
      "epoch": 5.453192564721738,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024665822574898344,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23322813510894774,
      "num_tokens": 108380487.0,
      "step": 48555
    },
    {
      "entropy": 5.453463220596314,
      "epoch": 5.453754141629696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024661839727057737,
      "loss": 4.8111,
      "mean_token_accuracy": 0.23695819526910783,
      "num_tokens": 108392221.0,
      "step": 48560
    },
    {
      "entropy": 5.545783281326294,
      "epoch": 5.454315718537654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024657856969585257,
      "loss": 4.92,
      "mean_token_accuracy": 0.22739667892456056,
      "num_tokens": 108404524.0,
      "step": 48565
    },
    {
      "entropy": 5.565169429779052,
      "epoch": 5.454877295445611,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000246538743026077,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21856648474931717,
      "num_tokens": 108416088.0,
      "step": 48570
    },
    {
      "entropy": 5.562700986862183,
      "epoch": 5.455438872353569,
      "grad_norm": 1.03125,
      "learning_rate": 0.000246498917262519,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21989426463842393,
      "num_tokens": 108428413.0,
      "step": 48575
    },
    {
      "entropy": 5.502653789520264,
      "epoch": 5.456000449261526,
      "grad_norm": 1.0,
      "learning_rate": 0.0002464590924064465,
      "loss": 4.8731,
      "mean_token_accuracy": 0.23422282338142394,
      "num_tokens": 108439505.0,
      "step": 48580
    },
    {
      "entropy": 5.495347023010254,
      "epoch": 5.456562026169484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024641926845912765,
      "loss": 4.9817,
      "mean_token_accuracy": 0.21856351792812348,
      "num_tokens": 108452849.0,
      "step": 48585
    },
    {
      "entropy": 5.43441333770752,
      "epoch": 5.4571236030774415,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024637944542183035,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23156758695840834,
      "num_tokens": 108463859.0,
      "step": 48590
    },
    {
      "entropy": 5.454178094863892,
      "epoch": 5.457685179985399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024633962329582255,
      "loss": 4.764,
      "mean_token_accuracy": 0.23611689507961273,
      "num_tokens": 108474972.0,
      "step": 48595
    },
    {
      "entropy": 5.466867303848266,
      "epoch": 5.458246756893357,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002462998020823723,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2395231157541275,
      "num_tokens": 108485490.0,
      "step": 48600
    },
    {
      "entropy": 5.487332105636597,
      "epoch": 5.458808333801314,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024625998178274747,
      "loss": 4.9621,
      "mean_token_accuracy": 0.23146501332521438,
      "num_tokens": 108496460.0,
      "step": 48605
    },
    {
      "entropy": 5.471648931503296,
      "epoch": 5.459369910709271,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002462201623982159,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22921790033578873,
      "num_tokens": 108508269.0,
      "step": 48610
    },
    {
      "entropy": 5.518256568908692,
      "epoch": 5.459931487617229,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002461803439300456,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23207414895296097,
      "num_tokens": 108519035.0,
      "step": 48615
    },
    {
      "entropy": 5.455826616287231,
      "epoch": 5.460493064525187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002461405263795044,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22795676589012145,
      "num_tokens": 108529646.0,
      "step": 48620
    },
    {
      "entropy": 5.5837725639343265,
      "epoch": 5.461054641433145,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002461007097478599,
      "loss": 5.0423,
      "mean_token_accuracy": 0.22334421873092652,
      "num_tokens": 108540888.0,
      "step": 48625
    },
    {
      "entropy": 5.372189998626709,
      "epoch": 5.461616218341102,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002460608940363801,
      "loss": 4.7082,
      "mean_token_accuracy": 0.2382564812898636,
      "num_tokens": 108552933.0,
      "step": 48630
    },
    {
      "entropy": 5.432645702362061,
      "epoch": 5.462177795249059,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024602107924633266,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23298147320747375,
      "num_tokens": 108564213.0,
      "step": 48635
    },
    {
      "entropy": 5.479667234420776,
      "epoch": 5.462739372157017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002459812653789851,
      "loss": 4.9447,
      "mean_token_accuracy": 0.23339366018772126,
      "num_tokens": 108575184.0,
      "step": 48640
    },
    {
      "entropy": 5.406191015243531,
      "epoch": 5.463300949064974,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002459414524356055,
      "loss": 4.8071,
      "mean_token_accuracy": 0.23927707970142365,
      "num_tokens": 108585226.0,
      "step": 48645
    },
    {
      "entropy": 5.424487590789795,
      "epoch": 5.463862525972932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002459016404174613,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24373016208410264,
      "num_tokens": 108595758.0,
      "step": 48650
    },
    {
      "entropy": 5.576717042922974,
      "epoch": 5.46442410288089,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024586182932582014,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22756571173667908,
      "num_tokens": 108606866.0,
      "step": 48655
    },
    {
      "entropy": 5.476849222183228,
      "epoch": 5.464985679788847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002458220191619496,
      "loss": 4.9252,
      "mean_token_accuracy": 0.23296440094709397,
      "num_tokens": 108618307.0,
      "step": 48660
    },
    {
      "entropy": 5.513866138458252,
      "epoch": 5.465547256696805,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024578220992711736,
      "loss": 4.9233,
      "mean_token_accuracy": 0.23065246045589446,
      "num_tokens": 108628669.0,
      "step": 48665
    },
    {
      "entropy": 5.485814380645752,
      "epoch": 5.466108833604762,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002457424016225909,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23852668553590775,
      "num_tokens": 108638943.0,
      "step": 48670
    },
    {
      "entropy": 5.463406610488891,
      "epoch": 5.4666704105127195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002457025942496376,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23402503430843352,
      "num_tokens": 108649177.0,
      "step": 48675
    },
    {
      "entropy": 5.484405612945556,
      "epoch": 5.467231987420678,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00024566278780952514,
      "loss": 4.8856,
      "mean_token_accuracy": 0.2275984138250351,
      "num_tokens": 108660387.0,
      "step": 48680
    },
    {
      "entropy": 5.488867330551147,
      "epoch": 5.467793564328635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002456229823035209,
      "loss": 4.8255,
      "mean_token_accuracy": 0.2380458727478981,
      "num_tokens": 108671327.0,
      "step": 48685
    },
    {
      "entropy": 5.489479351043701,
      "epoch": 5.468355141236592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024558317773289227,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22686739414930343,
      "num_tokens": 108681901.0,
      "step": 48690
    },
    {
      "entropy": 5.468024730682373,
      "epoch": 5.46891671814455,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002455433740989067,
      "loss": 4.9366,
      "mean_token_accuracy": 0.23435570299625397,
      "num_tokens": 108693205.0,
      "step": 48695
    },
    {
      "entropy": 5.501613330841065,
      "epoch": 5.469478295052507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002455035714028315,
      "loss": 4.9062,
      "mean_token_accuracy": 0.23647949248552322,
      "num_tokens": 108703651.0,
      "step": 48700
    },
    {
      "entropy": 5.4226093769073485,
      "epoch": 5.470039871960465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002454637696459341,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23426105380058287,
      "num_tokens": 108715811.0,
      "step": 48705
    },
    {
      "entropy": 5.4647211074829105,
      "epoch": 5.470601448868423,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024542396882948163,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22298502177000046,
      "num_tokens": 108727479.0,
      "step": 48710
    },
    {
      "entropy": 5.557422351837158,
      "epoch": 5.47116302577638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024538416895474144,
      "loss": 4.999,
      "mean_token_accuracy": 0.22326948046684264,
      "num_tokens": 108739083.0,
      "step": 48715
    },
    {
      "entropy": 5.601632452011108,
      "epoch": 5.471724602684338,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024534437002298083,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22125757336616517,
      "num_tokens": 108751596.0,
      "step": 48720
    },
    {
      "entropy": 5.563428783416748,
      "epoch": 5.472286179592295,
      "grad_norm": 1.0625,
      "learning_rate": 0.000245304572035467,
      "loss": 4.94,
      "mean_token_accuracy": 0.23493186235427857,
      "num_tokens": 108762856.0,
      "step": 48725
    },
    {
      "entropy": 5.476187896728516,
      "epoch": 5.4728477565002525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002452647749934671,
      "loss": 4.879,
      "mean_token_accuracy": 0.231661257147789,
      "num_tokens": 108773763.0,
      "step": 48730
    },
    {
      "entropy": 5.402482557296753,
      "epoch": 5.473409333408211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024522497889824837,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2338729113340378,
      "num_tokens": 108784803.0,
      "step": 48735
    },
    {
      "entropy": 5.458503532409668,
      "epoch": 5.473970910316168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002451851837510778,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22890920341014862,
      "num_tokens": 108795417.0,
      "step": 48740
    },
    {
      "entropy": 5.4636486053466795,
      "epoch": 5.474532487224125,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002451453895532226,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22898880541324615,
      "num_tokens": 108806382.0,
      "step": 48745
    },
    {
      "entropy": 5.499608945846558,
      "epoch": 5.475094064132083,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002451055963059496,
      "loss": 4.8316,
      "mean_token_accuracy": 0.24118355810642242,
      "num_tokens": 108816857.0,
      "step": 48750
    },
    {
      "entropy": 5.400055027008056,
      "epoch": 5.47565564104004,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024506580401052616,
      "loss": 4.8063,
      "mean_token_accuracy": 0.233623643219471,
      "num_tokens": 108828359.0,
      "step": 48755
    },
    {
      "entropy": 5.518301391601563,
      "epoch": 5.4762172179479975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002450260126682191,
      "loss": 4.9703,
      "mean_token_accuracy": 0.23315550833940507,
      "num_tokens": 108839520.0,
      "step": 48760
    },
    {
      "entropy": 5.525507307052612,
      "epoch": 5.476778794855956,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024498622228029545,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2294838547706604,
      "num_tokens": 108850529.0,
      "step": 48765
    },
    {
      "entropy": 5.4928957462310795,
      "epoch": 5.477340371763913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002449464328480221,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23758486211299895,
      "num_tokens": 108863452.0,
      "step": 48770
    },
    {
      "entropy": 5.442985105514526,
      "epoch": 5.477901948671871,
      "grad_norm": 1.0,
      "learning_rate": 0.000244906644372666,
      "loss": 4.8361,
      "mean_token_accuracy": 0.2356233224272728,
      "num_tokens": 108875050.0,
      "step": 48775
    },
    {
      "entropy": 5.502774858474732,
      "epoch": 5.478463525579828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000244866856855494,
      "loss": 4.8917,
      "mean_token_accuracy": 0.23313643634319306,
      "num_tokens": 108886386.0,
      "step": 48780
    },
    {
      "entropy": 5.529269552230835,
      "epoch": 5.479025102487785,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024482707029777284,
      "loss": 4.9849,
      "mean_token_accuracy": 0.23010611832141875,
      "num_tokens": 108899474.0,
      "step": 48785
    },
    {
      "entropy": 5.456200361251831,
      "epoch": 5.4795866793957435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024478728470076963,
      "loss": 4.8506,
      "mean_token_accuracy": 0.2402010440826416,
      "num_tokens": 108911714.0,
      "step": 48790
    },
    {
      "entropy": 5.480039739608765,
      "epoch": 5.480148256303701,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002447475000657509,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22998675853013992,
      "num_tokens": 108922500.0,
      "step": 48795
    },
    {
      "entropy": 5.467737483978271,
      "epoch": 5.480709833211658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002447077163939836,
      "loss": 4.944,
      "mean_token_accuracy": 0.22243638187646866,
      "num_tokens": 108933805.0,
      "step": 48800
    },
    {
      "entropy": 5.534436750411987,
      "epoch": 5.481271410119616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024466793368673427,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2216613471508026,
      "num_tokens": 108945538.0,
      "step": 48805
    },
    {
      "entropy": 5.5144371509552,
      "epoch": 5.481832987027573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002446281519452698,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22763215601444245,
      "num_tokens": 108957696.0,
      "step": 48810
    },
    {
      "entropy": 5.477454614639282,
      "epoch": 5.482394563935531,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002445883711708567,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23902850449085236,
      "num_tokens": 108967930.0,
      "step": 48815
    },
    {
      "entropy": 5.511138105392456,
      "epoch": 5.482956140843489,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00024454859136476155,
      "loss": 4.8944,
      "mean_token_accuracy": 0.23608840703964235,
      "num_tokens": 108978058.0,
      "step": 48820
    },
    {
      "entropy": 5.466859531402588,
      "epoch": 5.483517717751446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024450881252825117,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23584562093019484,
      "num_tokens": 108988854.0,
      "step": 48825
    },
    {
      "entropy": 5.489754009246826,
      "epoch": 5.484079294659404,
      "grad_norm": 1.140625,
      "learning_rate": 0.000244469034662592,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22986479252576827,
      "num_tokens": 109001060.0,
      "step": 48830
    },
    {
      "entropy": 5.487394094467163,
      "epoch": 5.484640871567361,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002444292577690506,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23352383375167846,
      "num_tokens": 109011943.0,
      "step": 48835
    },
    {
      "entropy": 5.437715291976929,
      "epoch": 5.485202448475318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024438948184889354,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23156629949808122,
      "num_tokens": 109022511.0,
      "step": 48840
    },
    {
      "entropy": 5.410395526885987,
      "epoch": 5.4857640253832765,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002443497069033872,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22721145004034043,
      "num_tokens": 109033757.0,
      "step": 48845
    },
    {
      "entropy": 5.396329164505005,
      "epoch": 5.486325602291234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002443099329337981,
      "loss": 4.7759,
      "mean_token_accuracy": 0.2391999900341034,
      "num_tokens": 109046075.0,
      "step": 48850
    },
    {
      "entropy": 5.442000055313111,
      "epoch": 5.486887179199192,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002442701599413927,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22190276980400087,
      "num_tokens": 109056768.0,
      "step": 48855
    },
    {
      "entropy": 5.525134801864624,
      "epoch": 5.487448756107149,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024423038792743715,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22564690113067626,
      "num_tokens": 109068929.0,
      "step": 48860
    },
    {
      "entropy": 5.51624174118042,
      "epoch": 5.488010333015106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002441906168931981,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22717302441596984,
      "num_tokens": 109080613.0,
      "step": 48865
    },
    {
      "entropy": 5.497553968429566,
      "epoch": 5.488571909923064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002441508468399417,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2298424005508423,
      "num_tokens": 109092742.0,
      "step": 48870
    },
    {
      "entropy": 5.480386066436767,
      "epoch": 5.4891334868310215,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024411107776893437,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2306819438934326,
      "num_tokens": 109103846.0,
      "step": 48875
    },
    {
      "entropy": 5.505369329452515,
      "epoch": 5.489695063738979,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002440713096814423,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22455752193927764,
      "num_tokens": 109115492.0,
      "step": 48880
    },
    {
      "entropy": 5.516284942626953,
      "epoch": 5.490256640646937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00024403154257873172,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22550592869520186,
      "num_tokens": 109126528.0,
      "step": 48885
    },
    {
      "entropy": 5.544899368286133,
      "epoch": 5.490818217554894,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024399177646206884,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22492387592792512,
      "num_tokens": 109138506.0,
      "step": 48890
    },
    {
      "entropy": 5.558720302581787,
      "epoch": 5.491379794462851,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024395201133271978,
      "loss": 4.9594,
      "mean_token_accuracy": 0.2323407083749771,
      "num_tokens": 109150380.0,
      "step": 48895
    },
    {
      "entropy": 5.571074199676514,
      "epoch": 5.491941371370809,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024391224719195077,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2323102504014969,
      "num_tokens": 109162369.0,
      "step": 48900
    },
    {
      "entropy": 5.468748712539673,
      "epoch": 5.492502948278767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024387248404102779,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2364979490637779,
      "num_tokens": 109173872.0,
      "step": 48905
    },
    {
      "entropy": 5.517179775238037,
      "epoch": 5.493064525186725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024383272188121708,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22766007781028746,
      "num_tokens": 109184316.0,
      "step": 48910
    },
    {
      "entropy": 5.5259379863739015,
      "epoch": 5.493626102094682,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002437929607137846,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2291978806257248,
      "num_tokens": 109196345.0,
      "step": 48915
    },
    {
      "entropy": 5.481290483474732,
      "epoch": 5.494187679002639,
      "grad_norm": 1.140625,
      "learning_rate": 0.00024375320053999627,
      "loss": 4.8282,
      "mean_token_accuracy": 0.24153456538915635,
      "num_tokens": 109206858.0,
      "step": 48920
    },
    {
      "entropy": 5.499751281738281,
      "epoch": 5.494749255910597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024371344136111823,
      "loss": 4.88,
      "mean_token_accuracy": 0.23030963242053987,
      "num_tokens": 109216866.0,
      "step": 48925
    },
    {
      "entropy": 5.430970191955566,
      "epoch": 5.4953108328185545,
      "grad_norm": 1.125,
      "learning_rate": 0.0002436736831784162,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23710994124412538,
      "num_tokens": 109226623.0,
      "step": 48930
    },
    {
      "entropy": 5.538551616668701,
      "epoch": 5.495872409726512,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024363392599315636,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21655757278203963,
      "num_tokens": 109237560.0,
      "step": 48935
    },
    {
      "entropy": 5.502460670471192,
      "epoch": 5.49643398663447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024359416980660454,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22573215812444686,
      "num_tokens": 109249010.0,
      "step": 48940
    },
    {
      "entropy": 5.5118495464324955,
      "epoch": 5.496995563542427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024355441462002648,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22918282598257064,
      "num_tokens": 109260328.0,
      "step": 48945
    },
    {
      "entropy": 5.544164657592773,
      "epoch": 5.497557140450384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024351466043468807,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2287612870335579,
      "num_tokens": 109270896.0,
      "step": 48950
    },
    {
      "entropy": 5.599378204345703,
      "epoch": 5.498118717358342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024347490725185506,
      "loss": 5.0662,
      "mean_token_accuracy": 0.22146632969379426,
      "num_tokens": 109282259.0,
      "step": 48955
    },
    {
      "entropy": 5.426846361160278,
      "epoch": 5.4986802942663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024343515507279325,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23243436366319656,
      "num_tokens": 109293871.0,
      "step": 48960
    },
    {
      "entropy": 5.51053204536438,
      "epoch": 5.499241871174258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002433954038987683,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22893541902303696,
      "num_tokens": 109305881.0,
      "step": 48965
    },
    {
      "entropy": 5.47735071182251,
      "epoch": 5.499803448082215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024335565373104594,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2346973419189453,
      "num_tokens": 109317088.0,
      "step": 48970
    },
    {
      "entropy": 5.529799509048462,
      "epoch": 5.500365024990172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024331590457089193,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22812590301036834,
      "num_tokens": 109328379.0,
      "step": 48975
    },
    {
      "entropy": 5.476178932189941,
      "epoch": 5.50092660189813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024327615641957175,
      "loss": 4.8323,
      "mean_token_accuracy": 0.23445447832345961,
      "num_tokens": 109339177.0,
      "step": 48980
    },
    {
      "entropy": 5.526739501953125,
      "epoch": 5.501488178806087,
      "grad_norm": 1.0,
      "learning_rate": 0.00024323640927835106,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2297579437494278,
      "num_tokens": 109350725.0,
      "step": 48985
    },
    {
      "entropy": 5.4915632724761965,
      "epoch": 5.5020497557140455,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024319666314849542,
      "loss": 4.897,
      "mean_token_accuracy": 0.2299300417304039,
      "num_tokens": 109361814.0,
      "step": 48990
    },
    {
      "entropy": 5.5119704246521,
      "epoch": 5.502611332622003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024315691803127037,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23931482881307603,
      "num_tokens": 109372920.0,
      "step": 48995
    },
    {
      "entropy": 5.556078147888184,
      "epoch": 5.50317290952996,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002431171739279413,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22204731404781342,
      "num_tokens": 109384415.0,
      "step": 49000
    },
    {
      "entropy": 5.515207147598266,
      "epoch": 5.503734486437917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024307743083977384,
      "loss": 4.9815,
      "mean_token_accuracy": 0.23305245786905288,
      "num_tokens": 109394696.0,
      "step": 49005
    },
    {
      "entropy": 5.553015518188476,
      "epoch": 5.504296063345875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024303768876803345,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22980528175830842,
      "num_tokens": 109405513.0,
      "step": 49010
    },
    {
      "entropy": 5.453128623962402,
      "epoch": 5.5048576402538325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024299794771398536,
      "loss": 4.8746,
      "mean_token_accuracy": 0.23320195972919464,
      "num_tokens": 109416693.0,
      "step": 49015
    },
    {
      "entropy": 5.515306758880615,
      "epoch": 5.505419217161791,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024295820767889504,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21658524572849275,
      "num_tokens": 109428128.0,
      "step": 49020
    },
    {
      "entropy": 5.480225658416748,
      "epoch": 5.505980794069748,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00024291846866402784,
      "loss": 4.908,
      "mean_token_accuracy": 0.23127192109823227,
      "num_tokens": 109441068.0,
      "step": 49025
    },
    {
      "entropy": 5.629679107666016,
      "epoch": 5.506542370977705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024287873067064904,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22673519998788833,
      "num_tokens": 109452352.0,
      "step": 49030
    },
    {
      "entropy": 5.472667455673218,
      "epoch": 5.507103947885663,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024283899370002393,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22513011246919631,
      "num_tokens": 109463742.0,
      "step": 49035
    },
    {
      "entropy": 5.524774980545044,
      "epoch": 5.50766552479362,
      "grad_norm": 1.125,
      "learning_rate": 0.00024279925775341765,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22911540865898133,
      "num_tokens": 109474521.0,
      "step": 49040
    },
    {
      "entropy": 5.503566455841065,
      "epoch": 5.5082271017015785,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024275952283209557,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2257555305957794,
      "num_tokens": 109486207.0,
      "step": 49045
    },
    {
      "entropy": 5.479707908630371,
      "epoch": 5.508788678609536,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024271978893732277,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2275390535593033,
      "num_tokens": 109497509.0,
      "step": 49050
    },
    {
      "entropy": 5.536691093444825,
      "epoch": 5.509350255517493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024268005607036442,
      "loss": 4.9484,
      "mean_token_accuracy": 0.23099400103092194,
      "num_tokens": 109509850.0,
      "step": 49055
    },
    {
      "entropy": 5.642967319488525,
      "epoch": 5.509911832425451,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024264032423248555,
      "loss": 5.003,
      "mean_token_accuracy": 0.225896418094635,
      "num_tokens": 109521098.0,
      "step": 49060
    },
    {
      "entropy": 5.49496636390686,
      "epoch": 5.510473409333408,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002426005934249514,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22601827085018159,
      "num_tokens": 109532563.0,
      "step": 49065
    },
    {
      "entropy": 5.5059596538543705,
      "epoch": 5.5110349862413655,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002425608636490269,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22552436590194702,
      "num_tokens": 109543793.0,
      "step": 49070
    },
    {
      "entropy": 5.509867238998413,
      "epoch": 5.511596563149324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024252113490597692,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22553712874650955,
      "num_tokens": 109554062.0,
      "step": 49075
    },
    {
      "entropy": 5.475304412841797,
      "epoch": 5.512158140057281,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002424814071970668,
      "loss": 4.9113,
      "mean_token_accuracy": 0.23531164079904557,
      "num_tokens": 109564404.0,
      "step": 49080
    },
    {
      "entropy": 5.629907274246216,
      "epoch": 5.512719716965238,
      "grad_norm": 1.15625,
      "learning_rate": 0.00024244168052356125,
      "loss": 5.0889,
      "mean_token_accuracy": 0.21920591294765474,
      "num_tokens": 109575080.0,
      "step": 49085
    },
    {
      "entropy": 5.600473213195801,
      "epoch": 5.513281293873196,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024240195488672516,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22784946411848067,
      "num_tokens": 109586653.0,
      "step": 49090
    },
    {
      "entropy": 5.532931661605835,
      "epoch": 5.513842870781153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024236223028782357,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22641454488039017,
      "num_tokens": 109598399.0,
      "step": 49095
    },
    {
      "entropy": 5.602181005477905,
      "epoch": 5.514404447689111,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024232250672812116,
      "loss": 5.0976,
      "mean_token_accuracy": 0.22055166810750962,
      "num_tokens": 109610946.0,
      "step": 49100
    },
    {
      "entropy": 5.411599779129029,
      "epoch": 5.514966024597069,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024228278420888283,
      "loss": 4.7524,
      "mean_token_accuracy": 0.2460738018155098,
      "num_tokens": 109621838.0,
      "step": 49105
    },
    {
      "entropy": 5.488338422775269,
      "epoch": 5.515527601505026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002422430627313733,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2329806163907051,
      "num_tokens": 109633897.0,
      "step": 49110
    },
    {
      "entropy": 5.554320383071899,
      "epoch": 5.516089178412984,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024220334229685736,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2230490803718567,
      "num_tokens": 109646012.0,
      "step": 49115
    },
    {
      "entropy": 5.506078004837036,
      "epoch": 5.516650755320941,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00024216362290659977,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23179092705249787,
      "num_tokens": 109658188.0,
      "step": 49120
    },
    {
      "entropy": 5.442566537857056,
      "epoch": 5.517212332228898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024212390456186512,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23851809948682784,
      "num_tokens": 109669219.0,
      "step": 49125
    },
    {
      "entropy": 5.5699115753173825,
      "epoch": 5.5177739091368565,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024208418726391813,
      "loss": 5.0672,
      "mean_token_accuracy": 0.22375694066286086,
      "num_tokens": 109681040.0,
      "step": 49130
    },
    {
      "entropy": 5.495637512207031,
      "epoch": 5.518335486044814,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024204447101402337,
      "loss": 4.8958,
      "mean_token_accuracy": 0.237813538312912,
      "num_tokens": 109692123.0,
      "step": 49135
    },
    {
      "entropy": 5.477726793289184,
      "epoch": 5.518897062952771,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024200475581344545,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23144594728946685,
      "num_tokens": 109702900.0,
      "step": 49140
    },
    {
      "entropy": 5.4698814868927,
      "epoch": 5.519458639860729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002419650416634488,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22475549727678298,
      "num_tokens": 109713301.0,
      "step": 49145
    },
    {
      "entropy": 5.431608533859253,
      "epoch": 5.520020216768686,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024192532856529814,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2313602015376091,
      "num_tokens": 109725272.0,
      "step": 49150
    },
    {
      "entropy": 5.500447797775268,
      "epoch": 5.520581793676644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024188561652025787,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22813388109207153,
      "num_tokens": 109736154.0,
      "step": 49155
    },
    {
      "entropy": 5.514536571502686,
      "epoch": 5.521143370584602,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024184590552959234,
      "loss": 4.922,
      "mean_token_accuracy": 0.22774300426244737,
      "num_tokens": 109748403.0,
      "step": 49160
    },
    {
      "entropy": 5.5046814441680905,
      "epoch": 5.521704947492559,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024180619559456608,
      "loss": 4.7901,
      "mean_token_accuracy": 0.24316680878400804,
      "num_tokens": 109758232.0,
      "step": 49165
    },
    {
      "entropy": 5.559400320053101,
      "epoch": 5.522266524400517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002417664867164434,
      "loss": 4.9834,
      "mean_token_accuracy": 0.2225027099251747,
      "num_tokens": 109770198.0,
      "step": 49170
    },
    {
      "entropy": 5.482193565368652,
      "epoch": 5.522828101308474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002417267788964887,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23252182006835936,
      "num_tokens": 109782564.0,
      "step": 49175
    },
    {
      "entropy": 5.445253705978393,
      "epoch": 5.523389678216432,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024168707213596626,
      "loss": 4.8584,
      "mean_token_accuracy": 0.23272517919540406,
      "num_tokens": 109793186.0,
      "step": 49180
    },
    {
      "entropy": 5.413465929031372,
      "epoch": 5.5239512551243894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024164736643614033,
      "loss": 4.8158,
      "mean_token_accuracy": 0.23052551448345185,
      "num_tokens": 109804799.0,
      "step": 49185
    },
    {
      "entropy": 5.456015491485596,
      "epoch": 5.524512832032347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002416076617982752,
      "loss": 4.7667,
      "mean_token_accuracy": 0.2343834862112999,
      "num_tokens": 109816031.0,
      "step": 49190
    },
    {
      "entropy": 5.603163051605224,
      "epoch": 5.525074408940304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024156795822363508,
      "loss": 5.0089,
      "mean_token_accuracy": 0.2204325884580612,
      "num_tokens": 109826952.0,
      "step": 49195
    },
    {
      "entropy": 5.557055902481079,
      "epoch": 5.525635985848262,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024152825571348418,
      "loss": 4.9927,
      "mean_token_accuracy": 0.22534118592739105,
      "num_tokens": 109839045.0,
      "step": 49200
    },
    {
      "entropy": 5.5217602252960205,
      "epoch": 5.526197562756219,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024148855426908656,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22777996957302094,
      "num_tokens": 109849812.0,
      "step": 49205
    },
    {
      "entropy": 5.5167036056518555,
      "epoch": 5.526759139664177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002414488538917064,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2297443464398384,
      "num_tokens": 109861368.0,
      "step": 49210
    },
    {
      "entropy": 5.50435791015625,
      "epoch": 5.5273207165721345,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002414091545826077,
      "loss": 4.938,
      "mean_token_accuracy": 0.2251734495162964,
      "num_tokens": 109872388.0,
      "step": 49215
    },
    {
      "entropy": 5.51364164352417,
      "epoch": 5.527882293480092,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024136945634305445,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22808264195919037,
      "num_tokens": 109883301.0,
      "step": 49220
    },
    {
      "entropy": 5.444095039367676,
      "epoch": 5.52844387038805,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024132975917431088,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23091235458850862,
      "num_tokens": 109894823.0,
      "step": 49225
    },
    {
      "entropy": 5.529217481613159,
      "epoch": 5.529005447296007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024129006307764084,
      "loss": 5.0024,
      "mean_token_accuracy": 0.22320341020822526,
      "num_tokens": 109905946.0,
      "step": 49230
    },
    {
      "entropy": 5.537041091918946,
      "epoch": 5.529567024203965,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024125036805430829,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22578869313001632,
      "num_tokens": 109917369.0,
      "step": 49235
    },
    {
      "entropy": 5.537041139602661,
      "epoch": 5.530128601111922,
      "grad_norm": 1.078125,
      "learning_rate": 0.000241210674105577,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22732775658369064,
      "num_tokens": 109929001.0,
      "step": 49240
    },
    {
      "entropy": 5.578138208389282,
      "epoch": 5.53069017801988,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024117098123271107,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22062288373708724,
      "num_tokens": 109941557.0,
      "step": 49245
    },
    {
      "entropy": 5.574688673019409,
      "epoch": 5.531251754927838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024113128943697415,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23062491565942764,
      "num_tokens": 109951603.0,
      "step": 49250
    },
    {
      "entropy": 5.508815765380859,
      "epoch": 5.531813331835795,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002410915987196301,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23036534190177918,
      "num_tokens": 109963037.0,
      "step": 49255
    },
    {
      "entropy": 5.493857431411743,
      "epoch": 5.532374908743752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00024105190908194275,
      "loss": 4.86,
      "mean_token_accuracy": 0.23665053844451905,
      "num_tokens": 109974267.0,
      "step": 49260
    },
    {
      "entropy": 5.4830498695373535,
      "epoch": 5.53293648565171,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024101222052517573,
      "loss": 4.8875,
      "mean_token_accuracy": 0.23710955083370208,
      "num_tokens": 109985591.0,
      "step": 49265
    },
    {
      "entropy": 5.3768579959869385,
      "epoch": 5.5334980625596675,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024097253305059287,
      "loss": 4.7906,
      "mean_token_accuracy": 0.2388907939195633,
      "num_tokens": 109997219.0,
      "step": 49270
    },
    {
      "entropy": 5.523189878463745,
      "epoch": 5.534059639467625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002409328466594577,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22069162726402283,
      "num_tokens": 110006775.0,
      "step": 49275
    },
    {
      "entropy": 5.511471509933472,
      "epoch": 5.534621216375583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002408931613530339,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2323594868183136,
      "num_tokens": 110018162.0,
      "step": 49280
    },
    {
      "entropy": 5.605857801437378,
      "epoch": 5.53518279328354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002408534771325851,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22082937955856324,
      "num_tokens": 110029960.0,
      "step": 49285
    },
    {
      "entropy": 5.540847444534302,
      "epoch": 5.535744370191498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002408137939993747,
      "loss": 4.9569,
      "mean_token_accuracy": 0.2241254836320877,
      "num_tokens": 110042043.0,
      "step": 49290
    },
    {
      "entropy": 5.439835500717163,
      "epoch": 5.536305947099455,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024077411195466648,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2259758949279785,
      "num_tokens": 110053094.0,
      "step": 49295
    },
    {
      "entropy": 5.389796209335327,
      "epoch": 5.5368675240074126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002407344309997238,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23808782398700715,
      "num_tokens": 110063555.0,
      "step": 49300
    },
    {
      "entropy": 5.513140392303467,
      "epoch": 5.537429100915371,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024069475113581014,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22685594260692596,
      "num_tokens": 110075089.0,
      "step": 49305
    },
    {
      "entropy": 5.5356649398803714,
      "epoch": 5.537990677823328,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024065507236418888,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22382721900939942,
      "num_tokens": 110086170.0,
      "step": 49310
    },
    {
      "entropy": 5.456835174560547,
      "epoch": 5.538552254731285,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024061539468612348,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23982564210891724,
      "num_tokens": 110096251.0,
      "step": 49315
    },
    {
      "entropy": 5.422799587249756,
      "epoch": 5.539113831639243,
      "grad_norm": 1.0,
      "learning_rate": 0.00024057571810287722,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2314571112394333,
      "num_tokens": 110107625.0,
      "step": 49320
    },
    {
      "entropy": 5.4699907302856445,
      "epoch": 5.5396754085472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024053604261571338,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2344010666012764,
      "num_tokens": 110116929.0,
      "step": 49325
    },
    {
      "entropy": 5.4768143653869625,
      "epoch": 5.540236985455158,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024049636822589537,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23291928172111512,
      "num_tokens": 110126975.0,
      "step": 49330
    },
    {
      "entropy": 5.5482926845550535,
      "epoch": 5.540798562363116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002404566949346864,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23256452828645707,
      "num_tokens": 110137817.0,
      "step": 49335
    },
    {
      "entropy": 5.570621824264526,
      "epoch": 5.541360139271073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002404170227433497,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22663910686969757,
      "num_tokens": 110149897.0,
      "step": 49340
    },
    {
      "entropy": 5.48655891418457,
      "epoch": 5.541921716179031,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024037735165314834,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22732877284288405,
      "num_tokens": 110161291.0,
      "step": 49345
    },
    {
      "entropy": 5.633256530761718,
      "epoch": 5.542483293086988,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00024033768166534553,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2322056233882904,
      "num_tokens": 110172670.0,
      "step": 49350
    },
    {
      "entropy": 5.641352272033691,
      "epoch": 5.5430448699949455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024029801278120438,
      "loss": 5.0359,
      "mean_token_accuracy": 0.22491966634988786,
      "num_tokens": 110183165.0,
      "step": 49355
    },
    {
      "entropy": 5.489501237869263,
      "epoch": 5.543606446902904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002402583450019879,
      "loss": 4.8741,
      "mean_token_accuracy": 0.22863700091838837,
      "num_tokens": 110194177.0,
      "step": 49360
    },
    {
      "entropy": 5.413175058364868,
      "epoch": 5.544168023810861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024021867832895927,
      "loss": 4.8003,
      "mean_token_accuracy": 0.24187544882297515,
      "num_tokens": 110204641.0,
      "step": 49365
    },
    {
      "entropy": 5.578957891464233,
      "epoch": 5.544729600718819,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002401790127633814,
      "loss": 5.1089,
      "mean_token_accuracy": 0.218769571185112,
      "num_tokens": 110216580.0,
      "step": 49370
    },
    {
      "entropy": 5.4895368099212645,
      "epoch": 5.545291177626776,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024013934830651728,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2275429978966713,
      "num_tokens": 110228731.0,
      "step": 49375
    },
    {
      "entropy": 5.549504995346069,
      "epoch": 5.545852754534733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024009968495962986,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2319965198636055,
      "num_tokens": 110241308.0,
      "step": 49380
    },
    {
      "entropy": 5.437914037704468,
      "epoch": 5.546414331442691,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024006002272398199,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23240613490343093,
      "num_tokens": 110253759.0,
      "step": 49385
    },
    {
      "entropy": 5.503518438339233,
      "epoch": 5.546975908350649,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002400203616008365,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23693045526742934,
      "num_tokens": 110265193.0,
      "step": 49390
    },
    {
      "entropy": 5.49438328742981,
      "epoch": 5.547537485258606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002399807015914563,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22652714252471923,
      "num_tokens": 110275777.0,
      "step": 49395
    },
    {
      "entropy": 5.503175735473633,
      "epoch": 5.548099062166564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002399410426971041,
      "loss": 4.8419,
      "mean_token_accuracy": 0.237531478703022,
      "num_tokens": 110286537.0,
      "step": 49400
    },
    {
      "entropy": 5.492461061477661,
      "epoch": 5.548660639074521,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023990138491904273,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23141546100378035,
      "num_tokens": 110297188.0,
      "step": 49405
    },
    {
      "entropy": 5.43338565826416,
      "epoch": 5.5492222159824784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023986172825853492,
      "loss": 4.7614,
      "mean_token_accuracy": 0.24244530349969864,
      "num_tokens": 110306355.0,
      "step": 49410
    },
    {
      "entropy": 5.457644367218018,
      "epoch": 5.5497837928904366,
      "grad_norm": 1.125,
      "learning_rate": 0.00023982207271684326,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22395133823156357,
      "num_tokens": 110317657.0,
      "step": 49415
    },
    {
      "entropy": 5.4574450016021725,
      "epoch": 5.550345369798394,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002397824182952305,
      "loss": 4.8503,
      "mean_token_accuracy": 0.2409653127193451,
      "num_tokens": 110327665.0,
      "step": 49420
    },
    {
      "entropy": 5.4028266906738285,
      "epoch": 5.550906946706352,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023974276499495918,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23370686024427414,
      "num_tokens": 110338054.0,
      "step": 49425
    },
    {
      "entropy": 5.5064293384552006,
      "epoch": 5.551468523614309,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023970311281729185,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22437463104724883,
      "num_tokens": 110349360.0,
      "step": 49430
    },
    {
      "entropy": 5.379789972305298,
      "epoch": 5.552030100522266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000239663461763491,
      "loss": 4.754,
      "mean_token_accuracy": 0.23574168086051941,
      "num_tokens": 110359336.0,
      "step": 49435
    },
    {
      "entropy": 5.462270259857178,
      "epoch": 5.552591677430224,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002396238118348194,
      "loss": 4.8542,
      "mean_token_accuracy": 0.2295437589287758,
      "num_tokens": 110371808.0,
      "step": 49440
    },
    {
      "entropy": 5.518948364257812,
      "epoch": 5.553153254338182,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023958416303253933,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22497284263372422,
      "num_tokens": 110383697.0,
      "step": 49445
    },
    {
      "entropy": 5.443642330169678,
      "epoch": 5.553714831246139,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023954451535791323,
      "loss": 4.8558,
      "mean_token_accuracy": 0.2335679665207863,
      "num_tokens": 110395310.0,
      "step": 49450
    },
    {
      "entropy": 5.473076725006104,
      "epoch": 5.554276408154097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023950486881220353,
      "loss": 4.9247,
      "mean_token_accuracy": 0.23345827162265778,
      "num_tokens": 110407140.0,
      "step": 49455
    },
    {
      "entropy": 5.535160636901855,
      "epoch": 5.554837985062054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023946522339667254,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23556961864233017,
      "num_tokens": 110418213.0,
      "step": 49460
    },
    {
      "entropy": 5.502024078369141,
      "epoch": 5.555399561970011,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002394255791125826,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23267967253923416,
      "num_tokens": 110428868.0,
      "step": 49465
    },
    {
      "entropy": 5.583645439147949,
      "epoch": 5.5559611388779695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000239385935961196,
      "loss": 5.0014,
      "mean_token_accuracy": 0.23020316064357757,
      "num_tokens": 110440156.0,
      "step": 49470
    },
    {
      "entropy": 5.511643362045288,
      "epoch": 5.556522715785927,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023934629394377512,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23093205243349074,
      "num_tokens": 110450889.0,
      "step": 49475
    },
    {
      "entropy": 5.545291996002197,
      "epoch": 5.557084292693885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000239306653061582,
      "loss": 4.9485,
      "mean_token_accuracy": 0.23852748721837996,
      "num_tokens": 110463557.0,
      "step": 49480
    },
    {
      "entropy": 5.539779901504517,
      "epoch": 5.557645869601842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023926701331587885,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2252531662583351,
      "num_tokens": 110473944.0,
      "step": 49485
    },
    {
      "entropy": 5.532902479171753,
      "epoch": 5.558207446509799,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023922737470792794,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2295096531510353,
      "num_tokens": 110484999.0,
      "step": 49490
    },
    {
      "entropy": 5.667097759246826,
      "epoch": 5.558769023417757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002391877372389913,
      "loss": 5.0658,
      "mean_token_accuracy": 0.21697979867458345,
      "num_tokens": 110498669.0,
      "step": 49495
    },
    {
      "entropy": 5.597529029846191,
      "epoch": 5.559330600325715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023914810091033096,
      "loss": 5.0355,
      "mean_token_accuracy": 0.21642813682556153,
      "num_tokens": 110511183.0,
      "step": 49500
    },
    {
      "entropy": 5.5448112964630125,
      "epoch": 5.559892177233672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002391084657232089,
      "loss": 4.92,
      "mean_token_accuracy": 0.23083713203668593,
      "num_tokens": 110521886.0,
      "step": 49505
    },
    {
      "entropy": 5.51102499961853,
      "epoch": 5.56045375414163,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023906883167888727,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22866471409797667,
      "num_tokens": 110532413.0,
      "step": 49510
    },
    {
      "entropy": 5.5450035572052006,
      "epoch": 5.561015331049587,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000239029198778628,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22694263011217117,
      "num_tokens": 110544900.0,
      "step": 49515
    },
    {
      "entropy": 5.4578320503234865,
      "epoch": 5.561576907957544,
      "grad_norm": 1.15625,
      "learning_rate": 0.000238989567023693,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23409500271081923,
      "num_tokens": 110556811.0,
      "step": 49520
    },
    {
      "entropy": 5.423512315750122,
      "epoch": 5.562138484865502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023894993641534418,
      "loss": 4.7993,
      "mean_token_accuracy": 0.23473928719758988,
      "num_tokens": 110567688.0,
      "step": 49525
    },
    {
      "entropy": 5.514882040023804,
      "epoch": 5.56270006177346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023891030695484333,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22757958024740219,
      "num_tokens": 110579821.0,
      "step": 49530
    },
    {
      "entropy": 5.434156990051269,
      "epoch": 5.563261638681418,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002388706786434523,
      "loss": 4.8188,
      "mean_token_accuracy": 0.2309257686138153,
      "num_tokens": 110589897.0,
      "step": 49535
    },
    {
      "entropy": 5.540633201599121,
      "epoch": 5.563823215589375,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002388310514824329,
      "loss": 4.9623,
      "mean_token_accuracy": 0.23056471347808838,
      "num_tokens": 110600826.0,
      "step": 49540
    },
    {
      "entropy": 5.474032402038574,
      "epoch": 5.564384792497332,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023879142547304677,
      "loss": 4.977,
      "mean_token_accuracy": 0.22263770997524263,
      "num_tokens": 110611138.0,
      "step": 49545
    },
    {
      "entropy": 5.519557666778565,
      "epoch": 5.56494636940529,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023875180061655576,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23141565471887587,
      "num_tokens": 110622967.0,
      "step": 49550
    },
    {
      "entropy": 5.5636626243591305,
      "epoch": 5.5655079463132475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023871217691422147,
      "loss": 4.934,
      "mean_token_accuracy": 0.23039920926094054,
      "num_tokens": 110635062.0,
      "step": 49555
    },
    {
      "entropy": 5.496329021453858,
      "epoch": 5.566069523221206,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023867255436730555,
      "loss": 4.866,
      "mean_token_accuracy": 0.24212087839841842,
      "num_tokens": 110646061.0,
      "step": 49560
    },
    {
      "entropy": 5.496082305908203,
      "epoch": 5.566631100129163,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002386329329770696,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22654022574424743,
      "num_tokens": 110658233.0,
      "step": 49565
    },
    {
      "entropy": 5.5654685497283936,
      "epoch": 5.56719267703712,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023859331274477514,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22957328706979752,
      "num_tokens": 110669253.0,
      "step": 49570
    },
    {
      "entropy": 5.414662551879883,
      "epoch": 5.567754253945077,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023855369367168378,
      "loss": 4.8185,
      "mean_token_accuracy": 0.234725484251976,
      "num_tokens": 110679386.0,
      "step": 49575
    },
    {
      "entropy": 5.508033037185669,
      "epoch": 5.568315830853035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00023851407575905682,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2359297439455986,
      "num_tokens": 110690517.0,
      "step": 49580
    },
    {
      "entropy": 5.475931406021118,
      "epoch": 5.568877407760993,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002384744590081559,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22714485377073287,
      "num_tokens": 110701023.0,
      "step": 49585
    },
    {
      "entropy": 5.496267890930175,
      "epoch": 5.569438984668951,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002384348434202424,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23027718663215638,
      "num_tokens": 110711530.0,
      "step": 49590
    },
    {
      "entropy": 5.511767721176147,
      "epoch": 5.570000561576908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023839522899657767,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2287864625453949,
      "num_tokens": 110723201.0,
      "step": 49595
    },
    {
      "entropy": 5.443825054168701,
      "epoch": 5.570562138484865,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023835561573842302,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23598258048295975,
      "num_tokens": 110733872.0,
      "step": 49600
    },
    {
      "entropy": 5.547453784942627,
      "epoch": 5.571123715392823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002383160036470398,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22076515406370162,
      "num_tokens": 110745038.0,
      "step": 49605
    },
    {
      "entropy": 5.467517328262329,
      "epoch": 5.5716852923007805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00023827639272368923,
      "loss": 4.922,
      "mean_token_accuracy": 0.22335768193006517,
      "num_tokens": 110756002.0,
      "step": 49610
    },
    {
      "entropy": 5.486204767227173,
      "epoch": 5.572246869208739,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023823678296963253,
      "loss": 4.8858,
      "mean_token_accuracy": 0.23065439462661744,
      "num_tokens": 110766462.0,
      "step": 49615
    },
    {
      "entropy": 5.498572587966919,
      "epoch": 5.572808446116696,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023819717438613097,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22348470240831375,
      "num_tokens": 110777764.0,
      "step": 49620
    },
    {
      "entropy": 5.504435729980469,
      "epoch": 5.573370023024653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023815756697444568,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23356441855430604,
      "num_tokens": 110788790.0,
      "step": 49625
    },
    {
      "entropy": 5.439608860015869,
      "epoch": 5.573931599932611,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023811796073583774,
      "loss": 4.8679,
      "mean_token_accuracy": 0.24015094488859176,
      "num_tokens": 110800178.0,
      "step": 49630
    },
    {
      "entropy": 5.475143623352051,
      "epoch": 5.574493176840568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023807835567156815,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22949534505605698,
      "num_tokens": 110811037.0,
      "step": 49635
    },
    {
      "entropy": 5.573132038116455,
      "epoch": 5.5750547537485255,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023803875178289813,
      "loss": 5.018,
      "mean_token_accuracy": 0.22584325820207596,
      "num_tokens": 110821368.0,
      "step": 49640
    },
    {
      "entropy": 5.504440975189209,
      "epoch": 5.575616330656484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023799914907108856,
      "loss": 4.8846,
      "mean_token_accuracy": 0.227621927857399,
      "num_tokens": 110832831.0,
      "step": 49645
    },
    {
      "entropy": 5.5050774097442625,
      "epoch": 5.576177907564441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00023795954753740034,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23208883851766587,
      "num_tokens": 110844280.0,
      "step": 49650
    },
    {
      "entropy": 5.428195762634277,
      "epoch": 5.576739484472398,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023791994718309455,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23718035370111465,
      "num_tokens": 110854882.0,
      "step": 49655
    },
    {
      "entropy": 5.5805933475494385,
      "epoch": 5.577301061380356,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023788034800943203,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22631799280643464,
      "num_tokens": 110865929.0,
      "step": 49660
    },
    {
      "entropy": 5.482133865356445,
      "epoch": 5.577862638288313,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023784075001767369,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23251930475234986,
      "num_tokens": 110875996.0,
      "step": 49665
    },
    {
      "entropy": 5.521721935272216,
      "epoch": 5.5784242151962715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023780115320908024,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22593717873096467,
      "num_tokens": 110887862.0,
      "step": 49670
    },
    {
      "entropy": 5.475393199920655,
      "epoch": 5.578985792104229,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023776155758491253,
      "loss": 4.8055,
      "mean_token_accuracy": 0.24094399958848953,
      "num_tokens": 110899340.0,
      "step": 49675
    },
    {
      "entropy": 5.5461564540863035,
      "epoch": 5.579547369012186,
      "grad_norm": 1.125,
      "learning_rate": 0.0002377219631464312,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2220518633723259,
      "num_tokens": 110910141.0,
      "step": 49680
    },
    {
      "entropy": 5.4757099628448485,
      "epoch": 5.580108945920144,
      "grad_norm": 1.09375,
      "learning_rate": 0.000237682369894897,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22761501371860504,
      "num_tokens": 110920488.0,
      "step": 49685
    },
    {
      "entropy": 5.543192529678345,
      "epoch": 5.580670522828101,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023764277783157068,
      "loss": 4.9993,
      "mean_token_accuracy": 0.23022578805685043,
      "num_tokens": 110933074.0,
      "step": 49690
    },
    {
      "entropy": 5.559882497787475,
      "epoch": 5.5812320997360585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023760318695771282,
      "loss": 4.9654,
      "mean_token_accuracy": 0.23127468824386596,
      "num_tokens": 110945251.0,
      "step": 49695
    },
    {
      "entropy": 5.4442853927612305,
      "epoch": 5.581793676644017,
      "grad_norm": 1.125,
      "learning_rate": 0.00023756359727458397,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23788377344608308,
      "num_tokens": 110955742.0,
      "step": 49700
    },
    {
      "entropy": 5.38725643157959,
      "epoch": 5.582355253551974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002375240087834447,
      "loss": 4.7711,
      "mean_token_accuracy": 0.23494660407304763,
      "num_tokens": 110965395.0,
      "step": 49705
    },
    {
      "entropy": 5.394710683822632,
      "epoch": 5.582916830459931,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023748442148555554,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23941817730665207,
      "num_tokens": 110976770.0,
      "step": 49710
    },
    {
      "entropy": 5.501466703414917,
      "epoch": 5.583478407367889,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023744483538217694,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22311840802431107,
      "num_tokens": 110987997.0,
      "step": 49715
    },
    {
      "entropy": 5.579561328887939,
      "epoch": 5.584039984275846,
      "grad_norm": 1.125,
      "learning_rate": 0.00023740525047456936,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2197435438632965,
      "num_tokens": 110998969.0,
      "step": 49720
    },
    {
      "entropy": 5.506470489501953,
      "epoch": 5.5846015611838045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023736566676399307,
      "loss": 4.8219,
      "mean_token_accuracy": 0.23472279012203218,
      "num_tokens": 111010481.0,
      "step": 49725
    },
    {
      "entropy": 5.47678804397583,
      "epoch": 5.585163138091762,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002373260842517086,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2335107833147049,
      "num_tokens": 111020355.0,
      "step": 49730
    },
    {
      "entropy": 5.48454737663269,
      "epoch": 5.585724714999719,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002372865029389763,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23547870218753814,
      "num_tokens": 111030044.0,
      "step": 49735
    },
    {
      "entropy": 5.513519191741944,
      "epoch": 5.586286291907677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023724692282705635,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22722022086381913,
      "num_tokens": 111041988.0,
      "step": 49740
    },
    {
      "entropy": 5.467833185195923,
      "epoch": 5.586847868815634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000237207343917209,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22525978535413743,
      "num_tokens": 111052841.0,
      "step": 49745
    },
    {
      "entropy": 5.475788927078247,
      "epoch": 5.587409445723592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023716776621069446,
      "loss": 4.895,
      "mean_token_accuracy": 0.22507154792547227,
      "num_tokens": 111063892.0,
      "step": 49750
    },
    {
      "entropy": 5.468601179122925,
      "epoch": 5.5879710226315495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002371281897087729,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23347117602825165,
      "num_tokens": 111074859.0,
      "step": 49755
    },
    {
      "entropy": 5.527934455871582,
      "epoch": 5.588532599539507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023708861441270452,
      "loss": 4.8803,
      "mean_token_accuracy": 0.23418593257665635,
      "num_tokens": 111087345.0,
      "step": 49760
    },
    {
      "entropy": 5.460082864761352,
      "epoch": 5.589094176447465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023704904032374936,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23555031418800354,
      "num_tokens": 111098698.0,
      "step": 49765
    },
    {
      "entropy": 5.520084381103516,
      "epoch": 5.589655753355422,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023700946744316747,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23675990998744964,
      "num_tokens": 111108793.0,
      "step": 49770
    },
    {
      "entropy": 5.571384716033935,
      "epoch": 5.590217330263379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002369698957722189,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22444604933261872,
      "num_tokens": 111120933.0,
      "step": 49775
    },
    {
      "entropy": 5.507222318649292,
      "epoch": 5.590778907171337,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023693032531216352,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23540295362472535,
      "num_tokens": 111132257.0,
      "step": 49780
    },
    {
      "entropy": 5.554266500473022,
      "epoch": 5.591340484079295,
      "grad_norm": 1.0,
      "learning_rate": 0.00023689075606426143,
      "loss": 4.9182,
      "mean_token_accuracy": 0.2300567477941513,
      "num_tokens": 111143976.0,
      "step": 49785
    },
    {
      "entropy": 5.4975831508636475,
      "epoch": 5.591902060987252,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002368511880297724,
      "loss": 4.821,
      "mean_token_accuracy": 0.2383611962199211,
      "num_tokens": 111155404.0,
      "step": 49790
    },
    {
      "entropy": 5.43618106842041,
      "epoch": 5.59246363789521,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023681162120995624,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23194658160209655,
      "num_tokens": 111166565.0,
      "step": 49795
    },
    {
      "entropy": 5.403244733810425,
      "epoch": 5.593025214803167,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023677205560607294,
      "loss": 4.8269,
      "mean_token_accuracy": 0.22653231620788575,
      "num_tokens": 111177433.0,
      "step": 49800
    },
    {
      "entropy": 5.462850427627563,
      "epoch": 5.593586791711125,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023673249121938228,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22487872391939162,
      "num_tokens": 111189839.0,
      "step": 49805
    },
    {
      "entropy": 5.522143077850342,
      "epoch": 5.5941483686190825,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023669292805114395,
      "loss": 4.9515,
      "mean_token_accuracy": 0.23122264444828033,
      "num_tokens": 111200395.0,
      "step": 49810
    },
    {
      "entropy": 5.656291723251343,
      "epoch": 5.59470994552704,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023665336610261762,
      "loss": 5.0409,
      "mean_token_accuracy": 0.2207055062055588,
      "num_tokens": 111212741.0,
      "step": 49815
    },
    {
      "entropy": 5.506521892547608,
      "epoch": 5.595271522434998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000236613805375063,
      "loss": 4.921,
      "mean_token_accuracy": 0.22579880505800248,
      "num_tokens": 111224448.0,
      "step": 49820
    },
    {
      "entropy": 5.553407526016235,
      "epoch": 5.595833099342955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002365742458697397,
      "loss": 4.9252,
      "mean_token_accuracy": 0.23110237568616868,
      "num_tokens": 111236022.0,
      "step": 49825
    },
    {
      "entropy": 5.435010004043579,
      "epoch": 5.596394676250912,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023653468758790726,
      "loss": 4.7992,
      "mean_token_accuracy": 0.2339686334133148,
      "num_tokens": 111246735.0,
      "step": 49830
    },
    {
      "entropy": 5.396459579467773,
      "epoch": 5.59695625315887,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023649513053082528,
      "loss": 4.8096,
      "mean_token_accuracy": 0.24219998568296433,
      "num_tokens": 111258102.0,
      "step": 49835
    },
    {
      "entropy": 5.473571348190307,
      "epoch": 5.597517830066828,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002364555746997533,
      "loss": 4.8443,
      "mean_token_accuracy": 0.24006955772638322,
      "num_tokens": 111268935.0,
      "step": 49840
    },
    {
      "entropy": 5.465491008758545,
      "epoch": 5.598079406974785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023641602009595087,
      "loss": 4.8505,
      "mean_token_accuracy": 0.2305327907204628,
      "num_tokens": 111281234.0,
      "step": 49845
    },
    {
      "entropy": 5.5015675067901615,
      "epoch": 5.598640983882743,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002363764667206772,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23346953690052033,
      "num_tokens": 111293176.0,
      "step": 49850
    },
    {
      "entropy": 5.566422367095948,
      "epoch": 5.5992025607907,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023633691457519191,
      "loss": 4.942,
      "mean_token_accuracy": 0.23221469074487686,
      "num_tokens": 111304378.0,
      "step": 49855
    },
    {
      "entropy": 5.539281368255615,
      "epoch": 5.599764137698658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023629736366075416,
      "loss": 4.9007,
      "mean_token_accuracy": 0.233448788523674,
      "num_tokens": 111316045.0,
      "step": 49860
    },
    {
      "entropy": 5.486573076248169,
      "epoch": 5.600325714606615,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023625781397862333,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2379889503121376,
      "num_tokens": 111326576.0,
      "step": 49865
    },
    {
      "entropy": 5.445421838760376,
      "epoch": 5.600887291514573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023621826553005883,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23701934814453124,
      "num_tokens": 111337649.0,
      "step": 49870
    },
    {
      "entropy": 5.452908611297607,
      "epoch": 5.601448868422531,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023617871831631976,
      "loss": 4.8462,
      "mean_token_accuracy": 0.2342609390616417,
      "num_tokens": 111348656.0,
      "step": 49875
    },
    {
      "entropy": 5.475807619094849,
      "epoch": 5.602010445330488,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023613917233866538,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23804262280464172,
      "num_tokens": 111359737.0,
      "step": 49880
    },
    {
      "entropy": 5.506835222244263,
      "epoch": 5.602572022238445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002360996275983548,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2306846261024475,
      "num_tokens": 111372074.0,
      "step": 49885
    },
    {
      "entropy": 5.51295919418335,
      "epoch": 5.603133599146403,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002360600840966472,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2250254914164543,
      "num_tokens": 111383996.0,
      "step": 49890
    },
    {
      "entropy": 5.5430981636047365,
      "epoch": 5.6036951760543605,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002360205418348016,
      "loss": 4.9378,
      "mean_token_accuracy": 0.23110331892967223,
      "num_tokens": 111394781.0,
      "step": 49895
    },
    {
      "entropy": 5.4944580078125,
      "epoch": 5.604256752962318,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023598100081407704,
      "loss": 4.8509,
      "mean_token_accuracy": 0.23340267539024354,
      "num_tokens": 111406056.0,
      "step": 49900
    },
    {
      "entropy": 5.493177700042724,
      "epoch": 5.604818329870276,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023594146103573255,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22364925295114518,
      "num_tokens": 111417020.0,
      "step": 49905
    },
    {
      "entropy": 5.431745910644532,
      "epoch": 5.605379906778233,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002359019225010271,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22940314412117005,
      "num_tokens": 111428085.0,
      "step": 49910
    },
    {
      "entropy": 5.501915788650512,
      "epoch": 5.605941483686191,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002358623852112196,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2258833095431328,
      "num_tokens": 111438016.0,
      "step": 49915
    },
    {
      "entropy": 5.42492642402649,
      "epoch": 5.606503060594148,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023582284916756898,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22980997264385222,
      "num_tokens": 111449040.0,
      "step": 49920
    },
    {
      "entropy": 5.427049779891968,
      "epoch": 5.607064637502106,
      "grad_norm": 1.109375,
      "learning_rate": 0.000235783314371334,
      "loss": 4.7954,
      "mean_token_accuracy": 0.23948002308607103,
      "num_tokens": 111459732.0,
      "step": 49925
    },
    {
      "entropy": 5.540278911590576,
      "epoch": 5.607626214410064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023574378082377351,
      "loss": 4.8732,
      "mean_token_accuracy": 0.24085599333047866,
      "num_tokens": 111470564.0,
      "step": 49930
    },
    {
      "entropy": 5.547036743164062,
      "epoch": 5.608187791318021,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002357042485261463,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22541179805994033,
      "num_tokens": 111481839.0,
      "step": 49935
    },
    {
      "entropy": 5.41504545211792,
      "epoch": 5.608749368225979,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002356647174797109,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23095788806676865,
      "num_tokens": 111492616.0,
      "step": 49940
    },
    {
      "entropy": 5.452085542678833,
      "epoch": 5.609310945133936,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023562518768572628,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2360859513282776,
      "num_tokens": 111503707.0,
      "step": 49945
    },
    {
      "entropy": 5.520689678192139,
      "epoch": 5.6098725220418935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023558565914545094,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22421319633722306,
      "num_tokens": 111513395.0,
      "step": 49950
    },
    {
      "entropy": 5.494087743759155,
      "epoch": 5.610434098949852,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023554613186014352,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23964318335056306,
      "num_tokens": 111523300.0,
      "step": 49955
    },
    {
      "entropy": 5.5007147789001465,
      "epoch": 5.610995675857809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023550660583106253,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23303583711385728,
      "num_tokens": 111534510.0,
      "step": 49960
    },
    {
      "entropy": 5.684980201721191,
      "epoch": 5.611557252765766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023546708105946657,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2150729164481163,
      "num_tokens": 111546351.0,
      "step": 49965
    },
    {
      "entropy": 5.514159393310547,
      "epoch": 5.612118829673724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023542755754661404,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22896862775087357,
      "num_tokens": 111557972.0,
      "step": 49970
    },
    {
      "entropy": 5.444034290313721,
      "epoch": 5.612680406581681,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002353880352937634,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23582083731889725,
      "num_tokens": 111569710.0,
      "step": 49975
    },
    {
      "entropy": 5.541096305847168,
      "epoch": 5.6132419834896385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023534851430217314,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2296427458524704,
      "num_tokens": 111580506.0,
      "step": 49980
    },
    {
      "entropy": 5.481709289550781,
      "epoch": 5.613803560397597,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023530899457310157,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23487478345632554,
      "num_tokens": 111590866.0,
      "step": 49985
    },
    {
      "entropy": 5.52684383392334,
      "epoch": 5.614365137305554,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023526947610780696,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2339150294661522,
      "num_tokens": 111600770.0,
      "step": 49990
    },
    {
      "entropy": 5.472396183013916,
      "epoch": 5.614926714213512,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023522995890754768,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22815098911523818,
      "num_tokens": 111612232.0,
      "step": 49995
    },
    {
      "entropy": 5.512876749038696,
      "epoch": 5.615488291121469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002351904429735819,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2292010337114334,
      "num_tokens": 111622542.0,
      "step": 50000
    },
    {
      "entropy": 5.479127216339111,
      "epoch": 5.616049868029426,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002351509283071679,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22715042531490326,
      "num_tokens": 111635991.0,
      "step": 50005
    },
    {
      "entropy": 5.440141010284424,
      "epoch": 5.6166114449373845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002351114149095636,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23949972093105315,
      "num_tokens": 111646232.0,
      "step": 50010
    },
    {
      "entropy": 5.466266250610351,
      "epoch": 5.617173021845342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023507190278202755,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2388244703412056,
      "num_tokens": 111657132.0,
      "step": 50015
    },
    {
      "entropy": 5.471819686889648,
      "epoch": 5.617734598753299,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023503239192581755,
      "loss": 4.8541,
      "mean_token_accuracy": 0.237163645029068,
      "num_tokens": 111667817.0,
      "step": 50020
    },
    {
      "entropy": 5.577879142761231,
      "epoch": 5.618296175661257,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002349928823421917,
      "loss": 4.9941,
      "mean_token_accuracy": 0.22709568440914155,
      "num_tokens": 111679899.0,
      "step": 50025
    },
    {
      "entropy": 5.484348297119141,
      "epoch": 5.618857752569214,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023495337403240795,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2285285010933876,
      "num_tokens": 111691303.0,
      "step": 50030
    },
    {
      "entropy": 5.580715751647949,
      "epoch": 5.6194193294771715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023491386699772432,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2310867875814438,
      "num_tokens": 111701726.0,
      "step": 50035
    },
    {
      "entropy": 5.437283897399903,
      "epoch": 5.61998090638513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002348743612393987,
      "loss": 4.8217,
      "mean_token_accuracy": 0.24010191410779952,
      "num_tokens": 111712100.0,
      "step": 50040
    },
    {
      "entropy": 5.537159109115601,
      "epoch": 5.620542483293087,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023483485675868899,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22203425765037538,
      "num_tokens": 111723396.0,
      "step": 50045
    },
    {
      "entropy": 5.441746568679809,
      "epoch": 5.621104060201045,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023479535355685306,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2290516749024391,
      "num_tokens": 111733870.0,
      "step": 50050
    },
    {
      "entropy": 5.508936786651612,
      "epoch": 5.621665637109002,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023475585163514862,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2315238192677498,
      "num_tokens": 111745581.0,
      "step": 50055
    },
    {
      "entropy": 5.463988447189331,
      "epoch": 5.622227214016959,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002347163509948335,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23779028207063674,
      "num_tokens": 111756284.0,
      "step": 50060
    },
    {
      "entropy": 5.548948431015015,
      "epoch": 5.6227887909249175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023467685163716534,
      "loss": 4.9091,
      "mean_token_accuracy": 0.22860272526741027,
      "num_tokens": 111768018.0,
      "step": 50065
    },
    {
      "entropy": 5.556025218963623,
      "epoch": 5.623350367832875,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023463735356340194,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22562988400459288,
      "num_tokens": 111779455.0,
      "step": 50070
    },
    {
      "entropy": 5.547200775146484,
      "epoch": 5.623911944740832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023459785677480078,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2402143508195877,
      "num_tokens": 111790474.0,
      "step": 50075
    },
    {
      "entropy": 5.527429819107056,
      "epoch": 5.62447352164879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002345583612726196,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22501947730779648,
      "num_tokens": 111801392.0,
      "step": 50080
    },
    {
      "entropy": 5.425311803817749,
      "epoch": 5.625035098556747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023451886705811576,
      "loss": 4.7737,
      "mean_token_accuracy": 0.23906650841236116,
      "num_tokens": 111812872.0,
      "step": 50085
    },
    {
      "entropy": 5.611092138290405,
      "epoch": 5.625596675464704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000234479374132547,
      "loss": 5.0621,
      "mean_token_accuracy": 0.22328969985246658,
      "num_tokens": 111824538.0,
      "step": 50090
    },
    {
      "entropy": 5.467000818252563,
      "epoch": 5.6261582523726625,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023443988249717063,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2340780481696129,
      "num_tokens": 111835070.0,
      "step": 50095
    },
    {
      "entropy": 5.517300176620483,
      "epoch": 5.62671982928062,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023440039215324415,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22915091514587402,
      "num_tokens": 111845919.0,
      "step": 50100
    },
    {
      "entropy": 5.460065793991089,
      "epoch": 5.627281406188578,
      "grad_norm": 1.125,
      "learning_rate": 0.00023436090310202497,
      "loss": 4.7621,
      "mean_token_accuracy": 0.24143423587083818,
      "num_tokens": 111855996.0,
      "step": 50105
    },
    {
      "entropy": 5.590693759918213,
      "epoch": 5.627842983096535,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023432141534477036,
      "loss": 5.0457,
      "mean_token_accuracy": 0.2229519620537758,
      "num_tokens": 111867323.0,
      "step": 50110
    },
    {
      "entropy": 5.521003293991089,
      "epoch": 5.628404560004492,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023428192888273768,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2295302450656891,
      "num_tokens": 111878070.0,
      "step": 50115
    },
    {
      "entropy": 5.553288745880127,
      "epoch": 5.62896613691245,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023424244371718418,
      "loss": 4.9779,
      "mean_token_accuracy": 0.220841583609581,
      "num_tokens": 111890163.0,
      "step": 50120
    },
    {
      "entropy": 5.545065307617188,
      "epoch": 5.629527713820408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00023420295984936707,
      "loss": 4.9549,
      "mean_token_accuracy": 0.23267349153757094,
      "num_tokens": 111901329.0,
      "step": 50125
    },
    {
      "entropy": 5.501566410064697,
      "epoch": 5.630089290728366,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023416347728054355,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2337649866938591,
      "num_tokens": 111911737.0,
      "step": 50130
    },
    {
      "entropy": 5.555122947692871,
      "epoch": 5.630650867636323,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023412399601197072,
      "loss": 5.0476,
      "mean_token_accuracy": 0.22070082575082778,
      "num_tokens": 111923375.0,
      "step": 50135
    },
    {
      "entropy": 5.4636787414550785,
      "epoch": 5.63121244454428,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023408451604490577,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2312339186668396,
      "num_tokens": 111935304.0,
      "step": 50140
    },
    {
      "entropy": 5.538064098358154,
      "epoch": 5.631774021452238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023404503738060566,
      "loss": 4.9941,
      "mean_token_accuracy": 0.22262648195028306,
      "num_tokens": 111946773.0,
      "step": 50145
    },
    {
      "entropy": 5.531956529617309,
      "epoch": 5.6323355983601955,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023400556002032747,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22745631635189056,
      "num_tokens": 111958963.0,
      "step": 50150
    },
    {
      "entropy": 5.619626235961914,
      "epoch": 5.632897175268153,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000233966083965328,
      "loss": 4.9951,
      "mean_token_accuracy": 0.23629034012556077,
      "num_tokens": 111971249.0,
      "step": 50155
    },
    {
      "entropy": 5.5284487247467045,
      "epoch": 5.633458752176111,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023392660921686448,
      "loss": 4.9401,
      "mean_token_accuracy": 0.23138049691915513,
      "num_tokens": 111982783.0,
      "step": 50160
    },
    {
      "entropy": 5.374984073638916,
      "epoch": 5.634020329084068,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023388713577619357,
      "loss": 4.7375,
      "mean_token_accuracy": 0.23725682646036148,
      "num_tokens": 111993653.0,
      "step": 50165
    },
    {
      "entropy": 5.467186450958252,
      "epoch": 5.634581905992025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023384766364457227,
      "loss": 4.9398,
      "mean_token_accuracy": 0.23080079406499862,
      "num_tokens": 112004332.0,
      "step": 50170
    },
    {
      "entropy": 5.476924753189087,
      "epoch": 5.635143482899983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023380819282325732,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2237631067633629,
      "num_tokens": 112015480.0,
      "step": 50175
    },
    {
      "entropy": 5.587428617477417,
      "epoch": 5.635705059807941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023376872331350553,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22967132776975632,
      "num_tokens": 112026437.0,
      "step": 50180
    },
    {
      "entropy": 5.509723281860351,
      "epoch": 5.636266636715899,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002337292551165735,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23422599881887435,
      "num_tokens": 112038183.0,
      "step": 50185
    },
    {
      "entropy": 5.497280883789062,
      "epoch": 5.636828213623856,
      "grad_norm": 1.046875,
      "learning_rate": 0.000233689788233718,
      "loss": 4.9021,
      "mean_token_accuracy": 0.22033994793891906,
      "num_tokens": 112050466.0,
      "step": 50190
    },
    {
      "entropy": 5.508537673950196,
      "epoch": 5.637389790531813,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023365032266619567,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22289575338363649,
      "num_tokens": 112060818.0,
      "step": 50195
    },
    {
      "entropy": 5.4421913623809814,
      "epoch": 5.637951367439771,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023361085841526314,
      "loss": 4.86,
      "mean_token_accuracy": 0.2376294896006584,
      "num_tokens": 112070980.0,
      "step": 50200
    },
    {
      "entropy": 5.501559257507324,
      "epoch": 5.638512944347728,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023357139548217692,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22740039676427842,
      "num_tokens": 112082725.0,
      "step": 50205
    },
    {
      "entropy": 5.37853512763977,
      "epoch": 5.639074521255686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023353193386819353,
      "loss": 4.7389,
      "mean_token_accuracy": 0.2400473952293396,
      "num_tokens": 112093900.0,
      "step": 50210
    },
    {
      "entropy": 5.458428859710693,
      "epoch": 5.639636098163644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023349247357456942,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2302079677581787,
      "num_tokens": 112104626.0,
      "step": 50215
    },
    {
      "entropy": 5.5363147258758545,
      "epoch": 5.640197675071601,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023345301460256113,
      "loss": 4.9376,
      "mean_token_accuracy": 0.2288116455078125,
      "num_tokens": 112116383.0,
      "step": 50220
    },
    {
      "entropy": 5.415137434005738,
      "epoch": 5.640759251979558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002334135569534248,
      "loss": 4.7905,
      "mean_token_accuracy": 0.24236671924591063,
      "num_tokens": 112126645.0,
      "step": 50225
    },
    {
      "entropy": 5.5165143489837645,
      "epoch": 5.641320828887516,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023337410062841706,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22742310613393785,
      "num_tokens": 112138740.0,
      "step": 50230
    },
    {
      "entropy": 5.490833711624146,
      "epoch": 5.6418824057954735,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023333464562879408,
      "loss": 4.855,
      "mean_token_accuracy": 0.23352675437927245,
      "num_tokens": 112150037.0,
      "step": 50235
    },
    {
      "entropy": 5.522901821136474,
      "epoch": 5.642443982703432,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023329519195581218,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2340655192732811,
      "num_tokens": 112161006.0,
      "step": 50240
    },
    {
      "entropy": 5.4710546970367435,
      "epoch": 5.643005559611389,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023325573961072754,
      "loss": 4.7986,
      "mean_token_accuracy": 0.23643695563077927,
      "num_tokens": 112172571.0,
      "step": 50245
    },
    {
      "entropy": 5.412079620361328,
      "epoch": 5.643567136519346,
      "grad_norm": 1.0,
      "learning_rate": 0.00023321628859479632,
      "loss": 4.7577,
      "mean_token_accuracy": 0.23980863839387895,
      "num_tokens": 112183745.0,
      "step": 50250
    },
    {
      "entropy": 5.486238241195679,
      "epoch": 5.644128713427304,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002331768389092747,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23013363182544708,
      "num_tokens": 112195200.0,
      "step": 50255
    },
    {
      "entropy": 5.555466222763061,
      "epoch": 5.644690290335261,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023313739055541878,
      "loss": 5.0695,
      "mean_token_accuracy": 0.21773400902748108,
      "num_tokens": 112207295.0,
      "step": 50260
    },
    {
      "entropy": 5.510489130020142,
      "epoch": 5.645251867243219,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002330979435344845,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22805554568767547,
      "num_tokens": 112218346.0,
      "step": 50265
    },
    {
      "entropy": 5.4937985897064205,
      "epoch": 5.645813444151177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023305849784772799,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22734954506158828,
      "num_tokens": 112228876.0,
      "step": 50270
    },
    {
      "entropy": 5.530193710327149,
      "epoch": 5.646375021059134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002330190534964051,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22649425268173218,
      "num_tokens": 112240643.0,
      "step": 50275
    },
    {
      "entropy": 5.455701017379761,
      "epoch": 5.646936597967091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023297961048177195,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22884737849235534,
      "num_tokens": 112251861.0,
      "step": 50280
    },
    {
      "entropy": 5.542558097839356,
      "epoch": 5.647498174875049,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023294016880508428,
      "loss": 4.9357,
      "mean_token_accuracy": 0.23589516431093216,
      "num_tokens": 112262948.0,
      "step": 50285
    },
    {
      "entropy": 5.540580177307129,
      "epoch": 5.6480597517830065,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023290072846759784,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2225437730550766,
      "num_tokens": 112274332.0,
      "step": 50290
    },
    {
      "entropy": 5.561758899688721,
      "epoch": 5.648621328690965,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023286128947056867,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22809417694807052,
      "num_tokens": 112286241.0,
      "step": 50295
    },
    {
      "entropy": 5.456953287124634,
      "epoch": 5.649182905598922,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023282185181525223,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2306885302066803,
      "num_tokens": 112297179.0,
      "step": 50300
    },
    {
      "entropy": 5.515915584564209,
      "epoch": 5.649744482506879,
      "grad_norm": 1.125,
      "learning_rate": 0.00023278241550290453,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23196350634098054,
      "num_tokens": 112307951.0,
      "step": 50305
    },
    {
      "entropy": 5.47362904548645,
      "epoch": 5.650306059414837,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023274298053478106,
      "loss": 4.857,
      "mean_token_accuracy": 0.22951384335756303,
      "num_tokens": 112319376.0,
      "step": 50310
    },
    {
      "entropy": 5.594174146652222,
      "epoch": 5.650867636322794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023270354691213742,
      "loss": 4.9568,
      "mean_token_accuracy": 0.23116144090890883,
      "num_tokens": 112330777.0,
      "step": 50315
    },
    {
      "entropy": 5.570613431930542,
      "epoch": 5.651429213230752,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023266411463622927,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2302822142839432,
      "num_tokens": 112340551.0,
      "step": 50320
    },
    {
      "entropy": 5.522701597213745,
      "epoch": 5.65199079013871,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002326246837083122,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2309802770614624,
      "num_tokens": 112353012.0,
      "step": 50325
    },
    {
      "entropy": 5.4826802730560305,
      "epoch": 5.652552367046667,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002325852541296416,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23294901996850967,
      "num_tokens": 112364031.0,
      "step": 50330
    },
    {
      "entropy": 5.513966131210327,
      "epoch": 5.653113943954625,
      "grad_norm": 1.0,
      "learning_rate": 0.00023254582590147288,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23439750224351882,
      "num_tokens": 112375421.0,
      "step": 50335
    },
    {
      "entropy": 5.511637830734253,
      "epoch": 5.653675520862582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002325063990250616,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22852743268013,
      "num_tokens": 112388023.0,
      "step": 50340
    },
    {
      "entropy": 5.406373023986816,
      "epoch": 5.654237097770539,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023246697350166306,
      "loss": 4.8152,
      "mean_token_accuracy": 0.23950564861297607,
      "num_tokens": 112397464.0,
      "step": 50345
    },
    {
      "entropy": 5.515755748748779,
      "epoch": 5.6547986746784975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023242754933253256,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2308337315917015,
      "num_tokens": 112408252.0,
      "step": 50350
    },
    {
      "entropy": 5.4481395244598385,
      "epoch": 5.655360251586455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002323881265189255,
      "loss": 4.814,
      "mean_token_accuracy": 0.23427259773015977,
      "num_tokens": 112419043.0,
      "step": 50355
    },
    {
      "entropy": 5.472545862197876,
      "epoch": 5.655921828494412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023234870506209693,
      "loss": 4.909,
      "mean_token_accuracy": 0.23392949998378754,
      "num_tokens": 112429488.0,
      "step": 50360
    },
    {
      "entropy": 5.519149303436279,
      "epoch": 5.65648340540237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002323092849633021,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2217687711119652,
      "num_tokens": 112441937.0,
      "step": 50365
    },
    {
      "entropy": 5.48123254776001,
      "epoch": 5.657044982310327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023226986622379615,
      "loss": 4.8369,
      "mean_token_accuracy": 0.2305375039577484,
      "num_tokens": 112453100.0,
      "step": 50370
    },
    {
      "entropy": 5.569051933288574,
      "epoch": 5.657606559218285,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023223044884483436,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2323813259601593,
      "num_tokens": 112462691.0,
      "step": 50375
    },
    {
      "entropy": 5.454617214202881,
      "epoch": 5.658168136126243,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023219103282767157,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23833703845739365,
      "num_tokens": 112473390.0,
      "step": 50380
    },
    {
      "entropy": 5.485606575012207,
      "epoch": 5.6587297130342,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000232151618173563,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2387339785695076,
      "num_tokens": 112482976.0,
      "step": 50385
    },
    {
      "entropy": 5.584595632553101,
      "epoch": 5.659291289942158,
      "grad_norm": 1.125,
      "learning_rate": 0.0002321122048837635,
      "loss": 5.0569,
      "mean_token_accuracy": 0.22811416834592818,
      "num_tokens": 112493565.0,
      "step": 50390
    },
    {
      "entropy": 5.529077672958374,
      "epoch": 5.659852866850115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023207279295952798,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22602058947086334,
      "num_tokens": 112504794.0,
      "step": 50395
    },
    {
      "entropy": 5.460667514801026,
      "epoch": 5.660414443758072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023203338240211142,
      "loss": 4.7938,
      "mean_token_accuracy": 0.23869824558496475,
      "num_tokens": 112515561.0,
      "step": 50400
    },
    {
      "entropy": 5.554173231124878,
      "epoch": 5.6609760206660305,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002319939732127686,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23257484585046767,
      "num_tokens": 112525534.0,
      "step": 50405
    },
    {
      "entropy": 5.467323303222656,
      "epoch": 5.661537597573988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023195456539275433,
      "loss": 4.8447,
      "mean_token_accuracy": 0.23361100852489472,
      "num_tokens": 112536790.0,
      "step": 50410
    },
    {
      "entropy": 5.5524131774902346,
      "epoch": 5.662099174481945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023191515894332345,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22807469218969345,
      "num_tokens": 112548613.0,
      "step": 50415
    },
    {
      "entropy": 5.567813396453857,
      "epoch": 5.662660751389903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023187575386573057,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22914978265762329,
      "num_tokens": 112558709.0,
      "step": 50420
    },
    {
      "entropy": 5.58377857208252,
      "epoch": 5.66322232829786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023183635016123044,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22836301624774932,
      "num_tokens": 112569850.0,
      "step": 50425
    },
    {
      "entropy": 5.445266914367676,
      "epoch": 5.663783905205818,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002317969478310777,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23841941803693772,
      "num_tokens": 112580842.0,
      "step": 50430
    },
    {
      "entropy": 5.430749225616455,
      "epoch": 5.6643454821137755,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023175754687652684,
      "loss": 4.7482,
      "mean_token_accuracy": 0.23693769425153732,
      "num_tokens": 112590879.0,
      "step": 50435
    },
    {
      "entropy": 5.411010885238648,
      "epoch": 5.664907059021733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002317181472988325,
      "loss": 4.8031,
      "mean_token_accuracy": 0.2378861278295517,
      "num_tokens": 112602327.0,
      "step": 50440
    },
    {
      "entropy": 5.428445148468017,
      "epoch": 5.665468635929691,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023167874909924902,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2320520967245102,
      "num_tokens": 112613406.0,
      "step": 50445
    },
    {
      "entropy": 5.450534009933472,
      "epoch": 5.666030212837648,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002316393522790311,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2335364043712616,
      "num_tokens": 112624286.0,
      "step": 50450
    },
    {
      "entropy": 5.513337135314941,
      "epoch": 5.666591789745605,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023159995683943297,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2348212406039238,
      "num_tokens": 112635097.0,
      "step": 50455
    },
    {
      "entropy": 5.520682621002197,
      "epoch": 5.667153366653563,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023156056278170905,
      "loss": 4.929,
      "mean_token_accuracy": 0.23321413695812226,
      "num_tokens": 112645994.0,
      "step": 50460
    },
    {
      "entropy": 5.582180118560791,
      "epoch": 5.667714943561521,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023152117010711372,
      "loss": 4.979,
      "mean_token_accuracy": 0.22776993215084076,
      "num_tokens": 112657902.0,
      "step": 50465
    },
    {
      "entropy": 5.459563779830932,
      "epoch": 5.668276520469478,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023148177881690112,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2371952012181282,
      "num_tokens": 112668495.0,
      "step": 50470
    },
    {
      "entropy": 5.481645488739014,
      "epoch": 5.668838097377436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002314423889123256,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22457174211740494,
      "num_tokens": 112679314.0,
      "step": 50475
    },
    {
      "entropy": 5.501887512207031,
      "epoch": 5.669399674285393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023140300039464125,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22815414071083068,
      "num_tokens": 112689114.0,
      "step": 50480
    },
    {
      "entropy": 5.533621501922608,
      "epoch": 5.669961251193351,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023136361326510236,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22803475856781005,
      "num_tokens": 112701548.0,
      "step": 50485
    },
    {
      "entropy": 5.600251340866089,
      "epoch": 5.6705228281013085,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002313242275249629,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22704421132802963,
      "num_tokens": 112714576.0,
      "step": 50490
    },
    {
      "entropy": 5.445908546447754,
      "epoch": 5.671084405009266,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023128484317547698,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23667290657758713,
      "num_tokens": 112725835.0,
      "step": 50495
    },
    {
      "entropy": 5.502953958511353,
      "epoch": 5.671645981917224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023124546021789866,
      "loss": 4.8959,
      "mean_token_accuracy": 0.22774239927530288,
      "num_tokens": 112737525.0,
      "step": 50500
    },
    {
      "entropy": 5.526046848297119,
      "epoch": 5.672207558825181,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002312060786534818,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2235347166657448,
      "num_tokens": 112750608.0,
      "step": 50505
    },
    {
      "entropy": 5.5390173435211185,
      "epoch": 5.672769135733139,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023116669848348043,
      "loss": 4.9769,
      "mean_token_accuracy": 0.23157512843608857,
      "num_tokens": 112761721.0,
      "step": 50510
    },
    {
      "entropy": 5.549457597732544,
      "epoch": 5.673330712641096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023112731970914823,
      "loss": 4.9757,
      "mean_token_accuracy": 0.227366641163826,
      "num_tokens": 112772797.0,
      "step": 50515
    },
    {
      "entropy": 5.520731973648071,
      "epoch": 5.673892289549054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002310879423317393,
      "loss": 4.9437,
      "mean_token_accuracy": 0.2255515933036804,
      "num_tokens": 112784433.0,
      "step": 50520
    },
    {
      "entropy": 5.504932832717896,
      "epoch": 5.674453866457012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023104856635250732,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2259750857949257,
      "num_tokens": 112795340.0,
      "step": 50525
    },
    {
      "entropy": 5.504950523376465,
      "epoch": 5.675015443364969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023100919177270607,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22894713133573533,
      "num_tokens": 112806779.0,
      "step": 50530
    },
    {
      "entropy": 5.540673494338989,
      "epoch": 5.675577020272926,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023096981859358918,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22619082480669023,
      "num_tokens": 112816844.0,
      "step": 50535
    },
    {
      "entropy": 5.41780309677124,
      "epoch": 5.676138597180884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023093044681641035,
      "loss": 4.7967,
      "mean_token_accuracy": 0.2340952128171921,
      "num_tokens": 112827647.0,
      "step": 50540
    },
    {
      "entropy": 5.387545537948609,
      "epoch": 5.676700174088841,
      "grad_norm": 1.0,
      "learning_rate": 0.0002308910764424232,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2355108380317688,
      "num_tokens": 112838658.0,
      "step": 50545
    },
    {
      "entropy": 5.464171838760376,
      "epoch": 5.677261750996799,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002308517074728812,
      "loss": 4.834,
      "mean_token_accuracy": 0.23763380199670792,
      "num_tokens": 112850093.0,
      "step": 50550
    },
    {
      "entropy": 5.370126819610595,
      "epoch": 5.677823327904757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023081233990903798,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2375270515680313,
      "num_tokens": 112860269.0,
      "step": 50555
    },
    {
      "entropy": 5.46386981010437,
      "epoch": 5.678384904812714,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023077297375214708,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23963345289230348,
      "num_tokens": 112869904.0,
      "step": 50560
    },
    {
      "entropy": 5.507802724838257,
      "epoch": 5.678946481720672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002307336090034618,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23669720739126204,
      "num_tokens": 112881384.0,
      "step": 50565
    },
    {
      "entropy": 5.467457485198975,
      "epoch": 5.679508058628629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023069424566423557,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23078600615262984,
      "num_tokens": 112891873.0,
      "step": 50570
    },
    {
      "entropy": 5.4927692890167235,
      "epoch": 5.6800696355365865,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023065488373572173,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2351877361536026,
      "num_tokens": 112903579.0,
      "step": 50575
    },
    {
      "entropy": 5.438537073135376,
      "epoch": 5.680631212444545,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023061552321917368,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23178581446409224,
      "num_tokens": 112913798.0,
      "step": 50580
    },
    {
      "entropy": 5.494939947128296,
      "epoch": 5.681192789352502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023057616411584453,
      "loss": 4.9184,
      "mean_token_accuracy": 0.2297137513756752,
      "num_tokens": 112924442.0,
      "step": 50585
    },
    {
      "entropy": 5.45796275138855,
      "epoch": 5.681754366260459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023053680642698743,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23240946978330612,
      "num_tokens": 112935347.0,
      "step": 50590
    },
    {
      "entropy": 5.443364095687866,
      "epoch": 5.682315943168417,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023049745015385576,
      "loss": 4.8225,
      "mean_token_accuracy": 0.2322368174791336,
      "num_tokens": 112945727.0,
      "step": 50595
    },
    {
      "entropy": 5.50997838973999,
      "epoch": 5.682877520076374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023045809529770252,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23061476051807403,
      "num_tokens": 112957998.0,
      "step": 50600
    },
    {
      "entropy": 5.375188970565796,
      "epoch": 5.683439096984332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023041874185978086,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23284327387809753,
      "num_tokens": 112969660.0,
      "step": 50605
    },
    {
      "entropy": 5.460199642181396,
      "epoch": 5.68400067389229,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002303793898413437,
      "loss": 4.9464,
      "mean_token_accuracy": 0.23123643398284913,
      "num_tokens": 112980996.0,
      "step": 50610
    },
    {
      "entropy": 5.427250576019287,
      "epoch": 5.684562250800247,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002303400392436441,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23921972662210464,
      "num_tokens": 112991200.0,
      "step": 50615
    },
    {
      "entropy": 5.528145027160645,
      "epoch": 5.685123827708205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002303006900679349,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22923928648233413,
      "num_tokens": 113003097.0,
      "step": 50620
    },
    {
      "entropy": 5.435457038879394,
      "epoch": 5.685685404616162,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023026134231546907,
      "loss": 4.8145,
      "mean_token_accuracy": 0.2316787362098694,
      "num_tokens": 113013767.0,
      "step": 50625
    },
    {
      "entropy": 5.557383346557617,
      "epoch": 5.6862469815241194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023022199598749955,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22781036347150802,
      "num_tokens": 113024731.0,
      "step": 50630
    },
    {
      "entropy": 5.425935125350952,
      "epoch": 5.6868085584320776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023018265108527903,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23576428443193437,
      "num_tokens": 113036289.0,
      "step": 50635
    },
    {
      "entropy": 5.511752891540527,
      "epoch": 5.687370135340035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023014330761006018,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22755714654922485,
      "num_tokens": 113047968.0,
      "step": 50640
    },
    {
      "entropy": 5.437666988372802,
      "epoch": 5.687931712247992,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023010396556309592,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2401668682694435,
      "num_tokens": 113059687.0,
      "step": 50645
    },
    {
      "entropy": 5.488844585418701,
      "epoch": 5.68849328915595,
      "grad_norm": 1.125,
      "learning_rate": 0.0002300646249456388,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22919604927301407,
      "num_tokens": 113071094.0,
      "step": 50650
    },
    {
      "entropy": 5.5026466846466064,
      "epoch": 5.689054866063907,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023002528575894138,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2426581159234047,
      "num_tokens": 113081600.0,
      "step": 50655
    },
    {
      "entropy": 5.4356523036956785,
      "epoch": 5.6896164429718645,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022998594800425627,
      "loss": 4.7607,
      "mean_token_accuracy": 0.23728619664907455,
      "num_tokens": 113092176.0,
      "step": 50660
    },
    {
      "entropy": 5.486303377151489,
      "epoch": 5.690178019879823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022994661168283608,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22854133248329161,
      "num_tokens": 113102745.0,
      "step": 50665
    },
    {
      "entropy": 5.559822082519531,
      "epoch": 5.69073959678778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022990727679593332,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2240733176469803,
      "num_tokens": 113115977.0,
      "step": 50670
    },
    {
      "entropy": 5.509330463409424,
      "epoch": 5.691301173695738,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022986794334480033,
      "loss": 4.9531,
      "mean_token_accuracy": 0.23182110041379927,
      "num_tokens": 113127089.0,
      "step": 50675
    },
    {
      "entropy": 5.505190992355347,
      "epoch": 5.691862750603695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002298286113306895,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23481871336698532,
      "num_tokens": 113137100.0,
      "step": 50680
    },
    {
      "entropy": 5.546192741394043,
      "epoch": 5.692424327511652,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022978928075485322,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23476732522249222,
      "num_tokens": 113148636.0,
      "step": 50685
    },
    {
      "entropy": 5.607618236541748,
      "epoch": 5.6929859044196105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002297499516185438,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2131972372531891,
      "num_tokens": 113160806.0,
      "step": 50690
    },
    {
      "entropy": 5.5632445335388185,
      "epoch": 5.693547481327568,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022971062392301335,
      "loss": 4.9877,
      "mean_token_accuracy": 0.22452317625284196,
      "num_tokens": 113172317.0,
      "step": 50695
    },
    {
      "entropy": 5.547025060653686,
      "epoch": 5.694109058235526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022967129766951429,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23799082189798354,
      "num_tokens": 113183325.0,
      "step": 50700
    },
    {
      "entropy": 5.505047178268432,
      "epoch": 5.694670635143483,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002296319728592987,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22907016277313233,
      "num_tokens": 113194897.0,
      "step": 50705
    },
    {
      "entropy": 5.487588214874267,
      "epoch": 5.69523221205144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002295926494936187,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22876883894205094,
      "num_tokens": 113205879.0,
      "step": 50710
    },
    {
      "entropy": 5.540397071838379,
      "epoch": 5.695793788959398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022955332757372633,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22703747898340226,
      "num_tokens": 113217101.0,
      "step": 50715
    },
    {
      "entropy": 5.499839639663696,
      "epoch": 5.696355365867356,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022951400710087368,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2255971610546112,
      "num_tokens": 113228264.0,
      "step": 50720
    },
    {
      "entropy": 5.499638843536377,
      "epoch": 5.696916942775313,
      "grad_norm": 1.0,
      "learning_rate": 0.0002294746880763126,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2263019546866417,
      "num_tokens": 113239246.0,
      "step": 50725
    },
    {
      "entropy": 5.546593904495239,
      "epoch": 5.697478519683271,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022943537050129504,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22767907679080962,
      "num_tokens": 113251761.0,
      "step": 50730
    },
    {
      "entropy": 5.506041479110718,
      "epoch": 5.698040096591228,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002293960543770731,
      "loss": 4.8747,
      "mean_token_accuracy": 0.23351125419139862,
      "num_tokens": 113262642.0,
      "step": 50735
    },
    {
      "entropy": 5.465931701660156,
      "epoch": 5.698601673499185,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022935673970489847,
      "loss": 4.896,
      "mean_token_accuracy": 0.23722633421421052,
      "num_tokens": 113272632.0,
      "step": 50740
    },
    {
      "entropy": 5.421657991409302,
      "epoch": 5.699163250407143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022931742648602294,
      "loss": 4.7818,
      "mean_token_accuracy": 0.24169245809316636,
      "num_tokens": 113283187.0,
      "step": 50745
    },
    {
      "entropy": 5.411830902099609,
      "epoch": 5.699724827315101,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022927811472169823,
      "loss": 4.861,
      "mean_token_accuracy": 0.23720872849226,
      "num_tokens": 113293195.0,
      "step": 50750
    },
    {
      "entropy": 5.477192449569702,
      "epoch": 5.700286404223059,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022923880441317606,
      "loss": 4.87,
      "mean_token_accuracy": 0.22748952656984328,
      "num_tokens": 113304238.0,
      "step": 50755
    },
    {
      "entropy": 5.487424659729004,
      "epoch": 5.700847981131016,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002291994955617082,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23380476981401443,
      "num_tokens": 113314688.0,
      "step": 50760
    },
    {
      "entropy": 5.505200338363648,
      "epoch": 5.701409558038973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002291601881685461,
      "loss": 4.9029,
      "mean_token_accuracy": 0.23383183479309083,
      "num_tokens": 113326402.0,
      "step": 50765
    },
    {
      "entropy": 5.4913019180297855,
      "epoch": 5.701971134946931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022912088223494138,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2284735068678856,
      "num_tokens": 113338265.0,
      "step": 50770
    },
    {
      "entropy": 5.44565978050232,
      "epoch": 5.7025327118548885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022908157776214556,
      "loss": 4.7971,
      "mean_token_accuracy": 0.2315584048628807,
      "num_tokens": 113348079.0,
      "step": 50775
    },
    {
      "entropy": 5.4605255126953125,
      "epoch": 5.703094288762846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022904227475141015,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2312790036201477,
      "num_tokens": 113359545.0,
      "step": 50780
    },
    {
      "entropy": 5.48437728881836,
      "epoch": 5.703655865670804,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022900297320398654,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23269954919815064,
      "num_tokens": 113370336.0,
      "step": 50785
    },
    {
      "entropy": 5.432611894607544,
      "epoch": 5.704217442578761,
      "grad_norm": 1.0,
      "learning_rate": 0.00022896367312112608,
      "loss": 4.8008,
      "mean_token_accuracy": 0.23537644147872924,
      "num_tokens": 113380850.0,
      "step": 50790
    },
    {
      "entropy": 5.515552663803101,
      "epoch": 5.704779019486718,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022892437450408015,
      "loss": 4.9076,
      "mean_token_accuracy": 0.22663138061761856,
      "num_tokens": 113392018.0,
      "step": 50795
    },
    {
      "entropy": 5.572317171096802,
      "epoch": 5.705340596394676,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022888507735409997,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2240215927362442,
      "num_tokens": 113403024.0,
      "step": 50800
    },
    {
      "entropy": 5.4788049221038815,
      "epoch": 5.705902173302634,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022884578167243675,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2325004979968071,
      "num_tokens": 113413602.0,
      "step": 50805
    },
    {
      "entropy": 5.461485767364502,
      "epoch": 5.706463750210592,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022880648746034183,
      "loss": 4.8643,
      "mean_token_accuracy": 0.2360743448138237,
      "num_tokens": 113423577.0,
      "step": 50810
    },
    {
      "entropy": 5.577311372756958,
      "epoch": 5.707025327118549,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022876719471906626,
      "loss": 4.9382,
      "mean_token_accuracy": 0.23224791288375854,
      "num_tokens": 113435282.0,
      "step": 50815
    },
    {
      "entropy": 5.446901750564575,
      "epoch": 5.707586904026506,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022872790344986117,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23691509217023848,
      "num_tokens": 113445950.0,
      "step": 50820
    },
    {
      "entropy": 5.508507871627808,
      "epoch": 5.708148480934464,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022868861365397752,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2263361006975174,
      "num_tokens": 113457825.0,
      "step": 50825
    },
    {
      "entropy": 5.504951095581054,
      "epoch": 5.7087100578424215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022864932533266642,
      "loss": 4.9017,
      "mean_token_accuracy": 0.228178933262825,
      "num_tokens": 113467890.0,
      "step": 50830
    },
    {
      "entropy": 5.482863712310791,
      "epoch": 5.709271634750379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022861003848717877,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22767933011054992,
      "num_tokens": 113480140.0,
      "step": 50835
    },
    {
      "entropy": 5.516789150238037,
      "epoch": 5.709833211658337,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022857075311876542,
      "loss": 4.9375,
      "mean_token_accuracy": 0.23040421307086945,
      "num_tokens": 113492888.0,
      "step": 50840
    },
    {
      "entropy": 5.525243473052979,
      "epoch": 5.710394788566294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022853146922867734,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2221596673130989,
      "num_tokens": 113505087.0,
      "step": 50845
    },
    {
      "entropy": 5.463366603851318,
      "epoch": 5.710956365474251,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002284921868181653,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2277344286441803,
      "num_tokens": 113516351.0,
      "step": 50850
    },
    {
      "entropy": 5.584154748916626,
      "epoch": 5.711517942382209,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022845290588848012,
      "loss": 5.0145,
      "mean_token_accuracy": 0.22660399228334427,
      "num_tokens": 113528959.0,
      "step": 50855
    },
    {
      "entropy": 5.455470705032349,
      "epoch": 5.7120795192901666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022841362644087244,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2366852954030037,
      "num_tokens": 113539445.0,
      "step": 50860
    },
    {
      "entropy": 5.634235191345215,
      "epoch": 5.712641096198125,
      "grad_norm": 1.0,
      "learning_rate": 0.00022837434847659295,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22851764261722565,
      "num_tokens": 113551293.0,
      "step": 50865
    },
    {
      "entropy": 5.482493352890015,
      "epoch": 5.713202673106082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002283350719968923,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23368156999349593,
      "num_tokens": 113560838.0,
      "step": 50870
    },
    {
      "entropy": 5.469582176208496,
      "epoch": 5.713764250014039,
      "grad_norm": 1.078125,
      "learning_rate": 0.000228295797003021,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23732036650180816,
      "num_tokens": 113571682.0,
      "step": 50875
    },
    {
      "entropy": 5.448435258865357,
      "epoch": 5.714325826921997,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022825652349622966,
      "loss": 4.8471,
      "mean_token_accuracy": 0.2351885676383972,
      "num_tokens": 113581612.0,
      "step": 50880
    },
    {
      "entropy": 5.539811658859253,
      "epoch": 5.714887403829954,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022821725147776877,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21997977048158646,
      "num_tokens": 113592674.0,
      "step": 50885
    },
    {
      "entropy": 5.52232346534729,
      "epoch": 5.7154489807379125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022817798094888875,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22376889884471893,
      "num_tokens": 113603752.0,
      "step": 50890
    },
    {
      "entropy": 5.495805168151856,
      "epoch": 5.71601055764587,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022813871191083995,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2333597406744957,
      "num_tokens": 113614939.0,
      "step": 50895
    },
    {
      "entropy": 5.56296124458313,
      "epoch": 5.716572134553827,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022809944436487278,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2358602911233902,
      "num_tokens": 113626153.0,
      "step": 50900
    },
    {
      "entropy": 5.513474225997925,
      "epoch": 5.717133711461785,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022806017831223746,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23710020780563354,
      "num_tokens": 113635846.0,
      "step": 50905
    },
    {
      "entropy": 5.415366125106812,
      "epoch": 5.717695288369742,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022802091375418422,
      "loss": 4.8177,
      "mean_token_accuracy": 0.23830520510673522,
      "num_tokens": 113647168.0,
      "step": 50910
    },
    {
      "entropy": 5.446756315231323,
      "epoch": 5.7182568652776995,
      "grad_norm": 1.125,
      "learning_rate": 0.00022798165069196335,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23749361038208008,
      "num_tokens": 113658613.0,
      "step": 50915
    },
    {
      "entropy": 5.523759412765503,
      "epoch": 5.718818442185658,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022794238912682497,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2330455556511879,
      "num_tokens": 113670491.0,
      "step": 50920
    },
    {
      "entropy": 5.541067695617675,
      "epoch": 5.719380019093615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022790312906001919,
      "loss": 4.877,
      "mean_token_accuracy": 0.23304493576288224,
      "num_tokens": 113681321.0,
      "step": 50925
    },
    {
      "entropy": 5.470807123184204,
      "epoch": 5.719941596001572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022786387049279606,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23189148157835007,
      "num_tokens": 113692428.0,
      "step": 50930
    },
    {
      "entropy": 5.480187177658081,
      "epoch": 5.72050317290953,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022782461342640554,
      "loss": 4.8341,
      "mean_token_accuracy": 0.24262651652097703,
      "num_tokens": 113703056.0,
      "step": 50935
    },
    {
      "entropy": 5.432170248031616,
      "epoch": 5.721064749817487,
      "grad_norm": 1.171875,
      "learning_rate": 0.00022778535786209758,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23368174880743026,
      "num_tokens": 113713545.0,
      "step": 50940
    },
    {
      "entropy": 5.4730414867401125,
      "epoch": 5.7216263267254455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002277461038011222,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2292021170258522,
      "num_tokens": 113724952.0,
      "step": 50945
    },
    {
      "entropy": 5.588611268997193,
      "epoch": 5.722187903633403,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022770685124472907,
      "loss": 5.0396,
      "mean_token_accuracy": 0.22379574924707413,
      "num_tokens": 113736378.0,
      "step": 50950
    },
    {
      "entropy": 5.435250520706177,
      "epoch": 5.72274948054136,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00022766760019416826,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22945650219917296,
      "num_tokens": 113748192.0,
      "step": 50955
    },
    {
      "entropy": 5.4714967727661135,
      "epoch": 5.723311057449318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022762835065068943,
      "loss": 4.9123,
      "mean_token_accuracy": 0.23065434992313386,
      "num_tokens": 113759448.0,
      "step": 50960
    },
    {
      "entropy": 5.4828534603118895,
      "epoch": 5.723872634357275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022758910261554228,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23330341875553132,
      "num_tokens": 113771216.0,
      "step": 50965
    },
    {
      "entropy": 5.446164464950561,
      "epoch": 5.724434211265232,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022754985608997648,
      "loss": 4.8324,
      "mean_token_accuracy": 0.23369636237621308,
      "num_tokens": 113782978.0,
      "step": 50970
    },
    {
      "entropy": 5.440762996673584,
      "epoch": 5.7249957881731905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022751061107524167,
      "loss": 4.8026,
      "mean_token_accuracy": 0.232102732360363,
      "num_tokens": 113793197.0,
      "step": 50975
    },
    {
      "entropy": 5.490477609634399,
      "epoch": 5.725557365081148,
      "grad_norm": 1.109375,
      "learning_rate": 0.00022747136757258742,
      "loss": 4.9,
      "mean_token_accuracy": 0.2335611715912819,
      "num_tokens": 113804336.0,
      "step": 50980
    },
    {
      "entropy": 5.4793774604797365,
      "epoch": 5.726118941989105,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002274321255832632,
      "loss": 4.912,
      "mean_token_accuracy": 0.23156969696283342,
      "num_tokens": 113814789.0,
      "step": 50985
    },
    {
      "entropy": 5.584000730514527,
      "epoch": 5.726680518897063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022739288510851863,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22840880304574968,
      "num_tokens": 113825741.0,
      "step": 50990
    },
    {
      "entropy": 5.565668344497681,
      "epoch": 5.72724209580502,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022735364614960302,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22623057663440704,
      "num_tokens": 113837099.0,
      "step": 50995
    },
    {
      "entropy": 5.551448106765747,
      "epoch": 5.727803672712978,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022731440870776582,
      "loss": 4.9044,
      "mean_token_accuracy": 0.2309229329228401,
      "num_tokens": 113848448.0,
      "step": 51000
    },
    {
      "epoch": 5.727803672712978,
      "eval_entropy": 5.3191273011342135,
      "eval_loss": 5.100851058959961,
      "eval_mean_token_accuracy": 0.2268330312271785,
      "eval_num_tokens": 113848448.0,
      "eval_runtime": 53.2467,
      "eval_samples_per_second": 582.44,
      "eval_steps_per_second": 72.812,
      "step": 51000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.78399368078336e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}