{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.064693659796709,
  "eval_steps": 3000,
  "global_step": 54000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742550659179688,
      "epoch": 0.0005615769079575448,
      "grad_norm": 5.46875,
      "learning_rate": 2e-06,
      "loss": 10.7833,
      "mean_token_accuracy": 9.587727836333216e-05,
      "num_tokens": 11659.0,
      "step": 5
    },
    {
      "entropy": 10.742562961578368,
      "epoch": 0.0011231538159150896,
      "grad_norm": 5.96875,
      "learning_rate": 4.5e-06,
      "loss": 10.7559,
      "mean_token_accuracy": 0.0,
      "num_tokens": 21794.0,
      "step": 10
    },
    {
      "entropy": 10.742598533630371,
      "epoch": 0.0016847307238726343,
      "grad_norm": 5.78125,
      "learning_rate": 7e-06,
      "loss": 10.7489,
      "mean_token_accuracy": 6.915629492141307e-05,
      "num_tokens": 33206.0,
      "step": 15
    },
    {
      "entropy": 10.742581844329834,
      "epoch": 0.0022463076318301792,
      "grad_norm": 5.3125,
      "learning_rate": 9.5e-06,
      "loss": 10.694,
      "mean_token_accuracy": 0.0004815136082470417,
      "num_tokens": 44487.0,
      "step": 20
    },
    {
      "entropy": 10.742487525939941,
      "epoch": 0.002807884539787724,
      "grad_norm": 4.875,
      "learning_rate": 1.2e-05,
      "loss": 10.5854,
      "mean_token_accuracy": 0.006200016639195382,
      "num_tokens": 56062.0,
      "step": 25
    },
    {
      "entropy": 10.742135620117187,
      "epoch": 0.0033694614477452686,
      "grad_norm": 3.984375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4464,
      "mean_token_accuracy": 0.028756615333259107,
      "num_tokens": 67453.0,
      "step": 30
    },
    {
      "entropy": 10.741005325317383,
      "epoch": 0.003931038355702813,
      "grad_norm": 3.109375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3034,
      "mean_token_accuracy": 0.03541945908218622,
      "num_tokens": 78527.0,
      "step": 35
    },
    {
      "entropy": 10.738347434997559,
      "epoch": 0.0044926152636603585,
      "grad_norm": 2.578125,
      "learning_rate": 1.95e-05,
      "loss": 10.1364,
      "mean_token_accuracy": 0.04151417501270771,
      "num_tokens": 89267.0,
      "step": 40
    },
    {
      "entropy": 10.734824466705323,
      "epoch": 0.005054192171617903,
      "grad_norm": 2.359375,
      "learning_rate": 2.2e-05,
      "loss": 10.0434,
      "mean_token_accuracy": 0.04415128082036972,
      "num_tokens": 102035.0,
      "step": 45
    },
    {
      "entropy": 10.732510948181153,
      "epoch": 0.005615769079575448,
      "grad_norm": 2.359375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9562,
      "mean_token_accuracy": 0.03771504815667868,
      "num_tokens": 112734.0,
      "step": 50
    },
    {
      "entropy": 10.731938362121582,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.046875,
      "learning_rate": 2.7e-05,
      "loss": 9.8743,
      "mean_token_accuracy": 0.045871196687221526,
      "num_tokens": 123854.0,
      "step": 55
    },
    {
      "entropy": 10.73061008453369,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.0625,
      "learning_rate": 2.95e-05,
      "loss": 9.8175,
      "mean_token_accuracy": 0.04188961498439312,
      "num_tokens": 133444.0,
      "step": 60
    },
    {
      "entropy": 10.72856788635254,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.078125,
      "learning_rate": 3.2e-05,
      "loss": 9.7416,
      "mean_token_accuracy": 0.04224029406905174,
      "num_tokens": 144678.0,
      "step": 65
    },
    {
      "entropy": 10.725998210906983,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.15625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.7079,
      "mean_token_accuracy": 0.04524468667805195,
      "num_tokens": 155866.0,
      "step": 70
    },
    {
      "entropy": 10.722082805633544,
      "epoch": 0.008423653619363172,
      "grad_norm": 2.015625,
      "learning_rate": 3.7e-05,
      "loss": 9.6415,
      "mean_token_accuracy": 0.04917885698378086,
      "num_tokens": 166965.0,
      "step": 75
    },
    {
      "entropy": 10.717686748504638,
      "epoch": 0.008985230527320717,
      "grad_norm": 1.9609375,
      "learning_rate": 3.95e-05,
      "loss": 9.5656,
      "mean_token_accuracy": 0.04601088799536228,
      "num_tokens": 177423.0,
      "step": 80
    },
    {
      "entropy": 10.713061332702637,
      "epoch": 0.00954680743527826,
      "grad_norm": 1.9453125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.4879,
      "mean_token_accuracy": 0.05287941135466099,
      "num_tokens": 188687.0,
      "step": 85
    },
    {
      "entropy": 10.705609130859376,
      "epoch": 0.010108384343235806,
      "grad_norm": 1.9140625,
      "learning_rate": 4.45e-05,
      "loss": 9.4583,
      "mean_token_accuracy": 0.04864194244146347,
      "num_tokens": 200173.0,
      "step": 90
    },
    {
      "entropy": 10.695794582366943,
      "epoch": 0.010669961251193351,
      "grad_norm": 1.984375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.3535,
      "mean_token_accuracy": 0.051952024549245836,
      "num_tokens": 210877.0,
      "step": 95
    },
    {
      "entropy": 10.683987998962403,
      "epoch": 0.011231538159150897,
      "grad_norm": 1.953125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.2644,
      "mean_token_accuracy": 0.04764242954552174,
      "num_tokens": 221770.0,
      "step": 100
    },
    {
      "entropy": 10.668627834320068,
      "epoch": 0.01179311506710844,
      "grad_norm": 1.9296875,
      "learning_rate": 5.2e-05,
      "loss": 9.1557,
      "mean_token_accuracy": 0.0619601234793663,
      "num_tokens": 232140.0,
      "step": 105
    },
    {
      "entropy": 10.643933582305909,
      "epoch": 0.012354691975065986,
      "grad_norm": 1.8125,
      "learning_rate": 5.45e-05,
      "loss": 9.1009,
      "mean_token_accuracy": 0.0608074888586998,
      "num_tokens": 244186.0,
      "step": 110
    },
    {
      "entropy": 10.613188362121582,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.8515625,
      "learning_rate": 5.7e-05,
      "loss": 8.9502,
      "mean_token_accuracy": 0.06373463608324528,
      "num_tokens": 254805.0,
      "step": 115
    },
    {
      "entropy": 10.57162914276123,
      "epoch": 0.013477845790981074,
      "grad_norm": 1.90625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.7802,
      "mean_token_accuracy": 0.06789718717336654,
      "num_tokens": 265055.0,
      "step": 120
    },
    {
      "entropy": 10.509242630004882,
      "epoch": 0.01403942269893862,
      "grad_norm": 1.7890625,
      "learning_rate": 6.2e-05,
      "loss": 8.6595,
      "mean_token_accuracy": 0.06531061716377735,
      "num_tokens": 275850.0,
      "step": 125
    },
    {
      "entropy": 10.434095478057861,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.5608,
      "mean_token_accuracy": 0.06572793014347553,
      "num_tokens": 286802.0,
      "step": 130
    },
    {
      "entropy": 10.361983966827392,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.6484375,
      "learning_rate": 6.7e-05,
      "loss": 8.4719,
      "mean_token_accuracy": 0.06788074560463428,
      "num_tokens": 297070.0,
      "step": 135
    },
    {
      "entropy": 10.27025728225708,
      "epoch": 0.015724153422811252,
      "grad_norm": 1.5859375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.3615,
      "mean_token_accuracy": 0.0678123127669096,
      "num_tokens": 309617.0,
      "step": 140
    },
    {
      "entropy": 10.180762577056885,
      "epoch": 0.0162857303307688,
      "grad_norm": 1.59375,
      "learning_rate": 7.2e-05,
      "loss": 8.2203,
      "mean_token_accuracy": 0.06984737552702427,
      "num_tokens": 321310.0,
      "step": 145
    },
    {
      "entropy": 10.010982036590576,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.3984375,
      "learning_rate": 7.45e-05,
      "loss": 8.0541,
      "mean_token_accuracy": 0.07262194454669953,
      "num_tokens": 332243.0,
      "step": 150
    },
    {
      "entropy": 9.835519409179687,
      "epoch": 0.017408884146683887,
      "grad_norm": 1.3359375,
      "learning_rate": 7.7e-05,
      "loss": 7.9575,
      "mean_token_accuracy": 0.06847953647375107,
      "num_tokens": 342847.0,
      "step": 155
    },
    {
      "entropy": 9.67606430053711,
      "epoch": 0.017970461054641434,
      "grad_norm": 1.3046875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.8107,
      "mean_token_accuracy": 0.07214009724557399,
      "num_tokens": 353570.0,
      "step": 160
    },
    {
      "entropy": 9.435340785980225,
      "epoch": 0.018532037962598977,
      "grad_norm": 1.2421875,
      "learning_rate": 8.2e-05,
      "loss": 7.6786,
      "mean_token_accuracy": 0.08545113056898117,
      "num_tokens": 364330.0,
      "step": 165
    },
    {
      "entropy": 9.163984489440917,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.203125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.637,
      "mean_token_accuracy": 0.08047130256891251,
      "num_tokens": 374930.0,
      "step": 170
    },
    {
      "entropy": 9.023346519470214,
      "epoch": 0.019655191778514068,
      "grad_norm": 1.140625,
      "learning_rate": 8.7e-05,
      "loss": 7.558,
      "mean_token_accuracy": 0.08092962652444839,
      "num_tokens": 384690.0,
      "step": 175
    },
    {
      "entropy": 8.729331684112548,
      "epoch": 0.020216768686471612,
      "grad_norm": 1.171875,
      "learning_rate": 8.95e-05,
      "loss": 7.5022,
      "mean_token_accuracy": 0.08200787231326104,
      "num_tokens": 395753.0,
      "step": 180
    },
    {
      "entropy": 8.560184669494628,
      "epoch": 0.020778345594429155,
      "grad_norm": 0.98046875,
      "learning_rate": 9.2e-05,
      "loss": 7.4189,
      "mean_token_accuracy": 0.0850575651973486,
      "num_tokens": 407049.0,
      "step": 185
    },
    {
      "entropy": 8.38232650756836,
      "epoch": 0.021339922502386702,
      "grad_norm": 0.984375,
      "learning_rate": 9.45e-05,
      "loss": 7.3876,
      "mean_token_accuracy": 0.08972770720720291,
      "num_tokens": 417942.0,
      "step": 190
    },
    {
      "entropy": 8.248396492004394,
      "epoch": 0.021901499410344246,
      "grad_norm": 0.93359375,
      "learning_rate": 9.7e-05,
      "loss": 7.3787,
      "mean_token_accuracy": 0.09045429974794388,
      "num_tokens": 428151.0,
      "step": 195
    },
    {
      "entropy": 8.15650987625122,
      "epoch": 0.022463076318301793,
      "grad_norm": 0.875,
      "learning_rate": 9.95e-05,
      "loss": 7.4337,
      "mean_token_accuracy": 0.07940280884504318,
      "num_tokens": 439837.0,
      "step": 200
    },
    {
      "entropy": 8.114033794403076,
      "epoch": 0.023024653226259337,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000102,
      "loss": 7.2614,
      "mean_token_accuracy": 0.08433199226856232,
      "num_tokens": 450357.0,
      "step": 205
    },
    {
      "entropy": 8.024269819259644,
      "epoch": 0.02358623013421688,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2624,
      "mean_token_accuracy": 0.08984772339463235,
      "num_tokens": 461174.0,
      "step": 210
    },
    {
      "entropy": 7.967411947250366,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.078125,
      "learning_rate": 0.000107,
      "loss": 7.307,
      "mean_token_accuracy": 0.08454840555787087,
      "num_tokens": 472785.0,
      "step": 215
    },
    {
      "entropy": 7.925465250015259,
      "epoch": 0.02470938395013197,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001095,
      "loss": 7.2511,
      "mean_token_accuracy": 0.09478851184248924,
      "num_tokens": 483248.0,
      "step": 220
    },
    {
      "entropy": 7.875831556320191,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000112,
      "loss": 7.142,
      "mean_token_accuracy": 0.09188979342579842,
      "num_tokens": 494690.0,
      "step": 225
    },
    {
      "entropy": 7.84493317604065,
      "epoch": 0.025832537766047062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001145,
      "loss": 7.3296,
      "mean_token_accuracy": 0.08929432481527329,
      "num_tokens": 506192.0,
      "step": 230
    },
    {
      "entropy": 7.927040433883667,
      "epoch": 0.026394114674004605,
      "grad_norm": 1.140625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1847,
      "mean_token_accuracy": 0.09605624824762345,
      "num_tokens": 517152.0,
      "step": 235
    },
    {
      "entropy": 7.799702072143555,
      "epoch": 0.02695569158196215,
      "grad_norm": 0.984375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1141,
      "mean_token_accuracy": 0.09965193793177604,
      "num_tokens": 527836.0,
      "step": 240
    },
    {
      "entropy": 7.821067762374878,
      "epoch": 0.027517268489919696,
      "grad_norm": 0.96875,
      "learning_rate": 0.000122,
      "loss": 7.2301,
      "mean_token_accuracy": 0.09543063268065452,
      "num_tokens": 539491.0,
      "step": 245
    },
    {
      "entropy": 7.813785552978516,
      "epoch": 0.02807884539787724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0001245,
      "loss": 7.2104,
      "mean_token_accuracy": 0.09490795731544495,
      "num_tokens": 550816.0,
      "step": 250
    },
    {
      "entropy": 7.805890893936157,
      "epoch": 0.028640422305834783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000127,
      "loss": 7.2396,
      "mean_token_accuracy": 0.09041734337806702,
      "num_tokens": 561566.0,
      "step": 255
    },
    {
      "entropy": 7.751222324371338,
      "epoch": 0.02920199921379233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0001295,
      "loss": 7.1521,
      "mean_token_accuracy": 0.09656212627887725,
      "num_tokens": 573032.0,
      "step": 260
    },
    {
      "entropy": 7.7648115158081055,
      "epoch": 0.029763576121749874,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000132,
      "loss": 7.0837,
      "mean_token_accuracy": 0.10204266905784606,
      "num_tokens": 584140.0,
      "step": 265
    },
    {
      "entropy": 7.7348565578460695,
      "epoch": 0.030325153029707418,
      "grad_norm": 1.046875,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1494,
      "mean_token_accuracy": 0.10442616268992425,
      "num_tokens": 595097.0,
      "step": 270
    },
    {
      "entropy": 7.7251269817352295,
      "epoch": 0.030886729937664965,
      "grad_norm": 0.96875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1622,
      "mean_token_accuracy": 0.097444436699152,
      "num_tokens": 606797.0,
      "step": 275
    },
    {
      "entropy": 7.689198732376099,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001395,
      "loss": 7.1191,
      "mean_token_accuracy": 0.09910647869110108,
      "num_tokens": 618187.0,
      "step": 280
    },
    {
      "entropy": 7.718263101577759,
      "epoch": 0.032009883753580055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0568,
      "mean_token_accuracy": 0.10645604655146598,
      "num_tokens": 629829.0,
      "step": 285
    },
    {
      "entropy": 7.608854055404663,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001445,
      "loss": 7.0604,
      "mean_token_accuracy": 0.10817909017205238,
      "num_tokens": 640914.0,
      "step": 290
    },
    {
      "entropy": 7.597271966934204,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000147,
      "loss": 7.0713,
      "mean_token_accuracy": 0.09871483817696572,
      "num_tokens": 653057.0,
      "step": 295
    },
    {
      "entropy": 7.650600433349609,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001495,
      "loss": 6.9991,
      "mean_token_accuracy": 0.10946055650711059,
      "num_tokens": 663437.0,
      "step": 300
    },
    {
      "entropy": 7.538260412216187,
      "epoch": 0.03425619138541023,
      "grad_norm": 0.875,
      "learning_rate": 0.000152,
      "loss": 7.0616,
      "mean_token_accuracy": 0.10260984227061272,
      "num_tokens": 675416.0,
      "step": 305
    },
    {
      "entropy": 7.554474115371704,
      "epoch": 0.03481776829336777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.8611,
      "mean_token_accuracy": 0.1113228090107441,
      "num_tokens": 686633.0,
      "step": 310
    },
    {
      "entropy": 7.547433662414551,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000157,
      "loss": 6.9481,
      "mean_token_accuracy": 0.10943943783640861,
      "num_tokens": 698065.0,
      "step": 315
    },
    {
      "entropy": 7.557992935180664,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001595,
      "loss": 6.9576,
      "mean_token_accuracy": 0.10788579881191254,
      "num_tokens": 708652.0,
      "step": 320
    },
    {
      "entropy": 7.549759817123413,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.03125,
      "learning_rate": 0.000162,
      "loss": 7.1014,
      "mean_token_accuracy": 0.104249157756567,
      "num_tokens": 720199.0,
      "step": 325
    },
    {
      "entropy": 7.516535854339599,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0884,
      "mean_token_accuracy": 0.10238429754972458,
      "num_tokens": 732964.0,
      "step": 330
    },
    {
      "entropy": 7.490964794158936,
      "epoch": 0.0376256528331555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8492,
      "mean_token_accuracy": 0.11204343438148498,
      "num_tokens": 743528.0,
      "step": 335
    },
    {
      "entropy": 7.517170190811157,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9139,
      "mean_token_accuracy": 0.11466537714004517,
      "num_tokens": 754365.0,
      "step": 340
    },
    {
      "entropy": 7.3772297382354735,
      "epoch": 0.03874880664907059,
      "grad_norm": 0.953125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8306,
      "mean_token_accuracy": 0.10668750330805779,
      "num_tokens": 766094.0,
      "step": 345
    },
    {
      "entropy": 7.388029289245606,
      "epoch": 0.039310383557028136,
      "grad_norm": 0.953125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.89,
      "mean_token_accuracy": 0.11124317348003387,
      "num_tokens": 777275.0,
      "step": 350
    },
    {
      "entropy": 7.42672815322876,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000177,
      "loss": 6.9507,
      "mean_token_accuracy": 0.11006278768181801,
      "num_tokens": 789436.0,
      "step": 355
    },
    {
      "entropy": 7.445900392532349,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001795,
      "loss": 6.9722,
      "mean_token_accuracy": 0.11323394179344178,
      "num_tokens": 800164.0,
      "step": 360
    },
    {
      "entropy": 7.275260972976684,
      "epoch": 0.04099511428090077,
      "grad_norm": 1.0,
      "learning_rate": 0.000182,
      "loss": 6.8431,
      "mean_token_accuracy": 0.11253381669521331,
      "num_tokens": 811797.0,
      "step": 365
    },
    {
      "entropy": 7.396295213699341,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.125,
      "learning_rate": 0.0001845,
      "loss": 6.8549,
      "mean_token_accuracy": 0.11424426287412644,
      "num_tokens": 824158.0,
      "step": 370
    },
    {
      "entropy": 7.365560102462768,
      "epoch": 0.04211826809681586,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000187,
      "loss": 6.8502,
      "mean_token_accuracy": 0.11723987609148026,
      "num_tokens": 835803.0,
      "step": 375
    },
    {
      "entropy": 7.352114200592041,
      "epoch": 0.042679845004773405,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0001895,
      "loss": 6.913,
      "mean_token_accuracy": 0.1121498741209507,
      "num_tokens": 848106.0,
      "step": 380
    },
    {
      "entropy": 7.327092075347901,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000192,
      "loss": 6.886,
      "mean_token_accuracy": 0.10864873975515366,
      "num_tokens": 859101.0,
      "step": 385
    },
    {
      "entropy": 7.300347328186035,
      "epoch": 0.04380299882068849,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001945,
      "loss": 6.7609,
      "mean_token_accuracy": 0.12088448479771614,
      "num_tokens": 870309.0,
      "step": 390
    },
    {
      "entropy": 7.309509944915772,
      "epoch": 0.044364575728646036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8472,
      "mean_token_accuracy": 0.11545197889208794,
      "num_tokens": 881515.0,
      "step": 395
    },
    {
      "entropy": 7.336116790771484,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.1875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8,
      "mean_token_accuracy": 0.11816975474357605,
      "num_tokens": 892129.0,
      "step": 400
    },
    {
      "entropy": 7.230046129226684,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000202,
      "loss": 6.8228,
      "mean_token_accuracy": 0.11072085574269294,
      "num_tokens": 903186.0,
      "step": 405
    },
    {
      "entropy": 7.260617780685425,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7811,
      "mean_token_accuracy": 0.12182695344090462,
      "num_tokens": 913471.0,
      "step": 410
    },
    {
      "entropy": 7.292573118209839,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.0625,
      "learning_rate": 0.000207,
      "loss": 6.7853,
      "mean_token_accuracy": 0.11431619748473168,
      "num_tokens": 924283.0,
      "step": 415
    },
    {
      "entropy": 7.2130533218383786,
      "epoch": 0.04717246026843376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002095,
      "loss": 6.8177,
      "mean_token_accuracy": 0.12043273225426673,
      "num_tokens": 934975.0,
      "step": 420
    },
    {
      "entropy": 7.203190803527832,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.171875,
      "learning_rate": 0.000212,
      "loss": 6.6934,
      "mean_token_accuracy": 0.12043462619185448,
      "num_tokens": 945587.0,
      "step": 425
    },
    {
      "entropy": 7.19977822303772,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002145,
      "loss": 6.7526,
      "mean_token_accuracy": 0.11818815246224404,
      "num_tokens": 955935.0,
      "step": 430
    },
    {
      "entropy": 7.208397436141968,
      "epoch": 0.0488571909923064,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7148,
      "mean_token_accuracy": 0.11614556312561035,
      "num_tokens": 967960.0,
      "step": 435
    },
    {
      "entropy": 7.185926675796509,
      "epoch": 0.04941876790026394,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002195,
      "loss": 6.7518,
      "mean_token_accuracy": 0.12180673331022263,
      "num_tokens": 979680.0,
      "step": 440
    },
    {
      "entropy": 7.172959804534912,
      "epoch": 0.049980344808221486,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000222,
      "loss": 6.8161,
      "mean_token_accuracy": 0.11992981359362602,
      "num_tokens": 990600.0,
      "step": 445
    },
    {
      "entropy": 7.308594560623169,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002245,
      "loss": 6.8481,
      "mean_token_accuracy": 0.12150532379746437,
      "num_tokens": 1002331.0,
      "step": 450
    },
    {
      "entropy": 7.118796253204346,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6594,
      "mean_token_accuracy": 0.12037151604890824,
      "num_tokens": 1013473.0,
      "step": 455
    },
    {
      "entropy": 7.1910758972167965,
      "epoch": 0.051665075532094124,
      "grad_norm": 0.921875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.749,
      "mean_token_accuracy": 0.11884191855788231,
      "num_tokens": 1024434.0,
      "step": 460
    },
    {
      "entropy": 7.1801292419433596,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7697,
      "mean_token_accuracy": 0.1166395016014576,
      "num_tokens": 1035452.0,
      "step": 465
    },
    {
      "entropy": 7.1035239696502686,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6438,
      "mean_token_accuracy": 0.11835466921329499,
      "num_tokens": 1046213.0,
      "step": 470
    },
    {
      "entropy": 7.088221454620362,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000237,
      "loss": 6.666,
      "mean_token_accuracy": 0.12225791588425636,
      "num_tokens": 1057469.0,
      "step": 475
    },
    {
      "entropy": 7.244529485702515,
      "epoch": 0.0539113831639243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002395,
      "loss": 6.8169,
      "mean_token_accuracy": 0.11744396165013313,
      "num_tokens": 1069543.0,
      "step": 480
    },
    {
      "entropy": 6.995753431320191,
      "epoch": 0.05447296007188184,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000242,
      "loss": 6.6818,
      "mean_token_accuracy": 0.12118739560246468,
      "num_tokens": 1080476.0,
      "step": 485
    },
    {
      "entropy": 7.242927026748657,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002445,
      "loss": 6.7896,
      "mean_token_accuracy": 0.12034596726298333,
      "num_tokens": 1091700.0,
      "step": 490
    },
    {
      "entropy": 7.122195816040039,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000247,
      "loss": 6.677,
      "mean_token_accuracy": 0.12182887271046638,
      "num_tokens": 1103056.0,
      "step": 495
    },
    {
      "entropy": 7.137054586410523,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002495,
      "loss": 6.7363,
      "mean_token_accuracy": 0.12379441931843757,
      "num_tokens": 1114836.0,
      "step": 500
    },
    {
      "entropy": 6.997432088851928,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.09375,
      "learning_rate": 0.000252,
      "loss": 6.6501,
      "mean_token_accuracy": 0.12145527005195618,
      "num_tokens": 1125993.0,
      "step": 505
    },
    {
      "entropy": 7.124313068389893,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002545,
      "loss": 6.7208,
      "mean_token_accuracy": 0.12192488759756089,
      "num_tokens": 1136719.0,
      "step": 510
    },
    {
      "entropy": 7.03917121887207,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.328125,
      "learning_rate": 0.000257,
      "loss": 6.5859,
      "mean_token_accuracy": 0.1255400411784649,
      "num_tokens": 1146947.0,
      "step": 515
    },
    {
      "entropy": 7.102051591873169,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002595,
      "loss": 6.7183,
      "mean_token_accuracy": 0.12346506491303444,
      "num_tokens": 1157974.0,
      "step": 520
    },
    {
      "entropy": 6.9870562076568605,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000262,
      "loss": 6.646,
      "mean_token_accuracy": 0.11965501159429551,
      "num_tokens": 1168573.0,
      "step": 525
    },
    {
      "entropy": 7.062936162948608,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5636,
      "mean_token_accuracy": 0.12382580935955048,
      "num_tokens": 1178446.0,
      "step": 530
    },
    {
      "entropy": 7.034807634353638,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.5794,
      "mean_token_accuracy": 0.1255895234644413,
      "num_tokens": 1190558.0,
      "step": 535
    },
    {
      "entropy": 6.959788131713867,
      "epoch": 0.060650306059414835,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.746,
      "mean_token_accuracy": 0.12257594242691994,
      "num_tokens": 1201929.0,
      "step": 540
    },
    {
      "entropy": 6.939215898513794,
      "epoch": 0.06121188296737238,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5748,
      "mean_token_accuracy": 0.12955221608281137,
      "num_tokens": 1213155.0,
      "step": 545
    },
    {
      "entropy": 7.138561248779297,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.25,
      "learning_rate": 0.0002745,
      "loss": 6.656,
      "mean_token_accuracy": 0.12121416926383972,
      "num_tokens": 1223961.0,
      "step": 550
    },
    {
      "entropy": 6.852568626403809,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000277,
      "loss": 6.5532,
      "mean_token_accuracy": 0.12889384627342224,
      "num_tokens": 1233888.0,
      "step": 555
    },
    {
      "entropy": 6.934140396118164,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.0,
      "learning_rate": 0.0002795,
      "loss": 6.6656,
      "mean_token_accuracy": 0.11890221461653709,
      "num_tokens": 1244405.0,
      "step": 560
    },
    {
      "entropy": 7.04783000946045,
      "epoch": 0.06345819059920256,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6081,
      "mean_token_accuracy": 0.12794038355350495,
      "num_tokens": 1257099.0,
      "step": 565
    },
    {
      "entropy": 7.002995395660401,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002845,
      "loss": 6.7054,
      "mean_token_accuracy": 0.1265850067138672,
      "num_tokens": 1267965.0,
      "step": 570
    },
    {
      "entropy": 6.995109415054321,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000287,
      "loss": 6.6099,
      "mean_token_accuracy": 0.12399538457393647,
      "num_tokens": 1280075.0,
      "step": 575
    },
    {
      "entropy": 7.007772064208984,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002895,
      "loss": 6.6188,
      "mean_token_accuracy": 0.12613040655851365,
      "num_tokens": 1290574.0,
      "step": 580
    },
    {
      "entropy": 6.981190586090088,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000292,
      "loss": 6.5851,
      "mean_token_accuracy": 0.12990300580859185,
      "num_tokens": 1302680.0,
      "step": 585
    },
    {
      "entropy": 6.931596183776856,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002945,
      "loss": 6.5766,
      "mean_token_accuracy": 0.12965173944830893,
      "num_tokens": 1314577.0,
      "step": 590
    },
    {
      "entropy": 6.966735315322876,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.109375,
      "learning_rate": 0.000297,
      "loss": 6.6977,
      "mean_token_accuracy": 0.1269223101437092,
      "num_tokens": 1326560.0,
      "step": 595
    },
    {
      "entropy": 7.021285438537598,
      "epoch": 0.06738922895490537,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002995,
      "loss": 6.7678,
      "mean_token_accuracy": 0.11841874867677689,
      "num_tokens": 1339105.0,
      "step": 600
    },
    {
      "entropy": 6.793500185012817,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000302,
      "loss": 6.4295,
      "mean_token_accuracy": 0.1327479861676693,
      "num_tokens": 1349914.0,
      "step": 605
    },
    {
      "entropy": 6.940507316589356,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003045,
      "loss": 6.6528,
      "mean_token_accuracy": 0.12645572870969773,
      "num_tokens": 1361700.0,
      "step": 610
    },
    {
      "entropy": 6.953217029571533,
      "epoch": 0.06907395967877801,
      "grad_norm": 0.921875,
      "learning_rate": 0.000307,
      "loss": 6.6735,
      "mean_token_accuracy": 0.12332428023219108,
      "num_tokens": 1373103.0,
      "step": 615
    },
    {
      "entropy": 6.914129304885864,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003095,
      "loss": 6.459,
      "mean_token_accuracy": 0.13300212621688842,
      "num_tokens": 1383742.0,
      "step": 620
    },
    {
      "entropy": 6.799388074874878,
      "epoch": 0.0701971134946931,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000312,
      "loss": 6.6126,
      "mean_token_accuracy": 0.12680527344346046,
      "num_tokens": 1396656.0,
      "step": 625
    },
    {
      "entropy": 6.921048307418824,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003145,
      "loss": 6.4886,
      "mean_token_accuracy": 0.14117394238710404,
      "num_tokens": 1406709.0,
      "step": 630
    },
    {
      "entropy": 6.875065326690674,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.015625,
      "learning_rate": 0.000317,
      "loss": 6.5279,
      "mean_token_accuracy": 0.13476523160934448,
      "num_tokens": 1417793.0,
      "step": 635
    },
    {
      "entropy": 6.934533977508545,
      "epoch": 0.07188184421856574,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003195,
      "loss": 6.6046,
      "mean_token_accuracy": 0.1279690831899643,
      "num_tokens": 1429093.0,
      "step": 640
    },
    {
      "entropy": 6.780666828155518,
      "epoch": 0.07244342112652327,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000322,
      "loss": 6.5326,
      "mean_token_accuracy": 0.13261018618941306,
      "num_tokens": 1440907.0,
      "step": 645
    },
    {
      "entropy": 6.917075872421265,
      "epoch": 0.07300499803448082,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6025,
      "mean_token_accuracy": 0.12796421274542807,
      "num_tokens": 1452089.0,
      "step": 650
    },
    {
      "entropy": 6.888136577606201,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6206,
      "mean_token_accuracy": 0.12372568026185035,
      "num_tokens": 1464378.0,
      "step": 655
    },
    {
      "entropy": 6.848985958099365,
      "epoch": 0.07412815185039591,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5431,
      "mean_token_accuracy": 0.1299239970743656,
      "num_tokens": 1476678.0,
      "step": 660
    },
    {
      "entropy": 6.798602628707886,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.3923,
      "mean_token_accuracy": 0.13382024466991424,
      "num_tokens": 1486808.0,
      "step": 665
    },
    {
      "entropy": 6.9307825565338135,
      "epoch": 0.075251305666311,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6765,
      "mean_token_accuracy": 0.12494136691093445,
      "num_tokens": 1499281.0,
      "step": 670
    },
    {
      "entropy": 6.8795429229736325,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000337,
      "loss": 6.5764,
      "mean_token_accuracy": 0.12798068225383757,
      "num_tokens": 1510202.0,
      "step": 675
    },
    {
      "entropy": 6.831705522537232,
      "epoch": 0.07637445948222608,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003395,
      "loss": 6.4526,
      "mean_token_accuracy": 0.1334339514374733,
      "num_tokens": 1521492.0,
      "step": 680
    },
    {
      "entropy": 6.733158874511719,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000342,
      "loss": 6.4612,
      "mean_token_accuracy": 0.13912885189056395,
      "num_tokens": 1532856.0,
      "step": 685
    },
    {
      "entropy": 6.818010997772217,
      "epoch": 0.07749761329814119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4673,
      "mean_token_accuracy": 0.1312962956726551,
      "num_tokens": 1543408.0,
      "step": 690
    },
    {
      "entropy": 6.730507946014404,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.046875,
      "learning_rate": 0.000347,
      "loss": 6.5037,
      "mean_token_accuracy": 0.12911636829376222,
      "num_tokens": 1555474.0,
      "step": 695
    },
    {
      "entropy": 6.848674011230469,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003495,
      "loss": 6.5921,
      "mean_token_accuracy": 0.13230398222804068,
      "num_tokens": 1567484.0,
      "step": 700
    },
    {
      "entropy": 6.715642690658569,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.0625,
      "learning_rate": 0.000352,
      "loss": 6.4247,
      "mean_token_accuracy": 0.14034333527088166,
      "num_tokens": 1578204.0,
      "step": 705
    },
    {
      "entropy": 6.863985157012939,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003545,
      "loss": 6.4245,
      "mean_token_accuracy": 0.1318908639252186,
      "num_tokens": 1589012.0,
      "step": 710
    },
    {
      "entropy": 6.696931076049805,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000357,
      "loss": 6.4363,
      "mean_token_accuracy": 0.13300200998783113,
      "num_tokens": 1601099.0,
      "step": 715
    },
    {
      "entropy": 6.80228910446167,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595,
      "loss": 6.5967,
      "mean_token_accuracy": 0.12784573882818223,
      "num_tokens": 1612814.0,
      "step": 720
    },
    {
      "entropy": 6.878298664093018,
      "epoch": 0.081428651653844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362,
      "loss": 6.4838,
      "mean_token_accuracy": 0.13338544070720673,
      "num_tokens": 1624953.0,
      "step": 725
    },
    {
      "entropy": 6.6930359363555905,
      "epoch": 0.08199022856180153,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003645,
      "loss": 6.5261,
      "mean_token_accuracy": 0.13288487046957015,
      "num_tokens": 1635673.0,
      "step": 730
    },
    {
      "entropy": 6.79468789100647,
      "epoch": 0.08255180546975908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000367,
      "loss": 6.4303,
      "mean_token_accuracy": 0.13556234017014504,
      "num_tokens": 1647282.0,
      "step": 735
    },
    {
      "entropy": 6.7501686096191404,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003695,
      "loss": 6.4432,
      "mean_token_accuracy": 0.13580124080181122,
      "num_tokens": 1658426.0,
      "step": 740
    },
    {
      "entropy": 6.72845983505249,
      "epoch": 0.08367495928567417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000372,
      "loss": 6.4705,
      "mean_token_accuracy": 0.13703056201338767,
      "num_tokens": 1669958.0,
      "step": 745
    },
    {
      "entropy": 6.7487999439239506,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003745,
      "loss": 6.5243,
      "mean_token_accuracy": 0.13213849067687988,
      "num_tokens": 1682409.0,
      "step": 750
    },
    {
      "entropy": 6.751234817504883,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.0,
      "learning_rate": 0.000377,
      "loss": 6.4479,
      "mean_token_accuracy": 0.1415566846728325,
      "num_tokens": 1692409.0,
      "step": 755
    },
    {
      "entropy": 6.7592613220214846,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003795,
      "loss": 6.5309,
      "mean_token_accuracy": 0.13545768857002258,
      "num_tokens": 1705018.0,
      "step": 760
    },
    {
      "entropy": 6.738600540161133,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000382,
      "loss": 6.4573,
      "mean_token_accuracy": 0.13337523862719536,
      "num_tokens": 1715625.0,
      "step": 765
    },
    {
      "entropy": 6.6791502952575685,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003845,
      "loss": 6.537,
      "mean_token_accuracy": 0.12219079732894897,
      "num_tokens": 1727045.0,
      "step": 770
    },
    {
      "entropy": 6.9133740901947025,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5994,
      "mean_token_accuracy": 0.12983450889587403,
      "num_tokens": 1737374.0,
      "step": 775
    },
    {
      "entropy": 6.78628134727478,
      "epoch": 0.08760599764137698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.5867,
      "mean_token_accuracy": 0.13110264837741853,
      "num_tokens": 1747788.0,
      "step": 780
    },
    {
      "entropy": 6.590769243240357,
      "epoch": 0.08816757454933453,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2882,
      "mean_token_accuracy": 0.1412363886833191,
      "num_tokens": 1757769.0,
      "step": 785
    },
    {
      "entropy": 6.66961555480957,
      "epoch": 0.08872915145729207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4112,
      "mean_token_accuracy": 0.13004442900419236,
      "num_tokens": 1769246.0,
      "step": 790
    },
    {
      "entropy": 6.755950880050659,
      "epoch": 0.08929072836524962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.5278,
      "mean_token_accuracy": 0.13320355713367463,
      "num_tokens": 1781306.0,
      "step": 795
    },
    {
      "entropy": 6.751245784759521,
      "epoch": 0.08985230527320717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003995,
      "loss": 6.429,
      "mean_token_accuracy": 0.1355741389095783,
      "num_tokens": 1792888.0,
      "step": 800
    },
    {
      "entropy": 6.691725397109986,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.109375,
      "learning_rate": 0.000402,
      "loss": 6.437,
      "mean_token_accuracy": 0.13696239069104194,
      "num_tokens": 1802938.0,
      "step": 805
    },
    {
      "entropy": 6.722809076309204,
      "epoch": 0.09097545908912226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004045,
      "loss": 6.4553,
      "mean_token_accuracy": 0.1335028439760208,
      "num_tokens": 1814738.0,
      "step": 810
    },
    {
      "entropy": 6.6880824089050295,
      "epoch": 0.0915370359970798,
      "grad_norm": 1.0,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5314,
      "mean_token_accuracy": 0.1331937789916992,
      "num_tokens": 1825100.0,
      "step": 815
    },
    {
      "entropy": 6.560724306106567,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004095,
      "loss": 6.3725,
      "mean_token_accuracy": 0.13867289274930955,
      "num_tokens": 1835879.0,
      "step": 820
    },
    {
      "entropy": 6.843134069442749,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000412,
      "loss": 6.5243,
      "mean_token_accuracy": 0.125446654856205,
      "num_tokens": 1848352.0,
      "step": 825
    },
    {
      "entropy": 6.709258794784546,
      "epoch": 0.09322176672095243,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004145,
      "loss": 6.4662,
      "mean_token_accuracy": 0.13462386131286622,
      "num_tokens": 1859687.0,
      "step": 830
    },
    {
      "entropy": 6.656016445159912,
      "epoch": 0.09378334362890998,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000417,
      "loss": 6.4593,
      "mean_token_accuracy": 0.13432704880833626,
      "num_tokens": 1871438.0,
      "step": 835
    },
    {
      "entropy": 6.657505941390991,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004195,
      "loss": 6.3499,
      "mean_token_accuracy": 0.1395242303609848,
      "num_tokens": 1881519.0,
      "step": 840
    },
    {
      "entropy": 6.7550750255584715,
      "epoch": 0.09490649744482507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000422,
      "loss": 6.519,
      "mean_token_accuracy": 0.1305337093770504,
      "num_tokens": 1892428.0,
      "step": 845
    },
    {
      "entropy": 6.742611217498779,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004245,
      "loss": 6.4219,
      "mean_token_accuracy": 0.13647260293364524,
      "num_tokens": 1903760.0,
      "step": 850
    },
    {
      "entropy": 6.685311698913575,
      "epoch": 0.09602965126074016,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000427,
      "loss": 6.47,
      "mean_token_accuracy": 0.130086899548769,
      "num_tokens": 1915430.0,
      "step": 855
    },
    {
      "entropy": 6.657859754562378,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004295,
      "loss": 6.4318,
      "mean_token_accuracy": 0.13017144575715064,
      "num_tokens": 1925709.0,
      "step": 860
    },
    {
      "entropy": 6.760464239120483,
      "epoch": 0.09715280507665525,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000432,
      "loss": 6.4697,
      "mean_token_accuracy": 0.1342718742787838,
      "num_tokens": 1938148.0,
      "step": 865
    },
    {
      "entropy": 6.464514970779419,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004345,
      "loss": 6.3438,
      "mean_token_accuracy": 0.13634837418794632,
      "num_tokens": 1949028.0,
      "step": 870
    },
    {
      "entropy": 6.70568528175354,
      "epoch": 0.09827595889257033,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000437,
      "loss": 6.4016,
      "mean_token_accuracy": 0.13514200374484062,
      "num_tokens": 1960353.0,
      "step": 875
    },
    {
      "entropy": 6.69066276550293,
      "epoch": 0.09883753580052788,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004395,
      "loss": 6.5016,
      "mean_token_accuracy": 0.13418799117207528,
      "num_tokens": 1972335.0,
      "step": 880
    },
    {
      "entropy": 6.591297006607055,
      "epoch": 0.09939911270848542,
      "grad_norm": 1.0625,
      "learning_rate": 0.000442,
      "loss": 6.3568,
      "mean_token_accuracy": 0.136691115796566,
      "num_tokens": 1982851.0,
      "step": 885
    },
    {
      "entropy": 6.583767509460449,
      "epoch": 0.09996068961644297,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004445,
      "loss": 6.3297,
      "mean_token_accuracy": 0.1408061757683754,
      "num_tokens": 1993931.0,
      "step": 890
    },
    {
      "entropy": 6.68323655128479,
      "epoch": 0.10052226652440052,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000447,
      "loss": 6.4678,
      "mean_token_accuracy": 0.135716450214386,
      "num_tokens": 2004838.0,
      "step": 895
    },
    {
      "entropy": 6.533196067810058,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2726,
      "mean_token_accuracy": 0.1382671043276787,
      "num_tokens": 2015620.0,
      "step": 900
    },
    {
      "entropy": 6.655929517745972,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4045,
      "mean_token_accuracy": 0.14243076518177986,
      "num_tokens": 2027489.0,
      "step": 905
    },
    {
      "entropy": 6.497221660614014,
      "epoch": 0.10220699724827315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2791,
      "mean_token_accuracy": 0.1438898392021656,
      "num_tokens": 2038475.0,
      "step": 910
    },
    {
      "entropy": 6.563181304931641,
      "epoch": 0.1027685741562307,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2808,
      "mean_token_accuracy": 0.14109615236520767,
      "num_tokens": 2048938.0,
      "step": 915
    },
    {
      "entropy": 6.534266376495362,
      "epoch": 0.10333015106418825,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3659,
      "mean_token_accuracy": 0.13268538564443588,
      "num_tokens": 2060372.0,
      "step": 920
    },
    {
      "entropy": 6.537831211090088,
      "epoch": 0.10389172797214578,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000462,
      "loss": 6.3626,
      "mean_token_accuracy": 0.13345034644007683,
      "num_tokens": 2071558.0,
      "step": 925
    },
    {
      "entropy": 6.619134902954102,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004645,
      "loss": 6.3985,
      "mean_token_accuracy": 0.13327228128910065,
      "num_tokens": 2082350.0,
      "step": 930
    },
    {
      "entropy": 6.5742902755737305,
      "epoch": 0.10501488178806087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000467,
      "loss": 6.4442,
      "mean_token_accuracy": 0.1369822859764099,
      "num_tokens": 2093448.0,
      "step": 935
    },
    {
      "entropy": 6.717497825622559,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004695,
      "loss": 6.4786,
      "mean_token_accuracy": 0.135139499604702,
      "num_tokens": 2104171.0,
      "step": 940
    },
    {
      "entropy": 6.5645914554595945,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000472,
      "loss": 6.3644,
      "mean_token_accuracy": 0.13385832235217093,
      "num_tokens": 2114496.0,
      "step": 945
    },
    {
      "entropy": 6.612611770629883,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004745,
      "loss": 6.3628,
      "mean_token_accuracy": 0.1407134100794792,
      "num_tokens": 2125930.0,
      "step": 950
    },
    {
      "entropy": 6.603974390029907,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000477,
      "loss": 6.3126,
      "mean_token_accuracy": 0.1430915966629982,
      "num_tokens": 2136452.0,
      "step": 955
    },
    {
      "entropy": 6.581751966476441,
      "epoch": 0.1078227663278486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004795,
      "loss": 6.4209,
      "mean_token_accuracy": 0.13645661920309066,
      "num_tokens": 2148233.0,
      "step": 960
    },
    {
      "entropy": 6.4400779724121096,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.03125,
      "learning_rate": 0.000482,
      "loss": 6.209,
      "mean_token_accuracy": 0.1424497328698635,
      "num_tokens": 2159475.0,
      "step": 965
    },
    {
      "entropy": 6.48789873123169,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004845,
      "loss": 6.3354,
      "mean_token_accuracy": 0.13974294438958168,
      "num_tokens": 2170042.0,
      "step": 970
    },
    {
      "entropy": 6.54668779373169,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.078125,
      "learning_rate": 0.000487,
      "loss": 6.3332,
      "mean_token_accuracy": 0.13519977256655694,
      "num_tokens": 2182234.0,
      "step": 975
    },
    {
      "entropy": 6.461383056640625,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004895,
      "loss": 6.3302,
      "mean_token_accuracy": 0.1389101855456829,
      "num_tokens": 2192477.0,
      "step": 980
    },
    {
      "entropy": 6.4945393085479735,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.15625,
      "learning_rate": 0.000492,
      "loss": 6.3111,
      "mean_token_accuracy": 0.13645306155085563,
      "num_tokens": 2203082.0,
      "step": 985
    },
    {
      "entropy": 6.488329744338989,
      "epoch": 0.11119222777559387,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945,
      "loss": 6.2808,
      "mean_token_accuracy": 0.1423295982182026,
      "num_tokens": 2213684.0,
      "step": 990
    },
    {
      "entropy": 6.481782674789429,
      "epoch": 0.11175380468355141,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497,
      "loss": 6.2903,
      "mean_token_accuracy": 0.13810332641005515,
      "num_tokens": 2225524.0,
      "step": 995
    },
    {
      "entropy": 6.548757648468017,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995,
      "loss": 6.2541,
      "mean_token_accuracy": 0.14044561684131623,
      "num_tokens": 2236004.0,
      "step": 1000
    },
    {
      "entropy": 6.52207841873169,
      "epoch": 0.1128769584994665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499999997707492,
      "loss": 6.4111,
      "mean_token_accuracy": 0.12956544756889343,
      "num_tokens": 2247730.0,
      "step": 1005
    },
    {
      "entropy": 6.574723386764527,
      "epoch": 0.11343853540742405,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.3719,
      "mean_token_accuracy": 0.14130636155605317,
      "num_tokens": 2260132.0,
      "step": 1010
    },
    {
      "entropy": 6.472715663909912,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.332,
      "mean_token_accuracy": 0.14104451537132262,
      "num_tokens": 2270060.0,
      "step": 1015
    },
    {
      "entropy": 6.53025188446045,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.3458,
      "mean_token_accuracy": 0.1401911422610283,
      "num_tokens": 2281568.0,
      "step": 1020
    },
    {
      "entropy": 6.55501389503479,
      "epoch": 0.11512326613129668,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.3655,
      "mean_token_accuracy": 0.12993661910295487,
      "num_tokens": 2293325.0,
      "step": 1025
    },
    {
      "entropy": 6.545099115371704,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.3734,
      "mean_token_accuracy": 0.1424901679158211,
      "num_tokens": 2304780.0,
      "step": 1030
    },
    {
      "entropy": 6.4766498565673825,
      "epoch": 0.11624641994721177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.3002,
      "mean_token_accuracy": 0.13595817238092422,
      "num_tokens": 2315999.0,
      "step": 1035
    },
    {
      "entropy": 6.538008546829223,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999997820685,
      "loss": 6.4523,
      "mean_token_accuracy": 0.1396072454750538,
      "num_tokens": 2326695.0,
      "step": 1040
    },
    {
      "entropy": 6.518211507797242,
      "epoch": 0.11736957376312686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.2754,
      "mean_token_accuracy": 0.1473626583814621,
      "num_tokens": 2336925.0,
      "step": 1045
    },
    {
      "entropy": 6.457152557373047,
      "epoch": 0.11793115067108441,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.1916,
      "mean_token_accuracy": 0.14323539957404136,
      "num_tokens": 2347431.0,
      "step": 1050
    },
    {
      "entropy": 6.450267267227173,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.2976,
      "mean_token_accuracy": 0.14234493300318718,
      "num_tokens": 2359894.0,
      "step": 1055
    },
    {
      "entropy": 6.444874811172485,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.3301,
      "mean_token_accuracy": 0.1389567509293556,
      "num_tokens": 2371181.0,
      "step": 1060
    },
    {
      "entropy": 6.532542181015015,
      "epoch": 0.11961588139495703,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.2249,
      "mean_token_accuracy": 0.14046408236026764,
      "num_tokens": 2381531.0,
      "step": 1065
    },
    {
      "entropy": 6.54881477355957,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.3564,
      "mean_token_accuracy": 0.14078477621078492,
      "num_tokens": 2391573.0,
      "step": 1070
    },
    {
      "entropy": 6.4431471824646,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.3146,
      "mean_token_accuracy": 0.1424903988838196,
      "num_tokens": 2402822.0,
      "step": 1075
    },
    {
      "entropy": 6.424335241317749,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.2338,
      "mean_token_accuracy": 0.14391232281923294,
      "num_tokens": 2414322.0,
      "step": 1080
    },
    {
      "entropy": 6.573728752136231,
      "epoch": 0.12186218902678722,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.4169,
      "mean_token_accuracy": 0.14226323813199998,
      "num_tokens": 2426560.0,
      "step": 1085
    },
    {
      "entropy": 6.537834119796753,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.3298,
      "mean_token_accuracy": 0.14009772390127181,
      "num_tokens": 2437723.0,
      "step": 1090
    },
    {
      "entropy": 6.393655395507812,
      "epoch": 0.12298534284270231,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.2686,
      "mean_token_accuracy": 0.1454232707619667,
      "num_tokens": 2448497.0,
      "step": 1095
    },
    {
      "entropy": 6.39782018661499,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.1461,
      "mean_token_accuracy": 0.15012311190366745,
      "num_tokens": 2459089.0,
      "step": 1100
    },
    {
      "entropy": 6.399022197723388,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.232,
      "mean_token_accuracy": 0.14566392898559571,
      "num_tokens": 2469882.0,
      "step": 1105
    },
    {
      "entropy": 6.414362478256225,
      "epoch": 0.12467007356657495,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.2702,
      "mean_token_accuracy": 0.14421137124300004,
      "num_tokens": 2482120.0,
      "step": 1110
    },
    {
      "entropy": 6.482989406585693,
      "epoch": 0.1252316504745325,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.281,
      "mean_token_accuracy": 0.14298845529556276,
      "num_tokens": 2492626.0,
      "step": 1115
    },
    {
      "entropy": 6.556837034225464,
      "epoch": 0.12579322738249002,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499997970990249,
      "loss": 6.3345,
      "mean_token_accuracy": 0.1420891359448433,
      "num_tokens": 2506347.0,
      "step": 1120
    },
    {
      "entropy": 6.283041858673096,
      "epoch": 0.12635480429044757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.2195,
      "mean_token_accuracy": 0.14754280745983123,
      "num_tokens": 2516584.0,
      "step": 1125
    },
    {
      "entropy": 6.48365969657898,
      "epoch": 0.12691638119840512,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.2571,
      "mean_token_accuracy": 0.14406370520591735,
      "num_tokens": 2526731.0,
      "step": 1130
    },
    {
      "entropy": 6.371777677536011,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.1664,
      "mean_token_accuracy": 0.14858057275414466,
      "num_tokens": 2537180.0,
      "step": 1135
    },
    {
      "entropy": 6.46185154914856,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.2588,
      "mean_token_accuracy": 0.14099827483296395,
      "num_tokens": 2547617.0,
      "step": 1140
    },
    {
      "entropy": 6.342129230499268,
      "epoch": 0.12860111192227774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.1259,
      "mean_token_accuracy": 0.1511496439576149,
      "num_tokens": 2557743.0,
      "step": 1145
    },
    {
      "entropy": 6.364615345001221,
      "epoch": 0.1291626888302353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.2275,
      "mean_token_accuracy": 0.14248820543289184,
      "num_tokens": 2568351.0,
      "step": 1150
    },
    {
      "entropy": 6.346706676483154,
      "epoch": 0.12972426573819285,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.284,
      "mean_token_accuracy": 0.1438751205801964,
      "num_tokens": 2579703.0,
      "step": 1155
    },
    {
      "entropy": 6.507516384124756,
      "epoch": 0.1302858426461504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.2616,
      "mean_token_accuracy": 0.14255546554923057,
      "num_tokens": 2591197.0,
      "step": 1160
    },
    {
      "entropy": 6.382220983505249,
      "epoch": 0.13084741955410795,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499996146305169,
      "loss": 6.2282,
      "mean_token_accuracy": 0.14040454253554344,
      "num_tokens": 2601836.0,
      "step": 1165
    },
    {
      "entropy": 6.333654737472534,
      "epoch": 0.13140899646206547,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.1738,
      "mean_token_accuracy": 0.14583537429571153,
      "num_tokens": 2613513.0,
      "step": 1170
    },
    {
      "entropy": 6.3707410335540775,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.2076,
      "mean_token_accuracy": 0.14694963172078132,
      "num_tokens": 2625246.0,
      "step": 1175
    },
    {
      "entropy": 6.359919452667237,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.1798,
      "mean_token_accuracy": 0.14660840332508088,
      "num_tokens": 2636375.0,
      "step": 1180
    },
    {
      "entropy": 6.385069990158081,
      "epoch": 0.13309372718593812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.2716,
      "mean_token_accuracy": 0.1412426635622978,
      "num_tokens": 2648586.0,
      "step": 1185
    },
    {
      "entropy": 6.442723178863526,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2785,
      "mean_token_accuracy": 0.1502460092306137,
      "num_tokens": 2660847.0,
      "step": 1190
    },
    {
      "entropy": 6.351391172409057,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049999460746976,
      "loss": 6.2095,
      "mean_token_accuracy": 0.14432434290647506,
      "num_tokens": 2671160.0,
      "step": 1195
    },
    {
      "entropy": 6.561935091018677,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.3171,
      "mean_token_accuracy": 0.14210065752267836,
      "num_tokens": 2681622.0,
      "step": 1200
    },
    {
      "entropy": 6.271227216720581,
      "epoch": 0.1353400348177683,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.1086,
      "mean_token_accuracy": 0.1514277994632721,
      "num_tokens": 2691503.0,
      "step": 1205
    },
    {
      "entropy": 6.338787794113159,
      "epoch": 0.13590161172572585,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.2185,
      "mean_token_accuracy": 0.14308153241872787,
      "num_tokens": 2702207.0,
      "step": 1210
    },
    {
      "entropy": 6.401724863052368,
      "epoch": 0.1364631886336834,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.2105,
      "mean_token_accuracy": 0.14515243843197823,
      "num_tokens": 2713178.0,
      "step": 1215
    },
    {
      "entropy": 6.416276550292968,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.2213,
      "mean_token_accuracy": 0.15066151767969133,
      "num_tokens": 2723717.0,
      "step": 1220
    },
    {
      "entropy": 6.349833822250366,
      "epoch": 0.13758634244959847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.2798,
      "mean_token_accuracy": 0.1415386199951172,
      "num_tokens": 2735669.0,
      "step": 1225
    },
    {
      "entropy": 6.367289066314697,
      "epoch": 0.13814791935755602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.1856,
      "mean_token_accuracy": 0.1438096582889557,
      "num_tokens": 2746087.0,
      "step": 1230
    },
    {
      "entropy": 6.332089853286743,
      "epoch": 0.13870949626551357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.2014,
      "mean_token_accuracy": 0.14393291622400284,
      "num_tokens": 2757143.0,
      "step": 1235
    },
    {
      "entropy": 6.428728485107422,
      "epoch": 0.1392710731734711,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.3014,
      "mean_token_accuracy": 0.14128980934619903,
      "num_tokens": 2769340.0,
      "step": 1240
    },
    {
      "entropy": 6.446348237991333,
      "epoch": 0.13983265008142864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.2146,
      "mean_token_accuracy": 0.14424442052841185,
      "num_tokens": 2779514.0,
      "step": 1245
    },
    {
      "entropy": 6.338409376144409,
      "epoch": 0.1403942269893862,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.1603,
      "mean_token_accuracy": 0.14622590690851212,
      "num_tokens": 2790801.0,
      "step": 1250
    },
    {
      "entropy": 6.357332468032837,
      "epoch": 0.14095580389734375,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.1487,
      "mean_token_accuracy": 0.146085686981678,
      "num_tokens": 2801082.0,
      "step": 1255
    },
    {
      "entropy": 6.335656404495239,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1549,
      "mean_token_accuracy": 0.14812700152397157,
      "num_tokens": 2812719.0,
      "step": 1260
    },
    {
      "entropy": 6.2651628971099855,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1605,
      "mean_token_accuracy": 0.14404110759496688,
      "num_tokens": 2823766.0,
      "step": 1265
    },
    {
      "entropy": 6.331585931777954,
      "epoch": 0.14264053462121637,
      "grad_norm": 1.0,
      "learning_rate": 0.000499989632069227,
      "loss": 6.2544,
      "mean_token_accuracy": 0.14194617569446563,
      "num_tokens": 2834801.0,
      "step": 1270
    },
    {
      "entropy": 6.331230735778808,
      "epoch": 0.14320211152917392,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.067,
      "mean_token_accuracy": 0.1510644406080246,
      "num_tokens": 2845769.0,
      "step": 1275
    },
    {
      "entropy": 6.218266439437866,
      "epoch": 0.14376368843713147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.1092,
      "mean_token_accuracy": 0.147148197889328,
      "num_tokens": 2856001.0,
      "step": 1280
    },
    {
      "entropy": 6.341158008575439,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.1444,
      "mean_token_accuracy": 0.14794962406158446,
      "num_tokens": 2867703.0,
      "step": 1285
    },
    {
      "entropy": 6.199103784561157,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999880330714,
      "loss": 6.0821,
      "mean_token_accuracy": 0.1512378692626953,
      "num_tokens": 2878671.0,
      "step": 1290
    },
    {
      "entropy": 6.32013258934021,
      "epoch": 0.1454484191610041,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.2049,
      "mean_token_accuracy": 0.14154347777366638,
      "num_tokens": 2890137.0,
      "step": 1295
    },
    {
      "entropy": 6.280849170684815,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.108,
      "mean_token_accuracy": 0.14886464178562164,
      "num_tokens": 2900344.0,
      "step": 1300
    },
    {
      "entropy": 6.263293170928955,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.1274,
      "mean_token_accuracy": 0.14697801023721696,
      "num_tokens": 2912270.0,
      "step": 1305
    },
    {
      "entropy": 6.3712437629699705,
      "epoch": 0.14713314988487675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.206,
      "mean_token_accuracy": 0.13799021914601325,
      "num_tokens": 2924677.0,
      "step": 1310
    },
    {
      "entropy": 6.215903329849243,
      "epoch": 0.14769472679283427,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.1221,
      "mean_token_accuracy": 0.15069050639867781,
      "num_tokens": 2935794.0,
      "step": 1315
    },
    {
      "entropy": 6.380605936050415,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.2322,
      "mean_token_accuracy": 0.14286753982305528,
      "num_tokens": 2945958.0,
      "step": 1320
    },
    {
      "entropy": 6.292030143737793,
      "epoch": 0.14881788060874937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.1825,
      "mean_token_accuracy": 0.14438321217894554,
      "num_tokens": 2957536.0,
      "step": 1325
    },
    {
      "entropy": 6.38133692741394,
      "epoch": 0.14937945751670692,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.1535,
      "mean_token_accuracy": 0.14794423878192903,
      "num_tokens": 2967787.0,
      "step": 1330
    },
    {
      "entropy": 6.29340763092041,
      "epoch": 0.14994103442466447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499984016250827,
      "loss": 6.1824,
      "mean_token_accuracy": 0.14508575797080994,
      "num_tokens": 2980522.0,
      "step": 1335
    },
    {
      "entropy": 6.2017439842224125,
      "epoch": 0.150502611332622,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499983534119342,
      "loss": 6.124,
      "mean_token_accuracy": 0.15016201436519622,
      "num_tokens": 2990497.0,
      "step": 1340
    },
    {
      "entropy": 6.38787727355957,
      "epoch": 0.15106418824057954,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.2135,
      "mean_token_accuracy": 0.14598810970783233,
      "num_tokens": 3002931.0,
      "step": 1345
    },
    {
      "entropy": 6.314008617401123,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.134,
      "mean_token_accuracy": 0.1540141612291336,
      "num_tokens": 3014137.0,
      "step": 1350
    },
    {
      "entropy": 6.420076322555542,
      "epoch": 0.15218734205649465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.309,
      "mean_token_accuracy": 0.13888537734746934,
      "num_tokens": 3026145.0,
      "step": 1355
    },
    {
      "entropy": 6.252240705490112,
      "epoch": 0.15274891896445217,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.1566,
      "mean_token_accuracy": 0.1455765649676323,
      "num_tokens": 3037068.0,
      "step": 1360
    },
    {
      "entropy": 6.313040637969971,
      "epoch": 0.15331049587240972,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.0889,
      "mean_token_accuracy": 0.1504852518439293,
      "num_tokens": 3047836.0,
      "step": 1365
    },
    {
      "entropy": 6.320889616012574,
      "epoch": 0.15387207278036727,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.2165,
      "mean_token_accuracy": 0.14922351911664009,
      "num_tokens": 3060309.0,
      "step": 1370
    },
    {
      "entropy": 6.3300347328186035,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.1815,
      "mean_token_accuracy": 0.14712618961930274,
      "num_tokens": 3071789.0,
      "step": 1375
    },
    {
      "entropy": 6.247071981430054,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499979419180532,
      "loss": 6.0938,
      "mean_token_accuracy": 0.15286702811717987,
      "num_tokens": 3081756.0,
      "step": 1380
    },
    {
      "entropy": 6.298181962966919,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.21,
      "mean_token_accuracy": 0.14156180024147033,
      "num_tokens": 3093456.0,
      "step": 1385
    },
    {
      "entropy": 6.382714557647705,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.1381,
      "mean_token_accuracy": 0.14954148083925248,
      "num_tokens": 3104565.0,
      "step": 1390
    },
    {
      "entropy": 6.256937789916992,
      "epoch": 0.156679957320155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999777578813904,
      "loss": 6.1531,
      "mean_token_accuracy": 0.14698541909456253,
      "num_tokens": 3115577.0,
      "step": 1395
    },
    {
      "entropy": 6.24382963180542,
      "epoch": 0.15724153422811254,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499977189788228,
      "loss": 6.0549,
      "mean_token_accuracy": 0.14838047921657563,
      "num_tokens": 3126019.0,
      "step": 1400
    },
    {
      "entropy": 6.324401712417602,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.2172,
      "mean_token_accuracy": 0.13645764887332917,
      "num_tokens": 3136191.0,
      "step": 1405
    },
    {
      "entropy": 6.369932317733765,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.0737,
      "mean_token_accuracy": 0.1490643858909607,
      "num_tokens": 3147737.0,
      "step": 1410
    },
    {
      "entropy": 6.144622421264648,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.1337,
      "mean_token_accuracy": 0.14620147198438643,
      "num_tokens": 3158494.0,
      "step": 1415
    },
    {
      "entropy": 6.409546136856079,
      "epoch": 0.15948784185994272,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.1659,
      "mean_token_accuracy": 0.14886748641729355,
      "num_tokens": 3170550.0,
      "step": 1420
    },
    {
      "entropy": 6.171326398849487,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1544,
      "mean_token_accuracy": 0.15117276310920716,
      "num_tokens": 3182860.0,
      "step": 1425
    },
    {
      "entropy": 6.3079229354858395,
      "epoch": 0.16061099567585782,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.1532,
      "mean_token_accuracy": 0.14341338276863097,
      "num_tokens": 3193807.0,
      "step": 1430
    },
    {
      "entropy": 6.1183257579803465,
      "epoch": 0.16117257258381534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.0216,
      "mean_token_accuracy": 0.14781626164913178,
      "num_tokens": 3203959.0,
      "step": 1435
    },
    {
      "entropy": 6.274162673950196,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.039,
      "mean_token_accuracy": 0.15461266785860062,
      "num_tokens": 3215416.0,
      "step": 1440
    },
    {
      "entropy": 6.329146146774292,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.1911,
      "mean_token_accuracy": 0.1445060230791569,
      "num_tokens": 3226938.0,
      "step": 1445
    },
    {
      "entropy": 6.1372476577758786,
      "epoch": 0.162857303307688,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999711148748476,
      "loss": 5.9971,
      "mean_token_accuracy": 0.1559592068195343,
      "num_tokens": 3238406.0,
      "step": 1450
    },
    {
      "entropy": 6.184646844863892,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.023,
      "mean_token_accuracy": 0.1487284079194069,
      "num_tokens": 3249695.0,
      "step": 1455
    },
    {
      "entropy": 6.337225914001465,
      "epoch": 0.16398045712360307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.1617,
      "mean_token_accuracy": 0.15224283635616304,
      "num_tokens": 3259656.0,
      "step": 1460
    },
    {
      "entropy": 6.184023857116699,
      "epoch": 0.16454203403156062,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.0,
      "mean_token_accuracy": 0.15234700739383697,
      "num_tokens": 3269712.0,
      "step": 1465
    },
    {
      "entropy": 6.068046569824219,
      "epoch": 0.16510361093951817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999684843396968,
      "loss": 5.9915,
      "mean_token_accuracy": 0.15861952006816865,
      "num_tokens": 3280343.0,
      "step": 1470
    },
    {
      "entropy": 6.2015856266021725,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.0279,
      "mean_token_accuracy": 0.1526598796248436,
      "num_tokens": 3291649.0,
      "step": 1475
    },
    {
      "entropy": 6.3056028366088865,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.2112,
      "mean_token_accuracy": 0.14913778901100158,
      "num_tokens": 3304190.0,
      "step": 1480
    },
    {
      "entropy": 6.126730823516846,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999664362258814,
      "loss": 5.9306,
      "mean_token_accuracy": 0.16392721086740494,
      "num_tokens": 3313966.0,
      "step": 1485
    },
    {
      "entropy": 6.21044430732727,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.1883,
      "mean_token_accuracy": 0.1507849559187889,
      "num_tokens": 3326017.0,
      "step": 1490
    },
    {
      "entropy": 6.236493635177612,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499965035001542,
      "loss": 6.0788,
      "mean_token_accuracy": 0.14554568827152253,
      "num_tokens": 3338264.0,
      "step": 1495
    },
    {
      "entropy": 6.257958507537841,
      "epoch": 0.16847307238726345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.1231,
      "mean_token_accuracy": 0.1576220154762268,
      "num_tokens": 3348901.0,
      "step": 1500
    },
    {
      "entropy": 6.217171573638916,
      "epoch": 0.16903464929522097,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.0476,
      "mean_token_accuracy": 0.15184800177812577,
      "num_tokens": 3358441.0,
      "step": 1505
    },
    {
      "entropy": 6.293072080612182,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.133,
      "mean_token_accuracy": 0.15614369064569472,
      "num_tokens": 3369334.0,
      "step": 1510
    },
    {
      "entropy": 6.208935117721557,
      "epoch": 0.17015780311113607,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.1283,
      "mean_token_accuracy": 0.15123554393649102,
      "num_tokens": 3380698.0,
      "step": 1515
    },
    {
      "entropy": 6.426233530044556,
      "epoch": 0.17071938001909362,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.1402,
      "mean_token_accuracy": 0.1447037622332573,
      "num_tokens": 3391061.0,
      "step": 1520
    },
    {
      "entropy": 6.1783136367797855,
      "epoch": 0.17128095692705117,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.1831,
      "mean_token_accuracy": 0.15091484934091567,
      "num_tokens": 3403983.0,
      "step": 1525
    },
    {
      "entropy": 6.342173147201538,
      "epoch": 0.1718425338350087,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049995990508388,
      "loss": 6.1231,
      "mean_token_accuracy": 0.15125393867492676,
      "num_tokens": 3413838.0,
      "step": 1530
    },
    {
      "entropy": 6.076597023010254,
      "epoch": 0.17240411074296624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999591435870674,
      "loss": 5.9419,
      "mean_token_accuracy": 0.15306319296360016,
      "num_tokens": 3423978.0,
      "step": 1535
    },
    {
      "entropy": 6.309637880325317,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.1192,
      "mean_token_accuracy": 0.15553057491779326,
      "num_tokens": 3433416.0,
      "step": 1540
    },
    {
      "entropy": 6.0907206535339355,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999575991051074,
      "loss": 5.97,
      "mean_token_accuracy": 0.15673767775297165,
      "num_tokens": 3444983.0,
      "step": 1545
    },
    {
      "entropy": 6.13786096572876,
      "epoch": 0.1740888414668389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999568161200093,
      "loss": 5.9552,
      "mean_token_accuracy": 0.16142054796218872,
      "num_tokens": 3455808.0,
      "step": 1550
    },
    {
      "entropy": 6.1685872077941895,
      "epoch": 0.17465041837479642,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.1104,
      "mean_token_accuracy": 0.14527395665645598,
      "num_tokens": 3467503.0,
      "step": 1555
    },
    {
      "entropy": 6.331913900375366,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.2931,
      "mean_token_accuracy": 0.14554888010025024,
      "num_tokens": 3479941.0,
      "step": 1560
    },
    {
      "entropy": 6.151180124282837,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999544241885421,
      "loss": 5.9826,
      "mean_token_accuracy": 0.15955932140350343,
      "num_tokens": 3491069.0,
      "step": 1565
    },
    {
      "entropy": 6.212237930297851,
      "epoch": 0.17633514909866907,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.1606,
      "mean_token_accuracy": 0.1459153823554516,
      "num_tokens": 3502065.0,
      "step": 1570
    },
    {
      "entropy": 6.338152885437012,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499952793754341,
      "loss": 6.1222,
      "mean_token_accuracy": 0.15344193577766418,
      "num_tokens": 3514228.0,
      "step": 1575
    },
    {
      "entropy": 6.192629480361939,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.0525,
      "mean_token_accuracy": 0.15235214978456496,
      "num_tokens": 3525801.0,
      "step": 1580
    },
    {
      "entropy": 6.1850494861602785,
      "epoch": 0.1780198798225417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999511346698035,
      "loss": 5.9837,
      "mean_token_accuracy": 0.1552681341767311,
      "num_tokens": 3536431.0,
      "step": 1585
    },
    {
      "entropy": 6.141394138336182,
      "epoch": 0.17858145673049924,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.0957,
      "mean_token_accuracy": 0.15150652602314948,
      "num_tokens": 3549098.0,
      "step": 1590
    },
    {
      "entropy": 6.296343088150024,
      "epoch": 0.1791430336384568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0648,
      "mean_token_accuracy": 0.14865811467170714,
      "num_tokens": 3560532.0,
      "step": 1595
    },
    {
      "entropy": 6.116127634048462,
      "epoch": 0.17970461054641435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0503,
      "mean_token_accuracy": 0.15432106032967569,
      "num_tokens": 3572175.0,
      "step": 1600
    },
    {
      "entropy": 6.233989953994751,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999477305505675,
      "loss": 5.9729,
      "mean_token_accuracy": 0.17390876933932303,
      "num_tokens": 3582862.0,
      "step": 1605
    },
    {
      "entropy": 6.209008359909058,
      "epoch": 0.18082776436232942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.1604,
      "mean_token_accuracy": 0.1523948520421982,
      "num_tokens": 3594685.0,
      "step": 1610
    },
    {
      "entropy": 6.118362140655518,
      "epoch": 0.18138934127028697,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9363,
      "mean_token_accuracy": 0.1629515752196312,
      "num_tokens": 3604847.0,
      "step": 1615
    },
    {
      "entropy": 6.185347604751587,
      "epoch": 0.18195091817824452,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499945102255605,
      "loss": 6.045,
      "mean_token_accuracy": 0.15654354467988013,
      "num_tokens": 3615549.0,
      "step": 1620
    },
    {
      "entropy": 6.171912479400635,
      "epoch": 0.18251249508620204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.9686,
      "mean_token_accuracy": 0.16261090338230133,
      "num_tokens": 3626300.0,
      "step": 1625
    },
    {
      "entropy": 6.0599151134490965,
      "epoch": 0.1830740719941596,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999433142472205,
      "loss": 5.9836,
      "mean_token_accuracy": 0.15348479002714158,
      "num_tokens": 3637832.0,
      "step": 1630
    },
    {
      "entropy": 6.173618030548096,
      "epoch": 0.18363564890211714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.0249,
      "mean_token_accuracy": 0.1517447590827942,
      "num_tokens": 3649179.0,
      "step": 1635
    },
    {
      "entropy": 6.176636791229248,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.0489,
      "mean_token_accuracy": 0.15198502838611602,
      "num_tokens": 3660011.0,
      "step": 1640
    },
    {
      "entropy": 6.254629039764405,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.1367,
      "mean_token_accuracy": 0.14882190227508546,
      "num_tokens": 3672791.0,
      "step": 1645
    },
    {
      "entropy": 6.142079782485962,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.9599,
      "mean_token_accuracy": 0.15162472128868104,
      "num_tokens": 3684709.0,
      "step": 1650
    },
    {
      "entropy": 6.06971001625061,
      "epoch": 0.18588195653394732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.9376,
      "mean_token_accuracy": 0.1539404273033142,
      "num_tokens": 3694675.0,
      "step": 1655
    },
    {
      "entropy": 6.053702163696289,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999377783282203,
      "loss": 5.9411,
      "mean_token_accuracy": 0.154044945538044,
      "num_tokens": 3706384.0,
      "step": 1660
    },
    {
      "entropy": 6.222718238830566,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.013,
      "mean_token_accuracy": 0.15491660088300704,
      "num_tokens": 3717419.0,
      "step": 1665
    },
    {
      "entropy": 6.1200758934021,
      "epoch": 0.18756668725781997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.0313,
      "mean_token_accuracy": 0.1560240775346756,
      "num_tokens": 3727306.0,
      "step": 1670
    },
    {
      "entropy": 6.299317455291748,
      "epoch": 0.1881282641657775,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.1666,
      "mean_token_accuracy": 0.15007805228233337,
      "num_tokens": 3739476.0,
      "step": 1675
    },
    {
      "entropy": 6.097905826568604,
      "epoch": 0.18868984107373504,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.9178,
      "mean_token_accuracy": 0.16339074969291686,
      "num_tokens": 3751035.0,
      "step": 1680
    },
    {
      "entropy": 6.131561183929444,
      "epoch": 0.1892514179816926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999329681042092,
      "loss": 6.0198,
      "mean_token_accuracy": 0.15711974799633027,
      "num_tokens": 3762881.0,
      "step": 1685
    },
    {
      "entropy": 6.139648151397705,
      "epoch": 0.18981299488965014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.0316,
      "mean_token_accuracy": 0.15230293124914168,
      "num_tokens": 3774180.0,
      "step": 1690
    },
    {
      "entropy": 6.116710186004639,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.0333,
      "mean_token_accuracy": 0.1552741691470146,
      "num_tokens": 3784864.0,
      "step": 1695
    },
    {
      "entropy": 6.22691388130188,
      "epoch": 0.19093614870556522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.0085,
      "mean_token_accuracy": 0.15241112262010575,
      "num_tokens": 3797693.0,
      "step": 1700
    },
    {
      "entropy": 6.1396164894104,
      "epoch": 0.19149772561352277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.0236,
      "mean_token_accuracy": 0.15405184477567674,
      "num_tokens": 3809379.0,
      "step": 1705
    },
    {
      "entropy": 6.053248691558838,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999279788313858,
      "loss": 5.8776,
      "mean_token_accuracy": 0.16238332390785218,
      "num_tokens": 3819473.0,
      "step": 1710
    },
    {
      "entropy": 6.033798170089722,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999269594913107,
      "loss": 5.8739,
      "mean_token_accuracy": 0.1607612520456314,
      "num_tokens": 3831037.0,
      "step": 1715
    },
    {
      "entropy": 6.053096103668213,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.0182,
      "mean_token_accuracy": 0.15346318036317824,
      "num_tokens": 3841767.0,
      "step": 1720
    },
    {
      "entropy": 6.032930660247803,
      "epoch": 0.19374403324535294,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999248993259081,
      "loss": 5.8539,
      "mean_token_accuracy": 0.16085348576307296,
      "num_tokens": 3851629.0,
      "step": 1725
    },
    {
      "entropy": 6.0723738193511965,
      "epoch": 0.1943056101533105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999238585006463,
      "loss": 5.9286,
      "mean_token_accuracy": 0.15617268905043602,
      "num_tokens": 3863247.0,
      "step": 1730
    },
    {
      "entropy": 6.095740032196045,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.9025,
      "mean_token_accuracy": 0.15338031351566314,
      "num_tokens": 3873949.0,
      "step": 1735
    },
    {
      "entropy": 6.136980056762695,
      "epoch": 0.1954287639692256,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999217553651676,
      "loss": 6.0144,
      "mean_token_accuracy": 0.15692252218723296,
      "num_tokens": 3884992.0,
      "step": 1740
    },
    {
      "entropy": 6.215441608428955,
      "epoch": 0.19599034087718312,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.1379,
      "mean_token_accuracy": 0.1581611804664135,
      "num_tokens": 3897029.0,
      "step": 1745
    },
    {
      "entropy": 6.093951034545898,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999196235833054,
      "loss": 5.9791,
      "mean_token_accuracy": 0.1530992403626442,
      "num_tokens": 3909510.0,
      "step": 1750
    },
    {
      "entropy": 6.112313222885132,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.8794,
      "mean_token_accuracy": 0.1585567131638527,
      "num_tokens": 3920593.0,
      "step": 1755
    },
    {
      "entropy": 6.073207998275757,
      "epoch": 0.19767507160105577,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499917463155331,
      "loss": 6.008,
      "mean_token_accuracy": 0.15076811462640763,
      "num_tokens": 3932215.0,
      "step": 1760
    },
    {
      "entropy": 6.162782859802246,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9614,
      "mean_token_accuracy": 0.1571688860654831,
      "num_tokens": 3943876.0,
      "step": 1765
    },
    {
      "entropy": 6.025495481491089,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049991527408152,
      "loss": 5.9544,
      "mean_token_accuracy": 0.15912806242704391,
      "num_tokens": 3954533.0,
      "step": 1770
    },
    {
      "entropy": 6.150787830352783,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.0102,
      "mean_token_accuracy": 0.1601330891251564,
      "num_tokens": 3965722.0,
      "step": 1775
    },
    {
      "entropy": 6.076108074188232,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.875,
      "learning_rate": 0.000499913056362151,
      "loss": 5.9457,
      "mean_token_accuracy": 0.15454069674015045,
      "num_tokens": 3976737.0,
      "step": 1780
    },
    {
      "entropy": 6.150193881988526,
      "epoch": 0.2004829561408435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999119367604703,
      "loss": 6.0181,
      "mean_token_accuracy": 0.1597037598490715,
      "num_tokens": 3986102.0,
      "step": 1785
    },
    {
      "entropy": 6.126420879364014,
      "epoch": 0.20104453304880104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999108099975064,
      "loss": 5.9904,
      "mean_token_accuracy": 0.15712075233459472,
      "num_tokens": 3996534.0,
      "step": 1790
    },
    {
      "entropy": 6.039429378509522,
      "epoch": 0.20160610995675857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499909676073295,
      "loss": 5.9055,
      "mean_token_accuracy": 0.1583872377872467,
      "num_tokens": 4006947.0,
      "step": 1795
    },
    {
      "entropy": 6.060782194137573,
      "epoch": 0.20216768686471612,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.9449,
      "mean_token_accuracy": 0.16321475505828859,
      "num_tokens": 4017723.0,
      "step": 1800
    },
    {
      "entropy": 6.0788586139678955,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999073867412746,
      "loss": 6.0245,
      "mean_token_accuracy": 0.1539995141327381,
      "num_tokens": 4028347.0,
      "step": 1805
    },
    {
      "entropy": 6.088930892944336,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.8783,
      "mean_token_accuracy": 0.16769046783447267,
      "num_tokens": 4039648.0,
      "step": 1810
    },
    {
      "entropy": 6.099359846115112,
      "epoch": 0.20385241758858877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9335,
      "mean_token_accuracy": 0.1550333634018898,
      "num_tokens": 4050786.0,
      "step": 1815
    },
    {
      "entropy": 6.12097282409668,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999038990347984,
      "loss": 5.9806,
      "mean_token_accuracy": 0.1614360526204109,
      "num_tokens": 4061184.0,
      "step": 1820
    },
    {
      "entropy": 6.204589939117431,
      "epoch": 0.20497557140450384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.1202,
      "mean_token_accuracy": 0.15479563772678376,
      "num_tokens": 4073607.0,
      "step": 1825
    },
    {
      "entropy": 6.022753477096558,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499901538091949,
      "loss": 5.9277,
      "mean_token_accuracy": 0.1573045566678047,
      "num_tokens": 4083924.0,
      "step": 1830
    },
    {
      "entropy": 6.086099147796631,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999003468790773,
      "loss": 5.9193,
      "mean_token_accuracy": 0.16215644925832748,
      "num_tokens": 4094083.0,
      "step": 1835
    },
    {
      "entropy": 6.078646850585938,
      "epoch": 0.2066603021283765,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.9532,
      "mean_token_accuracy": 0.16525087803602218,
      "num_tokens": 4105124.0,
      "step": 1840
    },
    {
      "entropy": 6.219698572158814,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499897942970629,
      "loss": 6.1184,
      "mean_token_accuracy": 0.15077613070607185,
      "num_tokens": 4116792.0,
      "step": 1845
    },
    {
      "entropy": 6.221433782577515,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499896730275129,
      "loss": 6.096,
      "mean_token_accuracy": 0.15168196707963943,
      "num_tokens": 4127441.0,
      "step": 1850
    },
    {
      "entropy": 6.097982692718506,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049989551041883,
      "loss": 6.0898,
      "mean_token_accuracy": 0.15172615349292756,
      "num_tokens": 4138696.0,
      "step": 1855
    },
    {
      "entropy": 6.1136155128479,
      "epoch": 0.20890660976020667,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998942834017707,
      "loss": 5.9484,
      "mean_token_accuracy": 0.1588933140039444,
      "num_tokens": 4151254.0,
      "step": 1860
    },
    {
      "entropy": 6.051373052597046,
      "epoch": 0.2094681866681642,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.9153,
      "mean_token_accuracy": 0.1542994648218155,
      "num_tokens": 4162046.0,
      "step": 1865
    },
    {
      "entropy": 6.029537487030029,
      "epoch": 0.21002976357612174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998918078855277,
      "loss": 5.8608,
      "mean_token_accuracy": 0.16153852343559266,
      "num_tokens": 4172409.0,
      "step": 1870
    },
    {
      "entropy": 6.062843751907349,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998905593864227,
      "loss": 6.0243,
      "mean_token_accuracy": 0.16137329041957854,
      "num_tokens": 4182949.0,
      "step": 1875
    },
    {
      "entropy": 6.069054412841797,
      "epoch": 0.21115291739203684,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998893037267153,
      "loss": 5.9882,
      "mean_token_accuracy": 0.1569446697831154,
      "num_tokens": 4194164.0,
      "step": 1880
    },
    {
      "entropy": 6.163977527618409,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998880409064452,
      "loss": 5.9555,
      "mean_token_accuracy": 0.15890365242958068,
      "num_tokens": 4206056.0,
      "step": 1885
    },
    {
      "entropy": 6.018830442428589,
      "epoch": 0.21227607120795192,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998867709256527,
      "loss": 6.0183,
      "mean_token_accuracy": 0.1555065929889679,
      "num_tokens": 4216660.0,
      "step": 1890
    },
    {
      "entropy": 6.058361387252807,
      "epoch": 0.21283764811590947,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499885493784378,
      "loss": 5.8838,
      "mean_token_accuracy": 0.1596136897802353,
      "num_tokens": 4228495.0,
      "step": 1895
    },
    {
      "entropy": 6.068961715698242,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.8992,
      "mean_token_accuracy": 0.15846104472875594,
      "num_tokens": 4238560.0,
      "step": 1900
    },
    {
      "entropy": 6.199825143814087,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.1062,
      "mean_token_accuracy": 0.1520417958498001,
      "num_tokens": 4250553.0,
      "step": 1905
    },
    {
      "entropy": 6.102573680877685,
      "epoch": 0.21452237883978212,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998816193980701,
      "loss": 6.0304,
      "mean_token_accuracy": 0.1546974703669548,
      "num_tokens": 4262050.0,
      "step": 1910
    },
    {
      "entropy": 6.224831295013428,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.975,
      "mean_token_accuracy": 0.14990944266319275,
      "num_tokens": 4273019.0,
      "step": 1915
    },
    {
      "entropy": 6.058222246170044,
      "epoch": 0.2156455326556972,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8647,
      "mean_token_accuracy": 0.1674592226743698,
      "num_tokens": 4283585.0,
      "step": 1920
    },
    {
      "entropy": 6.047630691528321,
      "epoch": 0.21620710956365474,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.937,
      "mean_token_accuracy": 0.16005693525075912,
      "num_tokens": 4294905.0,
      "step": 1925
    },
    {
      "entropy": 6.1609485149383545,
      "epoch": 0.2167686864716123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998763533054045,
      "loss": 6.0415,
      "mean_token_accuracy": 0.15440894961357116,
      "num_tokens": 4305926.0,
      "step": 1930
    },
    {
      "entropy": 5.985680294036865,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998750188817568,
      "loss": 6.0186,
      "mean_token_accuracy": 0.15445369929075242,
      "num_tokens": 4318297.0,
      "step": 1935
    },
    {
      "entropy": 6.032767105102539,
      "epoch": 0.21789184028752737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.84,
      "mean_token_accuracy": 0.16766211837530137,
      "num_tokens": 4328795.0,
      "step": 1940
    },
    {
      "entropy": 6.012192440032959,
      "epoch": 0.21845341719548492,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499872328554181,
      "loss": 5.8628,
      "mean_token_accuracy": 0.15800613760948182,
      "num_tokens": 4340279.0,
      "step": 1945
    },
    {
      "entropy": 6.122261905670166,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998709726503383,
      "loss": 6.0275,
      "mean_token_accuracy": 0.15114919617772102,
      "num_tokens": 4352157.0,
      "step": 1950
    },
    {
      "entropy": 6.07682991027832,
      "epoch": 0.21957657101140002,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998696095865168,
      "loss": 5.8643,
      "mean_token_accuracy": 0.16682649552822112,
      "num_tokens": 4361909.0,
      "step": 1955
    },
    {
      "entropy": 5.879463386535645,
      "epoch": 0.22013814791935757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.7878,
      "mean_token_accuracy": 0.1570300579071045,
      "num_tokens": 4372177.0,
      "step": 1960
    },
    {
      "entropy": 5.987885189056397,
      "epoch": 0.2206997248273151,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8391,
      "mean_token_accuracy": 0.16166026890277863,
      "num_tokens": 4382872.0,
      "step": 1965
    },
    {
      "entropy": 6.05077543258667,
      "epoch": 0.22126130173527264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998654774356131,
      "loss": 5.9305,
      "mean_token_accuracy": 0.15829528868198395,
      "num_tokens": 4393623.0,
      "step": 1970
    },
    {
      "entropy": 6.013175201416016,
      "epoch": 0.2218228786432302,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998640857323118,
      "loss": 5.8682,
      "mean_token_accuracy": 0.16012260168790818,
      "num_tokens": 4404215.0,
      "step": 1975
    },
    {
      "entropy": 6.117317819595337,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998626868692509,
      "loss": 6.1271,
      "mean_token_accuracy": 0.156798305362463,
      "num_tokens": 4415978.0,
      "step": 1980
    },
    {
      "entropy": 6.133135509490967,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998612808464748,
      "loss": 5.8725,
      "mean_token_accuracy": 0.16460159718990325,
      "num_tokens": 4427060.0,
      "step": 1985
    },
    {
      "entropy": 5.928405904769898,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.9522,
      "mean_token_accuracy": 0.15562185198068618,
      "num_tokens": 4438145.0,
      "step": 1990
    },
    {
      "entropy": 6.232151079177856,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.0299,
      "mean_token_accuracy": 0.15950280129909516,
      "num_tokens": 4449324.0,
      "step": 1995
    },
    {
      "entropy": 5.970883798599243,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.7709,
      "mean_token_accuracy": 0.16454007625579833,
      "num_tokens": 4459728.0,
      "step": 2000
    },
    {
      "entropy": 6.013013935089111,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9703,
      "mean_token_accuracy": 0.1580568253993988,
      "num_tokens": 4471164.0,
      "step": 2005
    },
    {
      "entropy": 6.116335105895996,
      "epoch": 0.225753916998933,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0193,
      "mean_token_accuracy": 0.1572705999016762,
      "num_tokens": 4483349.0,
      "step": 2010
    },
    {
      "entropy": 6.119635820388794,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998526943583226,
      "loss": 5.986,
      "mean_token_accuracy": 0.15683266669511794,
      "num_tokens": 4495311.0,
      "step": 2015
    },
    {
      "entropy": 5.996360874176025,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998512382188066,
      "loss": 5.9179,
      "mean_token_accuracy": 0.15888708531856538,
      "num_tokens": 4507017.0,
      "step": 2020
    },
    {
      "entropy": 6.032978343963623,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9484,
      "mean_token_accuracy": 0.15785446017980576,
      "num_tokens": 4517746.0,
      "step": 2025
    },
    {
      "entropy": 6.110689783096314,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998483044617694,
      "loss": 6.0512,
      "mean_token_accuracy": 0.14857769906520843,
      "num_tokens": 4529149.0,
      "step": 2030
    },
    {
      "entropy": 6.109778976440429,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.9216,
      "mean_token_accuracy": 0.15968868732452393,
      "num_tokens": 4540630.0,
      "step": 2035
    },
    {
      "entropy": 5.904194164276123,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.8583,
      "mean_token_accuracy": 0.1592124655842781,
      "num_tokens": 4552877.0,
      "step": 2040
    },
    {
      "entropy": 6.08227162361145,
      "epoch": 0.22968495535463582,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998438501319025,
      "loss": 5.953,
      "mean_token_accuracy": 0.16537819653749466,
      "num_tokens": 4563634.0,
      "step": 2045
    },
    {
      "entropy": 6.079555511474609,
      "epoch": 0.23024653226259337,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.8719,
      "mean_token_accuracy": 0.16541080325841903,
      "num_tokens": 4574007.0,
      "step": 2050
    },
    {
      "entropy": 5.959904050827026,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.8867,
      "mean_token_accuracy": 0.1608034625649452,
      "num_tokens": 4584146.0,
      "step": 2055
    },
    {
      "entropy": 6.00870270729065,
      "epoch": 0.23136968607850844,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.9059,
      "mean_token_accuracy": 0.1573877900838852,
      "num_tokens": 4594839.0,
      "step": 2060
    },
    {
      "entropy": 6.191771459579468,
      "epoch": 0.231931262986466,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998378107980233,
      "loss": 6.0678,
      "mean_token_accuracy": 0.1526424765586853,
      "num_tokens": 4608229.0,
      "step": 2065
    },
    {
      "entropy": 6.032376861572265,
      "epoch": 0.23249283989442354,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998362830671249,
      "loss": 5.8767,
      "mean_token_accuracy": 0.16054777950048446,
      "num_tokens": 4619206.0,
      "step": 2070
    },
    {
      "entropy": 5.951002788543701,
      "epoch": 0.2330544168023811,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998347481773522,
      "loss": 5.9053,
      "mean_token_accuracy": 0.16458992809057235,
      "num_tokens": 4628966.0,
      "step": 2075
    },
    {
      "entropy": 6.079168605804443,
      "epoch": 0.23361599371033864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998332061287538,
      "loss": 5.7883,
      "mean_token_accuracy": 0.1636739507317543,
      "num_tokens": 4639935.0,
      "step": 2080
    },
    {
      "entropy": 6.051212406158447,
      "epoch": 0.23417757061829617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499831656921379,
      "loss": 6.0266,
      "mean_token_accuracy": 0.15532584488391876,
      "num_tokens": 4651998.0,
      "step": 2085
    },
    {
      "entropy": 6.094482707977295,
      "epoch": 0.23473914752625372,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499830100555277,
      "loss": 6.0291,
      "mean_token_accuracy": 0.15595776289701463,
      "num_tokens": 4664002.0,
      "step": 2090
    },
    {
      "entropy": 6.037249612808227,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998285370304974,
      "loss": 5.8896,
      "mean_token_accuracy": 0.16437548100948335,
      "num_tokens": 4675343.0,
      "step": 2095
    },
    {
      "entropy": 6.085287809371948,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049982696634709,
      "loss": 5.9351,
      "mean_token_accuracy": 0.16324877589941025,
      "num_tokens": 4686410.0,
      "step": 2100
    },
    {
      "entropy": 6.030001497268676,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.9697,
      "mean_token_accuracy": 0.16037132889032363,
      "num_tokens": 4697768.0,
      "step": 2105
    },
    {
      "entropy": 6.1078695297241214,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499823803504592,
      "loss": 5.9731,
      "mean_token_accuracy": 0.1577034831047058,
      "num_tokens": 4709933.0,
      "step": 2110
    },
    {
      "entropy": 6.090886545181275,
      "epoch": 0.23754703206604144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998222113456022,
      "loss": 6.014,
      "mean_token_accuracy": 0.15369939506053926,
      "num_tokens": 4721385.0,
      "step": 2115
    },
    {
      "entropy": 5.996338558197022,
      "epoch": 0.238108608973999,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998206120281858,
      "loss": 5.8288,
      "mean_token_accuracy": 0.16938644796609878,
      "num_tokens": 4732562.0,
      "step": 2120
    },
    {
      "entropy": 5.999813508987427,
      "epoch": 0.23867018588195654,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499819005552394,
      "loss": 5.8352,
      "mean_token_accuracy": 0.16146625876426696,
      "num_tokens": 4742794.0,
      "step": 2125
    },
    {
      "entropy": 5.974365329742431,
      "epoch": 0.23923176278991407,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.814,
      "mean_token_accuracy": 0.16607148945331573,
      "num_tokens": 4754903.0,
      "step": 2130
    },
    {
      "entropy": 5.948705959320068,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499815771125889,
      "loss": 5.8377,
      "mean_token_accuracy": 0.16147608906030655,
      "num_tokens": 4766165.0,
      "step": 2135
    },
    {
      "entropy": 6.005882406234742,
      "epoch": 0.24035491660582917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998141431752785,
      "loss": 5.8587,
      "mean_token_accuracy": 0.16653385609388352,
      "num_tokens": 4776551.0,
      "step": 2140
    },
    {
      "entropy": 6.128078126907349,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998125080664987,
      "loss": 6.0525,
      "mean_token_accuracy": 0.1545305222272873,
      "num_tokens": 4788344.0,
      "step": 2145
    },
    {
      "entropy": 6.060748529434204,
      "epoch": 0.24147807042174427,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.8475,
      "mean_token_accuracy": 0.15840872079133989,
      "num_tokens": 4799314.0,
      "step": 2150
    },
    {
      "entropy": 5.97811918258667,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.8964,
      "mean_token_accuracy": 0.16318009197711944,
      "num_tokens": 4809588.0,
      "step": 2155
    },
    {
      "entropy": 5.96418137550354,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.7132,
      "mean_token_accuracy": 0.1681472808122635,
      "num_tokens": 4820076.0,
      "step": 2160
    },
    {
      "entropy": 6.002331733703613,
      "epoch": 0.2431628011456169,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998058960507286,
      "loss": 5.992,
      "mean_token_accuracy": 0.15780160278081895,
      "num_tokens": 4831504.0,
      "step": 2165
    },
    {
      "entropy": 6.020361518859863,
      "epoch": 0.24372437805357444,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.8343,
      "mean_token_accuracy": 0.16251724809408188,
      "num_tokens": 4842826.0,
      "step": 2170
    },
    {
      "entropy": 5.909654855728149,
      "epoch": 0.244285954961532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998025470951908,
      "loss": 5.824,
      "mean_token_accuracy": 0.16593606919050216,
      "num_tokens": 4853279.0,
      "step": 2175
    },
    {
      "entropy": 6.0372607707977295,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.8828,
      "mean_token_accuracy": 0.16533545404672623,
      "num_tokens": 4863332.0,
      "step": 2180
    },
    {
      "entropy": 5.999047374725341,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997991695084523,
      "loss": 6.0049,
      "mean_token_accuracy": 0.15240172892808915,
      "num_tokens": 4874217.0,
      "step": 2185
    },
    {
      "entropy": 6.115704011917114,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997974699785169,
      "loss": 5.8562,
      "mean_token_accuracy": 0.1665976122021675,
      "num_tokens": 4885322.0,
      "step": 2190
    },
    {
      "entropy": 5.871498727798462,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.8132,
      "mean_token_accuracy": 0.16305921524763106,
      "num_tokens": 4895546.0,
      "step": 2195
    },
    {
      "entropy": 6.153412675857544,
      "epoch": 0.24709383950131972,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997940494457846,
      "loss": 6.0032,
      "mean_token_accuracy": 0.15596965700387955,
      "num_tokens": 4906957.0,
      "step": 2200
    },
    {
      "entropy": 6.124114465713501,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9452,
      "mean_token_accuracy": 0.1516578808426857,
      "num_tokens": 4917965.0,
      "step": 2205
    },
    {
      "entropy": 5.93087248802185,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.8238,
      "mean_token_accuracy": 0.1671748161315918,
      "num_tokens": 4928585.0,
      "step": 2210
    },
    {
      "entropy": 5.9913819313049315,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499788864965351,
      "loss": 5.833,
      "mean_token_accuracy": 0.16251434683799743,
      "num_tokens": 4939818.0,
      "step": 2215
    },
    {
      "entropy": 6.075236415863037,
      "epoch": 0.2493401471331499,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997871224904036,
      "loss": 5.9279,
      "mean_token_accuracy": 0.160538013279438,
      "num_tokens": 4951752.0,
      "step": 2220
    },
    {
      "entropy": 5.8859203338623045,
      "epoch": 0.24990172404110744,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.7528,
      "mean_token_accuracy": 0.1742950603365898,
      "num_tokens": 4961846.0,
      "step": 2225
    },
    {
      "entropy": 5.972510623931885,
      "epoch": 0.250463300949065,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997836160686368,
      "loss": 5.9287,
      "mean_token_accuracy": 0.15616793632507325,
      "num_tokens": 4973693.0,
      "step": 2230
    },
    {
      "entropy": 5.999047088623047,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.8239,
      "mean_token_accuracy": 0.16658675223588942,
      "num_tokens": 4986156.0,
      "step": 2235
    },
    {
      "entropy": 5.945030450820923,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.9059,
      "mean_token_accuracy": 0.16279772520065308,
      "num_tokens": 4997615.0,
      "step": 2240
    },
    {
      "entropy": 6.058555555343628,
      "epoch": 0.2521480316729376,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.8432,
      "mean_token_accuracy": 0.16085562109947205,
      "num_tokens": 5007655.0,
      "step": 2245
    },
    {
      "entropy": 5.952951717376709,
      "epoch": 0.25270960858089514,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.9153,
      "mean_token_accuracy": 0.1652421697974205,
      "num_tokens": 5018933.0,
      "step": 2250
    },
    {
      "entropy": 6.1129766464233395,
      "epoch": 0.2532711854888527,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.9081,
      "mean_token_accuracy": 0.16212444603443146,
      "num_tokens": 5030495.0,
      "step": 2255
    },
    {
      "entropy": 6.000020503997803,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.96,
      "mean_token_accuracy": 0.15744963735342027,
      "num_tokens": 5041553.0,
      "step": 2260
    },
    {
      "entropy": 5.949009561538697,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.8867,
      "mean_token_accuracy": 0.16389947086572648,
      "num_tokens": 5052312.0,
      "step": 2265
    },
    {
      "entropy": 6.130438995361328,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997693040981922,
      "loss": 6.0774,
      "mean_token_accuracy": 0.15607239678502083,
      "num_tokens": 5063445.0,
      "step": 2270
    },
    {
      "entropy": 6.100005340576172,
      "epoch": 0.2555174931206829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8637,
      "mean_token_accuracy": 0.1615953415632248,
      "num_tokens": 5073052.0,
      "step": 2275
    },
    {
      "entropy": 5.964355039596557,
      "epoch": 0.25607907002864044,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.8996,
      "mean_token_accuracy": 0.162807098031044,
      "num_tokens": 5084174.0,
      "step": 2280
    },
    {
      "entropy": 6.064599895477295,
      "epoch": 0.256640646936598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499763819021459,
      "loss": 5.8753,
      "mean_token_accuracy": 0.162973290681839,
      "num_tokens": 5094106.0,
      "step": 2285
    },
    {
      "entropy": 6.088547229766846,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.9606,
      "mean_token_accuracy": 0.16254740357398986,
      "num_tokens": 5106963.0,
      "step": 2290
    },
    {
      "entropy": 5.932668828964234,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499760126520704,
      "loss": 5.7586,
      "mean_token_accuracy": 0.15899893045425414,
      "num_tokens": 5119476.0,
      "step": 2295
    },
    {
      "entropy": 6.0435731410980225,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997582695356225,
      "loss": 6.0162,
      "mean_token_accuracy": 0.15558764040470124,
      "num_tokens": 5132169.0,
      "step": 2300
    },
    {
      "entropy": 5.967125463485718,
      "epoch": 0.25888695456842814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.8449,
      "mean_token_accuracy": 0.1600498989224434,
      "num_tokens": 5143354.0,
      "step": 2305
    },
    {
      "entropy": 5.956926536560059,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.9234,
      "mean_token_accuracy": 0.15473302602767944,
      "num_tokens": 5154624.0,
      "step": 2310
    },
    {
      "entropy": 6.064026832580566,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997526556422733,
      "loss": 5.805,
      "mean_token_accuracy": 0.1624491482973099,
      "num_tokens": 5165349.0,
      "step": 2315
    },
    {
      "entropy": 5.973732566833496,
      "epoch": 0.2605716852923008,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.8869,
      "mean_token_accuracy": 0.16123015880584718,
      "num_tokens": 5175874.0,
      "step": 2320
    },
    {
      "entropy": 5.963538599014282,
      "epoch": 0.26113326220025834,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997488772655492,
      "loss": 5.7897,
      "mean_token_accuracy": 0.16631451696157457,
      "num_tokens": 5186591.0,
      "step": 2325
    },
    {
      "entropy": 5.951702833175659,
      "epoch": 0.2616948391082159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.8832,
      "mean_token_accuracy": 0.16798330396413802,
      "num_tokens": 5197307.0,
      "step": 2330
    },
    {
      "entropy": 5.961882829666138,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.8451,
      "mean_token_accuracy": 0.16027480661869048,
      "num_tokens": 5209868.0,
      "step": 2335
    },
    {
      "entropy": 5.992185497283936,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.8627,
      "mean_token_accuracy": 0.1646285265684128,
      "num_tokens": 5220365.0,
      "step": 2340
    },
    {
      "entropy": 5.977267789840698,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.7728,
      "mean_token_accuracy": 0.1643584579229355,
      "num_tokens": 5231575.0,
      "step": 2345
    },
    {
      "entropy": 5.980089855194092,
      "epoch": 0.26394114674004604,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.8646,
      "mean_token_accuracy": 0.16169388443231583,
      "num_tokens": 5241852.0,
      "step": 2350
    },
    {
      "entropy": 6.0129913806915285,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.9098,
      "mean_token_accuracy": 0.16636008620262147,
      "num_tokens": 5254536.0,
      "step": 2355
    },
    {
      "entropy": 6.019440126419068,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.8246,
      "mean_token_accuracy": 0.16354746818542482,
      "num_tokens": 5264946.0,
      "step": 2360
    },
    {
      "entropy": 5.750850582122803,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499733477519865,
      "loss": 5.6565,
      "mean_token_accuracy": 0.16620689481496811,
      "num_tokens": 5275527.0,
      "step": 2365
    },
    {
      "entropy": 5.940898036956787,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997315203508013,
      "loss": 5.8218,
      "mean_token_accuracy": 0.16179233640432358,
      "num_tokens": 5285471.0,
      "step": 2370
    },
    {
      "entropy": 5.861508893966675,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.7549,
      "mean_token_accuracy": 0.1623705729842186,
      "num_tokens": 5295743.0,
      "step": 2375
    },
    {
      "entropy": 6.0062055587768555,
      "epoch": 0.26731060818779134,
      "grad_norm": 1.0,
      "learning_rate": 0.00049972758454612,
      "loss": 5.8892,
      "mean_token_accuracy": 0.15389246344566346,
      "num_tokens": 5307963.0,
      "step": 2380
    },
    {
      "entropy": 5.945818948745727,
      "epoch": 0.26787218509574884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.7893,
      "mean_token_accuracy": 0.16452015787363053,
      "num_tokens": 5318394.0,
      "step": 2385
    },
    {
      "entropy": 5.984850311279297,
      "epoch": 0.2684337620037064,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004997236201197853,
      "loss": 6.0264,
      "mean_token_accuracy": 0.1485483929514885,
      "num_tokens": 5331728.0,
      "step": 2390
    },
    {
      "entropy": 6.003773021697998,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17206651121377944,
      "num_tokens": 5342825.0,
      "step": 2395
    },
    {
      "entropy": 6.011975288391113,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.8037,
      "mean_token_accuracy": 0.16357291787862777,
      "num_tokens": 5354232.0,
      "step": 2400
    },
    {
      "entropy": 5.938725376129151,
      "epoch": 0.27011849272757904,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.8948,
      "mean_token_accuracy": 0.15622272118926048,
      "num_tokens": 5364807.0,
      "step": 2405
    },
    {
      "entropy": 5.948577547073365,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.8751,
      "mean_token_accuracy": 0.15974455922842026,
      "num_tokens": 5375593.0,
      "step": 2410
    },
    {
      "entropy": 6.011623287200928,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.8221,
      "mean_token_accuracy": 0.16726389527320862,
      "num_tokens": 5386740.0,
      "step": 2415
    },
    {
      "entropy": 5.8543072700500485,
      "epoch": 0.2718032234514517,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.7997,
      "mean_token_accuracy": 0.16655707061290742,
      "num_tokens": 5397732.0,
      "step": 2420
    },
    {
      "entropy": 6.040705680847168,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499709519239414,
      "loss": 5.87,
      "mean_token_accuracy": 0.15803890973329543,
      "num_tokens": 5409392.0,
      "step": 2425
    },
    {
      "entropy": 5.810342502593994,
      "epoch": 0.2729263772673668,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.8158,
      "mean_token_accuracy": 0.17186738848686217,
      "num_tokens": 5420903.0,
      "step": 2430
    },
    {
      "entropy": 5.911894273757935,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499705426021939,
      "loss": 5.761,
      "mean_token_accuracy": 0.16358311474323273,
      "num_tokens": 5432631.0,
      "step": 2435
    },
    {
      "entropy": 5.893311023712158,
      "epoch": 0.27404953108328184,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499703368681107,
      "loss": 5.7113,
      "mean_token_accuracy": 0.17483898848295212,
      "num_tokens": 5443463.0,
      "step": 2440
    },
    {
      "entropy": 6.013640213012695,
      "epoch": 0.2746111079912394,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997013041856326,
      "loss": 6.0427,
      "mean_token_accuracy": 0.15905722007155418,
      "num_tokens": 5455903.0,
      "step": 2445
    },
    {
      "entropy": 6.00086669921875,
      "epoch": 0.27517268489919694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16078447848558425,
      "num_tokens": 5467201.0,
      "step": 2450
    },
    {
      "entropy": 5.952702760696411,
      "epoch": 0.2757342618071545,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7355,
      "mean_token_accuracy": 0.16615803241729737,
      "num_tokens": 5478109.0,
      "step": 2455
    },
    {
      "entropy": 5.957705736160278,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.8551,
      "mean_token_accuracy": 0.16423224210739135,
      "num_tokens": 5489621.0,
      "step": 2460
    },
    {
      "entropy": 5.93759388923645,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.875,
      "learning_rate": 0.000499692974658629,
      "loss": 5.815,
      "mean_token_accuracy": 0.16715700626373292,
      "num_tokens": 5500606.0,
      "step": 2465
    },
    {
      "entropy": 6.0008186340332035,
      "epoch": 0.27741899253102714,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.7597,
      "mean_token_accuracy": 0.1725642591714859,
      "num_tokens": 5510768.0,
      "step": 2470
    },
    {
      "entropy": 5.799460363388062,
      "epoch": 0.2779805694389847,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.781,
      "mean_token_accuracy": 0.16516642272472382,
      "num_tokens": 5522210.0,
      "step": 2475
    },
    {
      "entropy": 6.052443265914917,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.8989,
      "mean_token_accuracy": 0.1589887335896492,
      "num_tokens": 5533055.0,
      "step": 2480
    },
    {
      "entropy": 5.951822566986084,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.9103,
      "mean_token_accuracy": 0.1590309724211693,
      "num_tokens": 5544345.0,
      "step": 2485
    },
    {
      "entropy": 5.969441175460815,
      "epoch": 0.2796653001628573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.8232,
      "mean_token_accuracy": 0.16280512511730194,
      "num_tokens": 5556734.0,
      "step": 2490
    },
    {
      "entropy": 5.991024303436279,
      "epoch": 0.28022687707081484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.8603,
      "mean_token_accuracy": 0.162888403236866,
      "num_tokens": 5566874.0,
      "step": 2495
    },
    {
      "entropy": 5.989664602279663,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8781,
      "mean_token_accuracy": 0.1601580023765564,
      "num_tokens": 5578812.0,
      "step": 2500
    },
    {
      "entropy": 5.951496219635009,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996759722019872,
      "loss": 5.9058,
      "mean_token_accuracy": 0.16278692781925203,
      "num_tokens": 5590021.0,
      "step": 2505
    },
    {
      "entropy": 6.0353498458862305,
      "epoch": 0.2819116077946875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.9117,
      "mean_token_accuracy": 0.1564626947045326,
      "num_tokens": 5600921.0,
      "step": 2510
    },
    {
      "entropy": 5.905371189117432,
      "epoch": 0.28247318470264504,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.825,
      "mean_token_accuracy": 0.16611457020044326,
      "num_tokens": 5610615.0,
      "step": 2515
    },
    {
      "entropy": 5.86320858001709,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499669478241669,
      "loss": 5.7587,
      "mean_token_accuracy": 0.17136051654815673,
      "num_tokens": 5621596.0,
      "step": 2520
    },
    {
      "entropy": 6.044602966308593,
      "epoch": 0.28359633851856014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.8193,
      "mean_token_accuracy": 0.16327339112758638,
      "num_tokens": 5633047.0,
      "step": 2525
    },
    {
      "entropy": 5.8826415061950685,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.7823,
      "mean_token_accuracy": 0.1643436849117279,
      "num_tokens": 5644970.0,
      "step": 2530
    },
    {
      "entropy": 5.896431255340576,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996629198992823,
      "loss": 5.9127,
      "mean_token_accuracy": 0.15781734585762025,
      "num_tokens": 5656631.0,
      "step": 2535
    },
    {
      "entropy": 5.949198150634766,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.8638,
      "mean_token_accuracy": 0.15979290157556533,
      "num_tokens": 5668808.0,
      "step": 2540
    },
    {
      "entropy": 5.890338563919068,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.7855,
      "mean_token_accuracy": 0.16485803872346877,
      "num_tokens": 5679493.0,
      "step": 2545
    },
    {
      "entropy": 6.0119789123535154,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.8199,
      "mean_token_accuracy": 0.1721026435494423,
      "num_tokens": 5689879.0,
      "step": 2550
    },
    {
      "entropy": 5.97424168586731,
      "epoch": 0.2869657999663054,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499654075296133,
      "loss": 5.9128,
      "mean_token_accuracy": 0.16606319844722747,
      "num_tokens": 5699934.0,
      "step": 2555
    },
    {
      "entropy": 5.907690334320068,
      "epoch": 0.28752737687426294,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499651846262487,
      "loss": 5.8203,
      "mean_token_accuracy": 0.16842011064291001,
      "num_tokens": 5710171.0,
      "step": 2560
    },
    {
      "entropy": 6.016073274612427,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.8119,
      "mean_token_accuracy": 0.15843043178319932,
      "num_tokens": 5721190.0,
      "step": 2565
    },
    {
      "entropy": 5.969920349121094,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9343,
      "mean_token_accuracy": 0.15901276618242263,
      "num_tokens": 5733706.0,
      "step": 2570
    },
    {
      "entropy": 5.938461780548096,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.89,
      "mean_token_accuracy": 0.1645979881286621,
      "num_tokens": 5745712.0,
      "step": 2575
    },
    {
      "entropy": 5.821240472793579,
      "epoch": 0.2897736845060931,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.7068,
      "mean_token_accuracy": 0.17919964641332625,
      "num_tokens": 5757915.0,
      "step": 2580
    },
    {
      "entropy": 5.995480871200561,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.8531,
      "mean_token_accuracy": 0.16649403274059296,
      "num_tokens": 5768788.0,
      "step": 2585
    },
    {
      "entropy": 5.961299180984497,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.7885,
      "mean_token_accuracy": 0.16060467511415483,
      "num_tokens": 5779806.0,
      "step": 2590
    },
    {
      "entropy": 5.804096269607544,
      "epoch": 0.29145841522996574,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499636042746914,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1604410633444786,
      "num_tokens": 5790177.0,
      "step": 2595
    },
    {
      "entropy": 5.949581050872803,
      "epoch": 0.2920199921379233,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499633756491259,
      "loss": 5.7803,
      "mean_token_accuracy": 0.17098623216152192,
      "num_tokens": 5800251.0,
      "step": 2600
    },
    {
      "entropy": 5.966438150405883,
      "epoch": 0.29258156904588084,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.8654,
      "mean_token_accuracy": 0.16320379972457885,
      "num_tokens": 5812210.0,
      "step": 2605
    },
    {
      "entropy": 5.956333351135254,
      "epoch": 0.2931431459538384,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.8503,
      "mean_token_accuracy": 0.16726463437080383,
      "num_tokens": 5823280.0,
      "step": 2610
    },
    {
      "entropy": 6.0089202404022215,
      "epoch": 0.29370472286179594,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.9115,
      "mean_token_accuracy": 0.1580904632806778,
      "num_tokens": 5835076.0,
      "step": 2615
    },
    {
      "entropy": 5.815197277069092,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499624539945111,
      "loss": 5.7054,
      "mean_token_accuracy": 0.17123110741376876,
      "num_tokens": 5845376.0,
      "step": 2620
    },
    {
      "entropy": 5.862633275985718,
      "epoch": 0.294827876677711,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.7421,
      "mean_token_accuracy": 0.16225717961788177,
      "num_tokens": 5857124.0,
      "step": 2625
    },
    {
      "entropy": 5.990662527084351,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.8554,
      "mean_token_accuracy": 0.15759045258164406,
      "num_tokens": 5868681.0,
      "step": 2630
    },
    {
      "entropy": 5.914901399612427,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.838,
      "mean_token_accuracy": 0.16877058297395706,
      "num_tokens": 5880532.0,
      "step": 2635
    },
    {
      "entropy": 5.950133657455444,
      "epoch": 0.29651260740158364,
      "grad_norm": 1.0,
      "learning_rate": 0.000499615208964848,
      "loss": 5.7916,
      "mean_token_accuracy": 0.1670111671090126,
      "num_tokens": 5891582.0,
      "step": 2640
    },
    {
      "entropy": 5.908686256408691,
      "epoch": 0.2970741843095412,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.7954,
      "mean_token_accuracy": 0.16432597637176513,
      "num_tokens": 5903543.0,
      "step": 2645
    },
    {
      "entropy": 5.854915571212769,
      "epoch": 0.29763576121749874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.7566,
      "mean_token_accuracy": 0.17550766468048096,
      "num_tokens": 5913140.0,
      "step": 2650
    },
    {
      "entropy": 5.953357362747193,
      "epoch": 0.2981973381254563,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.7344,
      "mean_token_accuracy": 0.16562701612710953,
      "num_tokens": 5924010.0,
      "step": 2655
    },
    {
      "entropy": 5.931583070755005,
      "epoch": 0.29875891503341384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.9304,
      "mean_token_accuracy": 0.16155474483966828,
      "num_tokens": 5936246.0,
      "step": 2660
    },
    {
      "entropy": 5.950146579742432,
      "epoch": 0.2993204919413714,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.8149,
      "mean_token_accuracy": 0.1572188138961792,
      "num_tokens": 5947594.0,
      "step": 2665
    },
    {
      "entropy": 5.957934617996216,
      "epoch": 0.29988206884932894,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996009979409025,
      "loss": 5.7606,
      "mean_token_accuracy": 0.16158466041088104,
      "num_tokens": 5957660.0,
      "step": 2670
    },
    {
      "entropy": 5.738403224945069,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.5848,
      "mean_token_accuracy": 0.17791350483894347,
      "num_tokens": 5968881.0,
      "step": 2675
    },
    {
      "entropy": 5.894114017486572,
      "epoch": 0.301005222665244,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049959620372105,
      "loss": 5.7614,
      "mean_token_accuracy": 0.16772934794425964,
      "num_tokens": 5980032.0,
      "step": 2680
    },
    {
      "entropy": 5.77871789932251,
      "epoch": 0.30156679957320154,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.7557,
      "mean_token_accuracy": 0.16537490487098694,
      "num_tokens": 5991781.0,
      "step": 2685
    },
    {
      "entropy": 5.914988374710083,
      "epoch": 0.3021283764811591,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.7879,
      "mean_token_accuracy": 0.1649462789297104,
      "num_tokens": 6002303.0,
      "step": 2690
    },
    {
      "entropy": 6.009264516830444,
      "epoch": 0.30268995338911664,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.8814,
      "mean_token_accuracy": 0.1677633598446846,
      "num_tokens": 6011523.0,
      "step": 2695
    },
    {
      "entropy": 5.906981372833252,
      "epoch": 0.3032515302970742,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499586529467197,
      "loss": 5.6844,
      "mean_token_accuracy": 0.16914817541837693,
      "num_tokens": 6022169.0,
      "step": 2700
    },
    {
      "entropy": 5.74225082397461,
      "epoch": 0.30381310720503174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.7428,
      "mean_token_accuracy": 0.16678209528326987,
      "num_tokens": 6033393.0,
      "step": 2705
    },
    {
      "entropy": 5.971656274795532,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.8107,
      "mean_token_accuracy": 0.15811437368392944,
      "num_tokens": 6045365.0,
      "step": 2710
    },
    {
      "entropy": 5.846801328659057,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.8361,
      "mean_token_accuracy": 0.1662740647792816,
      "num_tokens": 6057482.0,
      "step": 2715
    },
    {
      "entropy": 5.941861152648926,
      "epoch": 0.30549783792890434,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.7375,
      "mean_token_accuracy": 0.17222627848386765,
      "num_tokens": 6068194.0,
      "step": 2720
    },
    {
      "entropy": 5.903083658218383,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.7946,
      "mean_token_accuracy": 0.1700889363884926,
      "num_tokens": 6080540.0,
      "step": 2725
    },
    {
      "entropy": 5.854546451568604,
      "epoch": 0.30662099174481944,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499571803560316,
      "loss": 5.8443,
      "mean_token_accuracy": 0.16272160336375235,
      "num_tokens": 6092464.0,
      "step": 2730
    },
    {
      "entropy": 5.971442317962646,
      "epoch": 0.307182568652777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.7935,
      "mean_token_accuracy": 0.16596091389656067,
      "num_tokens": 6103242.0,
      "step": 2735
    },
    {
      "entropy": 5.849540948867798,
      "epoch": 0.30774414556073454,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.8144,
      "mean_token_accuracy": 0.1654301702976227,
      "num_tokens": 6116383.0,
      "step": 2740
    },
    {
      "entropy": 5.819859552383423,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.6267,
      "mean_token_accuracy": 0.17252051830291748,
      "num_tokens": 6126883.0,
      "step": 2745
    },
    {
      "entropy": 5.9390562057495115,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.8212,
      "mean_token_accuracy": 0.16041162610054016,
      "num_tokens": 6139613.0,
      "step": 2750
    },
    {
      "entropy": 5.920775699615478,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.883,
      "mean_token_accuracy": 0.1595902293920517,
      "num_tokens": 6151950.0,
      "step": 2755
    },
    {
      "entropy": 5.84655532836914,
      "epoch": 0.30999045319256474,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.6638,
      "mean_token_accuracy": 0.17385415136814117,
      "num_tokens": 6161563.0,
      "step": 2760
    },
    {
      "entropy": 5.809947443008423,
      "epoch": 0.3105520301005223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.6793,
      "mean_token_accuracy": 0.17475582361221315,
      "num_tokens": 6170966.0,
      "step": 2765
    },
    {
      "entropy": 5.866269159317016,
      "epoch": 0.3111136070084798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.7548,
      "mean_token_accuracy": 0.16865830570459367,
      "num_tokens": 6182563.0,
      "step": 2770
    },
    {
      "entropy": 5.905898571014404,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.8659,
      "mean_token_accuracy": 0.1606164261698723,
      "num_tokens": 6194495.0,
      "step": 2775
    },
    {
      "entropy": 5.8810731887817385,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.7628,
      "mean_token_accuracy": 0.16512676775455476,
      "num_tokens": 6206594.0,
      "step": 2780
    },
    {
      "entropy": 5.955170059204102,
      "epoch": 0.31279833773235244,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.7886,
      "mean_token_accuracy": 0.16353470236063003,
      "num_tokens": 6218371.0,
      "step": 2785
    },
    {
      "entropy": 5.889581632614136,
      "epoch": 0.31335991464031,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.7207,
      "mean_token_accuracy": 0.17165483981370927,
      "num_tokens": 6228723.0,
      "step": 2790
    },
    {
      "entropy": 5.917896366119384,
      "epoch": 0.31392149154826754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.908,
      "mean_token_accuracy": 0.16421120017766952,
      "num_tokens": 6240535.0,
      "step": 2795
    },
    {
      "entropy": 5.842968416213989,
      "epoch": 0.3144830684562251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1732317879796028,
      "num_tokens": 6250643.0,
      "step": 2800
    },
    {
      "entropy": 5.829766130447387,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.7341,
      "mean_token_accuracy": 0.16942535787820817,
      "num_tokens": 6259569.0,
      "step": 2805
    },
    {
      "entropy": 5.917622089385986,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499531276042413,
      "loss": 5.9022,
      "mean_token_accuracy": 0.17028167098760605,
      "num_tokens": 6271108.0,
      "step": 2810
    },
    {
      "entropy": 5.926321697235108,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.738,
      "mean_token_accuracy": 0.15809677839279174,
      "num_tokens": 6282352.0,
      "step": 2815
    },
    {
      "entropy": 5.87783522605896,
      "epoch": 0.31672937608805524,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.8509,
      "mean_token_accuracy": 0.1608995482325554,
      "num_tokens": 6294946.0,
      "step": 2820
    },
    {
      "entropy": 5.803282737731934,
      "epoch": 0.3172909529960128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.6583,
      "mean_token_accuracy": 0.17535417079925536,
      "num_tokens": 6305552.0,
      "step": 2825
    },
    {
      "entropy": 5.835207843780518,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.7333,
      "mean_token_accuracy": 0.1743067905306816,
      "num_tokens": 6315815.0,
      "step": 2830
    },
    {
      "entropy": 5.907580804824829,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.7671,
      "mean_token_accuracy": 0.16423598527908326,
      "num_tokens": 6327384.0,
      "step": 2835
    },
    {
      "entropy": 5.8704465389251705,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.7459,
      "mean_token_accuracy": 0.15973282158374785,
      "num_tokens": 6338083.0,
      "step": 2840
    },
    {
      "entropy": 5.971950387954712,
      "epoch": 0.319537260627843,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.7529,
      "mean_token_accuracy": 0.17029481679201125,
      "num_tokens": 6349058.0,
      "step": 2845
    },
    {
      "entropy": 5.814688491821289,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.7498,
      "mean_token_accuracy": 0.1704786464571953,
      "num_tokens": 6359655.0,
      "step": 2850
    },
    {
      "entropy": 5.95844259262085,
      "epoch": 0.3206604144437581,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.8915,
      "mean_token_accuracy": 0.16460290402173997,
      "num_tokens": 6372088.0,
      "step": 2855
    },
    {
      "entropy": 5.860590982437134,
      "epoch": 0.32122199135171564,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.7776,
      "mean_token_accuracy": 0.16560305953025817,
      "num_tokens": 6382946.0,
      "step": 2860
    },
    {
      "entropy": 5.871022176742554,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.7859,
      "mean_token_accuracy": 0.16105389073491097,
      "num_tokens": 6392591.0,
      "step": 2865
    },
    {
      "entropy": 5.835945558547974,
      "epoch": 0.3223451451676307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.722,
      "mean_token_accuracy": 0.17206769585609435,
      "num_tokens": 6403159.0,
      "step": 2870
    },
    {
      "entropy": 5.90753059387207,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7444,
      "mean_token_accuracy": 0.16411917731165887,
      "num_tokens": 6414232.0,
      "step": 2875
    },
    {
      "entropy": 5.856598854064941,
      "epoch": 0.3234682989835458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.7139,
      "mean_token_accuracy": 0.17788330912590028,
      "num_tokens": 6425762.0,
      "step": 2880
    },
    {
      "entropy": 5.817793512344361,
      "epoch": 0.32402987589150334,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.7755,
      "mean_token_accuracy": 0.16134170442819595,
      "num_tokens": 6439077.0,
      "step": 2885
    },
    {
      "entropy": 5.9469287395477295,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.831,
      "mean_token_accuracy": 0.16600467413663864,
      "num_tokens": 6450425.0,
      "step": 2890
    },
    {
      "entropy": 5.849802923202515,
      "epoch": 0.32515302970741844,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.7945,
      "mean_token_accuracy": 0.16656314879655837,
      "num_tokens": 6461347.0,
      "step": 2895
    },
    {
      "entropy": 5.884488153457641,
      "epoch": 0.325714606615376,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.7833,
      "mean_token_accuracy": 0.1678066909313202,
      "num_tokens": 6472428.0,
      "step": 2900
    },
    {
      "entropy": 5.844870376586914,
      "epoch": 0.32627618352333354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.7445,
      "mean_token_accuracy": 0.17545183449983598,
      "num_tokens": 6483393.0,
      "step": 2905
    },
    {
      "entropy": 5.855503273010254,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.7243,
      "mean_token_accuracy": 0.16932949125766755,
      "num_tokens": 6493520.0,
      "step": 2910
    },
    {
      "entropy": 5.791838979721069,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.6659,
      "mean_token_accuracy": 0.1761859968304634,
      "num_tokens": 6505282.0,
      "step": 2915
    },
    {
      "entropy": 5.974535512924194,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.838,
      "mean_token_accuracy": 0.17055148035287857,
      "num_tokens": 6516808.0,
      "step": 2920
    },
    {
      "entropy": 5.885997676849366,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16254309713840484,
      "num_tokens": 6529184.0,
      "step": 2925
    },
    {
      "entropy": 5.788518714904785,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.6459,
      "mean_token_accuracy": 0.17723568826913833,
      "num_tokens": 6539909.0,
      "step": 2930
    },
    {
      "entropy": 5.893859624862671,
      "epoch": 0.3296456449710788,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.6959,
      "mean_token_accuracy": 0.1725839927792549,
      "num_tokens": 6550870.0,
      "step": 2935
    },
    {
      "entropy": 5.777626419067383,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.875,
      "learning_rate": 0.000499461515538303,
      "loss": 5.7488,
      "mean_token_accuracy": 0.17353251129388808,
      "num_tokens": 6561784.0,
      "step": 2940
    },
    {
      "entropy": 5.808140802383423,
      "epoch": 0.3307687987869939,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.7035,
      "mean_token_accuracy": 0.16959256380796434,
      "num_tokens": 6573422.0,
      "step": 2945
    },
    {
      "entropy": 5.940847444534302,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499455949211845,
      "loss": 5.7666,
      "mean_token_accuracy": 0.17316267341375352,
      "num_tokens": 6584565.0,
      "step": 2950
    },
    {
      "entropy": 5.905264854431152,
      "epoch": 0.331891952602909,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.6778,
      "mean_token_accuracy": 0.1691496714949608,
      "num_tokens": 6594238.0,
      "step": 2955
    },
    {
      "entropy": 5.750309991836548,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.7469,
      "mean_token_accuracy": 0.1695124924182892,
      "num_tokens": 6606309.0,
      "step": 2960
    },
    {
      "entropy": 5.996778583526611,
      "epoch": 0.33301510641882404,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.8541,
      "mean_token_accuracy": 0.16220774352550507,
      "num_tokens": 6617916.0,
      "step": 2965
    },
    {
      "entropy": 5.843599700927735,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.7021,
      "mean_token_accuracy": 0.1709466338157654,
      "num_tokens": 6627951.0,
      "step": 2970
    },
    {
      "entropy": 5.760891056060791,
      "epoch": 0.33413826023473914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.7622,
      "mean_token_accuracy": 0.16756141930818558,
      "num_tokens": 6639285.0,
      "step": 2975
    },
    {
      "entropy": 5.866618871688843,
      "epoch": 0.3346998371426967,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.6267,
      "mean_token_accuracy": 0.17637762576341628,
      "num_tokens": 6649973.0,
      "step": 2980
    },
    {
      "entropy": 5.75209150314331,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.648,
      "mean_token_accuracy": 0.1709219753742218,
      "num_tokens": 6660625.0,
      "step": 2985
    },
    {
      "entropy": 5.846449995040894,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00049943339804258,
      "loss": 5.7373,
      "mean_token_accuracy": 0.16657833606004716,
      "num_tokens": 6673191.0,
      "step": 2990
    },
    {
      "entropy": 5.888789987564087,
      "epoch": 0.33638456786656934,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.7279,
      "mean_token_accuracy": 0.17310751378536224,
      "num_tokens": 6683993.0,
      "step": 2995
    },
    {
      "entropy": 5.8018608570098875,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.7611,
      "mean_token_accuracy": 0.1704024627804756,
      "num_tokens": 6694881.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.657197658690343,
      "eval_loss": 5.7566962242126465,
      "eval_mean_token_accuracy": 0.17553335150354513,
      "eval_num_tokens": 6694881.0,
      "eval_runtime": 18.5774,
      "eval_samples_per_second": 1669.392,
      "eval_steps_per_second": 208.694,
      "step": 3000
    },
    {
      "entropy": 5.858089637756348,
      "epoch": 0.33750772168248444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.7458,
      "mean_token_accuracy": 0.17071258276700974,
      "num_tokens": 6706648.0,
      "step": 3005
    },
    {
      "entropy": 5.8388909816741945,
      "epoch": 0.33806929859044194,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499421950949953,
      "loss": 5.7625,
      "mean_token_accuracy": 0.16937266886234284,
      "num_tokens": 6716858.0,
      "step": 3010
    },
    {
      "entropy": 5.7809854507446286,
      "epoch": 0.3386308754983995,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499419071312363,
      "loss": 5.7096,
      "mean_token_accuracy": 0.17317416965961457,
      "num_tokens": 6726558.0,
      "step": 3015
    },
    {
      "entropy": 5.939871549606323,
      "epoch": 0.33919245240635704,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.7147,
      "mean_token_accuracy": 0.16636066883802414,
      "num_tokens": 6737819.0,
      "step": 3020
    },
    {
      "entropy": 5.799683427810669,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.664,
      "mean_token_accuracy": 0.17010829448699952,
      "num_tokens": 6748593.0,
      "step": 3025
    },
    {
      "entropy": 5.711278963088989,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.7694,
      "mean_token_accuracy": 0.16216303557157516,
      "num_tokens": 6760631.0,
      "step": 3030
    },
    {
      "entropy": 5.949622297286988,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.6601,
      "mean_token_accuracy": 0.1694652944803238,
      "num_tokens": 6772318.0,
      "step": 3035
    },
    {
      "entropy": 5.7371148586273195,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.7078,
      "mean_token_accuracy": 0.17238104939460755,
      "num_tokens": 6782684.0,
      "step": 3040
    },
    {
      "entropy": 5.820407104492188,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.5826,
      "mean_token_accuracy": 0.17427664399147033,
      "num_tokens": 6793895.0,
      "step": 3045
    },
    {
      "entropy": 5.696174097061157,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.653,
      "mean_token_accuracy": 0.1724589690566063,
      "num_tokens": 6804794.0,
      "step": 3050
    },
    {
      "entropy": 5.842650270462036,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499395776978143,
      "loss": 5.7775,
      "mean_token_accuracy": 0.16434261351823806,
      "num_tokens": 6815976.0,
      "step": 3055
    },
    {
      "entropy": 5.876256895065308,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.7591,
      "mean_token_accuracy": 0.16504910588264465,
      "num_tokens": 6828027.0,
      "step": 3060
    },
    {
      "entropy": 5.869896650314331,
      "epoch": 0.34424664457797494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.7378,
      "mean_token_accuracy": 0.17055140286684037,
      "num_tokens": 6839195.0,
      "step": 3065
    },
    {
      "entropy": 5.6817914009094235,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.6095,
      "mean_token_accuracy": 0.16941186636686326,
      "num_tokens": 6851731.0,
      "step": 3070
    },
    {
      "entropy": 5.866906595230103,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1743558958172798,
      "num_tokens": 6862916.0,
      "step": 3075
    },
    {
      "entropy": 5.924760532379151,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.8302,
      "mean_token_accuracy": 0.16059349328279496,
      "num_tokens": 6876003.0,
      "step": 3080
    },
    {
      "entropy": 5.792921638488769,
      "epoch": 0.34649295220980514,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.7973,
      "mean_token_accuracy": 0.16916952580213546,
      "num_tokens": 6886040.0,
      "step": 3085
    },
    {
      "entropy": 5.852839660644531,
      "epoch": 0.3470545291177627,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.7117,
      "mean_token_accuracy": 0.17139602601528167,
      "num_tokens": 6896758.0,
      "step": 3090
    },
    {
      "entropy": 5.8844445705413815,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.7593,
      "mean_token_accuracy": 0.17215778231620787,
      "num_tokens": 6907287.0,
      "step": 3095
    },
    {
      "entropy": 5.845010614395141,
      "epoch": 0.3481776829336778,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.7747,
      "mean_token_accuracy": 0.162651926279068,
      "num_tokens": 6918992.0,
      "step": 3100
    },
    {
      "entropy": 5.939760494232178,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.7962,
      "mean_token_accuracy": 0.16313686072826386,
      "num_tokens": 6931493.0,
      "step": 3105
    },
    {
      "entropy": 5.766195440292359,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.6154,
      "mean_token_accuracy": 0.17713082432746888,
      "num_tokens": 6941807.0,
      "step": 3110
    },
    {
      "entropy": 5.916352224349976,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.8215,
      "mean_token_accuracy": 0.1672494038939476,
      "num_tokens": 6953436.0,
      "step": 3115
    },
    {
      "entropy": 5.980728197097778,
      "epoch": 0.35042399056550794,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.935,
      "mean_token_accuracy": 0.16128722876310347,
      "num_tokens": 6965247.0,
      "step": 3120
    },
    {
      "entropy": 5.8007549285888675,
      "epoch": 0.3509855674734655,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.7124,
      "mean_token_accuracy": 0.17670172303915024,
      "num_tokens": 6976605.0,
      "step": 3125
    },
    {
      "entropy": 5.792571258544922,
      "epoch": 0.35154714438142304,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.6653,
      "mean_token_accuracy": 0.18100269883871078,
      "num_tokens": 6986205.0,
      "step": 3130
    },
    {
      "entropy": 5.800654935836792,
      "epoch": 0.3521087212893806,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.6362,
      "mean_token_accuracy": 0.1757133722305298,
      "num_tokens": 6997699.0,
      "step": 3135
    },
    {
      "entropy": 5.717818403244019,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.6568,
      "mean_token_accuracy": 0.1722394585609436,
      "num_tokens": 7008968.0,
      "step": 3140
    },
    {
      "entropy": 5.848590326309204,
      "epoch": 0.3532318751052957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499341692905686,
      "loss": 5.6869,
      "mean_token_accuracy": 0.16661932468414306,
      "num_tokens": 7020633.0,
      "step": 3145
    },
    {
      "entropy": 5.777771043777466,
      "epoch": 0.35379345201325324,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.5998,
      "mean_token_accuracy": 0.18080153167247773,
      "num_tokens": 7031187.0,
      "step": 3150
    },
    {
      "entropy": 5.867948865890503,
      "epoch": 0.35435502892121074,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.8282,
      "mean_token_accuracy": 0.16651090681552888,
      "num_tokens": 7042662.0,
      "step": 3155
    },
    {
      "entropy": 5.892845392227173,
      "epoch": 0.3549166058291683,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.8166,
      "mean_token_accuracy": 0.17055323272943496,
      "num_tokens": 7054297.0,
      "step": 3160
    },
    {
      "entropy": 5.839358615875244,
      "epoch": 0.35547818273712584,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.6624,
      "mean_token_accuracy": 0.17549268901348114,
      "num_tokens": 7064961.0,
      "step": 3165
    },
    {
      "entropy": 5.880225944519043,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.8119,
      "mean_token_accuracy": 0.16876525431871414,
      "num_tokens": 7076857.0,
      "step": 3170
    },
    {
      "entropy": 5.863937187194824,
      "epoch": 0.35660133655304094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499323150548625,
      "loss": 5.6935,
      "mean_token_accuracy": 0.17249748408794402,
      "num_tokens": 7088926.0,
      "step": 3175
    },
    {
      "entropy": 5.705014419555664,
      "epoch": 0.3571629134609985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17939966917037964,
      "num_tokens": 7099478.0,
      "step": 3180
    },
    {
      "entropy": 5.778427171707153,
      "epoch": 0.35772449036895604,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.6534,
      "mean_token_accuracy": 0.173947773873806,
      "num_tokens": 7111776.0,
      "step": 3185
    },
    {
      "entropy": 5.905063486099243,
      "epoch": 0.3582860672769136,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.7422,
      "mean_token_accuracy": 0.16857734620571135,
      "num_tokens": 7123084.0,
      "step": 3190
    },
    {
      "entropy": 5.805781412124634,
      "epoch": 0.35884764418487114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.7218,
      "mean_token_accuracy": 0.17042055875062942,
      "num_tokens": 7134973.0,
      "step": 3195
    },
    {
      "entropy": 5.732704782485962,
      "epoch": 0.3594092210928287,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.7469,
      "mean_token_accuracy": 0.1677598163485527,
      "num_tokens": 7147227.0,
      "step": 3200
    },
    {
      "entropy": 5.921637725830078,
      "epoch": 0.3599707980007862,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.7325,
      "mean_token_accuracy": 0.1675512045621872,
      "num_tokens": 7158201.0,
      "step": 3205
    },
    {
      "entropy": 5.748777055740357,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499301192814866,
      "loss": 5.6137,
      "mean_token_accuracy": 0.1745728835463524,
      "num_tokens": 7168909.0,
      "step": 3210
    },
    {
      "entropy": 5.843131256103516,
      "epoch": 0.3610939518167013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.775,
      "mean_token_accuracy": 0.1655922845005989,
      "num_tokens": 7179877.0,
      "step": 3215
    },
    {
      "entropy": 5.809346675872803,
      "epoch": 0.36165552872465884,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.6863,
      "mean_token_accuracy": 0.17956973612308502,
      "num_tokens": 7190823.0,
      "step": 3220
    },
    {
      "entropy": 5.810121297836304,
      "epoch": 0.3622171056326164,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.6106,
      "mean_token_accuracy": 0.1801090881228447,
      "num_tokens": 7201700.0,
      "step": 3225
    },
    {
      "entropy": 5.820446157455445,
      "epoch": 0.36277868254057394,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499288488415819,
      "loss": 5.6769,
      "mean_token_accuracy": 0.1763908237218857,
      "num_tokens": 7213618.0,
      "step": 3230
    },
    {
      "entropy": 5.91421275138855,
      "epoch": 0.3633402594485315,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.8594,
      "mean_token_accuracy": 0.1691361516714096,
      "num_tokens": 7226207.0,
      "step": 3235
    },
    {
      "entropy": 5.769177675247192,
      "epoch": 0.36390183635648904,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.6997,
      "mean_token_accuracy": 0.17611619979143142,
      "num_tokens": 7237545.0,
      "step": 3240
    },
    {
      "entropy": 5.743767881393433,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.6741,
      "mean_token_accuracy": 0.17366254925727845,
      "num_tokens": 7247799.0,
      "step": 3245
    },
    {
      "entropy": 5.803914546966553,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.6631,
      "mean_token_accuracy": 0.16791853904724122,
      "num_tokens": 7260229.0,
      "step": 3250
    },
    {
      "entropy": 5.759283065795898,
      "epoch": 0.36558656708036164,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.768,
      "mean_token_accuracy": 0.16590262502431868,
      "num_tokens": 7271457.0,
      "step": 3255
    },
    {
      "entropy": 5.822077703475952,
      "epoch": 0.3661481439883192,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17224242985248567,
      "num_tokens": 7282124.0,
      "step": 3260
    },
    {
      "entropy": 5.862492942810059,
      "epoch": 0.36670972089627674,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.7743,
      "mean_token_accuracy": 0.1691797837615013,
      "num_tokens": 7292981.0,
      "step": 3265
    },
    {
      "entropy": 5.9926777362823485,
      "epoch": 0.3672712978042343,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.9211,
      "mean_token_accuracy": 0.16726249903440477,
      "num_tokens": 7304591.0,
      "step": 3270
    },
    {
      "entropy": 5.924714994430542,
      "epoch": 0.36783287471219184,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17486585080623626,
      "num_tokens": 7315918.0,
      "step": 3275
    },
    {
      "entropy": 5.794388103485107,
      "epoch": 0.3683944516201494,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.768,
      "mean_token_accuracy": 0.1678200349211693,
      "num_tokens": 7327725.0,
      "step": 3280
    },
    {
      "entropy": 5.939551639556885,
      "epoch": 0.36895602852810694,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.769,
      "mean_token_accuracy": 0.16480651795864104,
      "num_tokens": 7339787.0,
      "step": 3285
    },
    {
      "entropy": 5.906058120727539,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.7585,
      "mean_token_accuracy": 0.17078427225351334,
      "num_tokens": 7351284.0,
      "step": 3290
    },
    {
      "entropy": 5.703143739700318,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.5781,
      "mean_token_accuracy": 0.17687998712062836,
      "num_tokens": 7362105.0,
      "step": 3295
    },
    {
      "entropy": 5.809231376647949,
      "epoch": 0.37064075925197953,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.6889,
      "mean_token_accuracy": 0.16779071241617202,
      "num_tokens": 7375204.0,
      "step": 3300
    },
    {
      "entropy": 5.918311738967896,
      "epoch": 0.3712023361599371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.823,
      "mean_token_accuracy": 0.16624482423067094,
      "num_tokens": 7385853.0,
      "step": 3305
    },
    {
      "entropy": 5.816821050643921,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.6956,
      "mean_token_accuracy": 0.16294027268886566,
      "num_tokens": 7397459.0,
      "step": 3310
    },
    {
      "entropy": 5.808701705932617,
      "epoch": 0.3723254899758522,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.6786,
      "mean_token_accuracy": 0.16604700088500976,
      "num_tokens": 7408996.0,
      "step": 3315
    },
    {
      "entropy": 5.867400789260865,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.7764,
      "mean_token_accuracy": 0.16139495223760605,
      "num_tokens": 7419938.0,
      "step": 3320
    },
    {
      "entropy": 5.851010036468506,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.7791,
      "mean_token_accuracy": 0.16566097140312194,
      "num_tokens": 7430181.0,
      "step": 3325
    },
    {
      "entropy": 5.846432256698608,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7311,
      "mean_token_accuracy": 0.17106661349534988,
      "num_tokens": 7440774.0,
      "step": 3330
    },
    {
      "entropy": 5.875411891937256,
      "epoch": 0.3745717976076824,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499219915847757,
      "loss": 5.6645,
      "mean_token_accuracy": 0.1727786511182785,
      "num_tokens": 7451744.0,
      "step": 3335
    },
    {
      "entropy": 5.661339139938354,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.7037,
      "mean_token_accuracy": 0.17100751996040345,
      "num_tokens": 7463731.0,
      "step": 3340
    },
    {
      "entropy": 5.863027715682984,
      "epoch": 0.37569495142359743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.6873,
      "mean_token_accuracy": 0.17933991998434068,
      "num_tokens": 7474392.0,
      "step": 3345
    },
    {
      "entropy": 5.759320640563965,
      "epoch": 0.376256528331555,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.6061,
      "mean_token_accuracy": 0.1835946798324585,
      "num_tokens": 7484860.0,
      "step": 3350
    },
    {
      "entropy": 5.734007120132446,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499206497413985,
      "loss": 5.6599,
      "mean_token_accuracy": 0.1750404268503189,
      "num_tokens": 7495877.0,
      "step": 3355
    },
    {
      "entropy": 5.874365949630738,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17955808490514755,
      "num_tokens": 7506879.0,
      "step": 3360
    },
    {
      "entropy": 5.757788562774659,
      "epoch": 0.37794125905542764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.563,
      "mean_token_accuracy": 0.17942067682743074,
      "num_tokens": 7517405.0,
      "step": 3365
    },
    {
      "entropy": 5.747072458267212,
      "epoch": 0.3785028359633852,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.6514,
      "mean_token_accuracy": 0.17287282198667525,
      "num_tokens": 7528832.0,
      "step": 3370
    },
    {
      "entropy": 5.802611684799194,
      "epoch": 0.37906441287134274,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.6303,
      "mean_token_accuracy": 0.1695370301604271,
      "num_tokens": 7541048.0,
      "step": 3375
    },
    {
      "entropy": 5.778376054763794,
      "epoch": 0.3796259897793003,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.6463,
      "mean_token_accuracy": 0.17571850568056108,
      "num_tokens": 7551849.0,
      "step": 3380
    },
    {
      "entropy": 5.687811470031738,
      "epoch": 0.38018756668725784,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.6563,
      "mean_token_accuracy": 0.17964737266302108,
      "num_tokens": 7563148.0,
      "step": 3385
    },
    {
      "entropy": 5.927419805526734,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.8546,
      "mean_token_accuracy": 0.16407665908336638,
      "num_tokens": 7575119.0,
      "step": 3390
    },
    {
      "entropy": 5.8714484691619875,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.6919,
      "mean_token_accuracy": 0.17286457866430283,
      "num_tokens": 7585899.0,
      "step": 3395
    },
    {
      "entropy": 5.822103357315063,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.6054,
      "mean_token_accuracy": 0.18119132965803147,
      "num_tokens": 7598938.0,
      "step": 3400
    },
    {
      "entropy": 5.653518533706665,
      "epoch": 0.382433874319088,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.6773,
      "mean_token_accuracy": 0.16799031496047973,
      "num_tokens": 7610087.0,
      "step": 3405
    },
    {
      "entropy": 5.919979476928711,
      "epoch": 0.38299545122704554,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.8033,
      "mean_token_accuracy": 0.1662902429699898,
      "num_tokens": 7622186.0,
      "step": 3410
    },
    {
      "entropy": 5.9257396221160885,
      "epoch": 0.3835570281350031,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.7522,
      "mean_token_accuracy": 0.16626912504434585,
      "num_tokens": 7633501.0,
      "step": 3415
    },
    {
      "entropy": 5.6931455612182615,
      "epoch": 0.38411860504296064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.6726,
      "mean_token_accuracy": 0.17202702462673186,
      "num_tokens": 7643656.0,
      "step": 3420
    },
    {
      "entropy": 5.858004999160767,
      "epoch": 0.3846801819509182,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.6736,
      "mean_token_accuracy": 0.17162217795848847,
      "num_tokens": 7654565.0,
      "step": 3425
    },
    {
      "entropy": 5.783432531356811,
      "epoch": 0.38524175885887574,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.597,
      "mean_token_accuracy": 0.18407575488090516,
      "num_tokens": 7664725.0,
      "step": 3430
    },
    {
      "entropy": 5.68522801399231,
      "epoch": 0.3858033357668333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.6849,
      "mean_token_accuracy": 0.17350034266710282,
      "num_tokens": 7674772.0,
      "step": 3435
    },
    {
      "entropy": 5.9145458221435545,
      "epoch": 0.38636491267479084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499148194874525,
      "loss": 5.7458,
      "mean_token_accuracy": 0.16479451060295106,
      "num_tokens": 7686488.0,
      "step": 3440
    },
    {
      "entropy": 5.87414517402649,
      "epoch": 0.38692648958274833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.7303,
      "mean_token_accuracy": 0.17141515165567398,
      "num_tokens": 7698842.0,
      "step": 3445
    },
    {
      "entropy": 5.709442234039306,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.5136,
      "mean_token_accuracy": 0.17301216572523118,
      "num_tokens": 7709255.0,
      "step": 3450
    },
    {
      "entropy": 5.668497705459595,
      "epoch": 0.38804964339866344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.5699,
      "mean_token_accuracy": 0.17693006694316865,
      "num_tokens": 7720182.0,
      "step": 3455
    },
    {
      "entropy": 5.734848499298096,
      "epoch": 0.388611220306621,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.6022,
      "mean_token_accuracy": 0.18371552228927612,
      "num_tokens": 7730944.0,
      "step": 3460
    },
    {
      "entropy": 5.799044370651245,
      "epoch": 0.38917279721457854,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049913065451149,
      "loss": 5.6763,
      "mean_token_accuracy": 0.17518288940191268,
      "num_tokens": 7741686.0,
      "step": 3465
    },
    {
      "entropy": 5.807097387313843,
      "epoch": 0.3897343741225361,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.6816,
      "mean_token_accuracy": 0.17521786093711852,
      "num_tokens": 7752245.0,
      "step": 3470
    },
    {
      "entropy": 5.700292348861694,
      "epoch": 0.39029595103049364,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.6544,
      "mean_token_accuracy": 0.16735104322433472,
      "num_tokens": 7762622.0,
      "step": 3475
    },
    {
      "entropy": 5.914432334899902,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.7749,
      "mean_token_accuracy": 0.17352273762226106,
      "num_tokens": 7773550.0,
      "step": 3480
    },
    {
      "entropy": 5.745480346679687,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.6523,
      "mean_token_accuracy": 0.1681404784321785,
      "num_tokens": 7784988.0,
      "step": 3485
    },
    {
      "entropy": 5.82211217880249,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.7329,
      "mean_token_accuracy": 0.17301115840673448,
      "num_tokens": 7797706.0,
      "step": 3490
    },
    {
      "entropy": 5.806065511703491,
      "epoch": 0.3925422586623238,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.6162,
      "mean_token_accuracy": 0.1772813007235527,
      "num_tokens": 7808960.0,
      "step": 3495
    },
    {
      "entropy": 5.7256844997406,
      "epoch": 0.39310383557028133,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.5811,
      "mean_token_accuracy": 0.18364969789981841,
      "num_tokens": 7819271.0,
      "step": 3500
    },
    {
      "entropy": 5.773703336715698,
      "epoch": 0.3936654124782389,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.7206,
      "mean_token_accuracy": 0.1715845599770546,
      "num_tokens": 7832257.0,
      "step": 3505
    },
    {
      "entropy": 5.871817970275879,
      "epoch": 0.39422698938619644,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049909863227389,
      "loss": 5.7153,
      "mean_token_accuracy": 0.16756437420845033,
      "num_tokens": 7842684.0,
      "step": 3510
    },
    {
      "entropy": 5.803624534606934,
      "epoch": 0.394788566294154,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.7325,
      "mean_token_accuracy": 0.16659511923789977,
      "num_tokens": 7855031.0,
      "step": 3515
    },
    {
      "entropy": 5.725556135177612,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.5347,
      "mean_token_accuracy": 0.17303010672330857,
      "num_tokens": 7865291.0,
      "step": 3520
    },
    {
      "entropy": 5.670720434188842,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.5068,
      "mean_token_accuracy": 0.17710875421762468,
      "num_tokens": 7876401.0,
      "step": 3525
    },
    {
      "entropy": 5.7161647319793705,
      "epoch": 0.39647329701802664,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.5561,
      "mean_token_accuracy": 0.17671194225549697,
      "num_tokens": 7887787.0,
      "step": 3530
    },
    {
      "entropy": 5.729979181289673,
      "epoch": 0.3970348739259842,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17679326683282853,
      "num_tokens": 7898550.0,
      "step": 3535
    },
    {
      "entropy": 5.715567111968994,
      "epoch": 0.3975964508339417,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.704,
      "mean_token_accuracy": 0.17211074084043504,
      "num_tokens": 7909254.0,
      "step": 3540
    },
    {
      "entropy": 5.9015869140625,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990733265032,
      "loss": 5.7227,
      "mean_token_accuracy": 0.17778433561325074,
      "num_tokens": 7920164.0,
      "step": 3545
    },
    {
      "entropy": 5.739831924438477,
      "epoch": 0.3987196046498568,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.6692,
      "mean_token_accuracy": 0.1687448263168335,
      "num_tokens": 7931390.0,
      "step": 3550
    },
    {
      "entropy": 5.791812801361084,
      "epoch": 0.39928118155781434,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499066032070036,
      "loss": 5.6842,
      "mean_token_accuracy": 0.17125159949064256,
      "num_tokens": 7943765.0,
      "step": 3555
    },
    {
      "entropy": 5.706348848342896,
      "epoch": 0.3998427584657719,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.5262,
      "mean_token_accuracy": 0.17829164266586303,
      "num_tokens": 7954621.0,
      "step": 3560
    },
    {
      "entropy": 5.703236484527588,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499058709099475,
      "loss": 5.7263,
      "mean_token_accuracy": 0.16872446686029435,
      "num_tokens": 7966063.0,
      "step": 3565
    },
    {
      "entropy": 5.8672455787658695,
      "epoch": 0.400965912281687,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.6455,
      "mean_token_accuracy": 0.17750734239816665,
      "num_tokens": 7977391.0,
      "step": 3570
    },
    {
      "entropy": 5.7744307041168215,
      "epoch": 0.40152748918964454,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.7566,
      "mean_token_accuracy": 0.17146649360656738,
      "num_tokens": 7988786.0,
      "step": 3575
    },
    {
      "entropy": 5.78408465385437,
      "epoch": 0.4020890660976021,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.5488,
      "mean_token_accuracy": 0.17537560313940048,
      "num_tokens": 8000848.0,
      "step": 3580
    },
    {
      "entropy": 5.782027244567871,
      "epoch": 0.40265064300555964,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.7072,
      "mean_token_accuracy": 0.17077694535255433,
      "num_tokens": 8012621.0,
      "step": 3585
    },
    {
      "entropy": 5.8876677513122555,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499040276828089,
      "loss": 5.7553,
      "mean_token_accuracy": 0.16768654882907869,
      "num_tokens": 8023599.0,
      "step": 3590
    },
    {
      "entropy": 5.675628709793091,
      "epoch": 0.4037737968214747,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.5721,
      "mean_token_accuracy": 0.17665697187185286,
      "num_tokens": 8037151.0,
      "step": 3595
    },
    {
      "entropy": 5.763100671768188,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.7561,
      "mean_token_accuracy": 0.17029832154512406,
      "num_tokens": 8048606.0,
      "step": 3600
    },
    {
      "entropy": 5.748295736312866,
      "epoch": 0.4048969506373898,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.5478,
      "mean_token_accuracy": 0.18396237939596177,
      "num_tokens": 8058430.0,
      "step": 3605
    },
    {
      "entropy": 5.647242832183838,
      "epoch": 0.40545852754534734,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.5607,
      "mean_token_accuracy": 0.17905217856168748,
      "num_tokens": 8069554.0,
      "step": 3610
    },
    {
      "entropy": 5.888359403610229,
      "epoch": 0.4060201044533049,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.7283,
      "mean_token_accuracy": 0.16652171611785888,
      "num_tokens": 8082131.0,
      "step": 3615
    },
    {
      "entropy": 5.721231842041016,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.5559,
      "mean_token_accuracy": 0.17986351102590561,
      "num_tokens": 8093454.0,
      "step": 3620
    },
    {
      "entropy": 5.828434181213379,
      "epoch": 0.40714325826922,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.8229,
      "mean_token_accuracy": 0.15695670247077942,
      "num_tokens": 8105502.0,
      "step": 3625
    },
    {
      "entropy": 5.798398780822754,
      "epoch": 0.40770483517717754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.6765,
      "mean_token_accuracy": 0.17159896045923234,
      "num_tokens": 8116524.0,
      "step": 3630
    },
    {
      "entropy": 5.787256288528442,
      "epoch": 0.40826641208513503,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499006649337227,
      "loss": 5.6609,
      "mean_token_accuracy": 0.18144576400518417,
      "num_tokens": 8126949.0,
      "step": 3635
    },
    {
      "entropy": 5.767451381683349,
      "epoch": 0.4088279889930926,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.6184,
      "mean_token_accuracy": 0.17964357882738113,
      "num_tokens": 8137346.0,
      "step": 3640
    },
    {
      "entropy": 5.686922883987426,
      "epoch": 0.40938956590105013,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.6491,
      "mean_token_accuracy": 0.16929376274347305,
      "num_tokens": 8148499.0,
      "step": 3645
    },
    {
      "entropy": 5.839772510528564,
      "epoch": 0.4099511428090077,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498995311785053,
      "loss": 5.677,
      "mean_token_accuracy": 0.17342871725559234,
      "num_tokens": 8159645.0,
      "step": 3650
    },
    {
      "entropy": 5.783364295959473,
      "epoch": 0.41051271971696524,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17835670709609985,
      "num_tokens": 8171033.0,
      "step": 3655
    },
    {
      "entropy": 5.714225387573242,
      "epoch": 0.4110742966249228,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.6354,
      "mean_token_accuracy": 0.17413499802350998,
      "num_tokens": 8181582.0,
      "step": 3660
    },
    {
      "entropy": 5.76922025680542,
      "epoch": 0.41163587353288034,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498983910044002,
      "loss": 5.5475,
      "mean_token_accuracy": 0.17680326849222183,
      "num_tokens": 8192432.0,
      "step": 3665
    },
    {
      "entropy": 5.778567695617676,
      "epoch": 0.4121974504408379,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.7374,
      "mean_token_accuracy": 0.17112922370433808,
      "num_tokens": 8203995.0,
      "step": 3670
    },
    {
      "entropy": 5.734625339508057,
      "epoch": 0.41275902734879544,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.5584,
      "mean_token_accuracy": 0.1776363492012024,
      "num_tokens": 8214323.0,
      "step": 3675
    },
    {
      "entropy": 5.681185245513916,
      "epoch": 0.413320604256753,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.6311,
      "mean_token_accuracy": 0.17019776701927186,
      "num_tokens": 8226285.0,
      "step": 3680
    },
    {
      "entropy": 5.8432228565216064,
      "epoch": 0.4138821811647105,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.7865,
      "mean_token_accuracy": 0.16805494725704193,
      "num_tokens": 8237961.0,
      "step": 3685
    },
    {
      "entropy": 5.714262676239014,
      "epoch": 0.41444375807266803,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.5428,
      "mean_token_accuracy": 0.18119778335094452,
      "num_tokens": 8248463.0,
      "step": 3690
    },
    {
      "entropy": 5.808351993560791,
      "epoch": 0.4150053349806256,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.6828,
      "mean_token_accuracy": 0.17469156086444854,
      "num_tokens": 8260342.0,
      "step": 3695
    },
    {
      "entropy": 5.797298431396484,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.7792,
      "mean_token_accuracy": 0.1726776584982872,
      "num_tokens": 8272055.0,
      "step": 3700
    },
    {
      "entropy": 5.935724067687988,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.7157,
      "mean_token_accuracy": 0.1744633972644806,
      "num_tokens": 8282129.0,
      "step": 3705
    },
    {
      "entropy": 5.790315341949463,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.7303,
      "mean_token_accuracy": 0.17217442095279695,
      "num_tokens": 8294663.0,
      "step": 3710
    },
    {
      "entropy": 5.709871244430542,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498945440696253,
      "loss": 5.5439,
      "mean_token_accuracy": 0.17666319161653518,
      "num_tokens": 8305882.0,
      "step": 3715
    },
    {
      "entropy": 5.779970645904541,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.6456,
      "mean_token_accuracy": 0.1782705307006836,
      "num_tokens": 8316604.0,
      "step": 3720
    },
    {
      "entropy": 5.824194240570068,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.7307,
      "mean_token_accuracy": 0.17325419783592225,
      "num_tokens": 8328126.0,
      "step": 3725
    },
    {
      "entropy": 5.738675975799561,
      "epoch": 0.4189363733363284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.658,
      "mean_token_accuracy": 0.17740787118673323,
      "num_tokens": 8339050.0,
      "step": 3730
    },
    {
      "entropy": 5.6924279689788815,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.4788,
      "mean_token_accuracy": 0.1815269708633423,
      "num_tokens": 8350216.0,
      "step": 3735
    },
    {
      "entropy": 5.758772468566894,
      "epoch": 0.4200595271522435,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.6503,
      "mean_token_accuracy": 0.177595092356205,
      "num_tokens": 8362311.0,
      "step": 3740
    },
    {
      "entropy": 5.801399850845337,
      "epoch": 0.42062110406020103,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.815,
      "mean_token_accuracy": 0.17430444210767745,
      "num_tokens": 8373751.0,
      "step": 3745
    },
    {
      "entropy": 5.851537799835205,
      "epoch": 0.4211826809681586,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.6377,
      "mean_token_accuracy": 0.17612933814525605,
      "num_tokens": 8384932.0,
      "step": 3750
    },
    {
      "entropy": 5.687012004852295,
      "epoch": 0.42174425787611614,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.5877,
      "mean_token_accuracy": 0.1752782717347145,
      "num_tokens": 8394973.0,
      "step": 3755
    },
    {
      "entropy": 5.810871505737305,
      "epoch": 0.4223058347840737,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.6826,
      "mean_token_accuracy": 0.17329255491495132,
      "num_tokens": 8406590.0,
      "step": 3760
    },
    {
      "entropy": 5.729452991485596,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.6105,
      "mean_token_accuracy": 0.1726459875702858,
      "num_tokens": 8417127.0,
      "step": 3765
    },
    {
      "entropy": 5.777321815490723,
      "epoch": 0.4234289885999888,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.7116,
      "mean_token_accuracy": 0.1736960381269455,
      "num_tokens": 8429442.0,
      "step": 3770
    },
    {
      "entropy": 5.730407667160034,
      "epoch": 0.42399056550794634,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.6144,
      "mean_token_accuracy": 0.1784180849790573,
      "num_tokens": 8439375.0,
      "step": 3775
    },
    {
      "entropy": 5.735183048248291,
      "epoch": 0.42455214241590383,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.6282,
      "mean_token_accuracy": 0.17678779661655425,
      "num_tokens": 8449817.0,
      "step": 3780
    },
    {
      "entropy": 5.69113392829895,
      "epoch": 0.4251137193238614,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.5298,
      "mean_token_accuracy": 0.18239751309156418,
      "num_tokens": 8460096.0,
      "step": 3785
    },
    {
      "entropy": 5.741287040710449,
      "epoch": 0.42567529623181893,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1769845187664032,
      "num_tokens": 8470241.0,
      "step": 3790
    },
    {
      "entropy": 5.782705640792846,
      "epoch": 0.4262368731397765,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.6562,
      "mean_token_accuracy": 0.16731305718421935,
      "num_tokens": 8481438.0,
      "step": 3795
    },
    {
      "entropy": 5.822871446609497,
      "epoch": 0.42679845004773403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498878406420253,
      "loss": 5.7188,
      "mean_token_accuracy": 0.17122531235218047,
      "num_tokens": 8492588.0,
      "step": 3800
    },
    {
      "entropy": 5.721942901611328,
      "epoch": 0.4273600269556916,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.6419,
      "mean_token_accuracy": 0.18052712380886077,
      "num_tokens": 8502298.0,
      "step": 3805
    },
    {
      "entropy": 5.66594967842102,
      "epoch": 0.42792160386364914,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.484,
      "mean_token_accuracy": 0.17950060963630676,
      "num_tokens": 8513507.0,
      "step": 3810
    },
    {
      "entropy": 5.774780654907227,
      "epoch": 0.4284831807716067,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.5844,
      "mean_token_accuracy": 0.18196558207273483,
      "num_tokens": 8523888.0,
      "step": 3815
    },
    {
      "entropy": 5.728365325927735,
      "epoch": 0.42904475767956424,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498862334234653,
      "loss": 5.5272,
      "mean_token_accuracy": 0.18457784056663512,
      "num_tokens": 8534032.0,
      "step": 3820
    },
    {
      "entropy": 5.720908832550049,
      "epoch": 0.4296063345875218,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.6687,
      "mean_token_accuracy": 0.17761805951595305,
      "num_tokens": 8545544.0,
      "step": 3825
    },
    {
      "entropy": 5.768918371200561,
      "epoch": 0.4301679114954793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.623,
      "mean_token_accuracy": 0.17694772332906722,
      "num_tokens": 8555428.0,
      "step": 3830
    },
    {
      "entropy": 5.666506671905518,
      "epoch": 0.43072948840343683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.5296,
      "mean_token_accuracy": 0.17718225121498107,
      "num_tokens": 8566456.0,
      "step": 3835
    },
    {
      "entropy": 5.7222388744354244,
      "epoch": 0.4312910653113944,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.5437,
      "mean_token_accuracy": 0.1785029098391533,
      "num_tokens": 8578125.0,
      "step": 3840
    },
    {
      "entropy": 5.729517602920533,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.6009,
      "mean_token_accuracy": 0.168761545419693,
      "num_tokens": 8589530.0,
      "step": 3845
    },
    {
      "entropy": 5.785266160964966,
      "epoch": 0.4324142191273095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.7159,
      "mean_token_accuracy": 0.175383123755455,
      "num_tokens": 8601655.0,
      "step": 3850
    },
    {
      "entropy": 5.71242470741272,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.6238,
      "mean_token_accuracy": 0.17698971331119537,
      "num_tokens": 8613355.0,
      "step": 3855
    },
    {
      "entropy": 5.744694662094116,
      "epoch": 0.4335373729432246,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.557,
      "mean_token_accuracy": 0.18139966130256652,
      "num_tokens": 8624154.0,
      "step": 3860
    },
    {
      "entropy": 5.737391424179077,
      "epoch": 0.43409894985118214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049882575484962,
      "loss": 5.646,
      "mean_token_accuracy": 0.1766352981328964,
      "num_tokens": 8633888.0,
      "step": 3865
    },
    {
      "entropy": 5.745247220993042,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.6155,
      "mean_token_accuracy": 0.17549385130405426,
      "num_tokens": 8644965.0,
      "step": 3870
    },
    {
      "entropy": 5.7478595733642575,
      "epoch": 0.4352221036670972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.5673,
      "mean_token_accuracy": 0.18230103850364685,
      "num_tokens": 8655483.0,
      "step": 3875
    },
    {
      "entropy": 5.729026031494141,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498813433436022,
      "loss": 5.5312,
      "mean_token_accuracy": 0.1756706491112709,
      "num_tokens": 8666443.0,
      "step": 3880
    },
    {
      "entropy": 5.772811222076416,
      "epoch": 0.4363452574830123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.6056,
      "mean_token_accuracy": 0.1834358587861061,
      "num_tokens": 8677198.0,
      "step": 3885
    },
    {
      "entropy": 5.677646493911743,
      "epoch": 0.43690683439096983,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.665,
      "mean_token_accuracy": 0.17260809540748595,
      "num_tokens": 8688303.0,
      "step": 3890
    },
    {
      "entropy": 5.7969519138336185,
      "epoch": 0.4374684112989274,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.6255,
      "mean_token_accuracy": 0.17326829582452774,
      "num_tokens": 8700326.0,
      "step": 3895
    },
    {
      "entropy": 5.675735950469971,
      "epoch": 0.43802998820688493,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.5863,
      "mean_token_accuracy": 0.18138237297534943,
      "num_tokens": 8711101.0,
      "step": 3900
    },
    {
      "entropy": 5.720053815841675,
      "epoch": 0.4385915651148425,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.7126,
      "mean_token_accuracy": 0.16878860592842101,
      "num_tokens": 8721751.0,
      "step": 3905
    },
    {
      "entropy": 5.8183698654174805,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.804,
      "mean_token_accuracy": 0.1679702877998352,
      "num_tokens": 8733550.0,
      "step": 3910
    },
    {
      "entropy": 5.783613157272339,
      "epoch": 0.4397147189307576,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17709120810031892,
      "num_tokens": 8744218.0,
      "step": 3915
    },
    {
      "entropy": 5.736134052276611,
      "epoch": 0.44027629583871514,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.6262,
      "mean_token_accuracy": 0.17595020532608033,
      "num_tokens": 8756531.0,
      "step": 3920
    },
    {
      "entropy": 5.849208211898803,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.6765,
      "mean_token_accuracy": 0.1736727774143219,
      "num_tokens": 8768114.0,
      "step": 3925
    },
    {
      "entropy": 5.723468542098999,
      "epoch": 0.4413994496546302,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.6677,
      "mean_token_accuracy": 0.17031940668821335,
      "num_tokens": 8780010.0,
      "step": 3930
    },
    {
      "entropy": 5.824918222427368,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.6229,
      "mean_token_accuracy": 0.16911651045084,
      "num_tokens": 8792183.0,
      "step": 3935
    },
    {
      "entropy": 5.795850419998169,
      "epoch": 0.4425226034705453,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.726,
      "mean_token_accuracy": 0.16847373098134993,
      "num_tokens": 8804513.0,
      "step": 3940
    },
    {
      "entropy": 5.689461374282837,
      "epoch": 0.44308418037850283,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.6025,
      "mean_token_accuracy": 0.17862226516008378,
      "num_tokens": 8814918.0,
      "step": 3945
    },
    {
      "entropy": 5.7146484375,
      "epoch": 0.4436457572864604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.5672,
      "mean_token_accuracy": 0.1769762635231018,
      "num_tokens": 8826176.0,
      "step": 3950
    },
    {
      "entropy": 5.658634424209595,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.5923,
      "mean_token_accuracy": 0.17806206941604613,
      "num_tokens": 8837242.0,
      "step": 3955
    },
    {
      "entropy": 5.828815221786499,
      "epoch": 0.4447689111023755,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.5586,
      "mean_token_accuracy": 0.17913035899400712,
      "num_tokens": 8848972.0,
      "step": 3960
    },
    {
      "entropy": 5.634966707229614,
      "epoch": 0.44533048801033304,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.5469,
      "mean_token_accuracy": 0.16954630315303804,
      "num_tokens": 8861108.0,
      "step": 3965
    },
    {
      "entropy": 5.719815015792847,
      "epoch": 0.4458920649182906,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.6526,
      "mean_token_accuracy": 0.1743015319108963,
      "num_tokens": 8872161.0,
      "step": 3970
    },
    {
      "entropy": 5.9182099342346195,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.718,
      "mean_token_accuracy": 0.17898963540792465,
      "num_tokens": 8882497.0,
      "step": 3975
    },
    {
      "entropy": 5.5615965843200685,
      "epoch": 0.44701521873420563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18167599588632583,
      "num_tokens": 8893155.0,
      "step": 3980
    },
    {
      "entropy": 5.687947511672974,
      "epoch": 0.4475767956421632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498725387911674,
      "loss": 5.5281,
      "mean_token_accuracy": 0.18152017295360565,
      "num_tokens": 8904478.0,
      "step": 3985
    },
    {
      "entropy": 5.703388118743897,
      "epoch": 0.44813837255012073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.6339,
      "mean_token_accuracy": 0.17210570424795152,
      "num_tokens": 8915470.0,
      "step": 3990
    },
    {
      "entropy": 5.788628959655762,
      "epoch": 0.4486999494580783,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498716838763451,
      "loss": 5.581,
      "mean_token_accuracy": 0.17711307406425475,
      "num_tokens": 8925644.0,
      "step": 3995
    },
    {
      "entropy": 5.747158479690552,
      "epoch": 0.44926152636603583,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498712553504425,
      "loss": 5.6075,
      "mean_token_accuracy": 0.1746936023235321,
      "num_tokens": 8937000.0,
      "step": 4000
    },
    {
      "entropy": 5.679166507720947,
      "epoch": 0.4498231032739934,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.6017,
      "mean_token_accuracy": 0.17030701637268067,
      "num_tokens": 8948258.0,
      "step": 4005
    },
    {
      "entropy": 5.750095653533935,
      "epoch": 0.45038468018195094,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.6016,
      "mean_token_accuracy": 0.17791567146778106,
      "num_tokens": 8960107.0,
      "step": 4010
    },
    {
      "entropy": 5.776009845733642,
      "epoch": 0.4509462570899085,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.5992,
      "mean_token_accuracy": 0.17729963958263398,
      "num_tokens": 8969612.0,
      "step": 4015
    },
    {
      "entropy": 5.78485336303711,
      "epoch": 0.451507833997866,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.7057,
      "mean_token_accuracy": 0.1708099812269211,
      "num_tokens": 8980970.0,
      "step": 4020
    },
    {
      "entropy": 5.759588193893433,
      "epoch": 0.45206941090582353,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.6215,
      "mean_token_accuracy": 0.17713819146156312,
      "num_tokens": 8992218.0,
      "step": 4025
    },
    {
      "entropy": 5.645942735671997,
      "epoch": 0.4526309878137811,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.5532,
      "mean_token_accuracy": 0.1778957411646843,
      "num_tokens": 9003737.0,
      "step": 4030
    },
    {
      "entropy": 5.763908529281617,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.6241,
      "mean_token_accuracy": 0.17733240127563477,
      "num_tokens": 9015427.0,
      "step": 4035
    },
    {
      "entropy": 5.7138035774230955,
      "epoch": 0.4537541416296962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17426750659942628,
      "num_tokens": 9025003.0,
      "step": 4040
    },
    {
      "entropy": 5.700944519042968,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.5783,
      "mean_token_accuracy": 0.17825210243463516,
      "num_tokens": 9036074.0,
      "step": 4045
    },
    {
      "entropy": 5.809437561035156,
      "epoch": 0.4548772954456113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.5882,
      "mean_token_accuracy": 0.17707725912332534,
      "num_tokens": 9046878.0,
      "step": 4050
    },
    {
      "entropy": 5.745957756042481,
      "epoch": 0.45543887235356884,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.6437,
      "mean_token_accuracy": 0.1804998129606247,
      "num_tokens": 9058366.0,
      "step": 4055
    },
    {
      "entropy": 5.77444281578064,
      "epoch": 0.4560004492615264,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498660574833995,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17413483560085297,
      "num_tokens": 9069542.0,
      "step": 4060
    },
    {
      "entropy": 5.803239107131958,
      "epoch": 0.45656202616948394,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.8492,
      "mean_token_accuracy": 0.1667573258280754,
      "num_tokens": 9081219.0,
      "step": 4065
    },
    {
      "entropy": 5.793753051757813,
      "epoch": 0.45712360307744143,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.5548,
      "mean_token_accuracy": 0.18062549084424973,
      "num_tokens": 9092101.0,
      "step": 4070
    },
    {
      "entropy": 5.7102467060089115,
      "epoch": 0.457685179985399,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498647419924648,
      "loss": 5.591,
      "mean_token_accuracy": 0.1737464874982834,
      "num_tokens": 9101900.0,
      "step": 4075
    },
    {
      "entropy": 5.804455852508545,
      "epoch": 0.45824675689335653,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498643020712638,
      "loss": 5.7361,
      "mean_token_accuracy": 0.1701218456029892,
      "num_tokens": 9114369.0,
      "step": 4080
    },
    {
      "entropy": 5.689722061157227,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.5326,
      "mean_token_accuracy": 0.17208467572927474,
      "num_tokens": 9126548.0,
      "step": 4085
    },
    {
      "entropy": 5.730998802185058,
      "epoch": 0.45936991070927163,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1791645899415016,
      "num_tokens": 9138070.0,
      "step": 4090
    },
    {
      "entropy": 5.786447286605835,
      "epoch": 0.4599314876172292,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.6207,
      "mean_token_accuracy": 0.17303537875413894,
      "num_tokens": 9149489.0,
      "step": 4095
    },
    {
      "entropy": 5.727853870391845,
      "epoch": 0.46049306452518673,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.6088,
      "mean_token_accuracy": 0.18384982645511627,
      "num_tokens": 9160363.0,
      "step": 4100
    },
    {
      "entropy": 5.788647603988648,
      "epoch": 0.4610546414331443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.6693,
      "mean_token_accuracy": 0.17254945486783982,
      "num_tokens": 9171619.0,
      "step": 4105
    },
    {
      "entropy": 5.696066236495971,
      "epoch": 0.46161621834110184,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498616475907006,
      "loss": 5.519,
      "mean_token_accuracy": 0.18080994933843614,
      "num_tokens": 9182828.0,
      "step": 4110
    },
    {
      "entropy": 5.776265478134155,
      "epoch": 0.46217779524905933,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1810159683227539,
      "num_tokens": 9195249.0,
      "step": 4115
    },
    {
      "entropy": 5.618781232833863,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.4829,
      "mean_token_accuracy": 0.18047850131988524,
      "num_tokens": 9205770.0,
      "step": 4120
    },
    {
      "entropy": 5.664587354660034,
      "epoch": 0.46330094906497443,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.5069,
      "mean_token_accuracy": 0.17904316931962966,
      "num_tokens": 9216367.0,
      "step": 4125
    },
    {
      "entropy": 5.721905612945557,
      "epoch": 0.463862525972932,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.6033,
      "mean_token_accuracy": 0.16826149672269822,
      "num_tokens": 9228524.0,
      "step": 4130
    },
    {
      "entropy": 5.706638765335083,
      "epoch": 0.46442410288088953,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.5777,
      "mean_token_accuracy": 0.18357411473989488,
      "num_tokens": 9238907.0,
      "step": 4135
    },
    {
      "entropy": 5.745681476593018,
      "epoch": 0.4649856797888471,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498589674780126,
      "loss": 5.6998,
      "mean_token_accuracy": 0.1771864339709282,
      "num_tokens": 9250491.0,
      "step": 4140
    },
    {
      "entropy": 5.7402605533599855,
      "epoch": 0.46554725669680463,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1833001047372818,
      "num_tokens": 9260984.0,
      "step": 4145
    },
    {
      "entropy": 5.760487413406372,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.6579,
      "mean_token_accuracy": 0.1782159239053726,
      "num_tokens": 9272416.0,
      "step": 4150
    },
    {
      "entropy": 5.78125696182251,
      "epoch": 0.46667041051271974,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498576178105807,
      "loss": 5.6436,
      "mean_token_accuracy": 0.17659080624580384,
      "num_tokens": 9284513.0,
      "step": 4155
    },
    {
      "entropy": 5.769126129150391,
      "epoch": 0.4672319874206773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.533,
      "mean_token_accuracy": 0.18448686301708223,
      "num_tokens": 9293422.0,
      "step": 4160
    },
    {
      "entropy": 5.6378617763519285,
      "epoch": 0.4677935643286348,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.5341,
      "mean_token_accuracy": 0.17681366354227065,
      "num_tokens": 9304735.0,
      "step": 4165
    },
    {
      "entropy": 5.693609189987183,
      "epoch": 0.46835514123659233,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.6367,
      "mean_token_accuracy": 0.17564833611249925,
      "num_tokens": 9315659.0,
      "step": 4170
    },
    {
      "entropy": 5.692082929611206,
      "epoch": 0.4689167181445499,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.659,
      "mean_token_accuracy": 0.17728588730096817,
      "num_tokens": 9327106.0,
      "step": 4175
    },
    {
      "entropy": 5.8515644550323485,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.6948,
      "mean_token_accuracy": 0.17592467963695527,
      "num_tokens": 9338762.0,
      "step": 4180
    },
    {
      "entropy": 5.749437999725342,
      "epoch": 0.470039871960465,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498548992554749,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17839116752147674,
      "num_tokens": 9349495.0,
      "step": 4185
    },
    {
      "entropy": 5.775325155258178,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.6523,
      "mean_token_accuracy": 0.1811431020498276,
      "num_tokens": 9360830.0,
      "step": 4190
    },
    {
      "entropy": 5.68581919670105,
      "epoch": 0.4711630257763801,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.5306,
      "mean_token_accuracy": 0.17705092281103135,
      "num_tokens": 9371668.0,
      "step": 4195
    },
    {
      "entropy": 5.737933111190796,
      "epoch": 0.47172460268433763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.6715,
      "mean_token_accuracy": 0.1726916804909706,
      "num_tokens": 9383437.0,
      "step": 4200
    },
    {
      "entropy": 5.683913516998291,
      "epoch": 0.4722861795922952,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.4891,
      "mean_token_accuracy": 0.18346946835517883,
      "num_tokens": 9394597.0,
      "step": 4205
    },
    {
      "entropy": 5.568347787857055,
      "epoch": 0.47284775650025274,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18162453025579453,
      "num_tokens": 9404322.0,
      "step": 4210
    },
    {
      "entropy": 5.797742080688477,
      "epoch": 0.47340933340821023,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.8002,
      "mean_token_accuracy": 0.16466799527406692,
      "num_tokens": 9416740.0,
      "step": 4215
    },
    {
      "entropy": 5.817827272415161,
      "epoch": 0.4739709103161678,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.6034,
      "mean_token_accuracy": 0.17978677302598953,
      "num_tokens": 9428617.0,
      "step": 4220
    },
    {
      "entropy": 5.642584180831909,
      "epoch": 0.47453248722412533,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.5538,
      "mean_token_accuracy": 0.18305636644363404,
      "num_tokens": 9439686.0,
      "step": 4225
    },
    {
      "entropy": 5.689055252075195,
      "epoch": 0.4750940641320829,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.5428,
      "mean_token_accuracy": 0.18234148174524306,
      "num_tokens": 9450378.0,
      "step": 4230
    },
    {
      "entropy": 5.745994567871094,
      "epoch": 0.47565564104004043,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.5745,
      "mean_token_accuracy": 0.17828824669122695,
      "num_tokens": 9461659.0,
      "step": 4235
    },
    {
      "entropy": 5.719333648681641,
      "epoch": 0.476217217947998,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.6143,
      "mean_token_accuracy": 0.18095430135726928,
      "num_tokens": 9472528.0,
      "step": 4240
    },
    {
      "entropy": 5.765508604049683,
      "epoch": 0.47677879485595553,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.5458,
      "mean_token_accuracy": 0.18705329298973083,
      "num_tokens": 9483388.0,
      "step": 4245
    },
    {
      "entropy": 5.550165224075317,
      "epoch": 0.4773403717639131,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498489211510521,
      "loss": 5.4057,
      "mean_token_accuracy": 0.19107298851013182,
      "num_tokens": 9495178.0,
      "step": 4250
    },
    {
      "entropy": 5.650314092636108,
      "epoch": 0.47790194867187064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.6387,
      "mean_token_accuracy": 0.17218846678733826,
      "num_tokens": 9506226.0,
      "step": 4255
    },
    {
      "entropy": 5.793583965301513,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5287,
      "mean_token_accuracy": 0.1777708724141121,
      "num_tokens": 9517671.0,
      "step": 4260
    },
    {
      "entropy": 5.645057678222656,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.6144,
      "mean_token_accuracy": 0.18139024674892426,
      "num_tokens": 9527637.0,
      "step": 4265
    },
    {
      "entropy": 5.77157940864563,
      "epoch": 0.47958667939574323,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498470575384638,
      "loss": 5.7075,
      "mean_token_accuracy": 0.1710361510515213,
      "num_tokens": 9539686.0,
      "step": 4270
    },
    {
      "entropy": 5.7558211326599125,
      "epoch": 0.4801482563037008,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.6532,
      "mean_token_accuracy": 0.18178366869688034,
      "num_tokens": 9550775.0,
      "step": 4275
    },
    {
      "entropy": 5.755832004547119,
      "epoch": 0.48070983321165833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.5609,
      "mean_token_accuracy": 0.1788163363933563,
      "num_tokens": 9561760.0,
      "step": 4280
    },
    {
      "entropy": 5.6204259395599365,
      "epoch": 0.4812714101196159,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.6082,
      "mean_token_accuracy": 0.17339764386415482,
      "num_tokens": 9573899.0,
      "step": 4285
    },
    {
      "entropy": 5.674421691894532,
      "epoch": 0.48183298702757343,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498451825412523,
      "loss": 5.5566,
      "mean_token_accuracy": 0.17854472994804382,
      "num_tokens": 9584336.0,
      "step": 4290
    },
    {
      "entropy": 5.838225173950195,
      "epoch": 0.482394563935531,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.7555,
      "mean_token_accuracy": 0.16703201532363893,
      "num_tokens": 9596614.0,
      "step": 4295
    },
    {
      "entropy": 5.769792699813843,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.5886,
      "mean_token_accuracy": 0.18015121370553971,
      "num_tokens": 9606796.0,
      "step": 4300
    },
    {
      "entropy": 5.730530166625977,
      "epoch": 0.4835177177514461,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.5764,
      "mean_token_accuracy": 0.17851542681455612,
      "num_tokens": 9617705.0,
      "step": 4305
    },
    {
      "entropy": 5.609689378738404,
      "epoch": 0.4840792946594036,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.5904,
      "mean_token_accuracy": 0.17903478741645812,
      "num_tokens": 9628098.0,
      "step": 4310
    },
    {
      "entropy": 5.732795381546021,
      "epoch": 0.48464087156736113,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.5365,
      "mean_token_accuracy": 0.18742966800928115,
      "num_tokens": 9637697.0,
      "step": 4315
    },
    {
      "entropy": 5.693505477905274,
      "epoch": 0.4852024484753187,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.4269,
      "mean_token_accuracy": 0.186981138586998,
      "num_tokens": 9648221.0,
      "step": 4320
    },
    {
      "entropy": 5.776789045333862,
      "epoch": 0.48576402538327623,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6255,
      "mean_token_accuracy": 0.17044762372970582,
      "num_tokens": 9660182.0,
      "step": 4325
    },
    {
      "entropy": 5.689573621749878,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.6341,
      "mean_token_accuracy": 0.17342920005321502,
      "num_tokens": 9672121.0,
      "step": 4330
    },
    {
      "entropy": 5.7519200325012205,
      "epoch": 0.48688717919919133,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.5277,
      "mean_token_accuracy": 0.17892132997512816,
      "num_tokens": 9682651.0,
      "step": 4335
    },
    {
      "entropy": 5.752449035644531,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.6762,
      "mean_token_accuracy": 0.17406619489192962,
      "num_tokens": 9693989.0,
      "step": 4340
    },
    {
      "entropy": 5.632310295104981,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1833195522427559,
      "num_tokens": 9703802.0,
      "step": 4345
    },
    {
      "entropy": 5.677293157577514,
      "epoch": 0.488571909923064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498394892514596,
      "loss": 5.485,
      "mean_token_accuracy": 0.19319503307342528,
      "num_tokens": 9714287.0,
      "step": 4350
    },
    {
      "entropy": 5.672032451629638,
      "epoch": 0.48913348683102154,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.5148,
      "mean_token_accuracy": 0.182816281914711,
      "num_tokens": 9724819.0,
      "step": 4355
    },
    {
      "entropy": 5.697476530075074,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.6603,
      "mean_token_accuracy": 0.17222750633955003,
      "num_tokens": 9736326.0,
      "step": 4360
    },
    {
      "entropy": 5.833402252197265,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.6767,
      "mean_token_accuracy": 0.17708243131637574,
      "num_tokens": 9747883.0,
      "step": 4365
    },
    {
      "entropy": 5.663409852981568,
      "epoch": 0.49081821755489413,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.5363,
      "mean_token_accuracy": 0.1819302424788475,
      "num_tokens": 9758191.0,
      "step": 4370
    },
    {
      "entropy": 5.772275733947754,
      "epoch": 0.4913797944628517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.6359,
      "mean_token_accuracy": 0.17441492080688475,
      "num_tokens": 9768852.0,
      "step": 4375
    },
    {
      "entropy": 5.769843196868896,
      "epoch": 0.49194137137080923,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.5795,
      "mean_token_accuracy": 0.17516547590494155,
      "num_tokens": 9779033.0,
      "step": 4380
    },
    {
      "entropy": 5.640242862701416,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.5043,
      "mean_token_accuracy": 0.18283630460500716,
      "num_tokens": 9790777.0,
      "step": 4385
    },
    {
      "entropy": 5.797633934020996,
      "epoch": 0.49306452518672433,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.7708,
      "mean_token_accuracy": 0.16839805692434312,
      "num_tokens": 9802305.0,
      "step": 4390
    },
    {
      "entropy": 5.754205942153931,
      "epoch": 0.4936261020946819,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.5553,
      "mean_token_accuracy": 0.175040665268898,
      "num_tokens": 9813590.0,
      "step": 4395
    },
    {
      "entropy": 5.7130505561828615,
      "epoch": 0.49418767900263944,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.6465,
      "mean_token_accuracy": 0.1715954750776291,
      "num_tokens": 9824080.0,
      "step": 4400
    },
    {
      "entropy": 5.7139181137084964,
      "epoch": 0.49474925591059693,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.5117,
      "mean_token_accuracy": 0.176900452375412,
      "num_tokens": 9835441.0,
      "step": 4405
    },
    {
      "entropy": 5.658385181427002,
      "epoch": 0.4953108328185545,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.5173,
      "mean_token_accuracy": 0.18041314631700517,
      "num_tokens": 9846750.0,
      "step": 4410
    },
    {
      "entropy": 5.771450090408325,
      "epoch": 0.49587240972651203,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.6682,
      "mean_token_accuracy": 0.17632983922958373,
      "num_tokens": 9857987.0,
      "step": 4415
    },
    {
      "entropy": 5.725841093063354,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.6206,
      "mean_token_accuracy": 0.17901892960071564,
      "num_tokens": 9869736.0,
      "step": 4420
    },
    {
      "entropy": 5.769300937652588,
      "epoch": 0.49699556354242713,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498322286045317,
      "loss": 5.5966,
      "mean_token_accuracy": 0.1730643853545189,
      "num_tokens": 9881610.0,
      "step": 4425
    },
    {
      "entropy": 5.704537391662598,
      "epoch": 0.4975571404503847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.5909,
      "mean_token_accuracy": 0.1746903657913208,
      "num_tokens": 9893203.0,
      "step": 4430
    },
    {
      "entropy": 5.687581205368042,
      "epoch": 0.49811871735834223,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.5053,
      "mean_token_accuracy": 0.17380916178226472,
      "num_tokens": 9904392.0,
      "step": 4435
    },
    {
      "entropy": 5.543652153015136,
      "epoch": 0.4986802942662998,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.5012,
      "mean_token_accuracy": 0.1814364716410637,
      "num_tokens": 9916184.0,
      "step": 4440
    },
    {
      "entropy": 5.751931428909302,
      "epoch": 0.49924187117425733,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.6216,
      "mean_token_accuracy": 0.18379164934158326,
      "num_tokens": 9927516.0,
      "step": 4445
    },
    {
      "entropy": 5.756189823150635,
      "epoch": 0.4998034480822149,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.6281,
      "mean_token_accuracy": 0.1757001519203186,
      "num_tokens": 9939182.0,
      "step": 4450
    },
    {
      "entropy": 5.734696912765503,
      "epoch": 0.5003650249901724,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498292795456261,
      "loss": 5.6503,
      "mean_token_accuracy": 0.17457239031791688,
      "num_tokens": 9950382.0,
      "step": 4455
    },
    {
      "entropy": 5.767827701568604,
      "epoch": 0.50092660189813,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.6374,
      "mean_token_accuracy": 0.1816966265439987,
      "num_tokens": 9961053.0,
      "step": 4460
    },
    {
      "entropy": 5.674236392974853,
      "epoch": 0.5014881788060875,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.5325,
      "mean_token_accuracy": 0.18293653875589372,
      "num_tokens": 9971313.0,
      "step": 4465
    },
    {
      "entropy": 5.694076919555664,
      "epoch": 0.5020497557140451,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.5678,
      "mean_token_accuracy": 0.1871047332882881,
      "num_tokens": 9981257.0,
      "step": 4470
    },
    {
      "entropy": 5.682608366012573,
      "epoch": 0.5026113326220026,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.4908,
      "mean_token_accuracy": 0.17670724242925645,
      "num_tokens": 9991860.0,
      "step": 4475
    },
    {
      "entropy": 5.669681549072266,
      "epoch": 0.5031729095299601,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.5845,
      "mean_token_accuracy": 0.1813073053956032,
      "num_tokens": 10002320.0,
      "step": 4480
    },
    {
      "entropy": 5.800026273727417,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1721622735261917,
      "num_tokens": 10013187.0,
      "step": 4485
    },
    {
      "entropy": 5.6245283603668215,
      "epoch": 0.5042960633458752,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1861062839627266,
      "num_tokens": 10024268.0,
      "step": 4490
    },
    {
      "entropy": 5.754734325408935,
      "epoch": 0.5048576402538327,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1772270083427429,
      "num_tokens": 10034738.0,
      "step": 4495
    },
    {
      "entropy": 5.817403793334961,
      "epoch": 0.5054192171617903,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.6225,
      "mean_token_accuracy": 0.17836388796567917,
      "num_tokens": 10045666.0,
      "step": 4500
    },
    {
      "entropy": 5.665848922729492,
      "epoch": 0.5059807940697478,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.4941,
      "mean_token_accuracy": 0.1802812859416008,
      "num_tokens": 10057084.0,
      "step": 4505
    },
    {
      "entropy": 5.653784704208374,
      "epoch": 0.5065423709777054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.4725,
      "mean_token_accuracy": 0.18146999478340148,
      "num_tokens": 10066754.0,
      "step": 4510
    },
    {
      "entropy": 5.593233346939087,
      "epoch": 0.5071039478856629,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1864606037735939,
      "num_tokens": 10078139.0,
      "step": 4515
    },
    {
      "entropy": 5.698436069488525,
      "epoch": 0.5076655247936205,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.5556,
      "mean_token_accuracy": 0.17708238065242768,
      "num_tokens": 10089652.0,
      "step": 4520
    },
    {
      "entropy": 5.551695680618286,
      "epoch": 0.508227101701578,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18821074217557907,
      "num_tokens": 10100642.0,
      "step": 4525
    },
    {
      "entropy": 5.725232076644898,
      "epoch": 0.5087886786095356,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.5971,
      "mean_token_accuracy": 0.1802185833454132,
      "num_tokens": 10112036.0,
      "step": 4530
    },
    {
      "entropy": 5.652781963348389,
      "epoch": 0.5093502555174931,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.466,
      "mean_token_accuracy": 0.18149771243333818,
      "num_tokens": 10121653.0,
      "step": 4535
    },
    {
      "entropy": 5.6757909774780275,
      "epoch": 0.5099118324254507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17870260030031204,
      "num_tokens": 10132271.0,
      "step": 4540
    },
    {
      "entropy": 5.588453149795532,
      "epoch": 0.5104734093334082,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.4761,
      "mean_token_accuracy": 0.17977570593357087,
      "num_tokens": 10143411.0,
      "step": 4545
    },
    {
      "entropy": 5.772822046279908,
      "epoch": 0.5110349862413658,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.5778,
      "mean_token_accuracy": 0.1801849126815796,
      "num_tokens": 10154338.0,
      "step": 4550
    },
    {
      "entropy": 5.696248483657837,
      "epoch": 0.5115965631493233,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.6099,
      "mean_token_accuracy": 0.1806991681456566,
      "num_tokens": 10164810.0,
      "step": 4555
    },
    {
      "entropy": 5.672287130355835,
      "epoch": 0.5121581400572809,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.5168,
      "mean_token_accuracy": 0.18060070723295213,
      "num_tokens": 10176323.0,
      "step": 4560
    },
    {
      "entropy": 5.689506912231446,
      "epoch": 0.5127197169652384,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.5549,
      "mean_token_accuracy": 0.1825497940182686,
      "num_tokens": 10187110.0,
      "step": 4565
    },
    {
      "entropy": 5.650188207626343,
      "epoch": 0.513281293873196,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.52,
      "mean_token_accuracy": 0.17509306371212005,
      "num_tokens": 10199099.0,
      "step": 4570
    },
    {
      "entropy": 5.722729301452636,
      "epoch": 0.5138428707811534,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.4956,
      "mean_token_accuracy": 0.183281572163105,
      "num_tokens": 10209726.0,
      "step": 4575
    },
    {
      "entropy": 5.730534839630127,
      "epoch": 0.514404447689111,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049816716337126,
      "loss": 5.6041,
      "mean_token_accuracy": 0.1800455316901207,
      "num_tokens": 10220858.0,
      "step": 4580
    },
    {
      "entropy": 5.560760450363159,
      "epoch": 0.5149660245970685,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.4425,
      "mean_token_accuracy": 0.18481195718050003,
      "num_tokens": 10232250.0,
      "step": 4585
    },
    {
      "entropy": 5.795879077911377,
      "epoch": 0.5155276015050261,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.6334,
      "mean_token_accuracy": 0.17141353636980056,
      "num_tokens": 10244584.0,
      "step": 4590
    },
    {
      "entropy": 5.726343107223511,
      "epoch": 0.5160891784129836,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.5027,
      "mean_token_accuracy": 0.1818414956331253,
      "num_tokens": 10255926.0,
      "step": 4595
    },
    {
      "entropy": 5.677523422241211,
      "epoch": 0.5166507553209412,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498146650039855,
      "loss": 5.5909,
      "mean_token_accuracy": 0.17883584201335906,
      "num_tokens": 10266691.0,
      "step": 4600
    },
    {
      "entropy": 5.737709665298462,
      "epoch": 0.5172123322288987,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.6506,
      "mean_token_accuracy": 0.17722098976373674,
      "num_tokens": 10277744.0,
      "step": 4605
    },
    {
      "entropy": 5.649214124679565,
      "epoch": 0.5177739091368563,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.4917,
      "mean_token_accuracy": 0.18419942855834961,
      "num_tokens": 10288792.0,
      "step": 4610
    },
    {
      "entropy": 5.734871959686279,
      "epoch": 0.5183354860448138,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.5492,
      "mean_token_accuracy": 0.18144079446792602,
      "num_tokens": 10299569.0,
      "step": 4615
    },
    {
      "entropy": 5.684595918655395,
      "epoch": 0.5188970629527714,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.5141,
      "mean_token_accuracy": 0.18706338256597518,
      "num_tokens": 10309742.0,
      "step": 4620
    },
    {
      "entropy": 5.733961725234986,
      "epoch": 0.5194586398607289,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.5659,
      "mean_token_accuracy": 0.17764922529458999,
      "num_tokens": 10320603.0,
      "step": 4625
    },
    {
      "entropy": 5.613412618637085,
      "epoch": 0.5200202167686865,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1833082213997841,
      "num_tokens": 10331498.0,
      "step": 4630
    },
    {
      "entropy": 5.683115816116333,
      "epoch": 0.520581793676644,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.5624,
      "mean_token_accuracy": 0.18289795219898225,
      "num_tokens": 10342002.0,
      "step": 4635
    },
    {
      "entropy": 5.66677508354187,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18084699511528016,
      "num_tokens": 10351797.0,
      "step": 4640
    },
    {
      "entropy": 5.647039461135864,
      "epoch": 0.5217049474925591,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.4525,
      "mean_token_accuracy": 0.18113619983196258,
      "num_tokens": 10362538.0,
      "step": 4645
    },
    {
      "entropy": 5.682296657562256,
      "epoch": 0.5222665244005167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18371141850948333,
      "num_tokens": 10373365.0,
      "step": 4650
    },
    {
      "entropy": 5.580294609069824,
      "epoch": 0.5228281013084742,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1784225136041641,
      "num_tokens": 10385087.0,
      "step": 4655
    },
    {
      "entropy": 5.6413490772247314,
      "epoch": 0.5233896782164318,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.5974,
      "mean_token_accuracy": 0.178703410923481,
      "num_tokens": 10396014.0,
      "step": 4660
    },
    {
      "entropy": 5.642305374145508,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.5234,
      "mean_token_accuracy": 0.1791493073105812,
      "num_tokens": 10407844.0,
      "step": 4665
    },
    {
      "entropy": 5.5928679466247555,
      "epoch": 0.5245128320323468,
      "grad_norm": 0.875,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.5867,
      "mean_token_accuracy": 0.17278059273958207,
      "num_tokens": 10419892.0,
      "step": 4670
    },
    {
      "entropy": 5.711860179901123,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.5343,
      "mean_token_accuracy": 0.175407874584198,
      "num_tokens": 10430686.0,
      "step": 4675
    },
    {
      "entropy": 5.680815744400024,
      "epoch": 0.5256359858482619,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.5715,
      "mean_token_accuracy": 0.17376552820205687,
      "num_tokens": 10443785.0,
      "step": 4680
    },
    {
      "entropy": 5.60487928390503,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.4533,
      "mean_token_accuracy": 0.18937646597623825,
      "num_tokens": 10454889.0,
      "step": 4685
    },
    {
      "entropy": 5.622950267791748,
      "epoch": 0.526759139664177,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.6252,
      "mean_token_accuracy": 0.17705041766166688,
      "num_tokens": 10466525.0,
      "step": 4690
    },
    {
      "entropy": 5.800705623626709,
      "epoch": 0.5273207165721345,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.573,
      "mean_token_accuracy": 0.18033843636512756,
      "num_tokens": 10476329.0,
      "step": 4695
    },
    {
      "entropy": 5.76061897277832,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.6276,
      "mean_token_accuracy": 0.17819187939167022,
      "num_tokens": 10487178.0,
      "step": 4700
    },
    {
      "entropy": 5.670125961303711,
      "epoch": 0.5284438703880496,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.5775,
      "mean_token_accuracy": 0.16941394060850143,
      "num_tokens": 10499293.0,
      "step": 4705
    },
    {
      "entropy": 5.805813312530518,
      "epoch": 0.5290054472960072,
      "grad_norm": 0.875,
      "learning_rate": 0.00049803179494862,
      "loss": 5.7464,
      "mean_token_accuracy": 0.1662947230041027,
      "num_tokens": 10511762.0,
      "step": 4710
    },
    {
      "entropy": 5.723719120025635,
      "epoch": 0.5295670242039647,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.4804,
      "mean_token_accuracy": 0.18739404976367952,
      "num_tokens": 10521725.0,
      "step": 4715
    },
    {
      "entropy": 5.624107551574707,
      "epoch": 0.5301286011119223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.4314,
      "mean_token_accuracy": 0.18282126039266586,
      "num_tokens": 10531764.0,
      "step": 4720
    },
    {
      "entropy": 5.694038391113281,
      "epoch": 0.5306901780198798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.5047,
      "mean_token_accuracy": 0.1832448124885559,
      "num_tokens": 10542127.0,
      "step": 4725
    },
    {
      "entropy": 5.727702331542969,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.875,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.5975,
      "mean_token_accuracy": 0.17381149977445604,
      "num_tokens": 10552580.0,
      "step": 4730
    },
    {
      "entropy": 5.709339141845703,
      "epoch": 0.5318133318357949,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.5354,
      "mean_token_accuracy": 0.1810144692659378,
      "num_tokens": 10562961.0,
      "step": 4735
    },
    {
      "entropy": 5.789928436279297,
      "epoch": 0.5323749087437525,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.5369,
      "mean_token_accuracy": 0.18183746337890624,
      "num_tokens": 10573210.0,
      "step": 4740
    },
    {
      "entropy": 5.543464326858521,
      "epoch": 0.53293648565171,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.4994,
      "mean_token_accuracy": 0.17975255250930786,
      "num_tokens": 10583878.0,
      "step": 4745
    },
    {
      "entropy": 5.69704794883728,
      "epoch": 0.5334980625596676,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.577,
      "mean_token_accuracy": 0.17803760021924972,
      "num_tokens": 10593420.0,
      "step": 4750
    },
    {
      "entropy": 5.768489360809326,
      "epoch": 0.5340596394676251,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497983818025976,
      "loss": 5.5838,
      "mean_token_accuracy": 0.17519904524087906,
      "num_tokens": 10604213.0,
      "step": 4755
    },
    {
      "entropy": 5.7158088207244875,
      "epoch": 0.5346212163755827,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.5655,
      "mean_token_accuracy": 0.17727186232805253,
      "num_tokens": 10615502.0,
      "step": 4760
    },
    {
      "entropy": 5.697181129455567,
      "epoch": 0.5351827932835401,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.6211,
      "mean_token_accuracy": 0.16839191168546677,
      "num_tokens": 10627589.0,
      "step": 4765
    },
    {
      "entropy": 5.622159051895141,
      "epoch": 0.5357443701914977,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.4478,
      "mean_token_accuracy": 0.18488689213991166,
      "num_tokens": 10638365.0,
      "step": 4770
    },
    {
      "entropy": 5.674831867218018,
      "epoch": 0.5363059470994552,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497962310344703,
      "loss": 5.5508,
      "mean_token_accuracy": 0.18342875242233275,
      "num_tokens": 10650664.0,
      "step": 4775
    },
    {
      "entropy": 5.705963945388794,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049795691567594,
      "loss": 5.6084,
      "mean_token_accuracy": 0.18426345139741898,
      "num_tokens": 10662418.0,
      "step": 4780
    },
    {
      "entropy": 5.66761965751648,
      "epoch": 0.5374291009153703,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.4826,
      "mean_token_accuracy": 0.1847769945859909,
      "num_tokens": 10672793.0,
      "step": 4785
    },
    {
      "entropy": 5.638708639144897,
      "epoch": 0.5379906778233279,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1878056287765503,
      "num_tokens": 10684082.0,
      "step": 4790
    },
    {
      "entropy": 5.561481666564942,
      "epoch": 0.5385522547312854,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.4683,
      "mean_token_accuracy": 0.18410870730876921,
      "num_tokens": 10694490.0,
      "step": 4795
    },
    {
      "entropy": 5.673333835601807,
      "epoch": 0.539113831639243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.4809,
      "mean_token_accuracy": 0.1789180412888527,
      "num_tokens": 10704731.0,
      "step": 4800
    },
    {
      "entropy": 5.652533054351807,
      "epoch": 0.5396754085472005,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.5039,
      "mean_token_accuracy": 0.1836671456694603,
      "num_tokens": 10715735.0,
      "step": 4805
    },
    {
      "entropy": 5.59511284828186,
      "epoch": 0.5402369854551581,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.4699,
      "mean_token_accuracy": 0.18215380162000655,
      "num_tokens": 10726951.0,
      "step": 4810
    },
    {
      "entropy": 5.646748495101929,
      "epoch": 0.5407985623631156,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.5296,
      "mean_token_accuracy": 0.17723255157470702,
      "num_tokens": 10737609.0,
      "step": 4815
    },
    {
      "entropy": 5.635796070098877,
      "epoch": 0.5413601392710732,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.4288,
      "mean_token_accuracy": 0.1895879924297333,
      "num_tokens": 10748957.0,
      "step": 4820
    },
    {
      "entropy": 5.712793827056885,
      "epoch": 0.5419217161790307,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.6145,
      "mean_token_accuracy": 0.1748154416680336,
      "num_tokens": 10759398.0,
      "step": 4825
    },
    {
      "entropy": 5.722065210342407,
      "epoch": 0.5424832930869883,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.5879,
      "mean_token_accuracy": 0.17405359745025634,
      "num_tokens": 10770436.0,
      "step": 4830
    },
    {
      "entropy": 5.600390243530273,
      "epoch": 0.5430448699949458,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.419,
      "mean_token_accuracy": 0.19431555569171904,
      "num_tokens": 10781160.0,
      "step": 4835
    },
    {
      "entropy": 5.723292827606201,
      "epoch": 0.5436064469029034,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.5481,
      "mean_token_accuracy": 0.1877936154603958,
      "num_tokens": 10793170.0,
      "step": 4840
    },
    {
      "entropy": 5.635513544082642,
      "epoch": 0.5441680238108609,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.4943,
      "mean_token_accuracy": 0.17929139882326126,
      "num_tokens": 10803515.0,
      "step": 4845
    },
    {
      "entropy": 5.6866954326629635,
      "epoch": 0.5447296007188185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.5275,
      "mean_token_accuracy": 0.17880094349384307,
      "num_tokens": 10814508.0,
      "step": 4850
    },
    {
      "entropy": 5.6985650062561035,
      "epoch": 0.545291177626776,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.627,
      "mean_token_accuracy": 0.17515473961830139,
      "num_tokens": 10825924.0,
      "step": 4855
    },
    {
      "entropy": 5.6857147216796875,
      "epoch": 0.5458527545347336,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.6024,
      "mean_token_accuracy": 0.17552653700113297,
      "num_tokens": 10837854.0,
      "step": 4860
    },
    {
      "entropy": 5.668138742446899,
      "epoch": 0.546414331442691,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.4941,
      "mean_token_accuracy": 0.18083750903606416,
      "num_tokens": 10848279.0,
      "step": 4865
    },
    {
      "entropy": 5.615544223785401,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.5109,
      "mean_token_accuracy": 0.17481240183115004,
      "num_tokens": 10859615.0,
      "step": 4870
    },
    {
      "entropy": 5.728345203399658,
      "epoch": 0.5475374852586061,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.4786,
      "mean_token_accuracy": 0.17889843583106996,
      "num_tokens": 10871388.0,
      "step": 4875
    },
    {
      "entropy": 5.633346271514893,
      "epoch": 0.5480990621665637,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.5364,
      "mean_token_accuracy": 0.17858044654130936,
      "num_tokens": 10882411.0,
      "step": 4880
    },
    {
      "entropy": 5.674109697341919,
      "epoch": 0.5486606390745212,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.662,
      "mean_token_accuracy": 0.1772112265229225,
      "num_tokens": 10894116.0,
      "step": 4885
    },
    {
      "entropy": 5.607702589035034,
      "epoch": 0.5492222159824788,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.4199,
      "mean_token_accuracy": 0.1881287917494774,
      "num_tokens": 10904514.0,
      "step": 4890
    },
    {
      "entropy": 5.59633469581604,
      "epoch": 0.5497837928904363,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.4677,
      "mean_token_accuracy": 0.18135118037462233,
      "num_tokens": 10915602.0,
      "step": 4895
    },
    {
      "entropy": 5.609324026107788,
      "epoch": 0.5503453697983939,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.4651,
      "mean_token_accuracy": 0.18813514858484268,
      "num_tokens": 10926172.0,
      "step": 4900
    },
    {
      "entropy": 5.73305835723877,
      "epoch": 0.5509069467063514,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.5719,
      "mean_token_accuracy": 0.17572105377912522,
      "num_tokens": 10938435.0,
      "step": 4905
    },
    {
      "entropy": 5.627225160598755,
      "epoch": 0.551468523614309,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1822732612490654,
      "num_tokens": 10949118.0,
      "step": 4910
    },
    {
      "entropy": 5.631764316558838,
      "epoch": 0.5520301005222665,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.5028,
      "mean_token_accuracy": 0.17702874690294265,
      "num_tokens": 10960120.0,
      "step": 4915
    },
    {
      "entropy": 5.690504455566407,
      "epoch": 0.5525916774302241,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497802983376339,
      "loss": 5.4937,
      "mean_token_accuracy": 0.182991823554039,
      "num_tokens": 10971210.0,
      "step": 4920
    },
    {
      "entropy": 5.6307860851287845,
      "epoch": 0.5531532543381816,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.4882,
      "mean_token_accuracy": 0.19323195070028304,
      "num_tokens": 10983446.0,
      "step": 4925
    },
    {
      "entropy": 5.675218486785889,
      "epoch": 0.5537148312461392,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17837650030851365,
      "num_tokens": 10994776.0,
      "step": 4930
    },
    {
      "entropy": 5.714323949813843,
      "epoch": 0.5542764081540967,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497786160684314,
      "loss": 5.5097,
      "mean_token_accuracy": 0.18002611100673677,
      "num_tokens": 11005280.0,
      "step": 4935
    },
    {
      "entropy": 5.581811714172363,
      "epoch": 0.5548379850620543,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.4967,
      "mean_token_accuracy": 0.17861787229776382,
      "num_tokens": 11017509.0,
      "step": 4940
    },
    {
      "entropy": 5.683772468566895,
      "epoch": 0.5553995619700118,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5224,
      "mean_token_accuracy": 0.18149121701717377,
      "num_tokens": 11029087.0,
      "step": 4945
    },
    {
      "entropy": 5.702160024642945,
      "epoch": 0.5559611388779694,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.5786,
      "mean_token_accuracy": 0.1756944939494133,
      "num_tokens": 11041144.0,
      "step": 4950
    },
    {
      "entropy": 5.779351806640625,
      "epoch": 0.5565227157859269,
      "grad_norm": 0.875,
      "learning_rate": 0.00049776363111875,
      "loss": 5.6344,
      "mean_token_accuracy": 0.1735006868839264,
      "num_tokens": 11054050.0,
      "step": 4955
    },
    {
      "entropy": 5.614022874832154,
      "epoch": 0.5570842926938844,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.4565,
      "mean_token_accuracy": 0.1862945720553398,
      "num_tokens": 11064257.0,
      "step": 4960
    },
    {
      "entropy": 5.688822221755982,
      "epoch": 0.5576458696018419,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.5732,
      "mean_token_accuracy": 0.180736143887043,
      "num_tokens": 11075568.0,
      "step": 4965
    },
    {
      "entropy": 5.666280508041382,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1826269268989563,
      "num_tokens": 11087078.0,
      "step": 4970
    },
    {
      "entropy": 5.811055707931518,
      "epoch": 0.558769023417757,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.6705,
      "mean_token_accuracy": 0.1719473510980606,
      "num_tokens": 11100655.0,
      "step": 4975
    },
    {
      "entropy": 5.71540937423706,
      "epoch": 0.5593306003257146,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1767576202750206,
      "num_tokens": 11110675.0,
      "step": 4980
    },
    {
      "entropy": 5.633408498764038,
      "epoch": 0.5598921772336721,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.5574,
      "mean_token_accuracy": 0.17852846831083297,
      "num_tokens": 11122337.0,
      "step": 4985
    },
    {
      "entropy": 5.752486705780029,
      "epoch": 0.5604537541416297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497723931310065,
      "loss": 5.4672,
      "mean_token_accuracy": 0.183281609416008,
      "num_tokens": 11132672.0,
      "step": 4990
    },
    {
      "entropy": 5.6341267108917235,
      "epoch": 0.5610153310495872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.5107,
      "mean_token_accuracy": 0.18576120287179948,
      "num_tokens": 11143740.0,
      "step": 4995
    },
    {
      "entropy": 5.60403003692627,
      "epoch": 0.5615769079575448,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.4525,
      "mean_token_accuracy": 0.18734156638383864,
      "num_tokens": 11153722.0,
      "step": 5000
    },
    {
      "entropy": 5.740697002410888,
      "epoch": 0.5621384848655023,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1734340563416481,
      "num_tokens": 11165042.0,
      "step": 5005
    },
    {
      "entropy": 5.654933977127075,
      "epoch": 0.5627000617734599,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1847301036119461,
      "num_tokens": 11176834.0,
      "step": 5010
    },
    {
      "entropy": 5.6152064323425295,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.5248,
      "mean_token_accuracy": 0.18113062381744385,
      "num_tokens": 11188442.0,
      "step": 5015
    },
    {
      "entropy": 5.599682092666626,
      "epoch": 0.563823215589375,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497689626327256,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1891382724046707,
      "num_tokens": 11197855.0,
      "step": 5020
    },
    {
      "entropy": 5.64382529258728,
      "epoch": 0.5643847924973325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497683884012223,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17903661131858825,
      "num_tokens": 11209300.0,
      "step": 5025
    },
    {
      "entropy": 5.649948167800903,
      "epoch": 0.5649463694052901,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.5437,
      "mean_token_accuracy": 0.18342785090208052,
      "num_tokens": 11222175.0,
      "step": 5030
    },
    {
      "entropy": 5.6943275928497314,
      "epoch": 0.5655079463132476,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.5536,
      "mean_token_accuracy": 0.17735838741064072,
      "num_tokens": 11233277.0,
      "step": 5035
    },
    {
      "entropy": 5.689804315567017,
      "epoch": 0.5660695232212052,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.5747,
      "mean_token_accuracy": 0.17973842918872834,
      "num_tokens": 11245241.0,
      "step": 5040
    },
    {
      "entropy": 5.683139896392822,
      "epoch": 0.5666311001291627,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1842031255364418,
      "num_tokens": 11257359.0,
      "step": 5045
    },
    {
      "entropy": 5.664784336090088,
      "epoch": 0.5671926770371203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.5261,
      "mean_token_accuracy": 0.18461871296167373,
      "num_tokens": 11270651.0,
      "step": 5050
    },
    {
      "entropy": 5.650564384460449,
      "epoch": 0.5677542539450777,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497649281231274,
      "loss": 5.4942,
      "mean_token_accuracy": 0.18811657130718232,
      "num_tokens": 11280164.0,
      "step": 5055
    },
    {
      "entropy": 5.616600275039673,
      "epoch": 0.5683158308530353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.4381,
      "mean_token_accuracy": 0.18068911731243134,
      "num_tokens": 11290156.0,
      "step": 5060
    },
    {
      "entropy": 5.654250001907348,
      "epoch": 0.5688774077609928,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.5027,
      "mean_token_accuracy": 0.18379531055688858,
      "num_tokens": 11301929.0,
      "step": 5065
    },
    {
      "entropy": 5.591801357269287,
      "epoch": 0.5694389846689504,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18950450867414476,
      "num_tokens": 11312805.0,
      "step": 5070
    },
    {
      "entropy": 5.494604301452637,
      "epoch": 0.5700005615769079,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19156534671783448,
      "num_tokens": 11322316.0,
      "step": 5075
    },
    {
      "entropy": 5.611537075042724,
      "epoch": 0.5705621384848655,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497620250624724,
      "loss": 5.5006,
      "mean_token_accuracy": 0.18360943347215652,
      "num_tokens": 11332994.0,
      "step": 5080
    },
    {
      "entropy": 5.64788966178894,
      "epoch": 0.571123715392823,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497614423237728,
      "loss": 5.4923,
      "mean_token_accuracy": 0.17883387207984924,
      "num_tokens": 11344355.0,
      "step": 5085
    },
    {
      "entropy": 5.624112510681153,
      "epoch": 0.5716852923007806,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.5366,
      "mean_token_accuracy": 0.17753890603780748,
      "num_tokens": 11356275.0,
      "step": 5090
    },
    {
      "entropy": 5.617595624923706,
      "epoch": 0.5722468692087381,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.5149,
      "mean_token_accuracy": 0.1771990954875946,
      "num_tokens": 11368624.0,
      "step": 5095
    },
    {
      "entropy": 5.735827016830444,
      "epoch": 0.5728084461166957,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000497596898548706,
      "loss": 5.6196,
      "mean_token_accuracy": 0.17638506442308427,
      "num_tokens": 11380520.0,
      "step": 5100
    },
    {
      "entropy": 5.643316745758057,
      "epoch": 0.5733700230246532,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.524,
      "mean_token_accuracy": 0.1832396864891052,
      "num_tokens": 11392293.0,
      "step": 5105
    },
    {
      "entropy": 5.6655292987823485,
      "epoch": 0.5739315999326108,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.5975,
      "mean_token_accuracy": 0.17412488609552385,
      "num_tokens": 11403866.0,
      "step": 5110
    },
    {
      "entropy": 5.632191896438599,
      "epoch": 0.5744931768405683,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000497579310071541,
      "loss": 5.4125,
      "mean_token_accuracy": 0.18685944229364396,
      "num_tokens": 11415334.0,
      "step": 5115
    },
    {
      "entropy": 5.625568485260009,
      "epoch": 0.5750547537485259,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.5615,
      "mean_token_accuracy": 0.1853780508041382,
      "num_tokens": 11426840.0,
      "step": 5120
    },
    {
      "entropy": 5.710986137390137,
      "epoch": 0.5756163306564834,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.7092,
      "mean_token_accuracy": 0.17285396456718444,
      "num_tokens": 11438497.0,
      "step": 5125
    },
    {
      "entropy": 5.636830997467041,
      "epoch": 0.576177907564441,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.422,
      "mean_token_accuracy": 0.1875576600432396,
      "num_tokens": 11448705.0,
      "step": 5130
    },
    {
      "entropy": 5.744878578186035,
      "epoch": 0.5767394844723985,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.6799,
      "mean_token_accuracy": 0.17194684743881225,
      "num_tokens": 11460610.0,
      "step": 5135
    },
    {
      "entropy": 5.654635953903198,
      "epoch": 0.5773010613803561,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18467996567487716,
      "num_tokens": 11471170.0,
      "step": 5140
    },
    {
      "entropy": 5.605772352218628,
      "epoch": 0.5778626382883136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1930658996105194,
      "num_tokens": 11482090.0,
      "step": 5145
    },
    {
      "entropy": 5.600895071029663,
      "epoch": 0.5784242151962711,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.5308,
      "mean_token_accuracy": 0.1807967096567154,
      "num_tokens": 11493650.0,
      "step": 5150
    },
    {
      "entropy": 5.655977535247803,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.5166,
      "mean_token_accuracy": 0.17395519316196442,
      "num_tokens": 11505903.0,
      "step": 5155
    },
    {
      "entropy": 5.729590225219726,
      "epoch": 0.5795473690121862,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.5538,
      "mean_token_accuracy": 0.17925923317670822,
      "num_tokens": 11517194.0,
      "step": 5160
    },
    {
      "entropy": 5.67247166633606,
      "epoch": 0.5801089459201437,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.5651,
      "mean_token_accuracy": 0.18369649052619935,
      "num_tokens": 11528683.0,
      "step": 5165
    },
    {
      "entropy": 5.634302377700806,
      "epoch": 0.5806705228281013,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.5827,
      "mean_token_accuracy": 0.18153703361749648,
      "num_tokens": 11541940.0,
      "step": 5170
    },
    {
      "entropy": 5.608839464187622,
      "epoch": 0.5812320997360588,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18154302686452867,
      "num_tokens": 11553045.0,
      "step": 5175
    },
    {
      "entropy": 5.605997133255005,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.4374,
      "mean_token_accuracy": 0.18805930465459825,
      "num_tokens": 11564401.0,
      "step": 5180
    },
    {
      "entropy": 5.664146041870117,
      "epoch": 0.5823552535519739,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.4843,
      "mean_token_accuracy": 0.17927162647247313,
      "num_tokens": 11576424.0,
      "step": 5185
    },
    {
      "entropy": 5.682039451599121,
      "epoch": 0.5829168304599315,
      "grad_norm": 0.796875,
      "learning_rate": 0.00049749041104062,
      "loss": 5.4976,
      "mean_token_accuracy": 0.1816076397895813,
      "num_tokens": 11587853.0,
      "step": 5190
    },
    {
      "entropy": 5.690493535995484,
      "epoch": 0.583478407367889,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.4956,
      "mean_token_accuracy": 0.18630104660987853,
      "num_tokens": 11597938.0,
      "step": 5195
    },
    {
      "entropy": 5.750465011596679,
      "epoch": 0.5840399842758466,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.66,
      "mean_token_accuracy": 0.18166985362768173,
      "num_tokens": 11610077.0,
      "step": 5200
    },
    {
      "entropy": 5.602049064636231,
      "epoch": 0.5846015611838041,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.5234,
      "mean_token_accuracy": 0.18146990537643432,
      "num_tokens": 11622079.0,
      "step": 5205
    },
    {
      "entropy": 5.563369989395142,
      "epoch": 0.5851631380917617,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.444,
      "mean_token_accuracy": 0.19037307500839235,
      "num_tokens": 11633047.0,
      "step": 5210
    },
    {
      "entropy": 5.604269695281983,
      "epoch": 0.5857247149997192,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.381,
      "mean_token_accuracy": 0.18752022832632065,
      "num_tokens": 11643858.0,
      "step": 5215
    },
    {
      "entropy": 5.580379056930542,
      "epoch": 0.5862862919076768,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.4198,
      "mean_token_accuracy": 0.18881472647190095,
      "num_tokens": 11654550.0,
      "step": 5220
    },
    {
      "entropy": 5.594556188583374,
      "epoch": 0.5868478688156343,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000497448379305403,
      "loss": 5.4572,
      "mean_token_accuracy": 0.1832424968481064,
      "num_tokens": 11665787.0,
      "step": 5225
    },
    {
      "entropy": 5.689477443695068,
      "epoch": 0.5874094457235919,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.5681,
      "mean_token_accuracy": 0.1849393665790558,
      "num_tokens": 11677717.0,
      "step": 5230
    },
    {
      "entropy": 5.644946050643921,
      "epoch": 0.5879710226315494,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497436306489741,
      "loss": 5.4549,
      "mean_token_accuracy": 0.18394995778799056,
      "num_tokens": 11687714.0,
      "step": 5235
    },
    {
      "entropy": 5.588587427139283,
      "epoch": 0.588532599539507,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.3444,
      "mean_token_accuracy": 0.1919522076845169,
      "num_tokens": 11698777.0,
      "step": 5240
    },
    {
      "entropy": 5.590402507781983,
      "epoch": 0.5890941764474645,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.451,
      "mean_token_accuracy": 0.1856154203414917,
      "num_tokens": 11710922.0,
      "step": 5245
    },
    {
      "entropy": 5.621795129776001,
      "epoch": 0.589655753355422,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.4946,
      "mean_token_accuracy": 0.18049226105213165,
      "num_tokens": 11721450.0,
      "step": 5250
    },
    {
      "entropy": 5.5702227592468265,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.4001,
      "mean_token_accuracy": 0.18821180015802383,
      "num_tokens": 11732458.0,
      "step": 5255
    },
    {
      "entropy": 5.550134181976318,
      "epoch": 0.5907789071713371,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.3723,
      "mean_token_accuracy": 0.18859368115663527,
      "num_tokens": 11742649.0,
      "step": 5260
    },
    {
      "entropy": 5.579849529266357,
      "epoch": 0.5913404840792946,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497399918069926,
      "loss": 5.395,
      "mean_token_accuracy": 0.19034955650568008,
      "num_tokens": 11753811.0,
      "step": 5265
    },
    {
      "entropy": 5.644034576416016,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.5059,
      "mean_token_accuracy": 0.17806767523288727,
      "num_tokens": 11763965.0,
      "step": 5270
    },
    {
      "entropy": 5.663493490219116,
      "epoch": 0.5924636378952097,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.5225,
      "mean_token_accuracy": 0.1867716819047928,
      "num_tokens": 11774323.0,
      "step": 5275
    },
    {
      "entropy": 5.654573392868042,
      "epoch": 0.5930252148031673,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.4907,
      "mean_token_accuracy": 0.18478358387947083,
      "num_tokens": 11786157.0,
      "step": 5280
    },
    {
      "entropy": 5.627108860015869,
      "epoch": 0.5935867917111248,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.4685,
      "mean_token_accuracy": 0.19186956882476808,
      "num_tokens": 11797344.0,
      "step": 5285
    },
    {
      "entropy": 5.478252601623535,
      "epoch": 0.5941483686190824,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.4105,
      "mean_token_accuracy": 0.1851436674594879,
      "num_tokens": 11808149.0,
      "step": 5290
    },
    {
      "entropy": 5.6173317432403564,
      "epoch": 0.5947099455270399,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.432,
      "mean_token_accuracy": 0.19321200102567673,
      "num_tokens": 11820203.0,
      "step": 5295
    },
    {
      "entropy": 5.651402282714844,
      "epoch": 0.5952715224349975,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1802552357316017,
      "num_tokens": 11830172.0,
      "step": 5300
    },
    {
      "entropy": 5.639830827713013,
      "epoch": 0.595833099342955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18880193680524826,
      "num_tokens": 11840203.0,
      "step": 5305
    },
    {
      "entropy": 5.635803365707398,
      "epoch": 0.5963946762509126,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.438,
      "mean_token_accuracy": 0.1821918308734894,
      "num_tokens": 11851002.0,
      "step": 5310
    },
    {
      "entropy": 5.624815845489502,
      "epoch": 0.5969562531588701,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.4368,
      "mean_token_accuracy": 0.1811597928404808,
      "num_tokens": 11861559.0,
      "step": 5315
    },
    {
      "entropy": 5.591391468048096,
      "epoch": 0.5975178300668277,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.5135,
      "mean_token_accuracy": 0.1832076296210289,
      "num_tokens": 11872535.0,
      "step": 5320
    },
    {
      "entropy": 5.686180877685547,
      "epoch": 0.5980794069747852,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.5098,
      "mean_token_accuracy": 0.18663521558046342,
      "num_tokens": 11884308.0,
      "step": 5325
    },
    {
      "entropy": 5.7295904636383055,
      "epoch": 0.5986409838827428,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.5846,
      "mean_token_accuracy": 0.17991345077753068,
      "num_tokens": 11896136.0,
      "step": 5330
    },
    {
      "entropy": 5.63388876914978,
      "epoch": 0.5992025607907003,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497314020447805,
      "loss": 5.415,
      "mean_token_accuracy": 0.19730381071567535,
      "num_tokens": 11906497.0,
      "step": 5335
    },
    {
      "entropy": 5.653187227249146,
      "epoch": 0.5997641376986579,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.5354,
      "mean_token_accuracy": 0.17559385746717454,
      "num_tokens": 11917872.0,
      "step": 5340
    },
    {
      "entropy": 5.677523183822632,
      "epoch": 0.6003257146066153,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497301636089344,
      "loss": 5.5549,
      "mean_token_accuracy": 0.18246214091777802,
      "num_tokens": 11929727.0,
      "step": 5345
    },
    {
      "entropy": 5.656993293762207,
      "epoch": 0.6008872915145729,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.3367,
      "mean_token_accuracy": 0.18527638912200928,
      "num_tokens": 11939996.0,
      "step": 5350
    },
    {
      "entropy": 5.556303691864014,
      "epoch": 0.6014488684225304,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.4267,
      "mean_token_accuracy": 0.1879384532570839,
      "num_tokens": 11950683.0,
      "step": 5355
    },
    {
      "entropy": 5.543930959701538,
      "epoch": 0.602010445330488,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.427,
      "mean_token_accuracy": 0.1841173693537712,
      "num_tokens": 11962447.0,
      "step": 5360
    },
    {
      "entropy": 5.661401414871216,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.5051,
      "mean_token_accuracy": 0.18415847569704055,
      "num_tokens": 11972740.0,
      "step": 5365
    },
    {
      "entropy": 5.612428188323975,
      "epoch": 0.6031335991464031,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.3873,
      "mean_token_accuracy": 0.19047181457281112,
      "num_tokens": 11983292.0,
      "step": 5370
    },
    {
      "entropy": 5.710958003997803,
      "epoch": 0.6036951760543606,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.6055,
      "mean_token_accuracy": 0.17753715217113494,
      "num_tokens": 11996384.0,
      "step": 5375
    },
    {
      "entropy": 5.722277307510376,
      "epoch": 0.6042567529623182,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.5277,
      "mean_token_accuracy": 0.19452916085720062,
      "num_tokens": 12006938.0,
      "step": 5380
    },
    {
      "entropy": 5.515638828277588,
      "epoch": 0.6048183298702757,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.3535,
      "mean_token_accuracy": 0.18811541199684143,
      "num_tokens": 12018729.0,
      "step": 5385
    },
    {
      "entropy": 5.749721479415894,
      "epoch": 0.6053799067782333,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.7007,
      "mean_token_accuracy": 0.17779027819633483,
      "num_tokens": 12029625.0,
      "step": 5390
    },
    {
      "entropy": 5.608671379089356,
      "epoch": 0.6059414836861908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.4148,
      "mean_token_accuracy": 0.17977323532104492,
      "num_tokens": 12040243.0,
      "step": 5395
    },
    {
      "entropy": 5.724008560180664,
      "epoch": 0.6065030605941484,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.6932,
      "mean_token_accuracy": 0.1744496300816536,
      "num_tokens": 12051961.0,
      "step": 5400
    },
    {
      "entropy": 5.685330486297607,
      "epoch": 0.6070646375021059,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.557,
      "mean_token_accuracy": 0.1856486365199089,
      "num_tokens": 12063969.0,
      "step": 5405
    },
    {
      "entropy": 5.660473442077636,
      "epoch": 0.6076262144100635,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1856810301542282,
      "num_tokens": 12075682.0,
      "step": 5410
    },
    {
      "entropy": 5.668943977355957,
      "epoch": 0.608187791318021,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.5772,
      "mean_token_accuracy": 0.18059138357639312,
      "num_tokens": 12086595.0,
      "step": 5415
    },
    {
      "entropy": 5.614640569686889,
      "epoch": 0.6087493682259786,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.4474,
      "mean_token_accuracy": 0.1859290048480034,
      "num_tokens": 12099421.0,
      "step": 5420
    },
    {
      "entropy": 5.706504201889038,
      "epoch": 0.6093109451339361,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.561,
      "mean_token_accuracy": 0.18208210915327072,
      "num_tokens": 12109680.0,
      "step": 5425
    },
    {
      "entropy": 5.56081190109253,
      "epoch": 0.6098725220418937,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.4377,
      "mean_token_accuracy": 0.18923980444669725,
      "num_tokens": 12121196.0,
      "step": 5430
    },
    {
      "entropy": 5.596039772033691,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000497188902983013,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18058521449565887,
      "num_tokens": 12132262.0,
      "step": 5435
    },
    {
      "entropy": 5.672529554367065,
      "epoch": 0.6109956758578087,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.5376,
      "mean_token_accuracy": 0.18426449596881866,
      "num_tokens": 12143747.0,
      "step": 5440
    },
    {
      "entropy": 5.7709879875183105,
      "epoch": 0.6115572527657662,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.6506,
      "mean_token_accuracy": 0.17480409294366836,
      "num_tokens": 12155887.0,
      "step": 5445
    },
    {
      "entropy": 5.604062223434449,
      "epoch": 0.6121188296737238,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18045834302902222,
      "num_tokens": 12166718.0,
      "step": 5450
    },
    {
      "entropy": 5.656278991699219,
      "epoch": 0.6126804065816813,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.5803,
      "mean_token_accuracy": 0.18506073504686354,
      "num_tokens": 12178540.0,
      "step": 5455
    },
    {
      "entropy": 5.756329917907715,
      "epoch": 0.6132419834896389,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1760148972272873,
      "num_tokens": 12189820.0,
      "step": 5460
    },
    {
      "entropy": 5.647144031524658,
      "epoch": 0.6138035603975964,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1846558704972267,
      "num_tokens": 12201679.0,
      "step": 5465
    },
    {
      "entropy": 5.637397241592407,
      "epoch": 0.614365137305554,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.5464,
      "mean_token_accuracy": 0.17856902182102202,
      "num_tokens": 12213793.0,
      "step": 5470
    },
    {
      "entropy": 5.7327272415161135,
      "epoch": 0.6149267142135115,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.6303,
      "mean_token_accuracy": 0.1834181472659111,
      "num_tokens": 12224519.0,
      "step": 5475
    },
    {
      "entropy": 5.794642162322998,
      "epoch": 0.6154882911214691,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.601,
      "mean_token_accuracy": 0.17089725136756898,
      "num_tokens": 12236401.0,
      "step": 5480
    },
    {
      "entropy": 5.654254293441772,
      "epoch": 0.6160498680294266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.6198,
      "mean_token_accuracy": 0.17620888203382493,
      "num_tokens": 12247670.0,
      "step": 5485
    },
    {
      "entropy": 5.631163263320923,
      "epoch": 0.6166114449373842,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.5,
      "mean_token_accuracy": 0.1862648382782936,
      "num_tokens": 12258877.0,
      "step": 5490
    },
    {
      "entropy": 5.645384931564331,
      "epoch": 0.6171730218453417,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.4703,
      "mean_token_accuracy": 0.18386943340301515,
      "num_tokens": 12268507.0,
      "step": 5495
    },
    {
      "entropy": 5.657395935058593,
      "epoch": 0.6177345987532993,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497106059033725,
      "loss": 5.4695,
      "mean_token_accuracy": 0.17794615924358367,
      "num_tokens": 12280420.0,
      "step": 5500
    },
    {
      "entropy": 5.588811731338501,
      "epoch": 0.6182961756612568,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.4607,
      "mean_token_accuracy": 0.17786094844341277,
      "num_tokens": 12290876.0,
      "step": 5505
    },
    {
      "entropy": 5.590397119522095,
      "epoch": 0.6188577525692144,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.3641,
      "mean_token_accuracy": 0.19031805992126466,
      "num_tokens": 12301831.0,
      "step": 5510
    },
    {
      "entropy": 5.531795454025269,
      "epoch": 0.6194193294771719,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497086771456353,
      "loss": 5.3883,
      "mean_token_accuracy": 0.19162252992391587,
      "num_tokens": 12312678.0,
      "step": 5515
    },
    {
      "entropy": 5.728099012374878,
      "epoch": 0.6199809063851295,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.5614,
      "mean_token_accuracy": 0.17756428718566894,
      "num_tokens": 12322905.0,
      "step": 5520
    },
    {
      "entropy": 5.61423659324646,
      "epoch": 0.620542483293087,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18627428263425827,
      "num_tokens": 12333091.0,
      "step": 5525
    },
    {
      "entropy": 5.611349964141846,
      "epoch": 0.6211040602010446,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.4394,
      "mean_token_accuracy": 0.18721218258142472,
      "num_tokens": 12346310.0,
      "step": 5530
    },
    {
      "entropy": 5.566066265106201,
      "epoch": 0.621665637109002,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.3653,
      "mean_token_accuracy": 0.1821967825293541,
      "num_tokens": 12357138.0,
      "step": 5535
    },
    {
      "entropy": 5.688712310791016,
      "epoch": 0.6222272140169596,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.4594,
      "mean_token_accuracy": 0.1864299327135086,
      "num_tokens": 12367382.0,
      "step": 5540
    },
    {
      "entropy": 5.58078293800354,
      "epoch": 0.6227887909249171,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.4449,
      "mean_token_accuracy": 0.1790702760219574,
      "num_tokens": 12380357.0,
      "step": 5545
    },
    {
      "entropy": 5.665856456756591,
      "epoch": 0.6233503678328747,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.6145,
      "mean_token_accuracy": 0.1763693630695343,
      "num_tokens": 12392187.0,
      "step": 5550
    },
    {
      "entropy": 5.67484827041626,
      "epoch": 0.6239119447408322,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.4795,
      "mean_token_accuracy": 0.18667882531881333,
      "num_tokens": 12402924.0,
      "step": 5555
    },
    {
      "entropy": 5.704171895980835,
      "epoch": 0.6244735216487898,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497028526894682,
      "loss": 5.6265,
      "mean_token_accuracy": 0.17679935991764067,
      "num_tokens": 12413554.0,
      "step": 5560
    },
    {
      "entropy": 5.624708557128907,
      "epoch": 0.6250350985567473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.494,
      "mean_token_accuracy": 0.18862047791481018,
      "num_tokens": 12424525.0,
      "step": 5565
    },
    {
      "entropy": 5.608011054992676,
      "epoch": 0.6255966754647049,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.4887,
      "mean_token_accuracy": 0.18696973323822022,
      "num_tokens": 12436405.0,
      "step": 5570
    },
    {
      "entropy": 5.664109802246093,
      "epoch": 0.6261582523726624,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.5354,
      "mean_token_accuracy": 0.18118735551834106,
      "num_tokens": 12446909.0,
      "step": 5575
    },
    {
      "entropy": 5.6651567935943605,
      "epoch": 0.62671982928062,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.5247,
      "mean_token_accuracy": 0.18565261065959932,
      "num_tokens": 12458739.0,
      "step": 5580
    },
    {
      "entropy": 5.625951433181763,
      "epoch": 0.6272814061885775,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.4692,
      "mean_token_accuracy": 0.18096400946378707,
      "num_tokens": 12469529.0,
      "step": 5585
    },
    {
      "entropy": 5.706216526031494,
      "epoch": 0.6278429830965351,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.4968,
      "mean_token_accuracy": 0.18015918731689454,
      "num_tokens": 12480067.0,
      "step": 5590
    },
    {
      "entropy": 5.612037277221679,
      "epoch": 0.6284045600044926,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.3705,
      "mean_token_accuracy": 0.1939912036061287,
      "num_tokens": 12491043.0,
      "step": 5595
    },
    {
      "entropy": 5.5298243999481205,
      "epoch": 0.6289661369124502,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1879752144217491,
      "num_tokens": 12501852.0,
      "step": 5600
    },
    {
      "entropy": 5.633104610443115,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.6362,
      "mean_token_accuracy": 0.1773230716586113,
      "num_tokens": 12512696.0,
      "step": 5605
    },
    {
      "entropy": 5.687568044662475,
      "epoch": 0.6300892907283653,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.4838,
      "mean_token_accuracy": 0.18605547547340393,
      "num_tokens": 12523643.0,
      "step": 5610
    },
    {
      "entropy": 5.637312746047973,
      "epoch": 0.6306508676363228,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.4531,
      "mean_token_accuracy": 0.18806927651166916,
      "num_tokens": 12534827.0,
      "step": 5615
    },
    {
      "entropy": 5.51575231552124,
      "epoch": 0.6312124445442804,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.3211,
      "mean_token_accuracy": 0.18738449215888978,
      "num_tokens": 12545977.0,
      "step": 5620
    },
    {
      "entropy": 5.510044574737549,
      "epoch": 0.6317740214522379,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.3967,
      "mean_token_accuracy": 0.18319795727729798,
      "num_tokens": 12555890.0,
      "step": 5625
    },
    {
      "entropy": 5.582265901565552,
      "epoch": 0.6323355983601955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.4559,
      "mean_token_accuracy": 0.1880454272031784,
      "num_tokens": 12566412.0,
      "step": 5630
    },
    {
      "entropy": 5.660838651657104,
      "epoch": 0.6328971752681529,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.4948,
      "mean_token_accuracy": 0.1843516856431961,
      "num_tokens": 12577175.0,
      "step": 5635
    },
    {
      "entropy": 5.573785352706909,
      "epoch": 0.6334587521761105,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000496923567209739,
      "loss": 5.418,
      "mean_token_accuracy": 0.18194658309221268,
      "num_tokens": 12588760.0,
      "step": 5640
    },
    {
      "entropy": 5.607411098480225,
      "epoch": 0.634020329084068,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.4933,
      "mean_token_accuracy": 0.184040305018425,
      "num_tokens": 12599899.0,
      "step": 5645
    },
    {
      "entropy": 5.666639137268066,
      "epoch": 0.6345819059920256,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1944987341761589,
      "num_tokens": 12610836.0,
      "step": 5650
    },
    {
      "entropy": 5.558991622924805,
      "epoch": 0.6351434828999831,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1889163836836815,
      "num_tokens": 12621855.0,
      "step": 5655
    },
    {
      "entropy": 5.622348356246948,
      "epoch": 0.6357050598079407,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.4939,
      "mean_token_accuracy": 0.18547171205282212,
      "num_tokens": 12632964.0,
      "step": 5660
    },
    {
      "entropy": 5.5253260135650635,
      "epoch": 0.6362666367158982,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496890396295883,
      "loss": 5.3481,
      "mean_token_accuracy": 0.18995504528284074,
      "num_tokens": 12643717.0,
      "step": 5665
    },
    {
      "entropy": 5.664473533630371,
      "epoch": 0.6368282136238558,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.4499,
      "mean_token_accuracy": 0.18521457761526108,
      "num_tokens": 12654491.0,
      "step": 5670
    },
    {
      "entropy": 5.633564758300781,
      "epoch": 0.6373897905318133,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.5143,
      "mean_token_accuracy": 0.18290475755929947,
      "num_tokens": 12666128.0,
      "step": 5675
    },
    {
      "entropy": 5.68228931427002,
      "epoch": 0.6379513674397709,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.5926,
      "mean_token_accuracy": 0.18103973865509032,
      "num_tokens": 12677957.0,
      "step": 5680
    },
    {
      "entropy": 5.62337007522583,
      "epoch": 0.6385129443477284,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18417457789182662,
      "num_tokens": 12687964.0,
      "step": 5685
    },
    {
      "entropy": 5.54130711555481,
      "epoch": 0.639074521255686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.4115,
      "mean_token_accuracy": 0.18455131351947784,
      "num_tokens": 12698805.0,
      "step": 5690
    },
    {
      "entropy": 5.651135206222534,
      "epoch": 0.6396360981636435,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.4429,
      "mean_token_accuracy": 0.1839181214570999,
      "num_tokens": 12709942.0,
      "step": 5695
    },
    {
      "entropy": 5.7260692596435545,
      "epoch": 0.6401976750716011,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.5058,
      "mean_token_accuracy": 0.18354142159223558,
      "num_tokens": 12721856.0,
      "step": 5700
    },
    {
      "entropy": 5.517352533340454,
      "epoch": 0.6407592519795586,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18346032798290252,
      "num_tokens": 12732619.0,
      "step": 5705
    },
    {
      "entropy": 5.582686567306519,
      "epoch": 0.6413208288875162,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.4436,
      "mean_token_accuracy": 0.18647329956293107,
      "num_tokens": 12743478.0,
      "step": 5710
    },
    {
      "entropy": 5.572625589370728,
      "epoch": 0.6418824057954737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18452126830816268,
      "num_tokens": 12754761.0,
      "step": 5715
    },
    {
      "entropy": 5.696010589599609,
      "epoch": 0.6424439827034313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.5034,
      "mean_token_accuracy": 0.18740470111370086,
      "num_tokens": 12765803.0,
      "step": 5720
    },
    {
      "entropy": 5.626456689834595,
      "epoch": 0.6430055596113888,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.5101,
      "mean_token_accuracy": 0.18201464265584946,
      "num_tokens": 12777882.0,
      "step": 5725
    },
    {
      "entropy": 5.663591432571411,
      "epoch": 0.6435671365193463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.5311,
      "mean_token_accuracy": 0.18065996319055558,
      "num_tokens": 12788268.0,
      "step": 5730
    },
    {
      "entropy": 5.607796144485474,
      "epoch": 0.6441287134273038,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.5249,
      "mean_token_accuracy": 0.18313902020454406,
      "num_tokens": 12800506.0,
      "step": 5735
    },
    {
      "entropy": 5.621351003646851,
      "epoch": 0.6446902903352614,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18507064878940582,
      "num_tokens": 12813040.0,
      "step": 5740
    },
    {
      "entropy": 5.6962645053863525,
      "epoch": 0.6452518672432189,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.5313,
      "mean_token_accuracy": 0.18668071627616883,
      "num_tokens": 12825534.0,
      "step": 5745
    },
    {
      "entropy": 5.6005199432373045,
      "epoch": 0.6458134441511765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1886446878314018,
      "num_tokens": 12835882.0,
      "step": 5750
    },
    {
      "entropy": 5.6180088996887205,
      "epoch": 0.646375021059134,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.5992,
      "mean_token_accuracy": 0.17473047077655793,
      "num_tokens": 12847663.0,
      "step": 5755
    },
    {
      "entropy": 5.800768613815308,
      "epoch": 0.6469365979670916,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.7284,
      "mean_token_accuracy": 0.1773770585656166,
      "num_tokens": 12860789.0,
      "step": 5760
    },
    {
      "entropy": 5.668123197555542,
      "epoch": 0.6474981748750491,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1829041913151741,
      "num_tokens": 12872509.0,
      "step": 5765
    },
    {
      "entropy": 5.59938793182373,
      "epoch": 0.6480597517830067,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.4634,
      "mean_token_accuracy": 0.18331168442964554,
      "num_tokens": 12884321.0,
      "step": 5770
    },
    {
      "entropy": 5.498247337341309,
      "epoch": 0.6486213286909642,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.3474,
      "mean_token_accuracy": 0.19287209957838058,
      "num_tokens": 12894559.0,
      "step": 5775
    },
    {
      "entropy": 5.591831207275391,
      "epoch": 0.6491829055989218,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.4583,
      "mean_token_accuracy": 0.18731842786073685,
      "num_tokens": 12904934.0,
      "step": 5780
    },
    {
      "entropy": 5.702964782714844,
      "epoch": 0.6497444825068793,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000496728717732762,
      "loss": 5.5775,
      "mean_token_accuracy": 0.17702377438545228,
      "num_tokens": 12916551.0,
      "step": 5785
    },
    {
      "entropy": 5.626647996902466,
      "epoch": 0.6503060594148369,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.4767,
      "mean_token_accuracy": 0.17969613373279572,
      "num_tokens": 12928097.0,
      "step": 5790
    },
    {
      "entropy": 5.616903018951416,
      "epoch": 0.6508676363227944,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.3537,
      "mean_token_accuracy": 0.19613699167966842,
      "num_tokens": 12939161.0,
      "step": 5795
    },
    {
      "entropy": 5.529639816284179,
      "epoch": 0.651429213230752,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.4048,
      "mean_token_accuracy": 0.1856609344482422,
      "num_tokens": 12950862.0,
      "step": 5800
    },
    {
      "entropy": 5.61510820388794,
      "epoch": 0.6519907901387095,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.4564,
      "mean_token_accuracy": 0.18268677443265915,
      "num_tokens": 12962333.0,
      "step": 5805
    },
    {
      "entropy": 5.6155867099761965,
      "epoch": 0.6525523670466671,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.4093,
      "mean_token_accuracy": 0.18707340359687805,
      "num_tokens": 12973136.0,
      "step": 5810
    },
    {
      "entropy": 5.61257791519165,
      "epoch": 0.6531139439546246,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.4633,
      "mean_token_accuracy": 0.18540784120559692,
      "num_tokens": 12983602.0,
      "step": 5815
    },
    {
      "entropy": 5.633553457260132,
      "epoch": 0.6536755208625822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.3689,
      "mean_token_accuracy": 0.19493580460548401,
      "num_tokens": 12994129.0,
      "step": 5820
    },
    {
      "entropy": 5.602857637405395,
      "epoch": 0.6542370977705396,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.4324,
      "mean_token_accuracy": 0.18489825129508972,
      "num_tokens": 13005071.0,
      "step": 5825
    },
    {
      "entropy": 5.700960159301758,
      "epoch": 0.6547986746784972,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.6004,
      "mean_token_accuracy": 0.18180658519268036,
      "num_tokens": 13016202.0,
      "step": 5830
    },
    {
      "entropy": 5.688804244995117,
      "epoch": 0.6553602515864547,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.59,
      "mean_token_accuracy": 0.17385727763175965,
      "num_tokens": 13028468.0,
      "step": 5835
    },
    {
      "entropy": 5.646315670013427,
      "epoch": 0.6559218284944123,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496653255842169,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18384341299533843,
      "num_tokens": 13039667.0,
      "step": 5840
    },
    {
      "entropy": 5.633904647827149,
      "epoch": 0.6564834054023698,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.5363,
      "mean_token_accuracy": 0.17847258746623992,
      "num_tokens": 13051583.0,
      "step": 5845
    },
    {
      "entropy": 5.640959644317627,
      "epoch": 0.6570449823103274,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.4753,
      "mean_token_accuracy": 0.18201233744621276,
      "num_tokens": 13062401.0,
      "step": 5850
    },
    {
      "entropy": 5.54059944152832,
      "epoch": 0.6576065592182849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.3585,
      "mean_token_accuracy": 0.1888939693570137,
      "num_tokens": 13074219.0,
      "step": 5855
    },
    {
      "entropy": 5.664067983627319,
      "epoch": 0.6581681361262425,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.5706,
      "mean_token_accuracy": 0.18419669717550277,
      "num_tokens": 13086073.0,
      "step": 5860
    },
    {
      "entropy": 5.633033895492554,
      "epoch": 0.6587297130342,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.4811,
      "mean_token_accuracy": 0.17839719504117965,
      "num_tokens": 13096340.0,
      "step": 5865
    },
    {
      "entropy": 5.5658525943756105,
      "epoch": 0.6592912899421576,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1857535645365715,
      "num_tokens": 13107273.0,
      "step": 5870
    },
    {
      "entropy": 5.569044017791748,
      "epoch": 0.6598528668501151,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.4231,
      "mean_token_accuracy": 0.1794855549931526,
      "num_tokens": 13117388.0,
      "step": 5875
    },
    {
      "entropy": 5.628138017654419,
      "epoch": 0.6604144437580727,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496597838078686,
      "loss": 5.5473,
      "mean_token_accuracy": 0.1797400563955307,
      "num_tokens": 13128720.0,
      "step": 5880
    },
    {
      "entropy": 5.648764038085938,
      "epoch": 0.6609760206660302,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496590879105005,
      "loss": 5.4603,
      "mean_token_accuracy": 0.18696583360433577,
      "num_tokens": 13139405.0,
      "step": 5885
    },
    {
      "entropy": 5.649448871612549,
      "epoch": 0.6615375975739878,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.5447,
      "mean_token_accuracy": 0.18331155627965928,
      "num_tokens": 13151408.0,
      "step": 5890
    },
    {
      "entropy": 5.6605949878692625,
      "epoch": 0.6620991744819453,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.4533,
      "mean_token_accuracy": 0.18801768273115158,
      "num_tokens": 13162592.0,
      "step": 5895
    },
    {
      "entropy": 5.6841260433197025,
      "epoch": 0.6626607513899029,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.4751,
      "mean_token_accuracy": 0.19073681086301802,
      "num_tokens": 13173040.0,
      "step": 5900
    },
    {
      "entropy": 5.599741792678833,
      "epoch": 0.6632223282978604,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.5629,
      "mean_token_accuracy": 0.17858159244060517,
      "num_tokens": 13185475.0,
      "step": 5905
    },
    {
      "entropy": 5.576128816604614,
      "epoch": 0.663783905205818,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.4244,
      "mean_token_accuracy": 0.19086278825998307,
      "num_tokens": 13196481.0,
      "step": 5910
    },
    {
      "entropy": 5.544696950912476,
      "epoch": 0.6643454821137755,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.3013,
      "mean_token_accuracy": 0.19332729279994965,
      "num_tokens": 13207947.0,
      "step": 5915
    },
    {
      "entropy": 5.4862476825714115,
      "epoch": 0.664907059021733,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.3627,
      "mean_token_accuracy": 0.19672193378210068,
      "num_tokens": 13219392.0,
      "step": 5920
    },
    {
      "entropy": 5.536784982681274,
      "epoch": 0.6654686359296905,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.3981,
      "mean_token_accuracy": 0.19011834710836412,
      "num_tokens": 13231556.0,
      "step": 5925
    },
    {
      "entropy": 5.600260257720947,
      "epoch": 0.6660302128376481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.3391,
      "mean_token_accuracy": 0.19259928464889525,
      "num_tokens": 13241287.0,
      "step": 5930
    },
    {
      "entropy": 5.543307828903198,
      "epoch": 0.6665917897456056,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1825370728969574,
      "num_tokens": 13254120.0,
      "step": 5935
    },
    {
      "entropy": 5.588070392608643,
      "epoch": 0.6671533666535632,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1794518396258354,
      "num_tokens": 13265745.0,
      "step": 5940
    },
    {
      "entropy": 5.714030075073242,
      "epoch": 0.6677149435615207,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496506821152345,
      "loss": 5.5249,
      "mean_token_accuracy": 0.18471344709396362,
      "num_tokens": 13277791.0,
      "step": 5945
    },
    {
      "entropy": 5.587875127792358,
      "epoch": 0.6682765204694783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.3866,
      "mean_token_accuracy": 0.18252573758363724,
      "num_tokens": 13288078.0,
      "step": 5950
    },
    {
      "entropy": 5.589505386352539,
      "epoch": 0.6688380973774358,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000496492712743046,
      "loss": 5.4309,
      "mean_token_accuracy": 0.19044736474752427,
      "num_tokens": 13300547.0,
      "step": 5955
    },
    {
      "entropy": 5.638584613800049,
      "epoch": 0.6693996742853934,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.4834,
      "mean_token_accuracy": 0.18632941842079162,
      "num_tokens": 13310339.0,
      "step": 5960
    },
    {
      "entropy": 5.692937517166138,
      "epoch": 0.6699612511933509,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.4707,
      "mean_token_accuracy": 0.18331103175878524,
      "num_tokens": 13321294.0,
      "step": 5965
    },
    {
      "entropy": 5.6274940967559814,
      "epoch": 0.6705228281013085,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.497,
      "mean_token_accuracy": 0.1822831779718399,
      "num_tokens": 13332608.0,
      "step": 5970
    },
    {
      "entropy": 5.599426651000977,
      "epoch": 0.671084405009266,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.4312,
      "mean_token_accuracy": 0.19674972593784332,
      "num_tokens": 13344169.0,
      "step": 5975
    },
    {
      "entropy": 5.5073963642120365,
      "epoch": 0.6716459819172236,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.3947,
      "mean_token_accuracy": 0.18853919208049774,
      "num_tokens": 13355713.0,
      "step": 5980
    },
    {
      "entropy": 5.647307634353638,
      "epoch": 0.6722075588251811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.4914,
      "mean_token_accuracy": 0.18118446171283722,
      "num_tokens": 13366644.0,
      "step": 5985
    },
    {
      "entropy": 5.610023880004883,
      "epoch": 0.6727691357331387,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.4569,
      "mean_token_accuracy": 0.18272807598114013,
      "num_tokens": 13378125.0,
      "step": 5990
    },
    {
      "entropy": 5.586076307296753,
      "epoch": 0.6733307126410962,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.5002,
      "mean_token_accuracy": 0.18203672617673874,
      "num_tokens": 13389681.0,
      "step": 5995
    },
    {
      "entropy": 5.703607749938965,
      "epoch": 0.6738922895490538,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.4965,
      "mean_token_accuracy": 0.17744856774806977,
      "num_tokens": 13399814.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.44550887278315,
      "eval_loss": 5.472325325012207,
      "eval_mean_token_accuracy": 0.19233731558723705,
      "eval_num_tokens": 13399814.0,
      "eval_runtime": 18.3219,
      "eval_samples_per_second": 1692.677,
      "eval_steps_per_second": 211.605,
      "step": 6000
    },
    {
      "entropy": 5.668199634552002,
      "epoch": 0.6744538664570113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.5071,
      "mean_token_accuracy": 0.1850060671567917,
      "num_tokens": 13410614.0,
      "step": 6005
    },
    {
      "entropy": 5.590621900558472,
      "epoch": 0.6750154433649689,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496414612292672,
      "loss": 5.5233,
      "mean_token_accuracy": 0.17850325107574463,
      "num_tokens": 13421392.0,
      "step": 6010
    },
    {
      "entropy": 5.691341209411621,
      "epoch": 0.6755770202729264,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.5468,
      "mean_token_accuracy": 0.1806807518005371,
      "num_tokens": 13432004.0,
      "step": 6015
    },
    {
      "entropy": 5.6276312351226805,
      "epoch": 0.6761385971808839,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.4082,
      "mean_token_accuracy": 0.18874668031930925,
      "num_tokens": 13442450.0,
      "step": 6020
    },
    {
      "entropy": 5.597358131408692,
      "epoch": 0.6767001740888414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.4531,
      "mean_token_accuracy": 0.18074196130037307,
      "num_tokens": 13453848.0,
      "step": 6025
    },
    {
      "entropy": 5.591590738296508,
      "epoch": 0.677261750996799,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.449,
      "mean_token_accuracy": 0.18535481244325638,
      "num_tokens": 13465119.0,
      "step": 6030
    },
    {
      "entropy": 5.5586638927459715,
      "epoch": 0.6778233279047565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.4349,
      "mean_token_accuracy": 0.1870497465133667,
      "num_tokens": 13475937.0,
      "step": 6035
    },
    {
      "entropy": 5.578341245651245,
      "epoch": 0.6783849048127141,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.4094,
      "mean_token_accuracy": 0.19031698256731033,
      "num_tokens": 13487430.0,
      "step": 6040
    },
    {
      "entropy": 5.6545226097106935,
      "epoch": 0.6789464817206716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18844267278909682,
      "num_tokens": 13499631.0,
      "step": 6045
    },
    {
      "entropy": 5.595108938217163,
      "epoch": 0.6795080586286292,
      "grad_norm": 0.875,
      "learning_rate": 0.000496357276153527,
      "loss": 5.4763,
      "mean_token_accuracy": 0.1863000974059105,
      "num_tokens": 13511753.0,
      "step": 6050
    },
    {
      "entropy": 5.6343714714050295,
      "epoch": 0.6800696355365867,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.5425,
      "mean_token_accuracy": 0.1786164313554764,
      "num_tokens": 13524228.0,
      "step": 6055
    },
    {
      "entropy": 5.71330943107605,
      "epoch": 0.6806312124445443,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000496342871633153,
      "loss": 5.5675,
      "mean_token_accuracy": 0.1814156576991081,
      "num_tokens": 13535700.0,
      "step": 6060
    },
    {
      "entropy": 5.652857494354248,
      "epoch": 0.6811927893525018,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.5413,
      "mean_token_accuracy": 0.1782984972000122,
      "num_tokens": 13547337.0,
      "step": 6065
    },
    {
      "entropy": 5.530259466171264,
      "epoch": 0.6817543662604594,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.3943,
      "mean_token_accuracy": 0.1889803573489189,
      "num_tokens": 13558350.0,
      "step": 6070
    },
    {
      "entropy": 5.625807762145996,
      "epoch": 0.6823159431684169,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.4355,
      "mean_token_accuracy": 0.19265064746141433,
      "num_tokens": 13569230.0,
      "step": 6075
    },
    {
      "entropy": 5.684139204025269,
      "epoch": 0.6828775200763745,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.5232,
      "mean_token_accuracy": 0.18758509904146195,
      "num_tokens": 13580502.0,
      "step": 6080
    },
    {
      "entropy": 5.6358067989349365,
      "epoch": 0.683439096984332,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.4468,
      "mean_token_accuracy": 0.19119317680597306,
      "num_tokens": 13591524.0,
      "step": 6085
    },
    {
      "entropy": 5.680996322631836,
      "epoch": 0.6840006738922896,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.4668,
      "mean_token_accuracy": 0.19468809962272643,
      "num_tokens": 13602432.0,
      "step": 6090
    },
    {
      "entropy": 5.517714738845825,
      "epoch": 0.6845622508002471,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.3883,
      "mean_token_accuracy": 0.19349505603313447,
      "num_tokens": 13613052.0,
      "step": 6095
    },
    {
      "entropy": 5.555488729476929,
      "epoch": 0.6851238277082047,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.4075,
      "mean_token_accuracy": 0.1888477623462677,
      "num_tokens": 13623961.0,
      "step": 6100
    },
    {
      "entropy": 5.560924768447876,
      "epoch": 0.6856854046161622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.4059,
      "mean_token_accuracy": 0.18469873815774918,
      "num_tokens": 13635285.0,
      "step": 6105
    },
    {
      "entropy": 5.734348249435425,
      "epoch": 0.6862469815241198,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.5471,
      "mean_token_accuracy": 0.18043653815984725,
      "num_tokens": 13647720.0,
      "step": 6110
    },
    {
      "entropy": 5.568621015548706,
      "epoch": 0.6868085584320772,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.4139,
      "mean_token_accuracy": 0.18512470573186873,
      "num_tokens": 13659726.0,
      "step": 6115
    },
    {
      "entropy": 5.693536186218262,
      "epoch": 0.6873701353400348,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496255852573374,
      "loss": 5.5933,
      "mean_token_accuracy": 0.1784685119986534,
      "num_tokens": 13670332.0,
      "step": 6120
    },
    {
      "entropy": 5.738230037689209,
      "epoch": 0.6879317122479923,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1777070701122284,
      "num_tokens": 13682419.0,
      "step": 6125
    },
    {
      "entropy": 5.503112745285034,
      "epoch": 0.6884932891559499,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19355863332748413,
      "num_tokens": 13692698.0,
      "step": 6130
    },
    {
      "entropy": 5.5520916938781735,
      "epoch": 0.6890548660639074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.495,
      "mean_token_accuracy": 0.1821081295609474,
      "num_tokens": 13704791.0,
      "step": 6135
    },
    {
      "entropy": 5.623216247558593,
      "epoch": 0.689616442971865,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496226620764454,
      "loss": 5.4057,
      "mean_token_accuracy": 0.1896865651011467,
      "num_tokens": 13715860.0,
      "step": 6140
    },
    {
      "entropy": 5.609112787246704,
      "epoch": 0.6901780198798225,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496219295201788,
      "loss": 5.3917,
      "mean_token_accuracy": 0.18857248574495317,
      "num_tokens": 13727029.0,
      "step": 6145
    },
    {
      "entropy": 5.553267765045166,
      "epoch": 0.6907395967877801,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1930687963962555,
      "num_tokens": 13737168.0,
      "step": 6150
    },
    {
      "entropy": 5.6214759826660154,
      "epoch": 0.6913011736957376,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.5415,
      "mean_token_accuracy": 0.17688292115926743,
      "num_tokens": 13749828.0,
      "step": 6155
    },
    {
      "entropy": 5.554470586776733,
      "epoch": 0.6918627506036952,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.3483,
      "mean_token_accuracy": 0.1854067325592041,
      "num_tokens": 13761617.0,
      "step": 6160
    },
    {
      "entropy": 5.659129571914673,
      "epoch": 0.6924243275116527,
      "grad_norm": 0.96875,
      "learning_rate": 0.000496189922516379,
      "loss": 5.4069,
      "mean_token_accuracy": 0.19488088339567183,
      "num_tokens": 13772721.0,
      "step": 6165
    },
    {
      "entropy": 5.538276052474975,
      "epoch": 0.6929859044196103,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.3999,
      "mean_token_accuracy": 0.1897376924753189,
      "num_tokens": 13783215.0,
      "step": 6170
    },
    {
      "entropy": 5.501791381835938,
      "epoch": 0.6935474813275678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.3418,
      "mean_token_accuracy": 0.19706242233514787,
      "num_tokens": 13793496.0,
      "step": 6175
    },
    {
      "entropy": 5.496315908432007,
      "epoch": 0.6941090582355254,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496167819052389,
      "loss": 5.2717,
      "mean_token_accuracy": 0.18807439655065536,
      "num_tokens": 13802818.0,
      "step": 6180
    },
    {
      "entropy": 5.716254568099975,
      "epoch": 0.6946706351434829,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.5712,
      "mean_token_accuracy": 0.18342462927103043,
      "num_tokens": 13814449.0,
      "step": 6185
    },
    {
      "entropy": 5.597082281112671,
      "epoch": 0.6952322120514405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.5049,
      "mean_token_accuracy": 0.18432164639234544,
      "num_tokens": 13825369.0,
      "step": 6190
    },
    {
      "entropy": 5.526138877868652,
      "epoch": 0.695793788959398,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18189728111028672,
      "num_tokens": 13837992.0,
      "step": 6195
    },
    {
      "entropy": 5.627196455001831,
      "epoch": 0.6963553658673556,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.4552,
      "mean_token_accuracy": 0.18379422277212143,
      "num_tokens": 13849771.0,
      "step": 6200
    },
    {
      "entropy": 5.633212995529175,
      "epoch": 0.6969169427753131,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1869375601410866,
      "num_tokens": 13861997.0,
      "step": 6205
    },
    {
      "entropy": 5.678434991836548,
      "epoch": 0.6974785196832706,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.5273,
      "mean_token_accuracy": 0.18440470844507217,
      "num_tokens": 13872972.0,
      "step": 6210
    },
    {
      "entropy": 5.497531890869141,
      "epoch": 0.6980400965912281,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.2919,
      "mean_token_accuracy": 0.19432213604450227,
      "num_tokens": 13883139.0,
      "step": 6215
    },
    {
      "entropy": 5.601830768585205,
      "epoch": 0.6986016734991857,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.4609,
      "mean_token_accuracy": 0.19036200940608977,
      "num_tokens": 13894326.0,
      "step": 6220
    },
    {
      "entropy": 5.6338025569915775,
      "epoch": 0.6991632504071432,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.4974,
      "mean_token_accuracy": 0.18872569501399994,
      "num_tokens": 13905525.0,
      "step": 6225
    },
    {
      "entropy": 5.552472829818726,
      "epoch": 0.6997248273151008,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18733143508434297,
      "num_tokens": 13915973.0,
      "step": 6230
    },
    {
      "entropy": 5.510089588165283,
      "epoch": 0.7002864042230583,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.232,
      "mean_token_accuracy": 0.19625985473394394,
      "num_tokens": 13926380.0,
      "step": 6235
    },
    {
      "entropy": 5.610934638977051,
      "epoch": 0.7008479811310159,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18449827134609223,
      "num_tokens": 13938523.0,
      "step": 6240
    },
    {
      "entropy": 5.520618486404419,
      "epoch": 0.7014095580389734,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1942552298307419,
      "num_tokens": 13949343.0,
      "step": 6245
    },
    {
      "entropy": 5.627010774612427,
      "epoch": 0.701971134946931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.5254,
      "mean_token_accuracy": 0.18694149255752562,
      "num_tokens": 13960056.0,
      "step": 6250
    },
    {
      "entropy": 5.678987169265747,
      "epoch": 0.7025327118548885,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.4943,
      "mean_token_accuracy": 0.18654798418283464,
      "num_tokens": 13971932.0,
      "step": 6255
    },
    {
      "entropy": 5.682872104644775,
      "epoch": 0.7030942887628461,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.4997,
      "mean_token_accuracy": 0.18324526399374008,
      "num_tokens": 13981992.0,
      "step": 6260
    },
    {
      "entropy": 5.619337034225464,
      "epoch": 0.7036558656708036,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1936004102230072,
      "num_tokens": 13992215.0,
      "step": 6265
    },
    {
      "entropy": 5.560675477981567,
      "epoch": 0.7042174425787612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19779611229896546,
      "num_tokens": 14002689.0,
      "step": 6270
    },
    {
      "entropy": 5.577189302444458,
      "epoch": 0.7047790194867187,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.4727,
      "mean_token_accuracy": 0.18506722003221512,
      "num_tokens": 14014951.0,
      "step": 6275
    },
    {
      "entropy": 5.595908546447754,
      "epoch": 0.7053405963946763,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.4075,
      "mean_token_accuracy": 0.18971808403730392,
      "num_tokens": 14025502.0,
      "step": 6280
    },
    {
      "entropy": 5.648522281646729,
      "epoch": 0.7059021733026338,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.4516,
      "mean_token_accuracy": 0.19311201721429824,
      "num_tokens": 14036576.0,
      "step": 6285
    },
    {
      "entropy": 5.650336408615113,
      "epoch": 0.7064637502105914,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.4894,
      "mean_token_accuracy": 0.191058212518692,
      "num_tokens": 14046493.0,
      "step": 6290
    },
    {
      "entropy": 5.646732902526855,
      "epoch": 0.7070253271185489,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18793125301599503,
      "num_tokens": 14057297.0,
      "step": 6295
    },
    {
      "entropy": 5.533433675765991,
      "epoch": 0.7075869040265065,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.3107,
      "mean_token_accuracy": 0.19207832366228103,
      "num_tokens": 14068947.0,
      "step": 6300
    },
    {
      "entropy": 5.586352205276489,
      "epoch": 0.7081484809344639,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.4111,
      "mean_token_accuracy": 0.18904066532850267,
      "num_tokens": 14078979.0,
      "step": 6305
    },
    {
      "entropy": 5.528048133850097,
      "epoch": 0.7087100578424215,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.3411,
      "mean_token_accuracy": 0.19574270248413086,
      "num_tokens": 14089785.0,
      "step": 6310
    },
    {
      "entropy": 5.581684637069702,
      "epoch": 0.709271634750379,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.4367,
      "mean_token_accuracy": 0.18712217658758162,
      "num_tokens": 14099921.0,
      "step": 6315
    },
    {
      "entropy": 5.501917314529419,
      "epoch": 0.7098332116583366,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.3415,
      "mean_token_accuracy": 0.1878671184182167,
      "num_tokens": 14110965.0,
      "step": 6320
    },
    {
      "entropy": 5.6401293754577635,
      "epoch": 0.7103947885662941,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.4787,
      "mean_token_accuracy": 0.18706678599119186,
      "num_tokens": 14122409.0,
      "step": 6325
    },
    {
      "entropy": 5.573147106170654,
      "epoch": 0.7109563654742517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.4247,
      "mean_token_accuracy": 0.19155364781618117,
      "num_tokens": 14134017.0,
      "step": 6330
    },
    {
      "entropy": 5.6098028182983395,
      "epoch": 0.7115179423822092,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1857508659362793,
      "num_tokens": 14144843.0,
      "step": 6335
    },
    {
      "entropy": 5.611140489578247,
      "epoch": 0.7120795192901668,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.3969,
      "mean_token_accuracy": 0.18700346797704698,
      "num_tokens": 14155944.0,
      "step": 6340
    },
    {
      "entropy": 5.579473495483398,
      "epoch": 0.7126410961981243,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.468,
      "mean_token_accuracy": 0.18665511757135392,
      "num_tokens": 14167064.0,
      "step": 6345
    },
    {
      "entropy": 5.61824140548706,
      "epoch": 0.7132026731060819,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.4338,
      "mean_token_accuracy": 0.18458014130592346,
      "num_tokens": 14178583.0,
      "step": 6350
    },
    {
      "entropy": 5.508779287338257,
      "epoch": 0.7137642500140394,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.3474,
      "mean_token_accuracy": 0.19063052386045456,
      "num_tokens": 14188752.0,
      "step": 6355
    },
    {
      "entropy": 5.559374761581421,
      "epoch": 0.714325826921997,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495897635781793,
      "loss": 5.3453,
      "mean_token_accuracy": 0.19423878490924834,
      "num_tokens": 14199962.0,
      "step": 6360
    },
    {
      "entropy": 5.539549922943115,
      "epoch": 0.7148874038299545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.3737,
      "mean_token_accuracy": 0.19133674651384353,
      "num_tokens": 14209877.0,
      "step": 6365
    },
    {
      "entropy": 5.5346839904785154,
      "epoch": 0.7154489807379121,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.423,
      "mean_token_accuracy": 0.1917804390192032,
      "num_tokens": 14220672.0,
      "step": 6370
    },
    {
      "entropy": 5.685506153106689,
      "epoch": 0.7160105576458696,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.4223,
      "mean_token_accuracy": 0.19130229204893112,
      "num_tokens": 14232348.0,
      "step": 6375
    },
    {
      "entropy": 5.5768579006195065,
      "epoch": 0.7165721345538272,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.4247,
      "mean_token_accuracy": 0.18018417805433273,
      "num_tokens": 14243985.0,
      "step": 6380
    },
    {
      "entropy": 5.48021969795227,
      "epoch": 0.7171337114617847,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000495859389142044,
      "loss": 5.3074,
      "mean_token_accuracy": 0.19575837552547454,
      "num_tokens": 14253548.0,
      "step": 6385
    },
    {
      "entropy": 5.652066326141357,
      "epoch": 0.7176952883697423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495851718716373,
      "loss": 5.4576,
      "mean_token_accuracy": 0.18776362687349318,
      "num_tokens": 14264757.0,
      "step": 6390
    },
    {
      "entropy": 5.481798267364502,
      "epoch": 0.7182568652776998,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.3653,
      "mean_token_accuracy": 0.19927841126918794,
      "num_tokens": 14275946.0,
      "step": 6395
    },
    {
      "entropy": 5.527574253082276,
      "epoch": 0.7188184421856574,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18056330680847169,
      "num_tokens": 14287006.0,
      "step": 6400
    },
    {
      "entropy": 5.640402889251709,
      "epoch": 0.7193800190936148,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.4736,
      "mean_token_accuracy": 0.18276593536138536,
      "num_tokens": 14299328.0,
      "step": 6405
    },
    {
      "entropy": 5.6107786178588865,
      "epoch": 0.7199415960015724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.4509,
      "mean_token_accuracy": 0.18899955451488495,
      "num_tokens": 14310048.0,
      "step": 6410
    },
    {
      "entropy": 5.501852035522461,
      "epoch": 0.7205031729095299,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1920309081673622,
      "num_tokens": 14321107.0,
      "step": 6415
    },
    {
      "entropy": 5.537177228927613,
      "epoch": 0.7210647498174875,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.3086,
      "mean_token_accuracy": 0.1903977796435356,
      "num_tokens": 14332561.0,
      "step": 6420
    },
    {
      "entropy": 5.645201253890991,
      "epoch": 0.721626326725445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.4993,
      "mean_token_accuracy": 0.18599895387887955,
      "num_tokens": 14343238.0,
      "step": 6425
    },
    {
      "entropy": 5.581672048568725,
      "epoch": 0.7221879036334026,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.389,
      "mean_token_accuracy": 0.19121433794498444,
      "num_tokens": 14356067.0,
      "step": 6430
    },
    {
      "entropy": 5.554627084732056,
      "epoch": 0.7227494805413601,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.4258,
      "mean_token_accuracy": 0.19321310967206956,
      "num_tokens": 14367302.0,
      "step": 6435
    },
    {
      "entropy": 5.529265832901001,
      "epoch": 0.7233110574493177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.3583,
      "mean_token_accuracy": 0.19456814080476761,
      "num_tokens": 14377535.0,
      "step": 6440
    },
    {
      "entropy": 5.611742925643921,
      "epoch": 0.7238726343572752,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.4935,
      "mean_token_accuracy": 0.1833214595913887,
      "num_tokens": 14389493.0,
      "step": 6445
    },
    {
      "entropy": 5.533410596847534,
      "epoch": 0.7244342112652328,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.3967,
      "mean_token_accuracy": 0.1925879552960396,
      "num_tokens": 14401564.0,
      "step": 6450
    },
    {
      "entropy": 5.626579666137696,
      "epoch": 0.7249957881731903,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.4597,
      "mean_token_accuracy": 0.19208807051181792,
      "num_tokens": 14412123.0,
      "step": 6455
    },
    {
      "entropy": 5.586443376541138,
      "epoch": 0.7255573650811479,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.4479,
      "mean_token_accuracy": 0.19102635085582734,
      "num_tokens": 14423614.0,
      "step": 6460
    },
    {
      "entropy": 5.610803413391113,
      "epoch": 0.7261189419891054,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.4299,
      "mean_token_accuracy": 0.18605317175388336,
      "num_tokens": 14434830.0,
      "step": 6465
    },
    {
      "entropy": 5.584013938903809,
      "epoch": 0.726680518897063,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.462,
      "mean_token_accuracy": 0.1924980103969574,
      "num_tokens": 14445015.0,
      "step": 6470
    },
    {
      "entropy": 5.612082672119141,
      "epoch": 0.7272420958050205,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.5518,
      "mean_token_accuracy": 0.17747223973274232,
      "num_tokens": 14455809.0,
      "step": 6475
    },
    {
      "entropy": 5.651576089859009,
      "epoch": 0.7278036727129781,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.3192,
      "mean_token_accuracy": 0.19817821234464644,
      "num_tokens": 14466565.0,
      "step": 6480
    },
    {
      "entropy": 5.6237757205963135,
      "epoch": 0.7283652496209356,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1949547603726387,
      "num_tokens": 14477383.0,
      "step": 6485
    },
    {
      "entropy": 5.542836380004883,
      "epoch": 0.7289268265288932,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.4096,
      "mean_token_accuracy": 0.19001729041337967,
      "num_tokens": 14489968.0,
      "step": 6490
    },
    {
      "entropy": 5.600843191146851,
      "epoch": 0.7294884034368507,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.4615,
      "mean_token_accuracy": 0.18952992409467698,
      "num_tokens": 14500808.0,
      "step": 6495
    },
    {
      "entropy": 5.584274339675903,
      "epoch": 0.7300499803448082,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.4676,
      "mean_token_accuracy": 0.18737349659204483,
      "num_tokens": 14511444.0,
      "step": 6500
    },
    {
      "entropy": 5.561411428451538,
      "epoch": 0.7306115572527657,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.3655,
      "mean_token_accuracy": 0.18902986198663713,
      "num_tokens": 14522562.0,
      "step": 6505
    },
    {
      "entropy": 5.558338356018067,
      "epoch": 0.7311731341607233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.411,
      "mean_token_accuracy": 0.18875286132097244,
      "num_tokens": 14533497.0,
      "step": 6510
    },
    {
      "entropy": 5.593453598022461,
      "epoch": 0.7317347110686808,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956576733121,
      "loss": 5.443,
      "mean_token_accuracy": 0.18193353563547135,
      "num_tokens": 14545637.0,
      "step": 6515
    },
    {
      "entropy": 5.65133376121521,
      "epoch": 0.7322962879766384,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.4322,
      "mean_token_accuracy": 0.1825462520122528,
      "num_tokens": 14556826.0,
      "step": 6520
    },
    {
      "entropy": 5.531628799438477,
      "epoch": 0.7328578648845959,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495641959931013,
      "loss": 5.3825,
      "mean_token_accuracy": 0.18547695726156235,
      "num_tokens": 14567315.0,
      "step": 6525
    },
    {
      "entropy": 5.579139471054077,
      "epoch": 0.7334194417925535,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18906217068433762,
      "num_tokens": 14577685.0,
      "step": 6530
    },
    {
      "entropy": 5.582767963409424,
      "epoch": 0.733981018700511,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495626218448624,
      "loss": 5.3942,
      "mean_token_accuracy": 0.19229510724544524,
      "num_tokens": 14588885.0,
      "step": 6535
    },
    {
      "entropy": 5.626885461807251,
      "epoch": 0.7345425956084686,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19379157125949859,
      "num_tokens": 14599774.0,
      "step": 6540
    },
    {
      "entropy": 5.663081216812134,
      "epoch": 0.7351041725164261,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.4643,
      "mean_token_accuracy": 0.18054187148809434,
      "num_tokens": 14612101.0,
      "step": 6545
    },
    {
      "entropy": 5.688703489303589,
      "epoch": 0.7356657494243837,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.5365,
      "mean_token_accuracy": 0.17170834988355638,
      "num_tokens": 14622151.0,
      "step": 6550
    },
    {
      "entropy": 5.620556735992432,
      "epoch": 0.7362273263323412,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.4322,
      "mean_token_accuracy": 0.1808737874031067,
      "num_tokens": 14634631.0,
      "step": 6555
    },
    {
      "entropy": 5.5858619689941404,
      "epoch": 0.7367889032402988,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1862860843539238,
      "num_tokens": 14645429.0,
      "step": 6560
    },
    {
      "entropy": 5.545288896560669,
      "epoch": 0.7373504801482563,
      "grad_norm": 0.875,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.3848,
      "mean_token_accuracy": 0.19431884437799454,
      "num_tokens": 14657309.0,
      "step": 6565
    },
    {
      "entropy": 5.543549394607544,
      "epoch": 0.7379120570562139,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19244820922613143,
      "num_tokens": 14668578.0,
      "step": 6570
    },
    {
      "entropy": 5.615769386291504,
      "epoch": 0.7384736339641714,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.5366,
      "mean_token_accuracy": 0.18634015321731567,
      "num_tokens": 14679120.0,
      "step": 6575
    },
    {
      "entropy": 5.674855279922485,
      "epoch": 0.739035210872129,
      "grad_norm": 0.875,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18364731669425965,
      "num_tokens": 14690849.0,
      "step": 6580
    },
    {
      "entropy": 5.628752136230469,
      "epoch": 0.7395967877800865,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.4397,
      "mean_token_accuracy": 0.1857118010520935,
      "num_tokens": 14702431.0,
      "step": 6585
    },
    {
      "entropy": 5.533368539810181,
      "epoch": 0.7401583646880441,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.3543,
      "mean_token_accuracy": 0.19142401665449144,
      "num_tokens": 14714271.0,
      "step": 6590
    },
    {
      "entropy": 5.559241676330567,
      "epoch": 0.7407199415960015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.4716,
      "mean_token_accuracy": 0.18924474120140075,
      "num_tokens": 14724796.0,
      "step": 6595
    },
    {
      "entropy": 5.568958806991577,
      "epoch": 0.7412815185039591,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495523213982657,
      "loss": 5.3998,
      "mean_token_accuracy": 0.18394528925418854,
      "num_tokens": 14735331.0,
      "step": 6600
    },
    {
      "entropy": 5.6803717613220215,
      "epoch": 0.7418430954119166,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.4691,
      "mean_token_accuracy": 0.18641417622566223,
      "num_tokens": 14748595.0,
      "step": 6605
    },
    {
      "entropy": 5.623095893859864,
      "epoch": 0.7424046723198742,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.3925,
      "mean_token_accuracy": 0.18890450745821,
      "num_tokens": 14759501.0,
      "step": 6610
    },
    {
      "entropy": 5.495756006240844,
      "epoch": 0.7429662492278317,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.4428,
      "mean_token_accuracy": 0.1917462557554245,
      "num_tokens": 14769752.0,
      "step": 6615
    },
    {
      "entropy": 5.613451910018921,
      "epoch": 0.7435278261357893,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.385,
      "mean_token_accuracy": 0.18974173665046692,
      "num_tokens": 14781166.0,
      "step": 6620
    },
    {
      "entropy": 5.551000595092773,
      "epoch": 0.7440894030437468,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495483280880511,
      "loss": 5.2813,
      "mean_token_accuracy": 0.19872922003269194,
      "num_tokens": 14790536.0,
      "step": 6625
    },
    {
      "entropy": 5.629636144638061,
      "epoch": 0.7446509799517044,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495475273198244,
      "loss": 5.4981,
      "mean_token_accuracy": 0.18538137376308442,
      "num_tokens": 14801673.0,
      "step": 6630
    },
    {
      "entropy": 5.585660171508789,
      "epoch": 0.7452125568596619,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18606215715408325,
      "num_tokens": 14812345.0,
      "step": 6635
    },
    {
      "entropy": 5.7000960350036625,
      "epoch": 0.7457741337676195,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.5657,
      "mean_token_accuracy": 0.18337281048297882,
      "num_tokens": 14823620.0,
      "step": 6640
    },
    {
      "entropy": 5.614903163909912,
      "epoch": 0.746335710675577,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.3679,
      "mean_token_accuracy": 0.19380548745393752,
      "num_tokens": 14833487.0,
      "step": 6645
    },
    {
      "entropy": 5.449532794952392,
      "epoch": 0.7468972875835346,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19434978812932968,
      "num_tokens": 14844421.0,
      "step": 6650
    },
    {
      "entropy": 5.514632654190064,
      "epoch": 0.7474588644914921,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1960705801844597,
      "num_tokens": 14855525.0,
      "step": 6655
    },
    {
      "entropy": 5.670510959625244,
      "epoch": 0.7480204413994497,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.4689,
      "mean_token_accuracy": 0.18677013665437697,
      "num_tokens": 14866717.0,
      "step": 6660
    },
    {
      "entropy": 5.537822818756103,
      "epoch": 0.7485820183074072,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.3779,
      "mean_token_accuracy": 0.18497093617916108,
      "num_tokens": 14877372.0,
      "step": 6665
    },
    {
      "entropy": 5.5434918880462645,
      "epoch": 0.7491435952153648,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.4238,
      "mean_token_accuracy": 0.18816833645105363,
      "num_tokens": 14888618.0,
      "step": 6670
    },
    {
      "entropy": 5.553262758255005,
      "epoch": 0.7497051721233223,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18732796013355255,
      "num_tokens": 14900421.0,
      "step": 6675
    },
    {
      "entropy": 5.6186586856842045,
      "epoch": 0.7502667490312799,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.4144,
      "mean_token_accuracy": 0.18278902471065522,
      "num_tokens": 14910634.0,
      "step": 6680
    },
    {
      "entropy": 5.6449973583221436,
      "epoch": 0.7508283259392374,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.4734,
      "mean_token_accuracy": 0.18465643227100373,
      "num_tokens": 14921408.0,
      "step": 6685
    },
    {
      "entropy": 5.6165954113006595,
      "epoch": 0.7513899028471949,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18984519988298415,
      "num_tokens": 14933435.0,
      "step": 6690
    },
    {
      "entropy": 5.563254642486572,
      "epoch": 0.7519514797551524,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.4729,
      "mean_token_accuracy": 0.1840805932879448,
      "num_tokens": 14946488.0,
      "step": 6695
    },
    {
      "entropy": 5.594253063201904,
      "epoch": 0.75251305666311,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.3354,
      "mean_token_accuracy": 0.19079768806695938,
      "num_tokens": 14958053.0,
      "step": 6700
    },
    {
      "entropy": 5.587183475494385,
      "epoch": 0.7530746335710675,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1829656183719635,
      "num_tokens": 14968922.0,
      "step": 6705
    },
    {
      "entropy": 5.601230096817017,
      "epoch": 0.7536362104790251,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.4114,
      "mean_token_accuracy": 0.18808330446481705,
      "num_tokens": 14979541.0,
      "step": 6710
    },
    {
      "entropy": 5.621957397460937,
      "epoch": 0.7541977873869826,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.4358,
      "mean_token_accuracy": 0.18540670722723007,
      "num_tokens": 14990009.0,
      "step": 6715
    },
    {
      "entropy": 5.545533752441406,
      "epoch": 0.7547593642949402,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.4199,
      "mean_token_accuracy": 0.18657217174768448,
      "num_tokens": 15001712.0,
      "step": 6720
    },
    {
      "entropy": 5.612843799591064,
      "epoch": 0.7553209412028977,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.3852,
      "mean_token_accuracy": 0.19382928162813187,
      "num_tokens": 15012574.0,
      "step": 6725
    },
    {
      "entropy": 5.527701044082642,
      "epoch": 0.7558825181108553,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18998706042766572,
      "num_tokens": 15023034.0,
      "step": 6730
    },
    {
      "entropy": 5.641063261032104,
      "epoch": 0.7564440950188128,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.4319,
      "mean_token_accuracy": 0.1854875788092613,
      "num_tokens": 15033925.0,
      "step": 6735
    },
    {
      "entropy": 5.602529144287109,
      "epoch": 0.7570056719267704,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.4226,
      "mean_token_accuracy": 0.19354378283023835,
      "num_tokens": 15044947.0,
      "step": 6740
    },
    {
      "entropy": 5.594545364379883,
      "epoch": 0.7575672488347279,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.3386,
      "mean_token_accuracy": 0.19448180943727494,
      "num_tokens": 15054784.0,
      "step": 6745
    },
    {
      "entropy": 5.56032075881958,
      "epoch": 0.7581288257426855,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18682491332292556,
      "num_tokens": 15066272.0,
      "step": 6750
    },
    {
      "entropy": 5.614891862869262,
      "epoch": 0.758690402650643,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.494,
      "mean_token_accuracy": 0.18463293015956878,
      "num_tokens": 15079504.0,
      "step": 6755
    },
    {
      "entropy": 5.629607343673706,
      "epoch": 0.7592519795586006,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.4165,
      "mean_token_accuracy": 0.19458481371402742,
      "num_tokens": 15089623.0,
      "step": 6760
    },
    {
      "entropy": 5.5151451587677,
      "epoch": 0.7598135564665581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19032284170389174,
      "num_tokens": 15100816.0,
      "step": 6765
    },
    {
      "entropy": 5.562209224700927,
      "epoch": 0.7603751333745157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18979522883892058,
      "num_tokens": 15111697.0,
      "step": 6770
    },
    {
      "entropy": 5.630496406555176,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18143586218357086,
      "num_tokens": 15122869.0,
      "step": 6775
    },
    {
      "entropy": 5.503409147262573,
      "epoch": 0.7614982871904308,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.2944,
      "mean_token_accuracy": 0.19230962097644805,
      "num_tokens": 15134709.0,
      "step": 6780
    },
    {
      "entropy": 5.530238771438599,
      "epoch": 0.7620598640983883,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19401519298553466,
      "num_tokens": 15145319.0,
      "step": 6785
    },
    {
      "entropy": 5.566782999038696,
      "epoch": 0.7626214410063458,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495215322197552,
      "loss": 5.3706,
      "mean_token_accuracy": 0.18686714619398118,
      "num_tokens": 15156463.0,
      "step": 6790
    },
    {
      "entropy": 5.654532527923584,
      "epoch": 0.7631830179143033,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.5185,
      "mean_token_accuracy": 0.17685183137655258,
      "num_tokens": 15167618.0,
      "step": 6795
    },
    {
      "entropy": 5.626258182525635,
      "epoch": 0.7637445948222609,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.4382,
      "mean_token_accuracy": 0.18959027379751206,
      "num_tokens": 15179312.0,
      "step": 6800
    },
    {
      "entropy": 5.519511651992798,
      "epoch": 0.7643061717302184,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.3953,
      "mean_token_accuracy": 0.1968903511762619,
      "num_tokens": 15191370.0,
      "step": 6805
    },
    {
      "entropy": 5.629844331741333,
      "epoch": 0.764867748638176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.3471,
      "mean_token_accuracy": 0.19216832816600798,
      "num_tokens": 15201060.0,
      "step": 6810
    },
    {
      "entropy": 5.5226799011230465,
      "epoch": 0.7654293255461335,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19263204485177993,
      "num_tokens": 15212186.0,
      "step": 6815
    },
    {
      "entropy": 5.4873881340026855,
      "epoch": 0.7659909024540911,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.3057,
      "mean_token_accuracy": 0.19683111011981963,
      "num_tokens": 15223301.0,
      "step": 6820
    },
    {
      "entropy": 5.546998691558838,
      "epoch": 0.7665524793620486,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18811178654432298,
      "num_tokens": 15234475.0,
      "step": 6825
    },
    {
      "entropy": 5.604031801223755,
      "epoch": 0.7671140562700062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.4221,
      "mean_token_accuracy": 0.18756646066904067,
      "num_tokens": 15245877.0,
      "step": 6830
    },
    {
      "entropy": 5.665262079238891,
      "epoch": 0.7676756331779637,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.4285,
      "mean_token_accuracy": 0.1926403671503067,
      "num_tokens": 15257375.0,
      "step": 6835
    },
    {
      "entropy": 5.602505874633789,
      "epoch": 0.7682372100859213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.4335,
      "mean_token_accuracy": 0.18843176066875458,
      "num_tokens": 15268079.0,
      "step": 6840
    },
    {
      "entropy": 5.542996072769165,
      "epoch": 0.7687987869938788,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.42,
      "mean_token_accuracy": 0.18988614678382873,
      "num_tokens": 15278678.0,
      "step": 6845
    },
    {
      "entropy": 5.6731055736541744,
      "epoch": 0.7693603639018364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18146705031394958,
      "num_tokens": 15290080.0,
      "step": 6850
    },
    {
      "entropy": 5.61447925567627,
      "epoch": 0.7699219408097939,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18456478267908097,
      "num_tokens": 15300602.0,
      "step": 6855
    },
    {
      "entropy": 5.634157419204712,
      "epoch": 0.7704835177177515,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.4486,
      "mean_token_accuracy": 0.1917038306593895,
      "num_tokens": 15312823.0,
      "step": 6860
    },
    {
      "entropy": 5.660559272766113,
      "epoch": 0.771045094625709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000495090998008418,
      "loss": 5.4541,
      "mean_token_accuracy": 0.1849473848938942,
      "num_tokens": 15323870.0,
      "step": 6865
    },
    {
      "entropy": 5.52766375541687,
      "epoch": 0.7716066715336666,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000495082653656997,
      "loss": 5.3532,
      "mean_token_accuracy": 0.19443190544843675,
      "num_tokens": 15334887.0,
      "step": 6870
    },
    {
      "entropy": 5.610599040985107,
      "epoch": 0.7721682484416241,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.4543,
      "mean_token_accuracy": 0.18433938026428223,
      "num_tokens": 15345871.0,
      "step": 6875
    },
    {
      "entropy": 5.543587970733642,
      "epoch": 0.7727298253495817,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.3026,
      "mean_token_accuracy": 0.19319830238819122,
      "num_tokens": 15356530.0,
      "step": 6880
    },
    {
      "entropy": 5.569575500488281,
      "epoch": 0.7732914022575391,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.3751,
      "mean_token_accuracy": 0.1902507573366165,
      "num_tokens": 15368756.0,
      "step": 6885
    },
    {
      "entropy": 5.638578414916992,
      "epoch": 0.7738529791654967,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.4616,
      "mean_token_accuracy": 0.19190478175878525,
      "num_tokens": 15379450.0,
      "step": 6890
    },
    {
      "entropy": 5.539693164825439,
      "epoch": 0.7744145560734542,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.3849,
      "mean_token_accuracy": 0.18365437686443328,
      "num_tokens": 15389501.0,
      "step": 6895
    },
    {
      "entropy": 5.4395287990570065,
      "epoch": 0.7749761329814118,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.3728,
      "mean_token_accuracy": 0.18910327106714248,
      "num_tokens": 15400608.0,
      "step": 6900
    },
    {
      "entropy": 5.673397874832153,
      "epoch": 0.7755377098893693,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.4494,
      "mean_token_accuracy": 0.18813574463129043,
      "num_tokens": 15411709.0,
      "step": 6905
    },
    {
      "entropy": 5.555392265319824,
      "epoch": 0.7760992867973269,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.3136,
      "mean_token_accuracy": 0.19901424646377563,
      "num_tokens": 15422292.0,
      "step": 6910
    },
    {
      "entropy": 5.580694627761841,
      "epoch": 0.7766608637052844,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495007239187896,
      "loss": 5.4693,
      "mean_token_accuracy": 0.1841059848666191,
      "num_tokens": 15434439.0,
      "step": 6915
    },
    {
      "entropy": 5.565446519851685,
      "epoch": 0.777222440613242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.4167,
      "mean_token_accuracy": 0.18889338672161102,
      "num_tokens": 15445480.0,
      "step": 6920
    },
    {
      "entropy": 5.529598999023437,
      "epoch": 0.7777840175211995,
      "grad_norm": 0.875,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.3936,
      "mean_token_accuracy": 0.18663587868213655,
      "num_tokens": 15456960.0,
      "step": 6925
    },
    {
      "entropy": 5.517472505569458,
      "epoch": 0.7783455944291571,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494981974929824,
      "loss": 5.399,
      "mean_token_accuracy": 0.18664600551128388,
      "num_tokens": 15468169.0,
      "step": 6930
    },
    {
      "entropy": 5.614720344543457,
      "epoch": 0.7789071713371146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494973539501486,
      "loss": 5.3652,
      "mean_token_accuracy": 0.19006241858005524,
      "num_tokens": 15478740.0,
      "step": 6935
    },
    {
      "entropy": 5.586401128768921,
      "epoch": 0.7794687482450722,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.4592,
      "mean_token_accuracy": 0.18360558152198792,
      "num_tokens": 15489705.0,
      "step": 6940
    },
    {
      "entropy": 5.51094069480896,
      "epoch": 0.7800303251530297,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19011501520872115,
      "num_tokens": 15500549.0,
      "step": 6945
    },
    {
      "entropy": 5.667606544494629,
      "epoch": 0.7805919020609873,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1869119331240654,
      "num_tokens": 15511124.0,
      "step": 6950
    },
    {
      "entropy": 5.551062488555909,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19355154186487197,
      "num_tokens": 15521810.0,
      "step": 6955
    },
    {
      "entropy": 5.5635956764221195,
      "epoch": 0.7817150558769024,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.4051,
      "mean_token_accuracy": 0.19026406556367875,
      "num_tokens": 15533108.0,
      "step": 6960
    },
    {
      "entropy": 5.569538164138794,
      "epoch": 0.7822766327848599,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.4251,
      "mean_token_accuracy": 0.18957830220460892,
      "num_tokens": 15544841.0,
      "step": 6965
    },
    {
      "entropy": 5.59359073638916,
      "epoch": 0.7828382096928175,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.4146,
      "mean_token_accuracy": 0.1840295135974884,
      "num_tokens": 15556297.0,
      "step": 6970
    },
    {
      "entropy": 5.56764965057373,
      "epoch": 0.783399786600775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.3102,
      "mean_token_accuracy": 0.19230929166078567,
      "num_tokens": 15566607.0,
      "step": 6975
    },
    {
      "entropy": 5.6005086421966555,
      "epoch": 0.7839613635087325,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.4902,
      "mean_token_accuracy": 0.18254853338003157,
      "num_tokens": 15577435.0,
      "step": 6980
    },
    {
      "entropy": 5.591782569885254,
      "epoch": 0.78452294041669,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.3788,
      "mean_token_accuracy": 0.19169819355010986,
      "num_tokens": 15587844.0,
      "step": 6985
    },
    {
      "entropy": 5.655375719070435,
      "epoch": 0.7850845173246476,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18751610517501832,
      "num_tokens": 15599586.0,
      "step": 6990
    },
    {
      "entropy": 5.535000419616699,
      "epoch": 0.7856460942326051,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.3292,
      "mean_token_accuracy": 0.18930536806583403,
      "num_tokens": 15610838.0,
      "step": 6995
    },
    {
      "entropy": 5.5763403415679935,
      "epoch": 0.7862076711405627,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.4144,
      "mean_token_accuracy": 0.19340723901987075,
      "num_tokens": 15622652.0,
      "step": 7000
    },
    {
      "entropy": 5.664337396621704,
      "epoch": 0.7867692480485202,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.5005,
      "mean_token_accuracy": 0.18615781664848327,
      "num_tokens": 15634280.0,
      "step": 7005
    },
    {
      "entropy": 5.5180103302001955,
      "epoch": 0.7873308249564778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.2351,
      "mean_token_accuracy": 0.20067231804132463,
      "num_tokens": 15644088.0,
      "step": 7010
    },
    {
      "entropy": 5.536578178405762,
      "epoch": 0.7878924018644353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.4688,
      "mean_token_accuracy": 0.17927106618881225,
      "num_tokens": 15654968.0,
      "step": 7015
    },
    {
      "entropy": 5.570732116699219,
      "epoch": 0.7884539787723929,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.3781,
      "mean_token_accuracy": 0.19106132090091704,
      "num_tokens": 15666758.0,
      "step": 7020
    },
    {
      "entropy": 5.486448097229004,
      "epoch": 0.7890155556803504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.2279,
      "mean_token_accuracy": 0.2017139509320259,
      "num_tokens": 15677295.0,
      "step": 7025
    },
    {
      "entropy": 5.533949089050293,
      "epoch": 0.789577132588308,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1913800612092018,
      "num_tokens": 15689508.0,
      "step": 7030
    },
    {
      "entropy": 5.598876428604126,
      "epoch": 0.7901387094962655,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1893519163131714,
      "num_tokens": 15701289.0,
      "step": 7035
    },
    {
      "entropy": 5.553190088272094,
      "epoch": 0.7907002864042231,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.3444,
      "mean_token_accuracy": 0.18960873633623124,
      "num_tokens": 15711579.0,
      "step": 7040
    },
    {
      "entropy": 5.568891429901123,
      "epoch": 0.7912618633121806,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.4407,
      "mean_token_accuracy": 0.1931941717863083,
      "num_tokens": 15722629.0,
      "step": 7045
    },
    {
      "entropy": 5.58252739906311,
      "epoch": 0.7918234402201382,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.3665,
      "mean_token_accuracy": 0.18674815595149993,
      "num_tokens": 15735818.0,
      "step": 7050
    },
    {
      "entropy": 5.557722091674805,
      "epoch": 0.7923850171280957,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18566744923591613,
      "num_tokens": 15746914.0,
      "step": 7055
    },
    {
      "entropy": 5.5220029830932615,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494760378181364,
      "loss": 5.3326,
      "mean_token_accuracy": 0.18730335533618928,
      "num_tokens": 15757597.0,
      "step": 7060
    },
    {
      "entropy": 5.510581827163696,
      "epoch": 0.7935081709440108,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.2568,
      "mean_token_accuracy": 0.19909459799528123,
      "num_tokens": 15768860.0,
      "step": 7065
    },
    {
      "entropy": 5.5844197273254395,
      "epoch": 0.7940697478519684,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.3742,
      "mean_token_accuracy": 0.1869338721036911,
      "num_tokens": 15779476.0,
      "step": 7070
    },
    {
      "entropy": 5.587221097946167,
      "epoch": 0.7946313247599258,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.4448,
      "mean_token_accuracy": 0.18548859357833863,
      "num_tokens": 15792142.0,
      "step": 7075
    },
    {
      "entropy": 5.594119930267334,
      "epoch": 0.7951929016678834,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.3996,
      "mean_token_accuracy": 0.19243068397045135,
      "num_tokens": 15802307.0,
      "step": 7080
    },
    {
      "entropy": 5.540694904327393,
      "epoch": 0.7957544785758409,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.3653,
      "mean_token_accuracy": 0.18889086097478866,
      "num_tokens": 15812986.0,
      "step": 7085
    },
    {
      "entropy": 5.499142837524414,
      "epoch": 0.7963160554837985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19277723580598832,
      "num_tokens": 15823938.0,
      "step": 7090
    },
    {
      "entropy": 5.5133613586425785,
      "epoch": 0.796877632391756,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19467879682779313,
      "num_tokens": 15835813.0,
      "step": 7095
    },
    {
      "entropy": 5.598083543777466,
      "epoch": 0.7974392092997136,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494691242717217,
      "loss": 5.3805,
      "mean_token_accuracy": 0.19346053898334503,
      "num_tokens": 15846203.0,
      "step": 7100
    },
    {
      "entropy": 5.575829029083252,
      "epoch": 0.7980007862076711,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.3301,
      "mean_token_accuracy": 0.19549839198589325,
      "num_tokens": 15856919.0,
      "step": 7105
    },
    {
      "entropy": 5.565961027145386,
      "epoch": 0.7985623631156287,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494673888895125,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18444461226463318,
      "num_tokens": 15867149.0,
      "step": 7110
    },
    {
      "entropy": 5.494357347488403,
      "epoch": 0.7991239400235862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19159308671951295,
      "num_tokens": 15877229.0,
      "step": 7115
    },
    {
      "entropy": 5.577346229553223,
      "epoch": 0.7996855169315438,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.3267,
      "mean_token_accuracy": 0.1931258499622345,
      "num_tokens": 15888799.0,
      "step": 7120
    },
    {
      "entropy": 5.492810440063477,
      "epoch": 0.8002470938395013,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.3198,
      "mean_token_accuracy": 0.204921917617321,
      "num_tokens": 15900144.0,
      "step": 7125
    },
    {
      "entropy": 5.567679214477539,
      "epoch": 0.8008086707474589,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18457708209753038,
      "num_tokens": 15912030.0,
      "step": 7130
    },
    {
      "entropy": 5.5671224117279055,
      "epoch": 0.8013702476554164,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18984089195728301,
      "num_tokens": 15923606.0,
      "step": 7135
    },
    {
      "entropy": 5.546453619003296,
      "epoch": 0.801931824563374,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.362,
      "mean_token_accuracy": 0.19773187935352327,
      "num_tokens": 15935077.0,
      "step": 7140
    },
    {
      "entropy": 5.489715337753296,
      "epoch": 0.8024934014713315,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.2895,
      "mean_token_accuracy": 0.20116740167140962,
      "num_tokens": 15945331.0,
      "step": 7145
    },
    {
      "entropy": 5.523759460449218,
      "epoch": 0.8030549783792891,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18508993983268737,
      "num_tokens": 15957031.0,
      "step": 7150
    },
    {
      "entropy": 5.591307258605957,
      "epoch": 0.8036165552872466,
      "grad_norm": 0.875,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.4123,
      "mean_token_accuracy": 0.19181235432624816,
      "num_tokens": 15967545.0,
      "step": 7155
    },
    {
      "entropy": 5.711222314834595,
      "epoch": 0.8041781321952042,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.4635,
      "mean_token_accuracy": 0.1803006187081337,
      "num_tokens": 15978700.0,
      "step": 7160
    },
    {
      "entropy": 5.555808877944946,
      "epoch": 0.8047397091031617,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1942754790186882,
      "num_tokens": 15989712.0,
      "step": 7165
    },
    {
      "entropy": 5.646197652816772,
      "epoch": 0.8053012860111193,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000494569178502031,
      "loss": 5.5747,
      "mean_token_accuracy": 0.18140885829925538,
      "num_tokens": 16002380.0,
      "step": 7170
    },
    {
      "entropy": 5.6009660243988035,
      "epoch": 0.8058628629190767,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19098512679338456,
      "num_tokens": 16013749.0,
      "step": 7175
    },
    {
      "entropy": 5.590990924835205,
      "epoch": 0.8064244398270343,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.5165,
      "mean_token_accuracy": 0.1854198843240738,
      "num_tokens": 16025131.0,
      "step": 7180
    },
    {
      "entropy": 5.608829498291016,
      "epoch": 0.8069860167349918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.3787,
      "mean_token_accuracy": 0.19198863208293915,
      "num_tokens": 16037269.0,
      "step": 7185
    },
    {
      "entropy": 5.717467403411865,
      "epoch": 0.8075475936429494,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.5755,
      "mean_token_accuracy": 0.17878250628709794,
      "num_tokens": 16050992.0,
      "step": 7190
    },
    {
      "entropy": 5.53238000869751,
      "epoch": 0.8081091705509069,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.3958,
      "mean_token_accuracy": 0.18481251448392869,
      "num_tokens": 16062309.0,
      "step": 7195
    },
    {
      "entropy": 5.6282390594482425,
      "epoch": 0.8086707474588645,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.432,
      "mean_token_accuracy": 0.1872140184044838,
      "num_tokens": 16073079.0,
      "step": 7200
    },
    {
      "entropy": 5.559512567520142,
      "epoch": 0.809232324366822,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.3355,
      "mean_token_accuracy": 0.19640682488679886,
      "num_tokens": 16082885.0,
      "step": 7205
    },
    {
      "entropy": 5.460223960876465,
      "epoch": 0.8097939012747796,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19837386310100555,
      "num_tokens": 16094173.0,
      "step": 7210
    },
    {
      "entropy": 5.522316980361938,
      "epoch": 0.8103554781827371,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494489985015675,
      "loss": 5.3435,
      "mean_token_accuracy": 0.19227355867624282,
      "num_tokens": 16106351.0,
      "step": 7215
    },
    {
      "entropy": 5.4737783908844,
      "epoch": 0.8109170550906947,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.3255,
      "mean_token_accuracy": 0.19104305356740953,
      "num_tokens": 16116892.0,
      "step": 7220
    },
    {
      "entropy": 5.637874269485474,
      "epoch": 0.8114786319986522,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.474,
      "mean_token_accuracy": 0.17940621078014374,
      "num_tokens": 16128316.0,
      "step": 7225
    },
    {
      "entropy": 5.546368074417114,
      "epoch": 0.8120402089066098,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.3077,
      "mean_token_accuracy": 0.19091015458106994,
      "num_tokens": 16139245.0,
      "step": 7230
    },
    {
      "entropy": 5.516021060943603,
      "epoch": 0.8126017858145673,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049445460619367,
      "loss": 5.362,
      "mean_token_accuracy": 0.1917641818523407,
      "num_tokens": 16150893.0,
      "step": 7235
    },
    {
      "entropy": 5.639879369735718,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.4799,
      "mean_token_accuracy": 0.17413984686136247,
      "num_tokens": 16163430.0,
      "step": 7240
    },
    {
      "entropy": 5.604622459411621,
      "epoch": 0.8137249396304824,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19260778725147248,
      "num_tokens": 16173954.0,
      "step": 7245
    },
    {
      "entropy": 5.59013991355896,
      "epoch": 0.81428651653844,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18455109894275665,
      "num_tokens": 16185491.0,
      "step": 7250
    },
    {
      "entropy": 5.5030594825744625,
      "epoch": 0.8148480934463975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.2788,
      "mean_token_accuracy": 0.20222099125385284,
      "num_tokens": 16195301.0,
      "step": 7255
    },
    {
      "entropy": 5.57300214767456,
      "epoch": 0.8154096703543551,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.4055,
      "mean_token_accuracy": 0.18483036756515503,
      "num_tokens": 16206620.0,
      "step": 7260
    },
    {
      "entropy": 5.58026647567749,
      "epoch": 0.8159712472623126,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.4952,
      "mean_token_accuracy": 0.1856081873178482,
      "num_tokens": 16218693.0,
      "step": 7265
    },
    {
      "entropy": 5.61599326133728,
      "epoch": 0.8165328241702701,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.3707,
      "mean_token_accuracy": 0.19189411401748657,
      "num_tokens": 16229124.0,
      "step": 7270
    },
    {
      "entropy": 5.516272020339966,
      "epoch": 0.8170944010782276,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.3427,
      "mean_token_accuracy": 0.19527129530906678,
      "num_tokens": 16239999.0,
      "step": 7275
    },
    {
      "entropy": 5.532791233062744,
      "epoch": 0.8176559779861852,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.4415,
      "mean_token_accuracy": 0.19038496315479278,
      "num_tokens": 16251027.0,
      "step": 7280
    },
    {
      "entropy": 5.6169427871704105,
      "epoch": 0.8182175548941427,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.3817,
      "mean_token_accuracy": 0.19348272532224656,
      "num_tokens": 16262924.0,
      "step": 7285
    },
    {
      "entropy": 5.658203744888306,
      "epoch": 0.8187791318021003,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.4789,
      "mean_token_accuracy": 0.1869910940527916,
      "num_tokens": 16274481.0,
      "step": 7290
    },
    {
      "entropy": 5.5553487777709964,
      "epoch": 0.8193407087100578,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.401,
      "mean_token_accuracy": 0.19489087015390397,
      "num_tokens": 16285214.0,
      "step": 7295
    },
    {
      "entropy": 5.581579065322876,
      "epoch": 0.8199022856180154,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.4822,
      "mean_token_accuracy": 0.18492209166288376,
      "num_tokens": 16296077.0,
      "step": 7300
    },
    {
      "entropy": 5.627924299240112,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1899802416563034,
      "num_tokens": 16307514.0,
      "step": 7305
    },
    {
      "entropy": 5.6731773853302006,
      "epoch": 0.8210254394339305,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.3923,
      "mean_token_accuracy": 0.19327206313610076,
      "num_tokens": 16318734.0,
      "step": 7310
    },
    {
      "entropy": 5.503451728820801,
      "epoch": 0.821587016341888,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494311973083621,
      "loss": 5.4258,
      "mean_token_accuracy": 0.189380219578743,
      "num_tokens": 16329057.0,
      "step": 7315
    },
    {
      "entropy": 5.535038614273072,
      "epoch": 0.8221485932498456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1946206659078598,
      "num_tokens": 16339646.0,
      "step": 7320
    },
    {
      "entropy": 5.632025575637817,
      "epoch": 0.8227101701578031,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.4259,
      "mean_token_accuracy": 0.186196868121624,
      "num_tokens": 16349783.0,
      "step": 7325
    },
    {
      "entropy": 5.6155784130096436,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.3407,
      "mean_token_accuracy": 0.19420970380306243,
      "num_tokens": 16361078.0,
      "step": 7330
    },
    {
      "entropy": 5.520710182189942,
      "epoch": 0.8238333239737182,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.3688,
      "mean_token_accuracy": 0.19222986698150635,
      "num_tokens": 16371308.0,
      "step": 7335
    },
    {
      "entropy": 5.539703369140625,
      "epoch": 0.8243949008816758,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.4004,
      "mean_token_accuracy": 0.19265177547931672,
      "num_tokens": 16382682.0,
      "step": 7340
    },
    {
      "entropy": 5.6106054306030275,
      "epoch": 0.8249564777896333,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.4254,
      "mean_token_accuracy": 0.1876666098833084,
      "num_tokens": 16393715.0,
      "step": 7345
    },
    {
      "entropy": 5.497316169738769,
      "epoch": 0.8255180546975909,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1883910521864891,
      "num_tokens": 16406332.0,
      "step": 7350
    },
    {
      "entropy": 5.529266405105591,
      "epoch": 0.8260796316055484,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.3417,
      "mean_token_accuracy": 0.19201942086219786,
      "num_tokens": 16417702.0,
      "step": 7355
    },
    {
      "entropy": 5.577368831634521,
      "epoch": 0.826641208513506,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.3874,
      "mean_token_accuracy": 0.1885086327791214,
      "num_tokens": 16428497.0,
      "step": 7360
    },
    {
      "entropy": 5.578837203979492,
      "epoch": 0.8272027854214634,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.3872,
      "mean_token_accuracy": 0.19048845916986465,
      "num_tokens": 16440637.0,
      "step": 7365
    },
    {
      "entropy": 5.5637633323669435,
      "epoch": 0.827764362329421,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.4059,
      "mean_token_accuracy": 0.18300192058086395,
      "num_tokens": 16453213.0,
      "step": 7370
    },
    {
      "entropy": 5.569536638259888,
      "epoch": 0.8283259392373785,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.4173,
      "mean_token_accuracy": 0.19181468039751054,
      "num_tokens": 16464345.0,
      "step": 7375
    },
    {
      "entropy": 5.545488023757935,
      "epoch": 0.8288875161453361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.305,
      "mean_token_accuracy": 0.19415025264024735,
      "num_tokens": 16475399.0,
      "step": 7380
    },
    {
      "entropy": 5.50160665512085,
      "epoch": 0.8294490930532936,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.2724,
      "mean_token_accuracy": 0.19475990235805513,
      "num_tokens": 16485439.0,
      "step": 7385
    },
    {
      "entropy": 5.531022214889527,
      "epoch": 0.8300106699612512,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.2743,
      "mean_token_accuracy": 0.19617179185152053,
      "num_tokens": 16496765.0,
      "step": 7390
    },
    {
      "entropy": 5.524594068527222,
      "epoch": 0.8305722468692087,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.3263,
      "mean_token_accuracy": 0.19536734521389007,
      "num_tokens": 16509199.0,
      "step": 7395
    },
    {
      "entropy": 5.547530031204223,
      "epoch": 0.8311338237771663,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.4733,
      "mean_token_accuracy": 0.1877291664481163,
      "num_tokens": 16520073.0,
      "step": 7400
    },
    {
      "entropy": 5.482045841217041,
      "epoch": 0.8316954006851238,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18801986277103425,
      "num_tokens": 16530919.0,
      "step": 7405
    },
    {
      "entropy": 5.571953630447387,
      "epoch": 0.8322569775930814,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.3761,
      "mean_token_accuracy": 0.18720916360616685,
      "num_tokens": 16542591.0,
      "step": 7410
    },
    {
      "entropy": 5.616376972198486,
      "epoch": 0.8328185545010389,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.3747,
      "mean_token_accuracy": 0.19256722331047058,
      "num_tokens": 16554984.0,
      "step": 7415
    },
    {
      "entropy": 5.497436285018921,
      "epoch": 0.8333801314089965,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494122054427151,
      "loss": 5.2769,
      "mean_token_accuracy": 0.19829754382371903,
      "num_tokens": 16566598.0,
      "step": 7420
    },
    {
      "entropy": 5.653843402862549,
      "epoch": 0.833941708316954,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.4238,
      "mean_token_accuracy": 0.18555232882499695,
      "num_tokens": 16577868.0,
      "step": 7425
    },
    {
      "entropy": 5.555824851989746,
      "epoch": 0.8345032852249116,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.3763,
      "mean_token_accuracy": 0.1900017738342285,
      "num_tokens": 16587891.0,
      "step": 7430
    },
    {
      "entropy": 5.5200104236602785,
      "epoch": 0.8350648621328691,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000494094671958544,
      "loss": 5.315,
      "mean_token_accuracy": 0.19897076189517976,
      "num_tokens": 16599924.0,
      "step": 7435
    },
    {
      "entropy": 5.509287738800049,
      "epoch": 0.8356264390408267,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.2603,
      "mean_token_accuracy": 0.19610760509967803,
      "num_tokens": 16609958.0,
      "step": 7440
    },
    {
      "entropy": 5.528403997421265,
      "epoch": 0.8361880159487842,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18935056328773497,
      "num_tokens": 16621159.0,
      "step": 7445
    },
    {
      "entropy": 5.6253071308135985,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.4907,
      "mean_token_accuracy": 0.18785437941551208,
      "num_tokens": 16632263.0,
      "step": 7450
    },
    {
      "entropy": 5.551202583312988,
      "epoch": 0.8373111697646993,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.3783,
      "mean_token_accuracy": 0.18347710520029067,
      "num_tokens": 16644117.0,
      "step": 7455
    },
    {
      "entropy": 5.457671022415161,
      "epoch": 0.8378727466726568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.2989,
      "mean_token_accuracy": 0.19613440930843354,
      "num_tokens": 16654529.0,
      "step": 7460
    },
    {
      "entropy": 5.52949743270874,
      "epoch": 0.8384343235806143,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.3147,
      "mean_token_accuracy": 0.19463446587324143,
      "num_tokens": 16666191.0,
      "step": 7465
    },
    {
      "entropy": 5.594643068313599,
      "epoch": 0.8389959004885719,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.2902,
      "mean_token_accuracy": 0.19543753266334535,
      "num_tokens": 16677252.0,
      "step": 7470
    },
    {
      "entropy": 5.479014873504639,
      "epoch": 0.8395574773965294,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494021345117024,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19304683804512024,
      "num_tokens": 16689487.0,
      "step": 7475
    },
    {
      "entropy": 5.508919811248779,
      "epoch": 0.840119054304487,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.2432,
      "mean_token_accuracy": 0.19911739677190782,
      "num_tokens": 16699955.0,
      "step": 7480
    },
    {
      "entropy": 5.438897275924683,
      "epoch": 0.8406806312124445,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.3046,
      "mean_token_accuracy": 0.18741327077150344,
      "num_tokens": 16711071.0,
      "step": 7485
    },
    {
      "entropy": 5.456135511398315,
      "epoch": 0.8412422081204021,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.3145,
      "mean_token_accuracy": 0.19677483588457106,
      "num_tokens": 16722115.0,
      "step": 7490
    },
    {
      "entropy": 5.522494697570801,
      "epoch": 0.8418037850283596,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18598861545324324,
      "num_tokens": 16733146.0,
      "step": 7495
    },
    {
      "entropy": 5.5032275199890135,
      "epoch": 0.8423653619363172,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.3271,
      "mean_token_accuracy": 0.19021343886852266,
      "num_tokens": 16745543.0,
      "step": 7500
    },
    {
      "entropy": 5.584363126754761,
      "epoch": 0.8429269388442747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.3835,
      "mean_token_accuracy": 0.19021010845899583,
      "num_tokens": 16755782.0,
      "step": 7505
    },
    {
      "entropy": 5.616142225265503,
      "epoch": 0.8434885157522323,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.4137,
      "mean_token_accuracy": 0.1914801388978958,
      "num_tokens": 16767509.0,
      "step": 7510
    },
    {
      "entropy": 5.47909951210022,
      "epoch": 0.8440500926601898,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19450779855251313,
      "num_tokens": 16779257.0,
      "step": 7515
    },
    {
      "entropy": 5.475256538391113,
      "epoch": 0.8446116695681474,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.3186,
      "mean_token_accuracy": 0.194671568274498,
      "num_tokens": 16790769.0,
      "step": 7520
    },
    {
      "entropy": 5.5638734817504885,
      "epoch": 0.8451732464761049,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.3202,
      "mean_token_accuracy": 0.19284795671701432,
      "num_tokens": 16800054.0,
      "step": 7525
    },
    {
      "entropy": 5.606045198440552,
      "epoch": 0.8457348233840625,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18594104498624803,
      "num_tokens": 16811485.0,
      "step": 7530
    },
    {
      "entropy": 5.660714817047119,
      "epoch": 0.84629640029202,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.4907,
      "mean_token_accuracy": 0.18876438289880754,
      "num_tokens": 16822683.0,
      "step": 7535
    },
    {
      "entropy": 5.589832353591919,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.3505,
      "mean_token_accuracy": 0.19204530268907546,
      "num_tokens": 16834200.0,
      "step": 7540
    },
    {
      "entropy": 5.569216537475586,
      "epoch": 0.8474195541079351,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18771866112947463,
      "num_tokens": 16846399.0,
      "step": 7545
    },
    {
      "entropy": 5.695494747161865,
      "epoch": 0.8479811310158927,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.5372,
      "mean_token_accuracy": 0.18349024653434753,
      "num_tokens": 16858322.0,
      "step": 7550
    },
    {
      "entropy": 5.572900295257568,
      "epoch": 0.8485427079238502,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.3639,
      "mean_token_accuracy": 0.19737725853919982,
      "num_tokens": 16869537.0,
      "step": 7555
    },
    {
      "entropy": 5.6444206714630125,
      "epoch": 0.8491042848318077,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.404,
      "mean_token_accuracy": 0.1946091189980507,
      "num_tokens": 16881566.0,
      "step": 7560
    },
    {
      "entropy": 5.507593297958374,
      "epoch": 0.8496658617397652,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.3286,
      "mean_token_accuracy": 0.19603571444749832,
      "num_tokens": 16892959.0,
      "step": 7565
    },
    {
      "entropy": 5.56847357749939,
      "epoch": 0.8502274386477228,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18607113510370255,
      "num_tokens": 16904809.0,
      "step": 7570
    },
    {
      "entropy": 5.601382875442505,
      "epoch": 0.8507890155556803,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.408,
      "mean_token_accuracy": 0.18271522223949432,
      "num_tokens": 16916969.0,
      "step": 7575
    },
    {
      "entropy": 5.498878240585327,
      "epoch": 0.8513505924636379,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.2891,
      "mean_token_accuracy": 0.19846532344818116,
      "num_tokens": 16928253.0,
      "step": 7580
    },
    {
      "entropy": 5.496449708938599,
      "epoch": 0.8519121693715954,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.3451,
      "mean_token_accuracy": 0.18728629499673843,
      "num_tokens": 16939875.0,
      "step": 7585
    },
    {
      "entropy": 5.6555074691772464,
      "epoch": 0.852473746279553,
      "grad_norm": 0.875,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.4755,
      "mean_token_accuracy": 0.18947706520557403,
      "num_tokens": 16950173.0,
      "step": 7590
    },
    {
      "entropy": 5.526337766647339,
      "epoch": 0.8530353231875105,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.3815,
      "mean_token_accuracy": 0.19198303520679474,
      "num_tokens": 16960933.0,
      "step": 7595
    },
    {
      "entropy": 5.5377367496490475,
      "epoch": 0.8535969000954681,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1913389340043068,
      "num_tokens": 16972284.0,
      "step": 7600
    },
    {
      "entropy": 5.545716142654419,
      "epoch": 0.8541584770034256,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19894578456878662,
      "num_tokens": 16983541.0,
      "step": 7605
    },
    {
      "entropy": 5.577863788604736,
      "epoch": 0.8547200539113832,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049377057282307,
      "loss": 5.4487,
      "mean_token_accuracy": 0.18269122242927552,
      "num_tokens": 16993773.0,
      "step": 7610
    },
    {
      "entropy": 5.650778532028198,
      "epoch": 0.8552816308193407,
      "grad_norm": 0.90625,
      "learning_rate": 0.000493761187403909,
      "loss": 5.4239,
      "mean_token_accuracy": 0.19039395302534104,
      "num_tokens": 17005287.0,
      "step": 7615
    },
    {
      "entropy": 5.564861488342285,
      "epoch": 0.8558432077272983,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493751795019307,
      "loss": 5.3596,
      "mean_token_accuracy": 0.1928644984960556,
      "num_tokens": 17015664.0,
      "step": 7620
    },
    {
      "entropy": 5.50378007888794,
      "epoch": 0.8564047846352558,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.2619,
      "mean_token_accuracy": 0.19736520051956177,
      "num_tokens": 17026876.0,
      "step": 7625
    },
    {
      "entropy": 5.4976475715637205,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.3111,
      "mean_token_accuracy": 0.2013237327337265,
      "num_tokens": 17037321.0,
      "step": 7630
    },
    {
      "entropy": 5.566529321670532,
      "epoch": 0.8575279384511709,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.3946,
      "mean_token_accuracy": 0.1911442130804062,
      "num_tokens": 17047506.0,
      "step": 7635
    },
    {
      "entropy": 5.622412395477295,
      "epoch": 0.8580895153591285,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.4374,
      "mean_token_accuracy": 0.19152735471725463,
      "num_tokens": 17059710.0,
      "step": 7640
    },
    {
      "entropy": 5.622036266326904,
      "epoch": 0.858651092267086,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.4495,
      "mean_token_accuracy": 0.18276994824409484,
      "num_tokens": 17070956.0,
      "step": 7645
    },
    {
      "entropy": 5.604185295104981,
      "epoch": 0.8592126691750436,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.4129,
      "mean_token_accuracy": 0.1950125738978386,
      "num_tokens": 17081404.0,
      "step": 7650
    },
    {
      "entropy": 5.657976388931274,
      "epoch": 0.859774246083001,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.4246,
      "mean_token_accuracy": 0.19007168859243392,
      "num_tokens": 17092692.0,
      "step": 7655
    },
    {
      "entropy": 5.5072746753692625,
      "epoch": 0.8603358229909586,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493676405222546,
      "loss": 5.2808,
      "mean_token_accuracy": 0.19722734838724137,
      "num_tokens": 17103730.0,
      "step": 7660
    },
    {
      "entropy": 5.5627227306365965,
      "epoch": 0.8608973998989161,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19256850630044936,
      "num_tokens": 17115086.0,
      "step": 7665
    },
    {
      "entropy": 5.582569026947022,
      "epoch": 0.8614589768068737,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.3511,
      "mean_token_accuracy": 0.19334861189126967,
      "num_tokens": 17125928.0,
      "step": 7670
    },
    {
      "entropy": 5.566898345947266,
      "epoch": 0.8620205537148312,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.3867,
      "mean_token_accuracy": 0.1884917438030243,
      "num_tokens": 17137328.0,
      "step": 7675
    },
    {
      "entropy": 5.546287584304809,
      "epoch": 0.8625821306227888,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1915384888648987,
      "num_tokens": 17148274.0,
      "step": 7680
    },
    {
      "entropy": 5.563704538345337,
      "epoch": 0.8631437075307463,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.4054,
      "mean_token_accuracy": 0.19138316810131073,
      "num_tokens": 17158834.0,
      "step": 7685
    },
    {
      "entropy": 5.554512214660645,
      "epoch": 0.8637052844387039,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.2634,
      "mean_token_accuracy": 0.19505443722009658,
      "num_tokens": 17169276.0,
      "step": 7690
    },
    {
      "entropy": 5.532989406585694,
      "epoch": 0.8642668613466614,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19803969264030458,
      "num_tokens": 17180642.0,
      "step": 7695
    },
    {
      "entropy": 5.512448787689209,
      "epoch": 0.864828438254619,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936005698104,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19688139855861664,
      "num_tokens": 17192477.0,
      "step": 7700
    },
    {
      "entropy": 5.617452144622803,
      "epoch": 0.8653900151625765,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.5085,
      "mean_token_accuracy": 0.18323329389095305,
      "num_tokens": 17202653.0,
      "step": 7705
    },
    {
      "entropy": 5.471633195877075,
      "epoch": 0.8659515920705341,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.242,
      "mean_token_accuracy": 0.20044415444135666,
      "num_tokens": 17212607.0,
      "step": 7710
    },
    {
      "entropy": 5.529920816421509,
      "epoch": 0.8665131689784916,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493572016677141,
      "loss": 5.3722,
      "mean_token_accuracy": 0.1968028724193573,
      "num_tokens": 17223701.0,
      "step": 7715
    },
    {
      "entropy": 5.585971736907959,
      "epoch": 0.8670747458864492,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.284,
      "mean_token_accuracy": 0.19454400092363358,
      "num_tokens": 17234878.0,
      "step": 7720
    },
    {
      "entropy": 5.520610523223877,
      "epoch": 0.8676363227944067,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.3589,
      "mean_token_accuracy": 0.194435778260231,
      "num_tokens": 17245813.0,
      "step": 7725
    },
    {
      "entropy": 5.546721935272217,
      "epoch": 0.8681978997023643,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.4326,
      "mean_token_accuracy": 0.1897881507873535,
      "num_tokens": 17257378.0,
      "step": 7730
    },
    {
      "entropy": 5.504445219039917,
      "epoch": 0.8687594766103218,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.3711,
      "mean_token_accuracy": 0.19056434333324432,
      "num_tokens": 17269108.0,
      "step": 7735
    },
    {
      "entropy": 5.590135860443115,
      "epoch": 0.8693210535182794,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.3434,
      "mean_token_accuracy": 0.1851867824792862,
      "num_tokens": 17280643.0,
      "step": 7740
    },
    {
      "entropy": 5.640087127685547,
      "epoch": 0.8698826304262369,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.4036,
      "mean_token_accuracy": 0.19341604858636857,
      "num_tokens": 17291236.0,
      "step": 7745
    },
    {
      "entropy": 5.613070297241211,
      "epoch": 0.8704442073341944,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.4812,
      "mean_token_accuracy": 0.18069635033607484,
      "num_tokens": 17303547.0,
      "step": 7750
    },
    {
      "entropy": 5.664354658126831,
      "epoch": 0.8710057842421519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.4351,
      "mean_token_accuracy": 0.1866860345005989,
      "num_tokens": 17314803.0,
      "step": 7755
    },
    {
      "entropy": 5.535034561157227,
      "epoch": 0.8715673611501095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.3207,
      "mean_token_accuracy": 0.19055616706609727,
      "num_tokens": 17326009.0,
      "step": 7760
    },
    {
      "entropy": 5.604511260986328,
      "epoch": 0.872128938058067,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.3281,
      "mean_token_accuracy": 0.1981400355696678,
      "num_tokens": 17337078.0,
      "step": 7765
    },
    {
      "entropy": 5.571507358551026,
      "epoch": 0.8726905149660246,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.4377,
      "mean_token_accuracy": 0.18731588870286942,
      "num_tokens": 17349266.0,
      "step": 7770
    },
    {
      "entropy": 5.407146739959717,
      "epoch": 0.8732520918739821,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.2525,
      "mean_token_accuracy": 0.19759435057640076,
      "num_tokens": 17360327.0,
      "step": 7775
    },
    {
      "entropy": 5.662004566192627,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.4077,
      "mean_token_accuracy": 0.1832973003387451,
      "num_tokens": 17372727.0,
      "step": 7780
    },
    {
      "entropy": 5.679233121871948,
      "epoch": 0.8743752456898972,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.4572,
      "mean_token_accuracy": 0.18479983061552047,
      "num_tokens": 17384160.0,
      "step": 7785
    },
    {
      "entropy": 5.540898561477661,
      "epoch": 0.8749368225978548,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.3887,
      "mean_token_accuracy": 0.1870481178164482,
      "num_tokens": 17396009.0,
      "step": 7790
    },
    {
      "entropy": 5.556040906906128,
      "epoch": 0.8754983995058123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.3837,
      "mean_token_accuracy": 0.191018246114254,
      "num_tokens": 17406903.0,
      "step": 7795
    },
    {
      "entropy": 5.541416358947754,
      "epoch": 0.8760599764137699,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.2492,
      "mean_token_accuracy": 0.1982877552509308,
      "num_tokens": 17417694.0,
      "step": 7800
    },
    {
      "entropy": 5.528493452072143,
      "epoch": 0.8766215533217274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.4868,
      "mean_token_accuracy": 0.19488344490528106,
      "num_tokens": 17428731.0,
      "step": 7805
    },
    {
      "entropy": 5.535378360748291,
      "epoch": 0.877183130229685,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.2347,
      "mean_token_accuracy": 0.19633994847536088,
      "num_tokens": 17439242.0,
      "step": 7810
    },
    {
      "entropy": 5.595781755447388,
      "epoch": 0.8777447071376425,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.3481,
      "mean_token_accuracy": 0.19086516946554183,
      "num_tokens": 17449980.0,
      "step": 7815
    },
    {
      "entropy": 5.615504312515259,
      "epoch": 0.8783062840456001,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.4365,
      "mean_token_accuracy": 0.18459131419658661,
      "num_tokens": 17464140.0,
      "step": 7820
    },
    {
      "entropy": 5.648376178741455,
      "epoch": 0.8788678609535576,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.4368,
      "mean_token_accuracy": 0.188435660302639,
      "num_tokens": 17476246.0,
      "step": 7825
    },
    {
      "entropy": 5.607189416885376,
      "epoch": 0.8794294378615152,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.4536,
      "mean_token_accuracy": 0.18098224550485612,
      "num_tokens": 17488169.0,
      "step": 7830
    },
    {
      "entropy": 5.549782371520996,
      "epoch": 0.8799910147694727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.2803,
      "mean_token_accuracy": 0.1958876684308052,
      "num_tokens": 17498465.0,
      "step": 7835
    },
    {
      "entropy": 5.58019061088562,
      "epoch": 0.8805525916774303,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.2727,
      "mean_token_accuracy": 0.20272148996591569,
      "num_tokens": 17509217.0,
      "step": 7840
    },
    {
      "entropy": 5.491253900527954,
      "epoch": 0.8811141685853877,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.3169,
      "mean_token_accuracy": 0.1963854193687439,
      "num_tokens": 17519851.0,
      "step": 7845
    },
    {
      "entropy": 5.530531406402588,
      "epoch": 0.8816757454933453,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.3738,
      "mean_token_accuracy": 0.19348302334547043,
      "num_tokens": 17532144.0,
      "step": 7850
    },
    {
      "entropy": 5.537290430068969,
      "epoch": 0.8822373224013028,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19952542930841446,
      "num_tokens": 17542929.0,
      "step": 7855
    },
    {
      "entropy": 5.651939535140992,
      "epoch": 0.8827988993092604,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.4744,
      "mean_token_accuracy": 0.18057009428739548,
      "num_tokens": 17554931.0,
      "step": 7860
    },
    {
      "entropy": 5.538858652114868,
      "epoch": 0.8833604762172179,
      "grad_norm": 0.875,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.2751,
      "mean_token_accuracy": 0.19320368617773057,
      "num_tokens": 17566026.0,
      "step": 7865
    },
    {
      "entropy": 5.521408987045288,
      "epoch": 0.8839220531251755,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049327330152104,
      "loss": 5.373,
      "mean_token_accuracy": 0.19261322170495987,
      "num_tokens": 17578396.0,
      "step": 7870
    },
    {
      "entropy": 5.568428182601929,
      "epoch": 0.884483630033133,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.3593,
      "mean_token_accuracy": 0.19598565846681595,
      "num_tokens": 17590559.0,
      "step": 7875
    },
    {
      "entropy": 5.5688090324401855,
      "epoch": 0.8850452069410906,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.3374,
      "mean_token_accuracy": 0.1932360514998436,
      "num_tokens": 17600833.0,
      "step": 7880
    },
    {
      "entropy": 5.483620738983154,
      "epoch": 0.8856067838490481,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.3213,
      "mean_token_accuracy": 0.19447025656700134,
      "num_tokens": 17612166.0,
      "step": 7885
    },
    {
      "entropy": 5.563448143005371,
      "epoch": 0.8861683607570057,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.4351,
      "mean_token_accuracy": 0.1893043413758278,
      "num_tokens": 17623070.0,
      "step": 7890
    },
    {
      "entropy": 5.639851903915405,
      "epoch": 0.8867299376649632,
      "grad_norm": 0.90625,
      "learning_rate": 0.000493224495924948,
      "loss": 5.4971,
      "mean_token_accuracy": 0.18729420006275177,
      "num_tokens": 17633319.0,
      "step": 7895
    },
    {
      "entropy": 5.564702129364013,
      "epoch": 0.8872915145729208,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1927963376045227,
      "num_tokens": 17643818.0,
      "step": 7900
    },
    {
      "entropy": 5.506327438354492,
      "epoch": 0.8878530914808783,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.3069,
      "mean_token_accuracy": 0.19169542342424392,
      "num_tokens": 17655665.0,
      "step": 7905
    },
    {
      "entropy": 5.478302431106568,
      "epoch": 0.8884146683888359,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.2333,
      "mean_token_accuracy": 0.20336542427539825,
      "num_tokens": 17666200.0,
      "step": 7910
    },
    {
      "entropy": 5.510001182556152,
      "epoch": 0.8889762452967934,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.4214,
      "mean_token_accuracy": 0.1845197170972824,
      "num_tokens": 17677299.0,
      "step": 7915
    },
    {
      "entropy": 5.589610528945923,
      "epoch": 0.889537822204751,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.2971,
      "mean_token_accuracy": 0.2044074386358261,
      "num_tokens": 17687870.0,
      "step": 7920
    },
    {
      "entropy": 5.507078218460083,
      "epoch": 0.8900993991127085,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19066539108753205,
      "num_tokens": 17698853.0,
      "step": 7925
    },
    {
      "entropy": 5.532127475738525,
      "epoch": 0.8906609760206661,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.3702,
      "mean_token_accuracy": 0.19769250750541686,
      "num_tokens": 17709897.0,
      "step": 7930
    },
    {
      "entropy": 5.617356014251709,
      "epoch": 0.8912225529286236,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.3607,
      "mean_token_accuracy": 0.19086020737886428,
      "num_tokens": 17720974.0,
      "step": 7935
    },
    {
      "entropy": 5.658880043029785,
      "epoch": 0.8917841298365812,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.5198,
      "mean_token_accuracy": 0.18324090242385865,
      "num_tokens": 17734449.0,
      "step": 7940
    },
    {
      "entropy": 5.440229034423828,
      "epoch": 0.8923457067445386,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.2542,
      "mean_token_accuracy": 0.19596020728349686,
      "num_tokens": 17744735.0,
      "step": 7945
    },
    {
      "entropy": 5.510704803466797,
      "epoch": 0.8929072836524962,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.276,
      "mean_token_accuracy": 0.1957583025097847,
      "num_tokens": 17755641.0,
      "step": 7950
    },
    {
      "entropy": 5.556438302993774,
      "epoch": 0.8934688605604537,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.3873,
      "mean_token_accuracy": 0.1877841904759407,
      "num_tokens": 17767621.0,
      "step": 7955
    },
    {
      "entropy": 5.686891555786133,
      "epoch": 0.8940304374684113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.4822,
      "mean_token_accuracy": 0.1870819538831711,
      "num_tokens": 17779507.0,
      "step": 7960
    },
    {
      "entropy": 5.555348491668701,
      "epoch": 0.8945920143763688,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1898730978369713,
      "num_tokens": 17790371.0,
      "step": 7965
    },
    {
      "entropy": 5.520899438858033,
      "epoch": 0.8951535912843264,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.3338,
      "mean_token_accuracy": 0.19737217873334884,
      "num_tokens": 17800903.0,
      "step": 7970
    },
    {
      "entropy": 5.628668117523193,
      "epoch": 0.8957151681922839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.5319,
      "mean_token_accuracy": 0.17857561707496644,
      "num_tokens": 17812053.0,
      "step": 7975
    },
    {
      "entropy": 5.528358459472656,
      "epoch": 0.8962767451002415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.2455,
      "mean_token_accuracy": 0.19843535572290422,
      "num_tokens": 17822113.0,
      "step": 7980
    },
    {
      "entropy": 5.516593885421753,
      "epoch": 0.896838322008199,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19497286677360534,
      "num_tokens": 17832741.0,
      "step": 7985
    },
    {
      "entropy": 5.3975913524627686,
      "epoch": 0.8973998989161566,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.226,
      "mean_token_accuracy": 0.19109468013048173,
      "num_tokens": 17842968.0,
      "step": 7990
    },
    {
      "entropy": 5.652894687652588,
      "epoch": 0.8979614758241141,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.4962,
      "mean_token_accuracy": 0.17819192707538606,
      "num_tokens": 17854295.0,
      "step": 7995
    },
    {
      "entropy": 5.680625343322754,
      "epoch": 0.8985230527320717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.4129,
      "mean_token_accuracy": 0.1910711407661438,
      "num_tokens": 17865597.0,
      "step": 8000
    },
    {
      "entropy": 5.4499085426330565,
      "epoch": 0.8990846296400292,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.1998,
      "mean_token_accuracy": 0.1972569301724434,
      "num_tokens": 17877405.0,
      "step": 8005
    },
    {
      "entropy": 5.638901233673096,
      "epoch": 0.8996462065479868,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.3867,
      "mean_token_accuracy": 0.18765722811222077,
      "num_tokens": 17889273.0,
      "step": 8010
    },
    {
      "entropy": 5.487770700454712,
      "epoch": 0.9002077834559443,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.2989,
      "mean_token_accuracy": 0.1971630036830902,
      "num_tokens": 17899761.0,
      "step": 8015
    },
    {
      "entropy": 5.461633014678955,
      "epoch": 0.9007693603639019,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.3049,
      "mean_token_accuracy": 0.19265760183334352,
      "num_tokens": 17910607.0,
      "step": 8020
    },
    {
      "entropy": 5.572126483917236,
      "epoch": 0.9013309372718594,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.3664,
      "mean_token_accuracy": 0.1873146638274193,
      "num_tokens": 17922538.0,
      "step": 8025
    },
    {
      "entropy": 5.5411042213439945,
      "epoch": 0.901892514179817,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1907604977488518,
      "num_tokens": 17933313.0,
      "step": 8030
    },
    {
      "entropy": 5.643222093582153,
      "epoch": 0.9024540910877745,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1843215212225914,
      "num_tokens": 17944893.0,
      "step": 8035
    },
    {
      "entropy": 5.579970836639404,
      "epoch": 0.903015667995732,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.3633,
      "mean_token_accuracy": 0.190171454846859,
      "num_tokens": 17955019.0,
      "step": 8040
    },
    {
      "entropy": 5.643844413757324,
      "epoch": 0.9035772449036895,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.39,
      "mean_token_accuracy": 0.187399822473526,
      "num_tokens": 17965793.0,
      "step": 8045
    },
    {
      "entropy": 5.520031690597534,
      "epoch": 0.9041388218116471,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000492918025562663,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1948908105492592,
      "num_tokens": 17975709.0,
      "step": 8050
    },
    {
      "entropy": 5.558075094223023,
      "epoch": 0.9047003987196046,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.4509,
      "mean_token_accuracy": 0.188435834646225,
      "num_tokens": 17985405.0,
      "step": 8055
    },
    {
      "entropy": 5.621336650848389,
      "epoch": 0.9052619756275622,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.4716,
      "mean_token_accuracy": 0.1862781122326851,
      "num_tokens": 17997105.0,
      "step": 8060
    },
    {
      "entropy": 5.570344400405884,
      "epoch": 0.9058235525355197,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19655727446079255,
      "num_tokens": 18007914.0,
      "step": 8065
    },
    {
      "entropy": 5.478234958648682,
      "epoch": 0.9063851294434773,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.3732,
      "mean_token_accuracy": 0.195900559425354,
      "num_tokens": 18018328.0,
      "step": 8070
    },
    {
      "entropy": 5.579414129257202,
      "epoch": 0.9069467063514348,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000492867970138744,
      "loss": 5.33,
      "mean_token_accuracy": 0.19166546165943146,
      "num_tokens": 18028568.0,
      "step": 8075
    },
    {
      "entropy": 5.585651397705078,
      "epoch": 0.9075082832593924,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.335,
      "mean_token_accuracy": 0.19079924374818802,
      "num_tokens": 18038655.0,
      "step": 8080
    },
    {
      "entropy": 5.526411294937134,
      "epoch": 0.9080698601673499,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492847899407939,
      "loss": 5.2887,
      "mean_token_accuracy": 0.19272549897432328,
      "num_tokens": 18049716.0,
      "step": 8085
    },
    {
      "entropy": 5.468786525726318,
      "epoch": 0.9086314370753075,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.2639,
      "mean_token_accuracy": 0.19567763507366182,
      "num_tokens": 18062530.0,
      "step": 8090
    },
    {
      "entropy": 5.514675855636597,
      "epoch": 0.909193013983265,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.3729,
      "mean_token_accuracy": 0.1907053917646408,
      "num_tokens": 18073076.0,
      "step": 8095
    },
    {
      "entropy": 5.59661283493042,
      "epoch": 0.9097545908912226,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.3168,
      "mean_token_accuracy": 0.19224281013011932,
      "num_tokens": 18083759.0,
      "step": 8100
    },
    {
      "entropy": 5.485331630706787,
      "epoch": 0.9103161677991801,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.252,
      "mean_token_accuracy": 0.19687026590108872,
      "num_tokens": 18095980.0,
      "step": 8105
    },
    {
      "entropy": 5.544535779953003,
      "epoch": 0.9108777447071377,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.3547,
      "mean_token_accuracy": 0.19675569534301757,
      "num_tokens": 18106672.0,
      "step": 8110
    },
    {
      "entropy": 5.526040840148926,
      "epoch": 0.9114393216150952,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.368,
      "mean_token_accuracy": 0.19050671607255937,
      "num_tokens": 18117481.0,
      "step": 8115
    },
    {
      "entropy": 5.502511310577392,
      "epoch": 0.9120008985230528,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.3049,
      "mean_token_accuracy": 0.19350315779447555,
      "num_tokens": 18127602.0,
      "step": 8120
    },
    {
      "entropy": 5.582899713516236,
      "epoch": 0.9125624754310103,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.3816,
      "mean_token_accuracy": 0.1934616819024086,
      "num_tokens": 18140109.0,
      "step": 8125
    },
    {
      "entropy": 5.517803144454956,
      "epoch": 0.9131240523389679,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.2461,
      "mean_token_accuracy": 0.20191185027360917,
      "num_tokens": 18150271.0,
      "step": 8130
    },
    {
      "entropy": 5.56629114151001,
      "epoch": 0.9136856292469253,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19615178257226945,
      "num_tokens": 18160746.0,
      "step": 8135
    },
    {
      "entropy": 5.595706129074097,
      "epoch": 0.9142472061548829,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.3711,
      "mean_token_accuracy": 0.1904945805668831,
      "num_tokens": 18172271.0,
      "step": 8140
    },
    {
      "entropy": 5.513562107086182,
      "epoch": 0.9148087830628404,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19334737062454224,
      "num_tokens": 18183116.0,
      "step": 8145
    },
    {
      "entropy": 5.505119514465332,
      "epoch": 0.915370359970798,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.2657,
      "mean_token_accuracy": 0.20379751175642014,
      "num_tokens": 18195147.0,
      "step": 8150
    },
    {
      "entropy": 5.548833131790161,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492706627563404,
      "loss": 5.3181,
      "mean_token_accuracy": 0.19461688697338103,
      "num_tokens": 18206217.0,
      "step": 8155
    },
    {
      "entropy": 5.5102691650390625,
      "epoch": 0.9164935137867131,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.3373,
      "mean_token_accuracy": 0.18679855465888978,
      "num_tokens": 18217434.0,
      "step": 8160
    },
    {
      "entropy": 5.555142021179199,
      "epoch": 0.9170550906946706,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18896551430225372,
      "num_tokens": 18229277.0,
      "step": 8165
    },
    {
      "entropy": 5.496541976928711,
      "epoch": 0.9176166676026282,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.2091,
      "mean_token_accuracy": 0.20136746317148208,
      "num_tokens": 18240320.0,
      "step": 8170
    },
    {
      "entropy": 5.513849449157715,
      "epoch": 0.9181782445105857,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.2424,
      "mean_token_accuracy": 0.20075628906488419,
      "num_tokens": 18250380.0,
      "step": 8175
    },
    {
      "entropy": 5.567536211013794,
      "epoch": 0.9187398214185433,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.4125,
      "mean_token_accuracy": 0.19566311836242675,
      "num_tokens": 18261703.0,
      "step": 8180
    },
    {
      "entropy": 5.569543838500977,
      "epoch": 0.9193013983265008,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.3426,
      "mean_token_accuracy": 0.18974880427122115,
      "num_tokens": 18272472.0,
      "step": 8185
    },
    {
      "entropy": 5.6758569240570065,
      "epoch": 0.9198629752344584,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18953519761562349,
      "num_tokens": 18282168.0,
      "step": 8190
    },
    {
      "entropy": 5.6537360668182375,
      "epoch": 0.9204245521424159,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18461560904979707,
      "num_tokens": 18293750.0,
      "step": 8195
    },
    {
      "entropy": 5.53588547706604,
      "epoch": 0.9209861290503735,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19447958320379258,
      "num_tokens": 18304568.0,
      "step": 8200
    },
    {
      "entropy": 5.479127359390259,
      "epoch": 0.921547705958331,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1962704688310623,
      "num_tokens": 18315129.0,
      "step": 8205
    },
    {
      "entropy": 5.465544700622559,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492594675081777,
      "loss": 5.2369,
      "mean_token_accuracy": 0.19667529314756393,
      "num_tokens": 18324628.0,
      "step": 8210
    },
    {
      "entropy": 5.453718471527099,
      "epoch": 0.9226708597742461,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000492584456007161,
      "loss": 5.2485,
      "mean_token_accuracy": 0.2012983024120331,
      "num_tokens": 18334899.0,
      "step": 8215
    },
    {
      "entropy": 5.5352483749389645,
      "epoch": 0.9232324366822037,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.2555,
      "mean_token_accuracy": 0.1985543802380562,
      "num_tokens": 18345719.0,
      "step": 8220
    },
    {
      "entropy": 5.57752628326416,
      "epoch": 0.9237940135901612,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.3535,
      "mean_token_accuracy": 0.19158326238393783,
      "num_tokens": 18355814.0,
      "step": 8225
    },
    {
      "entropy": 5.576590871810913,
      "epoch": 0.9243555904981187,
      "grad_norm": 0.875,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18854793161153793,
      "num_tokens": 18366746.0,
      "step": 8230
    },
    {
      "entropy": 5.489645481109619,
      "epoch": 0.9249171674060762,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.2318,
      "mean_token_accuracy": 0.20082572996616363,
      "num_tokens": 18375846.0,
      "step": 8235
    },
    {
      "entropy": 5.486934995651245,
      "epoch": 0.9254787443140338,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492533256720994,
      "loss": 5.3695,
      "mean_token_accuracy": 0.19297412931919097,
      "num_tokens": 18386850.0,
      "step": 8240
    },
    {
      "entropy": 5.493029356002808,
      "epoch": 0.9260403212219913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.2591,
      "mean_token_accuracy": 0.2010165646672249,
      "num_tokens": 18396347.0,
      "step": 8245
    },
    {
      "entropy": 5.512905645370483,
      "epoch": 0.9266018981299489,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.2902,
      "mean_token_accuracy": 0.19622570723295213,
      "num_tokens": 18406583.0,
      "step": 8250
    },
    {
      "entropy": 5.605619239807129,
      "epoch": 0.9271634750379064,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492502454030566,
      "loss": 5.4142,
      "mean_token_accuracy": 0.1884944334626198,
      "num_tokens": 18418586.0,
      "step": 8255
    },
    {
      "entropy": 5.555738925933838,
      "epoch": 0.927725051945864,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.2448,
      "mean_token_accuracy": 0.198564113676548,
      "num_tokens": 18429752.0,
      "step": 8260
    },
    {
      "entropy": 5.553796291351318,
      "epoch": 0.9282866288538215,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.3468,
      "mean_token_accuracy": 0.1901651978492737,
      "num_tokens": 18440663.0,
      "step": 8265
    },
    {
      "entropy": 5.482954597473144,
      "epoch": 0.9288482057617791,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1989026725292206,
      "num_tokens": 18451497.0,
      "step": 8270
    },
    {
      "entropy": 5.551775121688843,
      "epoch": 0.9294097826697366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.2774,
      "mean_token_accuracy": 0.18644112050533296,
      "num_tokens": 18462794.0,
      "step": 8275
    },
    {
      "entropy": 5.536880826950073,
      "epoch": 0.9299713595776942,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1928732216358185,
      "num_tokens": 18472973.0,
      "step": 8280
    },
    {
      "entropy": 5.574653911590576,
      "epoch": 0.9305329364856517,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.4485,
      "mean_token_accuracy": 0.19140620082616805,
      "num_tokens": 18484451.0,
      "step": 8285
    },
    {
      "entropy": 5.601592493057251,
      "epoch": 0.9310945133936093,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.2973,
      "mean_token_accuracy": 0.20106737464666366,
      "num_tokens": 18496412.0,
      "step": 8290
    },
    {
      "entropy": 5.55262451171875,
      "epoch": 0.9316560903015668,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19785861819982528,
      "num_tokens": 18508547.0,
      "step": 8295
    },
    {
      "entropy": 5.506743001937866,
      "epoch": 0.9322176672095244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492409672025112,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1919219821691513,
      "num_tokens": 18518954.0,
      "step": 8300
    },
    {
      "entropy": 5.494297456741333,
      "epoch": 0.9327792441174819,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.2173,
      "mean_token_accuracy": 0.20269113183021545,
      "num_tokens": 18530444.0,
      "step": 8305
    },
    {
      "entropy": 5.523416757583618,
      "epoch": 0.9333408210254395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.295,
      "mean_token_accuracy": 0.19449931681156157,
      "num_tokens": 18541612.0,
      "step": 8310
    },
    {
      "entropy": 5.557974433898925,
      "epoch": 0.933902397933397,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.4309,
      "mean_token_accuracy": 0.189580237865448,
      "num_tokens": 18553316.0,
      "step": 8315
    },
    {
      "entropy": 5.509123802185059,
      "epoch": 0.9344639748413546,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049236825558139,
      "loss": 5.3242,
      "mean_token_accuracy": 0.19414369016885757,
      "num_tokens": 18564468.0,
      "step": 8320
    },
    {
      "entropy": 5.490185689926148,
      "epoch": 0.9350255517493121,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19952441155910491,
      "num_tokens": 18575536.0,
      "step": 8325
    },
    {
      "entropy": 5.524473428726196,
      "epoch": 0.9355871286572696,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18121889978647232,
      "num_tokens": 18586906.0,
      "step": 8330
    },
    {
      "entropy": 5.602032470703125,
      "epoch": 0.9361487055652271,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1940658673644066,
      "num_tokens": 18599064.0,
      "step": 8335
    },
    {
      "entropy": 5.531958484649659,
      "epoch": 0.9367102824731847,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.3805,
      "mean_token_accuracy": 0.1895105093717575,
      "num_tokens": 18610057.0,
      "step": 8340
    },
    {
      "entropy": 5.440084457397461,
      "epoch": 0.9372718593811422,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.2017,
      "mean_token_accuracy": 0.1984959900379181,
      "num_tokens": 18620008.0,
      "step": 8345
    },
    {
      "entropy": 5.473634672164917,
      "epoch": 0.9378334362890998,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19321230202913284,
      "num_tokens": 18630599.0,
      "step": 8350
    },
    {
      "entropy": 5.522151184082031,
      "epoch": 0.9383950131970573,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.408,
      "mean_token_accuracy": 0.19337766319513322,
      "num_tokens": 18643898.0,
      "step": 8355
    },
    {
      "entropy": 5.588105630874634,
      "epoch": 0.9389565901050149,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492285090502816,
      "loss": 5.3766,
      "mean_token_accuracy": 0.18820469081401825,
      "num_tokens": 18654760.0,
      "step": 8360
    },
    {
      "entropy": 5.585411739349365,
      "epoch": 0.9395181670129724,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.3348,
      "mean_token_accuracy": 0.19308007955551149,
      "num_tokens": 18666036.0,
      "step": 8365
    },
    {
      "entropy": 5.555848884582519,
      "epoch": 0.94007974392093,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19247613549232484,
      "num_tokens": 18677806.0,
      "step": 8370
    },
    {
      "entropy": 5.621790885925293,
      "epoch": 0.9406413208288875,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000492253789434953,
      "loss": 5.5073,
      "mean_token_accuracy": 0.18745156973600388,
      "num_tokens": 18688971.0,
      "step": 8375
    },
    {
      "entropy": 5.5669755935668945,
      "epoch": 0.9412028977368451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.3837,
      "mean_token_accuracy": 0.18589464128017424,
      "num_tokens": 18700153.0,
      "step": 8380
    },
    {
      "entropy": 5.67041916847229,
      "epoch": 0.9417644746448026,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.4398,
      "mean_token_accuracy": 0.1897563800215721,
      "num_tokens": 18711935.0,
      "step": 8385
    },
    {
      "entropy": 5.465805768966675,
      "epoch": 0.9423260515527602,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19584644734859466,
      "num_tokens": 18722694.0,
      "step": 8390
    },
    {
      "entropy": 5.5705383777618405,
      "epoch": 0.9428876284607177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.3986,
      "mean_token_accuracy": 0.19401849508285524,
      "num_tokens": 18733443.0,
      "step": 8395
    },
    {
      "entropy": 5.596127033233643,
      "epoch": 0.9434492053686753,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.3049,
      "mean_token_accuracy": 0.1910915344953537,
      "num_tokens": 18743965.0,
      "step": 8400
    },
    {
      "entropy": 5.538570928573608,
      "epoch": 0.9440107822766328,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.4014,
      "mean_token_accuracy": 0.19617786705493928,
      "num_tokens": 18755712.0,
      "step": 8405
    },
    {
      "entropy": 5.5324314594268795,
      "epoch": 0.9445723591845904,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1932847797870636,
      "num_tokens": 18766756.0,
      "step": 8410
    },
    {
      "entropy": 5.554022932052613,
      "epoch": 0.9451339360925479,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1860532358288765,
      "num_tokens": 18776932.0,
      "step": 8415
    },
    {
      "entropy": 5.643239164352417,
      "epoch": 0.9456955130005055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.3268,
      "mean_token_accuracy": 0.1922232687473297,
      "num_tokens": 18787197.0,
      "step": 8420
    },
    {
      "entropy": 5.51154899597168,
      "epoch": 0.9462570899084629,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.2985,
      "mean_token_accuracy": 0.20033012926578522,
      "num_tokens": 18797861.0,
      "step": 8425
    },
    {
      "entropy": 5.491284275054932,
      "epoch": 0.9468186668164205,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19391173422336577,
      "num_tokens": 18809111.0,
      "step": 8430
    },
    {
      "entropy": 5.542313098907471,
      "epoch": 0.947380243724378,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.3123,
      "mean_token_accuracy": 0.20570485889911652,
      "num_tokens": 18820236.0,
      "step": 8435
    },
    {
      "entropy": 5.518101501464844,
      "epoch": 0.9479418206323356,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.3159,
      "mean_token_accuracy": 0.19251687973737716,
      "num_tokens": 18831942.0,
      "step": 8440
    },
    {
      "entropy": 5.556395530700684,
      "epoch": 0.9485033975402931,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.3328,
      "mean_token_accuracy": 0.1960123986005783,
      "num_tokens": 18843271.0,
      "step": 8445
    },
    {
      "entropy": 5.526339101791382,
      "epoch": 0.9490649744482507,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.3806,
      "mean_token_accuracy": 0.19405145198106766,
      "num_tokens": 18856285.0,
      "step": 8450
    },
    {
      "entropy": 5.5770423412323,
      "epoch": 0.9496265513562082,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049208579917436,
      "loss": 5.3861,
      "mean_token_accuracy": 0.19449511021375657,
      "num_tokens": 18867203.0,
      "step": 8455
    },
    {
      "entropy": 5.565312671661377,
      "epoch": 0.9501881282641658,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.4228,
      "mean_token_accuracy": 0.18535178005695344,
      "num_tokens": 18878598.0,
      "step": 8460
    },
    {
      "entropy": 5.6221788883209225,
      "epoch": 0.9507497051721233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.3444,
      "mean_token_accuracy": 0.1969468414783478,
      "num_tokens": 18890233.0,
      "step": 8465
    },
    {
      "entropy": 5.605530595779419,
      "epoch": 0.9513112820800809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.4504,
      "mean_token_accuracy": 0.18437018394470214,
      "num_tokens": 18900362.0,
      "step": 8470
    },
    {
      "entropy": 5.599183368682861,
      "epoch": 0.9518728589880384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19856763482093812,
      "num_tokens": 18911924.0,
      "step": 8475
    },
    {
      "entropy": 5.542718696594238,
      "epoch": 0.952434435895996,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.3763,
      "mean_token_accuracy": 0.18838248401880264,
      "num_tokens": 18923931.0,
      "step": 8480
    },
    {
      "entropy": 5.505335807800293,
      "epoch": 0.9529960128039535,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19574508666992188,
      "num_tokens": 18935859.0,
      "step": 8485
    },
    {
      "entropy": 5.553436803817749,
      "epoch": 0.9535575897119111,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1871015027165413,
      "num_tokens": 18946939.0,
      "step": 8490
    },
    {
      "entropy": 5.60821681022644,
      "epoch": 0.9541191666198686,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.3661,
      "mean_token_accuracy": 0.19118023812770843,
      "num_tokens": 18958067.0,
      "step": 8495
    },
    {
      "entropy": 5.5892986297607425,
      "epoch": 0.9546807435278262,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.2846,
      "mean_token_accuracy": 0.1986078828573227,
      "num_tokens": 18967920.0,
      "step": 8500
    },
    {
      "entropy": 5.570531988143921,
      "epoch": 0.9552423204357837,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491979906602092,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1987991601228714,
      "num_tokens": 18978244.0,
      "step": 8505
    },
    {
      "entropy": 5.545615530014038,
      "epoch": 0.9558038973437413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.2779,
      "mean_token_accuracy": 0.19204490035772323,
      "num_tokens": 18988464.0,
      "step": 8510
    },
    {
      "entropy": 5.468594932556153,
      "epoch": 0.9563654742516988,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19632799327373504,
      "num_tokens": 18998318.0,
      "step": 8515
    },
    {
      "entropy": 5.467539501190186,
      "epoch": 0.9569270511596563,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.2288,
      "mean_token_accuracy": 0.2000111699104309,
      "num_tokens": 19009802.0,
      "step": 8520
    },
    {
      "entropy": 5.617975854873658,
      "epoch": 0.9574886280676138,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.4417,
      "mean_token_accuracy": 0.19130353927612304,
      "num_tokens": 19020174.0,
      "step": 8525
    },
    {
      "entropy": 5.560629892349243,
      "epoch": 0.9580502049755714,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000491926701217703,
      "loss": 5.3714,
      "mean_token_accuracy": 0.18966914266347884,
      "num_tokens": 19031505.0,
      "step": 8530
    },
    {
      "entropy": 5.501335668563843,
      "epoch": 0.9586117818835289,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.2528,
      "mean_token_accuracy": 0.2018294647336006,
      "num_tokens": 19042052.0,
      "step": 8535
    },
    {
      "entropy": 5.663460206985474,
      "epoch": 0.9591733587914865,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491905370712362,
      "loss": 5.4604,
      "mean_token_accuracy": 0.18997792452573775,
      "num_tokens": 19052050.0,
      "step": 8540
    },
    {
      "entropy": 5.615803670883179,
      "epoch": 0.959734935699444,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19216306358575821,
      "num_tokens": 19062053.0,
      "step": 8545
    },
    {
      "entropy": 5.466895627975464,
      "epoch": 0.9602965126074016,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.3152,
      "mean_token_accuracy": 0.19456751942634581,
      "num_tokens": 19073095.0,
      "step": 8550
    },
    {
      "entropy": 5.5734251022338865,
      "epoch": 0.9608580895153591,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.3731,
      "mean_token_accuracy": 0.199061618745327,
      "num_tokens": 19084521.0,
      "step": 8555
    },
    {
      "entropy": 5.61777663230896,
      "epoch": 0.9614196664233167,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.368,
      "mean_token_accuracy": 0.19134332090616227,
      "num_tokens": 19096372.0,
      "step": 8560
    },
    {
      "entropy": 5.622645330429077,
      "epoch": 0.9619812433312742,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.4836,
      "mean_token_accuracy": 0.1852171689271927,
      "num_tokens": 19108535.0,
      "step": 8565
    },
    {
      "entropy": 5.659264802932739,
      "epoch": 0.9625428202392318,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.3267,
      "mean_token_accuracy": 0.19949966669082642,
      "num_tokens": 19119480.0,
      "step": 8570
    },
    {
      "entropy": 5.554690837860107,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.3303,
      "mean_token_accuracy": 0.19730740487575532,
      "num_tokens": 19131168.0,
      "step": 8575
    },
    {
      "entropy": 5.576368761062622,
      "epoch": 0.9636659740551469,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.4776,
      "mean_token_accuracy": 0.18987948298454285,
      "num_tokens": 19142454.0,
      "step": 8580
    },
    {
      "entropy": 5.574490308761597,
      "epoch": 0.9642275509631044,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000491809041613254,
      "loss": 5.3252,
      "mean_token_accuracy": 0.19607152342796325,
      "num_tokens": 19154682.0,
      "step": 8585
    },
    {
      "entropy": 5.514123106002808,
      "epoch": 0.964789127871062,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.2733,
      "mean_token_accuracy": 0.19470927268266677,
      "num_tokens": 19165190.0,
      "step": 8590
    },
    {
      "entropy": 5.569500207901001,
      "epoch": 0.9653507047790195,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.3971,
      "mean_token_accuracy": 0.19404277503490447,
      "num_tokens": 19176075.0,
      "step": 8595
    },
    {
      "entropy": 5.5756219863891605,
      "epoch": 0.9659122816869771,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.4005,
      "mean_token_accuracy": 0.19264953285455705,
      "num_tokens": 19188450.0,
      "step": 8600
    },
    {
      "entropy": 5.467321300506592,
      "epoch": 0.9664738585949346,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.2524,
      "mean_token_accuracy": 0.20545513927936554,
      "num_tokens": 19201254.0,
      "step": 8605
    },
    {
      "entropy": 5.482296943664551,
      "epoch": 0.9670354355028922,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.3445,
      "mean_token_accuracy": 0.19470448940992355,
      "num_tokens": 19214062.0,
      "step": 8610
    },
    {
      "entropy": 5.555178451538086,
      "epoch": 0.9675970124108496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20220063477754593,
      "num_tokens": 19224844.0,
      "step": 8615
    },
    {
      "entropy": 5.536800527572632,
      "epoch": 0.9681585893188072,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.3971,
      "mean_token_accuracy": 0.1888187289237976,
      "num_tokens": 19237347.0,
      "step": 8620
    },
    {
      "entropy": 5.561524677276611,
      "epoch": 0.9687201662267647,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19381256848573686,
      "num_tokens": 19248743.0,
      "step": 8625
    },
    {
      "entropy": 5.55904335975647,
      "epoch": 0.9692817431347223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.2969,
      "mean_token_accuracy": 0.19650973081588746,
      "num_tokens": 19259904.0,
      "step": 8630
    },
    {
      "entropy": 5.505545425415039,
      "epoch": 0.9698433200426798,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1984265238046646,
      "num_tokens": 19271388.0,
      "step": 8635
    },
    {
      "entropy": 5.493661785125733,
      "epoch": 0.9704048969506374,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.3597,
      "mean_token_accuracy": 0.19674887806177138,
      "num_tokens": 19282790.0,
      "step": 8640
    },
    {
      "entropy": 5.535542726516724,
      "epoch": 0.9709664738585949,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.2641,
      "mean_token_accuracy": 0.20085761994123458,
      "num_tokens": 19293389.0,
      "step": 8645
    },
    {
      "entropy": 5.528273296356201,
      "epoch": 0.9715280507665525,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1904158905148506,
      "num_tokens": 19304593.0,
      "step": 8650
    },
    {
      "entropy": 5.53565821647644,
      "epoch": 0.97208962767451,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.266,
      "mean_token_accuracy": 0.1928234338760376,
      "num_tokens": 19317388.0,
      "step": 8655
    },
    {
      "entropy": 5.538254308700561,
      "epoch": 0.9726512045824676,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.3137,
      "mean_token_accuracy": 0.19793542772531508,
      "num_tokens": 19329228.0,
      "step": 8660
    },
    {
      "entropy": 5.5746701717376705,
      "epoch": 0.9732127814904251,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.3018,
      "mean_token_accuracy": 0.198917718231678,
      "num_tokens": 19339421.0,
      "step": 8665
    },
    {
      "entropy": 5.43107852935791,
      "epoch": 0.9737743583983827,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.2071,
      "mean_token_accuracy": 0.20163071155548096,
      "num_tokens": 19349618.0,
      "step": 8670
    },
    {
      "entropy": 5.412902688980102,
      "epoch": 0.9743359353063402,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.2861,
      "mean_token_accuracy": 0.19824237525463104,
      "num_tokens": 19361324.0,
      "step": 8675
    },
    {
      "entropy": 5.554775667190552,
      "epoch": 0.9748975122142978,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.2959,
      "mean_token_accuracy": 0.19507164210081102,
      "num_tokens": 19373631.0,
      "step": 8680
    },
    {
      "entropy": 5.540220451354981,
      "epoch": 0.9754590891222553,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.2671,
      "mean_token_accuracy": 0.19020372331142427,
      "num_tokens": 19384682.0,
      "step": 8685
    },
    {
      "entropy": 5.597664451599121,
      "epoch": 0.9760206660302129,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.3721,
      "mean_token_accuracy": 0.18854376673698425,
      "num_tokens": 19396344.0,
      "step": 8690
    },
    {
      "entropy": 5.545240879058838,
      "epoch": 0.9765822429381704,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.2316,
      "mean_token_accuracy": 0.19282309263944625,
      "num_tokens": 19406921.0,
      "step": 8695
    },
    {
      "entropy": 5.485334396362305,
      "epoch": 0.977143819846128,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491560327430032,
      "loss": 5.3454,
      "mean_token_accuracy": 0.19403354376554488,
      "num_tokens": 19416851.0,
      "step": 8700
    },
    {
      "entropy": 5.531144380569458,
      "epoch": 0.9777053967540855,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.3228,
      "mean_token_accuracy": 0.19114968776702881,
      "num_tokens": 19428180.0,
      "step": 8705
    },
    {
      "entropy": 5.660508060455323,
      "epoch": 0.9782669736620431,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.3788,
      "mean_token_accuracy": 0.19371167123317717,
      "num_tokens": 19439575.0,
      "step": 8710
    },
    {
      "entropy": 5.5524050235748295,
      "epoch": 0.9788285505700005,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.3394,
      "mean_token_accuracy": 0.19327103793621064,
      "num_tokens": 19452074.0,
      "step": 8715
    },
    {
      "entropy": 5.49491057395935,
      "epoch": 0.9793901274779581,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.2665,
      "mean_token_accuracy": 0.1983703076839447,
      "num_tokens": 19463085.0,
      "step": 8720
    },
    {
      "entropy": 5.660985946655273,
      "epoch": 0.9799517043859156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.4285,
      "mean_token_accuracy": 0.18695472180843353,
      "num_tokens": 19472726.0,
      "step": 8725
    },
    {
      "entropy": 5.489388370513916,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049149484540214,
      "loss": 5.2673,
      "mean_token_accuracy": 0.19235725700855255,
      "num_tokens": 19482275.0,
      "step": 8730
    },
    {
      "entropy": 5.526563930511474,
      "epoch": 0.9810748582018307,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.366,
      "mean_token_accuracy": 0.19616764187812805,
      "num_tokens": 19492782.0,
      "step": 8735
    },
    {
      "entropy": 5.470396375656128,
      "epoch": 0.9816364351097883,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19502665102481842,
      "num_tokens": 19505709.0,
      "step": 8740
    },
    {
      "entropy": 5.574804067611694,
      "epoch": 0.9821980120177458,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19690555781126023,
      "num_tokens": 19516269.0,
      "step": 8745
    },
    {
      "entropy": 5.532788896560669,
      "epoch": 0.9827595889257034,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19426633417606354,
      "num_tokens": 19526786.0,
      "step": 8750
    },
    {
      "entropy": 5.521297597885132,
      "epoch": 0.9833211658336609,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.3288,
      "mean_token_accuracy": 0.1890870898962021,
      "num_tokens": 19537865.0,
      "step": 8755
    },
    {
      "entropy": 5.46358470916748,
      "epoch": 0.9838827427416185,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.2707,
      "mean_token_accuracy": 0.19979665726423262,
      "num_tokens": 19548320.0,
      "step": 8760
    },
    {
      "entropy": 5.561723947525024,
      "epoch": 0.984444319649576,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000491418136064015,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19528566002845765,
      "num_tokens": 19558856.0,
      "step": 8765
    },
    {
      "entropy": 5.570872831344604,
      "epoch": 0.9850058965575336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491407150020996,
      "loss": 5.3838,
      "mean_token_accuracy": 0.19990553110837936,
      "num_tokens": 19570792.0,
      "step": 8770
    },
    {
      "entropy": 5.553385591506958,
      "epoch": 0.9855674734654911,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.2816,
      "mean_token_accuracy": 0.2012216255068779,
      "num_tokens": 19581982.0,
      "step": 8775
    },
    {
      "entropy": 5.469842720031738,
      "epoch": 0.9861290503734487,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.2522,
      "mean_token_accuracy": 0.19338964074850082,
      "num_tokens": 19592882.0,
      "step": 8780
    },
    {
      "entropy": 5.595790576934815,
      "epoch": 0.9866906272814062,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.4713,
      "mean_token_accuracy": 0.19046589732170105,
      "num_tokens": 19605967.0,
      "step": 8785
    },
    {
      "entropy": 5.578424739837646,
      "epoch": 0.9872522041893638,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.2974,
      "mean_token_accuracy": 0.1976533368229866,
      "num_tokens": 19616494.0,
      "step": 8790
    },
    {
      "entropy": 5.574629545211792,
      "epoch": 0.9878137810973213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.3926,
      "mean_token_accuracy": 0.19181709587574006,
      "num_tokens": 19627013.0,
      "step": 8795
    },
    {
      "entropy": 5.555541324615478,
      "epoch": 0.9883753580052789,
      "grad_norm": 0.78125,
      "learning_rate": 0.000491341089074901,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19958296418190002,
      "num_tokens": 19639099.0,
      "step": 8800
    },
    {
      "entropy": 5.458128309249878,
      "epoch": 0.9889369349132364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000491330054805827,
      "loss": 5.2731,
      "mean_token_accuracy": 0.20282889604568483,
      "num_tokens": 19650650.0,
      "step": 8805
    },
    {
      "entropy": 5.507886457443237,
      "epoch": 0.9894985118211939,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.3301,
      "mean_token_accuracy": 0.19496433287858964,
      "num_tokens": 19662869.0,
      "step": 8810
    },
    {
      "entropy": 5.527589464187622,
      "epoch": 0.9900600887291514,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.316,
      "mean_token_accuracy": 0.19373108446598053,
      "num_tokens": 19673232.0,
      "step": 8815
    },
    {
      "entropy": 5.551449918746949,
      "epoch": 0.990621665637109,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.3277,
      "mean_token_accuracy": 0.1935148224234581,
      "num_tokens": 19685069.0,
      "step": 8820
    },
    {
      "entropy": 5.5793900966644285,
      "epoch": 0.9911832425450665,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000491285848852719,
      "loss": 5.4238,
      "mean_token_accuracy": 0.1850482076406479,
      "num_tokens": 19697018.0,
      "step": 8825
    },
    {
      "entropy": 5.5448802471160885,
      "epoch": 0.9917448194530241,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.2746,
      "mean_token_accuracy": 0.20018914639949797,
      "num_tokens": 19707922.0,
      "step": 8830
    },
    {
      "entropy": 5.435269546508789,
      "epoch": 0.9923063963609816,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491263704555005,
      "loss": 5.2324,
      "mean_token_accuracy": 0.1939116448163986,
      "num_tokens": 19718693.0,
      "step": 8835
    },
    {
      "entropy": 5.522778940200806,
      "epoch": 0.9928679732689392,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.3001,
      "mean_token_accuracy": 0.19419669806957246,
      "num_tokens": 19730846.0,
      "step": 8840
    },
    {
      "entropy": 5.551234006881714,
      "epoch": 0.9934295501768967,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.3121,
      "mean_token_accuracy": 0.20046380609273912,
      "num_tokens": 19742099.0,
      "step": 8845
    },
    {
      "entropy": 5.5506292343139645,
      "epoch": 0.9939911270848543,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000491230436464915,
      "loss": 5.3147,
      "mean_token_accuracy": 0.1951744556427002,
      "num_tokens": 19753216.0,
      "step": 8850
    },
    {
      "entropy": 5.508218908309937,
      "epoch": 0.9945527039928118,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.2653,
      "mean_token_accuracy": 0.19551834762096404,
      "num_tokens": 19765316.0,
      "step": 8855
    },
    {
      "entropy": 5.53702449798584,
      "epoch": 0.9951142809007694,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.2759,
      "mean_token_accuracy": 0.19814869165420532,
      "num_tokens": 19776208.0,
      "step": 8860
    },
    {
      "entropy": 5.59254560470581,
      "epoch": 0.9956758578087269,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.3079,
      "mean_token_accuracy": 0.1967771381139755,
      "num_tokens": 19788065.0,
      "step": 8865
    },
    {
      "entropy": 5.487558317184448,
      "epoch": 0.9962374347166845,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19341847598552703,
      "num_tokens": 19799388.0,
      "step": 8870
    },
    {
      "entropy": 5.552020025253296,
      "epoch": 0.996799011624642,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19033802151679993,
      "num_tokens": 19809933.0,
      "step": 8875
    },
    {
      "entropy": 5.656790447235108,
      "epoch": 0.9973605885325996,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.4298,
      "mean_token_accuracy": 0.19190446734428407,
      "num_tokens": 19821566.0,
      "step": 8880
    },
    {
      "entropy": 5.598303747177124,
      "epoch": 0.9979221654405571,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.3861,
      "mean_token_accuracy": 0.19424966722726822,
      "num_tokens": 19833727.0,
      "step": 8885
    },
    {
      "entropy": 5.606028318405151,
      "epoch": 0.9984837423485147,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.3919,
      "mean_token_accuracy": 0.1898197427392006,
      "num_tokens": 19845176.0,
      "step": 8890
    },
    {
      "entropy": 5.5207925796508786,
      "epoch": 0.9990453192564722,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1888708859682083,
      "num_tokens": 19857528.0,
      "step": 8895
    },
    {
      "entropy": 5.520441055297852,
      "epoch": 0.9996068961644298,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.228,
      "mean_token_accuracy": 0.19584738463163376,
      "num_tokens": 19868232.0,
      "step": 8900
    },
    {
      "entropy": 5.4868124855889215,
      "epoch": 1.0001123153815914,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.1628,
      "mean_token_accuracy": 0.19949835704432595,
      "num_tokens": 19877881.0,
      "step": 8905
    },
    {
      "entropy": 5.436063146591186,
      "epoch": 1.000673892289549,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.2598,
      "mean_token_accuracy": 0.20276890993118285,
      "num_tokens": 19888398.0,
      "step": 8910
    },
    {
      "entropy": 5.560799074172974,
      "epoch": 1.0012354691975065,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19680534899234772,
      "num_tokens": 19899661.0,
      "step": 8915
    },
    {
      "entropy": 5.588926029205322,
      "epoch": 1.0017970461054642,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.2798,
      "mean_token_accuracy": 0.19891129583120346,
      "num_tokens": 19911073.0,
      "step": 8920
    },
    {
      "entropy": 5.492541408538818,
      "epoch": 1.0023586230134216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.2151,
      "mean_token_accuracy": 0.20233505666255952,
      "num_tokens": 19922567.0,
      "step": 8925
    },
    {
      "entropy": 5.5742340087890625,
      "epoch": 1.0029201999213793,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.3035,
      "mean_token_accuracy": 0.20030399262905121,
      "num_tokens": 19934547.0,
      "step": 8930
    },
    {
      "entropy": 5.517861938476562,
      "epoch": 1.0034817768293367,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.302,
      "mean_token_accuracy": 0.1969043180346489,
      "num_tokens": 19947659.0,
      "step": 8935
    },
    {
      "entropy": 5.569216871261597,
      "epoch": 1.0040433537372944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491029527020594,
      "loss": 5.2096,
      "mean_token_accuracy": 0.1962577596306801,
      "num_tokens": 19958150.0,
      "step": 8940
    },
    {
      "entropy": 5.511219930648804,
      "epoch": 1.0046049306452518,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.1936,
      "mean_token_accuracy": 0.20453667640686035,
      "num_tokens": 19968173.0,
      "step": 8945
    },
    {
      "entropy": 5.460812282562256,
      "epoch": 1.0051665075532095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19375911653041838,
      "num_tokens": 19978543.0,
      "step": 8950
    },
    {
      "entropy": 5.531821155548096,
      "epoch": 1.005728084461167,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.2238,
      "mean_token_accuracy": 0.19551320672035216,
      "num_tokens": 19989586.0,
      "step": 8955
    },
    {
      "entropy": 5.539719104766846,
      "epoch": 1.0062896613691246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.2442,
      "mean_token_accuracy": 0.1953315407037735,
      "num_tokens": 20000524.0,
      "step": 8960
    },
    {
      "entropy": 5.581871604919433,
      "epoch": 1.006851238277082,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.3513,
      "mean_token_accuracy": 0.19357944577932357,
      "num_tokens": 20012935.0,
      "step": 8965
    },
    {
      "entropy": 5.533119583129883,
      "epoch": 1.0074128151850397,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.2893,
      "mean_token_accuracy": 0.19291466176509858,
      "num_tokens": 20024435.0,
      "step": 8970
    },
    {
      "entropy": 5.544022178649902,
      "epoch": 1.007974392092997,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19360746890306474,
      "num_tokens": 20036017.0,
      "step": 8975
    },
    {
      "entropy": 5.553669023513794,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19775602966547012,
      "num_tokens": 20047216.0,
      "step": 8980
    },
    {
      "entropy": 5.525852823257447,
      "epoch": 1.0090975459089122,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20178949236869811,
      "num_tokens": 20057439.0,
      "step": 8985
    },
    {
      "entropy": 5.520406723022461,
      "epoch": 1.0096591228168699,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.2993,
      "mean_token_accuracy": 0.19890758246183396,
      "num_tokens": 20069602.0,
      "step": 8990
    },
    {
      "entropy": 5.469711399078369,
      "epoch": 1.0102206997248273,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.2564,
      "mean_token_accuracy": 0.19699180573225022,
      "num_tokens": 20081366.0,
      "step": 8995
    },
    {
      "entropy": 5.543646812438965,
      "epoch": 1.010782276632785,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.2467,
      "mean_token_accuracy": 0.20182157158851624,
      "num_tokens": 20091384.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.345071268536732,
      "eval_loss": 5.335939884185791,
      "eval_mean_token_accuracy": 0.2012124814915617,
      "eval_num_tokens": 20091384.0,
      "eval_runtime": 18.4201,
      "eval_samples_per_second": 1683.649,
      "eval_steps_per_second": 210.477,
      "step": 9000
    },
    {
      "entropy": 5.596970748901367,
      "epoch": 1.0113438535407424,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.2679,
      "mean_token_accuracy": 0.19310101568698884,
      "num_tokens": 20102953.0,
      "step": 9005
    },
    {
      "entropy": 5.475909757614136,
      "epoch": 1.0119054304486998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19751104414463044,
      "num_tokens": 20113672.0,
      "step": 9010
    },
    {
      "entropy": 5.53746976852417,
      "epoch": 1.0124670073566575,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1988503783941269,
      "num_tokens": 20125143.0,
      "step": 9015
    },
    {
      "entropy": 5.556895732879639,
      "epoch": 1.013028584264615,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.2511,
      "mean_token_accuracy": 0.2013213276863098,
      "num_tokens": 20135489.0,
      "step": 9020
    },
    {
      "entropy": 5.471498441696167,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.2471,
      "mean_token_accuracy": 0.19856238216161728,
      "num_tokens": 20146768.0,
      "step": 9025
    },
    {
      "entropy": 5.364837217330932,
      "epoch": 1.01415173808053,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.0586,
      "mean_token_accuracy": 0.2066863387823105,
      "num_tokens": 20156894.0,
      "step": 9030
    },
    {
      "entropy": 5.532920217514038,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.268,
      "mean_token_accuracy": 0.20452864319086075,
      "num_tokens": 20168174.0,
      "step": 9035
    },
    {
      "entropy": 5.606223392486572,
      "epoch": 1.0152748918964452,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490803680581439,
      "loss": 5.3581,
      "mean_token_accuracy": 0.19166255444288255,
      "num_tokens": 20179962.0,
      "step": 9040
    },
    {
      "entropy": 5.563894557952881,
      "epoch": 1.0158364688044028,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000490792316087133,
      "loss": 5.263,
      "mean_token_accuracy": 0.19781433492898942,
      "num_tokens": 20189842.0,
      "step": 9045
    },
    {
      "entropy": 5.500795555114746,
      "epoch": 1.0163980457123603,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.1422,
      "mean_token_accuracy": 0.20262393355369568,
      "num_tokens": 20200471.0,
      "step": 9050
    },
    {
      "entropy": 5.511017990112305,
      "epoch": 1.016959622620318,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19372006207704545,
      "num_tokens": 20212145.0,
      "step": 9055
    },
    {
      "entropy": 5.546301174163818,
      "epoch": 1.0175211995282754,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.2144,
      "mean_token_accuracy": 0.2050517037510872,
      "num_tokens": 20222454.0,
      "step": 9060
    },
    {
      "entropy": 5.616752481460571,
      "epoch": 1.018082776436233,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19034057408571242,
      "num_tokens": 20233402.0,
      "step": 9065
    },
    {
      "entropy": 5.4688292980194095,
      "epoch": 1.0186443533441905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.1661,
      "mean_token_accuracy": 0.20426513254642487,
      "num_tokens": 20244102.0,
      "step": 9070
    },
    {
      "entropy": 5.590834617614746,
      "epoch": 1.0192059302521481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000490723984844845,
      "loss": 5.3671,
      "mean_token_accuracy": 0.19125825315713882,
      "num_tokens": 20255003.0,
      "step": 9075
    },
    {
      "entropy": 5.542279005050659,
      "epoch": 1.0197675071601056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.1926,
      "mean_token_accuracy": 0.1980121910572052,
      "num_tokens": 20265211.0,
      "step": 9080
    },
    {
      "entropy": 5.482794857025146,
      "epoch": 1.0203290840680632,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.1659,
      "mean_token_accuracy": 0.2007789507508278,
      "num_tokens": 20276552.0,
      "step": 9085
    },
    {
      "entropy": 5.566567039489746,
      "epoch": 1.0208906609760207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1920648843050003,
      "num_tokens": 20286805.0,
      "step": 9090
    },
    {
      "entropy": 5.5743073463439945,
      "epoch": 1.0214522378839783,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.2508,
      "mean_token_accuracy": 0.19493419975042342,
      "num_tokens": 20298536.0,
      "step": 9095
    },
    {
      "entropy": 5.498243141174316,
      "epoch": 1.0220138147919358,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.2509,
      "mean_token_accuracy": 0.19396184384822845,
      "num_tokens": 20309842.0,
      "step": 9100
    },
    {
      "entropy": 5.51289930343628,
      "epoch": 1.0225753916998932,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.1696,
      "mean_token_accuracy": 0.20073921084403992,
      "num_tokens": 20320413.0,
      "step": 9105
    },
    {
      "entropy": 5.537261867523194,
      "epoch": 1.0231369686078509,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1960890546441078,
      "num_tokens": 20331812.0,
      "step": 9110
    },
    {
      "entropy": 5.50948691368103,
      "epoch": 1.0236985455158083,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.168,
      "mean_token_accuracy": 0.20421408265829086,
      "num_tokens": 20342148.0,
      "step": 9115
    },
    {
      "entropy": 5.561124086380005,
      "epoch": 1.024260122423766,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.2778,
      "mean_token_accuracy": 0.19634966403245926,
      "num_tokens": 20353843.0,
      "step": 9120
    },
    {
      "entropy": 5.51594033241272,
      "epoch": 1.0248216993317234,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.2672,
      "mean_token_accuracy": 0.1993176221847534,
      "num_tokens": 20364846.0,
      "step": 9125
    },
    {
      "entropy": 5.53112530708313,
      "epoch": 1.025383276239681,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.2142,
      "mean_token_accuracy": 0.2015288695693016,
      "num_tokens": 20376874.0,
      "step": 9130
    },
    {
      "entropy": 5.499846887588501,
      "epoch": 1.0259448531476385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.2144,
      "mean_token_accuracy": 0.20141475796699523,
      "num_tokens": 20388189.0,
      "step": 9135
    },
    {
      "entropy": 5.5455484867095945,
      "epoch": 1.0265064300555962,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.2517,
      "mean_token_accuracy": 0.20053796023130416,
      "num_tokens": 20399155.0,
      "step": 9140
    },
    {
      "entropy": 5.486528015136718,
      "epoch": 1.0270680069635536,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.1311,
      "mean_token_accuracy": 0.2061092495918274,
      "num_tokens": 20410569.0,
      "step": 9145
    },
    {
      "entropy": 5.426243305206299,
      "epoch": 1.0276295838715113,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2058244377374649,
      "num_tokens": 20420438.0,
      "step": 9150
    },
    {
      "entropy": 5.536541223526001,
      "epoch": 1.0281911607794687,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1948956221342087,
      "num_tokens": 20431874.0,
      "step": 9155
    },
    {
      "entropy": 5.485208320617676,
      "epoch": 1.0287527376874264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.2367,
      "mean_token_accuracy": 0.1939605087041855,
      "num_tokens": 20442792.0,
      "step": 9160
    },
    {
      "entropy": 5.5400598526000975,
      "epoch": 1.0293143145953838,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490517507786001,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1949404090642929,
      "num_tokens": 20454410.0,
      "step": 9165
    },
    {
      "entropy": 5.63997631072998,
      "epoch": 1.0298758915033415,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.3842,
      "mean_token_accuracy": 0.18800537586212157,
      "num_tokens": 20467072.0,
      "step": 9170
    },
    {
      "entropy": 5.566827392578125,
      "epoch": 1.030437468411299,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.3043,
      "mean_token_accuracy": 0.19925779849290848,
      "num_tokens": 20477697.0,
      "step": 9175
    },
    {
      "entropy": 5.484126091003418,
      "epoch": 1.0309990453192566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20723719000816346,
      "num_tokens": 20489024.0,
      "step": 9180
    },
    {
      "entropy": 5.504212808609009,
      "epoch": 1.031560622227214,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19585105925798416,
      "num_tokens": 20499928.0,
      "step": 9185
    },
    {
      "entropy": 5.475916481018066,
      "epoch": 1.0321221991351717,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.2753,
      "mean_token_accuracy": 0.19215475767850876,
      "num_tokens": 20511085.0,
      "step": 9190
    },
    {
      "entropy": 5.534042596817017,
      "epoch": 1.032683776043129,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20305127054452896,
      "num_tokens": 20522164.0,
      "step": 9195
    },
    {
      "entropy": 5.586596488952637,
      "epoch": 1.0332453529510865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.3068,
      "mean_token_accuracy": 0.1998659312725067,
      "num_tokens": 20534187.0,
      "step": 9200
    },
    {
      "entropy": 5.573070049285889,
      "epoch": 1.0338069298590442,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.2864,
      "mean_token_accuracy": 0.2012860044836998,
      "num_tokens": 20545819.0,
      "step": 9205
    },
    {
      "entropy": 5.592633390426636,
      "epoch": 1.0343685067670016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.2825,
      "mean_token_accuracy": 0.19510758966207503,
      "num_tokens": 20556803.0,
      "step": 9210
    },
    {
      "entropy": 5.5788568496704105,
      "epoch": 1.0349300836749593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.2777,
      "mean_token_accuracy": 0.19718366861343384,
      "num_tokens": 20567366.0,
      "step": 9215
    },
    {
      "entropy": 5.549910974502564,
      "epoch": 1.0354916605829168,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.3081,
      "mean_token_accuracy": 0.19345057457685472,
      "num_tokens": 20579182.0,
      "step": 9220
    },
    {
      "entropy": 5.487379026412964,
      "epoch": 1.0360532374908744,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.165,
      "mean_token_accuracy": 0.1996963068842888,
      "num_tokens": 20590346.0,
      "step": 9225
    },
    {
      "entropy": 5.524546575546265,
      "epoch": 1.0366148143988319,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.2731,
      "mean_token_accuracy": 0.20224278122186662,
      "num_tokens": 20601279.0,
      "step": 9230
    },
    {
      "entropy": 5.508433389663696,
      "epoch": 1.0371763913067895,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490355377285356,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20041927993297576,
      "num_tokens": 20612163.0,
      "step": 9235
    },
    {
      "entropy": 5.545577526092529,
      "epoch": 1.037737968214747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.1438,
      "mean_token_accuracy": 0.2041214480996132,
      "num_tokens": 20622587.0,
      "step": 9240
    },
    {
      "entropy": 5.523426055908203,
      "epoch": 1.0382995451227046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.3103,
      "mean_token_accuracy": 0.18915583938360214,
      "num_tokens": 20633341.0,
      "step": 9245
    },
    {
      "entropy": 5.461141586303711,
      "epoch": 1.038861122030662,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.1974,
      "mean_token_accuracy": 0.19987042099237443,
      "num_tokens": 20644582.0,
      "step": 9250
    },
    {
      "entropy": 5.5433509826660154,
      "epoch": 1.0394226989386197,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.1766,
      "mean_token_accuracy": 0.20615993291139603,
      "num_tokens": 20654936.0,
      "step": 9255
    },
    {
      "entropy": 5.4880821228027346,
      "epoch": 1.0399842758465772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.1938,
      "mean_token_accuracy": 0.19732444137334823,
      "num_tokens": 20666348.0,
      "step": 9260
    },
    {
      "entropy": 5.590974807739258,
      "epoch": 1.0405458527545348,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19329182356595992,
      "num_tokens": 20677482.0,
      "step": 9265
    },
    {
      "entropy": 5.582634401321411,
      "epoch": 1.0411074296624923,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.2656,
      "mean_token_accuracy": 0.20015795081853865,
      "num_tokens": 20688369.0,
      "step": 9270
    },
    {
      "entropy": 5.484326887130737,
      "epoch": 1.04166900657045,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.288,
      "mean_token_accuracy": 0.18508729189634324,
      "num_tokens": 20698933.0,
      "step": 9275
    },
    {
      "entropy": 5.497766065597534,
      "epoch": 1.0422305834784074,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.1445,
      "mean_token_accuracy": 0.20259302407503127,
      "num_tokens": 20710192.0,
      "step": 9280
    },
    {
      "entropy": 5.499121904373169,
      "epoch": 1.042792160386365,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1956183597445488,
      "num_tokens": 20721408.0,
      "step": 9285
    },
    {
      "entropy": 5.453102922439575,
      "epoch": 1.0433537372943225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.1173,
      "mean_token_accuracy": 0.20515661239624022,
      "num_tokens": 20732155.0,
      "step": 9290
    },
    {
      "entropy": 5.501876974105835,
      "epoch": 1.04391531420228,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000490215338554978,
      "loss": 5.1129,
      "mean_token_accuracy": 0.20632632821798325,
      "num_tokens": 20741819.0,
      "step": 9295
    },
    {
      "entropy": 5.5616473197937015,
      "epoch": 1.0444768911102376,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.2538,
      "mean_token_accuracy": 0.19582687318325043,
      "num_tokens": 20754931.0,
      "step": 9300
    },
    {
      "entropy": 5.529901599884033,
      "epoch": 1.045038468018195,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.326,
      "mean_token_accuracy": 0.1938871219754219,
      "num_tokens": 20766533.0,
      "step": 9305
    },
    {
      "entropy": 5.512788724899292,
      "epoch": 1.0456000449261527,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20501883774995805,
      "num_tokens": 20777055.0,
      "step": 9310
    },
    {
      "entropy": 5.4719384670257565,
      "epoch": 1.04616162183411,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.1375,
      "mean_token_accuracy": 0.2040897712111473,
      "num_tokens": 20786871.0,
      "step": 9315
    },
    {
      "entropy": 5.56620421409607,
      "epoch": 1.0467231987420678,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.2954,
      "mean_token_accuracy": 0.19282669872045516,
      "num_tokens": 20798546.0,
      "step": 9320
    },
    {
      "entropy": 5.549797534942627,
      "epoch": 1.0472847756500252,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.2296,
      "mean_token_accuracy": 0.19723979979753495,
      "num_tokens": 20809657.0,
      "step": 9325
    },
    {
      "entropy": 5.471808815002442,
      "epoch": 1.0478463525579829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1972489222884178,
      "num_tokens": 20819662.0,
      "step": 9330
    },
    {
      "entropy": 5.423494577407837,
      "epoch": 1.0484079294659403,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000490121431158592,
      "loss": 5.1016,
      "mean_token_accuracy": 0.20698777437210084,
      "num_tokens": 20830554.0,
      "step": 9335
    },
    {
      "entropy": 5.536175012588501,
      "epoch": 1.048969506373898,
      "grad_norm": 0.875,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.2698,
      "mean_token_accuracy": 0.1924435317516327,
      "num_tokens": 20842003.0,
      "step": 9340
    },
    {
      "entropy": 5.5185455799102785,
      "epoch": 1.0495310832818554,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.2721,
      "mean_token_accuracy": 0.20159535259008407,
      "num_tokens": 20853453.0,
      "step": 9345
    },
    {
      "entropy": 5.611779260635376,
      "epoch": 1.050092660189813,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19494579881429672,
      "num_tokens": 20867296.0,
      "step": 9350
    },
    {
      "entropy": 5.474315357208252,
      "epoch": 1.0506542370977705,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.2158,
      "mean_token_accuracy": 0.20532965809106826,
      "num_tokens": 20878951.0,
      "step": 9355
    },
    {
      "entropy": 5.588257646560669,
      "epoch": 1.0512158140057282,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.3006,
      "mean_token_accuracy": 0.19790297448635102,
      "num_tokens": 20891103.0,
      "step": 9360
    },
    {
      "entropy": 5.597253513336182,
      "epoch": 1.0517773909136856,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19192941039800643,
      "num_tokens": 20903570.0,
      "step": 9365
    },
    {
      "entropy": 5.592698240280152,
      "epoch": 1.0523389678216433,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.3271,
      "mean_token_accuracy": 0.1891946539282799,
      "num_tokens": 20914156.0,
      "step": 9370
    },
    {
      "entropy": 5.490146207809448,
      "epoch": 1.0529005447296007,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000490027085391092,
      "loss": 5.2214,
      "mean_token_accuracy": 0.20149651765823365,
      "num_tokens": 20925192.0,
      "step": 9375
    },
    {
      "entropy": 5.487888956069947,
      "epoch": 1.0534621216375584,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490015261356751,
      "loss": 5.1053,
      "mean_token_accuracy": 0.211493019759655,
      "num_tokens": 20935281.0,
      "step": 9380
    },
    {
      "entropy": 5.634827327728272,
      "epoch": 1.0540236985455158,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.324,
      "mean_token_accuracy": 0.19794905036687852,
      "num_tokens": 20947397.0,
      "step": 9385
    },
    {
      "entropy": 5.484256649017334,
      "epoch": 1.0545852754534732,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000489991592749938,
      "loss": 5.184,
      "mean_token_accuracy": 0.20408945828676223,
      "num_tokens": 20958473.0,
      "step": 9390
    },
    {
      "entropy": 5.612189769744873,
      "epoch": 1.055146852361431,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.2887,
      "mean_token_accuracy": 0.19217817038297652,
      "num_tokens": 20970112.0,
      "step": 9395
    },
    {
      "entropy": 5.477616024017334,
      "epoch": 1.0557084292693883,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.2017,
      "mean_token_accuracy": 0.19814509004354477,
      "num_tokens": 20980762.0,
      "step": 9400
    },
    {
      "entropy": 5.442981624603272,
      "epoch": 1.056270006177346,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.1173,
      "mean_token_accuracy": 0.2074087977409363,
      "num_tokens": 20991663.0,
      "step": 9405
    },
    {
      "entropy": 5.491381645202637,
      "epoch": 1.0568315830853034,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489944173394342,
      "loss": 5.1345,
      "mean_token_accuracy": 0.2053958848118782,
      "num_tokens": 21002121.0,
      "step": 9410
    },
    {
      "entropy": 5.463469457626343,
      "epoch": 1.0573931599932611,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.1955,
      "mean_token_accuracy": 0.18889764845371246,
      "num_tokens": 21013791.0,
      "step": 9415
    },
    {
      "entropy": 5.457913541793824,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20537215620279312,
      "num_tokens": 21022977.0,
      "step": 9420
    },
    {
      "entropy": 5.542646503448486,
      "epoch": 1.0585163138091762,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19645731896162033,
      "num_tokens": 21036372.0,
      "step": 9425
    },
    {
      "entropy": 5.646893501281738,
      "epoch": 1.0590778907171337,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.3731,
      "mean_token_accuracy": 0.18222872018814087,
      "num_tokens": 21047984.0,
      "step": 9430
    },
    {
      "entropy": 5.586396789550781,
      "epoch": 1.0596394676250913,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19897354692220687,
      "num_tokens": 21060127.0,
      "step": 9435
    },
    {
      "entropy": 5.513791275024414,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.2282,
      "mean_token_accuracy": 0.19153477847576142,
      "num_tokens": 21070659.0,
      "step": 9440
    },
    {
      "entropy": 5.5013344287872314,
      "epoch": 1.0607626214410064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.162,
      "mean_token_accuracy": 0.2069990739226341,
      "num_tokens": 21081667.0,
      "step": 9445
    },
    {
      "entropy": 5.4936216354370115,
      "epoch": 1.0613241983489639,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.1901,
      "mean_token_accuracy": 0.20192742049694062,
      "num_tokens": 21094275.0,
      "step": 9450
    },
    {
      "entropy": 5.541127634048462,
      "epoch": 1.0618857752569215,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1975289151072502,
      "num_tokens": 21106360.0,
      "step": 9455
    },
    {
      "entropy": 5.535943508148193,
      "epoch": 1.062447352164879,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1947311908006668,
      "num_tokens": 21117647.0,
      "step": 9460
    },
    {
      "entropy": 5.426172924041748,
      "epoch": 1.0630089290728366,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21009614169597626,
      "num_tokens": 21129599.0,
      "step": 9465
    },
    {
      "entropy": 5.5112557888031,
      "epoch": 1.063570505980794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.169,
      "mean_token_accuracy": 0.1996450513601303,
      "num_tokens": 21140573.0,
      "step": 9470
    },
    {
      "entropy": 5.515107679367065,
      "epoch": 1.0641320828887517,
      "grad_norm": 0.875,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.2794,
      "mean_token_accuracy": 0.19868004322052002,
      "num_tokens": 21151899.0,
      "step": 9475
    },
    {
      "entropy": 5.6203735828399655,
      "epoch": 1.0646936597967092,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.3608,
      "mean_token_accuracy": 0.18995590955018998,
      "num_tokens": 21162748.0,
      "step": 9480
    },
    {
      "entropy": 5.549667835235596,
      "epoch": 1.0652552367046666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000489765375752192,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19841505140066146,
      "num_tokens": 21172962.0,
      "step": 9485
    },
    {
      "entropy": 5.519911241531372,
      "epoch": 1.0658168136126243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.2109,
      "mean_token_accuracy": 0.203855898976326,
      "num_tokens": 21183805.0,
      "step": 9490
    },
    {
      "entropy": 5.451312446594239,
      "epoch": 1.0663783905205817,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.1438,
      "mean_token_accuracy": 0.20594498068094252,
      "num_tokens": 21195129.0,
      "step": 9495
    },
    {
      "entropy": 5.457155847549439,
      "epoch": 1.0669399674285394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000489729431550941,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19502991139888765,
      "num_tokens": 21206300.0,
      "step": 9500
    },
    {
      "entropy": 5.530710077285766,
      "epoch": 1.0675015443364968,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1970871314406395,
      "num_tokens": 21217137.0,
      "step": 9505
    },
    {
      "entropy": 5.505410861968994,
      "epoch": 1.0680631212444545,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.1773,
      "mean_token_accuracy": 0.20232580006122589,
      "num_tokens": 21228399.0,
      "step": 9510
    },
    {
      "entropy": 5.539955377578735,
      "epoch": 1.068624698152412,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19609057158231735,
      "num_tokens": 21239396.0,
      "step": 9515
    },
    {
      "entropy": 5.51862063407898,
      "epoch": 1.0691862750603696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.1955,
      "mean_token_accuracy": 0.19819894134998323,
      "num_tokens": 21249966.0,
      "step": 9520
    },
    {
      "entropy": 5.494237852096558,
      "epoch": 1.069747851968327,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20288627445697785,
      "num_tokens": 21260157.0,
      "step": 9525
    },
    {
      "entropy": 5.518320894241333,
      "epoch": 1.0703094288762847,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.2622,
      "mean_token_accuracy": 0.20142678320407867,
      "num_tokens": 21270738.0,
      "step": 9530
    },
    {
      "entropy": 5.482416200637817,
      "epoch": 1.070871005784242,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.2173,
      "mean_token_accuracy": 0.19962265342473984,
      "num_tokens": 21282469.0,
      "step": 9535
    },
    {
      "entropy": 5.464159965515137,
      "epoch": 1.0714325826921998,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19722228050231932,
      "num_tokens": 21293474.0,
      "step": 9540
    },
    {
      "entropy": 5.544464874267578,
      "epoch": 1.0719941596001572,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.2928,
      "mean_token_accuracy": 0.19559511989355088,
      "num_tokens": 21304904.0,
      "step": 9545
    },
    {
      "entropy": 5.558366489410401,
      "epoch": 1.0725557365081149,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.2438,
      "mean_token_accuracy": 0.2001248210668564,
      "num_tokens": 21315407.0,
      "step": 9550
    },
    {
      "entropy": 5.475416803359986,
      "epoch": 1.0731173134160723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20174843668937684,
      "num_tokens": 21326496.0,
      "step": 9555
    },
    {
      "entropy": 5.5132591247558596,
      "epoch": 1.07367889032403,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1941083014011383,
      "num_tokens": 21339449.0,
      "step": 9560
    },
    {
      "entropy": 5.509999418258667,
      "epoch": 1.0742404672319874,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19096213281154634,
      "num_tokens": 21350431.0,
      "step": 9565
    },
    {
      "entropy": 5.481431674957276,
      "epoch": 1.074802044139945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.2037,
      "mean_token_accuracy": 0.20054869800806047,
      "num_tokens": 21361168.0,
      "step": 9570
    },
    {
      "entropy": 5.537471771240234,
      "epoch": 1.0753636210479025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.2082,
      "mean_token_accuracy": 0.20350566655397415,
      "num_tokens": 21372635.0,
      "step": 9575
    },
    {
      "entropy": 5.448892736434937,
      "epoch": 1.07592519795586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1972611740231514,
      "num_tokens": 21384051.0,
      "step": 9580
    },
    {
      "entropy": 5.464411020278931,
      "epoch": 1.0764867748638176,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19213889986276628,
      "num_tokens": 21397496.0,
      "step": 9585
    },
    {
      "entropy": 5.485758256912232,
      "epoch": 1.077048351771775,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.2911,
      "mean_token_accuracy": 0.19463024139404297,
      "num_tokens": 21409349.0,
      "step": 9590
    },
    {
      "entropy": 5.532784843444825,
      "epoch": 1.0776099286797327,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.2479,
      "mean_token_accuracy": 0.19828616231679916,
      "num_tokens": 21421843.0,
      "step": 9595
    },
    {
      "entropy": 5.472610187530518,
      "epoch": 1.0781715055876901,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.1146,
      "mean_token_accuracy": 0.20499437302350998,
      "num_tokens": 21433230.0,
      "step": 9600
    },
    {
      "entropy": 5.530306005477906,
      "epoch": 1.0787330824956478,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000489476099780026,
      "loss": 5.2465,
      "mean_token_accuracy": 0.19794631153345107,
      "num_tokens": 21444295.0,
      "step": 9605
    },
    {
      "entropy": 5.540285110473633,
      "epoch": 1.0792946594036052,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.1922,
      "mean_token_accuracy": 0.19946166574954988,
      "num_tokens": 21455721.0,
      "step": 9610
    },
    {
      "entropy": 5.522051763534546,
      "epoch": 1.079856236311563,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.2256,
      "mean_token_accuracy": 0.2014063909649849,
      "num_tokens": 21466514.0,
      "step": 9615
    },
    {
      "entropy": 5.516562747955322,
      "epoch": 1.0804178132195204,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000489439663599751,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19480774998664857,
      "num_tokens": 21478782.0,
      "step": 9620
    },
    {
      "entropy": 5.5354842185974125,
      "epoch": 1.080979390127478,
      "grad_norm": 0.90625,
      "learning_rate": 0.000489427504550126,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20682329684495926,
      "num_tokens": 21490428.0,
      "step": 9625
    },
    {
      "entropy": 5.396821737289429,
      "epoch": 1.0815409670354355,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.0782,
      "mean_token_accuracy": 0.20233680307865143,
      "num_tokens": 21501530.0,
      "step": 9630
    },
    {
      "entropy": 5.468472957611084,
      "epoch": 1.0821025439433931,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.1778,
      "mean_token_accuracy": 0.20531565696001053,
      "num_tokens": 21512122.0,
      "step": 9635
    },
    {
      "entropy": 5.455103635787964,
      "epoch": 1.0826641208513506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.1138,
      "mean_token_accuracy": 0.20158322006464005,
      "num_tokens": 21521896.0,
      "step": 9640
    },
    {
      "entropy": 5.485671758651733,
      "epoch": 1.0832256977593082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489378800080952,
      "loss": 5.277,
      "mean_token_accuracy": 0.19836003035306932,
      "num_tokens": 21533584.0,
      "step": 9645
    },
    {
      "entropy": 5.43019208908081,
      "epoch": 1.0837872746672657,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.2431,
      "mean_token_accuracy": 0.20011963844299316,
      "num_tokens": 21545432.0,
      "step": 9650
    },
    {
      "entropy": 5.517799377441406,
      "epoch": 1.0843488515752233,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.2375,
      "mean_token_accuracy": 0.1957999885082245,
      "num_tokens": 21556547.0,
      "step": 9655
    },
    {
      "entropy": 5.520965814590454,
      "epoch": 1.0849104284831808,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19733453392982483,
      "num_tokens": 21567354.0,
      "step": 9660
    },
    {
      "entropy": 5.516571998596191,
      "epoch": 1.0854720053911384,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.1653,
      "mean_token_accuracy": 0.1988843336701393,
      "num_tokens": 21578694.0,
      "step": 9665
    },
    {
      "entropy": 5.45679612159729,
      "epoch": 1.0860335822990959,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.1219,
      "mean_token_accuracy": 0.20868828594684602,
      "num_tokens": 21589365.0,
      "step": 9670
    },
    {
      "entropy": 5.517560720443726,
      "epoch": 1.0865951592070533,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19915731400251388,
      "num_tokens": 21600766.0,
      "step": 9675
    },
    {
      "entropy": 5.556910181045533,
      "epoch": 1.087156736115011,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18736910074949265,
      "num_tokens": 21613969.0,
      "step": 9680
    },
    {
      "entropy": 5.4648151874542235,
      "epoch": 1.0877183130229684,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.2036,
      "mean_token_accuracy": 0.20221873074769975,
      "num_tokens": 21626302.0,
      "step": 9685
    },
    {
      "entropy": 5.54852614402771,
      "epoch": 1.088279889930926,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.26,
      "mean_token_accuracy": 0.2003805473446846,
      "num_tokens": 21637058.0,
      "step": 9690
    },
    {
      "entropy": 5.586956262588501,
      "epoch": 1.0888414668388835,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.2692,
      "mean_token_accuracy": 0.1941409170627594,
      "num_tokens": 21648729.0,
      "step": 9695
    },
    {
      "entropy": 5.396589136123657,
      "epoch": 1.0894030437468412,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21044542491436005,
      "num_tokens": 21659961.0,
      "step": 9700
    },
    {
      "entropy": 5.470961093902588,
      "epoch": 1.0899646206547986,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.1976,
      "mean_token_accuracy": 0.20487565398216248,
      "num_tokens": 21670577.0,
      "step": 9705
    },
    {
      "entropy": 5.459893321990966,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.1412,
      "mean_token_accuracy": 0.20846119225025178,
      "num_tokens": 21679882.0,
      "step": 9710
    },
    {
      "entropy": 5.541026878356933,
      "epoch": 1.0910877744707137,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.2916,
      "mean_token_accuracy": 0.2032737985253334,
      "num_tokens": 21690595.0,
      "step": 9715
    },
    {
      "entropy": 5.465839290618897,
      "epoch": 1.0916493513786714,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.1423,
      "mean_token_accuracy": 0.2024049237370491,
      "num_tokens": 21701251.0,
      "step": 9720
    },
    {
      "entropy": 5.554806327819824,
      "epoch": 1.0922109282866288,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19772370606660844,
      "num_tokens": 21713286.0,
      "step": 9725
    },
    {
      "entropy": 5.582064628601074,
      "epoch": 1.0927725051945865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.3637,
      "mean_token_accuracy": 0.19730201065540315,
      "num_tokens": 21725508.0,
      "step": 9730
    },
    {
      "entropy": 5.479081916809082,
      "epoch": 1.093334082102544,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.1773,
      "mean_token_accuracy": 0.19692719578742982,
      "num_tokens": 21736467.0,
      "step": 9735
    },
    {
      "entropy": 5.555394554138184,
      "epoch": 1.0938956590105016,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.2696,
      "mean_token_accuracy": 0.2006936252117157,
      "num_tokens": 21746114.0,
      "step": 9740
    },
    {
      "entropy": 5.470128297805786,
      "epoch": 1.094457235918459,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000489133640016346,
      "loss": 5.2709,
      "mean_token_accuracy": 0.20017120391130447,
      "num_tokens": 21757145.0,
      "step": 9745
    },
    {
      "entropy": 5.507852983474732,
      "epoch": 1.0950188128264167,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.2426,
      "mean_token_accuracy": 0.19466722756624222,
      "num_tokens": 21768294.0,
      "step": 9750
    },
    {
      "entropy": 5.526945114135742,
      "epoch": 1.095580389734374,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.266,
      "mean_token_accuracy": 0.20226563215255738,
      "num_tokens": 21779667.0,
      "step": 9755
    },
    {
      "entropy": 5.433786201477051,
      "epoch": 1.0961419666423318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.1172,
      "mean_token_accuracy": 0.20274291336536407,
      "num_tokens": 21789635.0,
      "step": 9760
    },
    {
      "entropy": 5.438619661331177,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.015625,
      "learning_rate": 0.000489084280634532,
      "loss": 5.1031,
      "mean_token_accuracy": 0.2094033405184746,
      "num_tokens": 21799462.0,
      "step": 9765
    },
    {
      "entropy": 5.397061061859131,
      "epoch": 1.0972651204582466,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.1636,
      "mean_token_accuracy": 0.20064442306756974,
      "num_tokens": 21811520.0,
      "step": 9770
    },
    {
      "entropy": 5.591492605209351,
      "epoch": 1.0978266973662043,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.2354,
      "mean_token_accuracy": 0.1989922896027565,
      "num_tokens": 21821997.0,
      "step": 9775
    },
    {
      "entropy": 5.5327404022216795,
      "epoch": 1.0983882742741617,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.307,
      "mean_token_accuracy": 0.19593706876039504,
      "num_tokens": 21833353.0,
      "step": 9780
    },
    {
      "entropy": 5.465276479721069,
      "epoch": 1.0989498511821194,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19807688295841216,
      "num_tokens": 21843691.0,
      "step": 9785
    },
    {
      "entropy": 5.448185586929322,
      "epoch": 1.0995114280900768,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.1828,
      "mean_token_accuracy": 0.20047850012779236,
      "num_tokens": 21853771.0,
      "step": 9790
    },
    {
      "entropy": 5.376844835281372,
      "epoch": 1.1000730049980345,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.0672,
      "mean_token_accuracy": 0.2060468837618828,
      "num_tokens": 21863424.0,
      "step": 9795
    },
    {
      "entropy": 5.5019202709198,
      "epoch": 1.100634581905992,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20366901755332947,
      "num_tokens": 21875407.0,
      "step": 9800
    },
    {
      "entropy": 5.533499717712402,
      "epoch": 1.1011961588139496,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19600687325000762,
      "num_tokens": 21887081.0,
      "step": 9805
    },
    {
      "entropy": 5.473865365982055,
      "epoch": 1.101757735721907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.1576,
      "mean_token_accuracy": 0.19677783995866777,
      "num_tokens": 21897858.0,
      "step": 9810
    },
    {
      "entropy": 5.508574533462524,
      "epoch": 1.1023193126298647,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.2056,
      "mean_token_accuracy": 0.1996796101331711,
      "num_tokens": 21907985.0,
      "step": 9815
    },
    {
      "entropy": 5.477117490768433,
      "epoch": 1.1028808895378222,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048894798004696,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19353789687156678,
      "num_tokens": 21918726.0,
      "step": 9820
    },
    {
      "entropy": 5.457109546661377,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.1739,
      "mean_token_accuracy": 0.20030082613229752,
      "num_tokens": 21929027.0,
      "step": 9825
    },
    {
      "entropy": 5.453497838973999,
      "epoch": 1.1040040433537373,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20983829200267792,
      "num_tokens": 21938356.0,
      "step": 9830
    },
    {
      "entropy": 5.484285497665406,
      "epoch": 1.104565620261695,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.1508,
      "mean_token_accuracy": 0.19953617453575134,
      "num_tokens": 21947889.0,
      "step": 9835
    },
    {
      "entropy": 5.530827188491822,
      "epoch": 1.1051271971696524,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19934623837471008,
      "num_tokens": 21958409.0,
      "step": 9840
    },
    {
      "entropy": 5.448442029953003,
      "epoch": 1.10568877407761,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.232,
      "mean_token_accuracy": 0.19690271466970444,
      "num_tokens": 21970120.0,
      "step": 9845
    },
    {
      "entropy": 5.511752700805664,
      "epoch": 1.1062503509855675,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.194,
      "mean_token_accuracy": 0.20641541928052903,
      "num_tokens": 21981499.0,
      "step": 9850
    },
    {
      "entropy": 5.420591115951538,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.125,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.102,
      "mean_token_accuracy": 0.20733749568462373,
      "num_tokens": 21991263.0,
      "step": 9855
    },
    {
      "entropy": 5.44049220085144,
      "epoch": 1.1073735048014826,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20117707103490828,
      "num_tokens": 22001582.0,
      "step": 9860
    },
    {
      "entropy": 5.483178377151489,
      "epoch": 1.10793508170944,
      "grad_norm": 0.921875,
      "learning_rate": 0.000488835848264024,
      "loss": 5.166,
      "mean_token_accuracy": 0.20954571813344955,
      "num_tokens": 22012447.0,
      "step": 9865
    },
    {
      "entropy": 5.506381940841675,
      "epoch": 1.1084966586173977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.2548,
      "mean_token_accuracy": 0.19531938433647156,
      "num_tokens": 22024865.0,
      "step": 9870
    },
    {
      "entropy": 5.4304358005523685,
      "epoch": 1.109058235525355,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20006618797779083,
      "num_tokens": 22036298.0,
      "step": 9875
    },
    {
      "entropy": 5.5299865245819095,
      "epoch": 1.1096198124333128,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.1933,
      "mean_token_accuracy": 0.20357475131750108,
      "num_tokens": 22047499.0,
      "step": 9880
    },
    {
      "entropy": 5.533105707168579,
      "epoch": 1.1101813893412702,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.2167,
      "mean_token_accuracy": 0.2011225402355194,
      "num_tokens": 22057612.0,
      "step": 9885
    },
    {
      "entropy": 5.501762247085571,
      "epoch": 1.1107429662492279,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000488773314508967,
      "loss": 5.2565,
      "mean_token_accuracy": 0.20170857161283492,
      "num_tokens": 22068566.0,
      "step": 9890
    },
    {
      "entropy": 5.47289080619812,
      "epoch": 1.1113045431571853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20339447408914565,
      "num_tokens": 22079821.0,
      "step": 9895
    },
    {
      "entropy": 5.49556040763855,
      "epoch": 1.111866120065143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19686015993356704,
      "num_tokens": 22090143.0,
      "step": 9900
    },
    {
      "entropy": 5.429257249832153,
      "epoch": 1.1124276969731004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.2403,
      "mean_token_accuracy": 0.20217568278312684,
      "num_tokens": 22101601.0,
      "step": 9905
    },
    {
      "entropy": 5.446854066848755,
      "epoch": 1.112989273881058,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488723164983281,
      "loss": 5.1383,
      "mean_token_accuracy": 0.20753098726272584,
      "num_tokens": 22113102.0,
      "step": 9910
    },
    {
      "entropy": 5.560865211486816,
      "epoch": 1.1135508507890155,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.2819,
      "mean_token_accuracy": 0.1981932818889618,
      "num_tokens": 22124650.0,
      "step": 9915
    },
    {
      "entropy": 5.556131219863891,
      "epoch": 1.1141124276969732,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.2309,
      "mean_token_accuracy": 0.20397408157587052,
      "num_tokens": 22136637.0,
      "step": 9920
    },
    {
      "entropy": 5.4515608787536625,
      "epoch": 1.1146740046049306,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20321782529354096,
      "num_tokens": 22149318.0,
      "step": 9925
    },
    {
      "entropy": 5.359082889556885,
      "epoch": 1.1152355815128883,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20661814212799073,
      "num_tokens": 22159410.0,
      "step": 9930
    },
    {
      "entropy": 5.43456916809082,
      "epoch": 1.1157971584208457,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1990073949098587,
      "num_tokens": 22170017.0,
      "step": 9935
    },
    {
      "entropy": 5.460321092605591,
      "epoch": 1.1163587353288034,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.1562,
      "mean_token_accuracy": 0.1995518386363983,
      "num_tokens": 22179961.0,
      "step": 9940
    },
    {
      "entropy": 5.50778112411499,
      "epoch": 1.1169203122367608,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1962197482585907,
      "num_tokens": 22191415.0,
      "step": 9945
    },
    {
      "entropy": 5.550610589981079,
      "epoch": 1.1174818891447185,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.2381,
      "mean_token_accuracy": 0.20928673297166825,
      "num_tokens": 22205354.0,
      "step": 9950
    },
    {
      "entropy": 5.602804517745971,
      "epoch": 1.118043466052676,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.2567,
      "mean_token_accuracy": 0.20263593941926955,
      "num_tokens": 22216989.0,
      "step": 9955
    },
    {
      "entropy": 5.536322116851807,
      "epoch": 1.1186050429606333,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.2578,
      "mean_token_accuracy": 0.19654256701469422,
      "num_tokens": 22228421.0,
      "step": 9960
    },
    {
      "entropy": 5.491076898574829,
      "epoch": 1.119166619868591,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.1341,
      "mean_token_accuracy": 0.2005332887172699,
      "num_tokens": 22240621.0,
      "step": 9965
    },
    {
      "entropy": 5.419737052917481,
      "epoch": 1.1197281967765484,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20727255195379257,
      "num_tokens": 22250936.0,
      "step": 9970
    },
    {
      "entropy": 5.460672855377197,
      "epoch": 1.120289773684506,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.2474,
      "mean_token_accuracy": 0.20322750955820085,
      "num_tokens": 22261766.0,
      "step": 9975
    },
    {
      "entropy": 5.4934563636779785,
      "epoch": 1.1208513505924635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048854678437799,
      "loss": 5.1664,
      "mean_token_accuracy": 0.20223995447158813,
      "num_tokens": 22271831.0,
      "step": 9980
    },
    {
      "entropy": 5.4558759212493895,
      "epoch": 1.1214129275004212,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.1801,
      "mean_token_accuracy": 0.20288852751255035,
      "num_tokens": 22283756.0,
      "step": 9985
    },
    {
      "entropy": 5.4089593410491945,
      "epoch": 1.1219745044083786,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.145,
      "mean_token_accuracy": 0.20137999802827836,
      "num_tokens": 22294576.0,
      "step": 9990
    },
    {
      "entropy": 5.4498144626617435,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.2129,
      "mean_token_accuracy": 0.19997997879981994,
      "num_tokens": 22305003.0,
      "step": 9995
    },
    {
      "entropy": 5.447336196899414,
      "epoch": 1.1230976582242937,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.1775,
      "mean_token_accuracy": 0.20554780066013337,
      "num_tokens": 22316505.0,
      "step": 10000
    },
    {
      "entropy": 5.47895359992981,
      "epoch": 1.1236592351322514,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.2367,
      "mean_token_accuracy": 0.19690674394369126,
      "num_tokens": 22327853.0,
      "step": 10005
    },
    {
      "entropy": 5.505447769165039,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20144063830375672,
      "num_tokens": 22337386.0,
      "step": 10010
    },
    {
      "entropy": 5.508572578430176,
      "epoch": 1.1247823889481665,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.3256,
      "mean_token_accuracy": 0.193753445148468,
      "num_tokens": 22348742.0,
      "step": 10015
    },
    {
      "entropy": 5.5724211692810055,
      "epoch": 1.125343965856124,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.2551,
      "mean_token_accuracy": 0.19759654402732849,
      "num_tokens": 22360083.0,
      "step": 10020
    },
    {
      "entropy": 5.484494638442993,
      "epoch": 1.1259055427640816,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488432693037651,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19411914497613908,
      "num_tokens": 22370982.0,
      "step": 10025
    },
    {
      "entropy": 5.4492223262786865,
      "epoch": 1.126467119672039,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000488419982237771,
      "loss": 5.1627,
      "mean_token_accuracy": 0.20932293981313704,
      "num_tokens": 22382383.0,
      "step": 10030
    },
    {
      "entropy": 5.473843240737915,
      "epoch": 1.1270286965799967,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.1824,
      "mean_token_accuracy": 0.2009715273976326,
      "num_tokens": 22393994.0,
      "step": 10035
    },
    {
      "entropy": 5.3944480419158936,
      "epoch": 1.1275902734879542,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.0823,
      "mean_token_accuracy": 0.20899835079908372,
      "num_tokens": 22405354.0,
      "step": 10040
    },
    {
      "entropy": 5.408200025558472,
      "epoch": 1.1281518503959118,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.137,
      "mean_token_accuracy": 0.19862007945775986,
      "num_tokens": 22416708.0,
      "step": 10045
    },
    {
      "entropy": 5.417037487030029,
      "epoch": 1.1287134273038693,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20617434680461882,
      "num_tokens": 22427724.0,
      "step": 10050
    },
    {
      "entropy": 5.497153997421265,
      "epoch": 1.1292750042118267,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.2207,
      "mean_token_accuracy": 0.19686406552791597,
      "num_tokens": 22437823.0,
      "step": 10055
    },
    {
      "entropy": 5.509790229797363,
      "epoch": 1.1298365811197844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.2497,
      "mean_token_accuracy": 0.20112188309431075,
      "num_tokens": 22450155.0,
      "step": 10060
    },
    {
      "entropy": 5.524084520339966,
      "epoch": 1.130398158027742,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.1782,
      "mean_token_accuracy": 0.1990624412894249,
      "num_tokens": 22460484.0,
      "step": 10065
    },
    {
      "entropy": 5.450386047363281,
      "epoch": 1.1309597349356995,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1996014416217804,
      "num_tokens": 22472063.0,
      "step": 10070
    },
    {
      "entropy": 5.453121709823608,
      "epoch": 1.131521311843657,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.1758,
      "mean_token_accuracy": 0.2104387477040291,
      "num_tokens": 22483669.0,
      "step": 10075
    },
    {
      "entropy": 5.552246856689453,
      "epoch": 1.1320828887516146,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.2579,
      "mean_token_accuracy": 0.19650339037179948,
      "num_tokens": 22495505.0,
      "step": 10080
    },
    {
      "entropy": 5.502291345596314,
      "epoch": 1.132644465659572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.2393,
      "mean_token_accuracy": 0.19604319483041763,
      "num_tokens": 22505260.0,
      "step": 10085
    },
    {
      "entropy": 5.483895015716553,
      "epoch": 1.1332060425675297,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.2815,
      "mean_token_accuracy": 0.1973653182387352,
      "num_tokens": 22516362.0,
      "step": 10090
    },
    {
      "entropy": 5.556116390228271,
      "epoch": 1.133767619475487,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19741191267967223,
      "num_tokens": 22527760.0,
      "step": 10095
    },
    {
      "entropy": 5.664815473556518,
      "epoch": 1.1343291963834448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.3178,
      "mean_token_accuracy": 0.19242115914821625,
      "num_tokens": 22540601.0,
      "step": 10100
    },
    {
      "entropy": 5.45710244178772,
      "epoch": 1.1348907732914022,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.2306,
      "mean_token_accuracy": 0.2004949927330017,
      "num_tokens": 22552469.0,
      "step": 10105
    },
    {
      "entropy": 5.415681600570679,
      "epoch": 1.1354523501993599,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20124891102313996,
      "num_tokens": 22563552.0,
      "step": 10110
    },
    {
      "entropy": 5.472025537490845,
      "epoch": 1.1360139271073173,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20081318467855452,
      "num_tokens": 22575619.0,
      "step": 10115
    },
    {
      "entropy": 5.442496252059937,
      "epoch": 1.136575504015275,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20222095549106597,
      "num_tokens": 22585854.0,
      "step": 10120
    },
    {
      "entropy": 5.513937664031983,
      "epoch": 1.1371370809232324,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.1648,
      "mean_token_accuracy": 0.21144380867481233,
      "num_tokens": 22595899.0,
      "step": 10125
    },
    {
      "entropy": 5.467342376708984,
      "epoch": 1.13769865783119,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20492546558380126,
      "num_tokens": 22607517.0,
      "step": 10130
    },
    {
      "entropy": 5.495854234695434,
      "epoch": 1.1382602347391475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.246,
      "mean_token_accuracy": 0.2005260944366455,
      "num_tokens": 22618531.0,
      "step": 10135
    },
    {
      "entropy": 5.38078351020813,
      "epoch": 1.1388218116471052,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20562222450971604,
      "num_tokens": 22629193.0,
      "step": 10140
    },
    {
      "entropy": 5.512382984161377,
      "epoch": 1.1393833885550626,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.2186,
      "mean_token_accuracy": 0.2027831718325615,
      "num_tokens": 22639794.0,
      "step": 10145
    },
    {
      "entropy": 5.560671186447143,
      "epoch": 1.13994496546302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.1623,
      "mean_token_accuracy": 0.21101603358983995,
      "num_tokens": 22650061.0,
      "step": 10150
    },
    {
      "entropy": 5.471747016906738,
      "epoch": 1.1405065423709777,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048810000479995,
      "loss": 5.2298,
      "mean_token_accuracy": 0.19474509954452515,
      "num_tokens": 22661399.0,
      "step": 10155
    },
    {
      "entropy": 5.585216093063354,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.3226,
      "mean_token_accuracy": 0.19292656779289247,
      "num_tokens": 22672445.0,
      "step": 10160
    },
    {
      "entropy": 5.519457769393921,
      "epoch": 1.1416296961868928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20422002971172332,
      "num_tokens": 22682545.0,
      "step": 10165
    },
    {
      "entropy": 5.48565707206726,
      "epoch": 1.1421912730948502,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.1957,
      "mean_token_accuracy": 0.20079188346862792,
      "num_tokens": 22693885.0,
      "step": 10170
    },
    {
      "entropy": 5.467394590377808,
      "epoch": 1.142752850002808,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20520561933517456,
      "num_tokens": 22703841.0,
      "step": 10175
    },
    {
      "entropy": 5.5099140167236325,
      "epoch": 1.1433144269107653,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19712521135807037,
      "num_tokens": 22714142.0,
      "step": 10180
    },
    {
      "entropy": 5.5160821914672855,
      "epoch": 1.143876003818723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.1659,
      "mean_token_accuracy": 0.1982194647192955,
      "num_tokens": 22725056.0,
      "step": 10185
    },
    {
      "entropy": 5.4315660953521725,
      "epoch": 1.1444375807266804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000488009650901768,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20900379568338395,
      "num_tokens": 22735664.0,
      "step": 10190
    },
    {
      "entropy": 5.442817115783692,
      "epoch": 1.144999157634638,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1997764751315117,
      "num_tokens": 22748071.0,
      "step": 10195
    },
    {
      "entropy": 5.522923326492309,
      "epoch": 1.1455607345425955,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.2398,
      "mean_token_accuracy": 0.1998355805873871,
      "num_tokens": 22759477.0,
      "step": 10200
    },
    {
      "entropy": 5.465047550201416,
      "epoch": 1.1461223114505532,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.1039,
      "mean_token_accuracy": 0.2090495228767395,
      "num_tokens": 22769955.0,
      "step": 10205
    },
    {
      "entropy": 5.410012435913086,
      "epoch": 1.1466838883585107,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.1993,
      "mean_token_accuracy": 0.20063941478729247,
      "num_tokens": 22781216.0,
      "step": 10210
    },
    {
      "entropy": 5.462040615081787,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.1514,
      "mean_token_accuracy": 0.20110777467489244,
      "num_tokens": 22792581.0,
      "step": 10215
    },
    {
      "entropy": 5.425000190734863,
      "epoch": 1.1478070421744258,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.0983,
      "mean_token_accuracy": 0.2070978105068207,
      "num_tokens": 22803305.0,
      "step": 10220
    },
    {
      "entropy": 5.516083765029907,
      "epoch": 1.1483686190823834,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.2261,
      "mean_token_accuracy": 0.20040099322795868,
      "num_tokens": 22815558.0,
      "step": 10225
    },
    {
      "entropy": 5.445632791519165,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.1463,
      "mean_token_accuracy": 0.20543907433748246,
      "num_tokens": 22825428.0,
      "step": 10230
    },
    {
      "entropy": 5.515895652770996,
      "epoch": 1.1494917728982985,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.1878,
      "mean_token_accuracy": 0.20435583144426345,
      "num_tokens": 22836294.0,
      "step": 10235
    },
    {
      "entropy": 5.545832252502441,
      "epoch": 1.150053349806256,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.2725,
      "mean_token_accuracy": 0.2023238494992256,
      "num_tokens": 22847835.0,
      "step": 10240
    },
    {
      "entropy": 5.524635553359985,
      "epoch": 1.1506149267142134,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.2133,
      "mean_token_accuracy": 0.20386450737714767,
      "num_tokens": 22859726.0,
      "step": 10245
    },
    {
      "entropy": 5.409657144546509,
      "epoch": 1.151176503622171,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.1701,
      "mean_token_accuracy": 0.20074600130319595,
      "num_tokens": 22871250.0,
      "step": 10250
    },
    {
      "entropy": 5.535622262954712,
      "epoch": 1.1517380805301287,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.2418,
      "mean_token_accuracy": 0.19980781823396682,
      "num_tokens": 22882034.0,
      "step": 10255
    },
    {
      "entropy": 5.608191585540771,
      "epoch": 1.1522996574380862,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.3285,
      "mean_token_accuracy": 0.19436925202608107,
      "num_tokens": 22894606.0,
      "step": 10260
    },
    {
      "entropy": 5.416283321380615,
      "epoch": 1.1528612343460436,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.1472,
      "mean_token_accuracy": 0.20223457664251326,
      "num_tokens": 22905246.0,
      "step": 10265
    },
    {
      "entropy": 5.4595438003540036,
      "epoch": 1.1534228112540013,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.1008,
      "mean_token_accuracy": 0.20134578943252562,
      "num_tokens": 22916202.0,
      "step": 10270
    },
    {
      "entropy": 5.5063934326171875,
      "epoch": 1.1539843881619587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20342529863119124,
      "num_tokens": 22926451.0,
      "step": 10275
    },
    {
      "entropy": 5.522781896591186,
      "epoch": 1.1545459650699164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.1907,
      "mean_token_accuracy": 0.20174046903848647,
      "num_tokens": 22936618.0,
      "step": 10280
    },
    {
      "entropy": 5.5095234394073485,
      "epoch": 1.1551075419778738,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.235,
      "mean_token_accuracy": 0.19578957706689834,
      "num_tokens": 22948278.0,
      "step": 10285
    },
    {
      "entropy": 5.432264614105224,
      "epoch": 1.1556691188858315,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.12,
      "mean_token_accuracy": 0.20592789351940155,
      "num_tokens": 22959928.0,
      "step": 10290
    },
    {
      "entropy": 5.433085298538208,
      "epoch": 1.156230695793789,
      "grad_norm": 0.875,
      "learning_rate": 0.000487736595774831,
      "loss": 5.1827,
      "mean_token_accuracy": 0.1982198253273964,
      "num_tokens": 22971050.0,
      "step": 10295
    },
    {
      "entropy": 5.36858344078064,
      "epoch": 1.1567922727017466,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20067462623119353,
      "num_tokens": 22981471.0,
      "step": 10300
    },
    {
      "entropy": 5.573341178894043,
      "epoch": 1.157353849609704,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.3432,
      "mean_token_accuracy": 0.2015242800116539,
      "num_tokens": 22992526.0,
      "step": 10305
    },
    {
      "entropy": 5.50419340133667,
      "epoch": 1.1579154265176617,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000487697343959723,
      "loss": 5.1982,
      "mean_token_accuracy": 0.19864674359560014,
      "num_tokens": 23003408.0,
      "step": 10310
    },
    {
      "entropy": 5.457799243927002,
      "epoch": 1.158477003425619,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.107,
      "mean_token_accuracy": 0.2038427397608757,
      "num_tokens": 23013406.0,
      "step": 10315
    },
    {
      "entropy": 5.4152483463287355,
      "epoch": 1.1590385803335768,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.1737,
      "mean_token_accuracy": 0.2018618181347847,
      "num_tokens": 23024960.0,
      "step": 10320
    },
    {
      "entropy": 5.548586225509643,
      "epoch": 1.1596001572415342,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000487658031193323,
      "loss": 5.1913,
      "mean_token_accuracy": 0.195900060236454,
      "num_tokens": 23037514.0,
      "step": 10325
    },
    {
      "entropy": 5.550266551971435,
      "epoch": 1.1601617341494919,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.2967,
      "mean_token_accuracy": 0.19482381790876388,
      "num_tokens": 23048460.0,
      "step": 10330
    },
    {
      "entropy": 5.528992414474487,
      "epoch": 1.1607233110574493,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.157,
      "mean_token_accuracy": 0.20359973907470702,
      "num_tokens": 23058866.0,
      "step": 10335
    },
    {
      "entropy": 5.442178392410279,
      "epoch": 1.1612848879654067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19904339909553528,
      "num_tokens": 23069097.0,
      "step": 10340
    },
    {
      "entropy": 5.556402683258057,
      "epoch": 1.1618464648733644,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20066972821950912,
      "num_tokens": 23079553.0,
      "step": 10345
    },
    {
      "entropy": 5.465704679489136,
      "epoch": 1.162408041781322,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20302453637123108,
      "num_tokens": 23090554.0,
      "step": 10350
    },
    {
      "entropy": 5.417927026748657,
      "epoch": 1.1629696186892795,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2061321869492531,
      "num_tokens": 23101740.0,
      "step": 10355
    },
    {
      "entropy": 5.545480012893677,
      "epoch": 1.163531195597237,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20309724360704423,
      "num_tokens": 23112842.0,
      "step": 10360
    },
    {
      "entropy": 5.456405258178711,
      "epoch": 1.1640927725051946,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.1856,
      "mean_token_accuracy": 0.19895222634077073,
      "num_tokens": 23123822.0,
      "step": 10365
    },
    {
      "entropy": 5.522383117675782,
      "epoch": 1.164654349413152,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19544008672237395,
      "num_tokens": 23134045.0,
      "step": 10370
    },
    {
      "entropy": 5.501151275634766,
      "epoch": 1.1652159263211097,
      "grad_norm": 1.0,
      "learning_rate": 0.000487526548579656,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19219748675823212,
      "num_tokens": 23144200.0,
      "step": 10375
    },
    {
      "entropy": 5.462814235687256,
      "epoch": 1.1657775032290671,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20843852311372757,
      "num_tokens": 23155671.0,
      "step": 10380
    },
    {
      "entropy": 5.574885082244873,
      "epoch": 1.1663390801370248,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.3029,
      "mean_token_accuracy": 0.20232970118522645,
      "num_tokens": 23166644.0,
      "step": 10385
    },
    {
      "entropy": 5.46093053817749,
      "epoch": 1.1669006570449822,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.1884,
      "mean_token_accuracy": 0.19990547746419907,
      "num_tokens": 23177074.0,
      "step": 10390
    },
    {
      "entropy": 5.543512010574341,
      "epoch": 1.16746223395294,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19306806176900865,
      "num_tokens": 23188724.0,
      "step": 10395
    },
    {
      "entropy": 5.500392580032349,
      "epoch": 1.1680238108608973,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.2267,
      "mean_token_accuracy": 0.1992797076702118,
      "num_tokens": 23198459.0,
      "step": 10400
    },
    {
      "entropy": 5.502538013458252,
      "epoch": 1.168585387768855,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000487447334172127,
      "loss": 5.2168,
      "mean_token_accuracy": 0.2003210574388504,
      "num_tokens": 23208991.0,
      "step": 10405
    },
    {
      "entropy": 5.436027240753174,
      "epoch": 1.1691469646768125,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.1333,
      "mean_token_accuracy": 0.2041551649570465,
      "num_tokens": 23219309.0,
      "step": 10410
    },
    {
      "entropy": 5.5330126762390135,
      "epoch": 1.1697085415847701,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.2457,
      "mean_token_accuracy": 0.19402629286050796,
      "num_tokens": 23230551.0,
      "step": 10415
    },
    {
      "entropy": 5.574295139312744,
      "epoch": 1.1702701184927276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.3197,
      "mean_token_accuracy": 0.19698685258626938,
      "num_tokens": 23242799.0,
      "step": 10420
    },
    {
      "entropy": 5.581290435791016,
      "epoch": 1.1708316954006852,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19765586853027345,
      "num_tokens": 23255112.0,
      "step": 10425
    },
    {
      "entropy": 5.46783185005188,
      "epoch": 1.1713932723086427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.2071,
      "mean_token_accuracy": 0.20540781170129777,
      "num_tokens": 23264800.0,
      "step": 10430
    },
    {
      "entropy": 5.421821212768554,
      "epoch": 1.1719548492166,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000487367876246022,
      "loss": 5.1396,
      "mean_token_accuracy": 0.19992227703332902,
      "num_tokens": 23276672.0,
      "step": 10435
    },
    {
      "entropy": 5.486373805999756,
      "epoch": 1.1725164261245578,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.1794,
      "mean_token_accuracy": 0.20331962704658507,
      "num_tokens": 23288282.0,
      "step": 10440
    },
    {
      "entropy": 5.53466420173645,
      "epoch": 1.1730780030325154,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.197,
      "mean_token_accuracy": 0.19943659454584123,
      "num_tokens": 23299078.0,
      "step": 10445
    },
    {
      "entropy": 5.479045295715332,
      "epoch": 1.1736395799404729,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.1942,
      "mean_token_accuracy": 0.19931674748659134,
      "num_tokens": 23310036.0,
      "step": 10450
    },
    {
      "entropy": 5.458569431304932,
      "epoch": 1.1742011568484303,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20746329426765442,
      "num_tokens": 23321029.0,
      "step": 10455
    },
    {
      "entropy": 5.509467554092407,
      "epoch": 1.174762733756388,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.2252,
      "mean_token_accuracy": 0.20195245146751403,
      "num_tokens": 23331584.0,
      "step": 10460
    },
    {
      "entropy": 5.614092540740967,
      "epoch": 1.1753243106643454,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.28,
      "mean_token_accuracy": 0.19303126037120819,
      "num_tokens": 23343171.0,
      "step": 10465
    },
    {
      "entropy": 5.584438800811768,
      "epoch": 1.175885887572303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.2143,
      "mean_token_accuracy": 0.1980983003973961,
      "num_tokens": 23353047.0,
      "step": 10470
    },
    {
      "entropy": 5.503641653060913,
      "epoch": 1.1764474644802605,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.2112,
      "mean_token_accuracy": 0.20044744461774827,
      "num_tokens": 23364447.0,
      "step": 10475
    },
    {
      "entropy": 5.529523277282715,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487248232958846,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19701506793498993,
      "num_tokens": 23375616.0,
      "step": 10480
    },
    {
      "entropy": 5.597634649276733,
      "epoch": 1.1775706182961756,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.351,
      "mean_token_accuracy": 0.1909831017255783,
      "num_tokens": 23388176.0,
      "step": 10485
    },
    {
      "entropy": 5.584848690032959,
      "epoch": 1.1781321952041333,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.2723,
      "mean_token_accuracy": 0.20229338854551315,
      "num_tokens": 23400563.0,
      "step": 10490
    },
    {
      "entropy": 5.536035490036011,
      "epoch": 1.1786937721120907,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.218,
      "mean_token_accuracy": 0.1956962913274765,
      "num_tokens": 23410740.0,
      "step": 10495
    },
    {
      "entropy": 5.4853685855865475,
      "epoch": 1.1792553490200484,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19496024698019027,
      "num_tokens": 23423086.0,
      "step": 10500
    },
    {
      "entropy": 5.539540576934814,
      "epoch": 1.1798169259280058,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000487181527913898,
      "loss": 5.2848,
      "mean_token_accuracy": 0.1978589415550232,
      "num_tokens": 23435869.0,
      "step": 10505
    },
    {
      "entropy": 5.5027422428131105,
      "epoch": 1.1803785028359635,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.1525,
      "mean_token_accuracy": 0.2121325686573982,
      "num_tokens": 23446635.0,
      "step": 10510
    },
    {
      "entropy": 5.4411204814910885,
      "epoch": 1.180940079743921,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20586958080530166,
      "num_tokens": 23458214.0,
      "step": 10515
    },
    {
      "entropy": 5.509582996368408,
      "epoch": 1.1815016566518786,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.2138,
      "mean_token_accuracy": 0.2041307121515274,
      "num_tokens": 23469383.0,
      "step": 10520
    },
    {
      "entropy": 5.4540177345275875,
      "epoch": 1.182063233559836,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20947980284690856,
      "num_tokens": 23480841.0,
      "step": 10525
    },
    {
      "entropy": 5.438432025909424,
      "epoch": 1.1826248104677934,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.1432,
      "mean_token_accuracy": 0.20331989228725433,
      "num_tokens": 23490796.0,
      "step": 10530
    },
    {
      "entropy": 5.4738305568695065,
      "epoch": 1.183186387375751,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.1587,
      "mean_token_accuracy": 0.2014956533908844,
      "num_tokens": 23501759.0,
      "step": 10535
    },
    {
      "entropy": 5.492894887924194,
      "epoch": 1.1837479642837088,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20575653165578842,
      "num_tokens": 23512209.0,
      "step": 10540
    },
    {
      "entropy": 5.534551763534546,
      "epoch": 1.1843095411916662,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.2797,
      "mean_token_accuracy": 0.1959609404206276,
      "num_tokens": 23524316.0,
      "step": 10545
    },
    {
      "entropy": 5.502277421951294,
      "epoch": 1.1848711180996236,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.246,
      "mean_token_accuracy": 0.20394203215837478,
      "num_tokens": 23536199.0,
      "step": 10550
    },
    {
      "entropy": 5.527798414230347,
      "epoch": 1.1854326950075813,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19721332639455796,
      "num_tokens": 23548034.0,
      "step": 10555
    },
    {
      "entropy": 5.526521253585815,
      "epoch": 1.1859942719155387,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.2017,
      "mean_token_accuracy": 0.2062172293663025,
      "num_tokens": 23559280.0,
      "step": 10560
    },
    {
      "entropy": 5.496298742294312,
      "epoch": 1.1865558488234964,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20427300930023193,
      "num_tokens": 23569163.0,
      "step": 10565
    },
    {
      "entropy": 5.493490409851074,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.2417,
      "mean_token_accuracy": 0.1978980928659439,
      "num_tokens": 23580526.0,
      "step": 10570
    },
    {
      "entropy": 5.429447317123413,
      "epoch": 1.1876790026394115,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.1353,
      "mean_token_accuracy": 0.19842249155044556,
      "num_tokens": 23590552.0,
      "step": 10575
    },
    {
      "entropy": 5.390718793869018,
      "epoch": 1.188240579547369,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2079307496547699,
      "num_tokens": 23601889.0,
      "step": 10580
    },
    {
      "entropy": 5.532346725463867,
      "epoch": 1.1888021564553266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.2961,
      "mean_token_accuracy": 0.2045041009783745,
      "num_tokens": 23613833.0,
      "step": 10585
    },
    {
      "entropy": 5.550795555114746,
      "epoch": 1.189363733363284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.2441,
      "mean_token_accuracy": 0.20073410868644714,
      "num_tokens": 23625404.0,
      "step": 10590
    },
    {
      "entropy": 5.431831932067871,
      "epoch": 1.1899253102712417,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.1205,
      "mean_token_accuracy": 0.21142789870500564,
      "num_tokens": 23635670.0,
      "step": 10595
    },
    {
      "entropy": 5.443866205215454,
      "epoch": 1.1904868871791991,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.12,
      "mean_token_accuracy": 0.21075434386730194,
      "num_tokens": 23646893.0,
      "step": 10600
    },
    {
      "entropy": 5.415718650817871,
      "epoch": 1.1910484640871568,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.2214,
      "mean_token_accuracy": 0.2007893517613411,
      "num_tokens": 23657733.0,
      "step": 10605
    },
    {
      "entropy": 5.392384195327759,
      "epoch": 1.1916100409951143,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048689952297394,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20899626463651658,
      "num_tokens": 23668100.0,
      "step": 10610
    },
    {
      "entropy": 5.548053979873657,
      "epoch": 1.192171617903072,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.2443,
      "mean_token_accuracy": 0.193637914955616,
      "num_tokens": 23678295.0,
      "step": 10615
    },
    {
      "entropy": 5.461375284194946,
      "epoch": 1.1927331948110294,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.081,
      "mean_token_accuracy": 0.21138057857751846,
      "num_tokens": 23689310.0,
      "step": 10620
    },
    {
      "entropy": 5.411602210998535,
      "epoch": 1.1932947717189868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19535840451717376,
      "num_tokens": 23699475.0,
      "step": 10625
    },
    {
      "entropy": 5.60453200340271,
      "epoch": 1.1938563486269445,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.2335,
      "mean_token_accuracy": 0.20279577672481536,
      "num_tokens": 23710186.0,
      "step": 10630
    },
    {
      "entropy": 5.401488351821899,
      "epoch": 1.1944179255349021,
      "grad_norm": 1.0,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.0592,
      "mean_token_accuracy": 0.20986445099115372,
      "num_tokens": 23720449.0,
      "step": 10635
    },
    {
      "entropy": 5.520498037338257,
      "epoch": 1.1949795024428596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.2985,
      "mean_token_accuracy": 0.19768839776515962,
      "num_tokens": 23732295.0,
      "step": 10640
    },
    {
      "entropy": 5.431390237808228,
      "epoch": 1.195541079350817,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19689158350229263,
      "num_tokens": 23743616.0,
      "step": 10645
    },
    {
      "entropy": 5.540212249755859,
      "epoch": 1.1961026562587747,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19612667709589005,
      "num_tokens": 23755288.0,
      "step": 10650
    },
    {
      "entropy": 5.499504852294922,
      "epoch": 1.196664233166732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.1148,
      "mean_token_accuracy": 0.20572892129421233,
      "num_tokens": 23765916.0,
      "step": 10655
    },
    {
      "entropy": 5.379067087173462,
      "epoch": 1.1972258100746898,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.075,
      "mean_token_accuracy": 0.2067120537161827,
      "num_tokens": 23778891.0,
      "step": 10660
    },
    {
      "entropy": 5.506110715866089,
      "epoch": 1.1977873869826472,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.219,
      "mean_token_accuracy": 0.2073827639222145,
      "num_tokens": 23790402.0,
      "step": 10665
    },
    {
      "entropy": 5.502700614929199,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.1592,
      "mean_token_accuracy": 0.21002074778079988,
      "num_tokens": 23801928.0,
      "step": 10670
    },
    {
      "entropy": 5.5314232349395756,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.237,
      "mean_token_accuracy": 0.1959489107131958,
      "num_tokens": 23812302.0,
      "step": 10675
    },
    {
      "entropy": 5.506415939331054,
      "epoch": 1.19947211770652,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.2025,
      "mean_token_accuracy": 0.21072551012039184,
      "num_tokens": 23824033.0,
      "step": 10680
    },
    {
      "entropy": 5.445138883590698,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.1742,
      "mean_token_accuracy": 0.20385459512472154,
      "num_tokens": 23834907.0,
      "step": 10685
    },
    {
      "entropy": 5.566518974304199,
      "epoch": 1.200595271522435,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.3478,
      "mean_token_accuracy": 0.19692652821540832,
      "num_tokens": 23846418.0,
      "step": 10690
    },
    {
      "entropy": 5.630992412567139,
      "epoch": 1.2011568484303925,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1959493026137352,
      "num_tokens": 23858423.0,
      "step": 10695
    },
    {
      "entropy": 5.507239246368409,
      "epoch": 1.2017184253383502,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.1774,
      "mean_token_accuracy": 0.20183203667402266,
      "num_tokens": 23868340.0,
      "step": 10700
    },
    {
      "entropy": 5.41426100730896,
      "epoch": 1.2022800022463076,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.1057,
      "mean_token_accuracy": 0.2065039962530136,
      "num_tokens": 23878752.0,
      "step": 10705
    },
    {
      "entropy": 5.512552690505982,
      "epoch": 1.2028415791542653,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.1621,
      "mean_token_accuracy": 0.2099317044019699,
      "num_tokens": 23889374.0,
      "step": 10710
    },
    {
      "entropy": 5.568221378326416,
      "epoch": 1.2034031560622227,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.2244,
      "mean_token_accuracy": 0.1997619241476059,
      "num_tokens": 23901651.0,
      "step": 10715
    },
    {
      "entropy": 5.405527114868164,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.0,
      "learning_rate": 0.000486600897985503,
      "loss": 5.1569,
      "mean_token_accuracy": 0.2025499999523163,
      "num_tokens": 23911818.0,
      "step": 10720
    },
    {
      "entropy": 5.470909690856933,
      "epoch": 1.2045263098781378,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20432161539793015,
      "num_tokens": 23923840.0,
      "step": 10725
    },
    {
      "entropy": 5.43235673904419,
      "epoch": 1.2050878867860955,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21170912981033324,
      "num_tokens": 23935406.0,
      "step": 10730
    },
    {
      "entropy": 5.462775039672851,
      "epoch": 1.205649463694053,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.1762,
      "mean_token_accuracy": 0.2052365943789482,
      "num_tokens": 23946350.0,
      "step": 10735
    },
    {
      "entropy": 5.383950138092041,
      "epoch": 1.2062110406020103,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20410760194063188,
      "num_tokens": 23956315.0,
      "step": 10740
    },
    {
      "entropy": 5.478920412063599,
      "epoch": 1.206772617509968,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20044468343257904,
      "num_tokens": 23966649.0,
      "step": 10745
    },
    {
      "entropy": 5.57425217628479,
      "epoch": 1.2073341944179254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19220172017812728,
      "num_tokens": 23977674.0,
      "step": 10750
    },
    {
      "entropy": 5.451899623870849,
      "epoch": 1.207895771325883,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.211,
      "mean_token_accuracy": 0.19871015548706056,
      "num_tokens": 23987985.0,
      "step": 10755
    },
    {
      "entropy": 5.546921253204346,
      "epoch": 1.2084573482338405,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1912389174103737,
      "num_tokens": 23998730.0,
      "step": 10760
    },
    {
      "entropy": 5.4688927173614506,
      "epoch": 1.2090189251417982,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.144,
      "mean_token_accuracy": 0.19226997047662736,
      "num_tokens": 24008743.0,
      "step": 10765
    },
    {
      "entropy": 5.474906921386719,
      "epoch": 1.2095805020497556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20622960180044175,
      "num_tokens": 24019805.0,
      "step": 10770
    },
    {
      "entropy": 5.502226257324219,
      "epoch": 1.2101420789577133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.1696,
      "mean_token_accuracy": 0.199862040579319,
      "num_tokens": 24031437.0,
      "step": 10775
    },
    {
      "entropy": 5.559590005874634,
      "epoch": 1.2107036558656707,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.3058,
      "mean_token_accuracy": 0.19050522595643998,
      "num_tokens": 24043922.0,
      "step": 10780
    },
    {
      "entropy": 5.502380514144898,
      "epoch": 1.2112652327736284,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.1231,
      "mean_token_accuracy": 0.209582157433033,
      "num_tokens": 24054755.0,
      "step": 10785
    },
    {
      "entropy": 5.445281648635865,
      "epoch": 1.2118268096815858,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.1328,
      "mean_token_accuracy": 0.19305232018232346,
      "num_tokens": 24067347.0,
      "step": 10790
    },
    {
      "entropy": 5.484683561325073,
      "epoch": 1.2123883865895435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.1913,
      "mean_token_accuracy": 0.2029228389263153,
      "num_tokens": 24078938.0,
      "step": 10795
    },
    {
      "entropy": 5.504950046539307,
      "epoch": 1.212949963497501,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1979825183749199,
      "num_tokens": 24090275.0,
      "step": 10800
    },
    {
      "entropy": 5.455297517776489,
      "epoch": 1.2135115404054586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1998402953147888,
      "num_tokens": 24101518.0,
      "step": 10805
    },
    {
      "entropy": 5.539690780639648,
      "epoch": 1.214073117313416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.3035,
      "mean_token_accuracy": 0.19662148207426072,
      "num_tokens": 24113320.0,
      "step": 10810
    },
    {
      "entropy": 5.463568830490113,
      "epoch": 1.2146346942213737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20858280062675477,
      "num_tokens": 24124842.0,
      "step": 10815
    },
    {
      "entropy": 5.535233211517334,
      "epoch": 1.2151962711293312,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.2716,
      "mean_token_accuracy": 0.20383190661668776,
      "num_tokens": 24136391.0,
      "step": 10820
    },
    {
      "entropy": 5.5792961597442625,
      "epoch": 1.2157578480372888,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.2357,
      "mean_token_accuracy": 0.2029911309480667,
      "num_tokens": 24147194.0,
      "step": 10825
    },
    {
      "entropy": 5.532849407196045,
      "epoch": 1.2163194249452463,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.2921,
      "mean_token_accuracy": 0.19194599241018295,
      "num_tokens": 24158170.0,
      "step": 10830
    },
    {
      "entropy": 5.501841592788696,
      "epoch": 1.2168810018532037,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.1316,
      "mean_token_accuracy": 0.20615387260913848,
      "num_tokens": 24168671.0,
      "step": 10835
    },
    {
      "entropy": 5.499409198760986,
      "epoch": 1.2174425787611614,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.1892,
      "mean_token_accuracy": 0.2011120542883873,
      "num_tokens": 24179898.0,
      "step": 10840
    },
    {
      "entropy": 5.438292503356934,
      "epoch": 1.2180041556691188,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.1557,
      "mean_token_accuracy": 0.20252886414527893,
      "num_tokens": 24190622.0,
      "step": 10845
    },
    {
      "entropy": 5.610392475128174,
      "epoch": 1.2185657325770765,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.3368,
      "mean_token_accuracy": 0.19423122256994246,
      "num_tokens": 24202526.0,
      "step": 10850
    },
    {
      "entropy": 5.461727952957153,
      "epoch": 1.219127309485034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.1593,
      "mean_token_accuracy": 0.20182036012411117,
      "num_tokens": 24213186.0,
      "step": 10855
    },
    {
      "entropy": 5.463413715362549,
      "epoch": 1.2196888863929916,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.1854,
      "mean_token_accuracy": 0.2000411108136177,
      "num_tokens": 24225490.0,
      "step": 10860
    },
    {
      "entropy": 5.489551019668579,
      "epoch": 1.220250463300949,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.1595,
      "mean_token_accuracy": 0.20233822017908096,
      "num_tokens": 24236653.0,
      "step": 10865
    },
    {
      "entropy": 5.446139574050903,
      "epoch": 1.2208120402089067,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20879387259483337,
      "num_tokens": 24247843.0,
      "step": 10870
    },
    {
      "entropy": 5.437482309341431,
      "epoch": 1.221373617116864,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.2015,
      "mean_token_accuracy": 0.20367228835821152,
      "num_tokens": 24258332.0,
      "step": 10875
    },
    {
      "entropy": 5.52412691116333,
      "epoch": 1.2219351940248218,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.1923,
      "mean_token_accuracy": 0.203463551402092,
      "num_tokens": 24268685.0,
      "step": 10880
    },
    {
      "entropy": 5.489223527908325,
      "epoch": 1.2224967709327792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19571947306394577,
      "num_tokens": 24280660.0,
      "step": 10885
    },
    {
      "entropy": 5.507982492446899,
      "epoch": 1.2230583478407369,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.1542,
      "mean_token_accuracy": 0.20606174767017366,
      "num_tokens": 24291649.0,
      "step": 10890
    },
    {
      "entropy": 5.497207736968994,
      "epoch": 1.2236199247486943,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.1852,
      "mean_token_accuracy": 0.20491002798080443,
      "num_tokens": 24303108.0,
      "step": 10895
    },
    {
      "entropy": 5.449137020111084,
      "epoch": 1.224181501656652,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.2095,
      "mean_token_accuracy": 0.19586391299962996,
      "num_tokens": 24315939.0,
      "step": 10900
    },
    {
      "entropy": 5.485710668563843,
      "epoch": 1.2247430785646094,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20221732854843139,
      "num_tokens": 24326916.0,
      "step": 10905
    },
    {
      "entropy": 5.517705774307251,
      "epoch": 1.225304655472567,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.2265,
      "mean_token_accuracy": 0.20395824909210206,
      "num_tokens": 24337790.0,
      "step": 10910
    },
    {
      "entropy": 5.436698055267334,
      "epoch": 1.2258662323805245,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2084412008523941,
      "num_tokens": 24347930.0,
      "step": 10915
    },
    {
      "entropy": 5.466492700576782,
      "epoch": 1.2264278092884822,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.142,
      "mean_token_accuracy": 0.20436950623989106,
      "num_tokens": 24359213.0,
      "step": 10920
    },
    {
      "entropy": 5.445788335800171,
      "epoch": 1.2269893861964396,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20397409051656723,
      "num_tokens": 24371000.0,
      "step": 10925
    },
    {
      "entropy": 5.429629707336426,
      "epoch": 1.227550963104397,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21024152040481567,
      "num_tokens": 24381301.0,
      "step": 10930
    },
    {
      "entropy": 5.559954452514648,
      "epoch": 1.2281125400123547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.2843,
      "mean_token_accuracy": 0.19054571986198426,
      "num_tokens": 24391808.0,
      "step": 10935
    },
    {
      "entropy": 5.565803956985474,
      "epoch": 1.2286741169203121,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1993836373090744,
      "num_tokens": 24404220.0,
      "step": 10940
    },
    {
      "entropy": 5.5603865623474125,
      "epoch": 1.2292356938282698,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.2094,
      "mean_token_accuracy": 0.20521530508995056,
      "num_tokens": 24414906.0,
      "step": 10945
    },
    {
      "entropy": 5.470179796218872,
      "epoch": 1.2297972707362272,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.2058,
      "mean_token_accuracy": 0.19760701805353165,
      "num_tokens": 24426469.0,
      "step": 10950
    },
    {
      "entropy": 5.376231718063354,
      "epoch": 1.230358847644185,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.1692,
      "mean_token_accuracy": 0.20013339221477508,
      "num_tokens": 24437683.0,
      "step": 10955
    },
    {
      "entropy": 5.375754451751709,
      "epoch": 1.2309204245521423,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.0633,
      "mean_token_accuracy": 0.2078201502561569,
      "num_tokens": 24449038.0,
      "step": 10960
    },
    {
      "entropy": 5.561332654953003,
      "epoch": 1.2314820014601,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.2007,
      "mean_token_accuracy": 0.20481107383966446,
      "num_tokens": 24460228.0,
      "step": 10965
    },
    {
      "entropy": 5.452834701538086,
      "epoch": 1.2320435783680574,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.2564,
      "mean_token_accuracy": 0.19572933167219161,
      "num_tokens": 24471222.0,
      "step": 10970
    },
    {
      "entropy": 5.4334510326385494,
      "epoch": 1.232605155276015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19481053501367568,
      "num_tokens": 24482127.0,
      "step": 10975
    },
    {
      "entropy": 5.566746044158935,
      "epoch": 1.2331667321839725,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.1669,
      "mean_token_accuracy": 0.20080109387636186,
      "num_tokens": 24492963.0,
      "step": 10980
    },
    {
      "entropy": 5.573767423629761,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.2005,
      "mean_token_accuracy": 0.20147610902786256,
      "num_tokens": 24502885.0,
      "step": 10985
    },
    {
      "entropy": 5.5485392093658445,
      "epoch": 1.2342898859998876,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.2781,
      "mean_token_accuracy": 0.196953122317791,
      "num_tokens": 24514990.0,
      "step": 10990
    },
    {
      "entropy": 5.518796443939209,
      "epoch": 1.2348514629078453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.1993,
      "mean_token_accuracy": 0.19963733106851578,
      "num_tokens": 24525555.0,
      "step": 10995
    },
    {
      "entropy": 5.554047918319702,
      "epoch": 1.2354130398158027,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19470542818307876,
      "num_tokens": 24536350.0,
      "step": 11000
    },
    {
      "entropy": 5.603683280944824,
      "epoch": 1.2359746167237604,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.2386,
      "mean_token_accuracy": 0.19732715934515,
      "num_tokens": 24547756.0,
      "step": 11005
    },
    {
      "entropy": 5.424302387237549,
      "epoch": 1.2365361936317179,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.1465,
      "mean_token_accuracy": 0.2063658282160759,
      "num_tokens": 24559897.0,
      "step": 11010
    },
    {
      "entropy": 5.434448766708374,
      "epoch": 1.2370977705396755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.1621,
      "mean_token_accuracy": 0.195824334025383,
      "num_tokens": 24570168.0,
      "step": 11015
    },
    {
      "entropy": 5.475525999069214,
      "epoch": 1.237659347447633,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.1041,
      "mean_token_accuracy": 0.21376550644636155,
      "num_tokens": 24580918.0,
      "step": 11020
    },
    {
      "entropy": 5.427774238586426,
      "epoch": 1.2382209243555904,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.1436,
      "mean_token_accuracy": 0.1965084195137024,
      "num_tokens": 24592378.0,
      "step": 11025
    },
    {
      "entropy": 5.613003921508789,
      "epoch": 1.238782501263548,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.2349,
      "mean_token_accuracy": 0.20143068581819534,
      "num_tokens": 24603666.0,
      "step": 11030
    },
    {
      "entropy": 5.474401473999023,
      "epoch": 1.2393440781715055,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.1907,
      "mean_token_accuracy": 0.1979813203215599,
      "num_tokens": 24615299.0,
      "step": 11035
    },
    {
      "entropy": 5.470499324798584,
      "epoch": 1.2399056550794632,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.1544,
      "mean_token_accuracy": 0.20516038089990615,
      "num_tokens": 24626138.0,
      "step": 11040
    },
    {
      "entropy": 5.42411437034607,
      "epoch": 1.2404672319874206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20224527418613433,
      "num_tokens": 24636720.0,
      "step": 11045
    },
    {
      "entropy": 5.488111114501953,
      "epoch": 1.2410288088953783,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.1604,
      "mean_token_accuracy": 0.1978873908519745,
      "num_tokens": 24648147.0,
      "step": 11050
    },
    {
      "entropy": 5.496187782287597,
      "epoch": 1.2415903858033357,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.2102,
      "mean_token_accuracy": 0.20320375859737397,
      "num_tokens": 24659825.0,
      "step": 11055
    },
    {
      "entropy": 5.4546410083770756,
      "epoch": 1.2421519627112934,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.1695,
      "mean_token_accuracy": 0.20433487445116044,
      "num_tokens": 24671583.0,
      "step": 11060
    },
    {
      "entropy": 5.450158548355103,
      "epoch": 1.2427135396192508,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20187854170799255,
      "num_tokens": 24682034.0,
      "step": 11065
    },
    {
      "entropy": 5.488341331481934,
      "epoch": 1.2432751165272085,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.2026,
      "mean_token_accuracy": 0.20250927060842513,
      "num_tokens": 24693624.0,
      "step": 11070
    },
    {
      "entropy": 5.534673547744751,
      "epoch": 1.243836693435166,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20009965300559998,
      "num_tokens": 24705340.0,
      "step": 11075
    },
    {
      "entropy": 5.564755964279175,
      "epoch": 1.2443982703431236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19998407661914824,
      "num_tokens": 24717115.0,
      "step": 11080
    },
    {
      "entropy": 5.494892024993897,
      "epoch": 1.244959847251081,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.0914,
      "mean_token_accuracy": 0.20637310147285462,
      "num_tokens": 24727768.0,
      "step": 11085
    },
    {
      "entropy": 5.47627067565918,
      "epoch": 1.2455214241590387,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.2722,
      "mean_token_accuracy": 0.1922347664833069,
      "num_tokens": 24740054.0,
      "step": 11090
    },
    {
      "entropy": 5.524469327926636,
      "epoch": 1.246083001066996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20271396934986113,
      "num_tokens": 24750658.0,
      "step": 11095
    },
    {
      "entropy": 5.60982871055603,
      "epoch": 1.2466445779749538,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000485544223542299,
      "loss": 5.3353,
      "mean_token_accuracy": 0.19859305918216705,
      "num_tokens": 24762730.0,
      "step": 11100
    },
    {
      "entropy": 5.452606153488159,
      "epoch": 1.2472061548829112,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.0854,
      "mean_token_accuracy": 0.2059644117951393,
      "num_tokens": 24773865.0,
      "step": 11105
    },
    {
      "entropy": 5.431053400039673,
      "epoch": 1.2477677317908689,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.1969,
      "mean_token_accuracy": 0.20124391466379166,
      "num_tokens": 24783030.0,
      "step": 11110
    },
    {
      "entropy": 5.447950077056885,
      "epoch": 1.2483293086988263,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20429323464632035,
      "num_tokens": 24793177.0,
      "step": 11115
    },
    {
      "entropy": 5.571758031845093,
      "epoch": 1.2488908856067837,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.2219,
      "mean_token_accuracy": 0.20240359306335448,
      "num_tokens": 24804796.0,
      "step": 11120
    },
    {
      "entropy": 5.559151697158813,
      "epoch": 1.2494524625147414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.2232,
      "mean_token_accuracy": 0.2002537652850151,
      "num_tokens": 24816806.0,
      "step": 11125
    },
    {
      "entropy": 5.461002349853516,
      "epoch": 1.250014039422699,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.1756,
      "mean_token_accuracy": 0.2057817295193672,
      "num_tokens": 24826888.0,
      "step": 11130
    },
    {
      "entropy": 5.47648138999939,
      "epoch": 1.2505756163306565,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.132,
      "mean_token_accuracy": 0.20752278119325637,
      "num_tokens": 24837731.0,
      "step": 11135
    },
    {
      "entropy": 5.450400733947754,
      "epoch": 1.251137193238614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20421704649925232,
      "num_tokens": 24847900.0,
      "step": 11140
    },
    {
      "entropy": 5.490303659439087,
      "epoch": 1.2516987701465716,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.2414,
      "mean_token_accuracy": 0.19748635441064835,
      "num_tokens": 24858878.0,
      "step": 11145
    },
    {
      "entropy": 5.480385875701904,
      "epoch": 1.252260347054529,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20313088595867157,
      "num_tokens": 24870051.0,
      "step": 11150
    },
    {
      "entropy": 5.456357955932617,
      "epoch": 1.2528219239624867,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1979228898882866,
      "num_tokens": 24881671.0,
      "step": 11155
    },
    {
      "entropy": 5.517232465744018,
      "epoch": 1.2533835008704441,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.1256,
      "mean_token_accuracy": 0.20135408490896226,
      "num_tokens": 24892944.0,
      "step": 11160
    },
    {
      "entropy": 5.505953931808472,
      "epoch": 1.2539450777784018,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19954263865947724,
      "num_tokens": 24903346.0,
      "step": 11165
    },
    {
      "entropy": 5.480803680419922,
      "epoch": 1.2545066546863592,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.2052,
      "mean_token_accuracy": 0.20119234025478364,
      "num_tokens": 24913904.0,
      "step": 11170
    },
    {
      "entropy": 5.461050176620484,
      "epoch": 1.255068231594317,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20512817651033402,
      "num_tokens": 24924635.0,
      "step": 11175
    },
    {
      "entropy": 5.586445617675781,
      "epoch": 1.2556298085022743,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.3406,
      "mean_token_accuracy": 0.1903042286634445,
      "num_tokens": 24935466.0,
      "step": 11180
    },
    {
      "entropy": 5.552018928527832,
      "epoch": 1.256191385410232,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.2959,
      "mean_token_accuracy": 0.192475588619709,
      "num_tokens": 24948348.0,
      "step": 11185
    },
    {
      "entropy": 5.470663261413574,
      "epoch": 1.2567529623181894,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20616564452648162,
      "num_tokens": 24959242.0,
      "step": 11190
    },
    {
      "entropy": 5.43701958656311,
      "epoch": 1.2573145392261469,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2051642969250679,
      "num_tokens": 24969855.0,
      "step": 11195
    },
    {
      "entropy": 5.442464685440063,
      "epoch": 1.2578761161341045,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.1274,
      "mean_token_accuracy": 0.20372433066368104,
      "num_tokens": 24980378.0,
      "step": 11200
    },
    {
      "entropy": 5.537302732467651,
      "epoch": 1.2584376930420622,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.2035,
      "mean_token_accuracy": 0.19866700619459152,
      "num_tokens": 24990646.0,
      "step": 11205
    },
    {
      "entropy": 5.6357663631439205,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.3032,
      "mean_token_accuracy": 0.192381352186203,
      "num_tokens": 25002644.0,
      "step": 11210
    },
    {
      "entropy": 5.534732055664063,
      "epoch": 1.259560846857977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.1976,
      "mean_token_accuracy": 0.20510035902261733,
      "num_tokens": 25012635.0,
      "step": 11215
    },
    {
      "entropy": 5.484765434265137,
      "epoch": 1.2601224237659348,
      "grad_norm": 0.90625,
      "learning_rate": 0.000485202483023983,
      "loss": 5.3207,
      "mean_token_accuracy": 0.1922079086303711,
      "num_tokens": 25024289.0,
      "step": 11220
    },
    {
      "entropy": 5.508007955551148,
      "epoch": 1.2606840006738924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.081,
      "mean_token_accuracy": 0.20405347943305968,
      "num_tokens": 25035316.0,
      "step": 11225
    },
    {
      "entropy": 5.506869220733643,
      "epoch": 1.2612455775818499,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.1415,
      "mean_token_accuracy": 0.2054580494761467,
      "num_tokens": 25046011.0,
      "step": 11230
    },
    {
      "entropy": 5.515851926803589,
      "epoch": 1.2618071544898073,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.139,
      "mean_token_accuracy": 0.20588431060314177,
      "num_tokens": 25057402.0,
      "step": 11235
    },
    {
      "entropy": 5.459896850585937,
      "epoch": 1.262368731397765,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.1588,
      "mean_token_accuracy": 0.2041794925928116,
      "num_tokens": 25068287.0,
      "step": 11240
    },
    {
      "entropy": 5.405266761779785,
      "epoch": 1.2629303083057224,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20318641513586044,
      "num_tokens": 25079534.0,
      "step": 11245
    },
    {
      "entropy": 5.4034473419189455,
      "epoch": 1.26349188521368,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.0846,
      "mean_token_accuracy": 0.20523415207862855,
      "num_tokens": 25090825.0,
      "step": 11250
    },
    {
      "entropy": 5.5216649055480955,
      "epoch": 1.2640534621216375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.1954,
      "mean_token_accuracy": 0.20105952620506287,
      "num_tokens": 25101859.0,
      "step": 11255
    },
    {
      "entropy": 5.423482656478882,
      "epoch": 1.2646150390295952,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.1174,
      "mean_token_accuracy": 0.2088337153196335,
      "num_tokens": 25111416.0,
      "step": 11260
    },
    {
      "entropy": 5.47734055519104,
      "epoch": 1.2651766159375526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.2171,
      "mean_token_accuracy": 0.20360954850912094,
      "num_tokens": 25122129.0,
      "step": 11265
    },
    {
      "entropy": 5.502971601486206,
      "epoch": 1.2657381928455103,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.2048,
      "mean_token_accuracy": 0.20440458953380586,
      "num_tokens": 25132429.0,
      "step": 11270
    },
    {
      "entropy": 5.476453828811645,
      "epoch": 1.2662997697534677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.2437,
      "mean_token_accuracy": 0.20275018662214278,
      "num_tokens": 25144242.0,
      "step": 11275
    },
    {
      "entropy": 5.598007535934448,
      "epoch": 1.2668613466614254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.2369,
      "mean_token_accuracy": 0.20308931022882462,
      "num_tokens": 25155332.0,
      "step": 11280
    },
    {
      "entropy": 5.544496250152588,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.1435,
      "mean_token_accuracy": 0.20581945925951003,
      "num_tokens": 25166024.0,
      "step": 11285
    },
    {
      "entropy": 5.418857717514038,
      "epoch": 1.2679845004773402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20878749936819077,
      "num_tokens": 25176697.0,
      "step": 11290
    },
    {
      "entropy": 5.485689115524292,
      "epoch": 1.268546077385298,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.2022,
      "mean_token_accuracy": 0.20789519399404527,
      "num_tokens": 25187421.0,
      "step": 11295
    },
    {
      "entropy": 5.497169399261475,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.2068,
      "mean_token_accuracy": 0.20572234690189362,
      "num_tokens": 25198851.0,
      "step": 11300
    },
    {
      "entropy": 5.548741817474365,
      "epoch": 1.269669231201213,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.269,
      "mean_token_accuracy": 0.19668668806552886,
      "num_tokens": 25209984.0,
      "step": 11305
    },
    {
      "entropy": 5.518511962890625,
      "epoch": 1.2702308081091704,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.155,
      "mean_token_accuracy": 0.20332762897014617,
      "num_tokens": 25220672.0,
      "step": 11310
    },
    {
      "entropy": 5.515075063705444,
      "epoch": 1.270792385017128,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2010663166642189,
      "num_tokens": 25231906.0,
      "step": 11315
    },
    {
      "entropy": 5.507915496826172,
      "epoch": 1.2713539619250858,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.2136,
      "mean_token_accuracy": 0.19545953571796418,
      "num_tokens": 25242902.0,
      "step": 11320
    },
    {
      "entropy": 5.496995544433593,
      "epoch": 1.2719155388330432,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.2692,
      "mean_token_accuracy": 0.19443354457616807,
      "num_tokens": 25254363.0,
      "step": 11325
    },
    {
      "entropy": 5.500916767120361,
      "epoch": 1.2724771157410006,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20396453738212586,
      "num_tokens": 25267108.0,
      "step": 11330
    },
    {
      "entropy": 5.468712615966797,
      "epoch": 1.2730386926489583,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.1508,
      "mean_token_accuracy": 0.2056986927986145,
      "num_tokens": 25278583.0,
      "step": 11335
    },
    {
      "entropy": 5.580595827102661,
      "epoch": 1.2736002695569157,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.231,
      "mean_token_accuracy": 0.20201732516288756,
      "num_tokens": 25289813.0,
      "step": 11340
    },
    {
      "entropy": 5.450843095779419,
      "epoch": 1.2741618464648734,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.1399,
      "mean_token_accuracy": 0.19659793972969056,
      "num_tokens": 25301003.0,
      "step": 11345
    },
    {
      "entropy": 5.454554796218872,
      "epoch": 1.2747234233728308,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2077590599656105,
      "num_tokens": 25312535.0,
      "step": 11350
    },
    {
      "entropy": 5.466577768325806,
      "epoch": 1.2752850002807885,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.1594,
      "mean_token_accuracy": 0.20204212218523027,
      "num_tokens": 25323333.0,
      "step": 11355
    },
    {
      "entropy": 5.571787548065186,
      "epoch": 1.275846577188746,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.2743,
      "mean_token_accuracy": 0.19991803020238877,
      "num_tokens": 25334865.0,
      "step": 11360
    },
    {
      "entropy": 5.556667137145996,
      "epoch": 1.2764081540967036,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.2404,
      "mean_token_accuracy": 0.2015838637948036,
      "num_tokens": 25346672.0,
      "step": 11365
    },
    {
      "entropy": 5.458094263076783,
      "epoch": 1.276969731004661,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.2304,
      "mean_token_accuracy": 0.20394423604011536,
      "num_tokens": 25358702.0,
      "step": 11370
    },
    {
      "entropy": 5.508925247192383,
      "epoch": 1.2775313079126187,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.1685,
      "mean_token_accuracy": 0.2001793086528778,
      "num_tokens": 25369813.0,
      "step": 11375
    },
    {
      "entropy": 5.565791130065918,
      "epoch": 1.2780928848205761,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19597866982221604,
      "num_tokens": 25381042.0,
      "step": 11380
    },
    {
      "entropy": 5.480781841278076,
      "epoch": 1.2786544617285336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19683396369218825,
      "num_tokens": 25391286.0,
      "step": 11385
    },
    {
      "entropy": 5.4633729457855225,
      "epoch": 1.2792160386364912,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20117910653352739,
      "num_tokens": 25402335.0,
      "step": 11390
    },
    {
      "entropy": 5.450736951828003,
      "epoch": 1.279777615544449,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2132217451930046,
      "num_tokens": 25413484.0,
      "step": 11395
    },
    {
      "entropy": 5.531859016418457,
      "epoch": 1.2803391924524063,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.2223,
      "mean_token_accuracy": 0.19713054597377777,
      "num_tokens": 25424276.0,
      "step": 11400
    },
    {
      "entropy": 5.397913455963135,
      "epoch": 1.2809007693603638,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20603392273187637,
      "num_tokens": 25435178.0,
      "step": 11405
    },
    {
      "entropy": 5.3810608863830565,
      "epoch": 1.2814623462683215,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20900043696165085,
      "num_tokens": 25445792.0,
      "step": 11410
    },
    {
      "entropy": 5.444026756286621,
      "epoch": 1.2820239231762791,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.1335,
      "mean_token_accuracy": 0.2036905825138092,
      "num_tokens": 25456598.0,
      "step": 11415
    },
    {
      "entropy": 5.472315120697021,
      "epoch": 1.2825855000842366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20648545622825623,
      "num_tokens": 25468311.0,
      "step": 11420
    },
    {
      "entropy": 5.5864862442016605,
      "epoch": 1.283147076992194,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.2342,
      "mean_token_accuracy": 0.1980331778526306,
      "num_tokens": 25480680.0,
      "step": 11425
    },
    {
      "entropy": 5.596595191955567,
      "epoch": 1.2837086539001517,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.2037,
      "mean_token_accuracy": 0.2052680730819702,
      "num_tokens": 25492255.0,
      "step": 11430
    },
    {
      "entropy": 5.478435897827149,
      "epoch": 1.284270230808109,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000484580560960249,
      "loss": 5.2575,
      "mean_token_accuracy": 0.19775725901126862,
      "num_tokens": 25504530.0,
      "step": 11435
    },
    {
      "entropy": 5.42808780670166,
      "epoch": 1.2848318077160668,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.1896,
      "mean_token_accuracy": 0.2089684247970581,
      "num_tokens": 25515273.0,
      "step": 11440
    },
    {
      "entropy": 5.543203592300415,
      "epoch": 1.2853933846240242,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.3,
      "mean_token_accuracy": 0.19523099213838577,
      "num_tokens": 25526050.0,
      "step": 11445
    },
    {
      "entropy": 5.57367959022522,
      "epoch": 1.2859549615319819,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.206,
      "mean_token_accuracy": 0.19986650943756104,
      "num_tokens": 25538372.0,
      "step": 11450
    },
    {
      "entropy": 5.600811719894409,
      "epoch": 1.2865165384399393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.3051,
      "mean_token_accuracy": 0.2029346525669098,
      "num_tokens": 25549636.0,
      "step": 11455
    },
    {
      "entropy": 5.502301263809204,
      "epoch": 1.287078115347897,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.233,
      "mean_token_accuracy": 0.19834076464176179,
      "num_tokens": 25560850.0,
      "step": 11460
    },
    {
      "entropy": 5.470400953292847,
      "epoch": 1.2876396922558544,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.1713,
      "mean_token_accuracy": 0.20277154445648193,
      "num_tokens": 25572433.0,
      "step": 11465
    },
    {
      "entropy": 5.616461610794067,
      "epoch": 1.288201269163812,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.2164,
      "mean_token_accuracy": 0.2000472664833069,
      "num_tokens": 25582372.0,
      "step": 11470
    },
    {
      "entropy": 5.4393439292907715,
      "epoch": 1.2887628460717695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20855759531259538,
      "num_tokens": 25592570.0,
      "step": 11475
    },
    {
      "entropy": 5.474980449676513,
      "epoch": 1.289324422979727,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19402407109737396,
      "num_tokens": 25603573.0,
      "step": 11480
    },
    {
      "entropy": 5.56807541847229,
      "epoch": 1.2898859998876846,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.217,
      "mean_token_accuracy": 0.20349951833486557,
      "num_tokens": 25614233.0,
      "step": 11485
    },
    {
      "entropy": 5.495745086669922,
      "epoch": 1.2904475767956423,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19812342673540115,
      "num_tokens": 25625077.0,
      "step": 11490
    },
    {
      "entropy": 5.439569616317749,
      "epoch": 1.2910091537035997,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.2214,
      "mean_token_accuracy": 0.1963472530245781,
      "num_tokens": 25636693.0,
      "step": 11495
    },
    {
      "entropy": 5.416373538970947,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2026978075504303,
      "num_tokens": 25647644.0,
      "step": 11500
    },
    {
      "entropy": 5.453104782104492,
      "epoch": 1.2921323075195148,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2088281974196434,
      "num_tokens": 25659373.0,
      "step": 11505
    },
    {
      "entropy": 5.5191456317901615,
      "epoch": 1.2926938844274725,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.2326,
      "mean_token_accuracy": 0.20576125979423524,
      "num_tokens": 25670479.0,
      "step": 11510
    },
    {
      "entropy": 5.533950614929199,
      "epoch": 1.29325546133543,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.237,
      "mean_token_accuracy": 0.20345962643623353,
      "num_tokens": 25682748.0,
      "step": 11515
    },
    {
      "entropy": 5.440349292755127,
      "epoch": 1.2938170382433873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21103416979312897,
      "num_tokens": 25693170.0,
      "step": 11520
    },
    {
      "entropy": 5.467827081680298,
      "epoch": 1.294378615151345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1985013321042061,
      "num_tokens": 25706009.0,
      "step": 11525
    },
    {
      "entropy": 5.547778081893921,
      "epoch": 1.2949401920593024,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.227,
      "mean_token_accuracy": 0.20077787935733796,
      "num_tokens": 25716036.0,
      "step": 11530
    },
    {
      "entropy": 5.492983388900757,
      "epoch": 1.29550176896726,
      "grad_norm": 0.875,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.0364,
      "mean_token_accuracy": 0.20929135531187057,
      "num_tokens": 25726516.0,
      "step": 11535
    },
    {
      "entropy": 5.495220279693603,
      "epoch": 1.2960633458752175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19770278334617614,
      "num_tokens": 25736340.0,
      "step": 11540
    },
    {
      "entropy": 5.411406230926514,
      "epoch": 1.2966249227831752,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.1642,
      "mean_token_accuracy": 0.1971363604068756,
      "num_tokens": 25746439.0,
      "step": 11545
    },
    {
      "entropy": 5.481893634796142,
      "epoch": 1.2971864996911326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.074,
      "mean_token_accuracy": 0.20293221175670623,
      "num_tokens": 25757065.0,
      "step": 11550
    },
    {
      "entropy": 5.503316497802734,
      "epoch": 1.2977480765990903,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.121,
      "mean_token_accuracy": 0.2049147218465805,
      "num_tokens": 25768021.0,
      "step": 11555
    },
    {
      "entropy": 5.497176551818848,
      "epoch": 1.2983096535070477,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.2328,
      "mean_token_accuracy": 0.20426887571811675,
      "num_tokens": 25780197.0,
      "step": 11560
    },
    {
      "entropy": 5.471554327011108,
      "epoch": 1.2988712304150054,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000484198526925203,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20407297164201738,
      "num_tokens": 25790536.0,
      "step": 11565
    },
    {
      "entropy": 5.462974691390992,
      "epoch": 1.2994328073229628,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20496713519096374,
      "num_tokens": 25802397.0,
      "step": 11570
    },
    {
      "entropy": 5.513308429718018,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.2502,
      "mean_token_accuracy": 0.2044883444905281,
      "num_tokens": 25813961.0,
      "step": 11575
    },
    {
      "entropy": 5.525645971298218,
      "epoch": 1.300555961138878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.2639,
      "mean_token_accuracy": 0.1941633403301239,
      "num_tokens": 25825610.0,
      "step": 11580
    },
    {
      "entropy": 5.452743101119995,
      "epoch": 1.3011175380468356,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.0721,
      "mean_token_accuracy": 0.20595871210098265,
      "num_tokens": 25836076.0,
      "step": 11585
    },
    {
      "entropy": 5.471811580657959,
      "epoch": 1.301679114954793,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.2157,
      "mean_token_accuracy": 0.2046196222305298,
      "num_tokens": 25847341.0,
      "step": 11590
    },
    {
      "entropy": 5.384450197219849,
      "epoch": 1.3022406918627505,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20785040259361268,
      "num_tokens": 25856912.0,
      "step": 11595
    },
    {
      "entropy": 5.475660371780395,
      "epoch": 1.3028022687707081,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.1122,
      "mean_token_accuracy": 0.19965678304433823,
      "num_tokens": 25868370.0,
      "step": 11600
    },
    {
      "entropy": 5.448839139938355,
      "epoch": 1.3033638456786658,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.1589,
      "mean_token_accuracy": 0.2014479711651802,
      "num_tokens": 25879104.0,
      "step": 11605
    },
    {
      "entropy": 5.47381625175476,
      "epoch": 1.3039254225866233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19480928480625154,
      "num_tokens": 25889744.0,
      "step": 11610
    },
    {
      "entropy": 5.597229719161987,
      "epoch": 1.3044869994945807,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.3176,
      "mean_token_accuracy": 0.19588677436113358,
      "num_tokens": 25901615.0,
      "step": 11615
    },
    {
      "entropy": 5.53926420211792,
      "epoch": 1.3050485764025384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1978822723031044,
      "num_tokens": 25911418.0,
      "step": 11620
    },
    {
      "entropy": 5.532394027709961,
      "epoch": 1.3056101533104958,
      "grad_norm": 1.0,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.246,
      "mean_token_accuracy": 0.2066519856452942,
      "num_tokens": 25922141.0,
      "step": 11625
    },
    {
      "entropy": 5.509350824356079,
      "epoch": 1.3061717302184535,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.2724,
      "mean_token_accuracy": 0.1987565651535988,
      "num_tokens": 25933897.0,
      "step": 11630
    },
    {
      "entropy": 5.541150426864624,
      "epoch": 1.306733307126411,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.1969,
      "mean_token_accuracy": 0.20249079018831254,
      "num_tokens": 25943960.0,
      "step": 11635
    },
    {
      "entropy": 5.656686401367187,
      "epoch": 1.3072948840343686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19893847405910492,
      "num_tokens": 25955104.0,
      "step": 11640
    },
    {
      "entropy": 5.49756007194519,
      "epoch": 1.307856460942326,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.2054,
      "mean_token_accuracy": 0.20584158152341842,
      "num_tokens": 25965180.0,
      "step": 11645
    },
    {
      "entropy": 5.446691274642944,
      "epoch": 1.3084180378502837,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.1933,
      "mean_token_accuracy": 0.20273812860250473,
      "num_tokens": 25976600.0,
      "step": 11650
    },
    {
      "entropy": 5.4639276504516605,
      "epoch": 1.308979614758241,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19740075767040252,
      "num_tokens": 25987509.0,
      "step": 11655
    },
    {
      "entropy": 5.501996850967407,
      "epoch": 1.3095411916661988,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000483916500180593,
      "loss": 5.1058,
      "mean_token_accuracy": 0.208722123503685,
      "num_tokens": 25997962.0,
      "step": 11660
    },
    {
      "entropy": 5.495789623260498,
      "epoch": 1.3101027685741562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20562392920255662,
      "num_tokens": 26008932.0,
      "step": 11665
    },
    {
      "entropy": 5.472488832473755,
      "epoch": 1.3106643454821139,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.1524,
      "mean_token_accuracy": 0.19838826954364777,
      "num_tokens": 26020513.0,
      "step": 11670
    },
    {
      "entropy": 5.527810049057007,
      "epoch": 1.3112259223900713,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.2205,
      "mean_token_accuracy": 0.20001866668462753,
      "num_tokens": 26032526.0,
      "step": 11675
    },
    {
      "entropy": 5.515231800079346,
      "epoch": 1.311787499298029,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.1587,
      "mean_token_accuracy": 0.21011221557855606,
      "num_tokens": 26044216.0,
      "step": 11680
    },
    {
      "entropy": 5.461994361877442,
      "epoch": 1.3123490762059864,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.0739,
      "mean_token_accuracy": 0.2038852721452713,
      "num_tokens": 26054012.0,
      "step": 11685
    },
    {
      "entropy": 5.501443767547608,
      "epoch": 1.3129106531139438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.2245,
      "mean_token_accuracy": 0.20137694329023362,
      "num_tokens": 26065365.0,
      "step": 11690
    },
    {
      "entropy": 5.535269784927368,
      "epoch": 1.3134722300219015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.2227,
      "mean_token_accuracy": 0.19394707679748535,
      "num_tokens": 26075964.0,
      "step": 11695
    },
    {
      "entropy": 5.47238245010376,
      "epoch": 1.3140338069298592,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.1886,
      "mean_token_accuracy": 0.20323828011751174,
      "num_tokens": 26087572.0,
      "step": 11700
    },
    {
      "entropy": 5.456265449523926,
      "epoch": 1.3145953838378166,
      "grad_norm": 0.90625,
      "learning_rate": 0.000483782070216197,
      "loss": 5.0896,
      "mean_token_accuracy": 0.21203955560922622,
      "num_tokens": 26098612.0,
      "step": 11705
    },
    {
      "entropy": 5.524162340164184,
      "epoch": 1.315156960745774,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.2467,
      "mean_token_accuracy": 0.2019554853439331,
      "num_tokens": 26109491.0,
      "step": 11710
    },
    {
      "entropy": 5.488994026184082,
      "epoch": 1.3157185376537317,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.1586,
      "mean_token_accuracy": 0.2042241796851158,
      "num_tokens": 26120544.0,
      "step": 11715
    },
    {
      "entropy": 5.5634520053863525,
      "epoch": 1.3162801145616891,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.2326,
      "mean_token_accuracy": 0.19634515643119813,
      "num_tokens": 26131549.0,
      "step": 11720
    },
    {
      "entropy": 5.498715353012085,
      "epoch": 1.3168416914696468,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.2147,
      "mean_token_accuracy": 0.20142530500888825,
      "num_tokens": 26142871.0,
      "step": 11725
    },
    {
      "entropy": 5.578309440612793,
      "epoch": 1.3174032683776042,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.3125,
      "mean_token_accuracy": 0.20300493240356446,
      "num_tokens": 26155293.0,
      "step": 11730
    },
    {
      "entropy": 5.582346057891845,
      "epoch": 1.317964845285562,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.2311,
      "mean_token_accuracy": 0.20663432031869888,
      "num_tokens": 26167069.0,
      "step": 11735
    },
    {
      "entropy": 5.52286958694458,
      "epoch": 1.3185264221935193,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.2327,
      "mean_token_accuracy": 0.1932014837861061,
      "num_tokens": 26179142.0,
      "step": 11740
    },
    {
      "entropy": 5.519941949844361,
      "epoch": 1.319087999101477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.1793,
      "mean_token_accuracy": 0.20495942384004592,
      "num_tokens": 26190059.0,
      "step": 11745
    },
    {
      "entropy": 5.526188611984253,
      "epoch": 1.3196495760094344,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.1906,
      "mean_token_accuracy": 0.2040617510676384,
      "num_tokens": 26199859.0,
      "step": 11750
    },
    {
      "entropy": 5.504745721817017,
      "epoch": 1.320211152917392,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19765446931123734,
      "num_tokens": 26211920.0,
      "step": 11755
    },
    {
      "entropy": 5.483478021621704,
      "epoch": 1.3207727298253495,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20338927954435349,
      "num_tokens": 26222875.0,
      "step": 11760
    },
    {
      "entropy": 5.520140743255615,
      "epoch": 1.3213343067333072,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.2179,
      "mean_token_accuracy": 0.1984965369105339,
      "num_tokens": 26233549.0,
      "step": 11765
    },
    {
      "entropy": 5.468645620346069,
      "epoch": 1.3218958836412646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.0896,
      "mean_token_accuracy": 0.21030196249485017,
      "num_tokens": 26243886.0,
      "step": 11770
    },
    {
      "entropy": 5.453647518157959,
      "epoch": 1.3224574605492223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.1309,
      "mean_token_accuracy": 0.21073497235774993,
      "num_tokens": 26254475.0,
      "step": 11775
    },
    {
      "entropy": 5.56497368812561,
      "epoch": 1.3230190374571797,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.3081,
      "mean_token_accuracy": 0.1974778339266777,
      "num_tokens": 26265412.0,
      "step": 11780
    },
    {
      "entropy": 5.5869546890258786,
      "epoch": 1.3235806143651372,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.2449,
      "mean_token_accuracy": 0.19708897918462753,
      "num_tokens": 26276942.0,
      "step": 11785
    },
    {
      "entropy": 5.557177925109864,
      "epoch": 1.3241421912730948,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.194,
      "mean_token_accuracy": 0.2028286412358284,
      "num_tokens": 26288009.0,
      "step": 11790
    },
    {
      "entropy": 5.553801345825195,
      "epoch": 1.3247037681810525,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.3319,
      "mean_token_accuracy": 0.19255428910255432,
      "num_tokens": 26301583.0,
      "step": 11795
    },
    {
      "entropy": 5.543572092056275,
      "epoch": 1.32526534508901,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.2437,
      "mean_token_accuracy": 0.2023691326379776,
      "num_tokens": 26312922.0,
      "step": 11800
    },
    {
      "entropy": 5.5360023021698,
      "epoch": 1.3258269219969674,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20770903676748276,
      "num_tokens": 26324108.0,
      "step": 11805
    },
    {
      "entropy": 5.545470190048218,
      "epoch": 1.326388498904925,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.2262,
      "mean_token_accuracy": 0.19926516711711884,
      "num_tokens": 26335669.0,
      "step": 11810
    },
    {
      "entropy": 5.383867454528809,
      "epoch": 1.3269500758128827,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.1138,
      "mean_token_accuracy": 0.20779256224632264,
      "num_tokens": 26346065.0,
      "step": 11815
    },
    {
      "entropy": 5.525509405136108,
      "epoch": 1.3275116527208402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.102,
      "mean_token_accuracy": 0.1990676388144493,
      "num_tokens": 26357567.0,
      "step": 11820
    },
    {
      "entropy": 5.49427638053894,
      "epoch": 1.3280732296287976,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.1416,
      "mean_token_accuracy": 0.20554784685373306,
      "num_tokens": 26367805.0,
      "step": 11825
    },
    {
      "entropy": 5.418798971176147,
      "epoch": 1.3286348065367553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20781477838754653,
      "num_tokens": 26379343.0,
      "step": 11830
    },
    {
      "entropy": 5.536299562454223,
      "epoch": 1.3291963834447127,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19708786010742188,
      "num_tokens": 26392039.0,
      "step": 11835
    },
    {
      "entropy": 5.381959390640259,
      "epoch": 1.3297579603526704,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000483375550934239,
      "loss": 5.0458,
      "mean_token_accuracy": 0.21690324395895005,
      "num_tokens": 26403324.0,
      "step": 11840
    },
    {
      "entropy": 5.444871187210083,
      "epoch": 1.3303195372606278,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.1544,
      "mean_token_accuracy": 0.2060822293162346,
      "num_tokens": 26413435.0,
      "step": 11845
    },
    {
      "entropy": 5.49003038406372,
      "epoch": 1.3308811141685855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19642059355974198,
      "num_tokens": 26424254.0,
      "step": 11850
    },
    {
      "entropy": 5.557097291946411,
      "epoch": 1.331442691076543,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.1582,
      "mean_token_accuracy": 0.20589336156845092,
      "num_tokens": 26435658.0,
      "step": 11855
    },
    {
      "entropy": 5.5792840957641605,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.2195,
      "mean_token_accuracy": 0.20413391292095184,
      "num_tokens": 26447174.0,
      "step": 11860
    },
    {
      "entropy": 5.5840263843536375,
      "epoch": 1.332565844892458,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.3123,
      "mean_token_accuracy": 0.19879010617733,
      "num_tokens": 26459303.0,
      "step": 11865
    },
    {
      "entropy": 5.476587772369385,
      "epoch": 1.3331274218004157,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.1793,
      "mean_token_accuracy": 0.20003268718719483,
      "num_tokens": 26469981.0,
      "step": 11870
    },
    {
      "entropy": 5.435283756256103,
      "epoch": 1.333688998708373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.1758,
      "mean_token_accuracy": 0.2076283097267151,
      "num_tokens": 26481176.0,
      "step": 11875
    },
    {
      "entropy": 5.479259634017945,
      "epoch": 1.3342505756163305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20606766790151596,
      "num_tokens": 26492795.0,
      "step": 11880
    },
    {
      "entropy": 5.5474122047424315,
      "epoch": 1.3348121525242882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.286,
      "mean_token_accuracy": 0.19940145760774614,
      "num_tokens": 26504731.0,
      "step": 11885
    },
    {
      "entropy": 5.575257778167725,
      "epoch": 1.3353737294322459,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.1515,
      "mean_token_accuracy": 0.20093888342380523,
      "num_tokens": 26516816.0,
      "step": 11890
    },
    {
      "entropy": 5.506833791732788,
      "epoch": 1.3359353063402033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.2674,
      "mean_token_accuracy": 0.19896857142448426,
      "num_tokens": 26528861.0,
      "step": 11895
    },
    {
      "entropy": 5.551325893402099,
      "epoch": 1.3364968832481607,
      "grad_norm": 1.0,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20493878573179244,
      "num_tokens": 26538816.0,
      "step": 11900
    },
    {
      "entropy": 5.6159669876098635,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.276,
      "mean_token_accuracy": 0.19954812973737718,
      "num_tokens": 26548959.0,
      "step": 11905
    },
    {
      "entropy": 5.404068946838379,
      "epoch": 1.337620037064076,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.1287,
      "mean_token_accuracy": 0.2054669052362442,
      "num_tokens": 26559189.0,
      "step": 11910
    },
    {
      "entropy": 5.415679693222046,
      "epoch": 1.3381816139720335,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.0987,
      "mean_token_accuracy": 0.2140517607331276,
      "num_tokens": 26570066.0,
      "step": 11915
    },
    {
      "entropy": 5.536645984649658,
      "epoch": 1.338743190879991,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.272,
      "mean_token_accuracy": 0.2038959488272667,
      "num_tokens": 26581489.0,
      "step": 11920
    },
    {
      "entropy": 5.551652717590332,
      "epoch": 1.3393047677879486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.1944,
      "mean_token_accuracy": 0.19723350405693055,
      "num_tokens": 26592051.0,
      "step": 11925
    },
    {
      "entropy": 5.513260984420777,
      "epoch": 1.339866344695906,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.2013,
      "mean_token_accuracy": 0.2009888008236885,
      "num_tokens": 26602296.0,
      "step": 11930
    },
    {
      "entropy": 5.490020847320556,
      "epoch": 1.3404279216038637,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.1805,
      "mean_token_accuracy": 0.19477375745773315,
      "num_tokens": 26613837.0,
      "step": 11935
    },
    {
      "entropy": 5.498483037948608,
      "epoch": 1.3409894985118211,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.2277,
      "mean_token_accuracy": 0.20022510588169098,
      "num_tokens": 26624898.0,
      "step": 11940
    },
    {
      "entropy": 5.512249898910523,
      "epoch": 1.3415510754197788,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.218,
      "mean_token_accuracy": 0.2028593271970749,
      "num_tokens": 26636449.0,
      "step": 11945
    },
    {
      "entropy": 5.635387659072876,
      "epoch": 1.3421126523277362,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1958254098892212,
      "num_tokens": 26649192.0,
      "step": 11950
    },
    {
      "entropy": 5.497179365158081,
      "epoch": 1.342674229235694,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000483025441381746,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20541906505823135,
      "num_tokens": 26660005.0,
      "step": 11955
    },
    {
      "entropy": 5.572161245346069,
      "epoch": 1.3432358061436513,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.2693,
      "mean_token_accuracy": 0.19759068489074708,
      "num_tokens": 26672165.0,
      "step": 11960
    },
    {
      "entropy": 5.480876922607422,
      "epoch": 1.343797383051609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.1981,
      "mean_token_accuracy": 0.2068539321422577,
      "num_tokens": 26683135.0,
      "step": 11965
    },
    {
      "entropy": 5.464125299453736,
      "epoch": 1.3443589599595664,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.0889,
      "mean_token_accuracy": 0.2124148964881897,
      "num_tokens": 26693724.0,
      "step": 11970
    },
    {
      "entropy": 5.55276346206665,
      "epoch": 1.3449205368675239,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.3021,
      "mean_token_accuracy": 0.20056399405002595,
      "num_tokens": 26705880.0,
      "step": 11975
    },
    {
      "entropy": 5.565187931060791,
      "epoch": 1.3454821137754815,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.2742,
      "mean_token_accuracy": 0.20115237385034562,
      "num_tokens": 26716978.0,
      "step": 11980
    },
    {
      "entropy": 5.594359588623047,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.1958,
      "mean_token_accuracy": 0.21034796684980392,
      "num_tokens": 26728380.0,
      "step": 11985
    },
    {
      "entropy": 5.428484869003296,
      "epoch": 1.3466052675913966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20786663442850112,
      "num_tokens": 26739034.0,
      "step": 11990
    },
    {
      "entropy": 5.34709062576294,
      "epoch": 1.347166844499354,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004829028425172487,
      "loss": 4.9786,
      "mean_token_accuracy": 0.21418415904045104,
      "num_tokens": 26750123.0,
      "step": 11995
    },
    {
      "entropy": 5.446548175811768,
      "epoch": 1.3477284214073117,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.1512,
      "mean_token_accuracy": 0.21165154874324799,
      "num_tokens": 26760411.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.274218426926529,
      "eval_loss": 5.258620262145996,
      "eval_mean_token_accuracy": 0.20713500018229888,
      "eval_num_tokens": 26760411.0,
      "eval_runtime": 18.5091,
      "eval_samples_per_second": 1675.559,
      "eval_steps_per_second": 209.465,
      "step": 12000
    },
    {
      "entropy": 5.49584174156189,
      "epoch": 1.3482899983152694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.2167,
      "mean_token_accuracy": 0.20441518276929854,
      "num_tokens": 26770821.0,
      "step": 12005
    },
    {
      "entropy": 5.545044231414795,
      "epoch": 1.3488515752232269,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.186,
      "mean_token_accuracy": 0.20576249659061432,
      "num_tokens": 26782099.0,
      "step": 12010
    },
    {
      "entropy": 5.570070695877075,
      "epoch": 1.3494131521311843,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.3127,
      "mean_token_accuracy": 0.20108823627233505,
      "num_tokens": 26794056.0,
      "step": 12015
    },
    {
      "entropy": 5.472179985046386,
      "epoch": 1.349974729039142,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20144221186637878,
      "num_tokens": 26804501.0,
      "step": 12020
    },
    {
      "entropy": 5.588758373260498,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19861800372600555,
      "num_tokens": 26814996.0,
      "step": 12025
    },
    {
      "entropy": 5.513352680206299,
      "epoch": 1.351097882855057,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20483455210924148,
      "num_tokens": 26825838.0,
      "step": 12030
    },
    {
      "entropy": 5.505828189849853,
      "epoch": 1.3516594597630145,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.1729,
      "mean_token_accuracy": 0.20478046983480452,
      "num_tokens": 26836320.0,
      "step": 12035
    },
    {
      "entropy": 5.520681285858155,
      "epoch": 1.3522210366709722,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.1524,
      "mean_token_accuracy": 0.2102282702922821,
      "num_tokens": 26846485.0,
      "step": 12040
    },
    {
      "entropy": 5.596781015396118,
      "epoch": 1.3527826135789296,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.273,
      "mean_token_accuracy": 0.19390378296375274,
      "num_tokens": 26858554.0,
      "step": 12045
    },
    {
      "entropy": 5.5295757293701175,
      "epoch": 1.3533441904868873,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.2088,
      "mean_token_accuracy": 0.20490582287311554,
      "num_tokens": 26869538.0,
      "step": 12050
    },
    {
      "entropy": 5.5779510021209715,
      "epoch": 1.3539057673948447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.2778,
      "mean_token_accuracy": 0.195449997484684,
      "num_tokens": 26880284.0,
      "step": 12055
    },
    {
      "entropy": 5.5082540035247805,
      "epoch": 1.3544673443028024,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19591475278139114,
      "num_tokens": 26893180.0,
      "step": 12060
    },
    {
      "entropy": 5.455277299880981,
      "epoch": 1.3550289212107598,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.108,
      "mean_token_accuracy": 0.2091502860188484,
      "num_tokens": 26904213.0,
      "step": 12065
    },
    {
      "entropy": 5.504127120971679,
      "epoch": 1.3555904981187172,
      "grad_norm": 0.875,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.2092,
      "mean_token_accuracy": 0.2075573280453682,
      "num_tokens": 26915566.0,
      "step": 12070
    },
    {
      "entropy": 5.424219560623169,
      "epoch": 1.356152075026675,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.15,
      "mean_token_accuracy": 0.20781825482845306,
      "num_tokens": 26926253.0,
      "step": 12075
    },
    {
      "entropy": 5.456980562210083,
      "epoch": 1.3567136519346326,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.1403,
      "mean_token_accuracy": 0.21102577596902847,
      "num_tokens": 26936723.0,
      "step": 12080
    },
    {
      "entropy": 5.432686233520508,
      "epoch": 1.35727522884259,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20442180782556535,
      "num_tokens": 26947551.0,
      "step": 12085
    },
    {
      "entropy": 5.487112283706665,
      "epoch": 1.3578368057505474,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.242,
      "mean_token_accuracy": 0.20474966019392013,
      "num_tokens": 26959969.0,
      "step": 12090
    },
    {
      "entropy": 5.613993740081787,
      "epoch": 1.358398382658505,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.3215,
      "mean_token_accuracy": 0.19969538301229478,
      "num_tokens": 26972511.0,
      "step": 12095
    },
    {
      "entropy": 5.568914222717285,
      "epoch": 1.3589599595664628,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1948620930314064,
      "num_tokens": 26983827.0,
      "step": 12100
    },
    {
      "entropy": 5.600088834762573,
      "epoch": 1.3595215364744202,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1906919702887535,
      "num_tokens": 26994668.0,
      "step": 12105
    },
    {
      "entropy": 5.481191349029541,
      "epoch": 1.3600831133823776,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.1222,
      "mean_token_accuracy": 0.2137640029191971,
      "num_tokens": 27006313.0,
      "step": 12110
    },
    {
      "entropy": 5.497007656097412,
      "epoch": 1.3606446902903353,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19987176358699799,
      "num_tokens": 27017628.0,
      "step": 12115
    },
    {
      "entropy": 5.489603281021118,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.0,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20417182743549347,
      "num_tokens": 27028303.0,
      "step": 12120
    },
    {
      "entropy": 5.553450107574463,
      "epoch": 1.3617678441062504,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.2326,
      "mean_token_accuracy": 0.19935499429702758,
      "num_tokens": 27039307.0,
      "step": 12125
    },
    {
      "entropy": 5.484052276611328,
      "epoch": 1.3623294210142078,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20457243770360947,
      "num_tokens": 27049271.0,
      "step": 12130
    },
    {
      "entropy": 5.396960115432739,
      "epoch": 1.3628909979221655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2087179660797119,
      "num_tokens": 27059510.0,
      "step": 12135
    },
    {
      "entropy": 5.607878160476685,
      "epoch": 1.363452574830123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19065277874469758,
      "num_tokens": 27071685.0,
      "step": 12140
    },
    {
      "entropy": 5.460358619689941,
      "epoch": 1.3640141517380806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.0925,
      "mean_token_accuracy": 0.2005204364657402,
      "num_tokens": 27081957.0,
      "step": 12145
    },
    {
      "entropy": 5.44541563987732,
      "epoch": 1.364575728646038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20689482092857361,
      "num_tokens": 27091609.0,
      "step": 12150
    },
    {
      "entropy": 5.514796733856201,
      "epoch": 1.3651373055539957,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.2867,
      "mean_token_accuracy": 0.19851918369531632,
      "num_tokens": 27103889.0,
      "step": 12155
    },
    {
      "entropy": 5.573773431777954,
      "epoch": 1.3656988824619531,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000482392646735433,
      "loss": 5.224,
      "mean_token_accuracy": 0.20402745753526688,
      "num_tokens": 27114186.0,
      "step": 12160
    },
    {
      "entropy": 5.417332172393799,
      "epoch": 1.3662604593699106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20940093100070953,
      "num_tokens": 27124677.0,
      "step": 12165
    },
    {
      "entropy": 5.616716432571411,
      "epoch": 1.3668220362778682,
      "grad_norm": 1.0,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.2609,
      "mean_token_accuracy": 0.20400604009628295,
      "num_tokens": 27134873.0,
      "step": 12170
    },
    {
      "entropy": 5.516099214553833,
      "epoch": 1.367383613185826,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.249,
      "mean_token_accuracy": 0.20054423809051514,
      "num_tokens": 27146107.0,
      "step": 12175
    },
    {
      "entropy": 5.465147638320923,
      "epoch": 1.3679451900937833,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21363035291433335,
      "num_tokens": 27158821.0,
      "step": 12180
    },
    {
      "entropy": 5.4561059951782225,
      "epoch": 1.3685067670017408,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.128,
      "mean_token_accuracy": 0.20864364504814148,
      "num_tokens": 27169513.0,
      "step": 12185
    },
    {
      "entropy": 5.440722322463989,
      "epoch": 1.3690683439096984,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20908154249191285,
      "num_tokens": 27180059.0,
      "step": 12190
    },
    {
      "entropy": 5.451929140090942,
      "epoch": 1.369629920817656,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.1952,
      "mean_token_accuracy": 0.20354367643594742,
      "num_tokens": 27190738.0,
      "step": 12195
    },
    {
      "entropy": 5.478823852539063,
      "epoch": 1.3701914977256135,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.1031,
      "mean_token_accuracy": 0.20554327666759492,
      "num_tokens": 27200664.0,
      "step": 12200
    },
    {
      "entropy": 5.490411996841431,
      "epoch": 1.370753074633571,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20783050656318663,
      "num_tokens": 27211011.0,
      "step": 12205
    },
    {
      "entropy": 5.510783004760742,
      "epoch": 1.3713146515415287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.2246,
      "mean_token_accuracy": 0.20689368546009063,
      "num_tokens": 27221888.0,
      "step": 12210
    },
    {
      "entropy": 5.558411264419556,
      "epoch": 1.371876228449486,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.2224,
      "mean_token_accuracy": 0.2004938989877701,
      "num_tokens": 27234671.0,
      "step": 12215
    },
    {
      "entropy": 5.430681133270264,
      "epoch": 1.3724378053574438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.012,
      "mean_token_accuracy": 0.21147095710039138,
      "num_tokens": 27245283.0,
      "step": 12220
    },
    {
      "entropy": 5.477332019805909,
      "epoch": 1.3729993822654012,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.1789,
      "mean_token_accuracy": 0.20685552060604095,
      "num_tokens": 27255056.0,
      "step": 12225
    },
    {
      "entropy": 5.53238844871521,
      "epoch": 1.3735609591733589,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000482174026472802,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1991251900792122,
      "num_tokens": 27266893.0,
      "step": 12230
    },
    {
      "entropy": 5.480065822601318,
      "epoch": 1.3741225360813163,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.2418,
      "mean_token_accuracy": 0.20151694267988204,
      "num_tokens": 27279504.0,
      "step": 12235
    },
    {
      "entropy": 5.604198980331421,
      "epoch": 1.374684112989274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.2813,
      "mean_token_accuracy": 0.20343851894140244,
      "num_tokens": 27290525.0,
      "step": 12240
    },
    {
      "entropy": 5.549470663070679,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.1713,
      "mean_token_accuracy": 0.20547517985105515,
      "num_tokens": 27301404.0,
      "step": 12245
    },
    {
      "entropy": 5.490599536895752,
      "epoch": 1.375807266805189,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20582991987466812,
      "num_tokens": 27312635.0,
      "step": 12250
    },
    {
      "entropy": 5.429378271102905,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.0,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.1456,
      "mean_token_accuracy": 0.20946708023548127,
      "num_tokens": 27322485.0,
      "step": 12255
    },
    {
      "entropy": 5.5102893829345705,
      "epoch": 1.376930420621104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.2121,
      "mean_token_accuracy": 0.20140142142772674,
      "num_tokens": 27332568.0,
      "step": 12260
    },
    {
      "entropy": 5.633270788192749,
      "epoch": 1.3774919975290616,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19324231147766113,
      "num_tokens": 27343058.0,
      "step": 12265
    },
    {
      "entropy": 5.496754789352417,
      "epoch": 1.3780535744370193,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.1256,
      "mean_token_accuracy": 0.21085395365953447,
      "num_tokens": 27355316.0,
      "step": 12270
    },
    {
      "entropy": 5.436333703994751,
      "epoch": 1.3786151513449767,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21318896114826202,
      "num_tokens": 27366098.0,
      "step": 12275
    },
    {
      "entropy": 5.4514447212219235,
      "epoch": 1.3791767282529341,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21291046291589738,
      "num_tokens": 27376923.0,
      "step": 12280
    },
    {
      "entropy": 5.475803232192993,
      "epoch": 1.3797383051608918,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.2409,
      "mean_token_accuracy": 0.20171729028224944,
      "num_tokens": 27386789.0,
      "step": 12285
    },
    {
      "entropy": 5.492842960357666,
      "epoch": 1.3802998820688495,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000481985608570654,
      "loss": 5.1315,
      "mean_token_accuracy": 0.2056084468960762,
      "num_tokens": 27399241.0,
      "step": 12290
    },
    {
      "entropy": 5.4294211864471436,
      "epoch": 1.380861458976807,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21491445153951644,
      "num_tokens": 27409123.0,
      "step": 12295
    },
    {
      "entropy": 5.405773019790649,
      "epoch": 1.3814230358847643,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.1109,
      "mean_token_accuracy": 0.204316608607769,
      "num_tokens": 27419920.0,
      "step": 12300
    },
    {
      "entropy": 5.380084657669068,
      "epoch": 1.381984612792722,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21066481471061707,
      "num_tokens": 27429552.0,
      "step": 12305
    },
    {
      "entropy": 5.513218736648559,
      "epoch": 1.3825461897006794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.2212,
      "mean_token_accuracy": 0.20285790115594865,
      "num_tokens": 27440404.0,
      "step": 12310
    },
    {
      "entropy": 5.514640665054321,
      "epoch": 1.383107766608637,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19949701577425002,
      "num_tokens": 27452231.0,
      "step": 12315
    },
    {
      "entropy": 5.472192668914795,
      "epoch": 1.3836693435165945,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.1523,
      "mean_token_accuracy": 0.2082923337817192,
      "num_tokens": 27463658.0,
      "step": 12320
    },
    {
      "entropy": 5.50851469039917,
      "epoch": 1.3842309204245522,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.1874,
      "mean_token_accuracy": 0.20177329033613206,
      "num_tokens": 27475403.0,
      "step": 12325
    },
    {
      "entropy": 5.5778203964233395,
      "epoch": 1.3847924973325096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000481859469342418,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19922492653131485,
      "num_tokens": 27487056.0,
      "step": 12330
    },
    {
      "entropy": 5.472536849975586,
      "epoch": 1.3853540742404673,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.1364,
      "mean_token_accuracy": 0.21165980100631715,
      "num_tokens": 27498659.0,
      "step": 12335
    },
    {
      "entropy": 5.58970537185669,
      "epoch": 1.3859156511484247,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.273,
      "mean_token_accuracy": 0.19543916136026382,
      "num_tokens": 27510577.0,
      "step": 12340
    },
    {
      "entropy": 5.479802656173706,
      "epoch": 1.3864772280563824,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.1778,
      "mean_token_accuracy": 0.20541457682847977,
      "num_tokens": 27521441.0,
      "step": 12345
    },
    {
      "entropy": 5.505863189697266,
      "epoch": 1.3870388049643398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.1129,
      "mean_token_accuracy": 0.2090619370341301,
      "num_tokens": 27533443.0,
      "step": 12350
    },
    {
      "entropy": 5.5502166748046875,
      "epoch": 1.3876003818722973,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.1728,
      "mean_token_accuracy": 0.20822193920612336,
      "num_tokens": 27543872.0,
      "step": 12355
    },
    {
      "entropy": 5.608455562591553,
      "epoch": 1.388161958780255,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.3607,
      "mean_token_accuracy": 0.19590135514736176,
      "num_tokens": 27556556.0,
      "step": 12360
    },
    {
      "entropy": 5.528447914123535,
      "epoch": 1.3887235356882126,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.2253,
      "mean_token_accuracy": 0.1963939517736435,
      "num_tokens": 27570739.0,
      "step": 12365
    },
    {
      "entropy": 5.526506614685059,
      "epoch": 1.38928511259617,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.2756,
      "mean_token_accuracy": 0.19511748850345612,
      "num_tokens": 27582438.0,
      "step": 12370
    },
    {
      "entropy": 5.557900667190552,
      "epoch": 1.3898466895041275,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20703279674053193,
      "num_tokens": 27593854.0,
      "step": 12375
    },
    {
      "entropy": 5.553404235839844,
      "epoch": 1.3904082664120851,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20735314190387727,
      "num_tokens": 27604459.0,
      "step": 12380
    },
    {
      "entropy": 5.572381019592285,
      "epoch": 1.3909698433200428,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18455785512924194,
      "num_tokens": 27617901.0,
      "step": 12385
    },
    {
      "entropy": 5.5417852878570555,
      "epoch": 1.3915314202280002,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.2825,
      "mean_token_accuracy": 0.1957176372408867,
      "num_tokens": 27628574.0,
      "step": 12390
    },
    {
      "entropy": 5.530961227416992,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20285152047872543,
      "num_tokens": 27639195.0,
      "step": 12395
    },
    {
      "entropy": 5.4167218685150145,
      "epoch": 1.3926545740439153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2165919229388237,
      "num_tokens": 27649898.0,
      "step": 12400
    },
    {
      "entropy": 5.508417558670044,
      "epoch": 1.3932161509518728,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000481621822323602,
      "loss": 5.2052,
      "mean_token_accuracy": 0.2018896833062172,
      "num_tokens": 27663617.0,
      "step": 12405
    },
    {
      "entropy": 5.423588657379151,
      "epoch": 1.3937777278598305,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.0705,
      "mean_token_accuracy": 0.21269573420286178,
      "num_tokens": 27674823.0,
      "step": 12410
    },
    {
      "entropy": 5.504005384445191,
      "epoch": 1.394339304767788,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.1841,
      "mean_token_accuracy": 0.20515568256378175,
      "num_tokens": 27686483.0,
      "step": 12415
    },
    {
      "entropy": 5.490431642532348,
      "epoch": 1.3949008816757456,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.122,
      "mean_token_accuracy": 0.2079084113240242,
      "num_tokens": 27697081.0,
      "step": 12420
    },
    {
      "entropy": 5.507756662368775,
      "epoch": 1.395462458583703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.2112,
      "mean_token_accuracy": 0.19897021502256393,
      "num_tokens": 27706710.0,
      "step": 12425
    },
    {
      "entropy": 5.519643878936767,
      "epoch": 1.3960240354916607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.2398,
      "mean_token_accuracy": 0.1945173129439354,
      "num_tokens": 27718213.0,
      "step": 12430
    },
    {
      "entropy": 5.438825988769532,
      "epoch": 1.396585612399618,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19906750917434693,
      "num_tokens": 27729064.0,
      "step": 12435
    },
    {
      "entropy": 5.564300060272217,
      "epoch": 1.3971471893075758,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20962911248207092,
      "num_tokens": 27740121.0,
      "step": 12440
    },
    {
      "entropy": 5.45766315460205,
      "epoch": 1.3977087662155332,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.2155,
      "mean_token_accuracy": 0.21042629778385163,
      "num_tokens": 27751597.0,
      "step": 12445
    },
    {
      "entropy": 5.435930442810059,
      "epoch": 1.3982703431234906,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20711494088172913,
      "num_tokens": 27761539.0,
      "step": 12450
    },
    {
      "entropy": 5.517765665054322,
      "epoch": 1.3988319200314483,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20414216220378875,
      "num_tokens": 27772904.0,
      "step": 12455
    },
    {
      "entropy": 5.551731586456299,
      "epoch": 1.399393496939406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.13,
      "mean_token_accuracy": 0.20920319855213165,
      "num_tokens": 27783706.0,
      "step": 12460
    },
    {
      "entropy": 5.456592273712158,
      "epoch": 1.3999550738473634,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.1406,
      "mean_token_accuracy": 0.20146947503089904,
      "num_tokens": 27794503.0,
      "step": 12465
    },
    {
      "entropy": 5.456950998306274,
      "epoch": 1.4005166507553208,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19991386383771897,
      "num_tokens": 27805952.0,
      "step": 12470
    },
    {
      "entropy": 5.497203016281128,
      "epoch": 1.4010782276632785,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.1487,
      "mean_token_accuracy": 0.20707900822162628,
      "num_tokens": 27817530.0,
      "step": 12475
    },
    {
      "entropy": 5.503857421875,
      "epoch": 1.4016398045712362,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1953331157565117,
      "num_tokens": 27829932.0,
      "step": 12480
    },
    {
      "entropy": 5.493107748031616,
      "epoch": 1.4022013814791936,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.197,
      "mean_token_accuracy": 0.2065885215997696,
      "num_tokens": 27840993.0,
      "step": 12485
    },
    {
      "entropy": 5.469111394882202,
      "epoch": 1.402762958387151,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20760075002908707,
      "num_tokens": 27852727.0,
      "step": 12490
    },
    {
      "entropy": 5.51702733039856,
      "epoch": 1.4033245352951087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.2966,
      "mean_token_accuracy": 0.20161188393831253,
      "num_tokens": 27865183.0,
      "step": 12495
    },
    {
      "entropy": 5.433106803894043,
      "epoch": 1.4038861122030661,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21434956640005112,
      "num_tokens": 27877266.0,
      "step": 12500
    },
    {
      "entropy": 5.536842346191406,
      "epoch": 1.4044476891110238,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1966889724135399,
      "num_tokens": 27888863.0,
      "step": 12505
    },
    {
      "entropy": 5.505915689468384,
      "epoch": 1.4050092660189812,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.11,
      "mean_token_accuracy": 0.20910412520170213,
      "num_tokens": 27900443.0,
      "step": 12510
    },
    {
      "entropy": 5.461970472335816,
      "epoch": 1.405570842926939,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.1129,
      "mean_token_accuracy": 0.2036841481924057,
      "num_tokens": 27911241.0,
      "step": 12515
    },
    {
      "entropy": 5.492799711227417,
      "epoch": 1.4061324198348963,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.1398,
      "mean_token_accuracy": 0.21027563661336898,
      "num_tokens": 27922171.0,
      "step": 12520
    },
    {
      "entropy": 5.438047981262207,
      "epoch": 1.406693996742854,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.1115,
      "mean_token_accuracy": 0.2091817706823349,
      "num_tokens": 27933627.0,
      "step": 12525
    },
    {
      "entropy": 5.380120515823364,
      "epoch": 1.4072555736508114,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.0651,
      "mean_token_accuracy": 0.2088152915239334,
      "num_tokens": 27943585.0,
      "step": 12530
    },
    {
      "entropy": 5.521344137191773,
      "epoch": 1.407817150558769,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20321394205093385,
      "num_tokens": 27955554.0,
      "step": 12535
    },
    {
      "entropy": 5.52826156616211,
      "epoch": 1.4083787274667265,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20730353146791458,
      "num_tokens": 27966150.0,
      "step": 12540
    },
    {
      "entropy": 5.42886004447937,
      "epoch": 1.408940304374684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20883988887071608,
      "num_tokens": 27977436.0,
      "step": 12545
    },
    {
      "entropy": 5.51380558013916,
      "epoch": 1.4095018812826416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19582959562540053,
      "num_tokens": 27988325.0,
      "step": 12550
    },
    {
      "entropy": 5.548657083511353,
      "epoch": 1.4100634581905993,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20544904321432114,
      "num_tokens": 27999760.0,
      "step": 12555
    },
    {
      "entropy": 5.549814987182617,
      "epoch": 1.4106250350985567,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19435268491506577,
      "num_tokens": 28011333.0,
      "step": 12560
    },
    {
      "entropy": 5.490108585357666,
      "epoch": 1.4111866120065142,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.2129,
      "mean_token_accuracy": 0.20457708835601807,
      "num_tokens": 28023464.0,
      "step": 12565
    },
    {
      "entropy": 5.471909284591675,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000481093790794317,
      "loss": 5.1587,
      "mean_token_accuracy": 0.1978415697813034,
      "num_tokens": 28034205.0,
      "step": 12570
    },
    {
      "entropy": 5.5302345752716064,
      "epoch": 1.4123097658224295,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.1367,
      "mean_token_accuracy": 0.2024836927652359,
      "num_tokens": 28045846.0,
      "step": 12575
    },
    {
      "entropy": 5.526364135742187,
      "epoch": 1.412871342730387,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.1583,
      "mean_token_accuracy": 0.21020425707101822,
      "num_tokens": 28057365.0,
      "step": 12580
    },
    {
      "entropy": 5.5025975704193115,
      "epoch": 1.4134329196383444,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.2494,
      "mean_token_accuracy": 0.1981061577796936,
      "num_tokens": 28069098.0,
      "step": 12585
    },
    {
      "entropy": 5.500232601165772,
      "epoch": 1.413994496546302,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.1922,
      "mean_token_accuracy": 0.20091965049505234,
      "num_tokens": 28080008.0,
      "step": 12590
    },
    {
      "entropy": 5.5405395984649655,
      "epoch": 1.4145560734542595,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.2003,
      "mean_token_accuracy": 0.20463268905878068,
      "num_tokens": 28091657.0,
      "step": 12595
    },
    {
      "entropy": 5.4484052658081055,
      "epoch": 1.4151176503622171,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.102,
      "mean_token_accuracy": 0.20200641304254532,
      "num_tokens": 28102586.0,
      "step": 12600
    },
    {
      "entropy": 5.514485645294189,
      "epoch": 1.4156792272701746,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.1841,
      "mean_token_accuracy": 0.2082786574959755,
      "num_tokens": 28113066.0,
      "step": 12605
    },
    {
      "entropy": 5.494267177581787,
      "epoch": 1.4162408041781323,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20425644367933274,
      "num_tokens": 28124814.0,
      "step": 12610
    },
    {
      "entropy": 5.570365476608276,
      "epoch": 1.4168023810860897,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.234,
      "mean_token_accuracy": 0.19599348902702332,
      "num_tokens": 28138009.0,
      "step": 12615
    },
    {
      "entropy": 5.429140090942383,
      "epoch": 1.4173639579940474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.12,
      "mean_token_accuracy": 0.20635987520217897,
      "num_tokens": 28149222.0,
      "step": 12620
    },
    {
      "entropy": 5.478396081924439,
      "epoch": 1.4179255349020048,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.1646,
      "mean_token_accuracy": 0.2032903775572777,
      "num_tokens": 28161323.0,
      "step": 12625
    },
    {
      "entropy": 5.484135007858276,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20095252245664597,
      "num_tokens": 28173242.0,
      "step": 12630
    },
    {
      "entropy": 5.503753519058227,
      "epoch": 1.41904868871792,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.1144,
      "mean_token_accuracy": 0.20571043938398362,
      "num_tokens": 28183927.0,
      "step": 12635
    },
    {
      "entropy": 5.518493175506592,
      "epoch": 1.4196102656258773,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.1279,
      "mean_token_accuracy": 0.20653288513422013,
      "num_tokens": 28194375.0,
      "step": 12640
    },
    {
      "entropy": 5.49633936882019,
      "epoch": 1.420171842533835,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.1963,
      "mean_token_accuracy": 0.21066411584615707,
      "num_tokens": 28206492.0,
      "step": 12645
    },
    {
      "entropy": 5.497230100631714,
      "epoch": 1.4207334194417927,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.2054,
      "mean_token_accuracy": 0.19970814734697342,
      "num_tokens": 28217795.0,
      "step": 12650
    },
    {
      "entropy": 5.478448486328125,
      "epoch": 1.42129499634975,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.1575,
      "mean_token_accuracy": 0.2054598331451416,
      "num_tokens": 28228796.0,
      "step": 12655
    },
    {
      "entropy": 5.502583074569702,
      "epoch": 1.4218565732577075,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.2021,
      "mean_token_accuracy": 0.20309112071990967,
      "num_tokens": 28238873.0,
      "step": 12660
    },
    {
      "entropy": 5.43205246925354,
      "epoch": 1.4224181501656652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048078652978521914,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21628990322351455,
      "num_tokens": 28249832.0,
      "step": 12665
    },
    {
      "entropy": 5.490887022018432,
      "epoch": 1.4229797270736229,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.195,
      "mean_token_accuracy": 0.19961034059524535,
      "num_tokens": 28261532.0,
      "step": 12670
    },
    {
      "entropy": 5.594787883758545,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.0,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.2067,
      "mean_token_accuracy": 0.20420225113630294,
      "num_tokens": 28272008.0,
      "step": 12675
    },
    {
      "entropy": 5.494456052780151,
      "epoch": 1.4241028808895377,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.1799,
      "mean_token_accuracy": 0.2064451351761818,
      "num_tokens": 28283239.0,
      "step": 12680
    },
    {
      "entropy": 5.519506645202637,
      "epoch": 1.4246644577974954,
      "grad_norm": 0.953125,
      "learning_rate": 0.000480721541731488,
      "loss": 5.2277,
      "mean_token_accuracy": 0.20041002184152604,
      "num_tokens": 28293597.0,
      "step": 12685
    },
    {
      "entropy": 5.517670488357544,
      "epoch": 1.4252260347054528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20422108322381974,
      "num_tokens": 28305124.0,
      "step": 12690
    },
    {
      "entropy": 5.441624116897583,
      "epoch": 1.4257876116134105,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.16,
      "mean_token_accuracy": 0.20522437244653702,
      "num_tokens": 28315875.0,
      "step": 12695
    },
    {
      "entropy": 5.462756729125976,
      "epoch": 1.426349188521368,
      "grad_norm": 0.875,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.1532,
      "mean_token_accuracy": 0.20602448433637618,
      "num_tokens": 28327705.0,
      "step": 12700
    },
    {
      "entropy": 5.524948644638061,
      "epoch": 1.4269107654293256,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21232459247112273,
      "num_tokens": 28338300.0,
      "step": 12705
    },
    {
      "entropy": 5.473148488998413,
      "epoch": 1.427472342337283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.072,
      "mean_token_accuracy": 0.21225128918886185,
      "num_tokens": 28348485.0,
      "step": 12710
    },
    {
      "entropy": 5.459632110595703,
      "epoch": 1.4280339192452407,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.1,
      "mean_token_accuracy": 0.21192515939474105,
      "num_tokens": 28359118.0,
      "step": 12715
    },
    {
      "entropy": 5.4481199264526365,
      "epoch": 1.4285954961531981,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.1465,
      "mean_token_accuracy": 0.21156249493360518,
      "num_tokens": 28369608.0,
      "step": 12720
    },
    {
      "entropy": 5.541545152664185,
      "epoch": 1.4291570730611558,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.1329,
      "mean_token_accuracy": 0.21158370077610017,
      "num_tokens": 28379709.0,
      "step": 12725
    },
    {
      "entropy": 5.49413104057312,
      "epoch": 1.4297186499691132,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20483034998178482,
      "num_tokens": 28390758.0,
      "step": 12730
    },
    {
      "entropy": 5.660143709182739,
      "epoch": 1.4302802268770707,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000480558613151976,
      "loss": 5.4104,
      "mean_token_accuracy": 0.187820503115654,
      "num_tokens": 28403485.0,
      "step": 12735
    },
    {
      "entropy": 5.575120067596435,
      "epoch": 1.4308418037850283,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.2123,
      "mean_token_accuracy": 0.20165627896785737,
      "num_tokens": 28413627.0,
      "step": 12740
    },
    {
      "entropy": 5.5053736686706545,
      "epoch": 1.431403380692986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.1137,
      "mean_token_accuracy": 0.2040541648864746,
      "num_tokens": 28424202.0,
      "step": 12745
    },
    {
      "entropy": 5.37389907836914,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.0788,
      "mean_token_accuracy": 0.21305925995111466,
      "num_tokens": 28435136.0,
      "step": 12750
    },
    {
      "entropy": 5.453747844696045,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.015625,
      "learning_rate": 0.000480493258429251,
      "loss": 5.1113,
      "mean_token_accuracy": 0.20179128497838975,
      "num_tokens": 28444915.0,
      "step": 12755
    },
    {
      "entropy": 5.510418653488159,
      "epoch": 1.4330881114168585,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000480476903389806,
      "loss": 5.1768,
      "mean_token_accuracy": 0.19906016141176225,
      "num_tokens": 28458295.0,
      "step": 12760
    },
    {
      "entropy": 5.546279048919677,
      "epoch": 1.4336496883248162,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.2199,
      "mean_token_accuracy": 0.20799126029014586,
      "num_tokens": 28469587.0,
      "step": 12765
    },
    {
      "entropy": 5.5824261665344235,
      "epoch": 1.4342112652327736,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.1956,
      "mean_token_accuracy": 0.202959780395031,
      "num_tokens": 28481498.0,
      "step": 12770
    },
    {
      "entropy": 5.572489547729492,
      "epoch": 1.434772842140731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1933248370885849,
      "num_tokens": 28492626.0,
      "step": 12775
    },
    {
      "entropy": 5.545279026031494,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20513793975114822,
      "num_tokens": 28503882.0,
      "step": 12780
    },
    {
      "entropy": 5.492973470687867,
      "epoch": 1.4358959959566462,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.1208,
      "mean_token_accuracy": 0.21698886007070542,
      "num_tokens": 28514786.0,
      "step": 12785
    },
    {
      "entropy": 5.523755264282227,
      "epoch": 1.4364575728646038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.153,
      "mean_token_accuracy": 0.2071579724550247,
      "num_tokens": 28526045.0,
      "step": 12790
    },
    {
      "entropy": 5.518328285217285,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.1604,
      "mean_token_accuracy": 0.2059785768389702,
      "num_tokens": 28536773.0,
      "step": 12795
    },
    {
      "entropy": 5.454889869689941,
      "epoch": 1.437580726680519,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20139019638299943,
      "num_tokens": 28548099.0,
      "step": 12800
    },
    {
      "entropy": 5.516498470306397,
      "epoch": 1.4381423035884764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.157,
      "mean_token_accuracy": 0.20848567932844161,
      "num_tokens": 28559372.0,
      "step": 12805
    },
    {
      "entropy": 5.48066463470459,
      "epoch": 1.438703880496434,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20579466819763184,
      "num_tokens": 28570106.0,
      "step": 12810
    },
    {
      "entropy": 5.477960729598999,
      "epoch": 1.4392654574043915,
      "grad_norm": 0.890625,
      "learning_rate": 0.000480296566267982,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20573777407407762,
      "num_tokens": 28582417.0,
      "step": 12815
    },
    {
      "entropy": 5.555762767791748,
      "epoch": 1.4398270343123492,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.3051,
      "mean_token_accuracy": 0.19870916604995728,
      "num_tokens": 28594341.0,
      "step": 12820
    },
    {
      "entropy": 5.516356801986694,
      "epoch": 1.4403886112203066,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.1746,
      "mean_token_accuracy": 0.2082059159874916,
      "num_tokens": 28605580.0,
      "step": 12825
    },
    {
      "entropy": 5.574602890014648,
      "epoch": 1.440950188128264,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.2115,
      "mean_token_accuracy": 0.1996773436665535,
      "num_tokens": 28617779.0,
      "step": 12830
    },
    {
      "entropy": 5.556379652023315,
      "epoch": 1.4415117650362217,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.2142,
      "mean_token_accuracy": 0.1919758215546608,
      "num_tokens": 28629714.0,
      "step": 12835
    },
    {
      "entropy": 5.553221940994263,
      "epoch": 1.4420733419441794,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20141707956790925,
      "num_tokens": 28641381.0,
      "step": 12840
    },
    {
      "entropy": 5.518089437484742,
      "epoch": 1.4426349188521368,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.1968,
      "mean_token_accuracy": 0.20452302992343901,
      "num_tokens": 28653650.0,
      "step": 12845
    },
    {
      "entropy": 5.514556741714477,
      "epoch": 1.4431964957600942,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.1261,
      "mean_token_accuracy": 0.20813043862581254,
      "num_tokens": 28665252.0,
      "step": 12850
    },
    {
      "entropy": 5.48739333152771,
      "epoch": 1.443758072668052,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.1719,
      "mean_token_accuracy": 0.20587808042764663,
      "num_tokens": 28677072.0,
      "step": 12855
    },
    {
      "entropy": 5.427947568893432,
      "epoch": 1.4443196495760096,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2044305309653282,
      "num_tokens": 28688767.0,
      "step": 12860
    },
    {
      "entropy": 5.551112413406372,
      "epoch": 1.444881226483967,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.3443,
      "mean_token_accuracy": 0.20104330778121948,
      "num_tokens": 28699773.0,
      "step": 12865
    },
    {
      "entropy": 5.593872022628784,
      "epoch": 1.4454428033919244,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.1857,
      "mean_token_accuracy": 0.20449884235858917,
      "num_tokens": 28710971.0,
      "step": 12870
    },
    {
      "entropy": 5.480263185501099,
      "epoch": 1.446004380299882,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.0994,
      "mean_token_accuracy": 0.21018214672803878,
      "num_tokens": 28721544.0,
      "step": 12875
    },
    {
      "entropy": 5.553387689590454,
      "epoch": 1.4465659572078395,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.3051,
      "mean_token_accuracy": 0.20095628648996353,
      "num_tokens": 28733790.0,
      "step": 12880
    },
    {
      "entropy": 5.615977334976196,
      "epoch": 1.4471275341157972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.2211,
      "mean_token_accuracy": 0.20746534019708635,
      "num_tokens": 28745218.0,
      "step": 12885
    },
    {
      "entropy": 5.578668642044067,
      "epoch": 1.4476891110237546,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.255,
      "mean_token_accuracy": 0.1986464098095894,
      "num_tokens": 28758061.0,
      "step": 12890
    },
    {
      "entropy": 5.443735504150391,
      "epoch": 1.4482506879317123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.087,
      "mean_token_accuracy": 0.21252167224884033,
      "num_tokens": 28770037.0,
      "step": 12895
    },
    {
      "entropy": 5.564225578308106,
      "epoch": 1.4488122648396697,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19951967895030975,
      "num_tokens": 28781052.0,
      "step": 12900
    },
    {
      "entropy": 5.403512001037598,
      "epoch": 1.4493738417476274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.122,
      "mean_token_accuracy": 0.2020877033472061,
      "num_tokens": 28791225.0,
      "step": 12905
    },
    {
      "entropy": 5.509875297546387,
      "epoch": 1.4499354186555848,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.1414,
      "mean_token_accuracy": 0.2056181997060776,
      "num_tokens": 28803022.0,
      "step": 12910
    },
    {
      "entropy": 5.5057634830474855,
      "epoch": 1.4504969955635425,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20846739113330842,
      "num_tokens": 28813789.0,
      "step": 12915
    },
    {
      "entropy": 5.560517597198486,
      "epoch": 1.4510585724715,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20756234377622604,
      "num_tokens": 28825015.0,
      "step": 12920
    },
    {
      "entropy": 5.6226592540740965,
      "epoch": 1.4516201493794574,
      "grad_norm": 1.0,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.2822,
      "mean_token_accuracy": 0.20180982649326323,
      "num_tokens": 28835852.0,
      "step": 12925
    },
    {
      "entropy": 5.483248138427735,
      "epoch": 1.452181726287415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20172621458768844,
      "num_tokens": 28846684.0,
      "step": 12930
    },
    {
      "entropy": 5.552487468719482,
      "epoch": 1.4527433031953727,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.2004,
      "mean_token_accuracy": 0.19857454746961595,
      "num_tokens": 28857402.0,
      "step": 12935
    },
    {
      "entropy": 5.4710735321044925,
      "epoch": 1.4533048801033301,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21428440809249877,
      "num_tokens": 28867866.0,
      "step": 12940
    },
    {
      "entropy": 5.490120887756348,
      "epoch": 1.4538664570112876,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.1538,
      "mean_token_accuracy": 0.20599478483200073,
      "num_tokens": 28878102.0,
      "step": 12945
    },
    {
      "entropy": 5.591460180282593,
      "epoch": 1.4544280339192452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.2397,
      "mean_token_accuracy": 0.2003340244293213,
      "num_tokens": 28888981.0,
      "step": 12950
    },
    {
      "entropy": 5.466292238235473,
      "epoch": 1.454989610827203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.1155,
      "mean_token_accuracy": 0.2045106142759323,
      "num_tokens": 28899733.0,
      "step": 12955
    },
    {
      "entropy": 5.438828659057617,
      "epoch": 1.4555511877351603,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2016093075275421,
      "num_tokens": 28910701.0,
      "step": 12960
    },
    {
      "entropy": 5.553754615783691,
      "epoch": 1.4561127646431178,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.1885,
      "mean_token_accuracy": 0.2019767180085182,
      "num_tokens": 28922627.0,
      "step": 12965
    },
    {
      "entropy": 5.540034532546997,
      "epoch": 1.4566743415510754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.2265,
      "mean_token_accuracy": 0.2018926113843918,
      "num_tokens": 28933292.0,
      "step": 12970
    },
    {
      "entropy": 5.425068759918213,
      "epoch": 1.4572359184590329,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20696118772029876,
      "num_tokens": 28943898.0,
      "step": 12975
    },
    {
      "entropy": 5.4820404052734375,
      "epoch": 1.4577974953669905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20159929990768433,
      "num_tokens": 28955248.0,
      "step": 12980
    },
    {
      "entropy": 5.497602033615112,
      "epoch": 1.458359072274948,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21162678003311158,
      "num_tokens": 28966905.0,
      "step": 12985
    },
    {
      "entropy": 5.524466943740845,
      "epoch": 1.4589206491829056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.1997,
      "mean_token_accuracy": 0.20563169866800307,
      "num_tokens": 28978515.0,
      "step": 12990
    },
    {
      "entropy": 5.458061838150025,
      "epoch": 1.459482226090863,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.1712,
      "mean_token_accuracy": 0.196217380464077,
      "num_tokens": 28990185.0,
      "step": 12995
    },
    {
      "entropy": 5.487705516815185,
      "epoch": 1.4600438029988208,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.192,
      "mean_token_accuracy": 0.20145193934440614,
      "num_tokens": 29000779.0,
      "step": 13000
    },
    {
      "entropy": 5.5696252346038815,
      "epoch": 1.4606053799067782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.2612,
      "mean_token_accuracy": 0.2000700354576111,
      "num_tokens": 29012066.0,
      "step": 13005
    },
    {
      "entropy": 5.639033508300781,
      "epoch": 1.4611669568147359,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.308,
      "mean_token_accuracy": 0.19485906064510344,
      "num_tokens": 29023416.0,
      "step": 13010
    },
    {
      "entropy": 5.526591968536377,
      "epoch": 1.4617285337226933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.1387,
      "mean_token_accuracy": 0.2032766282558441,
      "num_tokens": 29034469.0,
      "step": 13015
    },
    {
      "entropy": 5.4981224060058596,
      "epoch": 1.4622901106306507,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19745652079582215,
      "num_tokens": 29045306.0,
      "step": 13020
    },
    {
      "entropy": 5.478452444076538,
      "epoch": 1.4628516875386084,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000479600737020506,
      "loss": 5.1053,
      "mean_token_accuracy": 0.2039801299571991,
      "num_tokens": 29057114.0,
      "step": 13025
    },
    {
      "entropy": 5.462406063079834,
      "epoch": 1.463413264446566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.0905,
      "mean_token_accuracy": 0.19879208356142045,
      "num_tokens": 29067250.0,
      "step": 13030
    },
    {
      "entropy": 5.471080493927002,
      "epoch": 1.4639748413545235,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.1234,
      "mean_token_accuracy": 0.21133094877004624,
      "num_tokens": 29078582.0,
      "step": 13035
    },
    {
      "entropy": 5.51628270149231,
      "epoch": 1.464536418262481,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.1597,
      "mean_token_accuracy": 0.2063067987561226,
      "num_tokens": 29090113.0,
      "step": 13040
    },
    {
      "entropy": 5.47154746055603,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20405905246734618,
      "num_tokens": 29101117.0,
      "step": 13045
    },
    {
      "entropy": 5.552691507339477,
      "epoch": 1.4656595720783963,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.1584,
      "mean_token_accuracy": 0.20199237614870072,
      "num_tokens": 29112569.0,
      "step": 13050
    },
    {
      "entropy": 5.545939111709595,
      "epoch": 1.4662211489863537,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.257,
      "mean_token_accuracy": 0.20114679187536239,
      "num_tokens": 29124462.0,
      "step": 13055
    },
    {
      "entropy": 5.495091390609741,
      "epoch": 1.4667827258943111,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.1294,
      "mean_token_accuracy": 0.2060330927371979,
      "num_tokens": 29135401.0,
      "step": 13060
    },
    {
      "entropy": 5.5179389953613285,
      "epoch": 1.4673443028022688,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.143,
      "mean_token_accuracy": 0.2070012718439102,
      "num_tokens": 29146601.0,
      "step": 13065
    },
    {
      "entropy": 5.512892961502075,
      "epoch": 1.4679058797102262,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.1311,
      "mean_token_accuracy": 0.20361219942569733,
      "num_tokens": 29157299.0,
      "step": 13070
    },
    {
      "entropy": 5.56696515083313,
      "epoch": 1.468467456618184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.2227,
      "mean_token_accuracy": 0.20172676891088487,
      "num_tokens": 29169101.0,
      "step": 13075
    },
    {
      "entropy": 5.407410907745361,
      "epoch": 1.4690290335261413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20720174610614778,
      "num_tokens": 29181286.0,
      "step": 13080
    },
    {
      "entropy": 5.510353231430054,
      "epoch": 1.469590610434099,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.2289,
      "mean_token_accuracy": 0.20066646486520767,
      "num_tokens": 29191181.0,
      "step": 13085
    },
    {
      "entropy": 5.448162031173706,
      "epoch": 1.4701521873420564,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.0947,
      "mean_token_accuracy": 0.2027595654129982,
      "num_tokens": 29202426.0,
      "step": 13090
    },
    {
      "entropy": 5.429555034637451,
      "epoch": 1.470713764250014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.0928,
      "mean_token_accuracy": 0.2053892582654953,
      "num_tokens": 29212817.0,
      "step": 13095
    },
    {
      "entropy": 5.48428955078125,
      "epoch": 1.4712753411579715,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.117,
      "mean_token_accuracy": 0.1993074521422386,
      "num_tokens": 29224644.0,
      "step": 13100
    },
    {
      "entropy": 5.511887836456299,
      "epoch": 1.4718369180659292,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.119,
      "mean_token_accuracy": 0.20782431811094285,
      "num_tokens": 29236443.0,
      "step": 13105
    },
    {
      "entropy": 5.550143194198609,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1964699849486351,
      "num_tokens": 29247687.0,
      "step": 13110
    },
    {
      "entropy": 5.549376201629639,
      "epoch": 1.4729600718818443,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.1742,
      "mean_token_accuracy": 0.2033306285738945,
      "num_tokens": 29259739.0,
      "step": 13115
    },
    {
      "entropy": 5.5659740447998045,
      "epoch": 1.4735216487898017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.2044,
      "mean_token_accuracy": 0.2034813418984413,
      "num_tokens": 29270834.0,
      "step": 13120
    },
    {
      "entropy": 5.523956394195556,
      "epoch": 1.4740832256977594,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.17,
      "mean_token_accuracy": 0.20406033843755722,
      "num_tokens": 29282291.0,
      "step": 13125
    },
    {
      "entropy": 5.500520420074463,
      "epoch": 1.4746448026057168,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.2037,
      "mean_token_accuracy": 0.20116291493177413,
      "num_tokens": 29293836.0,
      "step": 13130
    },
    {
      "entropy": 5.539033317565918,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.2091,
      "mean_token_accuracy": 0.1981520876288414,
      "num_tokens": 29305550.0,
      "step": 13135
    },
    {
      "entropy": 5.582889175415039,
      "epoch": 1.475767956421632,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.1902,
      "mean_token_accuracy": 0.20414477437734604,
      "num_tokens": 29316973.0,
      "step": 13140
    },
    {
      "entropy": 5.492055892944336,
      "epoch": 1.4763295333295896,
      "grad_norm": 0.875,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.064,
      "mean_token_accuracy": 0.21038700491189957,
      "num_tokens": 29327743.0,
      "step": 13145
    },
    {
      "entropy": 5.435461235046387,
      "epoch": 1.476891110237547,
      "grad_norm": 0.875,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20313042551279067,
      "num_tokens": 29338286.0,
      "step": 13150
    },
    {
      "entropy": 5.465268659591675,
      "epoch": 1.4774526871455045,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.2174,
      "mean_token_accuracy": 0.20017955601215362,
      "num_tokens": 29348877.0,
      "step": 13155
    },
    {
      "entropy": 5.443563652038574,
      "epoch": 1.4780142640534621,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047914734744424517,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21573174446821214,
      "num_tokens": 29360072.0,
      "step": 13160
    },
    {
      "entropy": 5.538565635681152,
      "epoch": 1.4785758409614196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.2195,
      "mean_token_accuracy": 0.20482390820980073,
      "num_tokens": 29370744.0,
      "step": 13165
    },
    {
      "entropy": 5.53998875617981,
      "epoch": 1.4791374178693772,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.1481,
      "mean_token_accuracy": 0.20351329147815705,
      "num_tokens": 29381463.0,
      "step": 13170
    },
    {
      "entropy": 5.525278377532959,
      "epoch": 1.4796989947773347,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.1938,
      "mean_token_accuracy": 0.20462251007556914,
      "num_tokens": 29394227.0,
      "step": 13175
    },
    {
      "entropy": 5.446918869018555,
      "epoch": 1.4802605716852923,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20696335881948472,
      "num_tokens": 29404754.0,
      "step": 13180
    },
    {
      "entropy": 5.452279043197632,
      "epoch": 1.4808221485932498,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.1528,
      "mean_token_accuracy": 0.21082573384046555,
      "num_tokens": 29416161.0,
      "step": 13185
    },
    {
      "entropy": 5.458856105804443,
      "epoch": 1.4813837255012074,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21318838447332383,
      "num_tokens": 29426587.0,
      "step": 13190
    },
    {
      "entropy": 5.501920986175537,
      "epoch": 1.4819453024091649,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1949346736073494,
      "num_tokens": 29437488.0,
      "step": 13195
    },
    {
      "entropy": 5.519732046127319,
      "epoch": 1.4825068793171226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20581742525100707,
      "num_tokens": 29448244.0,
      "step": 13200
    },
    {
      "entropy": 5.586218976974488,
      "epoch": 1.48306845622508,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.2538,
      "mean_token_accuracy": 0.19651770293712617,
      "num_tokens": 29458260.0,
      "step": 13205
    },
    {
      "entropy": 5.6076802730560305,
      "epoch": 1.4836300331330377,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.2066,
      "mean_token_accuracy": 0.2015571638941765,
      "num_tokens": 29470874.0,
      "step": 13210
    },
    {
      "entropy": 5.432733488082886,
      "epoch": 1.484191610040995,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20470332354307175,
      "num_tokens": 29481442.0,
      "step": 13215
    },
    {
      "entropy": 5.5742353916168215,
      "epoch": 1.4847531869489528,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.2683,
      "mean_token_accuracy": 0.20413292944431305,
      "num_tokens": 29492727.0,
      "step": 13220
    },
    {
      "entropy": 5.585005044937134,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.1918,
      "mean_token_accuracy": 0.19427585750818252,
      "num_tokens": 29502729.0,
      "step": 13225
    },
    {
      "entropy": 5.596617889404297,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20591949820518493,
      "num_tokens": 29514315.0,
      "step": 13230
    },
    {
      "entropy": 5.504071855545044,
      "epoch": 1.4864379176728253,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20799697935581207,
      "num_tokens": 29526030.0,
      "step": 13235
    },
    {
      "entropy": 5.460632801055908,
      "epoch": 1.486999494580783,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20458327680826188,
      "num_tokens": 29536972.0,
      "step": 13240
    },
    {
      "entropy": 5.434044742584229,
      "epoch": 1.4875610714887404,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20704464018344879,
      "num_tokens": 29547325.0,
      "step": 13245
    },
    {
      "entropy": 5.55746865272522,
      "epoch": 1.4881226483966978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.1802,
      "mean_token_accuracy": 0.2060597613453865,
      "num_tokens": 29557944.0,
      "step": 13250
    },
    {
      "entropy": 5.449753475189209,
      "epoch": 1.4886842253046555,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.0978,
      "mean_token_accuracy": 0.21049654483795166,
      "num_tokens": 29568618.0,
      "step": 13255
    },
    {
      "entropy": 5.544513511657715,
      "epoch": 1.489245802212613,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.1711,
      "mean_token_accuracy": 0.2013496607542038,
      "num_tokens": 29580682.0,
      "step": 13260
    },
    {
      "entropy": 5.55446228981018,
      "epoch": 1.4898073791205706,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.1973,
      "mean_token_accuracy": 0.20249828547239304,
      "num_tokens": 29592302.0,
      "step": 13265
    },
    {
      "entropy": 5.541396951675415,
      "epoch": 1.490368956028528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20154530853033065,
      "num_tokens": 29603591.0,
      "step": 13270
    },
    {
      "entropy": 5.479934167861939,
      "epoch": 1.4909305329364857,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.1764,
      "mean_token_accuracy": 0.2013258844614029,
      "num_tokens": 29615542.0,
      "step": 13275
    },
    {
      "entropy": 5.567523002624512,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.2116,
      "mean_token_accuracy": 0.205784410238266,
      "num_tokens": 29626650.0,
      "step": 13280
    },
    {
      "entropy": 5.500628280639648,
      "epoch": 1.4920536867524008,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.1506,
      "mean_token_accuracy": 0.2019352912902832,
      "num_tokens": 29637135.0,
      "step": 13285
    },
    {
      "entropy": 5.534039497375488,
      "epoch": 1.4926152636603582,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.1882,
      "mean_token_accuracy": 0.1967959702014923,
      "num_tokens": 29648862.0,
      "step": 13290
    },
    {
      "entropy": 5.512229871749878,
      "epoch": 1.493176840568316,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20311812460422515,
      "num_tokens": 29661015.0,
      "step": 13295
    },
    {
      "entropy": 5.5943684577941895,
      "epoch": 1.4937384174762733,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.1515,
      "mean_token_accuracy": 0.2002973660826683,
      "num_tokens": 29673521.0,
      "step": 13300
    },
    {
      "entropy": 5.546845197677612,
      "epoch": 1.494299994384231,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.2228,
      "mean_token_accuracy": 0.2089742749929428,
      "num_tokens": 29685379.0,
      "step": 13305
    },
    {
      "entropy": 5.472210216522217,
      "epoch": 1.4948615712921884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.1448,
      "mean_token_accuracy": 0.2004483848810196,
      "num_tokens": 29696372.0,
      "step": 13310
    },
    {
      "entropy": 5.540146112442017,
      "epoch": 1.495423148200146,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20663609802722932,
      "num_tokens": 29707401.0,
      "step": 13315
    },
    {
      "entropy": 5.500766849517822,
      "epoch": 1.4959847251081035,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20737934112548828,
      "num_tokens": 29717524.0,
      "step": 13320
    },
    {
      "entropy": 5.522632217407226,
      "epoch": 1.496546302016061,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.2516,
      "mean_token_accuracy": 0.199118909239769,
      "num_tokens": 29728235.0,
      "step": 13325
    },
    {
      "entropy": 5.572938251495361,
      "epoch": 1.4971078789240186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.2474,
      "mean_token_accuracy": 0.19530449360609053,
      "num_tokens": 29738800.0,
      "step": 13330
    },
    {
      "entropy": 5.55489673614502,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.0,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.1757,
      "mean_token_accuracy": 0.2024305447936058,
      "num_tokens": 29750346.0,
      "step": 13335
    },
    {
      "entropy": 5.485504293441773,
      "epoch": 1.4982310327399337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20853393375873566,
      "num_tokens": 29761931.0,
      "step": 13340
    },
    {
      "entropy": 5.438781595230102,
      "epoch": 1.4987926096478912,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.0923,
      "mean_token_accuracy": 0.20915288031101226,
      "num_tokens": 29773333.0,
      "step": 13345
    },
    {
      "entropy": 5.443496751785278,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20898902267217637,
      "num_tokens": 29785125.0,
      "step": 13350
    },
    {
      "entropy": 5.455953598022461,
      "epoch": 1.4999157634638065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.1878,
      "mean_token_accuracy": 0.21046700924634934,
      "num_tokens": 29796288.0,
      "step": 13355
    },
    {
      "entropy": 5.570097732543945,
      "epoch": 1.500477340371764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20278282314538956,
      "num_tokens": 29806348.0,
      "step": 13360
    },
    {
      "entropy": 5.536681413650513,
      "epoch": 1.5010389172797214,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.1107,
      "mean_token_accuracy": 0.2099132388830185,
      "num_tokens": 29816342.0,
      "step": 13365
    },
    {
      "entropy": 5.449040699005127,
      "epoch": 1.501600494187679,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.1053,
      "mean_token_accuracy": 0.2023299068212509,
      "num_tokens": 29826789.0,
      "step": 13370
    },
    {
      "entropy": 5.506910753250122,
      "epoch": 1.5021620710956367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.1898,
      "mean_token_accuracy": 0.20509798675775529,
      "num_tokens": 29837633.0,
      "step": 13375
    },
    {
      "entropy": 5.579867744445801,
      "epoch": 1.502723648003594,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.2509,
      "mean_token_accuracy": 0.20346703827381135,
      "num_tokens": 29848498.0,
      "step": 13380
    },
    {
      "entropy": 5.577423143386841,
      "epoch": 1.5032852249115516,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.2025,
      "mean_token_accuracy": 0.2009313389658928,
      "num_tokens": 29859777.0,
      "step": 13385
    },
    {
      "entropy": 5.734101247787476,
      "epoch": 1.5038468018195092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.4054,
      "mean_token_accuracy": 0.19479023069143295,
      "num_tokens": 29873373.0,
      "step": 13390
    },
    {
      "entropy": 5.553149223327637,
      "epoch": 1.5044083787274667,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20812069922685622,
      "num_tokens": 29884749.0,
      "step": 13395
    },
    {
      "entropy": 5.441007518768311,
      "epoch": 1.5049699556354241,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.085,
      "mean_token_accuracy": 0.2054499566555023,
      "num_tokens": 29895059.0,
      "step": 13400
    },
    {
      "entropy": 5.4207761764526365,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.1506,
      "mean_token_accuracy": 0.2003682255744934,
      "num_tokens": 29906611.0,
      "step": 13405
    },
    {
      "entropy": 5.406160688400268,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.0,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.022,
      "mean_token_accuracy": 0.21143694669008256,
      "num_tokens": 29918179.0,
      "step": 13410
    },
    {
      "entropy": 5.568183517456054,
      "epoch": 1.506654686359297,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.1888,
      "mean_token_accuracy": 0.203368616104126,
      "num_tokens": 29929612.0,
      "step": 13415
    },
    {
      "entropy": 5.531915664672852,
      "epoch": 1.5072162632672543,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.064,
      "mean_token_accuracy": 0.21058074831962587,
      "num_tokens": 29939569.0,
      "step": 13420
    },
    {
      "entropy": 5.4811841487884525,
      "epoch": 1.507777840175212,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.1051,
      "mean_token_accuracy": 0.20722654610872268,
      "num_tokens": 29950344.0,
      "step": 13425
    },
    {
      "entropy": 5.484924936294556,
      "epoch": 1.5083394170831697,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.1189,
      "mean_token_accuracy": 0.2118167147040367,
      "num_tokens": 29960748.0,
      "step": 13430
    },
    {
      "entropy": 5.620925951004028,
      "epoch": 1.508900993991127,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19196939766407012,
      "num_tokens": 29971678.0,
      "step": 13435
    },
    {
      "entropy": 5.460380125045776,
      "epoch": 1.5094625708990845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.129,
      "mean_token_accuracy": 0.2051836922764778,
      "num_tokens": 29982477.0,
      "step": 13440
    },
    {
      "entropy": 5.473675060272217,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.1762,
      "mean_token_accuracy": 0.2065405234694481,
      "num_tokens": 29994539.0,
      "step": 13445
    },
    {
      "entropy": 5.555587482452393,
      "epoch": 1.5105857247149999,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20993212461471558,
      "num_tokens": 30004821.0,
      "step": 13450
    },
    {
      "entropy": 5.613119649887085,
      "epoch": 1.5111473016229573,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.1836,
      "mean_token_accuracy": 0.1999419555068016,
      "num_tokens": 30016343.0,
      "step": 13455
    },
    {
      "entropy": 5.471673154830933,
      "epoch": 1.5117088785309147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.0946,
      "mean_token_accuracy": 0.2116156056523323,
      "num_tokens": 30027719.0,
      "step": 13460
    },
    {
      "entropy": 5.452823257446289,
      "epoch": 1.5122704554388724,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20878891944885253,
      "num_tokens": 30037988.0,
      "step": 13465
    },
    {
      "entropy": 5.4986106872558596,
      "epoch": 1.51283203234683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21202182322740554,
      "num_tokens": 30049736.0,
      "step": 13470
    },
    {
      "entropy": 5.562261247634888,
      "epoch": 1.5133936092547875,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.2157,
      "mean_token_accuracy": 0.20256965458393097,
      "num_tokens": 30061963.0,
      "step": 13475
    },
    {
      "entropy": 5.586698246002197,
      "epoch": 1.513955186162745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.1151,
      "mean_token_accuracy": 0.21958382278680802,
      "num_tokens": 30072256.0,
      "step": 13480
    },
    {
      "entropy": 5.466584205627441,
      "epoch": 1.5145167630707026,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.1793,
      "mean_token_accuracy": 0.20731011927127838,
      "num_tokens": 30084804.0,
      "step": 13485
    },
    {
      "entropy": 5.5143858909606935,
      "epoch": 1.51507833997866,
      "grad_norm": 1.0,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.1561,
      "mean_token_accuracy": 0.206527441740036,
      "num_tokens": 30096938.0,
      "step": 13490
    },
    {
      "entropy": 5.566529178619385,
      "epoch": 1.5156399168866175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.1337,
      "mean_token_accuracy": 0.1997727006673813,
      "num_tokens": 30108233.0,
      "step": 13495
    },
    {
      "entropy": 5.502043533325195,
      "epoch": 1.5162014937945751,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000477984505899334,
      "loss": 5.1547,
      "mean_token_accuracy": 0.20644167363643645,
      "num_tokens": 30120246.0,
      "step": 13500
    },
    {
      "entropy": 5.4806907176971436,
      "epoch": 1.5167630707025328,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047796718188901095,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21251375377178192,
      "num_tokens": 30131213.0,
      "step": 13505
    },
    {
      "entropy": 5.450793933868408,
      "epoch": 1.5173246476104902,
      "grad_norm": 0.875,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.0592,
      "mean_token_accuracy": 0.21112452894449235,
      "num_tokens": 30143300.0,
      "step": 13510
    },
    {
      "entropy": 5.477409887313843,
      "epoch": 1.5178862245184477,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.0963,
      "mean_token_accuracy": 0.20516830831766128,
      "num_tokens": 30153794.0,
      "step": 13515
    },
    {
      "entropy": 5.494986915588379,
      "epoch": 1.5184478014264053,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19827048778533934,
      "num_tokens": 30165152.0,
      "step": 13520
    },
    {
      "entropy": 5.500593042373657,
      "epoch": 1.519009378334363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20601859390735627,
      "num_tokens": 30176338.0,
      "step": 13525
    },
    {
      "entropy": 5.50661883354187,
      "epoch": 1.5195709552423204,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.188,
      "mean_token_accuracy": 0.19990554302930832,
      "num_tokens": 30187666.0,
      "step": 13530
    },
    {
      "entropy": 5.519746446609497,
      "epoch": 1.5201325321502779,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19866112917661666,
      "num_tokens": 30198892.0,
      "step": 13535
    },
    {
      "entropy": 5.486255168914795,
      "epoch": 1.5206941090582355,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.172,
      "mean_token_accuracy": 0.20193684846162796,
      "num_tokens": 30210677.0,
      "step": 13540
    },
    {
      "entropy": 5.533208417892456,
      "epoch": 1.5212556859661932,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2090500995516777,
      "num_tokens": 30221217.0,
      "step": 13545
    },
    {
      "entropy": 5.562088060379028,
      "epoch": 1.5218172628741506,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19776107370853424,
      "num_tokens": 30233576.0,
      "step": 13550
    },
    {
      "entropy": 5.47334623336792,
      "epoch": 1.522378839782108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.0902,
      "mean_token_accuracy": 0.20376465022563933,
      "num_tokens": 30244787.0,
      "step": 13555
    },
    {
      "entropy": 5.527383804321289,
      "epoch": 1.5229404166900657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000477776191368407,
      "loss": 5.1443,
      "mean_token_accuracy": 0.2076149582862854,
      "num_tokens": 30256656.0,
      "step": 13560
    },
    {
      "entropy": 5.554253816604614,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.2147,
      "mean_token_accuracy": 0.20448839664459229,
      "num_tokens": 30267715.0,
      "step": 13565
    },
    {
      "entropy": 5.563332176208496,
      "epoch": 1.5240635705059808,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.2062,
      "mean_token_accuracy": 0.2025810956954956,
      "num_tokens": 30278119.0,
      "step": 13570
    },
    {
      "entropy": 5.570546197891235,
      "epoch": 1.5246251474139383,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.1285,
      "mean_token_accuracy": 0.19979864805936814,
      "num_tokens": 30289899.0,
      "step": 13575
    },
    {
      "entropy": 5.444940137863159,
      "epoch": 1.525186724321896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20863568484783174,
      "num_tokens": 30300799.0,
      "step": 13580
    },
    {
      "entropy": 5.581645965576172,
      "epoch": 1.5257483012298534,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19904510527849198,
      "num_tokens": 30312955.0,
      "step": 13585
    },
    {
      "entropy": 5.577303743362426,
      "epoch": 1.5263098781378108,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20475982278585433,
      "num_tokens": 30323089.0,
      "step": 13590
    },
    {
      "entropy": 5.492038631439209,
      "epoch": 1.5268714550457685,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000477654245141998,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21648856699466706,
      "num_tokens": 30333926.0,
      "step": 13595
    },
    {
      "entropy": 5.426802778244019,
      "epoch": 1.5274330319537262,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20605419874191283,
      "num_tokens": 30346347.0,
      "step": 13600
    },
    {
      "entropy": 5.488521575927734,
      "epoch": 1.5279946088616836,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.1525,
      "mean_token_accuracy": 0.20099833458662034,
      "num_tokens": 30357757.0,
      "step": 13605
    },
    {
      "entropy": 5.614195871353149,
      "epoch": 1.528556185769641,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.1946,
      "mean_token_accuracy": 0.21140614598989488,
      "num_tokens": 30369323.0,
      "step": 13610
    },
    {
      "entropy": 5.482068252563477,
      "epoch": 1.5291177626775987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20954455584287643,
      "num_tokens": 30380335.0,
      "step": 13615
    },
    {
      "entropy": 5.419958877563476,
      "epoch": 1.5296793395855564,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.0875,
      "mean_token_accuracy": 0.20452880263328552,
      "num_tokens": 30391338.0,
      "step": 13620
    },
    {
      "entropy": 5.509082078933716,
      "epoch": 1.5302409164935138,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000477549468146883,
      "loss": 5.205,
      "mean_token_accuracy": 0.19613491743803024,
      "num_tokens": 30402917.0,
      "step": 13625
    },
    {
      "entropy": 5.44488492012024,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004775319827387309,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21413268148899078,
      "num_tokens": 30412690.0,
      "step": 13630
    },
    {
      "entropy": 5.476470041275024,
      "epoch": 1.531364070309429,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.1895,
      "mean_token_accuracy": 0.20066993683576584,
      "num_tokens": 30423020.0,
      "step": 13635
    },
    {
      "entropy": 5.62052206993103,
      "epoch": 1.5319256472173866,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.3926,
      "mean_token_accuracy": 0.18681813776493073,
      "num_tokens": 30434039.0,
      "step": 13640
    },
    {
      "entropy": 5.588673782348633,
      "epoch": 1.532487224125344,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20178985893726348,
      "num_tokens": 30446743.0,
      "step": 13645
    },
    {
      "entropy": 5.5280348777771,
      "epoch": 1.5330488010333014,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000477461976624724,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20643433183431625,
      "num_tokens": 30456969.0,
      "step": 13650
    },
    {
      "entropy": 5.431046342849731,
      "epoch": 1.533610377941259,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.2048,
      "mean_token_accuracy": 0.19871219247579575,
      "num_tokens": 30467497.0,
      "step": 13655
    },
    {
      "entropy": 5.490679454803467,
      "epoch": 1.5341719548492168,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000477426934886684,
      "loss": 5.1465,
      "mean_token_accuracy": 0.20694950073957444,
      "num_tokens": 30479428.0,
      "step": 13660
    },
    {
      "entropy": 5.544655513763428,
      "epoch": 1.5347335317571742,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.1174,
      "mean_token_accuracy": 0.2107458770275116,
      "num_tokens": 30490193.0,
      "step": 13665
    },
    {
      "entropy": 5.586663436889649,
      "epoch": 1.5352951086651316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.1888,
      "mean_token_accuracy": 0.20256867110729218,
      "num_tokens": 30500311.0,
      "step": 13670
    },
    {
      "entropy": 5.394921827316284,
      "epoch": 1.5358566855730893,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.0912,
      "mean_token_accuracy": 0.20886549949645997,
      "num_tokens": 30512485.0,
      "step": 13675
    },
    {
      "entropy": 5.50133056640625,
      "epoch": 1.5364182624810467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.0841,
      "mean_token_accuracy": 0.2084451988339424,
      "num_tokens": 30523495.0,
      "step": 13680
    },
    {
      "entropy": 5.49138503074646,
      "epoch": 1.5369798393890042,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.1443,
      "mean_token_accuracy": 0.21129620373249053,
      "num_tokens": 30533964.0,
      "step": 13685
    },
    {
      "entropy": 5.4022057056427,
      "epoch": 1.5375414162969618,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21302701234817506,
      "num_tokens": 30544798.0,
      "step": 13690
    },
    {
      "entropy": 5.49497127532959,
      "epoch": 1.5381029932049195,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.2082,
      "mean_token_accuracy": 0.20119480341672896,
      "num_tokens": 30555991.0,
      "step": 13695
    },
    {
      "entropy": 5.542425918579101,
      "epoch": 1.538664570112877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.0749,
      "mean_token_accuracy": 0.2084500476717949,
      "num_tokens": 30566848.0,
      "step": 13700
    },
    {
      "entropy": 5.544796371459961,
      "epoch": 1.5392261470208344,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.0842,
      "mean_token_accuracy": 0.21806326359510422,
      "num_tokens": 30577361.0,
      "step": 13705
    },
    {
      "entropy": 5.484889030456543,
      "epoch": 1.539787723928792,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.1355,
      "mean_token_accuracy": 0.2069666028022766,
      "num_tokens": 30587685.0,
      "step": 13710
    },
    {
      "entropy": 5.464565372467041,
      "epoch": 1.5403493008367497,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.1784,
      "mean_token_accuracy": 0.20487663745880128,
      "num_tokens": 30599864.0,
      "step": 13715
    },
    {
      "entropy": 5.511539316177368,
      "epoch": 1.5409108777447071,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20523305237293243,
      "num_tokens": 30610885.0,
      "step": 13720
    },
    {
      "entropy": 5.518194150924683,
      "epoch": 1.5414724546526646,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.1265,
      "mean_token_accuracy": 0.20369516164064408,
      "num_tokens": 30622537.0,
      "step": 13725
    },
    {
      "entropy": 5.574005079269409,
      "epoch": 1.5420340315606222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1963867038488388,
      "num_tokens": 30635045.0,
      "step": 13730
    },
    {
      "entropy": 5.471251058578491,
      "epoch": 1.54259560846858,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.0916,
      "mean_token_accuracy": 0.2010647863149643,
      "num_tokens": 30645998.0,
      "step": 13735
    },
    {
      "entropy": 5.496492576599121,
      "epoch": 1.5431571853765373,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.1469,
      "mean_token_accuracy": 0.2026340737938881,
      "num_tokens": 30658744.0,
      "step": 13740
    },
    {
      "entropy": 5.446831798553466,
      "epoch": 1.5437187622844948,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21553710997104644,
      "num_tokens": 30669420.0,
      "step": 13745
    },
    {
      "entropy": 5.3946631908416744,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.1099,
      "mean_token_accuracy": 0.20690556168556212,
      "num_tokens": 30680037.0,
      "step": 13750
    },
    {
      "entropy": 5.575890636444091,
      "epoch": 1.54484191610041,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.1505,
      "mean_token_accuracy": 0.2114671066403389,
      "num_tokens": 30691650.0,
      "step": 13755
    },
    {
      "entropy": 5.507847929000855,
      "epoch": 1.5454034930083675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.1629,
      "mean_token_accuracy": 0.20654657632112502,
      "num_tokens": 30701876.0,
      "step": 13760
    },
    {
      "entropy": 5.63815541267395,
      "epoch": 1.545965069916325,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19816827178001403,
      "num_tokens": 30714484.0,
      "step": 13765
    },
    {
      "entropy": 5.530080080032349,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477039775179498,
      "loss": 5.1653,
      "mean_token_accuracy": 0.19893455654382705,
      "num_tokens": 30725185.0,
      "step": 13770
    },
    {
      "entropy": 5.52663950920105,
      "epoch": 1.54708822373224,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20600399374961853,
      "num_tokens": 30735522.0,
      "step": 13775
    },
    {
      "entropy": 5.473762083053589,
      "epoch": 1.5476498006401975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.1497,
      "mean_token_accuracy": 0.2072127178311348,
      "num_tokens": 30746353.0,
      "step": 13780
    },
    {
      "entropy": 5.4942466735839846,
      "epoch": 1.5482113775481552,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000476986739302201,
      "loss": 5.106,
      "mean_token_accuracy": 0.20564247518777848,
      "num_tokens": 30756245.0,
      "step": 13785
    },
    {
      "entropy": 5.536868000030518,
      "epoch": 1.5487729544561128,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.2289,
      "mean_token_accuracy": 0.20538665354251862,
      "num_tokens": 30767387.0,
      "step": 13790
    },
    {
      "entropy": 5.507962369918824,
      "epoch": 1.5493345313640703,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20495103895664216,
      "num_tokens": 30780140.0,
      "step": 13795
    },
    {
      "entropy": 5.500946712493897,
      "epoch": 1.5498961082720277,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047693364554288,
      "loss": 5.0946,
      "mean_token_accuracy": 0.2153229758143425,
      "num_tokens": 30791137.0,
      "step": 13800
    },
    {
      "entropy": 5.499880695343018,
      "epoch": 1.5504576851799854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.0968,
      "mean_token_accuracy": 0.21359084844589232,
      "num_tokens": 30801708.0,
      "step": 13805
    },
    {
      "entropy": 5.57369441986084,
      "epoch": 1.551019262087943,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.1763,
      "mean_token_accuracy": 0.20434479415416718,
      "num_tokens": 30813519.0,
      "step": 13810
    },
    {
      "entropy": 5.570215940475464,
      "epoch": 1.5515808389959005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.2431,
      "mean_token_accuracy": 0.20303847193717955,
      "num_tokens": 30823786.0,
      "step": 13815
    },
    {
      "entropy": 5.457481527328492,
      "epoch": 1.552142415903858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20730505585670472,
      "num_tokens": 30834381.0,
      "step": 13820
    },
    {
      "entropy": 5.502739572525025,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.1222,
      "mean_token_accuracy": 0.20743568986654282,
      "num_tokens": 30844865.0,
      "step": 13825
    },
    {
      "entropy": 5.577833461761474,
      "epoch": 1.5532655697197733,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.1972,
      "mean_token_accuracy": 0.20870999991893768,
      "num_tokens": 30857088.0,
      "step": 13830
    },
    {
      "entropy": 5.491534042358398,
      "epoch": 1.5538271466277307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.2404,
      "mean_token_accuracy": 0.1946061819791794,
      "num_tokens": 30868320.0,
      "step": 13835
    },
    {
      "entropy": 5.505542945861817,
      "epoch": 1.5543887235356881,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20536718517541885,
      "num_tokens": 30878598.0,
      "step": 13840
    },
    {
      "entropy": 5.5238720893859865,
      "epoch": 1.5549503004436458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20900287926197053,
      "num_tokens": 30889254.0,
      "step": 13845
    },
    {
      "entropy": 5.54768385887146,
      "epoch": 1.5555118773516035,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.2102,
      "mean_token_accuracy": 0.2057593822479248,
      "num_tokens": 30901164.0,
      "step": 13850
    },
    {
      "entropy": 5.4767991065979,
      "epoch": 1.556073454259561,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.0853,
      "mean_token_accuracy": 0.2070373237133026,
      "num_tokens": 30911828.0,
      "step": 13855
    },
    {
      "entropy": 5.479254245758057,
      "epoch": 1.5566350311675183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004767206919899014,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2164432406425476,
      "num_tokens": 30922050.0,
      "step": 13860
    },
    {
      "entropy": 5.417798089981079,
      "epoch": 1.557196608075476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000476702904098382,
      "loss": 5.0599,
      "mean_token_accuracy": 0.20986329913139343,
      "num_tokens": 30932723.0,
      "step": 13865
    },
    {
      "entropy": 5.459502792358398,
      "epoch": 1.5577581849834334,
      "grad_norm": 0.953125,
      "learning_rate": 0.000476685109784564,
      "loss": 5.0832,
      "mean_token_accuracy": 0.2100524440407753,
      "num_tokens": 30943699.0,
      "step": 13870
    },
    {
      "entropy": 5.52390570640564,
      "epoch": 1.5583197618913909,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20195478945970535,
      "num_tokens": 30954315.0,
      "step": 13875
    },
    {
      "entropy": 5.587045955657959,
      "epoch": 1.5588813387993485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.3222,
      "mean_token_accuracy": 0.20200438797473907,
      "num_tokens": 30965840.0,
      "step": 13880
    },
    {
      "entropy": 5.572679090499878,
      "epoch": 1.5594429157073062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.23,
      "mean_token_accuracy": 0.2035712018609047,
      "num_tokens": 30977051.0,
      "step": 13885
    },
    {
      "entropy": 5.471611785888672,
      "epoch": 1.5600044926152636,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.075,
      "mean_token_accuracy": 0.20643924176692963,
      "num_tokens": 30988730.0,
      "step": 13890
    },
    {
      "entropy": 5.4927205562591555,
      "epoch": 1.560566069523221,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20704305917024612,
      "num_tokens": 31000571.0,
      "step": 13895
    },
    {
      "entropy": 5.455665493011475,
      "epoch": 1.5611276464311787,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.1073,
      "mean_token_accuracy": 0.21378732919692994,
      "num_tokens": 31011003.0,
      "step": 13900
    },
    {
      "entropy": 5.5556834697723385,
      "epoch": 1.5616892233391364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.1266,
      "mean_token_accuracy": 0.21040140986442565,
      "num_tokens": 31022510.0,
      "step": 13905
    },
    {
      "entropy": 5.568290090560913,
      "epoch": 1.5622508002470938,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.1667,
      "mean_token_accuracy": 0.2082556277513504,
      "num_tokens": 31033149.0,
      "step": 13910
    },
    {
      "entropy": 5.548035955429077,
      "epoch": 1.5628123771550513,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.205,
      "mean_token_accuracy": 0.21040696352720262,
      "num_tokens": 31043626.0,
      "step": 13915
    },
    {
      "entropy": 5.549443626403809,
      "epoch": 1.563373954063009,
      "grad_norm": 1.0,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.2528,
      "mean_token_accuracy": 0.20009284913539888,
      "num_tokens": 31055083.0,
      "step": 13920
    },
    {
      "entropy": 5.5060193061828615,
      "epoch": 1.5639355309709666,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20298336446285248,
      "num_tokens": 31066763.0,
      "step": 13925
    },
    {
      "entropy": 5.53149561882019,
      "epoch": 1.564497107878924,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.1666,
      "mean_token_accuracy": 0.20081799179315568,
      "num_tokens": 31078722.0,
      "step": 13930
    },
    {
      "entropy": 5.562806034088135,
      "epoch": 1.5650586847868815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.2148,
      "mean_token_accuracy": 0.20478310137987138,
      "num_tokens": 31090207.0,
      "step": 13935
    },
    {
      "entropy": 5.541351556777954,
      "epoch": 1.5656202616948391,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.2345,
      "mean_token_accuracy": 0.2069670781493187,
      "num_tokens": 31101943.0,
      "step": 13940
    },
    {
      "entropy": 5.5199360847473145,
      "epoch": 1.5661818386027968,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19824124574661256,
      "num_tokens": 31115032.0,
      "step": 13945
    },
    {
      "entropy": 5.589374637603759,
      "epoch": 1.5667434155107542,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.1462,
      "mean_token_accuracy": 0.20419761091470717,
      "num_tokens": 31124965.0,
      "step": 13950
    },
    {
      "entropy": 5.559849596023559,
      "epoch": 1.5673049924187117,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20736181885004043,
      "num_tokens": 31136692.0,
      "step": 13955
    },
    {
      "entropy": 5.50918288230896,
      "epoch": 1.5678665693266693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20278095602989196,
      "num_tokens": 31148352.0,
      "step": 13960
    },
    {
      "entropy": 5.472424173355103,
      "epoch": 1.5684281462346268,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000476345798340041,
      "loss": 5.0945,
      "mean_token_accuracy": 0.2086835876107216,
      "num_tokens": 31159892.0,
      "step": 13965
    },
    {
      "entropy": 5.601951980590821,
      "epoch": 1.5689897231425842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.2434,
      "mean_token_accuracy": 0.20175336450338363,
      "num_tokens": 31170087.0,
      "step": 13970
    },
    {
      "entropy": 5.410476970672607,
      "epoch": 1.5695513000505419,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.0704,
      "mean_token_accuracy": 0.20607765913009643,
      "num_tokens": 31180687.0,
      "step": 13975
    },
    {
      "entropy": 5.572989225387573,
      "epoch": 1.5701128769584995,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.234,
      "mean_token_accuracy": 0.20023520588874816,
      "num_tokens": 31192513.0,
      "step": 13980
    },
    {
      "entropy": 5.532577276229858,
      "epoch": 1.570674453866457,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.1251,
      "mean_token_accuracy": 0.2049111470580101,
      "num_tokens": 31203432.0,
      "step": 13985
    },
    {
      "entropy": 5.5790708541870115,
      "epoch": 1.5712360307744144,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.0994,
      "mean_token_accuracy": 0.21238565891981126,
      "num_tokens": 31214474.0,
      "step": 13990
    },
    {
      "entropy": 5.513875484466553,
      "epoch": 1.571797607682372,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.1728,
      "mean_token_accuracy": 0.2011403724551201,
      "num_tokens": 31226489.0,
      "step": 13995
    },
    {
      "entropy": 5.442663097381592,
      "epoch": 1.5723591845903298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20568770170211792,
      "num_tokens": 31237568.0,
      "step": 14000
    },
    {
      "entropy": 5.496985864639282,
      "epoch": 1.5729207614982872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2177225798368454,
      "num_tokens": 31248587.0,
      "step": 14005
    },
    {
      "entropy": 5.534311532974243,
      "epoch": 1.5734823384062446,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.1811,
      "mean_token_accuracy": 0.20286870002746582,
      "num_tokens": 31259513.0,
      "step": 14010
    },
    {
      "entropy": 5.521858882904053,
      "epoch": 1.5740439153142023,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.1852,
      "mean_token_accuracy": 0.1992175430059433,
      "num_tokens": 31270984.0,
      "step": 14015
    },
    {
      "entropy": 5.476067399978637,
      "epoch": 1.57460549222216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20905451029539107,
      "num_tokens": 31281907.0,
      "step": 14020
    },
    {
      "entropy": 5.430789947509766,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.1382,
      "mean_token_accuracy": 0.20333491414785385,
      "num_tokens": 31292903.0,
      "step": 14025
    },
    {
      "entropy": 5.562652492523194,
      "epoch": 1.5757286460380748,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.1443,
      "mean_token_accuracy": 0.20979639142751694,
      "num_tokens": 31304185.0,
      "step": 14030
    },
    {
      "entropy": 5.526289987564087,
      "epoch": 1.5762902229460325,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.1303,
      "mean_token_accuracy": 0.2072632759809494,
      "num_tokens": 31314806.0,
      "step": 14035
    },
    {
      "entropy": 5.468735599517823,
      "epoch": 1.5768517998539902,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.1777,
      "mean_token_accuracy": 0.2036278784275055,
      "num_tokens": 31326361.0,
      "step": 14040
    },
    {
      "entropy": 5.591930294036866,
      "epoch": 1.5774133767619476,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.1556,
      "mean_token_accuracy": 0.2066408723592758,
      "num_tokens": 31336549.0,
      "step": 14045
    },
    {
      "entropy": 5.638899278640747,
      "epoch": 1.577974953669905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.2742,
      "mean_token_accuracy": 0.20070070624351502,
      "num_tokens": 31346994.0,
      "step": 14050
    },
    {
      "entropy": 5.544733333587646,
      "epoch": 1.5785365305778627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.1596,
      "mean_token_accuracy": 0.20502906888723374,
      "num_tokens": 31358449.0,
      "step": 14055
    },
    {
      "entropy": 5.571099185943604,
      "epoch": 1.5790981074858201,
      "grad_norm": 1.0,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.187,
      "mean_token_accuracy": 0.2026542067527771,
      "num_tokens": 31368964.0,
      "step": 14060
    },
    {
      "entropy": 5.5426260471344,
      "epoch": 1.5796596843937776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.1286,
      "mean_token_accuracy": 0.21134089380502702,
      "num_tokens": 31379388.0,
      "step": 14065
    },
    {
      "entropy": 5.486283445358277,
      "epoch": 1.5802212613017352,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.1915,
      "mean_token_accuracy": 0.2047051191329956,
      "num_tokens": 31390757.0,
      "step": 14070
    },
    {
      "entropy": 5.482009410858154,
      "epoch": 1.580782838209693,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.1219,
      "mean_token_accuracy": 0.20434726923704147,
      "num_tokens": 31402309.0,
      "step": 14075
    },
    {
      "entropy": 5.584751987457276,
      "epoch": 1.5813444151176503,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.1427,
      "mean_token_accuracy": 0.2115712508559227,
      "num_tokens": 31412559.0,
      "step": 14080
    },
    {
      "entropy": 5.529061412811279,
      "epoch": 1.5819059920256078,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.2788,
      "mean_token_accuracy": 0.1996172472834587,
      "num_tokens": 31424982.0,
      "step": 14085
    },
    {
      "entropy": 5.4861784934997555,
      "epoch": 1.5824675689335654,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2133765384554863,
      "num_tokens": 31434472.0,
      "step": 14090
    },
    {
      "entropy": 5.536374807357788,
      "epoch": 1.583029145841523,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.2377,
      "mean_token_accuracy": 0.20860762596130372,
      "num_tokens": 31445458.0,
      "step": 14095
    },
    {
      "entropy": 5.516174411773681,
      "epoch": 1.5835907227494805,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.1112,
      "mean_token_accuracy": 0.21201658844947815,
      "num_tokens": 31455851.0,
      "step": 14100
    },
    {
      "entropy": 5.413447475433349,
      "epoch": 1.584152299657438,
      "grad_norm": 1.0,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20738999992609025,
      "num_tokens": 31466279.0,
      "step": 14105
    },
    {
      "entropy": 5.560923528671265,
      "epoch": 1.5847138765653956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.1465,
      "mean_token_accuracy": 0.21184105575084686,
      "num_tokens": 31477933.0,
      "step": 14110
    },
    {
      "entropy": 5.565778684616089,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.1756,
      "mean_token_accuracy": 0.20060184448957444,
      "num_tokens": 31487640.0,
      "step": 14115
    },
    {
      "entropy": 5.583529806137085,
      "epoch": 1.5858370303813107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.119,
      "mean_token_accuracy": 0.21317314058542253,
      "num_tokens": 31499059.0,
      "step": 14120
    },
    {
      "entropy": 5.562082386016845,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.1469,
      "mean_token_accuracy": 0.20614176541566848,
      "num_tokens": 31510264.0,
      "step": 14125
    },
    {
      "entropy": 5.503358364105225,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.1443,
      "mean_token_accuracy": 0.20803655982017516,
      "num_tokens": 31520873.0,
      "step": 14130
    },
    {
      "entropy": 5.468314075469971,
      "epoch": 1.5875217611051835,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.1693,
      "mean_token_accuracy": 0.20396882444620132,
      "num_tokens": 31533836.0,
      "step": 14135
    },
    {
      "entropy": 5.498449516296387,
      "epoch": 1.588083338013141,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.1796,
      "mean_token_accuracy": 0.20280317962169647,
      "num_tokens": 31545091.0,
      "step": 14140
    },
    {
      "entropy": 5.5161748886108395,
      "epoch": 1.5886449149210984,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.0585,
      "mean_token_accuracy": 0.2096870630979538,
      "num_tokens": 31557139.0,
      "step": 14145
    },
    {
      "entropy": 5.577733755111694,
      "epoch": 1.589206491829056,
      "grad_norm": 1.0,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.1659,
      "mean_token_accuracy": 0.2066556006669998,
      "num_tokens": 31567928.0,
      "step": 14150
    },
    {
      "entropy": 5.462486839294433,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.1568,
      "mean_token_accuracy": 0.20705496221780778,
      "num_tokens": 31578081.0,
      "step": 14155
    },
    {
      "entropy": 5.44721736907959,
      "epoch": 1.590329645644971,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.0598,
      "mean_token_accuracy": 0.2118351548910141,
      "num_tokens": 31589648.0,
      "step": 14160
    },
    {
      "entropy": 5.5252820491790775,
      "epoch": 1.5908912225529286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20247109681367875,
      "num_tokens": 31601085.0,
      "step": 14165
    },
    {
      "entropy": 5.564582443237304,
      "epoch": 1.5914527994608862,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.2082,
      "mean_token_accuracy": 0.21005199402570723,
      "num_tokens": 31612005.0,
      "step": 14170
    },
    {
      "entropy": 5.526444959640503,
      "epoch": 1.5920143763688437,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.1589,
      "mean_token_accuracy": 0.20356072187423707,
      "num_tokens": 31623427.0,
      "step": 14175
    },
    {
      "entropy": 5.6283997058868405,
      "epoch": 1.5925759532768011,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.3073,
      "mean_token_accuracy": 0.19590914994478226,
      "num_tokens": 31635086.0,
      "step": 14180
    },
    {
      "entropy": 5.599794626235962,
      "epoch": 1.5931375301847588,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.1449,
      "mean_token_accuracy": 0.2067814975976944,
      "num_tokens": 31646707.0,
      "step": 14185
    },
    {
      "entropy": 5.456398010253906,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21116466373205184,
      "num_tokens": 31657327.0,
      "step": 14190
    },
    {
      "entropy": 5.476096296310425,
      "epoch": 1.5942606840006739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.0577,
      "mean_token_accuracy": 0.20986416786909104,
      "num_tokens": 31668075.0,
      "step": 14195
    },
    {
      "entropy": 5.513457345962524,
      "epoch": 1.5948222609086313,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19779542684555054,
      "num_tokens": 31680270.0,
      "step": 14200
    },
    {
      "entropy": 5.4570276737213135,
      "epoch": 1.595383837816589,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21382302790880203,
      "num_tokens": 31691271.0,
      "step": 14205
    },
    {
      "entropy": 5.475865411758423,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21226844936609268,
      "num_tokens": 31702982.0,
      "step": 14210
    },
    {
      "entropy": 5.525867080688476,
      "epoch": 1.596506991632504,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.1638,
      "mean_token_accuracy": 0.2070437416434288,
      "num_tokens": 31715129.0,
      "step": 14215
    },
    {
      "entropy": 5.547640132904053,
      "epoch": 1.5970685685404615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.2275,
      "mean_token_accuracy": 0.20543714761734008,
      "num_tokens": 31726392.0,
      "step": 14220
    },
    {
      "entropy": 5.565173006057739,
      "epoch": 1.5976301454484192,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.1945,
      "mean_token_accuracy": 0.2066534012556076,
      "num_tokens": 31738143.0,
      "step": 14225
    },
    {
      "entropy": 5.5848479747772215,
      "epoch": 1.5981917223563769,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.1579,
      "mean_token_accuracy": 0.2045840620994568,
      "num_tokens": 31748510.0,
      "step": 14230
    },
    {
      "entropy": 5.428285026550293,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047536881675345,
      "loss": 5.0489,
      "mean_token_accuracy": 0.212905615568161,
      "num_tokens": 31759378.0,
      "step": 14235
    },
    {
      "entropy": 5.631343221664428,
      "epoch": 1.5993148761722917,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19852713495492935,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.481202793121338,
      "epoch": 1.5998764530802494,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.0936,
      "mean_token_accuracy": 0.21173202991485596,
      "num_tokens": 31781990.0,
      "step": 14245
    },
    {
      "entropy": 5.492170000076294,
      "epoch": 1.6004380299882068,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.1388,
      "mean_token_accuracy": 0.2043069764971733,
      "num_tokens": 31792144.0,
      "step": 14250
    },
    {
      "entropy": 5.564090728759766,
      "epoch": 1.6009996068961643,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.1958,
      "mean_token_accuracy": 0.20518141835927964,
      "num_tokens": 31804022.0,
      "step": 14255
    },
    {
      "entropy": 5.614117431640625,
      "epoch": 1.601561183804122,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000475277413077462,
      "loss": 5.2139,
      "mean_token_accuracy": 0.2038545623421669,
      "num_tokens": 31815108.0,
      "step": 14260
    },
    {
      "entropy": 5.450232219696045,
      "epoch": 1.6021227607120796,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.1907,
      "mean_token_accuracy": 0.2060265690088272,
      "num_tokens": 31825727.0,
      "step": 14265
    },
    {
      "entropy": 5.498540639877319,
      "epoch": 1.602684337620037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.1422,
      "mean_token_accuracy": 0.21269199699163438,
      "num_tokens": 31836168.0,
      "step": 14270
    },
    {
      "entropy": 5.612802934646607,
      "epoch": 1.6032459145279945,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20002222061157227,
      "num_tokens": 31847383.0,
      "step": 14275
    },
    {
      "entropy": 5.483266735076905,
      "epoch": 1.6038074914359521,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.056,
      "mean_token_accuracy": 0.2167121335864067,
      "num_tokens": 31858679.0,
      "step": 14280
    },
    {
      "entropy": 5.505045938491821,
      "epoch": 1.6043690683439098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.1401,
      "mean_token_accuracy": 0.21236609518527985,
      "num_tokens": 31870195.0,
      "step": 14285
    },
    {
      "entropy": 5.593772459030151,
      "epoch": 1.6049306452518672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.1547,
      "mean_token_accuracy": 0.20616501718759536,
      "num_tokens": 31881175.0,
      "step": 14290
    },
    {
      "entropy": 5.489932727813721,
      "epoch": 1.6054922221598247,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.0911,
      "mean_token_accuracy": 0.21190257072448732,
      "num_tokens": 31891199.0,
      "step": 14295
    },
    {
      "entropy": 5.648785734176636,
      "epoch": 1.6060537990677823,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.2701,
      "mean_token_accuracy": 0.19871019572019577,
      "num_tokens": 31903379.0,
      "step": 14300
    },
    {
      "entropy": 5.561472988128662,
      "epoch": 1.60661537597574,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21333224773406984,
      "num_tokens": 31913958.0,
      "step": 14305
    },
    {
      "entropy": 5.422557926177978,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.0405,
      "mean_token_accuracy": 0.20705051720142365,
      "num_tokens": 31924547.0,
      "step": 14310
    },
    {
      "entropy": 5.588853549957276,
      "epoch": 1.6077385297916549,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.2535,
      "mean_token_accuracy": 0.2012529194355011,
      "num_tokens": 31936987.0,
      "step": 14315
    },
    {
      "entropy": 5.518535137176514,
      "epoch": 1.6083001066996125,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.1025,
      "mean_token_accuracy": 0.20575579553842543,
      "num_tokens": 31948165.0,
      "step": 14320
    },
    {
      "entropy": 5.572640466690063,
      "epoch": 1.6088616836075702,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.1788,
      "mean_token_accuracy": 0.20657920241355895,
      "num_tokens": 31959206.0,
      "step": 14325
    },
    {
      "entropy": 5.516082143783569,
      "epoch": 1.6094232605155276,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20659202784299852,
      "num_tokens": 31970804.0,
      "step": 14330
    },
    {
      "entropy": 5.466564178466797,
      "epoch": 1.609984837423485,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20953229069709778,
      "num_tokens": 31982961.0,
      "step": 14335
    },
    {
      "entropy": 5.544533634185791,
      "epoch": 1.6105464143314427,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.1266,
      "mean_token_accuracy": 0.20665303021669387,
      "num_tokens": 31993812.0,
      "step": 14340
    },
    {
      "entropy": 5.516867589950562,
      "epoch": 1.6111079912394002,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.177,
      "mean_token_accuracy": 0.20833884328603744,
      "num_tokens": 32004688.0,
      "step": 14345
    },
    {
      "entropy": 5.497816801071167,
      "epoch": 1.6116695681473576,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.0585,
      "mean_token_accuracy": 0.20898520350456237,
      "num_tokens": 32014653.0,
      "step": 14350
    },
    {
      "entropy": 5.458018207550049,
      "epoch": 1.6122311450553153,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20157098919153213,
      "num_tokens": 32025638.0,
      "step": 14355
    },
    {
      "entropy": 5.4036033153533936,
      "epoch": 1.612792721963273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21399628669023513,
      "num_tokens": 32036316.0,
      "step": 14360
    },
    {
      "entropy": 5.508919143676758,
      "epoch": 1.6133542988712304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.0696,
      "mean_token_accuracy": 0.21064760833978652,
      "num_tokens": 32047355.0,
      "step": 14365
    },
    {
      "entropy": 5.5209455490112305,
      "epoch": 1.6139158757791878,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.1777,
      "mean_token_accuracy": 0.19715942144393922,
      "num_tokens": 32058897.0,
      "step": 14370
    },
    {
      "entropy": 5.560714912414551,
      "epoch": 1.6144774526871455,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.1664,
      "mean_token_accuracy": 0.20463836193084717,
      "num_tokens": 32069932.0,
      "step": 14375
    },
    {
      "entropy": 5.506041812896728,
      "epoch": 1.6150390295951031,
      "grad_norm": 0.875,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.0717,
      "mean_token_accuracy": 0.2121717393398285,
      "num_tokens": 32081359.0,
      "step": 14380
    },
    {
      "entropy": 5.43913254737854,
      "epoch": 1.6156006065030606,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21074463278055192,
      "num_tokens": 32093050.0,
      "step": 14385
    },
    {
      "entropy": 5.566392135620117,
      "epoch": 1.616162183411018,
      "grad_norm": 1.0,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.186,
      "mean_token_accuracy": 0.20373099744319917,
      "num_tokens": 32104623.0,
      "step": 14390
    },
    {
      "entropy": 5.484420585632324,
      "epoch": 1.6167237603189757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.1351,
      "mean_token_accuracy": 0.19966809153556825,
      "num_tokens": 32115845.0,
      "step": 14395
    },
    {
      "entropy": 5.497075939178467,
      "epoch": 1.6172853372269334,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.0987,
      "mean_token_accuracy": 0.20521976053714752,
      "num_tokens": 32127066.0,
      "step": 14400
    },
    {
      "entropy": 5.52498779296875,
      "epoch": 1.6178469141348908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.227,
      "mean_token_accuracy": 0.2041291654109955,
      "num_tokens": 32138344.0,
      "step": 14405
    },
    {
      "entropy": 5.582254886627197,
      "epoch": 1.6184084910428482,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.183,
      "mean_token_accuracy": 0.1959277421236038,
      "num_tokens": 32149054.0,
      "step": 14410
    },
    {
      "entropy": 5.547164106369019,
      "epoch": 1.618970067950806,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.0674,
      "mean_token_accuracy": 0.20239914059638978,
      "num_tokens": 32160385.0,
      "step": 14415
    },
    {
      "entropy": 5.575809144973755,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19909197092056274,
      "num_tokens": 32171856.0,
      "step": 14420
    },
    {
      "entropy": 5.501068449020385,
      "epoch": 1.620093221766721,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.1627,
      "mean_token_accuracy": 0.20347215086221696,
      "num_tokens": 32182690.0,
      "step": 14425
    },
    {
      "entropy": 5.507587909698486,
      "epoch": 1.6206547986746784,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20808456987142562,
      "num_tokens": 32194045.0,
      "step": 14430
    },
    {
      "entropy": 5.519456481933593,
      "epoch": 1.621216375582636,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19897712767124176,
      "num_tokens": 32205699.0,
      "step": 14435
    },
    {
      "entropy": 5.575470066070556,
      "epoch": 1.6217779524905935,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.1678,
      "mean_token_accuracy": 0.2004276245832443,
      "num_tokens": 32217974.0,
      "step": 14440
    },
    {
      "entropy": 5.403668546676636,
      "epoch": 1.622339529398551,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.0836,
      "mean_token_accuracy": 0.20436247289180756,
      "num_tokens": 32228723.0,
      "step": 14445
    },
    {
      "entropy": 5.503937196731568,
      "epoch": 1.6229011063065086,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.2218,
      "mean_token_accuracy": 0.20240015238523484,
      "num_tokens": 32241048.0,
      "step": 14450
    },
    {
      "entropy": 5.501192235946656,
      "epoch": 1.6234626832144663,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.0958,
      "mean_token_accuracy": 0.21367053389549256,
      "num_tokens": 32251255.0,
      "step": 14455
    },
    {
      "entropy": 5.514215660095215,
      "epoch": 1.6240242601224237,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.1443,
      "mean_token_accuracy": 0.2045670509338379,
      "num_tokens": 32262954.0,
      "step": 14460
    },
    {
      "entropy": 5.512150287628174,
      "epoch": 1.6245858370303812,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.1427,
      "mean_token_accuracy": 0.2098095029592514,
      "num_tokens": 32275445.0,
      "step": 14465
    },
    {
      "entropy": 5.56142029762268,
      "epoch": 1.6251474139383388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20970887839794158,
      "num_tokens": 32285976.0,
      "step": 14470
    },
    {
      "entropy": 5.435481405258178,
      "epoch": 1.6257089908462965,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2117967948317528,
      "num_tokens": 32296304.0,
      "step": 14475
    },
    {
      "entropy": 5.478853178024292,
      "epoch": 1.626270567754254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.116,
      "mean_token_accuracy": 0.20966135263442992,
      "num_tokens": 32307085.0,
      "step": 14480
    },
    {
      "entropy": 5.502233076095581,
      "epoch": 1.6268321446622114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21873703151941298,
      "num_tokens": 32316814.0,
      "step": 14485
    },
    {
      "entropy": 5.489801502227783,
      "epoch": 1.627393721570169,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.1277,
      "mean_token_accuracy": 0.20210335701704024,
      "num_tokens": 32327820.0,
      "step": 14490
    },
    {
      "entropy": 5.573613166809082,
      "epoch": 1.6279552984781267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.115,
      "mean_token_accuracy": 0.20945221483707427,
      "num_tokens": 32338185.0,
      "step": 14495
    },
    {
      "entropy": 5.463333463668823,
      "epoch": 1.6285168753860841,
      "grad_norm": 1.0,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.0375,
      "mean_token_accuracy": 0.2103608950972557,
      "num_tokens": 32348537.0,
      "step": 14500
    },
    {
      "entropy": 5.5007518291473385,
      "epoch": 1.6290784522940416,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.1825,
      "mean_token_accuracy": 0.20297686606645585,
      "num_tokens": 32359094.0,
      "step": 14505
    },
    {
      "entropy": 5.403300762176514,
      "epoch": 1.6296400292019992,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047435462013406554,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22224227637052535,
      "num_tokens": 32370811.0,
      "step": 14510
    },
    {
      "entropy": 5.5734696865081785,
      "epoch": 1.630201606109957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.1956,
      "mean_token_accuracy": 0.20500603765249253,
      "num_tokens": 32381522.0,
      "step": 14515
    },
    {
      "entropy": 5.507951116561889,
      "epoch": 1.6307631830179143,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.1753,
      "mean_token_accuracy": 0.20020594894886018,
      "num_tokens": 32393221.0,
      "step": 14520
    },
    {
      "entropy": 5.5465960025787355,
      "epoch": 1.6313247599258718,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.1302,
      "mean_token_accuracy": 0.20057762861251832,
      "num_tokens": 32404007.0,
      "step": 14525
    },
    {
      "entropy": 5.54197506904602,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.2217,
      "mean_token_accuracy": 0.19898286908864976,
      "num_tokens": 32414318.0,
      "step": 14530
    },
    {
      "entropy": 5.4904083728790285,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.1421,
      "mean_token_accuracy": 0.21019266545772552,
      "num_tokens": 32425493.0,
      "step": 14535
    },
    {
      "entropy": 5.485447359085083,
      "epoch": 1.6330094906497443,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20637193322181702,
      "num_tokens": 32435765.0,
      "step": 14540
    },
    {
      "entropy": 5.608353948593139,
      "epoch": 1.633571067557702,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.1832,
      "mean_token_accuracy": 0.2112825706601143,
      "num_tokens": 32446478.0,
      "step": 14545
    },
    {
      "entropy": 5.45617561340332,
      "epoch": 1.6341326444656596,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.1856,
      "mean_token_accuracy": 0.2106230691075325,
      "num_tokens": 32456866.0,
      "step": 14550
    },
    {
      "entropy": 5.5487651348114015,
      "epoch": 1.634694221373617,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000474186829994371,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20999484211206437,
      "num_tokens": 32466973.0,
      "step": 14555
    },
    {
      "entropy": 5.6173012256622314,
      "epoch": 1.6352557982815745,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1980729416012764,
      "num_tokens": 32478118.0,
      "step": 14560
    },
    {
      "entropy": 5.4519189357757565,
      "epoch": 1.6358173751895322,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.0495,
      "mean_token_accuracy": 0.2145690754055977,
      "num_tokens": 32488255.0,
      "step": 14565
    },
    {
      "entropy": 5.531585931777954,
      "epoch": 1.6363789520974898,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.1481,
      "mean_token_accuracy": 0.20579595267772674,
      "num_tokens": 32498685.0,
      "step": 14570
    },
    {
      "entropy": 5.535092210769653,
      "epoch": 1.6369405290054473,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.1546,
      "mean_token_accuracy": 0.20406717658042908,
      "num_tokens": 32510457.0,
      "step": 14575
    },
    {
      "entropy": 5.49478178024292,
      "epoch": 1.6375021059134047,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.128,
      "mean_token_accuracy": 0.2062384232878685,
      "num_tokens": 32522495.0,
      "step": 14580
    },
    {
      "entropy": 5.62724289894104,
      "epoch": 1.6380636828213624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.2095,
      "mean_token_accuracy": 0.2029316782951355,
      "num_tokens": 32534891.0,
      "step": 14585
    },
    {
      "entropy": 5.5218908309936525,
      "epoch": 1.63862525972932,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.1058,
      "mean_token_accuracy": 0.20941483825445176,
      "num_tokens": 32545555.0,
      "step": 14590
    },
    {
      "entropy": 5.455060815811157,
      "epoch": 1.6391868366372775,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20962076187133788,
      "num_tokens": 32556851.0,
      "step": 14595
    },
    {
      "entropy": 5.5974822521209715,
      "epoch": 1.639748413545235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19591506570577621,
      "num_tokens": 32569028.0,
      "step": 14600
    },
    {
      "entropy": 5.695719623565674,
      "epoch": 1.6403099904531926,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20810769200325013,
      "num_tokens": 32581510.0,
      "step": 14605
    },
    {
      "entropy": 5.467253828048706,
      "epoch": 1.6408715673611503,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20671173334121704,
      "num_tokens": 32592914.0,
      "step": 14610
    },
    {
      "entropy": 5.4435234546661375,
      "epoch": 1.6414331442691077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21497218012809755,
      "num_tokens": 32603931.0,
      "step": 14615
    },
    {
      "entropy": 5.5915077209472654,
      "epoch": 1.6419947211770651,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.1883,
      "mean_token_accuracy": 0.20751513987779618,
      "num_tokens": 32614508.0,
      "step": 14620
    },
    {
      "entropy": 5.470487022399903,
      "epoch": 1.6425562980850228,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21915442794561385,
      "num_tokens": 32625393.0,
      "step": 14625
    },
    {
      "entropy": 5.408475351333618,
      "epoch": 1.6431178749929805,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.0375,
      "mean_token_accuracy": 0.21229948699474335,
      "num_tokens": 32636645.0,
      "step": 14630
    },
    {
      "entropy": 5.622018194198608,
      "epoch": 1.6436794519009377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.302,
      "mean_token_accuracy": 0.20249833762645722,
      "num_tokens": 32648792.0,
      "step": 14635
    },
    {
      "entropy": 5.574437284469605,
      "epoch": 1.6442410288088953,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.1235,
      "mean_token_accuracy": 0.21310717463493348,
      "num_tokens": 32658825.0,
      "step": 14640
    },
    {
      "entropy": 5.5891255855560305,
      "epoch": 1.644802605716853,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.1504,
      "mean_token_accuracy": 0.21385838687419892,
      "num_tokens": 32670141.0,
      "step": 14645
    },
    {
      "entropy": 5.340501928329468,
      "epoch": 1.6453641826248104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004738309198730219,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21885974556207657,
      "num_tokens": 32681804.0,
      "step": 14650
    },
    {
      "entropy": 5.624820899963379,
      "epoch": 1.6459257595327679,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.2328,
      "mean_token_accuracy": 0.19918057024478913,
      "num_tokens": 32692550.0,
      "step": 14655
    },
    {
      "entropy": 5.526116418838501,
      "epoch": 1.6464873364407255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21132478266954421,
      "num_tokens": 32703833.0,
      "step": 14660
    },
    {
      "entropy": 5.466999340057373,
      "epoch": 1.6470489133486832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.154,
      "mean_token_accuracy": 0.20270257145166398,
      "num_tokens": 32713928.0,
      "step": 14665
    },
    {
      "entropy": 5.567133092880249,
      "epoch": 1.6476104902566406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20907633751630783,
      "num_tokens": 32725393.0,
      "step": 14670
    },
    {
      "entropy": 5.605475616455078,
      "epoch": 1.648172067164598,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.173,
      "mean_token_accuracy": 0.20266627073287963,
      "num_tokens": 32736047.0,
      "step": 14675
    },
    {
      "entropy": 5.495849847793579,
      "epoch": 1.6487336440725557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.2059,
      "mean_token_accuracy": 0.19812060594558717,
      "num_tokens": 32747960.0,
      "step": 14680
    },
    {
      "entropy": 5.485668611526489,
      "epoch": 1.6492952209805134,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.1108,
      "mean_token_accuracy": 0.2108248144388199,
      "num_tokens": 32757851.0,
      "step": 14685
    },
    {
      "entropy": 5.497148227691651,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20798286348581313,
      "num_tokens": 32767317.0,
      "step": 14690
    },
    {
      "entropy": 5.583369398117066,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20787992775440217,
      "num_tokens": 32777729.0,
      "step": 14695
    },
    {
      "entropy": 5.5554908275604244,
      "epoch": 1.650979951704386,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1986207351088524,
      "num_tokens": 32789547.0,
      "step": 14700
    },
    {
      "entropy": 5.528878545761108,
      "epoch": 1.6515415286123436,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.1253,
      "mean_token_accuracy": 0.2078517496585846,
      "num_tokens": 32801487.0,
      "step": 14705
    },
    {
      "entropy": 5.58141508102417,
      "epoch": 1.652103105520301,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.1696,
      "mean_token_accuracy": 0.21396719068288803,
      "num_tokens": 32813042.0,
      "step": 14710
    },
    {
      "entropy": 5.58083028793335,
      "epoch": 1.6526646824282585,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20571644008159637,
      "num_tokens": 32825467.0,
      "step": 14715
    },
    {
      "entropy": 5.517534589767456,
      "epoch": 1.6532262593362161,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20248352736234665,
      "num_tokens": 32837113.0,
      "step": 14720
    },
    {
      "entropy": 5.624235534667969,
      "epoch": 1.6537878362441738,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.2349,
      "mean_token_accuracy": 0.19774530380964278,
      "num_tokens": 32849779.0,
      "step": 14725
    },
    {
      "entropy": 5.497866106033325,
      "epoch": 1.654349413152131,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.1517,
      "mean_token_accuracy": 0.20153157711029052,
      "num_tokens": 32861626.0,
      "step": 14730
    },
    {
      "entropy": 5.528966331481934,
      "epoch": 1.6549109900600887,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.1744,
      "mean_token_accuracy": 0.208600814640522,
      "num_tokens": 32872875.0,
      "step": 14735
    },
    {
      "entropy": 5.5064130306243895,
      "epoch": 1.6554725669680463,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20948169082403184,
      "num_tokens": 32883899.0,
      "step": 14740
    },
    {
      "entropy": 5.449746894836426,
      "epoch": 1.6560341438760038,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21500381082296371,
      "num_tokens": 32894255.0,
      "step": 14745
    },
    {
      "entropy": 5.5075325012207035,
      "epoch": 1.6565957207839612,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.123,
      "mean_token_accuracy": 0.20941878110170364,
      "num_tokens": 32903822.0,
      "step": 14750
    },
    {
      "entropy": 5.4984479427337645,
      "epoch": 1.6571572976919189,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.0823,
      "mean_token_accuracy": 0.21446591168642043,
      "num_tokens": 32915879.0,
      "step": 14755
    },
    {
      "entropy": 5.4515033721923825,
      "epoch": 1.6577188745998765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.1091,
      "mean_token_accuracy": 0.20575200468301774,
      "num_tokens": 32928161.0,
      "step": 14760
    },
    {
      "entropy": 5.501524639129639,
      "epoch": 1.658280451507834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.1001,
      "mean_token_accuracy": 0.206801176071167,
      "num_tokens": 32938399.0,
      "step": 14765
    },
    {
      "entropy": 5.481723737716675,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.064,
      "mean_token_accuracy": 0.210494664311409,
      "num_tokens": 32948221.0,
      "step": 14770
    },
    {
      "entropy": 5.4934789657592775,
      "epoch": 1.659403605323749,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.1318,
      "mean_token_accuracy": 0.21218484938144683,
      "num_tokens": 32959802.0,
      "step": 14775
    },
    {
      "entropy": 5.536651134490967,
      "epoch": 1.6599651822317067,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.1614,
      "mean_token_accuracy": 0.20603793412446975,
      "num_tokens": 32970222.0,
      "step": 14780
    },
    {
      "entropy": 5.526614618301392,
      "epoch": 1.6605267591396642,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21383655816316605,
      "num_tokens": 32982241.0,
      "step": 14785
    },
    {
      "entropy": 5.507622623443604,
      "epoch": 1.6610883360476216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21030376702547074,
      "num_tokens": 32992584.0,
      "step": 14790
    },
    {
      "entropy": 5.551192855834961,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.2437,
      "mean_token_accuracy": 0.20860573649406433,
      "num_tokens": 33004080.0,
      "step": 14795
    },
    {
      "entropy": 5.474023723602295,
      "epoch": 1.662211489863537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.085,
      "mean_token_accuracy": 0.2088903605937958,
      "num_tokens": 33015290.0,
      "step": 14800
    },
    {
      "entropy": 5.4976911544799805,
      "epoch": 1.6627730667714944,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21072752475738527,
      "num_tokens": 33026470.0,
      "step": 14805
    },
    {
      "entropy": 5.45592999458313,
      "epoch": 1.6633346436794518,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2189531669020653,
      "num_tokens": 33036766.0,
      "step": 14810
    },
    {
      "entropy": 5.448187160491943,
      "epoch": 1.6638962205874095,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.0638,
      "mean_token_accuracy": 0.20279725641012192,
      "num_tokens": 33047675.0,
      "step": 14815
    },
    {
      "entropy": 5.543229007720948,
      "epoch": 1.6644577974953672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000473188326686571,
      "loss": 5.1166,
      "mean_token_accuracy": 0.2134767085313797,
      "num_tokens": 33057613.0,
      "step": 14820
    },
    {
      "entropy": 5.60366644859314,
      "epoch": 1.6650193744033244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.2357,
      "mean_token_accuracy": 0.19590090364217758,
      "num_tokens": 33070498.0,
      "step": 14825
    },
    {
      "entropy": 5.48177809715271,
      "epoch": 1.665580951311282,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.0927,
      "mean_token_accuracy": 0.20773256421089173,
      "num_tokens": 33082026.0,
      "step": 14830
    },
    {
      "entropy": 5.506904363632202,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20448995679616927,
      "num_tokens": 33093841.0,
      "step": 14835
    },
    {
      "entropy": 5.569602108001709,
      "epoch": 1.6667041051271971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20851803719997405,
      "num_tokens": 33103817.0,
      "step": 14840
    },
    {
      "entropy": 5.566472482681275,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047309321185648,
      "loss": 5.168,
      "mean_token_accuracy": 0.20754391103982925,
      "num_tokens": 33116006.0,
      "step": 14845
    },
    {
      "entropy": 5.511340570449829,
      "epoch": 1.6678272589431122,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.1692,
      "mean_token_accuracy": 0.2033284291625023,
      "num_tokens": 33127649.0,
      "step": 14850
    },
    {
      "entropy": 5.414474439620972,
      "epoch": 1.66838883585107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004730551217686495,
      "loss": 5.0066,
      "mean_token_accuracy": 0.21824029237031936,
      "num_tokens": 33138748.0,
      "step": 14855
    },
    {
      "entropy": 5.566191101074219,
      "epoch": 1.6689504127590273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20400191098451614,
      "num_tokens": 33150504.0,
      "step": 14860
    },
    {
      "entropy": 5.510572814941407,
      "epoch": 1.6695119896669848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.208,
      "mean_token_accuracy": 0.19853705018758774,
      "num_tokens": 33160625.0,
      "step": 14865
    },
    {
      "entropy": 5.539209079742432,
      "epoch": 1.6700735665749424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.0915,
      "mean_token_accuracy": 0.21554587185382842,
      "num_tokens": 33172229.0,
      "step": 14870
    },
    {
      "entropy": 5.507149410247803,
      "epoch": 1.6706351434829,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20499349683523177,
      "num_tokens": 33183234.0,
      "step": 14875
    },
    {
      "entropy": 5.53568205833435,
      "epoch": 1.6711967203908575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.1712,
      "mean_token_accuracy": 0.2054738223552704,
      "num_tokens": 33193976.0,
      "step": 14880
    },
    {
      "entropy": 5.503268098831176,
      "epoch": 1.671758297298815,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.1303,
      "mean_token_accuracy": 0.21446665823459626,
      "num_tokens": 33205959.0,
      "step": 14885
    },
    {
      "entropy": 5.504506063461304,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21398627609014512,
      "num_tokens": 33218405.0,
      "step": 14890
    },
    {
      "entropy": 5.483625030517578,
      "epoch": 1.6728814511147303,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.129,
      "mean_token_accuracy": 0.20857305079698563,
      "num_tokens": 33230778.0,
      "step": 14895
    },
    {
      "entropy": 5.5011238098144535,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20994947999715804,
      "num_tokens": 33242069.0,
      "step": 14900
    },
    {
      "entropy": 5.5308201789855955,
      "epoch": 1.6740046049306452,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20929866135120392,
      "num_tokens": 33252942.0,
      "step": 14905
    },
    {
      "entropy": 5.6256012439727785,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.2058,
      "mean_token_accuracy": 0.20223486423492432,
      "num_tokens": 33263892.0,
      "step": 14910
    },
    {
      "entropy": 5.469962310791016,
      "epoch": 1.6751277587465605,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20481978207826615,
      "num_tokens": 33274383.0,
      "step": 14915
    },
    {
      "entropy": 5.554306268692017,
      "epoch": 1.6756893356545177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.0946,
      "mean_token_accuracy": 0.2082730233669281,
      "num_tokens": 33285897.0,
      "step": 14920
    },
    {
      "entropy": 5.569066524505615,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.1756,
      "mean_token_accuracy": 0.20453460961580278,
      "num_tokens": 33297192.0,
      "step": 14925
    },
    {
      "entropy": 5.4711119651794435,
      "epoch": 1.676812489470433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21253782212734224,
      "num_tokens": 33308250.0,
      "step": 14930
    },
    {
      "entropy": 5.4573499202728275,
      "epoch": 1.6773740663783905,
      "grad_norm": 1.0,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.0907,
      "mean_token_accuracy": 0.20850565135478974,
      "num_tokens": 33318515.0,
      "step": 14935
    },
    {
      "entropy": 5.543845558166504,
      "epoch": 1.677935643286348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.2006,
      "mean_token_accuracy": 0.20632969886064528,
      "num_tokens": 33329762.0,
      "step": 14940
    },
    {
      "entropy": 5.558840894699097,
      "epoch": 1.6784972201943056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20625799298286437,
      "num_tokens": 33341054.0,
      "step": 14945
    },
    {
      "entropy": 5.479928207397461,
      "epoch": 1.6790587971022632,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.0699,
      "mean_token_accuracy": 0.21087110042572021,
      "num_tokens": 33351524.0,
      "step": 14950
    },
    {
      "entropy": 5.49660210609436,
      "epoch": 1.6796203740102207,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20191365629434585,
      "num_tokens": 33362953.0,
      "step": 14955
    },
    {
      "entropy": 5.4467674732208256,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21632658839225768,
      "num_tokens": 33373209.0,
      "step": 14960
    },
    {
      "entropy": 5.4142402648925785,
      "epoch": 1.6807435278261358,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21061183512210846,
      "num_tokens": 33384775.0,
      "step": 14965
    },
    {
      "entropy": 5.420619487762451,
      "epoch": 1.6813051047340934,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20839086174964905,
      "num_tokens": 33395457.0,
      "step": 14970
    },
    {
      "entropy": 5.626844215393066,
      "epoch": 1.6818666816420509,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.2332,
      "mean_token_accuracy": 0.2096819117665291,
      "num_tokens": 33407506.0,
      "step": 14975
    },
    {
      "entropy": 5.5659871101379395,
      "epoch": 1.6824282585500083,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.1108,
      "mean_token_accuracy": 0.213632969558239,
      "num_tokens": 33418154.0,
      "step": 14980
    },
    {
      "entropy": 5.456899356842041,
      "epoch": 1.682989835457966,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.1387,
      "mean_token_accuracy": 0.21294301599264145,
      "num_tokens": 33428319.0,
      "step": 14985
    },
    {
      "entropy": 5.453951025009156,
      "epoch": 1.6835514123659236,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.0589,
      "mean_token_accuracy": 0.2113415852189064,
      "num_tokens": 33440429.0,
      "step": 14990
    },
    {
      "entropy": 5.671184921264649,
      "epoch": 1.684112989273881,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.285,
      "mean_token_accuracy": 0.1977385625243187,
      "num_tokens": 33451736.0,
      "step": 14995
    },
    {
      "entropy": 5.551936769485474,
      "epoch": 1.6846745661818385,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.057,
      "mean_token_accuracy": 0.20319695472717286,
      "num_tokens": 33461991.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.285287425602285,
      "eval_loss": 5.196067810058594,
      "eval_mean_token_accuracy": 0.21205611763339943,
      "eval_num_tokens": 33461991.0,
      "eval_runtime": 18.3302,
      "eval_samples_per_second": 1691.909,
      "eval_steps_per_second": 211.509,
      "step": 15000
    },
    {
      "entropy": 5.488405466079712,
      "epoch": 1.6852361430897962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.1872,
      "mean_token_accuracy": 0.2051481381058693,
      "num_tokens": 33473646.0,
      "step": 15005
    },
    {
      "entropy": 5.46118950843811,
      "epoch": 1.6857977199977539,
      "grad_norm": 0.921875,
      "learning_rate": 0.000472461502817644,
      "loss": 5.1171,
      "mean_token_accuracy": 0.20636711418628692,
      "num_tokens": 33485444.0,
      "step": 15010
    },
    {
      "entropy": 5.683063268661499,
      "epoch": 1.6863592969057113,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.2917,
      "mean_token_accuracy": 0.20151777565479279,
      "num_tokens": 33497199.0,
      "step": 15015
    },
    {
      "entropy": 5.614034986495971,
      "epoch": 1.6869208738136687,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.2226,
      "mean_token_accuracy": 0.20040614008903504,
      "num_tokens": 33509045.0,
      "step": 15020
    },
    {
      "entropy": 5.457744932174682,
      "epoch": 1.6874824507216264,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21683614253997802,
      "num_tokens": 33519730.0,
      "step": 15025
    },
    {
      "entropy": 5.498045539855957,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.109375,
      "learning_rate": 0.000472384466331328,
      "loss": 5.2122,
      "mean_token_accuracy": 0.19963295608758927,
      "num_tokens": 33530393.0,
      "step": 15030
    },
    {
      "entropy": 5.537956237792969,
      "epoch": 1.6886056045375413,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.1701,
      "mean_token_accuracy": 0.2074519544839859,
      "num_tokens": 33543592.0,
      "step": 15035
    },
    {
      "entropy": 5.568791246414184,
      "epoch": 1.689167181445499,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.1113,
      "mean_token_accuracy": 0.2074056401848793,
      "num_tokens": 33554736.0,
      "step": 15040
    },
    {
      "entropy": 5.515632295608521,
      "epoch": 1.6897287583534566,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21906666606664657,
      "num_tokens": 33565428.0,
      "step": 15045
    },
    {
      "entropy": 5.568249845504761,
      "epoch": 1.690290335261414,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.333,
      "mean_token_accuracy": 0.20112321078777312,
      "num_tokens": 33577102.0,
      "step": 15050
    },
    {
      "entropy": 5.584223985671997,
      "epoch": 1.6908519121693715,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.1819,
      "mean_token_accuracy": 0.21207108497619628,
      "num_tokens": 33587325.0,
      "step": 15055
    },
    {
      "entropy": 5.596794986724854,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.0,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.2007,
      "mean_token_accuracy": 0.20055781602859496,
      "num_tokens": 33599294.0,
      "step": 15060
    },
    {
      "entropy": 5.576909923553467,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.2655,
      "mean_token_accuracy": 0.2025859385728836,
      "num_tokens": 33609025.0,
      "step": 15065
    },
    {
      "entropy": 5.620525360107422,
      "epoch": 1.6925366428932442,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.172,
      "mean_token_accuracy": 0.20634727627038957,
      "num_tokens": 33620881.0,
      "step": 15070
    },
    {
      "entropy": 5.585985517501831,
      "epoch": 1.6930982198012017,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.1245,
      "mean_token_accuracy": 0.21221617609262466,
      "num_tokens": 33631505.0,
      "step": 15075
    },
    {
      "entropy": 5.566060304641724,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.2894,
      "mean_token_accuracy": 0.2000936508178711,
      "num_tokens": 33642560.0,
      "step": 15080
    },
    {
      "entropy": 5.556210565567016,
      "epoch": 1.694221373617117,
      "grad_norm": 1.0,
      "learning_rate": 0.000472172097616375,
      "loss": 5.1605,
      "mean_token_accuracy": 0.20886228531599044,
      "num_tokens": 33654706.0,
      "step": 15085
    },
    {
      "entropy": 5.598141050338745,
      "epoch": 1.6947829505250744,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19825393110513687,
      "num_tokens": 33665392.0,
      "step": 15090
    },
    {
      "entropy": 5.406635713577271,
      "epoch": 1.6953445274330319,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.0336,
      "mean_token_accuracy": 0.2157024547457695,
      "num_tokens": 33675504.0,
      "step": 15095
    },
    {
      "entropy": 5.537990665435791,
      "epoch": 1.6959061043409895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.1617,
      "mean_token_accuracy": 0.21118904799222946,
      "num_tokens": 33686505.0,
      "step": 15100
    },
    {
      "entropy": 5.599983215332031,
      "epoch": 1.6964676812489472,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.1656,
      "mean_token_accuracy": 0.2079371616244316,
      "num_tokens": 33698752.0,
      "step": 15105
    },
    {
      "entropy": 5.59257755279541,
      "epoch": 1.6970292581569046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.2168,
      "mean_token_accuracy": 0.19959573000669478,
      "num_tokens": 33710323.0,
      "step": 15110
    },
    {
      "entropy": 5.607518053054809,
      "epoch": 1.697590835064862,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19143851697444916,
      "num_tokens": 33721875.0,
      "step": 15115
    },
    {
      "entropy": 5.514672470092774,
      "epoch": 1.6981524119728197,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004720365583207737,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2176821455359459,
      "num_tokens": 33731520.0,
      "step": 15120
    },
    {
      "entropy": 5.5842174053192135,
      "epoch": 1.6987139888807772,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.2537,
      "mean_token_accuracy": 0.2009032502770424,
      "num_tokens": 33743351.0,
      "step": 15125
    },
    {
      "entropy": 5.5804157733917235,
      "epoch": 1.6992755657887346,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.1781,
      "mean_token_accuracy": 0.20825228989124298,
      "num_tokens": 33754879.0,
      "step": 15130
    },
    {
      "entropy": 5.495001888275146,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2077656626701355,
      "num_tokens": 33766671.0,
      "step": 15135
    },
    {
      "entropy": 5.572140789031982,
      "epoch": 1.70039871960465,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.1425,
      "mean_token_accuracy": 0.2050899237394333,
      "num_tokens": 33778236.0,
      "step": 15140
    },
    {
      "entropy": 5.517440557479858,
      "epoch": 1.7009602965126074,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2109249711036682,
      "num_tokens": 33788898.0,
      "step": 15145
    },
    {
      "entropy": 5.529820585250855,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.2083,
      "mean_token_accuracy": 0.20293100625276567,
      "num_tokens": 33800641.0,
      "step": 15150
    },
    {
      "entropy": 5.604246997833252,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.1213,
      "mean_token_accuracy": 0.20573474168777467,
      "num_tokens": 33812581.0,
      "step": 15155
    },
    {
      "entropy": 5.468530368804932,
      "epoch": 1.7026450272364801,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21762901395559311,
      "num_tokens": 33823378.0,
      "step": 15160
    },
    {
      "entropy": 5.525546979904175,
      "epoch": 1.7032066041444376,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.1804,
      "mean_token_accuracy": 0.20358660966157913,
      "num_tokens": 33834804.0,
      "step": 15165
    },
    {
      "entropy": 5.493289232254028,
      "epoch": 1.703768181052395,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.1338,
      "mean_token_accuracy": 0.20673038810491562,
      "num_tokens": 33847044.0,
      "step": 15170
    },
    {
      "entropy": 5.565756845474243,
      "epoch": 1.7043297579603527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20107275396585464,
      "num_tokens": 33858077.0,
      "step": 15175
    },
    {
      "entropy": 5.507013273239136,
      "epoch": 1.7048913348683103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20795485228300095,
      "num_tokens": 33868765.0,
      "step": 15180
    },
    {
      "entropy": 5.499920463562011,
      "epoch": 1.7054529117762678,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.0813,
      "mean_token_accuracy": 0.2124208465218544,
      "num_tokens": 33880415.0,
      "step": 15185
    },
    {
      "entropy": 5.544648313522339,
      "epoch": 1.7060144886842252,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20401975959539415,
      "num_tokens": 33892081.0,
      "step": 15190
    },
    {
      "entropy": 5.584453916549682,
      "epoch": 1.706576065592183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.2645,
      "mean_token_accuracy": 0.1991408497095108,
      "num_tokens": 33903433.0,
      "step": 15195
    },
    {
      "entropy": 5.555294322967529,
      "epoch": 1.7071376425001406,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.1312,
      "mean_token_accuracy": 0.20572934299707413,
      "num_tokens": 33914706.0,
      "step": 15200
    },
    {
      "entropy": 5.519938659667969,
      "epoch": 1.707699219408098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.2431,
      "mean_token_accuracy": 0.2060904935002327,
      "num_tokens": 33927195.0,
      "step": 15205
    },
    {
      "entropy": 5.563093852996826,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.128,
      "mean_token_accuracy": 0.20726257115602492,
      "num_tokens": 33938422.0,
      "step": 15210
    },
    {
      "entropy": 5.6106781482696535,
      "epoch": 1.708822373224013,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.1633,
      "mean_token_accuracy": 0.21017622500658034,
      "num_tokens": 33949775.0,
      "step": 15215
    },
    {
      "entropy": 5.51922836303711,
      "epoch": 1.7093839501319705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20598908066749572,
      "num_tokens": 33960371.0,
      "step": 15220
    },
    {
      "entropy": 5.428254652023315,
      "epoch": 1.709945527039928,
      "grad_norm": 1.0,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20852178633213042,
      "num_tokens": 33971326.0,
      "step": 15225
    },
    {
      "entropy": 5.568927574157715,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20287554264068602,
      "num_tokens": 33982744.0,
      "step": 15230
    },
    {
      "entropy": 5.594589948654175,
      "epoch": 1.7110686808558433,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.229,
      "mean_token_accuracy": 0.19981864839792252,
      "num_tokens": 33993647.0,
      "step": 15235
    },
    {
      "entropy": 5.495735549926758,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.092,
      "mean_token_accuracy": 0.20581635981798171,
      "num_tokens": 34004026.0,
      "step": 15240
    },
    {
      "entropy": 5.504519128799439,
      "epoch": 1.7121918346717582,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.1674,
      "mean_token_accuracy": 0.2096847727894783,
      "num_tokens": 34014882.0,
      "step": 15245
    },
    {
      "entropy": 5.521789073944092,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.1832,
      "mean_token_accuracy": 0.2043699637055397,
      "num_tokens": 34026548.0,
      "step": 15250
    },
    {
      "entropy": 5.507537126541138,
      "epoch": 1.7133149884876735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000471510886150377,
      "loss": 5.047,
      "mean_token_accuracy": 0.21090672314167022,
      "num_tokens": 34038393.0,
      "step": 15255
    },
    {
      "entropy": 5.56649923324585,
      "epoch": 1.713876565395631,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.1586,
      "mean_token_accuracy": 0.20567502826452255,
      "num_tokens": 34050257.0,
      "step": 15260
    },
    {
      "entropy": 5.537067890167236,
      "epoch": 1.7144381423035884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.131,
      "mean_token_accuracy": 0.20895869731903077,
      "num_tokens": 34061860.0,
      "step": 15265
    },
    {
      "entropy": 5.595859384536743,
      "epoch": 1.714999719211546,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20558800995349885,
      "num_tokens": 34075247.0,
      "step": 15270
    },
    {
      "entropy": 5.535283994674683,
      "epoch": 1.7155612961195037,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20510610789060593,
      "num_tokens": 34085801.0,
      "step": 15275
    },
    {
      "entropy": 5.554241180419922,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20347414463758468,
      "num_tokens": 34095686.0,
      "step": 15280
    },
    {
      "entropy": 5.499098300933838,
      "epoch": 1.7166844499354186,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.0954,
      "mean_token_accuracy": 0.2141059458255768,
      "num_tokens": 34106518.0,
      "step": 15285
    },
    {
      "entropy": 5.540509986877441,
      "epoch": 1.7172460268433762,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.1411,
      "mean_token_accuracy": 0.21340920329093932,
      "num_tokens": 34117793.0,
      "step": 15290
    },
    {
      "entropy": 5.586933135986328,
      "epoch": 1.717807603751334,
      "grad_norm": 1.078125,
      "learning_rate": 0.000471354254904924,
      "loss": 5.2177,
      "mean_token_accuracy": 0.20775901079177855,
      "num_tokens": 34128748.0,
      "step": 15295
    },
    {
      "entropy": 5.559786701202393,
      "epoch": 1.7183691806592913,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.1472,
      "mean_token_accuracy": 0.21191691756248474,
      "num_tokens": 34140597.0,
      "step": 15300
    },
    {
      "entropy": 5.616422843933106,
      "epoch": 1.7189307575672488,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000471315034561129,
      "loss": 5.125,
      "mean_token_accuracy": 0.19880042374134063,
      "num_tokens": 34151823.0,
      "step": 15305
    },
    {
      "entropy": 5.46340651512146,
      "epoch": 1.7194923344752064,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.073,
      "mean_token_accuracy": 0.21487921625375747,
      "num_tokens": 34163697.0,
      "step": 15310
    },
    {
      "entropy": 5.5131231307983395,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21221717000007628,
      "num_tokens": 34175058.0,
      "step": 15315
    },
    {
      "entropy": 5.569633531570434,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.2398,
      "mean_token_accuracy": 0.2053506478667259,
      "num_tokens": 34186883.0,
      "step": 15320
    },
    {
      "entropy": 5.535079908370972,
      "epoch": 1.721177065199079,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20897285491228104,
      "num_tokens": 34198103.0,
      "step": 15325
    },
    {
      "entropy": 5.5458731174469,
      "epoch": 1.7217386421070366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.0922,
      "mean_token_accuracy": 0.20799963772296906,
      "num_tokens": 34208747.0,
      "step": 15330
    },
    {
      "entropy": 5.530550003051758,
      "epoch": 1.722300219014994,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.2298,
      "mean_token_accuracy": 0.20584967881441116,
      "num_tokens": 34220775.0,
      "step": 15335
    },
    {
      "entropy": 5.556913280487061,
      "epoch": 1.7228617959229515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.075,
      "mean_token_accuracy": 0.21244683116674423,
      "num_tokens": 34231031.0,
      "step": 15340
    },
    {
      "entropy": 5.587532234191895,
      "epoch": 1.7234233728309092,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.1889,
      "mean_token_accuracy": 0.20807191133499145,
      "num_tokens": 34241850.0,
      "step": 15345
    },
    {
      "entropy": 5.527195596694947,
      "epoch": 1.7239849497388668,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.0893,
      "mean_token_accuracy": 0.21200267225503922,
      "num_tokens": 34252732.0,
      "step": 15350
    },
    {
      "entropy": 5.469162654876709,
      "epoch": 1.7245465266468243,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.08,
      "mean_token_accuracy": 0.20750088691711427,
      "num_tokens": 34264066.0,
      "step": 15355
    },
    {
      "entropy": 5.599398136138916,
      "epoch": 1.7251081035547817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.2007,
      "mean_token_accuracy": 0.19786975681781768,
      "num_tokens": 34275239.0,
      "step": 15360
    },
    {
      "entropy": 5.484118890762329,
      "epoch": 1.7256696804627394,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20345867425203323,
      "num_tokens": 34285723.0,
      "step": 15365
    },
    {
      "entropy": 5.533049201965332,
      "epoch": 1.726231257370697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.1149,
      "mean_token_accuracy": 0.2092798575758934,
      "num_tokens": 34296292.0,
      "step": 15370
    },
    {
      "entropy": 5.54659333229065,
      "epoch": 1.7267928342786545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.1398,
      "mean_token_accuracy": 0.21544404029846193,
      "num_tokens": 34307232.0,
      "step": 15375
    },
    {
      "entropy": 5.524669933319092,
      "epoch": 1.727354411186612,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.1458,
      "mean_token_accuracy": 0.21583929359912873,
      "num_tokens": 34318933.0,
      "step": 15380
    },
    {
      "entropy": 5.541958045959473,
      "epoch": 1.7279159880945696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.0764,
      "mean_token_accuracy": 0.21485261470079423,
      "num_tokens": 34330431.0,
      "step": 15385
    },
    {
      "entropy": 5.5624940395355225,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.1741,
      "mean_token_accuracy": 0.20924707055091857,
      "num_tokens": 34342006.0,
      "step": 15390
    },
    {
      "entropy": 5.564741611480713,
      "epoch": 1.7290391419104847,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.0918,
      "mean_token_accuracy": 0.208769354224205,
      "num_tokens": 34353390.0,
      "step": 15395
    },
    {
      "entropy": 5.468025588989258,
      "epoch": 1.7296007188184421,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.0917,
      "mean_token_accuracy": 0.20867200195789337,
      "num_tokens": 34364211.0,
      "step": 15400
    },
    {
      "entropy": 5.523447513580322,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.099,
      "mean_token_accuracy": 0.20473264902830124,
      "num_tokens": 34375362.0,
      "step": 15405
    },
    {
      "entropy": 5.448624849319458,
      "epoch": 1.7307238726343572,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004709017123528462,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21700740456581116,
      "num_tokens": 34385511.0,
      "step": 15410
    },
    {
      "entropy": 5.521172094345093,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.2322,
      "mean_token_accuracy": 0.20273891538381578,
      "num_tokens": 34397060.0,
      "step": 15415
    },
    {
      "entropy": 5.52694616317749,
      "epoch": 1.7318470264502723,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.1265,
      "mean_token_accuracy": 0.21140242069959642,
      "num_tokens": 34407703.0,
      "step": 15420
    },
    {
      "entropy": 5.548181629180908,
      "epoch": 1.73240860335823,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.1593,
      "mean_token_accuracy": 0.2107311710715294,
      "num_tokens": 34418929.0,
      "step": 15425
    },
    {
      "entropy": 5.471923065185547,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21309733539819717,
      "num_tokens": 34430693.0,
      "step": 15430
    },
    {
      "entropy": 5.468740177154541,
      "epoch": 1.7335317571741449,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20820690244436263,
      "num_tokens": 34442299.0,
      "step": 15435
    },
    {
      "entropy": 5.559539794921875,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.076,
      "mean_token_accuracy": 0.21020232886075974,
      "num_tokens": 34453426.0,
      "step": 15440
    },
    {
      "entropy": 5.511450624465942,
      "epoch": 1.7346549109900602,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21371775567531587,
      "num_tokens": 34463601.0,
      "step": 15445
    },
    {
      "entropy": 5.533650732040405,
      "epoch": 1.7352164878980176,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21388920843601228,
      "num_tokens": 34474938.0,
      "step": 15450
    },
    {
      "entropy": 5.46483154296875,
      "epoch": 1.735778064805975,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21737215667963028,
      "num_tokens": 34485244.0,
      "step": 15455
    },
    {
      "entropy": 5.448960876464843,
      "epoch": 1.7363396417139327,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21710579693317414,
      "num_tokens": 34496912.0,
      "step": 15460
    },
    {
      "entropy": 5.506086540222168,
      "epoch": 1.7369012186218904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000470684112060605,
      "loss": 5.121,
      "mean_token_accuracy": 0.20519276857376098,
      "num_tokens": 34508122.0,
      "step": 15465
    },
    {
      "entropy": 5.553223752975464,
      "epoch": 1.7374627955298478,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.1131,
      "mean_token_accuracy": 0.21308030039072037,
      "num_tokens": 34519522.0,
      "step": 15470
    },
    {
      "entropy": 5.513301706314087,
      "epoch": 1.7380243724378053,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000470644467347967,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21242327094078065,
      "num_tokens": 34532832.0,
      "step": 15475
    },
    {
      "entropy": 5.551942634582519,
      "epoch": 1.738585949345763,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.2244,
      "mean_token_accuracy": 0.20605125278234482,
      "num_tokens": 34544817.0,
      "step": 15480
    },
    {
      "entropy": 5.462670516967774,
      "epoch": 1.7391475262537206,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.0479,
      "mean_token_accuracy": 0.20990079194307326,
      "num_tokens": 34555391.0,
      "step": 15485
    },
    {
      "entropy": 5.448474502563476,
      "epoch": 1.739709103161678,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.0781,
      "mean_token_accuracy": 0.2070907920598984,
      "num_tokens": 34567531.0,
      "step": 15490
    },
    {
      "entropy": 5.545671510696411,
      "epoch": 1.7402706800696355,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.1355,
      "mean_token_accuracy": 0.2084236115217209,
      "num_tokens": 34578212.0,
      "step": 15495
    },
    {
      "entropy": 5.4476744651794435,
      "epoch": 1.7408322569775931,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21814000457525254,
      "num_tokens": 34589331.0,
      "step": 15500
    },
    {
      "entropy": 5.455878448486328,
      "epoch": 1.7413938338855506,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20962044447660447,
      "num_tokens": 34601467.0,
      "step": 15505
    },
    {
      "entropy": 5.480067300796509,
      "epoch": 1.741955410793508,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.0927,
      "mean_token_accuracy": 0.2180152490735054,
      "num_tokens": 34611608.0,
      "step": 15510
    },
    {
      "entropy": 5.5293854713439945,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21660980582237244,
      "num_tokens": 34622495.0,
      "step": 15515
    },
    {
      "entropy": 5.605437469482422,
      "epoch": 1.7430785646094233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.2544,
      "mean_token_accuracy": 0.20256226807832717,
      "num_tokens": 34633279.0,
      "step": 15520
    },
    {
      "entropy": 5.4556904315948485,
      "epoch": 1.7436401415173808,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20959815979003907,
      "num_tokens": 34644030.0,
      "step": 15525
    },
    {
      "entropy": 5.619700574874878,
      "epoch": 1.7442017184253382,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.1948,
      "mean_token_accuracy": 0.2011444538831711,
      "num_tokens": 34655540.0,
      "step": 15530
    },
    {
      "entropy": 5.55294280052185,
      "epoch": 1.7447632953332959,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20596397221088408,
      "num_tokens": 34666752.0,
      "step": 15535
    },
    {
      "entropy": 5.550954389572143,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.1198,
      "mean_token_accuracy": 0.2051733285188675,
      "num_tokens": 34678670.0,
      "step": 15540
    },
    {
      "entropy": 5.529381179809571,
      "epoch": 1.745886449149211,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.1545,
      "mean_token_accuracy": 0.2047477051615715,
      "num_tokens": 34690848.0,
      "step": 15545
    },
    {
      "entropy": 5.552477025985718,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20235341191291809,
      "num_tokens": 34703135.0,
      "step": 15550
    },
    {
      "entropy": 5.534253406524658,
      "epoch": 1.747009602965126,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.0939,
      "mean_token_accuracy": 0.20791334807872772,
      "num_tokens": 34714329.0,
      "step": 15555
    },
    {
      "entropy": 5.5364068984985355,
      "epoch": 1.7475711798730837,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.162,
      "mean_token_accuracy": 0.20220322161912918,
      "num_tokens": 34727032.0,
      "step": 15560
    },
    {
      "entropy": 5.554045391082764,
      "epoch": 1.7481327567810412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20482802838087083,
      "num_tokens": 34738600.0,
      "step": 15565
    },
    {
      "entropy": 5.599101305007935,
      "epoch": 1.7486943336889986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.1952,
      "mean_token_accuracy": 0.2024545356631279,
      "num_tokens": 34749810.0,
      "step": 15570
    },
    {
      "entropy": 5.708774089813232,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.125,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.3372,
      "mean_token_accuracy": 0.1986189678311348,
      "num_tokens": 34762029.0,
      "step": 15575
    },
    {
      "entropy": 5.497018527984619,
      "epoch": 1.749817487504914,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.1918,
      "mean_token_accuracy": 0.20458855479955673,
      "num_tokens": 34773160.0,
      "step": 15580
    },
    {
      "entropy": 5.490473842620849,
      "epoch": 1.7503790644128714,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20299303978681565,
      "num_tokens": 34784659.0,
      "step": 15585
    },
    {
      "entropy": 5.614253759384155,
      "epoch": 1.7509406413208288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000470186763474261,
      "loss": 5.1468,
      "mean_token_accuracy": 0.21239375025033952,
      "num_tokens": 34796976.0,
      "step": 15590
    },
    {
      "entropy": 5.453258275985718,
      "epoch": 1.7515022182287865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.0763,
      "mean_token_accuracy": 0.20890699774026872,
      "num_tokens": 34809271.0,
      "step": 15595
    },
    {
      "entropy": 5.523576879501343,
      "epoch": 1.752063795136744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.1586,
      "mean_token_accuracy": 0.20885020047426223,
      "num_tokens": 34820291.0,
      "step": 15600
    },
    {
      "entropy": 5.481470251083374,
      "epoch": 1.7526253720447014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21541340947151183,
      "num_tokens": 34831224.0,
      "step": 15605
    },
    {
      "entropy": 5.572166919708252,
      "epoch": 1.753186948952659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.1392,
      "mean_token_accuracy": 0.20931598246097566,
      "num_tokens": 34842947.0,
      "step": 15610
    },
    {
      "entropy": 5.488855218887329,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.0868,
      "mean_token_accuracy": 0.21649779379367828,
      "num_tokens": 34853422.0,
      "step": 15615
    },
    {
      "entropy": 5.456126689910889,
      "epoch": 1.7543101027685741,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.1044,
      "mean_token_accuracy": 0.20488331615924835,
      "num_tokens": 34864082.0,
      "step": 15620
    },
    {
      "entropy": 5.498503923416138,
      "epoch": 1.7548716796765316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.0884,
      "mean_token_accuracy": 0.22038241773843764,
      "num_tokens": 34874094.0,
      "step": 15625
    },
    {
      "entropy": 5.484879875183106,
      "epoch": 1.7554332565844892,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.1205,
      "mean_token_accuracy": 0.21539092808961868,
      "num_tokens": 34884597.0,
      "step": 15630
    },
    {
      "entropy": 5.413533496856689,
      "epoch": 1.755994833492447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047000676659768477,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21671170443296434,
      "num_tokens": 34894746.0,
      "step": 15635
    },
    {
      "entropy": 5.4755956649780275,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21731432676315307,
      "num_tokens": 34906250.0,
      "step": 15640
    },
    {
      "entropy": 5.585721778869629,
      "epoch": 1.7571179873083618,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.2862,
      "mean_token_accuracy": 0.19965091347694397,
      "num_tokens": 34917772.0,
      "step": 15645
    },
    {
      "entropy": 5.562360668182373,
      "epoch": 1.7576795642163194,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.1134,
      "mean_token_accuracy": 0.21236780434846877,
      "num_tokens": 34930243.0,
      "step": 15650
    },
    {
      "entropy": 5.521789789199829,
      "epoch": 1.758241141124277,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20728515684604645,
      "num_tokens": 34941357.0,
      "step": 15655
    },
    {
      "entropy": 5.625071001052857,
      "epoch": 1.7588027180322345,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.2508,
      "mean_token_accuracy": 0.19521945267915725,
      "num_tokens": 34952116.0,
      "step": 15660
    },
    {
      "entropy": 5.550064468383789,
      "epoch": 1.759364294940192,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.1593,
      "mean_token_accuracy": 0.20651935786008835,
      "num_tokens": 34963707.0,
      "step": 15665
    },
    {
      "entropy": 5.605381917953491,
      "epoch": 1.7599258718481496,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.2543,
      "mean_token_accuracy": 0.2031252086162567,
      "num_tokens": 34974376.0,
      "step": 15670
    },
    {
      "entropy": 5.559389400482178,
      "epoch": 1.7604874487561073,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.135,
      "mean_token_accuracy": 0.20830572992563248,
      "num_tokens": 34986183.0,
      "step": 15675
    },
    {
      "entropy": 5.622667121887207,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.1572,
      "mean_token_accuracy": 0.211855186522007,
      "num_tokens": 34997137.0,
      "step": 15680
    },
    {
      "entropy": 5.6183342933654785,
      "epoch": 1.7616106025720222,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.3041,
      "mean_token_accuracy": 0.20137612223625184,
      "num_tokens": 35008944.0,
      "step": 15685
    },
    {
      "entropy": 5.572197914123535,
      "epoch": 1.7621721794799798,
      "grad_norm": 1.0,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.2018,
      "mean_token_accuracy": 0.1983625516295433,
      "num_tokens": 35019972.0,
      "step": 15690
    },
    {
      "entropy": 5.5080718994140625,
      "epoch": 1.7627337563879373,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.0775,
      "mean_token_accuracy": 0.21350693851709365,
      "num_tokens": 35031285.0,
      "step": 15695
    },
    {
      "entropy": 5.485590600967408,
      "epoch": 1.7632953332958947,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.0508,
      "mean_token_accuracy": 0.20915182828903198,
      "num_tokens": 35042553.0,
      "step": 15700
    },
    {
      "entropy": 5.646862936019898,
      "epoch": 1.7638569102038524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.2673,
      "mean_token_accuracy": 0.19416423738002778,
      "num_tokens": 35053637.0,
      "step": 15705
    },
    {
      "entropy": 5.547137355804443,
      "epoch": 1.76441848711181,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.2019,
      "mean_token_accuracy": 0.20696668475866317,
      "num_tokens": 35064754.0,
      "step": 15710
    },
    {
      "entropy": 5.55148549079895,
      "epoch": 1.7649800640197675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.1059,
      "mean_token_accuracy": 0.20681236833333969,
      "num_tokens": 35075816.0,
      "step": 15715
    },
    {
      "entropy": 5.565423202514649,
      "epoch": 1.765541640927725,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.21,
      "mean_token_accuracy": 0.2051266387104988,
      "num_tokens": 35088647.0,
      "step": 15720
    },
    {
      "entropy": 5.548673009872436,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.0864,
      "mean_token_accuracy": 0.20676093697547912,
      "num_tokens": 35099372.0,
      "step": 15725
    },
    {
      "entropy": 5.58281364440918,
      "epoch": 1.7666647947436402,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046962512213995,
      "loss": 5.2083,
      "mean_token_accuracy": 0.20361634194850922,
      "num_tokens": 35110663.0,
      "step": 15730
    },
    {
      "entropy": 5.699960803985595,
      "epoch": 1.7672263716515977,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.3113,
      "mean_token_accuracy": 0.19347182363271714,
      "num_tokens": 35122406.0,
      "step": 15735
    },
    {
      "entropy": 5.507610654830932,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20949216932058334,
      "num_tokens": 35133444.0,
      "step": 15740
    },
    {
      "entropy": 5.476340627670288,
      "epoch": 1.7683495254675128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.115,
      "mean_token_accuracy": 0.20817748308181763,
      "num_tokens": 35143840.0,
      "step": 15745
    },
    {
      "entropy": 5.609024477005005,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.2838,
      "mean_token_accuracy": 0.19736655801534653,
      "num_tokens": 35155050.0,
      "step": 15750
    },
    {
      "entropy": 5.590265274047852,
      "epoch": 1.7694726792834279,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.2027,
      "mean_token_accuracy": 0.20334538966417312,
      "num_tokens": 35167649.0,
      "step": 15755
    },
    {
      "entropy": 5.540337610244751,
      "epoch": 1.7700342561913853,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.1248,
      "mean_token_accuracy": 0.21156666427850723,
      "num_tokens": 35179238.0,
      "step": 15760
    },
    {
      "entropy": 5.596830224990844,
      "epoch": 1.770595833099343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.1859,
      "mean_token_accuracy": 0.206203655898571,
      "num_tokens": 35189510.0,
      "step": 15765
    },
    {
      "entropy": 5.607501268386841,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20511122792959213,
      "num_tokens": 35199727.0,
      "step": 15770
    },
    {
      "entropy": 5.488244247436524,
      "epoch": 1.771718986915258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.1177,
      "mean_token_accuracy": 0.2096966177225113,
      "num_tokens": 35210311.0,
      "step": 15775
    },
    {
      "entropy": 5.552104711532593,
      "epoch": 1.7722805638232155,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.1485,
      "mean_token_accuracy": 0.21224112659692765,
      "num_tokens": 35222755.0,
      "step": 15780
    },
    {
      "entropy": 5.575040626525879,
      "epoch": 1.7728421407311732,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.178,
      "mean_token_accuracy": 0.20058861821889878,
      "num_tokens": 35234337.0,
      "step": 15785
    },
    {
      "entropy": 5.535239791870117,
      "epoch": 1.7734037176391306,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.1432,
      "mean_token_accuracy": 0.2034648522734642,
      "num_tokens": 35245736.0,
      "step": 15790
    },
    {
      "entropy": 5.516910982131958,
      "epoch": 1.773965294547088,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.0893,
      "mean_token_accuracy": 0.2156677946448326,
      "num_tokens": 35256368.0,
      "step": 15795
    },
    {
      "entropy": 5.534982538223266,
      "epoch": 1.7745268714550457,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.1601,
      "mean_token_accuracy": 0.20812107026576995,
      "num_tokens": 35267238.0,
      "step": 15800
    },
    {
      "entropy": 5.537866258621216,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.1116,
      "mean_token_accuracy": 0.2074248507618904,
      "num_tokens": 35277113.0,
      "step": 15805
    },
    {
      "entropy": 5.549642181396484,
      "epoch": 1.7756500252709608,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.052,
      "mean_token_accuracy": 0.2052018091082573,
      "num_tokens": 35288504.0,
      "step": 15810
    },
    {
      "entropy": 5.50852403640747,
      "epoch": 1.7762116021789183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.0829,
      "mean_token_accuracy": 0.2163776308298111,
      "num_tokens": 35301215.0,
      "step": 15815
    },
    {
      "entropy": 5.472559070587158,
      "epoch": 1.776773179086876,
      "grad_norm": 1.015625,
      "learning_rate": 0.000469261500582484,
      "loss": 5.0787,
      "mean_token_accuracy": 0.20995114743709564,
      "num_tokens": 35311997.0,
      "step": 15820
    },
    {
      "entropy": 5.56088695526123,
      "epoch": 1.7773347559948336,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.1555,
      "mean_token_accuracy": 0.21026714742183686,
      "num_tokens": 35323217.0,
      "step": 15825
    },
    {
      "entropy": 5.566448926925659,
      "epoch": 1.777896332902791,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.1892,
      "mean_token_accuracy": 0.20944911986589432,
      "num_tokens": 35333102.0,
      "step": 15830
    },
    {
      "entropy": 5.616394519805908,
      "epoch": 1.7784579098107485,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.1942,
      "mean_token_accuracy": 0.2021913304924965,
      "num_tokens": 35346394.0,
      "step": 15835
    },
    {
      "entropy": 5.456839323043823,
      "epoch": 1.7790194867187061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004691804235037845,
      "loss": 5.0202,
      "mean_token_accuracy": 0.2155841052532196,
      "num_tokens": 35357827.0,
      "step": 15840
    },
    {
      "entropy": 5.521256780624389,
      "epoch": 1.7795810636266638,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.1385,
      "mean_token_accuracy": 0.20600610822439194,
      "num_tokens": 35370517.0,
      "step": 15845
    },
    {
      "entropy": 5.554784870147705,
      "epoch": 1.7801426405346212,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.1636,
      "mean_token_accuracy": 0.20827459990978242,
      "num_tokens": 35381830.0,
      "step": 15850
    },
    {
      "entropy": 5.623511266708374,
      "epoch": 1.7807042174425787,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.2006,
      "mean_token_accuracy": 0.1995435655117035,
      "num_tokens": 35393161.0,
      "step": 15855
    },
    {
      "entropy": 5.586424827575684,
      "epoch": 1.7812657943505363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.0992,
      "mean_token_accuracy": 0.21456411182880403,
      "num_tokens": 35403338.0,
      "step": 15860
    },
    {
      "entropy": 5.554392099380493,
      "epoch": 1.781827371258494,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.197,
      "mean_token_accuracy": 0.20029408484697342,
      "num_tokens": 35415862.0,
      "step": 15865
    },
    {
      "entropy": 5.490436458587647,
      "epoch": 1.7823889481664514,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20466220378875732,
      "num_tokens": 35426696.0,
      "step": 15870
    },
    {
      "entropy": 5.537225914001465,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.1173,
      "mean_token_accuracy": 0.21246024668216706,
      "num_tokens": 35438624.0,
      "step": 15875
    },
    {
      "entropy": 5.596300315856934,
      "epoch": 1.7835121019823665,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.2673,
      "mean_token_accuracy": 0.20027589052915573,
      "num_tokens": 35449841.0,
      "step": 15880
    },
    {
      "entropy": 5.731958961486816,
      "epoch": 1.784073678890324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.3478,
      "mean_token_accuracy": 0.19502552449703217,
      "num_tokens": 35462081.0,
      "step": 15885
    },
    {
      "entropy": 5.480266523361206,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.0,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.0554,
      "mean_token_accuracy": 0.21102778017520904,
      "num_tokens": 35472834.0,
      "step": 15890
    },
    {
      "entropy": 5.477356433868408,
      "epoch": 1.785196832706239,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21321683824062349,
      "num_tokens": 35482975.0,
      "step": 15895
    },
    {
      "entropy": 5.467747402191162,
      "epoch": 1.7857584096141967,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20469704419374465,
      "num_tokens": 35494470.0,
      "step": 15900
    },
    {
      "entropy": 5.505661201477051,
      "epoch": 1.7863199865221542,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.063,
      "mean_token_accuracy": 0.21006495058536528,
      "num_tokens": 35505464.0,
      "step": 15905
    },
    {
      "entropy": 5.50462965965271,
      "epoch": 1.7868815634301116,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20181670933961868,
      "num_tokens": 35517609.0,
      "step": 15910
    },
    {
      "entropy": 5.527349519729614,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21299619376659393,
      "num_tokens": 35528994.0,
      "step": 15915
    },
    {
      "entropy": 5.603882694244385,
      "epoch": 1.788004717246027,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.2224,
      "mean_token_accuracy": 0.20156983882188798,
      "num_tokens": 35541044.0,
      "step": 15920
    },
    {
      "entropy": 5.521676635742187,
      "epoch": 1.7885662941539844,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21300457268953324,
      "num_tokens": 35552660.0,
      "step": 15925
    },
    {
      "entropy": 5.609447240829468,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.1536,
      "mean_token_accuracy": 0.2033581703901291,
      "num_tokens": 35565144.0,
      "step": 15930
    },
    {
      "entropy": 5.5493556499481205,
      "epoch": 1.7896894479698995,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.0968,
      "mean_token_accuracy": 0.20968289971351622,
      "num_tokens": 35577067.0,
      "step": 15935
    },
    {
      "entropy": 5.466688203811645,
      "epoch": 1.7902510248778571,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.0665,
      "mean_token_accuracy": 0.21116508096456527,
      "num_tokens": 35588712.0,
      "step": 15940
    },
    {
      "entropy": 5.553652334213257,
      "epoch": 1.7908126017858146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21427206248044967,
      "num_tokens": 35599212.0,
      "step": 15945
    },
    {
      "entropy": 5.551005172729492,
      "epoch": 1.791374178693772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.1201,
      "mean_token_accuracy": 0.21352460086345673,
      "num_tokens": 35609397.0,
      "step": 15950
    },
    {
      "entropy": 5.5086719512939455,
      "epoch": 1.7919357556017297,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.0792,
      "mean_token_accuracy": 0.2070300504565239,
      "num_tokens": 35620615.0,
      "step": 15955
    },
    {
      "entropy": 5.559685134887696,
      "epoch": 1.7924973325096873,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.129,
      "mean_token_accuracy": 0.2079283818602562,
      "num_tokens": 35632684.0,
      "step": 15960
    },
    {
      "entropy": 5.542967081069946,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.046875,
      "learning_rate": 0.000468671452146028,
      "loss": 5.139,
      "mean_token_accuracy": 0.21152493059635163,
      "num_tokens": 35643896.0,
      "step": 15965
    },
    {
      "entropy": 5.53941764831543,
      "epoch": 1.7936204863256022,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2104354903101921,
      "num_tokens": 35655752.0,
      "step": 15970
    },
    {
      "entropy": 5.458705902099609,
      "epoch": 1.7941820632335599,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000468630567805468,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20734648406505585,
      "num_tokens": 35665870.0,
      "step": 15975
    },
    {
      "entropy": 5.515145015716553,
      "epoch": 1.7947436401415173,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000468610116386954,
      "loss": 5.1837,
      "mean_token_accuracy": 0.2046958848834038,
      "num_tokens": 35677792.0,
      "step": 15980
    },
    {
      "entropy": 5.497518825531006,
      "epoch": 1.7953052170494748,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20749557316303252,
      "num_tokens": 35689059.0,
      "step": 15985
    },
    {
      "entropy": 5.560515403747559,
      "epoch": 1.7958667939574324,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.1814,
      "mean_token_accuracy": 0.2046581283211708,
      "num_tokens": 35699953.0,
      "step": 15990
    },
    {
      "entropy": 5.5813439846038815,
      "epoch": 1.79642837086539,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20730597376823426,
      "num_tokens": 35710894.0,
      "step": 15995
    },
    {
      "entropy": 5.532754898071289,
      "epoch": 1.7969899477733475,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20677562057971954,
      "num_tokens": 35722440.0,
      "step": 16000
    },
    {
      "entropy": 5.4981255531311035,
      "epoch": 1.797551524681305,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.1648,
      "mean_token_accuracy": 0.20817843675613404,
      "num_tokens": 35733768.0,
      "step": 16005
    },
    {
      "entropy": 5.5243692874908445,
      "epoch": 1.7981131015892626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20820131450891494,
      "num_tokens": 35744211.0,
      "step": 16010
    },
    {
      "entropy": 5.559757947921753,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.1214,
      "mean_token_accuracy": 0.21088675856590272,
      "num_tokens": 35755498.0,
      "step": 16015
    },
    {
      "entropy": 5.5569947242736815,
      "epoch": 1.7992362554051777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.1419,
      "mean_token_accuracy": 0.21241978257894517,
      "num_tokens": 35767442.0,
      "step": 16020
    },
    {
      "entropy": 5.504355621337891,
      "epoch": 1.7997978323131352,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.0999,
      "mean_token_accuracy": 0.211040098965168,
      "num_tokens": 35778040.0,
      "step": 16025
    },
    {
      "entropy": 5.509951162338257,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.059,
      "mean_token_accuracy": 0.21813765317201614,
      "num_tokens": 35788139.0,
      "step": 16030
    },
    {
      "entropy": 5.510417079925537,
      "epoch": 1.8009209861290505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20653367191553115,
      "num_tokens": 35799077.0,
      "step": 16035
    },
    {
      "entropy": 5.564406156539917,
      "epoch": 1.801482563037008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20847166180610657,
      "num_tokens": 35809748.0,
      "step": 16040
    },
    {
      "entropy": 5.652004384994507,
      "epoch": 1.8020441399449654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.1567,
      "mean_token_accuracy": 0.21287556737661362,
      "num_tokens": 35821621.0,
      "step": 16045
    },
    {
      "entropy": 5.539083480834961,
      "epoch": 1.802605716852923,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.1217,
      "mean_token_accuracy": 0.21265282332897187,
      "num_tokens": 35833497.0,
      "step": 16050
    },
    {
      "entropy": 5.611848449707031,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.2328,
      "mean_token_accuracy": 0.20935529917478563,
      "num_tokens": 35844266.0,
      "step": 16055
    },
    {
      "entropy": 5.639096450805664,
      "epoch": 1.8037288706688381,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.2211,
      "mean_token_accuracy": 0.2044106811285019,
      "num_tokens": 35855260.0,
      "step": 16060
    },
    {
      "entropy": 5.567238616943359,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.1652,
      "mean_token_accuracy": 0.21188524812459947,
      "num_tokens": 35865950.0,
      "step": 16065
    },
    {
      "entropy": 5.583362865447998,
      "epoch": 1.8048520244847532,
      "grad_norm": 1.0,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.2949,
      "mean_token_accuracy": 0.19802670925855637,
      "num_tokens": 35877028.0,
      "step": 16070
    },
    {
      "entropy": 5.542549133300781,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.1526,
      "mean_token_accuracy": 0.2020435243844986,
      "num_tokens": 35887831.0,
      "step": 16075
    },
    {
      "entropy": 5.5656005382537845,
      "epoch": 1.8059751783006681,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.1321,
      "mean_token_accuracy": 0.21055003702640535,
      "num_tokens": 35897946.0,
      "step": 16080
    },
    {
      "entropy": 5.62137360572815,
      "epoch": 1.8065367552086258,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.2643,
      "mean_token_accuracy": 0.20307594090700148,
      "num_tokens": 35910973.0,
      "step": 16085
    },
    {
      "entropy": 5.581250381469727,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.1792,
      "mean_token_accuracy": 0.20447730273008347,
      "num_tokens": 35921478.0,
      "step": 16090
    },
    {
      "entropy": 5.539097738265991,
      "epoch": 1.8076599090245409,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.0614,
      "mean_token_accuracy": 0.21530839055776596,
      "num_tokens": 35932211.0,
      "step": 16095
    },
    {
      "entropy": 5.507720470428467,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.2217,
      "mean_token_accuracy": 0.20191744416952134,
      "num_tokens": 35944350.0,
      "step": 16100
    },
    {
      "entropy": 5.6125000476837155,
      "epoch": 1.808783062840456,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.2149,
      "mean_token_accuracy": 0.2026865601539612,
      "num_tokens": 35955590.0,
      "step": 16105
    },
    {
      "entropy": 5.4943009376525875,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.0,
      "learning_rate": 0.0004680762176322701,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22385379374027253,
      "num_tokens": 35965872.0,
      "step": 16110
    },
    {
      "entropy": 5.466759443283081,
      "epoch": 1.809906216656371,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20214666426181793,
      "num_tokens": 35976851.0,
      "step": 16115
    },
    {
      "entropy": 5.554028844833374,
      "epoch": 1.8104677935643285,
      "grad_norm": 1.0,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20810256749391556,
      "num_tokens": 35986984.0,
      "step": 16120
    },
    {
      "entropy": 5.587913084030151,
      "epoch": 1.8110293704722862,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20628853738307953,
      "num_tokens": 35999349.0,
      "step": 16125
    },
    {
      "entropy": 5.501755142211914,
      "epoch": 1.8115909473802438,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.1617,
      "mean_token_accuracy": 0.2119600683450699,
      "num_tokens": 36011122.0,
      "step": 16130
    },
    {
      "entropy": 5.5628783226013185,
      "epoch": 1.8121525242882013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046797306799879,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2079499363899231,
      "num_tokens": 36022729.0,
      "step": 16135
    },
    {
      "entropy": 5.522836065292358,
      "epoch": 1.8127141011961587,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20629012882709502,
      "num_tokens": 36034042.0,
      "step": 16140
    },
    {
      "entropy": 5.522409725189209,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.1398,
      "mean_token_accuracy": 0.2059260204434395,
      "num_tokens": 36045357.0,
      "step": 16145
    },
    {
      "entropy": 5.4188957691192625,
      "epoch": 1.813837255012074,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004679111044813987,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21463740170001983,
      "num_tokens": 36055180.0,
      "step": 16150
    },
    {
      "entropy": 5.476156997680664,
      "epoch": 1.8143988319200315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.042,
      "mean_token_accuracy": 0.21280794739723205,
      "num_tokens": 36065574.0,
      "step": 16155
    },
    {
      "entropy": 5.509959936141968,
      "epoch": 1.814960408827989,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.1351,
      "mean_token_accuracy": 0.21192506402730943,
      "num_tokens": 36077316.0,
      "step": 16160
    },
    {
      "entropy": 5.506747245788574,
      "epoch": 1.8155219857359466,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20973020792007446,
      "num_tokens": 36088735.0,
      "step": 16165
    },
    {
      "entropy": 5.557045364379883,
      "epoch": 1.8160835626439042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467828400468182,
      "loss": 5.1261,
      "mean_token_accuracy": 0.20709376484155656,
      "num_tokens": 36098806.0,
      "step": 16170
    },
    {
      "entropy": 5.50244026184082,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.1831,
      "mean_token_accuracy": 0.20532937496900558,
      "num_tokens": 36110333.0,
      "step": 16175
    },
    {
      "entropy": 5.565231609344482,
      "epoch": 1.8172067164598191,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20878666639328003,
      "num_tokens": 36121719.0,
      "step": 16180
    },
    {
      "entropy": 5.5523268699646,
      "epoch": 1.8177682933677768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.0311,
      "mean_token_accuracy": 0.21674093902111052,
      "num_tokens": 36132129.0,
      "step": 16185
    },
    {
      "entropy": 5.613173151016236,
      "epoch": 1.8183298702757342,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.3139,
      "mean_token_accuracy": 0.20399825870990754,
      "num_tokens": 36145189.0,
      "step": 16190
    },
    {
      "entropy": 5.5120491027832035,
      "epoch": 1.8188914471836917,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20740067213773727,
      "num_tokens": 36157896.0,
      "step": 16195
    },
    {
      "entropy": 5.566295146942139,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.1764,
      "mean_token_accuracy": 0.20699597746133805,
      "num_tokens": 36168842.0,
      "step": 16200
    },
    {
      "entropy": 5.468304443359375,
      "epoch": 1.820014600999607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2147214114665985,
      "num_tokens": 36180075.0,
      "step": 16205
    },
    {
      "entropy": 5.48673644065857,
      "epoch": 1.8205761779075644,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.0701,
      "mean_token_accuracy": 0.21327518224716185,
      "num_tokens": 36190589.0,
      "step": 16210
    },
    {
      "entropy": 5.460137271881104,
      "epoch": 1.8211377548155219,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046764195732846013,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21655097752809524,
      "num_tokens": 36201145.0,
      "step": 16215
    },
    {
      "entropy": 5.459126663208008,
      "epoch": 1.8216993317234795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.095,
      "mean_token_accuracy": 0.2121373802423477,
      "num_tokens": 36213243.0,
      "step": 16220
    },
    {
      "entropy": 5.43475832939148,
      "epoch": 1.8222609086314372,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21341248452663422,
      "num_tokens": 36224491.0,
      "step": 16225
    },
    {
      "entropy": 5.48335223197937,
      "epoch": 1.8228224855393946,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.1056,
      "mean_token_accuracy": 0.21120526939630507,
      "num_tokens": 36236006.0,
      "step": 16230
    },
    {
      "entropy": 5.636699581146241,
      "epoch": 1.823384062447352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046755893420355,
      "loss": 5.2072,
      "mean_token_accuracy": 0.20229344815015793,
      "num_tokens": 36246925.0,
      "step": 16235
    },
    {
      "entropy": 5.547803640365601,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20051463693380356,
      "num_tokens": 36258449.0,
      "step": 16240
    },
    {
      "entropy": 5.538269901275635,
      "epoch": 1.8245072162632674,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.1229,
      "mean_token_accuracy": 0.2100173130631447,
      "num_tokens": 36270820.0,
      "step": 16245
    },
    {
      "entropy": 5.598525142669677,
      "epoch": 1.8250687931712248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.187,
      "mean_token_accuracy": 0.20589418560266495,
      "num_tokens": 36282346.0,
      "step": 16250
    },
    {
      "entropy": 5.58179497718811,
      "epoch": 1.8256303700791823,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.2037,
      "mean_token_accuracy": 0.20074910670518875,
      "num_tokens": 36294111.0,
      "step": 16255
    },
    {
      "entropy": 5.52326979637146,
      "epoch": 1.82619194698714,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.0965,
      "mean_token_accuracy": 0.2069130524992943,
      "num_tokens": 36304899.0,
      "step": 16260
    },
    {
      "entropy": 5.578506851196289,
      "epoch": 1.8267535238950976,
      "grad_norm": 1.0,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.2108,
      "mean_token_accuracy": 0.20362971276044844,
      "num_tokens": 36316272.0,
      "step": 16265
    },
    {
      "entropy": 5.52227668762207,
      "epoch": 1.8273151008030548,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.0137,
      "mean_token_accuracy": 0.21745406687259675,
      "num_tokens": 36326624.0,
      "step": 16270
    },
    {
      "entropy": 5.463012933731079,
      "epoch": 1.8278766777110125,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21933161467313766,
      "num_tokens": 36338178.0,
      "step": 16275
    },
    {
      "entropy": 5.5840229988098145,
      "epoch": 1.8284382546189701,
      "grad_norm": 1.0,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21008913218975067,
      "num_tokens": 36349993.0,
      "step": 16280
    },
    {
      "entropy": 5.60274109840393,
      "epoch": 1.8289998315269276,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.1864,
      "mean_token_accuracy": 0.2073033258318901,
      "num_tokens": 36361002.0,
      "step": 16285
    },
    {
      "entropy": 5.553527688980102,
      "epoch": 1.829561408434885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.1341,
      "mean_token_accuracy": 0.21003983467817305,
      "num_tokens": 36371688.0,
      "step": 16290
    },
    {
      "entropy": 5.484617471694946,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.0811,
      "mean_token_accuracy": 0.21331827491521835,
      "num_tokens": 36382683.0,
      "step": 16295
    },
    {
      "entropy": 5.586533069610596,
      "epoch": 1.8306845622508003,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.2222,
      "mean_token_accuracy": 0.2004428833723068,
      "num_tokens": 36394170.0,
      "step": 16300
    },
    {
      "entropy": 5.605193567276001,
      "epoch": 1.8312461391587578,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20278556495904923,
      "num_tokens": 36405613.0,
      "step": 16305
    },
    {
      "entropy": 5.587423086166382,
      "epoch": 1.8318077160667152,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.2056,
      "mean_token_accuracy": 0.20887486785650253,
      "num_tokens": 36416282.0,
      "step": 16310
    },
    {
      "entropy": 5.573940086364746,
      "epoch": 1.8323692929746729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19974818974733352,
      "num_tokens": 36427624.0,
      "step": 16315
    },
    {
      "entropy": 5.517344522476196,
      "epoch": 1.8329308698826305,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2231230616569519,
      "num_tokens": 36438230.0,
      "step": 16320
    },
    {
      "entropy": 5.526496505737304,
      "epoch": 1.833492446790588,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.153,
      "mean_token_accuracy": 0.20782502144575118,
      "num_tokens": 36448854.0,
      "step": 16325
    },
    {
      "entropy": 5.56245265007019,
      "epoch": 1.8340540236985454,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20650095641613006,
      "num_tokens": 36460426.0,
      "step": 16330
    },
    {
      "entropy": 5.598569107055664,
      "epoch": 1.834615600606503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20883156806230546,
      "num_tokens": 36470914.0,
      "step": 16335
    },
    {
      "entropy": 5.516495227813721,
      "epoch": 1.8351771775144607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.0582,
      "mean_token_accuracy": 0.2117043823003769,
      "num_tokens": 36480753.0,
      "step": 16340
    },
    {
      "entropy": 5.4965132713317875,
      "epoch": 1.8357387544224182,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21171022355556487,
      "num_tokens": 36492858.0,
      "step": 16345
    },
    {
      "entropy": 5.4419841289520265,
      "epoch": 1.8363003313303756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046707964881554005,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21847275346517564,
      "num_tokens": 36503912.0,
      "step": 16350
    },
    {
      "entropy": 5.473934364318848,
      "epoch": 1.8368619082383333,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.0996,
      "mean_token_accuracy": 0.20875011682510375,
      "num_tokens": 36515410.0,
      "step": 16355
    },
    {
      "entropy": 5.520986557006836,
      "epoch": 1.837423485146291,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21521356999874114,
      "num_tokens": 36526310.0,
      "step": 16360
    },
    {
      "entropy": 5.497657775878906,
      "epoch": 1.8379850620542482,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046701689463761947,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2118170902132988,
      "num_tokens": 36537081.0,
      "step": 16365
    },
    {
      "entropy": 5.510953283309936,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21122519224882125,
      "num_tokens": 36549419.0,
      "step": 16370
    },
    {
      "entropy": 5.444672918319702,
      "epoch": 1.8391082158701635,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.0781,
      "mean_token_accuracy": 0.21695982813835143,
      "num_tokens": 36560842.0,
      "step": 16375
    },
    {
      "entropy": 5.504959678649902,
      "epoch": 1.839669792778121,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21273245066404342,
      "num_tokens": 36571717.0,
      "step": 16380
    },
    {
      "entropy": 5.557960367202758,
      "epoch": 1.8402313696860784,
      "grad_norm": 1.0,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.1,
      "mean_token_accuracy": 0.20946543514728547,
      "num_tokens": 36581709.0,
      "step": 16385
    },
    {
      "entropy": 5.5071851253509525,
      "epoch": 1.840792946594036,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.1108,
      "mean_token_accuracy": 0.20555109679698944,
      "num_tokens": 36593033.0,
      "step": 16390
    },
    {
      "entropy": 5.55257658958435,
      "epoch": 1.8413545235019937,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.1798,
      "mean_token_accuracy": 0.208029168844223,
      "num_tokens": 36603605.0,
      "step": 16395
    },
    {
      "entropy": 5.575206518173218,
      "epoch": 1.8419161004099511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20561684519052506,
      "num_tokens": 36616219.0,
      "step": 16400
    },
    {
      "entropy": 5.621106576919556,
      "epoch": 1.8424776773179086,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2116603285074234,
      "num_tokens": 36627777.0,
      "step": 16405
    },
    {
      "entropy": 5.547340726852417,
      "epoch": 1.8430392542258662,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.1074,
      "mean_token_accuracy": 0.2088935434818268,
      "num_tokens": 36638792.0,
      "step": 16410
    },
    {
      "entropy": 5.532141637802124,
      "epoch": 1.843600831133824,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.1309,
      "mean_token_accuracy": 0.2126806080341339,
      "num_tokens": 36650218.0,
      "step": 16415
    },
    {
      "entropy": 5.559504842758178,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21041698157787322,
      "num_tokens": 36660987.0,
      "step": 16420
    },
    {
      "entropy": 5.52473931312561,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20393646210432054,
      "num_tokens": 36671270.0,
      "step": 16425
    },
    {
      "entropy": 5.501770830154419,
      "epoch": 1.8452855618576964,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.086,
      "mean_token_accuracy": 0.2147757664322853,
      "num_tokens": 36682484.0,
      "step": 16430
    },
    {
      "entropy": 5.56200065612793,
      "epoch": 1.845847138765654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.153,
      "mean_token_accuracy": 0.21047349274158478,
      "num_tokens": 36693723.0,
      "step": 16435
    },
    {
      "entropy": 5.465136098861694,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004667022987327732,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21866004914045334,
      "num_tokens": 36704166.0,
      "step": 16440
    },
    {
      "entropy": 5.472092390060425,
      "epoch": 1.846970292581569,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21416597068309784,
      "num_tokens": 36714139.0,
      "step": 16445
    },
    {
      "entropy": 5.5302447319030765,
      "epoch": 1.8475318694895266,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.0786,
      "mean_token_accuracy": 0.214362969994545,
      "num_tokens": 36724273.0,
      "step": 16450
    },
    {
      "entropy": 5.53122615814209,
      "epoch": 1.8480934463974843,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.1433,
      "mean_token_accuracy": 0.2062819182872772,
      "num_tokens": 36735176.0,
      "step": 16455
    },
    {
      "entropy": 5.478092050552368,
      "epoch": 1.8486550233054415,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2115882322192192,
      "num_tokens": 36746170.0,
      "step": 16460
    },
    {
      "entropy": 5.6308471202850345,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.2332,
      "mean_token_accuracy": 0.2036072388291359,
      "num_tokens": 36756969.0,
      "step": 16465
    },
    {
      "entropy": 5.612758588790894,
      "epoch": 1.8497781771213568,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.0876,
      "mean_token_accuracy": 0.21479337960481643,
      "num_tokens": 36767892.0,
      "step": 16470
    },
    {
      "entropy": 5.513358879089355,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20914121419191362,
      "num_tokens": 36778967.0,
      "step": 16475
    },
    {
      "entropy": 5.421139287948608,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004665339525502236,
      "loss": 4.9763,
      "mean_token_accuracy": 0.20990423411130904,
      "num_tokens": 36789204.0,
      "step": 16480
    },
    {
      "entropy": 5.520025110244751,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21905306875705718,
      "num_tokens": 36798990.0,
      "step": 16485
    },
    {
      "entropy": 5.583359050750732,
      "epoch": 1.852024484753187,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.1357,
      "mean_token_accuracy": 0.2104924276471138,
      "num_tokens": 36809988.0,
      "step": 16490
    },
    {
      "entropy": 5.503124380111695,
      "epoch": 1.8525860616611445,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2169835925102234,
      "num_tokens": 36821579.0,
      "step": 16495
    },
    {
      "entropy": 5.60382194519043,
      "epoch": 1.853147638569102,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20810676515102386,
      "num_tokens": 36833141.0,
      "step": 16500
    },
    {
      "entropy": 5.601029682159424,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.1821,
      "mean_token_accuracy": 0.2118373140692711,
      "num_tokens": 36843347.0,
      "step": 16505
    },
    {
      "entropy": 5.5648987770080565,
      "epoch": 1.8542707923850172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.2426,
      "mean_token_accuracy": 0.1956004396080971,
      "num_tokens": 36854459.0,
      "step": 16510
    },
    {
      "entropy": 5.5041258335113525,
      "epoch": 1.8548323692929747,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004663863294567462,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21323901712894439,
      "num_tokens": 36864761.0,
      "step": 16515
    },
    {
      "entropy": 5.651431322097778,
      "epoch": 1.8553939462009321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.242,
      "mean_token_accuracy": 0.2011815682053566,
      "num_tokens": 36875413.0,
      "step": 16520
    },
    {
      "entropy": 5.527823972702026,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.1273,
      "mean_token_accuracy": 0.2089690700173378,
      "num_tokens": 36885350.0,
      "step": 16525
    },
    {
      "entropy": 5.492122936248779,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2142224907875061,
      "num_tokens": 36895355.0,
      "step": 16530
    },
    {
      "entropy": 5.567118692398071,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.1283,
      "mean_token_accuracy": 0.21483739614486694,
      "num_tokens": 36907454.0,
      "step": 16535
    },
    {
      "entropy": 5.549027681350708,
      "epoch": 1.8576402538327623,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.1241,
      "mean_token_accuracy": 0.2042805016040802,
      "num_tokens": 36919199.0,
      "step": 16540
    },
    {
      "entropy": 5.633251762390136,
      "epoch": 1.85820183074072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.2272,
      "mean_token_accuracy": 0.20432233512401582,
      "num_tokens": 36930499.0,
      "step": 16545
    },
    {
      "entropy": 5.5547199726104735,
      "epoch": 1.8587634076486776,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.1474,
      "mean_token_accuracy": 0.20221250057220458,
      "num_tokens": 36943082.0,
      "step": 16550
    },
    {
      "entropy": 5.484726762771606,
      "epoch": 1.859324984556635,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.046,
      "mean_token_accuracy": 0.21432142108678817,
      "num_tokens": 36954591.0,
      "step": 16555
    },
    {
      "entropy": 5.523643493652344,
      "epoch": 1.8598865614645925,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.1203,
      "mean_token_accuracy": 0.21007298976182937,
      "num_tokens": 36967028.0,
      "step": 16560
    },
    {
      "entropy": 5.532343435287475,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20340125411748886,
      "num_tokens": 36978716.0,
      "step": 16565
    },
    {
      "entropy": 5.519258832931518,
      "epoch": 1.8610097152805076,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21277046650648118,
      "num_tokens": 36989416.0,
      "step": 16570
    },
    {
      "entropy": 5.5362715244293215,
      "epoch": 1.861571292188465,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21235738396644593,
      "num_tokens": 37002197.0,
      "step": 16575
    },
    {
      "entropy": 5.674366283416748,
      "epoch": 1.8621328690964227,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19783121049404145,
      "num_tokens": 37012977.0,
      "step": 16580
    },
    {
      "entropy": 5.544118213653564,
      "epoch": 1.8626944460043804,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20846157073974608,
      "num_tokens": 37024141.0,
      "step": 16585
    },
    {
      "entropy": 5.501484632492065,
      "epoch": 1.8632560229123378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.046,
      "mean_token_accuracy": 0.2153637394309044,
      "num_tokens": 37034471.0,
      "step": 16590
    },
    {
      "entropy": 5.543417358398438,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.108,
      "mean_token_accuracy": 0.21154415607452393,
      "num_tokens": 37046375.0,
      "step": 16595
    },
    {
      "entropy": 5.5620263576507565,
      "epoch": 1.864379176728253,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.125,
      "mean_token_accuracy": 0.21386207193136214,
      "num_tokens": 37056617.0,
      "step": 16600
    },
    {
      "entropy": 5.435905981063843,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.046875,
      "learning_rate": 0.000466005356663438,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2201532617211342,
      "num_tokens": 37066477.0,
      "step": 16605
    },
    {
      "entropy": 5.481939888000488,
      "epoch": 1.865502330544168,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.1497,
      "mean_token_accuracy": 0.20296623259782792,
      "num_tokens": 37077150.0,
      "step": 16610
    },
    {
      "entropy": 5.587792253494262,
      "epoch": 1.8660639074521255,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20729612708091735,
      "num_tokens": 37089961.0,
      "step": 16615
    },
    {
      "entropy": 5.635315370559693,
      "epoch": 1.8666254843600831,
      "grad_norm": 0.984375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.1851,
      "mean_token_accuracy": 0.2058102682232857,
      "num_tokens": 37100779.0,
      "step": 16620
    },
    {
      "entropy": 5.509862089157105,
      "epoch": 1.8671870612680408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.0445,
      "mean_token_accuracy": 0.21108440011739732,
      "num_tokens": 37111293.0,
      "step": 16625
    },
    {
      "entropy": 5.533247613906861,
      "epoch": 1.8677486381759982,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.1402,
      "mean_token_accuracy": 0.21164448261260987,
      "num_tokens": 37123087.0,
      "step": 16630
    },
    {
      "entropy": 5.560720539093017,
      "epoch": 1.8683102150839557,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.1069,
      "mean_token_accuracy": 0.21125419586896896,
      "num_tokens": 37135227.0,
      "step": 16635
    },
    {
      "entropy": 5.534476327896118,
      "epoch": 1.8688717919919133,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.0926,
      "mean_token_accuracy": 0.21373541206121444,
      "num_tokens": 37146970.0,
      "step": 16640
    },
    {
      "entropy": 5.544089460372925,
      "epoch": 1.869433368899871,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.084,
      "mean_token_accuracy": 0.20534783750772476,
      "num_tokens": 37158144.0,
      "step": 16645
    },
    {
      "entropy": 5.581535768508911,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.2062,
      "mean_token_accuracy": 0.200848813354969,
      "num_tokens": 37169265.0,
      "step": 16650
    },
    {
      "entropy": 5.574963140487671,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.1607,
      "mean_token_accuracy": 0.21057904213666917,
      "num_tokens": 37179186.0,
      "step": 16655
    },
    {
      "entropy": 5.6192549705505375,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2104545757174492,
      "num_tokens": 37188906.0,
      "step": 16660
    },
    {
      "entropy": 5.493058395385742,
      "epoch": 1.871679676531701,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000465750279855493,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21996092051267624,
      "num_tokens": 37200879.0,
      "step": 16665
    },
    {
      "entropy": 5.512601280212403,
      "epoch": 1.8722412534396584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21698102802038194,
      "num_tokens": 37211648.0,
      "step": 16670
    },
    {
      "entropy": 5.629525470733642,
      "epoch": 1.872802830347616,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.2303,
      "mean_token_accuracy": 0.20782027244567872,
      "num_tokens": 37223164.0,
      "step": 16675
    },
    {
      "entropy": 5.583185005187988,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.1027,
      "mean_token_accuracy": 0.2087726667523384,
      "num_tokens": 37233860.0,
      "step": 16680
    },
    {
      "entropy": 5.602839803695678,
      "epoch": 1.8739259841635312,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.0664,
      "mean_token_accuracy": 0.21311709135770798,
      "num_tokens": 37245431.0,
      "step": 16685
    },
    {
      "entropy": 5.580342054367065,
      "epoch": 1.8744875610714886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.2053,
      "mean_token_accuracy": 0.20472179651260375,
      "num_tokens": 37256394.0,
      "step": 16690
    },
    {
      "entropy": 5.536282348632812,
      "epoch": 1.8750491379794463,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.1605,
      "mean_token_accuracy": 0.21069211959838868,
      "num_tokens": 37268367.0,
      "step": 16695
    },
    {
      "entropy": 5.546684217453003,
      "epoch": 1.875610714887404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20671092122793197,
      "num_tokens": 37279019.0,
      "step": 16700
    },
    {
      "entropy": 5.590096521377563,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20960246175527572,
      "num_tokens": 37289806.0,
      "step": 16705
    },
    {
      "entropy": 5.471644878387451,
      "epoch": 1.8767338687033188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20591232925653458,
      "num_tokens": 37300607.0,
      "step": 16710
    },
    {
      "entropy": 5.513507843017578,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.1787,
      "mean_token_accuracy": 0.20676879584789276,
      "num_tokens": 37311966.0,
      "step": 16715
    },
    {
      "entropy": 5.434850120544434,
      "epoch": 1.8778570225192341,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20851288139820098,
      "num_tokens": 37324459.0,
      "step": 16720
    },
    {
      "entropy": 5.589182996749878,
      "epoch": 1.8784185994271916,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2075342833995819,
      "num_tokens": 37335276.0,
      "step": 16725
    },
    {
      "entropy": 5.550718832015991,
      "epoch": 1.878980176335149,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21858253479003906,
      "num_tokens": 37347274.0,
      "step": 16730
    },
    {
      "entropy": 5.5434599876403805,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.0,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20366531759500503,
      "num_tokens": 37359789.0,
      "step": 16735
    },
    {
      "entropy": 5.539094543457031,
      "epoch": 1.8801033301510643,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.1072,
      "mean_token_accuracy": 0.2170163169503212,
      "num_tokens": 37369851.0,
      "step": 16740
    },
    {
      "entropy": 5.620258378982544,
      "epoch": 1.8806649070590218,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.2644,
      "mean_token_accuracy": 0.20121966004371644,
      "num_tokens": 37381008.0,
      "step": 16745
    },
    {
      "entropy": 5.516312932968139,
      "epoch": 1.8812264839669792,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20453935563564302,
      "num_tokens": 37392110.0,
      "step": 16750
    },
    {
      "entropy": 5.587967205047607,
      "epoch": 1.8817880608749369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21372598260641099,
      "num_tokens": 37402995.0,
      "step": 16755
    },
    {
      "entropy": 5.570173072814941,
      "epoch": 1.8823496377828943,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.1217,
      "mean_token_accuracy": 0.21016641110181808,
      "num_tokens": 37415072.0,
      "step": 16760
    },
    {
      "entropy": 5.625506162643433,
      "epoch": 1.8829112146908518,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19457338452339173,
      "num_tokens": 37427375.0,
      "step": 16765
    },
    {
      "entropy": 5.464407300949096,
      "epoch": 1.8834727915988094,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21248177736997603,
      "num_tokens": 37436829.0,
      "step": 16770
    },
    {
      "entropy": 5.56703462600708,
      "epoch": 1.884034368506767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19810497164726257,
      "num_tokens": 37447875.0,
      "step": 16775
    },
    {
      "entropy": 5.503093814849853,
      "epoch": 1.8845959454147245,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20960627794265746,
      "num_tokens": 37458408.0,
      "step": 16780
    },
    {
      "entropy": 5.5358555793762205,
      "epoch": 1.885157522322682,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046523750363497,
      "loss": 5.1073,
      "mean_token_accuracy": 0.2063271552324295,
      "num_tokens": 37468437.0,
      "step": 16785
    },
    {
      "entropy": 5.590099048614502,
      "epoch": 1.8857190992306396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.1817,
      "mean_token_accuracy": 0.20877105593681336,
      "num_tokens": 37479359.0,
      "step": 16790
    },
    {
      "entropy": 5.500168895721435,
      "epoch": 1.8862806761385973,
      "grad_norm": 1.0,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2118855118751526,
      "num_tokens": 37491321.0,
      "step": 16795
    },
    {
      "entropy": 5.592943382263184,
      "epoch": 1.8868422530465547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.0985,
      "mean_token_accuracy": 0.20760951936244965,
      "num_tokens": 37501828.0,
      "step": 16800
    },
    {
      "entropy": 5.66723051071167,
      "epoch": 1.8874038299545122,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.2423,
      "mean_token_accuracy": 0.2036977991461754,
      "num_tokens": 37514525.0,
      "step": 16805
    },
    {
      "entropy": 5.547118091583252,
      "epoch": 1.8879654068624698,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.045,
      "mean_token_accuracy": 0.20776873975992202,
      "num_tokens": 37525588.0,
      "step": 16810
    },
    {
      "entropy": 5.526061296463013,
      "epoch": 1.8885269837704275,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000465108764060676,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2115371897816658,
      "num_tokens": 37537941.0,
      "step": 16815
    },
    {
      "entropy": 5.553654289245605,
      "epoch": 1.889088560678385,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20563834607601167,
      "num_tokens": 37549663.0,
      "step": 16820
    },
    {
      "entropy": 5.6016092777252195,
      "epoch": 1.8896501375863424,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.18,
      "mean_token_accuracy": 0.20446603894233703,
      "num_tokens": 37559793.0,
      "step": 16825
    },
    {
      "entropy": 5.510056877136231,
      "epoch": 1.8902117144943,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.039,
      "mean_token_accuracy": 0.21429328620433807,
      "num_tokens": 37570203.0,
      "step": 16830
    },
    {
      "entropy": 5.48742561340332,
      "epoch": 1.8907732914022577,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.1346,
      "mean_token_accuracy": 0.2114340603351593,
      "num_tokens": 37581858.0,
      "step": 16835
    },
    {
      "entropy": 5.509918308258056,
      "epoch": 1.8913348683102151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.0811,
      "mean_token_accuracy": 0.2111032247543335,
      "num_tokens": 37592311.0,
      "step": 16840
    },
    {
      "entropy": 5.619149541854858,
      "epoch": 1.8918964452181726,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.2408,
      "mean_token_accuracy": 0.20205678939819335,
      "num_tokens": 37604721.0,
      "step": 16845
    },
    {
      "entropy": 5.66050705909729,
      "epoch": 1.8924580221261302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19795581698417664,
      "num_tokens": 37616044.0,
      "step": 16850
    },
    {
      "entropy": 5.5544227123260494,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.1815,
      "mean_token_accuracy": 0.21222018599510192,
      "num_tokens": 37626501.0,
      "step": 16855
    },
    {
      "entropy": 5.615000867843628,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20831548124551774,
      "num_tokens": 37637861.0,
      "step": 16860
    },
    {
      "entropy": 5.516228532791137,
      "epoch": 1.8941427528500028,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21240439414978027,
      "num_tokens": 37649288.0,
      "step": 16865
    },
    {
      "entropy": 5.47109694480896,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21746535003185272,
      "num_tokens": 37659700.0,
      "step": 16870
    },
    {
      "entropy": 5.543674516677856,
      "epoch": 1.8952659066659179,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.1714,
      "mean_token_accuracy": 0.20450260788202285,
      "num_tokens": 37671739.0,
      "step": 16875
    },
    {
      "entropy": 5.559766578674316,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2211561381816864,
      "num_tokens": 37682250.0,
      "step": 16880
    },
    {
      "entropy": 5.547030115127564,
      "epoch": 1.896389060481833,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046480752453711,
      "loss": 5.1328,
      "mean_token_accuracy": 0.21356362998485565,
      "num_tokens": 37694052.0,
      "step": 16885
    },
    {
      "entropy": 5.610237789154053,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.1773,
      "mean_token_accuracy": 0.21178375631570817,
      "num_tokens": 37704818.0,
      "step": 16890
    },
    {
      "entropy": 5.64116530418396,
      "epoch": 1.897512214297748,
      "grad_norm": 1.0,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.2352,
      "mean_token_accuracy": 0.20039330124855043,
      "num_tokens": 37717679.0,
      "step": 16895
    },
    {
      "entropy": 5.514208698272705,
      "epoch": 1.8980737912057055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20735617578029633,
      "num_tokens": 37728711.0,
      "step": 16900
    },
    {
      "entropy": 5.559884834289551,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.0926,
      "mean_token_accuracy": 0.21026391088962554,
      "num_tokens": 37740082.0,
      "step": 16905
    },
    {
      "entropy": 5.622815990447998,
      "epoch": 1.8991969450216208,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000464699651826876,
      "loss": 5.2248,
      "mean_token_accuracy": 0.20210323631763458,
      "num_tokens": 37750715.0,
      "step": 16910
    },
    {
      "entropy": 5.588087272644043,
      "epoch": 1.8997585219295783,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.1817,
      "mean_token_accuracy": 0.20266517996788025,
      "num_tokens": 37763230.0,
      "step": 16915
    },
    {
      "entropy": 5.536771965026856,
      "epoch": 1.9003200988375357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.0827,
      "mean_token_accuracy": 0.20298218131065368,
      "num_tokens": 37773621.0,
      "step": 16920
    },
    {
      "entropy": 5.581118822097778,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.1482,
      "mean_token_accuracy": 0.20363586992025376,
      "num_tokens": 37785068.0,
      "step": 16925
    },
    {
      "entropy": 5.562619924545288,
      "epoch": 1.901443252653451,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.1413,
      "mean_token_accuracy": 0.2095977246761322,
      "num_tokens": 37796925.0,
      "step": 16930
    },
    {
      "entropy": 5.538260889053345,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20840151011943817,
      "num_tokens": 37808291.0,
      "step": 16935
    },
    {
      "entropy": 5.4659546375274655,
      "epoch": 1.902566406469366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004645700052585267,
      "loss": 5.0083,
      "mean_token_accuracy": 0.22569208592176437,
      "num_tokens": 37819337.0,
      "step": 16940
    },
    {
      "entropy": 5.56496033668518,
      "epoch": 1.9031279833773236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20596638023853303,
      "num_tokens": 37831832.0,
      "step": 16945
    },
    {
      "entropy": 5.590729904174805,
      "epoch": 1.903689560285281,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.125,
      "mean_token_accuracy": 0.20754328668117522,
      "num_tokens": 37842586.0,
      "step": 16950
    },
    {
      "entropy": 5.54504017829895,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.0398,
      "mean_token_accuracy": 0.2175325408577919,
      "num_tokens": 37852198.0,
      "step": 16955
    },
    {
      "entropy": 5.53588342666626,
      "epoch": 1.9048127141011961,
      "grad_norm": 1.0,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21114291697740556,
      "num_tokens": 37862967.0,
      "step": 16960
    },
    {
      "entropy": 5.538812160491943,
      "epoch": 1.9053742910091538,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.2103,
      "mean_token_accuracy": 0.19744621217250824,
      "num_tokens": 37874501.0,
      "step": 16965
    },
    {
      "entropy": 5.533835458755493,
      "epoch": 1.9059358679171112,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004644401413948279,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21848829686641694,
      "num_tokens": 37886404.0,
      "step": 16970
    },
    {
      "entropy": 5.619067859649658,
      "epoch": 1.9064974448250687,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.1928,
      "mean_token_accuracy": 0.20142446160316468,
      "num_tokens": 37897344.0,
      "step": 16975
    },
    {
      "entropy": 5.4872698307037355,
      "epoch": 1.9070590217330263,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21592144072055816,
      "num_tokens": 37909226.0,
      "step": 16980
    },
    {
      "entropy": 5.579238414764404,
      "epoch": 1.907620598640984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20900387316942215,
      "num_tokens": 37920578.0,
      "step": 16985
    },
    {
      "entropy": 5.614823198318481,
      "epoch": 1.9081821755489414,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.0605,
      "mean_token_accuracy": 0.20762941241264343,
      "num_tokens": 37931814.0,
      "step": 16990
    },
    {
      "entropy": 5.515163326263428,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.098,
      "mean_token_accuracy": 0.2086511880159378,
      "num_tokens": 37942345.0,
      "step": 16995
    },
    {
      "entropy": 5.489299249649048,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046431006038463653,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2160610646009445,
      "num_tokens": 37953027.0,
      "step": 17000
    },
    {
      "entropy": 5.57960991859436,
      "epoch": 1.9098669062728142,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20352329611778258,
      "num_tokens": 37963290.0,
      "step": 17005
    },
    {
      "entropy": 5.586730003356934,
      "epoch": 1.9104284831807716,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1982607811689377,
      "num_tokens": 37974424.0,
      "step": 17010
    },
    {
      "entropy": 5.6751312732696535,
      "epoch": 1.910990060088729,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.2592,
      "mean_token_accuracy": 0.20096279978752135,
      "num_tokens": 37987993.0,
      "step": 17015
    },
    {
      "entropy": 5.549593305587768,
      "epoch": 1.9115516369966867,
      "grad_norm": 1.0,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21128463596105576,
      "num_tokens": 37999340.0,
      "step": 17020
    },
    {
      "entropy": 5.525022220611572,
      "epoch": 1.9121132139046444,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20240429788827896,
      "num_tokens": 38009737.0,
      "step": 17025
    },
    {
      "entropy": 5.550814580917359,
      "epoch": 1.9126747908126018,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21402548402547836,
      "num_tokens": 38021283.0,
      "step": 17030
    },
    {
      "entropy": 5.657793283462524,
      "epoch": 1.9132363677205593,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20219065248966217,
      "num_tokens": 38034044.0,
      "step": 17035
    },
    {
      "entropy": 5.524997901916504,
      "epoch": 1.913797944628517,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2120611697435379,
      "num_tokens": 38044585.0,
      "step": 17040
    },
    {
      "entropy": 5.497483158111573,
      "epoch": 1.9143595215364744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21264301240444183,
      "num_tokens": 38055771.0,
      "step": 17045
    },
    {
      "entropy": 5.499868440628052,
      "epoch": 1.9149210984444318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.1165,
      "mean_token_accuracy": 0.20274513363838195,
      "num_tokens": 38066344.0,
      "step": 17050
    },
    {
      "entropy": 5.563672637939453,
      "epoch": 1.9154826753523895,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21734229177236558,
      "num_tokens": 38077785.0,
      "step": 17055
    },
    {
      "entropy": 5.5184094429016115,
      "epoch": 1.9160442522603471,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.1577,
      "mean_token_accuracy": 0.20551990270614623,
      "num_tokens": 38089023.0,
      "step": 17060
    },
    {
      "entropy": 5.529807615280151,
      "epoch": 1.9166058291683046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046402747397356666,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21741097569465637,
      "num_tokens": 38099428.0,
      "step": 17065
    },
    {
      "entropy": 5.462765312194824,
      "epoch": 1.917167406076262,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21186559200286864,
      "num_tokens": 38110757.0,
      "step": 17070
    },
    {
      "entropy": 5.586038589477539,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.1149,
      "mean_token_accuracy": 0.20783121585845948,
      "num_tokens": 38121442.0,
      "step": 17075
    },
    {
      "entropy": 5.556519269943237,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.1032,
      "mean_token_accuracy": 0.21056283563375472,
      "num_tokens": 38132917.0,
      "step": 17080
    },
    {
      "entropy": 5.464198875427246,
      "epoch": 1.9188521368001348,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21216889172792436,
      "num_tokens": 38143188.0,
      "step": 17085
    },
    {
      "entropy": 5.636513710021973,
      "epoch": 1.9194137137080922,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.1693,
      "mean_token_accuracy": 0.21167919784784317,
      "num_tokens": 38154300.0,
      "step": 17090
    },
    {
      "entropy": 5.575698375701904,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20696183741092683,
      "num_tokens": 38165697.0,
      "step": 17095
    },
    {
      "entropy": 5.547935104370117,
      "epoch": 1.9205368675240075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.0918,
      "mean_token_accuracy": 0.2090992197394371,
      "num_tokens": 38176601.0,
      "step": 17100
    },
    {
      "entropy": 5.609787178039551,
      "epoch": 1.921098444431965,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20900874435901642,
      "num_tokens": 38188334.0,
      "step": 17105
    },
    {
      "entropy": 5.624440145492554,
      "epoch": 1.9216600213399224,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.141,
      "mean_token_accuracy": 0.20782422572374343,
      "num_tokens": 38199483.0,
      "step": 17110
    },
    {
      "entropy": 5.5819614887237545,
      "epoch": 1.92222159824788,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.1842,
      "mean_token_accuracy": 0.20421977192163468,
      "num_tokens": 38211391.0,
      "step": 17115
    },
    {
      "entropy": 5.547573566436768,
      "epoch": 1.9227831751558377,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2102360889315605,
      "num_tokens": 38223415.0,
      "step": 17120
    },
    {
      "entropy": 5.484861421585083,
      "epoch": 1.9233447520637952,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21555303335189818,
      "num_tokens": 38235093.0,
      "step": 17125
    },
    {
      "entropy": 5.516800308227539,
      "epoch": 1.9239063289717526,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.1893,
      "mean_token_accuracy": 0.20380346179008485,
      "num_tokens": 38247192.0,
      "step": 17130
    },
    {
      "entropy": 5.615087461471558,
      "epoch": 1.9244679058797103,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.1333,
      "mean_token_accuracy": 0.2048703536391258,
      "num_tokens": 38257965.0,
      "step": 17135
    },
    {
      "entropy": 5.617480659484864,
      "epoch": 1.9250294827876677,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000463700148899063,
      "loss": 5.2426,
      "mean_token_accuracy": 0.20490214824676514,
      "num_tokens": 38268529.0,
      "step": 17140
    },
    {
      "entropy": 5.537068843841553,
      "epoch": 1.9255910596956252,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.1135,
      "mean_token_accuracy": 0.20522623211145402,
      "num_tokens": 38279480.0,
      "step": 17145
    },
    {
      "entropy": 5.622207164764404,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.1842,
      "mean_token_accuracy": 0.20717252492904664,
      "num_tokens": 38291670.0,
      "step": 17150
    },
    {
      "entropy": 5.488466262817383,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046363452158532857,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22514206767082215,
      "num_tokens": 38303073.0,
      "step": 17155
    },
    {
      "entropy": 5.580797100067139,
      "epoch": 1.927275790419498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.1138,
      "mean_token_accuracy": 0.2030535876750946,
      "num_tokens": 38313915.0,
      "step": 17160
    },
    {
      "entropy": 5.51703896522522,
      "epoch": 1.9278373673274554,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20307507812976838,
      "num_tokens": 38324845.0,
      "step": 17165
    },
    {
      "entropy": 5.525485181808472,
      "epoch": 1.928398944235413,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.093,
      "mean_token_accuracy": 0.21484927088022232,
      "num_tokens": 38335096.0,
      "step": 17170
    },
    {
      "entropy": 5.5696759700775145,
      "epoch": 1.9289605211433707,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.0859,
      "mean_token_accuracy": 0.20981388092041015,
      "num_tokens": 38345559.0,
      "step": 17175
    },
    {
      "entropy": 5.494486093521118,
      "epoch": 1.9295220980513281,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.031,
      "mean_token_accuracy": 0.22386527359485625,
      "num_tokens": 38356318.0,
      "step": 17180
    },
    {
      "entropy": 5.528080320358276,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.0798,
      "mean_token_accuracy": 0.21025200933218002,
      "num_tokens": 38367246.0,
      "step": 17185
    },
    {
      "entropy": 5.529863071441651,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20839520543813705,
      "num_tokens": 38378767.0,
      "step": 17190
    },
    {
      "entropy": 5.561200475692749,
      "epoch": 1.931206828775201,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21581949144601822,
      "num_tokens": 38390414.0,
      "step": 17195
    },
    {
      "entropy": 5.5392005920410154,
      "epoch": 1.9317684056831583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21654677242040635,
      "num_tokens": 38401569.0,
      "step": 17200
    },
    {
      "entropy": 5.597090387344361,
      "epoch": 1.9323299825911158,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.2079,
      "mean_token_accuracy": 0.2076840043067932,
      "num_tokens": 38413561.0,
      "step": 17205
    },
    {
      "entropy": 5.580898904800415,
      "epoch": 1.9328915594990734,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.134,
      "mean_token_accuracy": 0.21324655115604402,
      "num_tokens": 38423965.0,
      "step": 17210
    },
    {
      "entropy": 5.495054721832275,
      "epoch": 1.933453136407031,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.0595,
      "mean_token_accuracy": 0.20792473703622819,
      "num_tokens": 38434158.0,
      "step": 17215
    },
    {
      "entropy": 5.474375295639038,
      "epoch": 1.9340147133149885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20841258019208908,
      "num_tokens": 38445449.0,
      "step": 17220
    },
    {
      "entropy": 5.594013595581055,
      "epoch": 1.934576290222946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.1776,
      "mean_token_accuracy": 0.21004260182380677,
      "num_tokens": 38457601.0,
      "step": 17225
    },
    {
      "entropy": 5.559880876541138,
      "epoch": 1.9351378671309036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21740063428878784,
      "num_tokens": 38468631.0,
      "step": 17230
    },
    {
      "entropy": 5.489247035980225,
      "epoch": 1.935699444038861,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.125,
      "mean_token_accuracy": 0.208731210231781,
      "num_tokens": 38479567.0,
      "step": 17235
    },
    {
      "entropy": 5.477107000350952,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000463261612974268,
      "loss": 5.0972,
      "mean_token_accuracy": 0.21030912846326827,
      "num_tokens": 38489934.0,
      "step": 17240
    },
    {
      "entropy": 5.641141557693482,
      "epoch": 1.9368225978547762,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.2773,
      "mean_token_accuracy": 0.1997739180922508,
      "num_tokens": 38502214.0,
      "step": 17245
    },
    {
      "entropy": 5.612338209152222,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.1098,
      "mean_token_accuracy": 0.2069053068757057,
      "num_tokens": 38513915.0,
      "step": 17250
    },
    {
      "entropy": 5.54060959815979,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21377045810222625,
      "num_tokens": 38525039.0,
      "step": 17255
    },
    {
      "entropy": 5.484276247024536,
      "epoch": 1.9385073285786487,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.062,
      "mean_token_accuracy": 0.21758541315793992,
      "num_tokens": 38537264.0,
      "step": 17260
    },
    {
      "entropy": 5.562069940567016,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.1446,
      "mean_token_accuracy": 0.2138465866446495,
      "num_tokens": 38547926.0,
      "step": 17265
    },
    {
      "entropy": 5.578473281860352,
      "epoch": 1.939630482394564,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2165961354970932,
      "num_tokens": 38558745.0,
      "step": 17270
    },
    {
      "entropy": 5.51697301864624,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000463107558650873,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21311751157045364,
      "num_tokens": 38568190.0,
      "step": 17275
    },
    {
      "entropy": 5.59080491065979,
      "epoch": 1.940753636210479,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20837933719158172,
      "num_tokens": 38578932.0,
      "step": 17280
    },
    {
      "entropy": 5.544730615615845,
      "epoch": 1.9413152131184366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.1315,
      "mean_token_accuracy": 0.21001126766204833,
      "num_tokens": 38590501.0,
      "step": 17285
    },
    {
      "entropy": 5.508183193206787,
      "epoch": 1.9418767900263942,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.044,
      "mean_token_accuracy": 0.21760032773017884,
      "num_tokens": 38600498.0,
      "step": 17290
    },
    {
      "entropy": 5.507646656036377,
      "epoch": 1.9424383669343517,
      "grad_norm": 0.90625,
      "learning_rate": 0.000463019395826327,
      "loss": 5.0382,
      "mean_token_accuracy": 0.2106712654232979,
      "num_tokens": 38611991.0,
      "step": 17295
    },
    {
      "entropy": 5.477294206619263,
      "epoch": 1.9429999438423091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21664314568042756,
      "num_tokens": 38622284.0,
      "step": 17300
    },
    {
      "entropy": 5.56667127609253,
      "epoch": 1.9435615207502668,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.1608,
      "mean_token_accuracy": 0.20919508934020997,
      "num_tokens": 38633777.0,
      "step": 17305
    },
    {
      "entropy": 5.591518831253052,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.1405,
      "mean_token_accuracy": 0.20262035131454467,
      "num_tokens": 38644360.0,
      "step": 17310
    },
    {
      "entropy": 5.533865833282471,
      "epoch": 1.9446846745661819,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.1197,
      "mean_token_accuracy": 0.21016398668289185,
      "num_tokens": 38655242.0,
      "step": 17315
    },
    {
      "entropy": 5.529800033569336,
      "epoch": 1.9452462514741393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046290905760218193,
      "loss": 4.979,
      "mean_token_accuracy": 0.2239682763814926,
      "num_tokens": 38667792.0,
      "step": 17320
    },
    {
      "entropy": 5.489251661300659,
      "epoch": 1.945807828382097,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046288697200524964,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21804354339838028,
      "num_tokens": 38677735.0,
      "step": 17325
    },
    {
      "entropy": 5.559644031524658,
      "epoch": 1.9463694052900544,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.0803,
      "mean_token_accuracy": 0.21174318790435792,
      "num_tokens": 38689556.0,
      "step": 17330
    },
    {
      "entropy": 5.52203221321106,
      "epoch": 1.9469309821980119,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.0968,
      "mean_token_accuracy": 0.21413452476263045,
      "num_tokens": 38700556.0,
      "step": 17335
    },
    {
      "entropy": 5.529509496688843,
      "epoch": 1.9474925591059695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.0834,
      "mean_token_accuracy": 0.20894801765680313,
      "num_tokens": 38711019.0,
      "step": 17340
    },
    {
      "entropy": 5.527403974533081,
      "epoch": 1.9480541360139272,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.1,
      "mean_token_accuracy": 0.21058001071214677,
      "num_tokens": 38722228.0,
      "step": 17345
    },
    {
      "entropy": 5.489513969421386,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462776454298748,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21683382987976074,
      "num_tokens": 38732183.0,
      "step": 17350
    },
    {
      "entropy": 5.431792688369751,
      "epoch": 1.949177289829842,
      "grad_norm": 1.0,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21554741263389587,
      "num_tokens": 38743659.0,
      "step": 17355
    },
    {
      "entropy": 5.541941213607788,
      "epoch": 1.9497388667377997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20806533992290496,
      "num_tokens": 38755032.0,
      "step": 17360
    },
    {
      "entropy": 5.565519571304321,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.0587,
      "mean_token_accuracy": 0.21583800464868547,
      "num_tokens": 38766126.0,
      "step": 17365
    },
    {
      "entropy": 5.621300554275512,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.1913,
      "mean_token_accuracy": 0.21140002012252807,
      "num_tokens": 38776921.0,
      "step": 17370
    },
    {
      "entropy": 5.511100006103516,
      "epoch": 1.9514235974616723,
      "grad_norm": 1.0,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.0763,
      "mean_token_accuracy": 0.2092418536543846,
      "num_tokens": 38788884.0,
      "step": 17375
    },
    {
      "entropy": 5.641822338104248,
      "epoch": 1.95198517436963,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19803064912557602,
      "num_tokens": 38800505.0,
      "step": 17380
    },
    {
      "entropy": 5.536871290206909,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21095919311046601,
      "num_tokens": 38811132.0,
      "step": 17385
    },
    {
      "entropy": 5.60416808128357,
      "epoch": 1.953108328185545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.1238,
      "mean_token_accuracy": 0.21546659469604493,
      "num_tokens": 38821349.0,
      "step": 17390
    },
    {
      "entropy": 5.535746240615845,
      "epoch": 1.9536699050935025,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.1429,
      "mean_token_accuracy": 0.21040274053812028,
      "num_tokens": 38832970.0,
      "step": 17395
    },
    {
      "entropy": 5.504929113388061,
      "epoch": 1.9542314820014601,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.1176,
      "mean_token_accuracy": 0.210134357213974,
      "num_tokens": 38844220.0,
      "step": 17400
    },
    {
      "entropy": 5.528324699401855,
      "epoch": 1.9547930589094178,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20482401996850969,
      "num_tokens": 38854848.0,
      "step": 17405
    },
    {
      "entropy": 5.5476428985595705,
      "epoch": 1.9553546358173752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.2172,
      "mean_token_accuracy": 0.2033805653452873,
      "num_tokens": 38864827.0,
      "step": 17410
    },
    {
      "entropy": 5.570899581909179,
      "epoch": 1.9559162127253327,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20879019796848297,
      "num_tokens": 38875914.0,
      "step": 17415
    },
    {
      "entropy": 5.481240272521973,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.049,
      "mean_token_accuracy": 0.21483028531074524,
      "num_tokens": 38886565.0,
      "step": 17420
    },
    {
      "entropy": 5.593453407287598,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.294,
      "mean_token_accuracy": 0.1987745851278305,
      "num_tokens": 38897319.0,
      "step": 17425
    },
    {
      "entropy": 5.519160795211792,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.0,
      "learning_rate": 0.000462421793621278,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2143239051103592,
      "num_tokens": 38908824.0,
      "step": 17430
    },
    {
      "entropy": 5.554297542572021,
      "epoch": 1.9581625203571629,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000462399576574456,
      "loss": 5.1361,
      "mean_token_accuracy": 0.2127651661634445,
      "num_tokens": 38921800.0,
      "step": 17435
    },
    {
      "entropy": 5.528083562850952,
      "epoch": 1.9587240972651205,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.0979,
      "mean_token_accuracy": 0.21169813126325607,
      "num_tokens": 38932284.0,
      "step": 17440
    },
    {
      "entropy": 5.516578626632691,
      "epoch": 1.959285674173078,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2116766095161438,
      "num_tokens": 38943632.0,
      "step": 17445
    },
    {
      "entropy": 5.568294715881348,
      "epoch": 1.9598472510810354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21391698569059373,
      "num_tokens": 38954701.0,
      "step": 17450
    },
    {
      "entropy": 5.60570125579834,
      "epoch": 1.960408827988993,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.1387,
      "mean_token_accuracy": 0.208309343457222,
      "num_tokens": 38965625.0,
      "step": 17455
    },
    {
      "entropy": 5.5366309642791744,
      "epoch": 1.9609704048969507,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21059037744998932,
      "num_tokens": 38976989.0,
      "step": 17460
    },
    {
      "entropy": 5.445692682266236,
      "epoch": 1.9615319818049082,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.1352,
      "mean_token_accuracy": 0.21837786287069322,
      "num_tokens": 38987661.0,
      "step": 17465
    },
    {
      "entropy": 5.5091893672943115,
      "epoch": 1.9620935587128656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.0927,
      "mean_token_accuracy": 0.21209360808134078,
      "num_tokens": 38998014.0,
      "step": 17470
    },
    {
      "entropy": 5.5147889137268065,
      "epoch": 1.9626551356208233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2176202118396759,
      "num_tokens": 39008843.0,
      "step": 17475
    },
    {
      "entropy": 5.570705366134644,
      "epoch": 1.963216712528781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.1605,
      "mean_token_accuracy": 0.20591076016426085,
      "num_tokens": 39020173.0,
      "step": 17480
    },
    {
      "entropy": 5.54777045249939,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.1337,
      "mean_token_accuracy": 0.20553926080465318,
      "num_tokens": 39033087.0,
      "step": 17485
    },
    {
      "entropy": 5.612965202331543,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.111,
      "mean_token_accuracy": 0.21533661633729934,
      "num_tokens": 39045252.0,
      "step": 17490
    },
    {
      "entropy": 5.577392101287842,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.244,
      "mean_token_accuracy": 0.2048220381140709,
      "num_tokens": 39056755.0,
      "step": 17495
    },
    {
      "entropy": 5.511656808853149,
      "epoch": 1.9654630201606111,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21796070486307145,
      "num_tokens": 39066993.0,
      "step": 17500
    },
    {
      "entropy": 5.527868700027466,
      "epoch": 1.9660245970685686,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.0875,
      "mean_token_accuracy": 0.21312518715858458,
      "num_tokens": 39078671.0,
      "step": 17505
    },
    {
      "entropy": 5.577115774154663,
      "epoch": 1.966586173976526,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.1045,
      "mean_token_accuracy": 0.20414740294218064,
      "num_tokens": 39089563.0,
      "step": 17510
    },
    {
      "entropy": 5.614460754394531,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.1551,
      "mean_token_accuracy": 0.2042163074016571,
      "num_tokens": 39099837.0,
      "step": 17515
    },
    {
      "entropy": 5.603175401687622,
      "epoch": 1.9677093277924411,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.1575,
      "mean_token_accuracy": 0.2034602388739586,
      "num_tokens": 39112061.0,
      "step": 17520
    },
    {
      "entropy": 5.537866735458374,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.0833,
      "mean_token_accuracy": 0.21599120944738387,
      "num_tokens": 39122777.0,
      "step": 17525
    },
    {
      "entropy": 5.453569746017456,
      "epoch": 1.9688324816083562,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21262011379003526,
      "num_tokens": 39134827.0,
      "step": 17530
    },
    {
      "entropy": 5.524830961227417,
      "epoch": 1.9693940585163139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.0949,
      "mean_token_accuracy": 0.21093212217092513,
      "num_tokens": 39144352.0,
      "step": 17535
    },
    {
      "entropy": 5.4799196243286135,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21047249138355256,
      "num_tokens": 39155610.0,
      "step": 17540
    },
    {
      "entropy": 5.4797008514404295,
      "epoch": 1.9705172123322288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.0913,
      "mean_token_accuracy": 0.21067326813936232,
      "num_tokens": 39167058.0,
      "step": 17545
    },
    {
      "entropy": 5.605982446670533,
      "epoch": 1.9710787892401864,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20445574969053268,
      "num_tokens": 39178797.0,
      "step": 17550
    },
    {
      "entropy": 5.552499675750733,
      "epoch": 1.971640366148144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.045,
      "mean_token_accuracy": 0.21403322517871856,
      "num_tokens": 39188980.0,
      "step": 17555
    },
    {
      "entropy": 5.50287356376648,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.0468,
      "mean_token_accuracy": 0.20819973349571227,
      "num_tokens": 39200058.0,
      "step": 17560
    },
    {
      "entropy": 5.501881647109985,
      "epoch": 1.972763519964059,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.1344,
      "mean_token_accuracy": 0.21161461174488067,
      "num_tokens": 39211369.0,
      "step": 17565
    },
    {
      "entropy": 5.561885166168213,
      "epoch": 1.9733250968720166,
      "grad_norm": 1.0,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.1759,
      "mean_token_accuracy": 0.20572018772363662,
      "num_tokens": 39222438.0,
      "step": 17570
    },
    {
      "entropy": 5.602379226684571,
      "epoch": 1.9738866737799743,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.1432,
      "mean_token_accuracy": 0.20663530230522156,
      "num_tokens": 39233047.0,
      "step": 17575
    },
    {
      "entropy": 5.555154085159302,
      "epoch": 1.9744482506879317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.0691,
      "mean_token_accuracy": 0.21437205672264098,
      "num_tokens": 39244547.0,
      "step": 17580
    },
    {
      "entropy": 5.5591511726379395,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2079404816031456,
      "num_tokens": 39255549.0,
      "step": 17585
    },
    {
      "entropy": 5.498625755310059,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000461707892069342,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21309639662504196,
      "num_tokens": 39266527.0,
      "step": 17590
    },
    {
      "entropy": 5.5182281017303465,
      "epoch": 1.9761329814118045,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.1157,
      "mean_token_accuracy": 0.21229381263256072,
      "num_tokens": 39278013.0,
      "step": 17595
    },
    {
      "entropy": 5.575626087188721,
      "epoch": 1.976694558319762,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.0878,
      "mean_token_accuracy": 0.2108127787709236,
      "num_tokens": 39289349.0,
      "step": 17600
    },
    {
      "entropy": 5.532868337631226,
      "epoch": 1.9772561352277194,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2141982614994049,
      "num_tokens": 39301213.0,
      "step": 17605
    },
    {
      "entropy": 5.460589742660522,
      "epoch": 1.977817712135677,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004616182258676765,
      "loss": 4.9854,
      "mean_token_accuracy": 0.21805476844310762,
      "num_tokens": 39310960.0,
      "step": 17610
    },
    {
      "entropy": 5.551234102249145,
      "epoch": 1.9783792890436345,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.134,
      "mean_token_accuracy": 0.2087684154510498,
      "num_tokens": 39322603.0,
      "step": 17615
    },
    {
      "entropy": 5.550279521942139,
      "epoch": 1.978940865951592,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20941422432661055,
      "num_tokens": 39333811.0,
      "step": 17620
    },
    {
      "entropy": 5.616265201568604,
      "epoch": 1.9795024428595496,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.1269,
      "mean_token_accuracy": 0.20754290223121644,
      "num_tokens": 39345218.0,
      "step": 17625
    },
    {
      "entropy": 5.487884616851806,
      "epoch": 1.9800640197675072,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046152846459406765,
      "loss": 4.9515,
      "mean_token_accuracy": 0.21459653228521347,
      "num_tokens": 39356777.0,
      "step": 17630
    },
    {
      "entropy": 5.470108222961426,
      "epoch": 1.9806255966754647,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.0879,
      "mean_token_accuracy": 0.21187360137701033,
      "num_tokens": 39368288.0,
      "step": 17635
    },
    {
      "entropy": 5.585501575469971,
      "epoch": 1.9811871735834221,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.0878,
      "mean_token_accuracy": 0.21639569401741027,
      "num_tokens": 39379398.0,
      "step": 17640
    },
    {
      "entropy": 5.496791505813599,
      "epoch": 1.9817487504913798,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.0755,
      "mean_token_accuracy": 0.21543890684843064,
      "num_tokens": 39391129.0,
      "step": 17645
    },
    {
      "entropy": 5.547719955444336,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.1718,
      "mean_token_accuracy": 0.19846502542495728,
      "num_tokens": 39402442.0,
      "step": 17650
    },
    {
      "entropy": 5.601680755615234,
      "epoch": 1.9828719043072949,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.1121,
      "mean_token_accuracy": 0.21447552144527435,
      "num_tokens": 39413044.0,
      "step": 17655
    },
    {
      "entropy": 5.567352294921875,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.0,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.0838,
      "mean_token_accuracy": 0.21579736173152925,
      "num_tokens": 39423739.0,
      "step": 17660
    },
    {
      "entropy": 5.56474814414978,
      "epoch": 1.98399505812321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000461371153735959,
      "loss": 5.2158,
      "mean_token_accuracy": 0.2066743940114975,
      "num_tokens": 39436199.0,
      "step": 17665
    },
    {
      "entropy": 5.579655742645263,
      "epoch": 1.9845566350311676,
      "grad_norm": 0.875,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.1061,
      "mean_token_accuracy": 0.21502996534109114,
      "num_tokens": 39449302.0,
      "step": 17670
    },
    {
      "entropy": 5.576826667785644,
      "epoch": 1.985118211939125,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20939110070466996,
      "num_tokens": 39460723.0,
      "step": 17675
    },
    {
      "entropy": 5.511474370956421,
      "epoch": 1.9856797888470825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21747067272663118,
      "num_tokens": 39471251.0,
      "step": 17680
    },
    {
      "entropy": 5.553354358673095,
      "epoch": 1.9862413657550402,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.0704,
      "mean_token_accuracy": 0.22107060700654985,
      "num_tokens": 39482613.0,
      "step": 17685
    },
    {
      "entropy": 5.524911546707154,
      "epoch": 1.9868029426629978,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.0877,
      "mean_token_accuracy": 0.22049636840820314,
      "num_tokens": 39494626.0,
      "step": 17690
    },
    {
      "entropy": 5.474266862869262,
      "epoch": 1.9873645195709553,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2084072381258011,
      "num_tokens": 39505302.0,
      "step": 17695
    },
    {
      "entropy": 5.5574671745300295,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046121355210552626,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21467339545488356,
      "num_tokens": 39515854.0,
      "step": 17700
    },
    {
      "entropy": 5.580496501922608,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.1698,
      "mean_token_accuracy": 0.20428482443094254,
      "num_tokens": 39527504.0,
      "step": 17705
    },
    {
      "entropy": 5.574518299102783,
      "epoch": 1.989049250294828,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.0796,
      "mean_token_accuracy": 0.21308551579713822,
      "num_tokens": 39538885.0,
      "step": 17710
    },
    {
      "entropy": 5.522453737258911,
      "epoch": 1.9896108272027853,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.061,
      "mean_token_accuracy": 0.2142333984375,
      "num_tokens": 39549401.0,
      "step": 17715
    },
    {
      "entropy": 5.511954593658447,
      "epoch": 1.990172404110743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.137,
      "mean_token_accuracy": 0.20859352201223375,
      "num_tokens": 39560814.0,
      "step": 17720
    },
    {
      "entropy": 5.551943445205689,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20487646758556366,
      "num_tokens": 39571490.0,
      "step": 17725
    },
    {
      "entropy": 5.498092699050903,
      "epoch": 1.991295557926658,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22168999016284943,
      "num_tokens": 39582815.0,
      "step": 17730
    },
    {
      "entropy": 5.568689250946045,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.1041,
      "mean_token_accuracy": 0.21206796318292617,
      "num_tokens": 39593712.0,
      "step": 17735
    },
    {
      "entropy": 5.53940601348877,
      "epoch": 1.9924187117425731,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.0771,
      "mean_token_accuracy": 0.21189842671155928,
      "num_tokens": 39604504.0,
      "step": 17740
    },
    {
      "entropy": 5.537401580810547,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.1135,
      "mean_token_accuracy": 0.21371131539344787,
      "num_tokens": 39614971.0,
      "step": 17745
    },
    {
      "entropy": 5.534714460372925,
      "epoch": 1.9935418655584882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000460987903005792,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21461047381162643,
      "num_tokens": 39626647.0,
      "step": 17750
    },
    {
      "entropy": 5.571485185623169,
      "epoch": 1.9941034424664457,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.1714,
      "mean_token_accuracy": 0.2062537357211113,
      "num_tokens": 39636432.0,
      "step": 17755
    },
    {
      "entropy": 5.573623704910278,
      "epoch": 1.9946650193744033,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.1046,
      "mean_token_accuracy": 0.21211764514446257,
      "num_tokens": 39647055.0,
      "step": 17760
    },
    {
      "entropy": 5.556338691711426,
      "epoch": 1.995226596282361,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20910577178001405,
      "num_tokens": 39657891.0,
      "step": 17765
    },
    {
      "entropy": 5.479522562026977,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004608974775116888,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2222324788570404,
      "num_tokens": 39668023.0,
      "step": 17770
    },
    {
      "entropy": 5.492808389663696,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.1045,
      "mean_token_accuracy": 0.2145366072654724,
      "num_tokens": 39678398.0,
      "step": 17775
    },
    {
      "entropy": 5.475552654266357,
      "epoch": 1.9969113270062335,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.089,
      "mean_token_accuracy": 0.207901731133461,
      "num_tokens": 39690619.0,
      "step": 17780
    },
    {
      "entropy": 5.451734972000122,
      "epoch": 1.9974729039141912,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004608295962385911,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21642621904611586,
      "num_tokens": 39701690.0,
      "step": 17785
    },
    {
      "entropy": 5.536186552047729,
      "epoch": 1.9980344808221486,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.1645,
      "mean_token_accuracy": 0.2027362734079361,
      "num_tokens": 39713878.0,
      "step": 17790
    },
    {
      "entropy": 5.597382259368897,
      "epoch": 1.998596057730106,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000460784312470893,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20794167667627333,
      "num_tokens": 39725164.0,
      "step": 17795
    },
    {
      "entropy": 5.521429109573364,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.021,
      "mean_token_accuracy": 0.20922671109437943,
      "num_tokens": 39735354.0,
      "step": 17800
    },
    {
      "entropy": 5.607086515426635,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000460739005041417,
      "loss": 5.1812,
      "mean_token_accuracy": 0.20790481120347976,
      "num_tokens": 39746618.0,
      "step": 17805
    },
    {
      "entropy": 5.533457226223415,
      "epoch": 2.000224630763183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.115,
      "mean_token_accuracy": 0.209270684255494,
      "num_tokens": 39755697.0,
      "step": 17810
    },
    {
      "entropy": 5.451167869567871,
      "epoch": 2.0007862076711405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.019,
      "mean_token_accuracy": 0.21543059647083282,
      "num_tokens": 39766311.0,
      "step": 17815
    },
    {
      "entropy": 5.569439458847046,
      "epoch": 2.001347784579098,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20841307193040848,
      "num_tokens": 39777504.0,
      "step": 17820
    },
    {
      "entropy": 5.583997678756714,
      "epoch": 2.001909361487056,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046064831922021534,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21300911605358125,
      "num_tokens": 39789190.0,
      "step": 17825
    },
    {
      "entropy": 5.519847249984741,
      "epoch": 2.002470938395013,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004606256329853235,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2121793657541275,
      "num_tokens": 39801534.0,
      "step": 17830
    },
    {
      "entropy": 5.5759259223937985,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21558524370193483,
      "num_tokens": 39812353.0,
      "step": 17835
    },
    {
      "entropy": 5.4746650695800785,
      "epoch": 2.0035940922109283,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004605802427850861,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22116704881191254,
      "num_tokens": 39823313.0,
      "step": 17840
    },
    {
      "entropy": 5.461472702026367,
      "epoch": 2.004155669118886,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046055753882118564,
      "loss": 4.942,
      "mean_token_accuracy": 0.225359308719635,
      "num_tokens": 39833966.0,
      "step": 17845
    },
    {
      "entropy": 5.554065084457397,
      "epoch": 2.004717246026843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21651779860258102,
      "num_tokens": 39843963.0,
      "step": 17850
    },
    {
      "entropy": 5.563182020187378,
      "epoch": 2.005278822934801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21475581526756288,
      "num_tokens": 39855237.0,
      "step": 17855
    },
    {
      "entropy": 5.59713134765625,
      "epoch": 2.0058403998427585,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.1447,
      "mean_token_accuracy": 0.21314780712127684,
      "num_tokens": 39867270.0,
      "step": 17860
    },
    {
      "entropy": 5.602164697647095,
      "epoch": 2.006401976750716,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.0546,
      "mean_token_accuracy": 0.22170350402593614,
      "num_tokens": 39878103.0,
      "step": 17865
    },
    {
      "entropy": 5.504422330856324,
      "epoch": 2.0069635536586734,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000460443930392788,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21270087063312532,
      "num_tokens": 39888945.0,
      "step": 17870
    },
    {
      "entropy": 5.592469167709351,
      "epoch": 2.007525130566631,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.1777,
      "mean_token_accuracy": 0.20929083377122878,
      "num_tokens": 39899848.0,
      "step": 17875
    },
    {
      "entropy": 5.574280738830566,
      "epoch": 2.0080867074745887,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21353320628404618,
      "num_tokens": 39911824.0,
      "step": 17880
    },
    {
      "entropy": 5.545816898345947,
      "epoch": 2.0086482843825464,
      "grad_norm": 1.0,
      "learning_rate": 0.0004603756944746813,
      "loss": 4.9953,
      "mean_token_accuracy": 0.2146693915128708,
      "num_tokens": 39923257.0,
      "step": 17885
    },
    {
      "entropy": 5.509811449050903,
      "epoch": 2.0092098612905036,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046035293736281185,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22528252601623536,
      "num_tokens": 39935312.0,
      "step": 17890
    },
    {
      "entropy": 5.514961528778076,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004603301743492331,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21576204150915146,
      "num_tokens": 39946468.0,
      "step": 17895
    },
    {
      "entropy": 5.414021062850952,
      "epoch": 2.010333015106419,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004603074054346696,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22445444613695145,
      "num_tokens": 39957517.0,
      "step": 17900
    },
    {
      "entropy": 5.510267448425293,
      "epoch": 2.0108945920143766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004602846306198465,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21266891211271285,
      "num_tokens": 39968385.0,
      "step": 17905
    },
    {
      "entropy": 5.469725704193115,
      "epoch": 2.011456168922334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004602618499054889,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21681830435991287,
      "num_tokens": 39979393.0,
      "step": 17910
    },
    {
      "entropy": 5.553671073913574,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21304552108049393,
      "num_tokens": 39989709.0,
      "step": 17915
    },
    {
      "entropy": 5.549003934860229,
      "epoch": 2.012579322738249,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046021627078107177,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21550816148519517,
      "num_tokens": 40000472.0,
      "step": 17920
    },
    {
      "entropy": 5.385231399536133,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046019347237246343,
      "loss": 4.842,
      "mean_token_accuracy": 0.22408753335475923,
      "num_tokens": 40011062.0,
      "step": 17925
    },
    {
      "entropy": 5.593262958526611,
      "epoch": 2.013702476554164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004601706680672232,
      "loss": 5.1091,
      "mean_token_accuracy": 0.2023676872253418,
      "num_tokens": 40022430.0,
      "step": 17930
    },
    {
      "entropy": 5.499789810180664,
      "epoch": 2.0142640534621217,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004601478578660771,
      "loss": 4.9878,
      "mean_token_accuracy": 0.216521954536438,
      "num_tokens": 40033077.0,
      "step": 17935
    },
    {
      "entropy": 5.553524160385132,
      "epoch": 2.0148256303700793,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004601250417697513,
      "loss": 5.0108,
      "mean_token_accuracy": 0.2204272821545601,
      "num_tokens": 40044258.0,
      "step": 17940
    },
    {
      "entropy": 5.574278116226196,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21157810389995574,
      "num_tokens": 40055876.0,
      "step": 17945
    },
    {
      "entropy": 5.502738666534424,
      "epoch": 2.015948784185994,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004600793918944671,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2228882759809494,
      "num_tokens": 40066691.0,
      "step": 17950
    },
    {
      "entropy": 5.512238788604736,
      "epoch": 2.016510361093952,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004600565581169621,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21588892489671707,
      "num_tokens": 40077335.0,
      "step": 17955
    },
    {
      "entropy": 5.535549259185791,
      "epoch": 2.0170719380019095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20765413641929625,
      "num_tokens": 40089053.0,
      "step": 17960
    },
    {
      "entropy": 5.496164703369141,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046001087288586164,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22210579514503478,
      "num_tokens": 40099371.0,
      "step": 17965
    },
    {
      "entropy": 5.550470972061158,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.0703,
      "mean_token_accuracy": 0.2090276673436165,
      "num_tokens": 40110274.0,
      "step": 17970
    },
    {
      "entropy": 5.5838541984558105,
      "epoch": 2.018756668725782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21264989525079728,
      "num_tokens": 40120364.0,
      "step": 17975
    },
    {
      "entropy": 5.52621054649353,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21338126063346863,
      "num_tokens": 40132400.0,
      "step": 17980
    },
    {
      "entropy": 5.550860261917114,
      "epoch": 2.019879822541697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21379424631595612,
      "num_tokens": 40142820.0,
      "step": 17985
    },
    {
      "entropy": 5.572525215148926,
      "epoch": 2.0204413994496546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20750982612371444,
      "num_tokens": 40153620.0,
      "step": 17990
    },
    {
      "entropy": 5.554708003997803,
      "epoch": 2.0210029763576123,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004598736758362198,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21535060852766036,
      "num_tokens": 40164250.0,
      "step": 17995
    },
    {
      "entropy": 5.5052910327911375,
      "epoch": 2.02156455326557,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045985078905445734,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21716532111167908,
      "num_tokens": 40175809.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.334257876962185,
      "eval_loss": 5.158364772796631,
      "eval_mean_token_accuracy": 0.21636118419105194,
      "eval_num_tokens": 40175809.0,
      "eval_runtime": 18.5119,
      "eval_samples_per_second": 1675.299,
      "eval_steps_per_second": 209.433,
      "step": 18000
    },
    {
      "entropy": 5.419641017913818,
      "epoch": 2.022126130173527,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045982789638697414,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22538330554962158,
      "num_tokens": 40187855.0,
      "step": 18005
    },
    {
      "entropy": 5.531759405136109,
      "epoch": 2.022687707081485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045980499783449897,
      "loss": 4.982,
      "mean_token_accuracy": 0.2212439477443695,
      "num_tokens": 40198802.0,
      "step": 18010
    },
    {
      "entropy": 5.645687913894653,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.1793,
      "mean_token_accuracy": 0.2011903151869774,
      "num_tokens": 40210636.0,
      "step": 18015
    },
    {
      "entropy": 5.57664270401001,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21706817150115967,
      "num_tokens": 40221216.0,
      "step": 18020
    },
    {
      "entropy": 5.493118238449097,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004597362668744141,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22497875094413758,
      "num_tokens": 40231988.0,
      "step": 18025
    },
    {
      "entropy": 5.62480845451355,
      "epoch": 2.024934014713315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.208,
      "mean_token_accuracy": 0.20432210713624954,
      "num_tokens": 40243789.0,
      "step": 18030
    },
    {
      "entropy": 5.548117160797119,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21950574219226837,
      "num_tokens": 40254771.0,
      "step": 18035
    },
    {
      "entropy": 5.541691398620605,
      "epoch": 2.02605716852923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045966748297566085,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2201815202832222,
      "num_tokens": 40265444.0,
      "step": 18040
    },
    {
      "entropy": 5.538756942749023,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2110884055495262,
      "num_tokens": 40277146.0,
      "step": 18045
    },
    {
      "entropy": 5.557550668716431,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.0901,
      "mean_token_accuracy": 0.2146453380584717,
      "num_tokens": 40286493.0,
      "step": 18050
    },
    {
      "entropy": 5.569970178604126,
      "epoch": 2.027741899253103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21552975624799728,
      "num_tokens": 40297209.0,
      "step": 18055
    },
    {
      "entropy": 5.447643566131592,
      "epoch": 2.02830347616106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045957568879568736,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21825084537267686,
      "num_tokens": 40308666.0,
      "step": 18060
    },
    {
      "entropy": 5.567036294937134,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.0906,
      "mean_token_accuracy": 0.20901870876550674,
      "num_tokens": 40321145.0,
      "step": 18065
    },
    {
      "entropy": 5.581127405166626,
      "epoch": 2.0294266299769754,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21430826485157012,
      "num_tokens": 40332418.0,
      "step": 18070
    },
    {
      "entropy": 5.480958890914917,
      "epoch": 2.029988206884933,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045950678145005456,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21771776229143142,
      "num_tokens": 40342838.0,
      "step": 18075
    },
    {
      "entropy": 5.499071884155273,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004594838005843333,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2185700938105583,
      "num_tokens": 40352940.0,
      "step": 18080
    },
    {
      "entropy": 5.620015954971313,
      "epoch": 2.031111360700848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.0903,
      "mean_token_accuracy": 0.20844756960868835,
      "num_tokens": 40364273.0,
      "step": 18085
    },
    {
      "entropy": 5.5350569725036625,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045943782123151556,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22511158287525176,
      "num_tokens": 40374617.0,
      "step": 18090
    },
    {
      "entropy": 5.51204137802124,
      "epoch": 2.0322345145167633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21240984350442887,
      "num_tokens": 40385398.0,
      "step": 18095
    },
    {
      "entropy": 5.664219045639038,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.1314,
      "mean_token_accuracy": 0.2089846760034561,
      "num_tokens": 40397030.0,
      "step": 18100
    },
    {
      "entropy": 5.61735315322876,
      "epoch": 2.033357668332678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20691438466310502,
      "num_tokens": 40408027.0,
      "step": 18105
    },
    {
      "entropy": 5.603941202163696,
      "epoch": 2.033919245240636,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.035,
      "mean_token_accuracy": 0.21354867219924928,
      "num_tokens": 40418755.0,
      "step": 18110
    },
    {
      "entropy": 5.591098213195801,
      "epoch": 2.034480822148593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.0593,
      "mean_token_accuracy": 0.21527769267559052,
      "num_tokens": 40430694.0,
      "step": 18115
    },
    {
      "entropy": 5.538786840438843,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21735875755548478,
      "num_tokens": 40442019.0,
      "step": 18120
    },
    {
      "entropy": 5.5661121845245365,
      "epoch": 2.0356039759645084,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20454643964767455,
      "num_tokens": 40454120.0,
      "step": 18125
    },
    {
      "entropy": 5.550861787796021,
      "epoch": 2.036165552872466,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.1637,
      "mean_token_accuracy": 0.21220882683992387,
      "num_tokens": 40467445.0,
      "step": 18130
    },
    {
      "entropy": 5.481247234344482,
      "epoch": 2.0367271297804233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045923062353615167,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21782450675964354,
      "num_tokens": 40477512.0,
      "step": 18135
    },
    {
      "entropy": 5.450270652770996,
      "epoch": 2.037288706688381,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045920757223034833,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21996873021125793,
      "num_tokens": 40488209.0,
      "step": 18140
    },
    {
      "entropy": 5.574331521987915,
      "epoch": 2.0378502835963386,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.034,
      "mean_token_accuracy": 0.212552447617054,
      "num_tokens": 40499793.0,
      "step": 18145
    },
    {
      "entropy": 5.5816065788269045,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.1326,
      "mean_token_accuracy": 0.20790785402059556,
      "num_tokens": 40510598.0,
      "step": 18150
    },
    {
      "entropy": 5.63364953994751,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.0717,
      "mean_token_accuracy": 0.2183879181742668,
      "num_tokens": 40521289.0,
      "step": 18155
    },
    {
      "entropy": 5.613236904144287,
      "epoch": 2.039535014320211,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.1007,
      "mean_token_accuracy": 0.21238788068294526,
      "num_tokens": 40532048.0,
      "step": 18160
    },
    {
      "entropy": 5.540859937667847,
      "epoch": 2.0400965912281688,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21120023876428604,
      "num_tokens": 40543699.0,
      "step": 18165
    },
    {
      "entropy": 5.587678146362305,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.1034,
      "mean_token_accuracy": 0.2152194619178772,
      "num_tokens": 40554626.0,
      "step": 18170
    },
    {
      "entropy": 5.555510473251343,
      "epoch": 2.0412197450440837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045904604890411476,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21356632709503173,
      "num_tokens": 40567046.0,
      "step": 18175
    },
    {
      "entropy": 5.6751251220703125,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.1352,
      "mean_token_accuracy": 0.20509109348058702,
      "num_tokens": 40578472.0,
      "step": 18180
    },
    {
      "entropy": 5.548484563827515,
      "epoch": 2.042342898859999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.047,
      "mean_token_accuracy": 0.21397585272789002,
      "num_tokens": 40589012.0,
      "step": 18185
    },
    {
      "entropy": 5.455786466598511,
      "epoch": 2.0429044757679566,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045897673670528086,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21633989959955216,
      "num_tokens": 40600807.0,
      "step": 18190
    },
    {
      "entropy": 5.626452112197876,
      "epoch": 2.043466052675914,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.1461,
      "mean_token_accuracy": 0.21260230988264084,
      "num_tokens": 40612578.0,
      "step": 18195
    },
    {
      "entropy": 5.533149862289429,
      "epoch": 2.0440276295838715,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004589304992820454,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22182343304157257,
      "num_tokens": 40623504.0,
      "step": 18200
    },
    {
      "entropy": 5.574591636657715,
      "epoch": 2.044589206491829,
      "grad_norm": 1.046875,
      "learning_rate": 0.000458907371785432,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2141539454460144,
      "num_tokens": 40633676.0,
      "step": 18205
    },
    {
      "entropy": 5.555432605743408,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.1156,
      "mean_token_accuracy": 0.21100273728370667,
      "num_tokens": 40645991.0,
      "step": 18210
    },
    {
      "entropy": 5.536661529541016,
      "epoch": 2.045712360307744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045886109922589564,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21809422075748444,
      "num_tokens": 40656617.0,
      "step": 18215
    },
    {
      "entropy": 5.5460446834564205,
      "epoch": 2.0462739372157017,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.0756,
      "mean_token_accuracy": 0.21840692162513733,
      "num_tokens": 40669708.0,
      "step": 18220
    },
    {
      "entropy": 5.502389860153198,
      "epoch": 2.0468355141236594,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21136400997638702,
      "num_tokens": 40680248.0,
      "step": 18225
    },
    {
      "entropy": 5.615874576568603,
      "epoch": 2.0473970910316166,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.1242,
      "mean_token_accuracy": 0.20721979439258575,
      "num_tokens": 40692704.0,
      "step": 18230
    },
    {
      "entropy": 5.580901098251343,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.033,
      "mean_token_accuracy": 0.2221447929739952,
      "num_tokens": 40703156.0,
      "step": 18235
    },
    {
      "entropy": 5.503661584854126,
      "epoch": 2.048520244847532,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045874531539130204,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2162831246852875,
      "num_tokens": 40713996.0,
      "step": 18240
    },
    {
      "entropy": 5.498610973358154,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045872214106986684,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2227897047996521,
      "num_tokens": 40724532.0,
      "step": 18245
    },
    {
      "entropy": 5.597820615768432,
      "epoch": 2.049643398663447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.078,
      "mean_token_accuracy": 0.21177640855312346,
      "num_tokens": 40736045.0,
      "step": 18250
    },
    {
      "entropy": 5.432802152633667,
      "epoch": 2.0502049755714045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004586757748783817,
      "loss": 4.9402,
      "mean_token_accuracy": 0.21847816705703735,
      "num_tokens": 40747398.0,
      "step": 18255
    },
    {
      "entropy": 5.416982555389405,
      "epoch": 2.050766552479362,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045865258300980817,
      "loss": 4.8856,
      "mean_token_accuracy": 0.2260712519288063,
      "num_tokens": 40758219.0,
      "step": 18260
    },
    {
      "entropy": 5.562691020965576,
      "epoch": 2.05132812938732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045862938529366514,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22397554218769072,
      "num_tokens": 40770002.0,
      "step": 18265
    },
    {
      "entropy": 5.500147819519043,
      "epoch": 2.051889706295277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004586061817306913,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21905215531587602,
      "num_tokens": 40780087.0,
      "step": 18270
    },
    {
      "entropy": 5.521575117111206,
      "epoch": 2.0524512832032347,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2081024169921875,
      "num_tokens": 40792225.0,
      "step": 18275
    },
    {
      "entropy": 5.559395837783813,
      "epoch": 2.0530128601111923,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20567791610956193,
      "num_tokens": 40804437.0,
      "step": 18280
    },
    {
      "entropy": 5.467989587783814,
      "epoch": 2.05357443701915,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004585365359681736,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23042981922626496,
      "num_tokens": 40814625.0,
      "step": 18285
    },
    {
      "entropy": 5.451507472991944,
      "epoch": 2.054136013927107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045851330902526633,
      "loss": 4.9641,
      "mean_token_accuracy": 0.21802471131086348,
      "num_tokens": 40825843.0,
      "step": 18290
    },
    {
      "entropy": 5.527272081375122,
      "epoch": 2.054697590835065,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.1453,
      "mean_token_accuracy": 0.20133074671030043,
      "num_tokens": 40837146.0,
      "step": 18295
    },
    {
      "entropy": 5.658598184585571,
      "epoch": 2.0552591677430225,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.2027,
      "mean_token_accuracy": 0.20035164207220077,
      "num_tokens": 40849109.0,
      "step": 18300
    },
    {
      "entropy": 5.520171451568603,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.015625,
      "learning_rate": 0.000458443593140694,
      "loss": 4.9781,
      "mean_token_accuracy": 0.20988082587718965,
      "num_tokens": 40861168.0,
      "step": 18305
    },
    {
      "entropy": 5.584891891479492,
      "epoch": 2.0563823215589374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.0861,
      "mean_token_accuracy": 0.212784481048584,
      "num_tokens": 40873403.0,
      "step": 18310
    },
    {
      "entropy": 5.525013971328735,
      "epoch": 2.056943898466895,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004583970866785035,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2129603296518326,
      "num_tokens": 40883434.0,
      "step": 18315
    },
    {
      "entropy": 5.544161415100097,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.0216,
      "mean_token_accuracy": 0.21142987310886383,
      "num_tokens": 40894742.0,
      "step": 18320
    },
    {
      "entropy": 5.502384328842163,
      "epoch": 2.05806705228281,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004583505568585757,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21947601437568665,
      "num_tokens": 40905233.0,
      "step": 18325
    },
    {
      "entropy": 5.537048625946045,
      "epoch": 2.0586286291907676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045832728319131205,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22312068641185762,
      "num_tokens": 40915897.0,
      "step": 18330
    },
    {
      "entropy": 5.495863962173462,
      "epoch": 2.0591902060987253,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004583040036868367,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21405690163373947,
      "num_tokens": 40928178.0,
      "step": 18335
    },
    {
      "entropy": 5.498478555679322,
      "epoch": 2.059751783006683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20849822759628295,
      "num_tokens": 40938988.0,
      "step": 18340
    },
    {
      "entropy": 5.486923122406006,
      "epoch": 2.06031335991464,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045825742716921553,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22056949138641357,
      "num_tokens": 40949671.0,
      "step": 18345
    },
    {
      "entropy": 5.437801790237427,
      "epoch": 2.060874936822598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004582341301575529,
      "loss": 4.9149,
      "mean_token_accuracy": 0.21660062968730925,
      "num_tokens": 40960818.0,
      "step": 18350
    },
    {
      "entropy": 5.624400281906128,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20221328884363174,
      "num_tokens": 40973866.0,
      "step": 18355
    },
    {
      "entropy": 5.5339373588562015,
      "epoch": 2.061998090638513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.0346,
      "mean_token_accuracy": 0.21197231113910675,
      "num_tokens": 40983355.0,
      "step": 18360
    },
    {
      "entropy": 5.472381353378296,
      "epoch": 2.0625596675464704,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045816420412005826,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2203068420290947,
      "num_tokens": 40994045.0,
      "step": 18365
    },
    {
      "entropy": 5.447839546203613,
      "epoch": 2.063121244454428,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045814088377586505,
      "loss": 4.942,
      "mean_token_accuracy": 0.2215038314461708,
      "num_tokens": 41003974.0,
      "step": 18370
    },
    {
      "entropy": 5.514606523513794,
      "epoch": 2.0636828213623857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.0898,
      "mean_token_accuracy": 0.2143349751830101,
      "num_tokens": 41015262.0,
      "step": 18375
    },
    {
      "entropy": 5.589032793045044,
      "epoch": 2.0642443982703433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.1274,
      "mean_token_accuracy": 0.21062336564064027,
      "num_tokens": 41027103.0,
      "step": 18380
    },
    {
      "entropy": 5.582323741912842,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21459201276302337,
      "num_tokens": 41037240.0,
      "step": 18385
    },
    {
      "entropy": 5.6005034923553465,
      "epoch": 2.065367552086258,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.0049,
      "mean_token_accuracy": 0.22095420956611633,
      "num_tokens": 41048258.0,
      "step": 18390
    },
    {
      "entropy": 5.55772032737732,
      "epoch": 2.065929128994216,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20876812040805817,
      "num_tokens": 41058204.0,
      "step": 18395
    },
    {
      "entropy": 5.623005008697509,
      "epoch": 2.066490705902173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.0835,
      "mean_token_accuracy": 0.21040922850370408,
      "num_tokens": 41070019.0,
      "step": 18400
    },
    {
      "entropy": 5.599216318130493,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21564376801252366,
      "num_tokens": 41081503.0,
      "step": 18405
    },
    {
      "entropy": 5.580285263061524,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20955599397420882,
      "num_tokens": 41091959.0,
      "step": 18410
    },
    {
      "entropy": 5.566930913925171,
      "epoch": 2.068175436626046,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21094672828912736,
      "num_tokens": 41103193.0,
      "step": 18415
    },
    {
      "entropy": 5.595669937133789,
      "epoch": 2.0687370135340033,
      "grad_norm": 1.0,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2184932693839073,
      "num_tokens": 41115886.0,
      "step": 18420
    },
    {
      "entropy": 5.538093423843383,
      "epoch": 2.069298590441961,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045788397528896224,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2169068083167076,
      "num_tokens": 41126868.0,
      "step": 18425
    },
    {
      "entropy": 5.582740640640258,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2217319831252098,
      "num_tokens": 41136417.0,
      "step": 18430
    },
    {
      "entropy": 5.544301319122314,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21168864965438844,
      "num_tokens": 41149515.0,
      "step": 18435
    },
    {
      "entropy": 5.611599445343018,
      "epoch": 2.0709833211658335,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000457813787011288,
      "loss": 5.1201,
      "mean_token_accuracy": 0.21515440344810485,
      "num_tokens": 41160023.0,
      "step": 18440
    },
    {
      "entropy": 5.446491670608521,
      "epoch": 2.071544898073791,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004577903792760127,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22639045864343643,
      "num_tokens": 41170909.0,
      "step": 18445
    },
    {
      "entropy": 5.5852948188781735,
      "epoch": 2.072106474981749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20891152024269105,
      "num_tokens": 41181960.0,
      "step": 18450
    },
    {
      "entropy": 5.624131965637207,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.1071,
      "mean_token_accuracy": 0.21428196877241135,
      "num_tokens": 41192292.0,
      "step": 18455
    },
    {
      "entropy": 5.60480523109436,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2180459201335907,
      "num_tokens": 41203826.0,
      "step": 18460
    },
    {
      "entropy": 5.507061004638672,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.0885,
      "mean_token_accuracy": 0.21428006291389465,
      "num_tokens": 41215095.0,
      "step": 18465
    },
    {
      "entropy": 5.5521927833557125,
      "epoch": 2.074352782613579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045767325331280044,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21446262300014496,
      "num_tokens": 41224953.0,
      "step": 18470
    },
    {
      "entropy": 5.577698993682861,
      "epoch": 2.0749143595215367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.0673,
      "mean_token_accuracy": 0.20524761378765105,
      "num_tokens": 41235479.0,
      "step": 18475
    },
    {
      "entropy": 5.563669490814209,
      "epoch": 2.075475936429494,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21387392282485962,
      "num_tokens": 41248520.0,
      "step": 18480
    },
    {
      "entropy": 5.631129026412964,
      "epoch": 2.0760375133374516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000457602907932259,
      "loss": 5.1656,
      "mean_token_accuracy": 0.2062474250793457,
      "num_tokens": 41259112.0,
      "step": 18485
    },
    {
      "entropy": 5.5459761142730715,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.125,
      "learning_rate": 0.00045757944784278803,
      "loss": 5.003,
      "mean_token_accuracy": 0.2152387022972107,
      "num_tokens": 41269554.0,
      "step": 18490
    },
    {
      "entropy": 5.563116931915284,
      "epoch": 2.0771606671533664,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2157930091023445,
      "num_tokens": 41281339.0,
      "step": 18495
    },
    {
      "entropy": 5.4908682346344,
      "epoch": 2.077722244061324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045753251022439195,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22298128753900529,
      "num_tokens": 41291629.0,
      "step": 18500
    },
    {
      "entropy": 5.591483545303345,
      "epoch": 2.0782838209692818,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20964066982269286,
      "num_tokens": 41302162.0,
      "step": 18505
    },
    {
      "entropy": 5.538320398330688,
      "epoch": 2.0788453978772394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21771809607744216,
      "num_tokens": 41313667.0,
      "step": 18510
    },
    {
      "entropy": 5.538956260681152,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045746206020937276,
      "loss": 5.005,
      "mean_token_accuracy": 0.2130013272166252,
      "num_tokens": 41325063.0,
      "step": 18515
    },
    {
      "entropy": 5.519079685211182,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.0712,
      "mean_token_accuracy": 0.2148708701133728,
      "num_tokens": 41335974.0,
      "step": 18520
    },
    {
      "entropy": 5.483977746963501,
      "epoch": 2.080530128601112,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045741506448313327,
      "loss": 5.0097,
      "mean_token_accuracy": 0.2195491075515747,
      "num_tokens": 41347469.0,
      "step": 18525
    },
    {
      "entropy": 5.615661668777466,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.0,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21746228039264678,
      "num_tokens": 41358418.0,
      "step": 18530
    },
    {
      "entropy": 5.584331130981445,
      "epoch": 2.081653282417027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.0983,
      "mean_token_accuracy": 0.21439199000597,
      "num_tokens": 41368859.0,
      "step": 18535
    },
    {
      "entropy": 5.619612121582032,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.1469,
      "mean_token_accuracy": 0.204937744140625,
      "num_tokens": 41380929.0,
      "step": 18540
    },
    {
      "entropy": 5.55598087310791,
      "epoch": 2.082776436232942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004573210033386464,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2247625172138214,
      "num_tokens": 41392190.0,
      "step": 18545
    },
    {
      "entropy": 5.619199180603028,
      "epoch": 2.0833380131409,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21872944831848146,
      "num_tokens": 41402646.0,
      "step": 18550
    },
    {
      "entropy": 5.5332451343536375,
      "epoch": 2.083899590048857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21755874454975127,
      "num_tokens": 41415179.0,
      "step": 18555
    },
    {
      "entropy": 5.580738544464111,
      "epoch": 2.0844611669568147,
      "grad_norm": 1.0,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.085,
      "mean_token_accuracy": 0.20952620953321457,
      "num_tokens": 41426695.0,
      "step": 18560
    },
    {
      "entropy": 5.510734605789184,
      "epoch": 2.0850227438647724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2140922501683235,
      "num_tokens": 41438078.0,
      "step": 18565
    },
    {
      "entropy": 5.57774806022644,
      "epoch": 2.08558432077273,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21235553473234176,
      "num_tokens": 41448862.0,
      "step": 18570
    },
    {
      "entropy": 5.490666294097901,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045717973748176695,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22526055276393891,
      "num_tokens": 41458528.0,
      "step": 18575
    },
    {
      "entropy": 5.485172605514526,
      "epoch": 2.086707474588645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045715617286555516,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21403821408748627,
      "num_tokens": 41469553.0,
      "step": 18580
    },
    {
      "entropy": 5.626990699768067,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.1556,
      "mean_token_accuracy": 0.2133496880531311,
      "num_tokens": 41482327.0,
      "step": 18585
    },
    {
      "entropy": 5.499331665039063,
      "epoch": 2.08783062840456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571090262341125,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22859012186527253,
      "num_tokens": 41493126.0,
      "step": 18590
    },
    {
      "entropy": 5.561182451248169,
      "epoch": 2.0883922053125175,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.103,
      "mean_token_accuracy": 0.20945305824279786,
      "num_tokens": 41505597.0,
      "step": 18595
    },
    {
      "entropy": 5.642197608947754,
      "epoch": 2.088953782220475,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.13,
      "mean_token_accuracy": 0.21093325316905975,
      "num_tokens": 41517296.0,
      "step": 18600
    },
    {
      "entropy": 5.584100246429443,
      "epoch": 2.089515359128433,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21629886627197265,
      "num_tokens": 41529900.0,
      "step": 18605
    },
    {
      "entropy": 5.557789945602417,
      "epoch": 2.09007693603639,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.0681,
      "mean_token_accuracy": 0.21501244753599166,
      "num_tokens": 41540641.0,
      "step": 18610
    },
    {
      "entropy": 5.484941053390503,
      "epoch": 2.0906385129443477,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20818408131599425,
      "num_tokens": 41551335.0,
      "step": 18615
    },
    {
      "entropy": 5.5314312934875485,
      "epoch": 2.0912000898523053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21950836777687072,
      "num_tokens": 41562265.0,
      "step": 18620
    },
    {
      "entropy": 5.545453977584839,
      "epoch": 2.091761666760263,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004569438304132031,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22009489238262175,
      "num_tokens": 41574758.0,
      "step": 18625
    },
    {
      "entropy": 5.473229646682739,
      "epoch": 2.09232324366822,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045692020783155376,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21851201206445695,
      "num_tokens": 41585007.0,
      "step": 18630
    },
    {
      "entropy": 5.48208270072937,
      "epoch": 2.092884820576178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004568965794574943,
      "loss": 4.9599,
      "mean_token_accuracy": 0.2151700660586357,
      "num_tokens": 41596026.0,
      "step": 18635
    },
    {
      "entropy": 5.606670713424682,
      "epoch": 2.0934463974841355,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.09,
      "mean_token_accuracy": 0.21522781103849412,
      "num_tokens": 41608440.0,
      "step": 18640
    },
    {
      "entropy": 5.551240158081055,
      "epoch": 2.094007974392093,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045684930533515506,
      "loss": 4.9893,
      "mean_token_accuracy": 0.21555105894804,
      "num_tokens": 41620467.0,
      "step": 18645
    },
    {
      "entropy": 5.483487272262574,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045682565958838045,
      "loss": 5.006,
      "mean_token_accuracy": 0.2133037582039833,
      "num_tokens": 41630918.0,
      "step": 18650
    },
    {
      "entropy": 5.533989143371582,
      "epoch": 2.095131128208008,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004568020080522063,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2105986714363098,
      "num_tokens": 41641148.0,
      "step": 18655
    },
    {
      "entropy": 5.613045740127563,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21939192414283754,
      "num_tokens": 41652148.0,
      "step": 18660
    },
    {
      "entropy": 5.525360727310181,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045675468761467204,
      "loss": 5.01,
      "mean_token_accuracy": 0.21960366368293763,
      "num_tokens": 41662275.0,
      "step": 18665
    },
    {
      "entropy": 5.562108087539673,
      "epoch": 2.0968158589318806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21044763475656508,
      "num_tokens": 41673141.0,
      "step": 18670
    },
    {
      "entropy": 5.664791679382324,
      "epoch": 2.0973774358398383,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.1827,
      "mean_token_accuracy": 0.2053629368543625,
      "num_tokens": 41683473.0,
      "step": 18675
    },
    {
      "entropy": 5.460963916778565,
      "epoch": 2.097939012747796,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045668366355670697,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22168046832084656,
      "num_tokens": 41693847.0,
      "step": 18680
    },
    {
      "entropy": 5.527900362014771,
      "epoch": 2.098500589655753,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21920809596776963,
      "num_tokens": 41705810.0,
      "step": 18685
    },
    {
      "entropy": 5.540380907058716,
      "epoch": 2.099062166563711,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045663628525908237,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2222769856452942,
      "num_tokens": 41717160.0,
      "step": 18690
    },
    {
      "entropy": 5.5995183944702145,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.1093,
      "mean_token_accuracy": 0.21248626559972764,
      "num_tokens": 41728631.0,
      "step": 18695
    },
    {
      "entropy": 5.621408271789551,
      "epoch": 2.100185320379626,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.1471,
      "mean_token_accuracy": 0.2120542585849762,
      "num_tokens": 41738933.0,
      "step": 18700
    },
    {
      "entropy": 5.6149030208587645,
      "epoch": 2.1007468972875833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.1345,
      "mean_token_accuracy": 0.21550002098083496,
      "num_tokens": 41750386.0,
      "step": 18705
    },
    {
      "entropy": 5.520841932296753,
      "epoch": 2.101308474195541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21219655126333237,
      "num_tokens": 41760711.0,
      "step": 18710
    },
    {
      "entropy": 5.55823860168457,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.1003,
      "mean_token_accuracy": 0.2083415910601616,
      "num_tokens": 41772147.0,
      "step": 18715
    },
    {
      "entropy": 5.576873254776001,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045649401158948664,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22232021689414977,
      "num_tokens": 41783265.0,
      "step": 18720
    },
    {
      "entropy": 5.562469387054444,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004564702790808758,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22281046956777573,
      "num_tokens": 41793977.0,
      "step": 18725
    },
    {
      "entropy": 5.575196933746338,
      "epoch": 2.103554781827371,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004564465407941809,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21851780265569687,
      "num_tokens": 41804790.0,
      "step": 18730
    },
    {
      "entropy": 5.4644654273986815,
      "epoch": 2.104116358735329,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045642279673015787,
      "loss": 4.98,
      "mean_token_accuracy": 0.21095648258924485,
      "num_tokens": 41815611.0,
      "step": 18735
    },
    {
      "entropy": 5.469069957733154,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.014,
      "mean_token_accuracy": 0.21651919633150102,
      "num_tokens": 41826935.0,
      "step": 18740
    },
    {
      "entropy": 5.491155052185059,
      "epoch": 2.1052395125512438,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004563752912731512,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21770624369382857,
      "num_tokens": 41838932.0,
      "step": 18745
    },
    {
      "entropy": 5.4488622665405275,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.0349,
      "mean_token_accuracy": 0.2130894735455513,
      "num_tokens": 41850771.0,
      "step": 18750
    },
    {
      "entropy": 5.488556861877441,
      "epoch": 2.106362666367159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045632776271590624,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22131444215774537,
      "num_tokens": 41861515.0,
      "step": 18755
    },
    {
      "entropy": 5.523909378051758,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004563039897765859,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23386247903108598,
      "num_tokens": 41872195.0,
      "step": 18760
    },
    {
      "entropy": 5.46957368850708,
      "epoch": 2.107485820183074,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004562802110644764,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22413865327835084,
      "num_tokens": 41882496.0,
      "step": 18765
    },
    {
      "entropy": 5.640412092208862,
      "epoch": 2.1080473970910316,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20482517927885055,
      "num_tokens": 41893078.0,
      "step": 18770
    },
    {
      "entropy": 5.602809810638428,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21554078310728073,
      "num_tokens": 41905062.0,
      "step": 18775
    },
    {
      "entropy": 5.476525926589966,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.027,
      "mean_token_accuracy": 0.21718530654907225,
      "num_tokens": 41916086.0,
      "step": 18780
    },
    {
      "entropy": 5.544496154785156,
      "epoch": 2.109732127814904,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.0931,
      "mean_token_accuracy": 0.21687526106834412,
      "num_tokens": 41925466.0,
      "step": 18785
    },
    {
      "entropy": 5.6007777690887455,
      "epoch": 2.110293704722862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20871905088424683,
      "num_tokens": 41936567.0,
      "step": 18790
    },
    {
      "entropy": 5.530610656738281,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045613741760565385,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21544190496206284,
      "num_tokens": 41947021.0,
      "step": 18795
    },
    {
      "entropy": 5.59056453704834,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21275424510240554,
      "num_tokens": 41957929.0,
      "step": 18800
    },
    {
      "entropy": 5.526122570037842,
      "epoch": 2.1119784354467344,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000456089773638076,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22123714685440063,
      "num_tokens": 41968677.0,
      "step": 18805
    },
    {
      "entropy": 5.514961910247803,
      "epoch": 2.112540012354692,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.1129,
      "mean_token_accuracy": 0.2133169412612915,
      "num_tokens": 41980330.0,
      "step": 18810
    },
    {
      "entropy": 5.541747522354126,
      "epoch": 2.1131015892626497,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004560421066066237,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21616625487804414,
      "num_tokens": 41991258.0,
      "step": 18815
    },
    {
      "entropy": 5.569202041625976,
      "epoch": 2.113663166170607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.0984,
      "mean_token_accuracy": 0.22309866845607756,
      "num_tokens": 42002679.0,
      "step": 18820
    },
    {
      "entropy": 5.50654149055481,
      "epoch": 2.1142247430785646,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045599441651736805,
      "loss": 4.9842,
      "mean_token_accuracy": 0.21473367214202882,
      "num_tokens": 42012845.0,
      "step": 18825
    },
    {
      "entropy": 5.605464124679566,
      "epoch": 2.1147863199865222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.0636,
      "mean_token_accuracy": 0.21679504960775375,
      "num_tokens": 42024461.0,
      "step": 18830
    },
    {
      "entropy": 5.586025381088257,
      "epoch": 2.11534789689448,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004559467033763829,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21257071644067765,
      "num_tokens": 42036996.0,
      "step": 18835
    },
    {
      "entropy": 5.6022552967071535,
      "epoch": 2.115909473802437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.1208,
      "mean_token_accuracy": 0.21309621632099152,
      "num_tokens": 42048731.0,
      "step": 18840
    },
    {
      "entropy": 5.568261814117432,
      "epoch": 2.1164710507103948,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21573513746261597,
      "num_tokens": 42060396.0,
      "step": 18845
    },
    {
      "entropy": 5.4232769966125485,
      "epoch": 2.1170326276183524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004558750904562063,
      "loss": 4.9544,
      "mean_token_accuracy": 0.21528733521699905,
      "num_tokens": 42072037.0,
      "step": 18850
    },
    {
      "entropy": 5.499650573730468,
      "epoch": 2.11759420452631,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004558512079635345,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2204629272222519,
      "num_tokens": 42082577.0,
      "step": 18855
    },
    {
      "entropy": 5.439575481414795,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004558273197124899,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22096109688282012,
      "num_tokens": 42093978.0,
      "step": 18860
    },
    {
      "entropy": 5.564894294738769,
      "epoch": 2.118717358342225,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.1191,
      "mean_token_accuracy": 0.20804581344127654,
      "num_tokens": 42105773.0,
      "step": 18865
    },
    {
      "entropy": 5.513844966888428,
      "epoch": 2.1192789352501826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.0428,
      "mean_token_accuracy": 0.21704984754323958,
      "num_tokens": 42116232.0,
      "step": 18870
    },
    {
      "entropy": 5.488379907608032,
      "epoch": 2.11984051215814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21343366652727128,
      "num_tokens": 42127813.0,
      "step": 18875
    },
    {
      "entropy": 5.512840270996094,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045573170913980094,
      "loss": 5.0001,
      "mean_token_accuracy": 0.21526493430137633,
      "num_tokens": 42138054.0,
      "step": 18880
    },
    {
      "entropy": 5.601580953598022,
      "epoch": 2.120963665974055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21122063547372819,
      "num_tokens": 42149575.0,
      "step": 18885
    },
    {
      "entropy": 5.538790035247803,
      "epoch": 2.121525242882013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21681202203035355,
      "num_tokens": 42160824.0,
      "step": 18890
    },
    {
      "entropy": 5.564800024032593,
      "epoch": 2.12208681978997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.1019,
      "mean_token_accuracy": 0.2111995115876198,
      "num_tokens": 42171088.0,
      "step": 18895
    },
    {
      "entropy": 5.5296838760375975,
      "epoch": 2.1226483966979277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.0312,
      "mean_token_accuracy": 0.21706053018569946,
      "num_tokens": 42182112.0,
      "step": 18900
    },
    {
      "entropy": 5.5649144649505615,
      "epoch": 2.1232099736058854,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004556120664518838,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22470432370901108,
      "num_tokens": 42193563.0,
      "step": 18905
    },
    {
      "entropy": 5.51471791267395,
      "epoch": 2.123771550513843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.0069,
      "mean_token_accuracy": 0.21602774858474733,
      "num_tokens": 42204891.0,
      "step": 18910
    },
    {
      "entropy": 5.581862545013427,
      "epoch": 2.1243331274218002,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2205270767211914,
      "num_tokens": 42217226.0,
      "step": 18915
    },
    {
      "entropy": 5.5973481178283695,
      "epoch": 2.124894704329758,
      "grad_norm": 0.96875,
      "learning_rate": 0.000455540211823939,
      "loss": 5.0413,
      "mean_token_accuracy": 0.21689275056123733,
      "num_tokens": 42228134.0,
      "step": 18920
    },
    {
      "entropy": 5.55786623954773,
      "epoch": 2.1254562812377156,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.0784,
      "mean_token_accuracy": 0.20524224638938904,
      "num_tokens": 42239868.0,
      "step": 18925
    },
    {
      "entropy": 5.608202743530273,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.1144,
      "mean_token_accuracy": 0.21472693830728531,
      "num_tokens": 42250297.0,
      "step": 18930
    },
    {
      "entropy": 5.48277096748352,
      "epoch": 2.1265794350536305,
      "grad_norm": 1.0,
      "learning_rate": 0.00045546830545976203,
      "loss": 4.7769,
      "mean_token_accuracy": 0.24096340984106063,
      "num_tokens": 42261263.0,
      "step": 18935
    },
    {
      "entropy": 5.5844676971435545,
      "epoch": 2.127141011961588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.0859,
      "mean_token_accuracy": 0.21803128272294997,
      "num_tokens": 42272357.0,
      "step": 18940
    },
    {
      "entropy": 5.602430725097657,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.1191,
      "mean_token_accuracy": 0.2149945691227913,
      "num_tokens": 42283630.0,
      "step": 18945
    },
    {
      "entropy": 5.5620057582855225,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.129,
      "mean_token_accuracy": 0.19933973401784896,
      "num_tokens": 42294981.0,
      "step": 18950
    },
    {
      "entropy": 5.566015291213989,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045537234986455964,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21830561459064485,
      "num_tokens": 42304877.0,
      "step": 18955
    },
    {
      "entropy": 5.540729093551636,
      "epoch": 2.1293873195934183,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.093,
      "mean_token_accuracy": 0.20755600780248643,
      "num_tokens": 42316325.0,
      "step": 18960
    },
    {
      "entropy": 5.514732599258423,
      "epoch": 2.129948896501376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21307916492223739,
      "num_tokens": 42327437.0,
      "step": 18965
    },
    {
      "entropy": 5.458932781219483,
      "epoch": 2.130510473409333,
      "grad_norm": 1.0,
      "learning_rate": 0.0004553003228626665,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21910125762224197,
      "num_tokens": 42337742.0,
      "step": 18970
    },
    {
      "entropy": 5.517786645889283,
      "epoch": 2.131072050317291,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2148967280983925,
      "num_tokens": 42348511.0,
      "step": 18975
    },
    {
      "entropy": 5.535186052322388,
      "epoch": 2.1316336272252485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004552522761559694,
      "loss": 5.011,
      "mean_token_accuracy": 0.2207840636372566,
      "num_tokens": 42358977.0,
      "step": 18980
    },
    {
      "entropy": 5.591801500320434,
      "epoch": 2.132195204133206,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21440659165382386,
      "num_tokens": 42371609.0,
      "step": 18985
    },
    {
      "entropy": 5.595178556442261,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22016884833574296,
      "num_tokens": 42383150.0,
      "step": 18990
    },
    {
      "entropy": 5.528708410263062,
      "epoch": 2.133318357949121,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004551801630530692,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21736479103565215,
      "num_tokens": 42395150.0,
      "step": 18995
    },
    {
      "entropy": 5.554240369796753,
      "epoch": 2.1338799348570787,
      "grad_norm": 1.0,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21568124890327453,
      "num_tokens": 42406348.0,
      "step": 19000
    },
    {
      "entropy": 5.587265539169311,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.0596,
      "mean_token_accuracy": 0.20822735428810119,
      "num_tokens": 42417183.0,
      "step": 19005
    },
    {
      "entropy": 5.5621730327606205,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045510799831711366,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22623004168272018,
      "num_tokens": 42427805.0,
      "step": 19010
    },
    {
      "entropy": 5.555837345123291,
      "epoch": 2.1355646655809513,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045508393193469,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21316086053848265,
      "num_tokens": 42439521.0,
      "step": 19015
    },
    {
      "entropy": 5.6286335468292235,
      "epoch": 2.136126242488909,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.1156,
      "mean_token_accuracy": 0.21682822108268737,
      "num_tokens": 42451201.0,
      "step": 19020
    },
    {
      "entropy": 5.546107482910156,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004550357819687825,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21569205522537233,
      "num_tokens": 42462315.0,
      "step": 19025
    },
    {
      "entropy": 5.659907484054566,
      "epoch": 2.137249396304824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.1816,
      "mean_token_accuracy": 0.20507620722055436,
      "num_tokens": 42473993.0,
      "step": 19030
    },
    {
      "entropy": 5.676066493988037,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20797190368175505,
      "num_tokens": 42484307.0,
      "step": 19035
    },
    {
      "entropy": 5.562943077087402,
      "epoch": 2.138372550120739,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.0749,
      "mean_token_accuracy": 0.2140333905816078,
      "num_tokens": 42497305.0,
      "step": 19040
    },
    {
      "entropy": 5.577332067489624,
      "epoch": 2.138934127028697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.0945,
      "mean_token_accuracy": 0.21101474463939668,
      "num_tokens": 42508307.0,
      "step": 19045
    },
    {
      "entropy": 5.557398653030395,
      "epoch": 2.139495703936654,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.1608,
      "mean_token_accuracy": 0.20556483119726182,
      "num_tokens": 42519946.0,
      "step": 19050
    },
    {
      "entropy": 5.5132886409759525,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045489119451778645,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2191242054104805,
      "num_tokens": 42530737.0,
      "step": 19055
    },
    {
      "entropy": 5.436007833480835,
      "epoch": 2.1406188577525693,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004548670765574908,
      "loss": 4.9244,
      "mean_token_accuracy": 0.2207363173365593,
      "num_tokens": 42542184.0,
      "step": 19060
    },
    {
      "entropy": 5.531171607971191,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21014407873153687,
      "num_tokens": 42553816.0,
      "step": 19065
    },
    {
      "entropy": 5.611481809616089,
      "epoch": 2.141742011568484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.0705,
      "mean_token_accuracy": 0.21818750500679016,
      "num_tokens": 42564685.0,
      "step": 19070
    },
    {
      "entropy": 5.541925716400146,
      "epoch": 2.142303588476442,
      "grad_norm": 1.0,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21246128678321838,
      "num_tokens": 42574864.0,
      "step": 19075
    },
    {
      "entropy": 5.556358337402344,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004547705474536152,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2090007930994034,
      "num_tokens": 42585214.0,
      "step": 19080
    },
    {
      "entropy": 5.579411315917969,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.117,
      "mean_token_accuracy": 0.21023037135601044,
      "num_tokens": 42596017.0,
      "step": 19085
    },
    {
      "entropy": 5.55363974571228,
      "epoch": 2.1439883192003144,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21124672144651413,
      "num_tokens": 42607903.0,
      "step": 19090
    },
    {
      "entropy": 5.5491039752960205,
      "epoch": 2.144549896108272,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2154656246304512,
      "num_tokens": 42618690.0,
      "step": 19095
    },
    {
      "entropy": 5.5145446300506595,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045467392676632016,
      "loss": 4.9787,
      "mean_token_accuracy": 0.21744531244039536,
      "num_tokens": 42629655.0,
      "step": 19100
    },
    {
      "entropy": 5.556855630874634,
      "epoch": 2.145673049924187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.109,
      "mean_token_accuracy": 0.2129283145070076,
      "num_tokens": 42641214.0,
      "step": 19105
    },
    {
      "entropy": 5.636620378494262,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.1336,
      "mean_token_accuracy": 0.2068973794579506,
      "num_tokens": 42653829.0,
      "step": 19110
    },
    {
      "entropy": 5.6331702709198,
      "epoch": 2.1467962037401023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20961866676807403,
      "num_tokens": 42665245.0,
      "step": 19115
    },
    {
      "entropy": 5.522520589828491,
      "epoch": 2.14735778064806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004545772145448284,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2163372442126274,
      "num_tokens": 42676409.0,
      "step": 19120
    },
    {
      "entropy": 5.487315654754639,
      "epoch": 2.147919357556017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045455302219301,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22381097227334976,
      "num_tokens": 42686586.0,
      "step": 19125
    },
    {
      "entropy": 5.451073741912841,
      "epoch": 2.148480934463975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545288241241538,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21838153302669525,
      "num_tokens": 42696597.0,
      "step": 19130
    },
    {
      "entropy": 5.600845193862915,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.1248,
      "mean_token_accuracy": 0.21650497615337372,
      "num_tokens": 42707447.0,
      "step": 19135
    },
    {
      "entropy": 5.691606569290161,
      "epoch": 2.14960408827989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20294069796800612,
      "num_tokens": 42717533.0,
      "step": 19140
    },
    {
      "entropy": 5.590234422683716,
      "epoch": 2.1501656651878474,
      "grad_norm": 0.875,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.1273,
      "mean_token_accuracy": 0.2134213551878929,
      "num_tokens": 42730563.0,
      "step": 19145
    },
    {
      "entropy": 5.534512186050415,
      "epoch": 2.150727242095805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.0219,
      "mean_token_accuracy": 0.219589464366436,
      "num_tokens": 42740301.0,
      "step": 19150
    },
    {
      "entropy": 5.571725654602051,
      "epoch": 2.1512888190037627,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21233469992876053,
      "num_tokens": 42752296.0,
      "step": 19155
    },
    {
      "entropy": 5.519036769866943,
      "epoch": 2.15185039591172,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045438351569638055,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21756040602922438,
      "num_tokens": 42762530.0,
      "step": 19160
    },
    {
      "entropy": 5.535397243499756,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045435927762984277,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21824690848588943,
      "num_tokens": 42773059.0,
      "step": 19165
    },
    {
      "entropy": 5.508025407791138,
      "epoch": 2.152973549727635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000454335033852436,
      "loss": 5.0009,
      "mean_token_accuracy": 0.220305697619915,
      "num_tokens": 42783832.0,
      "step": 19170
    },
    {
      "entropy": 5.541471338272094,
      "epoch": 2.153535126635593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.1064,
      "mean_token_accuracy": 0.21402810961008073,
      "num_tokens": 42794565.0,
      "step": 19175
    },
    {
      "entropy": 5.6409523487091064,
      "epoch": 2.15409670354355,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20821309238672256,
      "num_tokens": 42805536.0,
      "step": 19180
    },
    {
      "entropy": 5.567105770111084,
      "epoch": 2.1546582804515078,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.0778,
      "mean_token_accuracy": 0.21090813726186752,
      "num_tokens": 42816718.0,
      "step": 19185
    },
    {
      "entropy": 5.568763017654419,
      "epoch": 2.1552198573594654,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21393198817968367,
      "num_tokens": 42830183.0,
      "step": 19190
    },
    {
      "entropy": 5.493221712112427,
      "epoch": 2.155781434267423,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2201959878206253,
      "num_tokens": 42841241.0,
      "step": 19195
    },
    {
      "entropy": 5.705216693878174,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.1378,
      "mean_token_accuracy": 0.21053665280342101,
      "num_tokens": 42852775.0,
      "step": 19200
    },
    {
      "entropy": 5.7534722805023195,
      "epoch": 2.156904588083338,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.2413,
      "mean_token_accuracy": 0.20581746548414231,
      "num_tokens": 42863947.0,
      "step": 19205
    },
    {
      "entropy": 5.560336399078369,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.0878,
      "mean_token_accuracy": 0.2190374180674553,
      "num_tokens": 42874794.0,
      "step": 19210
    },
    {
      "entropy": 5.5485443592071535,
      "epoch": 2.1580277418992533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.0586,
      "mean_token_accuracy": 0.2116953730583191,
      "num_tokens": 42886038.0,
      "step": 19215
    },
    {
      "entropy": 5.562234163284302,
      "epoch": 2.1585893188072105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21083592474460602,
      "num_tokens": 42897243.0,
      "step": 19220
    },
    {
      "entropy": 5.520388793945313,
      "epoch": 2.159150895715168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004540679756045125,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2148055240511894,
      "num_tokens": 42908802.0,
      "step": 19225
    },
    {
      "entropy": 5.613173294067383,
      "epoch": 2.159712472623126,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2058844208717346,
      "num_tokens": 42921156.0,
      "step": 19230
    },
    {
      "entropy": 5.493680667877197,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045401934540854937,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22285173535346986,
      "num_tokens": 42931336.0,
      "step": 19235
    },
    {
      "entropy": 5.562589693069458,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.0232,
      "mean_token_accuracy": 0.22291235774755477,
      "num_tokens": 42942312.0,
      "step": 19240
    },
    {
      "entropy": 5.588434553146362,
      "epoch": 2.1613972033469984,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.0831,
      "mean_token_accuracy": 0.21336940675973892,
      "num_tokens": 42953429.0,
      "step": 19245
    },
    {
      "entropy": 5.486431550979614,
      "epoch": 2.161958780254956,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000453946357366201,
      "loss": 4.961,
      "mean_token_accuracy": 0.2176554471254349,
      "num_tokens": 42963354.0,
      "step": 19250
    },
    {
      "entropy": 5.627550268173218,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.1143,
      "mean_token_accuracy": 0.21519964039325715,
      "num_tokens": 42974643.0,
      "step": 19255
    },
    {
      "entropy": 5.587266206741333,
      "epoch": 2.163081934070871,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.097,
      "mean_token_accuracy": 0.21105421036481858,
      "num_tokens": 42985821.0,
      "step": 19260
    },
    {
      "entropy": 5.626778173446655,
      "epoch": 2.1636435109788286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2115868717432022,
      "num_tokens": 42998258.0,
      "step": 19265
    },
    {
      "entropy": 5.553482151031494,
      "epoch": 2.1642050878867862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21404251158237458,
      "num_tokens": 43009117.0,
      "step": 19270
    },
    {
      "entropy": 5.5099584579467775,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.0312,
      "mean_token_accuracy": 0.21382674872875213,
      "num_tokens": 43020437.0,
      "step": 19275
    },
    {
      "entropy": 5.544236707687378,
      "epoch": 2.165328241702701,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2127535805106163,
      "num_tokens": 43031722.0,
      "step": 19280
    },
    {
      "entropy": 5.548121595382691,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045377585254998743,
      "loss": 4.9973,
      "mean_token_accuracy": 0.224325692653656,
      "num_tokens": 43042957.0,
      "step": 19285
    },
    {
      "entropy": 5.5446085929870605,
      "epoch": 2.1664513955186164,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21341798454523087,
      "num_tokens": 43054086.0,
      "step": 19290
    },
    {
      "entropy": 5.636159610748291,
      "epoch": 2.1670129724265736,
      "grad_norm": 1.0,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.1674,
      "mean_token_accuracy": 0.20577600300312043,
      "num_tokens": 43066259.0,
      "step": 19295
    },
    {
      "entropy": 5.499989700317383,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045370269365742143,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22147487699985505,
      "num_tokens": 43076837.0,
      "step": 19300
    },
    {
      "entropy": 5.521359395980835,
      "epoch": 2.168136126242489,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045367829597893776,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22152996212244033,
      "num_tokens": 43087802.0,
      "step": 19305
    },
    {
      "entropy": 5.56063232421875,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045365389261126793,
      "loss": 4.9734,
      "mean_token_accuracy": 0.21772230565547943,
      "num_tokens": 43097562.0,
      "step": 19310
    },
    {
      "entropy": 5.483265161514282,
      "epoch": 2.169259280058404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045362948355518883,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22191784679889678,
      "num_tokens": 43107973.0,
      "step": 19315
    },
    {
      "entropy": 5.491032075881958,
      "epoch": 2.1698208569663615,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004536050688114777,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2217192232608795,
      "num_tokens": 43118923.0,
      "step": 19320
    },
    {
      "entropy": 5.526493215560913,
      "epoch": 2.170382433874319,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00045358064838091187,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21835330724716187,
      "num_tokens": 43130904.0,
      "step": 19325
    },
    {
      "entropy": 5.552015018463135,
      "epoch": 2.170944010782277,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045355622226426903,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20952733904123305,
      "num_tokens": 43142293.0,
      "step": 19330
    },
    {
      "entropy": 5.56937689781189,
      "epoch": 2.171505587690234,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.0669,
      "mean_token_accuracy": 0.2183936804533005,
      "num_tokens": 43154340.0,
      "step": 19335
    },
    {
      "entropy": 5.563002824783325,
      "epoch": 2.1720671645981917,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000453507352975863,
      "loss": 5.0159,
      "mean_token_accuracy": 0.22311041951179506,
      "num_tokens": 43165924.0,
      "step": 19340
    },
    {
      "entropy": 5.5236307144165036,
      "epoch": 2.1726287415061494,
      "grad_norm": 1.0,
      "learning_rate": 0.00045348290980565595,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21846989989280702,
      "num_tokens": 43177394.0,
      "step": 19345
    },
    {
      "entropy": 5.549494791030884,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045345846095248376,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21963704973459244,
      "num_tokens": 43188043.0,
      "step": 19350
    },
    {
      "entropy": 5.593376636505127,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.1555,
      "mean_token_accuracy": 0.20760195851325988,
      "num_tokens": 43199466.0,
      "step": 19355
    },
    {
      "entropy": 5.459299468994141,
      "epoch": 2.174313472230022,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045340954620035827,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22089126259088515,
      "num_tokens": 43210673.0,
      "step": 19360
    },
    {
      "entropy": 5.564438819885254,
      "epoch": 2.1748750491379796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21280408799648284,
      "num_tokens": 43221263.0,
      "step": 19365
    },
    {
      "entropy": 5.599343109130859,
      "epoch": 2.175436626045937,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.084,
      "mean_token_accuracy": 0.22124669700860977,
      "num_tokens": 43232590.0,
      "step": 19370
    },
    {
      "entropy": 5.652501201629638,
      "epoch": 2.1759982029538945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.0986,
      "mean_token_accuracy": 0.21868456304073333,
      "num_tokens": 43243693.0,
      "step": 19375
    },
    {
      "entropy": 5.576356267929077,
      "epoch": 2.176559779861852,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045331164853479084,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21763329654932023,
      "num_tokens": 43254526.0,
      "step": 19380
    },
    {
      "entropy": 5.696829843521118,
      "epoch": 2.17712135676981,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.1249,
      "mean_token_accuracy": 0.21074826270341873,
      "num_tokens": 43264490.0,
      "step": 19385
    },
    {
      "entropy": 5.560895013809204,
      "epoch": 2.177682933677767,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045326266563381734,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21854330748319625,
      "num_tokens": 43275012.0,
      "step": 19390
    },
    {
      "entropy": 5.532060384750366,
      "epoch": 2.1782445105857247,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045323816566901205,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22072146087884903,
      "num_tokens": 43285194.0,
      "step": 19395
    },
    {
      "entropy": 5.564247035980225,
      "epoch": 2.1788060874936823,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.0674,
      "mean_token_accuracy": 0.2197772666811943,
      "num_tokens": 43298159.0,
      "step": 19400
    },
    {
      "entropy": 5.5877601146698,
      "epoch": 2.17936766440164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.1191,
      "mean_token_accuracy": 0.20598918348550796,
      "num_tokens": 43309664.0,
      "step": 19405
    },
    {
      "entropy": 5.552412605285644,
      "epoch": 2.179929241309597,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21478799730539322,
      "num_tokens": 43320399.0,
      "step": 19410
    },
    {
      "entropy": 5.481645393371582,
      "epoch": 2.180490818217555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045314010906587096,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2147519186139107,
      "num_tokens": 43331510.0,
      "step": 19415
    },
    {
      "entropy": 5.632701253890991,
      "epoch": 2.1810523951255125,
      "grad_norm": 1.0,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.112,
      "mean_token_accuracy": 0.21555982679128646,
      "num_tokens": 43342844.0,
      "step": 19420
    },
    {
      "entropy": 5.5105979442596436,
      "epoch": 2.18161397203347,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004530910467288769,
      "loss": 4.9698,
      "mean_token_accuracy": 0.2150255858898163,
      "num_tokens": 43352556.0,
      "step": 19425
    },
    {
      "entropy": 5.463360595703125,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045306650705425727,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22259082943201064,
      "num_tokens": 43363357.0,
      "step": 19430
    },
    {
      "entropy": 5.581704235076904,
      "epoch": 2.182737125849385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.0445,
      "mean_token_accuracy": 0.2128152370452881,
      "num_tokens": 43374697.0,
      "step": 19435
    },
    {
      "entropy": 5.679049348831176,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.0,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.1167,
      "mean_token_accuracy": 0.21011421382427214,
      "num_tokens": 43386826.0,
      "step": 19440
    },
    {
      "entropy": 5.557777023315429,
      "epoch": 2.1838602796653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.0722,
      "mean_token_accuracy": 0.21580599695444108,
      "num_tokens": 43398352.0,
      "step": 19445
    },
    {
      "entropy": 5.559335851669312,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.021,
      "mean_token_accuracy": 0.21774175614118577,
      "num_tokens": 43409509.0,
      "step": 19450
    },
    {
      "entropy": 5.538899946212768,
      "epoch": 2.1849834334812153,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045294372365119276,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21488408148288726,
      "num_tokens": 43420195.0,
      "step": 19455
    },
    {
      "entropy": 5.558478116989136,
      "epoch": 2.185545010389173,
      "grad_norm": 1.0,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.077,
      "mean_token_accuracy": 0.21415614038705827,
      "num_tokens": 43430599.0,
      "step": 19460
    },
    {
      "entropy": 5.457676935195923,
      "epoch": 2.18610658729713,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004528945706239122,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21648177355527878,
      "num_tokens": 43442212.0,
      "step": 19465
    },
    {
      "entropy": 5.519367122650147,
      "epoch": 2.186668164205088,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045286998561353236,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21573808491230012,
      "num_tokens": 43452631.0,
      "step": 19470
    },
    {
      "entropy": 5.533348321914673,
      "epoch": 2.1872297411130455,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.0312,
      "mean_token_accuracy": 0.2147737681865692,
      "num_tokens": 43463323.0,
      "step": 19475
    },
    {
      "entropy": 5.531951379776001,
      "epoch": 2.187791318021003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20848031640052794,
      "num_tokens": 43474451.0,
      "step": 19480
    },
    {
      "entropy": 5.501145601272583,
      "epoch": 2.1883528949289603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045279619660482925,
      "loss": 4.937,
      "mean_token_accuracy": 0.22151163816452027,
      "num_tokens": 43485948.0,
      "step": 19485
    },
    {
      "entropy": 5.605907297134399,
      "epoch": 2.188914471836918,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.043,
      "mean_token_accuracy": 0.21904576420783997,
      "num_tokens": 43498955.0,
      "step": 19490
    },
    {
      "entropy": 5.578834342956543,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2248677760362625,
      "num_tokens": 43509416.0,
      "step": 19495
    },
    {
      "entropy": 5.527827739715576,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21914236843585969,
      "num_tokens": 43520387.0,
      "step": 19500
    },
    {
      "entropy": 5.48080792427063,
      "epoch": 2.1905992025607905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.023,
      "mean_token_accuracy": 0.21255133748054506,
      "num_tokens": 43532006.0,
      "step": 19505
    },
    {
      "entropy": 5.4930257320404055,
      "epoch": 2.191160779468748,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004526731017121111,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21788736432790756,
      "num_tokens": 43542303.0,
      "step": 19510
    },
    {
      "entropy": 5.619324398040772,
      "epoch": 2.191722356376706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.1106,
      "mean_token_accuracy": 0.21225156486034394,
      "num_tokens": 43553711.0,
      "step": 19515
    },
    {
      "entropy": 5.499596118927002,
      "epoch": 2.1922839332846635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045262382414927235,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2154448002576828,
      "num_tokens": 43564488.0,
      "step": 19520
    },
    {
      "entropy": 5.531378984451294,
      "epoch": 2.1928455101926208,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045259917688404336,
      "loss": 4.99,
      "mean_token_accuracy": 0.21906263679265975,
      "num_tokens": 43575414.0,
      "step": 19525
    },
    {
      "entropy": 5.505902719497681,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004525745239639877,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21381391137838363,
      "num_tokens": 43586789.0,
      "step": 19530
    },
    {
      "entropy": 5.528029203414917,
      "epoch": 2.193968664008536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045254986538989045,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2289793998003006,
      "num_tokens": 43597423.0,
      "step": 19535
    },
    {
      "entropy": 5.565798234939575,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004525252011625366,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21448690593242645,
      "num_tokens": 43606703.0,
      "step": 19540
    },
    {
      "entropy": 5.628390645980835,
      "epoch": 2.195091817824451,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21392180174589157,
      "num_tokens": 43617869.0,
      "step": 19545
    },
    {
      "entropy": 5.616559171676636,
      "epoch": 2.1956533947324086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.1807,
      "mean_token_accuracy": 0.21235133409500123,
      "num_tokens": 43629102.0,
      "step": 19550
    },
    {
      "entropy": 5.569294738769531,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004524511745687895,
      "loss": 5.0739,
      "mean_token_accuracy": 0.21845516562461853,
      "num_tokens": 43640421.0,
      "step": 19555
    },
    {
      "entropy": 5.6019055366516115,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004524264877362643,
      "loss": 4.9849,
      "mean_token_accuracy": 0.2204217478632927,
      "num_tokens": 43650516.0,
      "step": 19560
    },
    {
      "entropy": 5.539620018005371,
      "epoch": 2.197338125456281,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045240179525441094,
      "loss": 5.003,
      "mean_token_accuracy": 0.22057234346866608,
      "num_tokens": 43661439.0,
      "step": 19565
    },
    {
      "entropy": 5.51871337890625,
      "epoch": 2.197899702364239,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.0833,
      "mean_token_accuracy": 0.21552836745977402,
      "num_tokens": 43672137.0,
      "step": 19570
    },
    {
      "entropy": 5.569383716583252,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21173907071352005,
      "num_tokens": 43683612.0,
      "step": 19575
    },
    {
      "entropy": 5.562359428405761,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2191691905260086,
      "num_tokens": 43694331.0,
      "step": 19580
    },
    {
      "entropy": 5.589565658569336,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.0957,
      "mean_token_accuracy": 0.21537536531686782,
      "num_tokens": 43706572.0,
      "step": 19585
    },
    {
      "entropy": 5.521126985549927,
      "epoch": 2.200146009996069,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045227824813274517,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2268496960401535,
      "num_tokens": 43718472.0,
      "step": 19590
    },
    {
      "entropy": 5.606297779083252,
      "epoch": 2.2007075869040267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.0888,
      "mean_token_accuracy": 0.21538886725902556,
      "num_tokens": 43729642.0,
      "step": 19595
    },
    {
      "entropy": 5.545511150360108,
      "epoch": 2.201269163811984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.0989,
      "mean_token_accuracy": 0.20666202455759047,
      "num_tokens": 43739846.0,
      "step": 19600
    },
    {
      "entropy": 5.5078990936279295,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045220405211815327,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21947708874940872,
      "num_tokens": 43750265.0,
      "step": 19605
    },
    {
      "entropy": 5.600264549255371,
      "epoch": 2.202392317627899,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.0438,
      "mean_token_accuracy": 0.20963021963834763,
      "num_tokens": 43761245.0,
      "step": 19610
    },
    {
      "entropy": 5.60635666847229,
      "epoch": 2.202953894535857,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21567571759223939,
      "num_tokens": 43772677.0,
      "step": 19615
    },
    {
      "entropy": 5.578537178039551,
      "epoch": 2.203515471443814,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.095,
      "mean_token_accuracy": 0.21014566719532013,
      "num_tokens": 43783128.0,
      "step": 19620
    },
    {
      "entropy": 5.493904685974121,
      "epoch": 2.2040770483517718,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2178509220480919,
      "num_tokens": 43795323.0,
      "step": 19625
    },
    {
      "entropy": 5.681013822555542,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.1113,
      "mean_token_accuracy": 0.21259612888097762,
      "num_tokens": 43806917.0,
      "step": 19630
    },
    {
      "entropy": 5.677973651885987,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.1008,
      "mean_token_accuracy": 0.21276296228170394,
      "num_tokens": 43818963.0,
      "step": 19635
    },
    {
      "entropy": 5.491607046127319,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004520307306422844,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2125406950712204,
      "num_tokens": 43830163.0,
      "step": 19640
    },
    {
      "entropy": 5.534350395202637,
      "epoch": 2.206323355983602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045200594787822795,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21788241416215898,
      "num_tokens": 43840434.0,
      "step": 19645
    },
    {
      "entropy": 5.420539426803589,
      "epoch": 2.2068849328915596,
      "grad_norm": 1.0,
      "learning_rate": 0.0004519811594782335,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22413428574800492,
      "num_tokens": 43851086.0,
      "step": 19650
    },
    {
      "entropy": 5.5392430305480955,
      "epoch": 2.207446509799517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20897450298070908,
      "num_tokens": 43862834.0,
      "step": 19655
    },
    {
      "entropy": 5.554836797714233,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20336219370365144,
      "num_tokens": 43874868.0,
      "step": 19660
    },
    {
      "entropy": 5.583745431900025,
      "epoch": 2.208569663615432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.0413,
      "mean_token_accuracy": 0.22144425362348558,
      "num_tokens": 43886175.0,
      "step": 19665
    },
    {
      "entropy": 5.520596837997436,
      "epoch": 2.20913124052339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045188194953466363,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2211938664317131,
      "num_tokens": 43896489.0,
      "step": 19670
    },
    {
      "entropy": 5.502599287033081,
      "epoch": 2.209692817431347,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045185713296682175,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21735432893037795,
      "num_tokens": 43907759.0,
      "step": 19675
    },
    {
      "entropy": 5.527173852920532,
      "epoch": 2.2102543943393047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2156906768679619,
      "num_tokens": 43918136.0,
      "step": 19680
    },
    {
      "entropy": 5.584428501129151,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21055467575788497,
      "num_tokens": 43929662.0,
      "step": 19685
    },
    {
      "entropy": 5.539862632751465,
      "epoch": 2.21137754815522,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.0301,
      "mean_token_accuracy": 0.223611319065094,
      "num_tokens": 43939599.0,
      "step": 19690
    },
    {
      "entropy": 5.543952894210816,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045175781039136133,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21646052747964858,
      "num_tokens": 43949728.0,
      "step": 19695
    },
    {
      "entropy": 5.61400351524353,
      "epoch": 2.212500701971135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21800603121519088,
      "num_tokens": 43961147.0,
      "step": 19700
    },
    {
      "entropy": 5.580230760574341,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.0437,
      "mean_token_accuracy": 0.22525279968976974,
      "num_tokens": 43972541.0,
      "step": 19705
    },
    {
      "entropy": 5.504610633850097,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.125,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.0654,
      "mean_token_accuracy": 0.2155996099114418,
      "num_tokens": 43983540.0,
      "step": 19710
    },
    {
      "entropy": 5.502172136306763,
      "epoch": 2.2141854326950074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004516583977673087,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22251654267311097,
      "num_tokens": 43994255.0,
      "step": 19715
    },
    {
      "entropy": 5.650884962081909,
      "epoch": 2.214747009602965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.1414,
      "mean_token_accuracy": 0.21258847266435624,
      "num_tokens": 44004364.0,
      "step": 19720
    },
    {
      "entropy": 5.618717002868652,
      "epoch": 2.2153085865109228,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.115,
      "mean_token_accuracy": 0.21166601926088333,
      "num_tokens": 44015487.0,
      "step": 19725
    },
    {
      "entropy": 5.669394874572754,
      "epoch": 2.21587016341888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21873199045658112,
      "num_tokens": 44026974.0,
      "step": 19730
    },
    {
      "entropy": 5.574199295043945,
      "epoch": 2.2164317403268377,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.018,
      "mean_token_accuracy": 0.21676014959812165,
      "num_tokens": 44038718.0,
      "step": 19735
    },
    {
      "entropy": 5.497802829742431,
      "epoch": 2.2169933172347953,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004515340054335696,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2205212041735649,
      "num_tokens": 44048821.0,
      "step": 19740
    },
    {
      "entropy": 5.555745553970337,
      "epoch": 2.217554894142753,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.1173,
      "mean_token_accuracy": 0.2080742433667183,
      "num_tokens": 44060973.0,
      "step": 19745
    },
    {
      "entropy": 5.679597234725952,
      "epoch": 2.21811647105071,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.0908,
      "mean_token_accuracy": 0.2215743288397789,
      "num_tokens": 44073595.0,
      "step": 19750
    },
    {
      "entropy": 5.513116073608399,
      "epoch": 2.218678047958668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.0124,
      "mean_token_accuracy": 0.2122385695576668,
      "num_tokens": 44084860.0,
      "step": 19755
    },
    {
      "entropy": 5.53724160194397,
      "epoch": 2.2192396248666255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20838320702314378,
      "num_tokens": 44096026.0,
      "step": 19760
    },
    {
      "entropy": 5.570093107223511,
      "epoch": 2.219801201774583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2134440541267395,
      "num_tokens": 44107981.0,
      "step": 19765
    },
    {
      "entropy": 5.566717290878296,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.0648,
      "mean_token_accuracy": 0.21377577483654023,
      "num_tokens": 44119589.0,
      "step": 19770
    },
    {
      "entropy": 5.524425268173218,
      "epoch": 2.220924355590498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21102766394615174,
      "num_tokens": 44131094.0,
      "step": 19775
    },
    {
      "entropy": 5.594834566116333,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.0693,
      "mean_token_accuracy": 0.2070179060101509,
      "num_tokens": 44141348.0,
      "step": 19780
    },
    {
      "entropy": 5.679374361038208,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.1939,
      "mean_token_accuracy": 0.20830076187849045,
      "num_tokens": 44152085.0,
      "step": 19785
    },
    {
      "entropy": 5.55826849937439,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045128479929690756,
      "loss": 4.9678,
      "mean_token_accuracy": 0.22939949929714204,
      "num_tokens": 44163518.0,
      "step": 19790
    },
    {
      "entropy": 5.489483118057251,
      "epoch": 2.2231706632223283,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2088601678609848,
      "num_tokens": 44175513.0,
      "step": 19795
    },
    {
      "entropy": 5.599001550674439,
      "epoch": 2.223732240130286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21453016698360444,
      "num_tokens": 44186660.0,
      "step": 19800
    },
    {
      "entropy": 5.595495986938476,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.0973,
      "mean_token_accuracy": 0.21123612821102142,
      "num_tokens": 44197613.0,
      "step": 19805
    },
    {
      "entropy": 5.518256187438965,
      "epoch": 2.224855393946201,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.0299,
      "mean_token_accuracy": 0.20505275428295136,
      "num_tokens": 44209469.0,
      "step": 19810
    },
    {
      "entropy": 5.564903926849365,
      "epoch": 2.2254169708541585,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.045,
      "mean_token_accuracy": 0.21316471844911575,
      "num_tokens": 44220672.0,
      "step": 19815
    },
    {
      "entropy": 5.59509129524231,
      "epoch": 2.225978547762116,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20989060401916504,
      "num_tokens": 44232867.0,
      "step": 19820
    },
    {
      "entropy": 5.564734745025635,
      "epoch": 2.2265401246700733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21301808953285217,
      "num_tokens": 44245567.0,
      "step": 19825
    },
    {
      "entropy": 5.554443502426148,
      "epoch": 2.227101701578031,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045108503021523737,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22407880425453186,
      "num_tokens": 44257449.0,
      "step": 19830
    },
    {
      "entropy": 5.558816051483154,
      "epoch": 2.2276632784859887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045106003384190604,
      "loss": 5.1021,
      "mean_token_accuracy": 0.21169852167367936,
      "num_tokens": 44268890.0,
      "step": 19835
    },
    {
      "entropy": 5.58437557220459,
      "epoch": 2.2282248553939463,
      "grad_norm": 1.0,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20712777078151703,
      "num_tokens": 44280719.0,
      "step": 19840
    },
    {
      "entropy": 5.666105890274048,
      "epoch": 2.2287864323019035,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.153,
      "mean_token_accuracy": 0.20440329909324645,
      "num_tokens": 44292567.0,
      "step": 19845
    },
    {
      "entropy": 5.609812927246094,
      "epoch": 2.229348009209861,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.0787,
      "mean_token_accuracy": 0.21372351795434952,
      "num_tokens": 44304177.0,
      "step": 19850
    },
    {
      "entropy": 5.524691820144653,
      "epoch": 2.229909586117819,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045095999229834437,
      "loss": 4.971,
      "mean_token_accuracy": 0.22028746008872985,
      "num_tokens": 44315142.0,
      "step": 19855
    },
    {
      "entropy": 5.5317676067352295,
      "epoch": 2.2304711630257765,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.1498,
      "mean_token_accuracy": 0.21328200548887252,
      "num_tokens": 44327457.0,
      "step": 19860
    },
    {
      "entropy": 5.6635359764099125,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.1357,
      "mean_token_accuracy": 0.20960438549518584,
      "num_tokens": 44338978.0,
      "step": 19865
    },
    {
      "entropy": 5.65834321975708,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.1262,
      "mean_token_accuracy": 0.21538204401731492,
      "num_tokens": 44349121.0,
      "step": 19870
    },
    {
      "entropy": 5.534888124465942,
      "epoch": 2.232155893749649,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004508598611126359,
      "loss": 5.0131,
      "mean_token_accuracy": 0.22554313987493516,
      "num_tokens": 44360753.0,
      "step": 19875
    },
    {
      "entropy": 5.631625175476074,
      "epoch": 2.2327174706576067,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2180103227496147,
      "num_tokens": 44371977.0,
      "step": 19880
    },
    {
      "entropy": 5.522323703765869,
      "epoch": 2.233279047565564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045080976191991587,
      "loss": 4.9825,
      "mean_token_accuracy": 0.2187456801533699,
      "num_tokens": 44383077.0,
      "step": 19885
    },
    {
      "entropy": 5.55124192237854,
      "epoch": 2.2338406244735216,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004507847039263805,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2134784057736397,
      "num_tokens": 44394827.0,
      "step": 19890
    },
    {
      "entropy": 5.509810972213745,
      "epoch": 2.2344022013814793,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000450759640335792,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2194083422422409,
      "num_tokens": 44406594.0,
      "step": 19895
    },
    {
      "entropy": 5.508883047103882,
      "epoch": 2.234963778289437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004507345711489484,
      "loss": 4.9642,
      "mean_token_accuracy": 0.21768635809421538,
      "num_tokens": 44417842.0,
      "step": 19900
    },
    {
      "entropy": 5.553750801086426,
      "epoch": 2.235525355197394,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21233983784914018,
      "num_tokens": 44428234.0,
      "step": 19905
    },
    {
      "entropy": 5.489419603347779,
      "epoch": 2.236086932105352,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004506844159896888,
      "loss": 4.9863,
      "mean_token_accuracy": 0.2169657826423645,
      "num_tokens": 44437504.0,
      "step": 19910
    },
    {
      "entropy": 5.595154523849487,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21796657592058183,
      "num_tokens": 44447873.0,
      "step": 19915
    },
    {
      "entropy": 5.596170043945312,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.0992,
      "mean_token_accuracy": 0.21654627472162247,
      "num_tokens": 44458990.0,
      "step": 19920
    },
    {
      "entropy": 5.543861341476441,
      "epoch": 2.2377716628292244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21756665259599686,
      "num_tokens": 44469374.0,
      "step": 19925
    },
    {
      "entropy": 5.53139362335205,
      "epoch": 2.238333239737182,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21836907416582108,
      "num_tokens": 44479629.0,
      "step": 19930
    },
    {
      "entropy": 5.586995220184326,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.0811,
      "mean_token_accuracy": 0.21098481565713884,
      "num_tokens": 44490603.0,
      "step": 19935
    },
    {
      "entropy": 5.484796142578125,
      "epoch": 2.239456393553097,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004505338162848382,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2251945123076439,
      "num_tokens": 44501527.0,
      "step": 19940
    },
    {
      "entropy": 5.6269646167755125,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.1506,
      "mean_token_accuracy": 0.21024064123630523,
      "num_tokens": 44512619.0,
      "step": 19945
    },
    {
      "entropy": 5.679042816162109,
      "epoch": 2.240579547369012,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2070343792438507,
      "num_tokens": 44524639.0,
      "step": 19950
    },
    {
      "entropy": 5.552363443374634,
      "epoch": 2.24114112427697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22128241509199142,
      "num_tokens": 44535013.0,
      "step": 19955
    },
    {
      "entropy": 5.6789312839508055,
      "epoch": 2.241702701184927,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.1283,
      "mean_token_accuracy": 0.20774058103561402,
      "num_tokens": 44546838.0,
      "step": 19960
    },
    {
      "entropy": 5.541607093811035,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21803765296936034,
      "num_tokens": 44557984.0,
      "step": 19965
    },
    {
      "entropy": 5.531014060974121,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.125,
      "learning_rate": 0.00045038301537367374,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22408093959093095,
      "num_tokens": 44569029.0,
      "step": 19970
    },
    {
      "entropy": 5.5450798034667965,
      "epoch": 2.2433874319088,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21749174296855928,
      "num_tokens": 44578804.0,
      "step": 19975
    },
    {
      "entropy": 5.588059949874878,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045033270372064123,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21741889417171478,
      "num_tokens": 44589362.0,
      "step": 19980
    },
    {
      "entropy": 5.609087753295898,
      "epoch": 2.244510585724715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.0739,
      "mean_token_accuracy": 0.2180183544754982,
      "num_tokens": 44600137.0,
      "step": 19985
    },
    {
      "entropy": 5.566953372955322,
      "epoch": 2.2450721626326726,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21739982962608337,
      "num_tokens": 44611707.0,
      "step": 19990
    },
    {
      "entropy": 5.563157796859741,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.043,
      "mean_token_accuracy": 0.21339641660451888,
      "num_tokens": 44621141.0,
      "step": 19995
    },
    {
      "entropy": 5.581042194366455,
      "epoch": 2.2461953164485875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21454594284296036,
      "num_tokens": 44632043.0,
      "step": 20000
    },
    {
      "entropy": 5.598944664001465,
      "epoch": 2.246756893356545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20647457987070084,
      "num_tokens": 44643905.0,
      "step": 20005
    },
    {
      "entropy": 5.5588874340057375,
      "epoch": 2.247318470264503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.0458,
      "mean_token_accuracy": 0.20355794578790665,
      "num_tokens": 44655105.0,
      "step": 20010
    },
    {
      "entropy": 5.523007726669311,
      "epoch": 2.24788004717246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045015643712327465,
      "loss": 4.993,
      "mean_token_accuracy": 0.21826727390289308,
      "num_tokens": 44667096.0,
      "step": 20015
    },
    {
      "entropy": 5.6283416748046875,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.045,
      "mean_token_accuracy": 0.2166026011109352,
      "num_tokens": 44678206.0,
      "step": 20020
    },
    {
      "entropy": 5.54194598197937,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501060250328419,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21855185329914092,
      "num_tokens": 44688619.0,
      "step": 20025
    },
    {
      "entropy": 5.540648126602173,
      "epoch": 2.249564777896333,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21129293441772462,
      "num_tokens": 44700757.0,
      "step": 20030
    },
    {
      "entropy": 5.466196012496948,
      "epoch": 2.2501263548042907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004500555906406343,
      "loss": 4.9066,
      "mean_token_accuracy": 0.225569386780262,
      "num_tokens": 44710822.0,
      "step": 20035
    },
    {
      "entropy": 5.440199995040894,
      "epoch": 2.250687931712248,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004500303650833722,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2233733519911766,
      "num_tokens": 44722163.0,
      "step": 20040
    },
    {
      "entropy": 5.614795160293579,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21222877651453018,
      "num_tokens": 44734271.0,
      "step": 20045
    },
    {
      "entropy": 5.6472821712493895,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2157983273267746,
      "num_tokens": 44745345.0,
      "step": 20050
    },
    {
      "entropy": 5.5189781665802,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21251294761896133,
      "num_tokens": 44756986.0,
      "step": 20055
    },
    {
      "entropy": 5.481416940689087,
      "epoch": 2.252934239344078,
      "grad_norm": 1.0,
      "learning_rate": 0.0004499294071320112,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22094498723745346,
      "num_tokens": 44767259.0,
      "step": 20060
    },
    {
      "entropy": 5.541097164154053,
      "epoch": 2.2534958162520358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000449904153717611,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2130111947655678,
      "num_tokens": 44778884.0,
      "step": 20065
    },
    {
      "entropy": 5.561671257019043,
      "epoch": 2.2540573931599934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004498788947341947,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21920643150806426,
      "num_tokens": 44789216.0,
      "step": 20070
    },
    {
      "entropy": 5.4591961860656735,
      "epoch": 2.2546189700679506,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004498536301825665,
      "loss": 4.9305,
      "mean_token_accuracy": 0.225284406542778,
      "num_tokens": 44799357.0,
      "step": 20075
    },
    {
      "entropy": 5.51366229057312,
      "epoch": 2.2551805469759083,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000449828360063531,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21524444073438645,
      "num_tokens": 44811040.0,
      "step": 20080
    },
    {
      "entropy": 5.612596702575684,
      "epoch": 2.255742123883866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044980308437789267,
      "loss": 5.128,
      "mean_token_accuracy": 0.20377812087535857,
      "num_tokens": 44822103.0,
      "step": 20085
    },
    {
      "entropy": 5.558966588973999,
      "epoch": 2.2563037007918236,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21482159346342086,
      "num_tokens": 44834712.0,
      "step": 20090
    },
    {
      "entropy": 5.55741024017334,
      "epoch": 2.256865277699781,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044975251631002694,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21829248666763307,
      "num_tokens": 44845412.0,
      "step": 20095
    },
    {
      "entropy": 5.548772859573364,
      "epoch": 2.2574268546077385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2158433362841606,
      "num_tokens": 44857720.0,
      "step": 20100
    },
    {
      "entropy": 5.551323461532593,
      "epoch": 2.257988431515696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21842246055603026,
      "num_tokens": 44868253.0,
      "step": 20105
    },
    {
      "entropy": 5.543375062942505,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044967662247883295,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2127770885825157,
      "num_tokens": 44878525.0,
      "step": 20110
    },
    {
      "entropy": 5.444416046142578,
      "epoch": 2.259111585331611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044965131341048454,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22253538370132447,
      "num_tokens": 44889565.0,
      "step": 20115
    },
    {
      "entropy": 5.495766687393188,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004496259987811707,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21973766535520553,
      "num_tokens": 44901191.0,
      "step": 20120
    },
    {
      "entropy": 5.5368434429168705,
      "epoch": 2.2602347391475264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004496006785916973,
      "loss": 4.942,
      "mean_token_accuracy": 0.21933159977197647,
      "num_tokens": 44911483.0,
      "step": 20125
    },
    {
      "entropy": 5.591509819030762,
      "epoch": 2.260796316055484,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.1157,
      "mean_token_accuracy": 0.2146093875169754,
      "num_tokens": 44923537.0,
      "step": 20130
    },
    {
      "entropy": 5.598501110076905,
      "epoch": 2.2613578929634413,
      "grad_norm": 1.0,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.03,
      "mean_token_accuracy": 0.216207917034626,
      "num_tokens": 44935780.0,
      "step": 20135
    },
    {
      "entropy": 5.584345054626465,
      "epoch": 2.261919469871399,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21909110397100448,
      "num_tokens": 44946879.0,
      "step": 20140
    },
    {
      "entropy": 5.6073753356933596,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.0507,
      "mean_token_accuracy": 0.22201086282730104,
      "num_tokens": 44958385.0,
      "step": 20145
    },
    {
      "entropy": 5.572727298736572,
      "epoch": 2.263042623687314,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.1304,
      "mean_token_accuracy": 0.21094195544719696,
      "num_tokens": 44970597.0,
      "step": 20150
    },
    {
      "entropy": 5.541462182998657,
      "epoch": 2.2636042005952715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21954300552606582,
      "num_tokens": 44983123.0,
      "step": 20155
    },
    {
      "entropy": 5.591849899291992,
      "epoch": 2.264165777503229,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044942328164866,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21586065590381623,
      "num_tokens": 44993378.0,
      "step": 20160
    },
    {
      "entropy": 5.533859348297119,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004493979170069486,
      "loss": 4.9572,
      "mean_token_accuracy": 0.212008997797966,
      "num_tokens": 45004905.0,
      "step": 20165
    },
    {
      "entropy": 5.520791053771973,
      "epoch": 2.265288931319144,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.016,
      "mean_token_accuracy": 0.2186549037694931,
      "num_tokens": 45016640.0,
      "step": 20170
    },
    {
      "entropy": 5.62679877281189,
      "epoch": 2.2658505082271017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20677691847085952,
      "num_tokens": 45028285.0,
      "step": 20175
    },
    {
      "entropy": 5.637751483917237,
      "epoch": 2.2664120851350593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21435697972774506,
      "num_tokens": 45038924.0,
      "step": 20180
    },
    {
      "entropy": 5.571620655059815,
      "epoch": 2.266973662043017,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044929640291920865,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21692424416542053,
      "num_tokens": 45050483.0,
      "step": 20185
    },
    {
      "entropy": 5.6206910610198975,
      "epoch": 2.267535238950974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004492710105210904,
      "loss": 5.178,
      "mean_token_accuracy": 0.20670901387929916,
      "num_tokens": 45063242.0,
      "step": 20190
    },
    {
      "entropy": 5.529792642593383,
      "epoch": 2.268096815858932,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21073022335767747,
      "num_tokens": 45074057.0,
      "step": 20195
    },
    {
      "entropy": 5.561744594573975,
      "epoch": 2.2686583927668895,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20804219245910643,
      "num_tokens": 45085488.0,
      "step": 20200
    },
    {
      "entropy": 5.639613056182862,
      "epoch": 2.2692199696748467,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.1359,
      "mean_token_accuracy": 0.21687105149030686,
      "num_tokens": 45096303.0,
      "step": 20205
    },
    {
      "entropy": 5.548724603652954,
      "epoch": 2.2697815465828044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044916938544813936,
      "loss": 4.9138,
      "mean_token_accuracy": 0.2180546924471855,
      "num_tokens": 45106411.0,
      "step": 20210
    },
    {
      "entropy": 5.538466358184815,
      "epoch": 2.270343123490762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.0848,
      "mean_token_accuracy": 0.21298103928565978,
      "num_tokens": 45117355.0,
      "step": 20215
    },
    {
      "entropy": 5.515856313705444,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044911853963470295,
      "loss": 4.9671,
      "mean_token_accuracy": 0.21655595898628235,
      "num_tokens": 45127612.0,
      "step": 20220
    },
    {
      "entropy": 5.617521476745606,
      "epoch": 2.2714662773066774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.141,
      "mean_token_accuracy": 0.21408609449863433,
      "num_tokens": 45138495.0,
      "step": 20225
    },
    {
      "entropy": 5.556015157699585,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044906767164525936,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22517688870429992,
      "num_tokens": 45149727.0,
      "step": 20230
    },
    {
      "entropy": 5.549364519119263,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21121534556150437,
      "num_tokens": 45160921.0,
      "step": 20235
    },
    {
      "entropy": 5.492720603942871,
      "epoch": 2.27315100803055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004490167814862872,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21933651566505433,
      "num_tokens": 45172607.0,
      "step": 20240
    },
    {
      "entropy": 5.615347576141358,
      "epoch": 2.273712584938507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.1455,
      "mean_token_accuracy": 0.21565479636192322,
      "num_tokens": 45186330.0,
      "step": 20245
    },
    {
      "entropy": 5.441721391677857,
      "epoch": 2.274274161846465,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004489658691642681,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22831785380840303,
      "num_tokens": 45198239.0,
      "step": 20250
    },
    {
      "entropy": 5.474574279785156,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044894040469414207,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22363467365503312,
      "num_tokens": 45209295.0,
      "step": 20255
    },
    {
      "entropy": 5.587273263931275,
      "epoch": 2.27539731566238,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21424532681703568,
      "num_tokens": 45220210.0,
      "step": 20260
    },
    {
      "entropy": 5.582193231582641,
      "epoch": 2.2759588925703373,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.0714,
      "mean_token_accuracy": 0.21551869362592696,
      "num_tokens": 45231809.0,
      "step": 20265
    },
    {
      "entropy": 5.563639354705811,
      "epoch": 2.276520469478295,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.0337,
      "mean_token_accuracy": 0.22337889075279235,
      "num_tokens": 45242847.0,
      "step": 20270
    },
    {
      "entropy": 5.505256223678589,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044883849143844914,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22452423721551895,
      "num_tokens": 45253357.0,
      "step": 20275
    },
    {
      "entropy": 5.456206941604615,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488129992847848,
      "loss": 4.908,
      "mean_token_accuracy": 0.21905696541070938,
      "num_tokens": 45264607.0,
      "step": 20280
    },
    {
      "entropy": 5.596430730819702,
      "epoch": 2.2782052002021675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.071,
      "mean_token_accuracy": 0.21662406325340272,
      "num_tokens": 45275211.0,
      "step": 20285
    },
    {
      "entropy": 5.507759618759155,
      "epoch": 2.278766777110125,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004487619983754483,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2262323021888733,
      "num_tokens": 45286665.0,
      "step": 20290
    },
    {
      "entropy": 5.516144847869873,
      "epoch": 2.279328354018083,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.07,
      "mean_token_accuracy": 0.21683111935853958,
      "num_tokens": 45298066.0,
      "step": 20295
    },
    {
      "entropy": 5.534784460067749,
      "epoch": 2.27988993092604,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2095528647303581,
      "num_tokens": 45308799.0,
      "step": 20300
    },
    {
      "entropy": 5.510761785507202,
      "epoch": 2.2804515078339977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004486854555186037,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2246817246079445,
      "num_tokens": 45319932.0,
      "step": 20305
    },
    {
      "entropy": 5.580576992034912,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.0502,
      "mean_token_accuracy": 0.22258580923080445,
      "num_tokens": 45330008.0,
      "step": 20310
    },
    {
      "entropy": 5.579342365264893,
      "epoch": 2.281574661649913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20708697140216828,
      "num_tokens": 45340877.0,
      "step": 20315
    },
    {
      "entropy": 5.566436910629273,
      "epoch": 2.2821362385578707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21946957260370253,
      "num_tokens": 45351956.0,
      "step": 20320
    },
    {
      "entropy": 5.610913848876953,
      "epoch": 2.282697815465828,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.1211,
      "mean_token_accuracy": 0.20493345260620116,
      "num_tokens": 45363133.0,
      "step": 20325
    },
    {
      "entropy": 5.566841745376587,
      "epoch": 2.2832593923737856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004485577734971247,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21748705506324767,
      "num_tokens": 45373917.0,
      "step": 20330
    },
    {
      "entropy": 5.5820897102355955,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.0812,
      "mean_token_accuracy": 0.21580351293087005,
      "num_tokens": 45385575.0,
      "step": 20335
    },
    {
      "entropy": 5.545222330093384,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044850666199981244,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21779112815856932,
      "num_tokens": 45396615.0,
      "step": 20340
    },
    {
      "entropy": 5.565461540222168,
      "epoch": 2.284944123097658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044848109796397035,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21578286290168763,
      "num_tokens": 45408931.0,
      "step": 20345
    },
    {
      "entropy": 5.630756902694702,
      "epoch": 2.285505700005616,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.0713,
      "mean_token_accuracy": 0.21396509557962418,
      "num_tokens": 45419993.0,
      "step": 20350
    },
    {
      "entropy": 5.566761207580567,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.1077,
      "mean_token_accuracy": 0.2066176041960716,
      "num_tokens": 45431889.0,
      "step": 20355
    },
    {
      "entropy": 5.456265640258789,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21775930374860764,
      "num_tokens": 45444233.0,
      "step": 20360
    },
    {
      "entropy": 5.502660894393921,
      "epoch": 2.2871904307294884,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004483787865916898,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21656214892864228,
      "num_tokens": 45455743.0,
      "step": 20365
    },
    {
      "entropy": 5.533611917495728,
      "epoch": 2.287752007637446,
      "grad_norm": 1.0,
      "learning_rate": 0.00044835319494546366,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2140478953719139,
      "num_tokens": 45466272.0,
      "step": 20370
    },
    {
      "entropy": 5.576041889190674,
      "epoch": 2.2883135845454037,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044832759777960456,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21914386749267578,
      "num_tokens": 45477828.0,
      "step": 20375
    },
    {
      "entropy": 5.5511890888214115,
      "epoch": 2.288875161453361,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004483019950949276,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21633769124746322,
      "num_tokens": 45489315.0,
      "step": 20380
    },
    {
      "entropy": 5.487712860107422,
      "epoch": 2.2894367383613186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004482763868922478,
      "loss": 4.987,
      "mean_token_accuracy": 0.22231616228818893,
      "num_tokens": 45499836.0,
      "step": 20385
    },
    {
      "entropy": 5.564362478256226,
      "epoch": 2.289998315269276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.0454,
      "mean_token_accuracy": 0.21476729661226274,
      "num_tokens": 45511106.0,
      "step": 20390
    },
    {
      "entropy": 5.5496924877166744,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004482251539361418,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21394300013780593,
      "num_tokens": 45521314.0,
      "step": 20395
    },
    {
      "entropy": 5.595235347747803,
      "epoch": 2.291121469085191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2214626058936119,
      "num_tokens": 45531675.0,
      "step": 20400
    },
    {
      "entropy": 5.59092116355896,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.0609,
      "mean_token_accuracy": 0.22041004449129104,
      "num_tokens": 45542064.0,
      "step": 20405
    },
    {
      "entropy": 5.552023267745971,
      "epoch": 2.2922446229011064,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.0857,
      "mean_token_accuracy": 0.21748349964618682,
      "num_tokens": 45553764.0,
      "step": 20410
    },
    {
      "entropy": 5.6326967716217045,
      "epoch": 2.292806199809064,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044812262184378503,
      "loss": 5.083,
      "mean_token_accuracy": 0.21780811697244645,
      "num_tokens": 45565302.0,
      "step": 20415
    },
    {
      "entropy": 5.446408271789551,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044809697503792647,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22583274394273758,
      "num_tokens": 45577165.0,
      "step": 20420
    },
    {
      "entropy": 5.524872541427612,
      "epoch": 2.293929353624979,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044807132272059306,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22041620761156083,
      "num_tokens": 45587771.0,
      "step": 20425
    },
    {
      "entropy": 5.545679664611816,
      "epoch": 2.2944909305329366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004480456648926016,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22803145051002502,
      "num_tokens": 45598815.0,
      "step": 20430
    },
    {
      "entropy": 5.53143277168274,
      "epoch": 2.295052507440894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004480200015547691,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21826599091291427,
      "num_tokens": 45609007.0,
      "step": 20435
    },
    {
      "entropy": 5.547308111190796,
      "epoch": 2.2956140843488515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.0814,
      "mean_token_accuracy": 0.21829911321401596,
      "num_tokens": 45621323.0,
      "step": 20440
    },
    {
      "entropy": 5.476387071609497,
      "epoch": 2.296175661256809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004479686583528494,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22197080701589583,
      "num_tokens": 45631895.0,
      "step": 20445
    },
    {
      "entropy": 5.57951545715332,
      "epoch": 2.296737238164767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21964376419782639,
      "num_tokens": 45643204.0,
      "step": 20450
    },
    {
      "entropy": 5.599549293518066,
      "epoch": 2.297298815072724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2151808947324753,
      "num_tokens": 45654607.0,
      "step": 20455
    },
    {
      "entropy": 5.472682046890259,
      "epoch": 2.2978603919806817,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004478916022465958,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22247938215732574,
      "num_tokens": 45666746.0,
      "step": 20460
    },
    {
      "entropy": 5.52073655128479,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004478659058668827,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22619403004646302,
      "num_tokens": 45676663.0,
      "step": 20465
    },
    {
      "entropy": 5.485407066345215,
      "epoch": 2.298983545796597,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044784020398305204,
      "loss": 4.9617,
      "mean_token_accuracy": 0.21707216501235962,
      "num_tokens": 45688746.0,
      "step": 20470
    },
    {
      "entropy": 5.640700817108154,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.1479,
      "mean_token_accuracy": 0.20968689620494843,
      "num_tokens": 45700199.0,
      "step": 20475
    },
    {
      "entropy": 5.547812986373901,
      "epoch": 2.300106699612512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044778878370631183,
      "loss": 5.0043,
      "mean_token_accuracy": 0.21039715856313707,
      "num_tokens": 45710771.0,
      "step": 20480
    },
    {
      "entropy": 5.593759202957154,
      "epoch": 2.3006682765204696,
      "grad_norm": 1.0,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.1583,
      "mean_token_accuracy": 0.2125330701470375,
      "num_tokens": 45721957.0,
      "step": 20485
    },
    {
      "entropy": 5.572651052474976,
      "epoch": 2.301229853428427,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21614271104335786,
      "num_tokens": 45733904.0,
      "step": 20490
    },
    {
      "entropy": 5.626233625411987,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.0439,
      "mean_token_accuracy": 0.2158783942461014,
      "num_tokens": 45744004.0,
      "step": 20495
    },
    {
      "entropy": 5.566758060455323,
      "epoch": 2.302353007244342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004476858771394412,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2299055963754654,
      "num_tokens": 45755130.0,
      "step": 20500
    },
    {
      "entropy": 5.540961885452271,
      "epoch": 2.3029145841522998,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.132,
      "mean_token_accuracy": 0.20976939797401428,
      "num_tokens": 45767353.0,
      "step": 20505
    },
    {
      "entropy": 5.600943899154663,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000447634390862417,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20062874555587767,
      "num_tokens": 45779266.0,
      "step": 20510
    },
    {
      "entropy": 5.652445363998413,
      "epoch": 2.3040377379682146,
      "grad_norm": 1.0,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.1209,
      "mean_token_accuracy": 0.21222143918275832,
      "num_tokens": 45790601.0,
      "step": 20515
    },
    {
      "entropy": 5.567757320404053,
      "epoch": 2.3045993148761723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044758288259840904,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22123689353466033,
      "num_tokens": 45801449.0,
      "step": 20520
    },
    {
      "entropy": 5.4978687286376955,
      "epoch": 2.30516089178413,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004475571202233363,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21775230169296264,
      "num_tokens": 45811737.0,
      "step": 20525
    },
    {
      "entropy": 5.557856512069702,
      "epoch": 2.305722468692087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21920127421617508,
      "num_tokens": 45822955.0,
      "step": 20530
    },
    {
      "entropy": 5.5447766304016115,
      "epoch": 2.306284045600045,
      "grad_norm": 1.0,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.094,
      "mean_token_accuracy": 0.21994825154542924,
      "num_tokens": 45834697.0,
      "step": 20535
    },
    {
      "entropy": 5.48887996673584,
      "epoch": 2.3068456225080025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004474798001356859,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2154937282204628,
      "num_tokens": 45846869.0,
      "step": 20540
    },
    {
      "entropy": 5.4811066627502445,
      "epoch": 2.30740719941596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000447454015788394,
      "loss": 4.996,
      "mean_token_accuracy": 0.21849816143512726,
      "num_tokens": 45858171.0,
      "step": 20545
    },
    {
      "entropy": 5.527911567687989,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004474282259500992,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21488165855407715,
      "num_tokens": 45868715.0,
      "step": 20550
    },
    {
      "entropy": 5.572001791000366,
      "epoch": 2.308530353231875,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000447402430621623,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2222256064414978,
      "num_tokens": 45880569.0,
      "step": 20555
    },
    {
      "entropy": 5.566620349884033,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004473766298037865,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2187553882598877,
      "num_tokens": 45890452.0,
      "step": 20560
    },
    {
      "entropy": 5.619940519332886,
      "epoch": 2.3096535070477904,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.1001,
      "mean_token_accuracy": 0.21549041718244552,
      "num_tokens": 45903125.0,
      "step": 20565
    },
    {
      "entropy": 5.534860515594483,
      "epoch": 2.3102150839557476,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044732501170331886,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2234114646911621,
      "num_tokens": 45914961.0,
      "step": 20570
    },
    {
      "entropy": 5.635332250595093,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20963136553764344,
      "num_tokens": 45927198.0,
      "step": 20575
    },
    {
      "entropy": 5.584187650680542,
      "epoch": 2.311338237771663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2174232318997383,
      "num_tokens": 45937937.0,
      "step": 20580
    },
    {
      "entropy": 5.660223007202148,
      "epoch": 2.31189981467962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.052,
      "mean_token_accuracy": 0.20956479609012604,
      "num_tokens": 45948983.0,
      "step": 20585
    },
    {
      "entropy": 5.549386119842529,
      "epoch": 2.312461391587578,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2146148979663849,
      "num_tokens": 45960674.0,
      "step": 20590
    },
    {
      "entropy": 5.550292444229126,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21366385370492935,
      "num_tokens": 45971260.0,
      "step": 20595
    },
    {
      "entropy": 5.643903684616089,
      "epoch": 2.313584545403493,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.0239,
      "mean_token_accuracy": 0.2218250334262848,
      "num_tokens": 45983942.0,
      "step": 20600
    },
    {
      "entropy": 5.563588809967041,
      "epoch": 2.314146122311451,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20474110394716263,
      "num_tokens": 45995983.0,
      "step": 20605
    },
    {
      "entropy": 5.541653442382812,
      "epoch": 2.314707699219408,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044711831989142534,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2204861655831337,
      "num_tokens": 46007074.0,
      "step": 20610
    },
    {
      "entropy": 5.606692314147949,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21870445162057878,
      "num_tokens": 46018376.0,
      "step": 20615
    },
    {
      "entropy": 5.553802490234375,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004470665921188519,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21515269279479982,
      "num_tokens": 46029192.0,
      "step": 20620
    },
    {
      "entropy": 5.530257940292358,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044704072001415285,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21515427082777022,
      "num_tokens": 46040735.0,
      "step": 20625
    },
    {
      "entropy": 5.510697078704834,
      "epoch": 2.316954006851238,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044701484243161055,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22104890942573546,
      "num_tokens": 46050859.0,
      "step": 20630
    },
    {
      "entropy": 5.463787174224853,
      "epoch": 2.317515583759196,
      "grad_norm": 1.0,
      "learning_rate": 0.000446988959372049,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21798447668552398,
      "num_tokens": 46060704.0,
      "step": 20635
    },
    {
      "entropy": 5.479064273834228,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21448990404605867,
      "num_tokens": 46072317.0,
      "step": 20640
    },
    {
      "entropy": 5.632720994949341,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.129,
      "mean_token_accuracy": 0.21106757670640947,
      "num_tokens": 46083823.0,
      "step": 20645
    },
    {
      "entropy": 5.68085823059082,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21443405300378798,
      "num_tokens": 46093920.0,
      "step": 20650
    },
    {
      "entropy": 5.496704339981079,
      "epoch": 2.319761891391026,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004468853723800948,
      "loss": 5.0112,
      "mean_token_accuracy": 0.22241187691688538,
      "num_tokens": 46104618.0,
      "step": 20655
    },
    {
      "entropy": 5.502344274520874,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468594619478021,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22287059277296067,
      "num_tokens": 46114748.0,
      "step": 20660
    },
    {
      "entropy": 5.567959022521973,
      "epoch": 2.320885045206941,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.0628,
      "mean_token_accuracy": 0.22061190605163575,
      "num_tokens": 46125125.0,
      "step": 20665
    },
    {
      "entropy": 5.46665620803833,
      "epoch": 2.3214466221148986,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004468076246678261,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21278192102909088,
      "num_tokens": 46135877.0,
      "step": 20670
    },
    {
      "entropy": 5.618646430969238,
      "epoch": 2.3220081990228563,
      "grad_norm": 1.0,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.1563,
      "mean_token_accuracy": 0.20922595560550689,
      "num_tokens": 46146769.0,
      "step": 20675
    },
    {
      "entropy": 5.54126877784729,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.0,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2162444695830345,
      "num_tokens": 46158408.0,
      "step": 20680
    },
    {
      "entropy": 5.618822050094605,
      "epoch": 2.323131352838771,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.125,
      "mean_token_accuracy": 0.20835967361927032,
      "num_tokens": 46169961.0,
      "step": 20685
    },
    {
      "entropy": 5.615501976013183,
      "epoch": 2.323692929746729,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.0889,
      "mean_token_accuracy": 0.21710782945156099,
      "num_tokens": 46181750.0,
      "step": 20690
    },
    {
      "entropy": 5.565064334869385,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004466779357499612,
      "loss": 4.9732,
      "mean_token_accuracy": 0.22400198727846146,
      "num_tokens": 46192745.0,
      "step": 20695
    },
    {
      "entropy": 5.550230836868286,
      "epoch": 2.324816083562644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21534472703933716,
      "num_tokens": 46203267.0,
      "step": 20700
    },
    {
      "entropy": 5.544508838653565,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2236128866672516,
      "num_tokens": 46214966.0,
      "step": 20705
    },
    {
      "entropy": 5.532233142852784,
      "epoch": 2.325939237378559,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044660005679713027,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21984640657901763,
      "num_tokens": 46225755.0,
      "step": 20710
    },
    {
      "entropy": 5.572015380859375,
      "epoch": 2.3265008142865167,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.0425,
      "mean_token_accuracy": 0.21849248111248015,
      "num_tokens": 46237091.0,
      "step": 20715
    },
    {
      "entropy": 5.586054944992066,
      "epoch": 2.327062391194474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21506887078285217,
      "num_tokens": 46249250.0,
      "step": 20720
    },
    {
      "entropy": 5.5476579666137695,
      "epoch": 2.3276239681024316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044652212866998914,
      "loss": 4.9931,
      "mean_token_accuracy": 0.2184322401881218,
      "num_tokens": 46259866.0,
      "step": 20725
    },
    {
      "entropy": 5.509886360168457,
      "epoch": 2.328185545010389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044649614170384354,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2259687215089798,
      "num_tokens": 46270888.0,
      "step": 20730
    },
    {
      "entropy": 5.531872415542603,
      "epoch": 2.328747121918347,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21379695683717728,
      "num_tokens": 46281109.0,
      "step": 20735
    },
    {
      "entropy": 5.562994480133057,
      "epoch": 2.329308698826304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044644415139086913,
      "loss": 4.8982,
      "mean_token_accuracy": 0.22558528631925584,
      "num_tokens": 46292171.0,
      "step": 20740
    },
    {
      "entropy": 5.552895355224609,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004464181480456955,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21549065560102462,
      "num_tokens": 46302919.0,
      "step": 20745
    },
    {
      "entropy": 5.4073436737060545,
      "epoch": 2.3304318526422194,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004463921392425017,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22527623325586318,
      "num_tokens": 46313199.0,
      "step": 20750
    },
    {
      "entropy": 5.60562744140625,
      "epoch": 2.330993429550177,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21719380021095275,
      "num_tokens": 46324316.0,
      "step": 20755
    },
    {
      "entropy": 5.547980070114136,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.125,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.0503,
      "mean_token_accuracy": 0.2110810935497284,
      "num_tokens": 46335767.0,
      "step": 20760
    },
    {
      "entropy": 5.525160455703736,
      "epoch": 2.332116583366092,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004463140800930804,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21188642382621764,
      "num_tokens": 46347840.0,
      "step": 20765
    },
    {
      "entropy": 5.548009443283081,
      "epoch": 2.3326781602740496,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.1141,
      "mean_token_accuracy": 0.20905691534280776,
      "num_tokens": 46359766.0,
      "step": 20770
    },
    {
      "entropy": 5.574706745147705,
      "epoch": 2.333239737182007,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044626201338521834,
      "loss": 4.9877,
      "mean_token_accuracy": 0.22901145964860917,
      "num_tokens": 46371426.0,
      "step": 20775
    },
    {
      "entropy": 5.563321113586426,
      "epoch": 2.3338013140899645,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.0712,
      "mean_token_accuracy": 0.21051281392574311,
      "num_tokens": 46382257.0,
      "step": 20780
    },
    {
      "entropy": 5.59622631072998,
      "epoch": 2.334362890997922,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2140096455812454,
      "num_tokens": 46393689.0,
      "step": 20785
    },
    {
      "entropy": 5.578646564483643,
      "epoch": 2.33492446790588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21530118882656096,
      "num_tokens": 46404856.0,
      "step": 20790
    },
    {
      "entropy": 5.65553388595581,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.0742,
      "mean_token_accuracy": 0.21044759452342987,
      "num_tokens": 46415395.0,
      "step": 20795
    },
    {
      "entropy": 5.643966770172119,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.0883,
      "mean_token_accuracy": 0.2132789745926857,
      "num_tokens": 46426773.0,
      "step": 20800
    },
    {
      "entropy": 5.503091621398926,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004461056824149979,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2219191387295723,
      "num_tokens": 46438206.0,
      "step": 20805
    },
    {
      "entropy": 5.567049312591553,
      "epoch": 2.33717077553771,
      "grad_norm": 1.0625,
      "learning_rate": 0.000446079608180204,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21476563513278962,
      "num_tokens": 46449922.0,
      "step": 20810
    },
    {
      "entropy": 5.526745557785034,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004460535284981692,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22897884249687195,
      "num_tokens": 46461715.0,
      "step": 20815
    },
    {
      "entropy": 5.505333375930786,
      "epoch": 2.338293929353625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004460274433697235,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21457634270191192,
      "num_tokens": 46473136.0,
      "step": 20820
    },
    {
      "entropy": 5.505570745468139,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044600135279569784,
      "loss": 4.9107,
      "mean_token_accuracy": 0.22645069807767867,
      "num_tokens": 46483644.0,
      "step": 20825
    },
    {
      "entropy": 5.574982357025147,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.0742,
      "mean_token_accuracy": 0.2173317089676857,
      "num_tokens": 46494896.0,
      "step": 20830
    },
    {
      "entropy": 5.589889335632324,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004459491553142291,
      "loss": 5.0548,
      "mean_token_accuracy": 0.21539081782102584,
      "num_tokens": 46505415.0,
      "step": 20835
    },
    {
      "entropy": 5.532310628890992,
      "epoch": 2.340540236985455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.0338,
      "mean_token_accuracy": 0.2183132141828537,
      "num_tokens": 46516280.0,
      "step": 20840
    },
    {
      "entropy": 5.484214448928833,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000445896936060411,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21869853734970093,
      "num_tokens": 46526456.0,
      "step": 20845
    },
    {
      "entropy": 5.583834791183472,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.043,
      "mean_token_accuracy": 0.22055585086345672,
      "num_tokens": 46537547.0,
      "step": 20850
    },
    {
      "entropy": 5.456442403793335,
      "epoch": 2.3422249677093276,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044584469504089407,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22030052691698074,
      "num_tokens": 46549424.0,
      "step": 20855
    },
    {
      "entropy": 5.553288459777832,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004458185663710777,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21694985181093215,
      "num_tokens": 46560234.0,
      "step": 20860
    },
    {
      "entropy": 5.583620262145996,
      "epoch": 2.343348121525243,
      "grad_norm": 1.125,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21740693598985672,
      "num_tokens": 46570847.0,
      "step": 20865
    },
    {
      "entropy": 5.463548231124878,
      "epoch": 2.3439096984332,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044576629271548884,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21751579642295837,
      "num_tokens": 46582379.0,
      "step": 20870
    },
    {
      "entropy": 5.52239875793457,
      "epoch": 2.344471275341158,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.003,
      "mean_token_accuracy": 0.21263401657342912,
      "num_tokens": 46594669.0,
      "step": 20875
    },
    {
      "entropy": 5.462040185928345,
      "epoch": 2.3450328522491155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004457139973108403,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22139841616153716,
      "num_tokens": 46605834.0,
      "step": 20880
    },
    {
      "entropy": 5.550636100769043,
      "epoch": 2.345594429157073,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.0148,
      "mean_token_accuracy": 0.22142380177974702,
      "num_tokens": 46616684.0,
      "step": 20885
    },
    {
      "entropy": 5.540843391418457,
      "epoch": 2.346156006065031,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.0368,
      "mean_token_accuracy": 0.22062717527151107,
      "num_tokens": 46628045.0,
      "step": 20890
    },
    {
      "entropy": 5.5419590950012205,
      "epoch": 2.346717582972988,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.02,
      "mean_token_accuracy": 0.21409319043159486,
      "num_tokens": 46639455.0,
      "step": 20895
    },
    {
      "entropy": 5.636029672622681,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.2062,
      "mean_token_accuracy": 0.2065692737698555,
      "num_tokens": 46651025.0,
      "step": 20900
    },
    {
      "entropy": 5.5961908340454105,
      "epoch": 2.3478407367889034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.021,
      "mean_token_accuracy": 0.21742363423109054,
      "num_tokens": 46661859.0,
      "step": 20905
    },
    {
      "entropy": 5.591137552261353,
      "epoch": 2.3484023136968606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044555698066915385,
      "loss": 4.9778,
      "mean_token_accuracy": 0.217545722424984,
      "num_tokens": 46673441.0,
      "step": 20910
    },
    {
      "entropy": 5.600366592407227,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004455307922169092,
      "loss": 5.077,
      "mean_token_accuracy": 0.21524248868227006,
      "num_tokens": 46684959.0,
      "step": 20915
    },
    {
      "entropy": 5.589709949493408,
      "epoch": 2.349525467512776,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.1099,
      "mean_token_accuracy": 0.22262055426836014,
      "num_tokens": 46694940.0,
      "step": 20920
    },
    {
      "entropy": 5.555403327941894,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.125,
      "learning_rate": 0.0004454783990239542,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2208619922399521,
      "num_tokens": 46706184.0,
      "step": 20925
    },
    {
      "entropy": 5.554701042175293,
      "epoch": 2.350648621328691,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.0067,
      "mean_token_accuracy": 0.22455042898654937,
      "num_tokens": 46718267.0,
      "step": 20930
    },
    {
      "entropy": 5.5336850643157955,
      "epoch": 2.3512101982366485,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21690160930156707,
      "num_tokens": 46729293.0,
      "step": 20935
    },
    {
      "entropy": 5.51219744682312,
      "epoch": 2.351771775144606,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22259681969881057,
      "num_tokens": 46740214.0,
      "step": 20940
    },
    {
      "entropy": 5.62998685836792,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.0786,
      "mean_token_accuracy": 0.21124810129404067,
      "num_tokens": 46751269.0,
      "step": 20945
    },
    {
      "entropy": 5.514626407623291,
      "epoch": 2.352894928960521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004453473210644503,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22854215204715728,
      "num_tokens": 46761851.0,
      "step": 20950
    },
    {
      "entropy": 5.494147777557373,
      "epoch": 2.3534565058684787,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004453210891974355,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21906809508800507,
      "num_tokens": 46773441.0,
      "step": 20955
    },
    {
      "entropy": 5.432849168777466,
      "epoch": 2.3540180827764363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044529485190733113,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21639605462551117,
      "num_tokens": 46785016.0,
      "step": 20960
    },
    {
      "entropy": 5.591449022293091,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2107321187853813,
      "num_tokens": 46796059.0,
      "step": 20965
    },
    {
      "entropy": 5.522956037521363,
      "epoch": 2.355141236592351,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044524236106119563,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22196345031261444,
      "num_tokens": 46805725.0,
      "step": 20970
    },
    {
      "entropy": 5.520514869689942,
      "epoch": 2.355702813500309,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044521610750683586,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2213359847664833,
      "num_tokens": 46816300.0,
      "step": 20975
    },
    {
      "entropy": 5.5293677806854244,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004451898485327291,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2154019609093666,
      "num_tokens": 46827498.0,
      "step": 20980
    },
    {
      "entropy": 5.523615980148316,
      "epoch": 2.356825967316224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004451635841397116,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21961453855037688,
      "num_tokens": 46838416.0,
      "step": 20985
    },
    {
      "entropy": 5.586485528945923,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.0971,
      "mean_token_accuracy": 0.21427804827690125,
      "num_tokens": 46849371.0,
      "step": 20990
    },
    {
      "entropy": 5.5036732196807865,
      "epoch": 2.357949121132139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21663839966058732,
      "num_tokens": 46860095.0,
      "step": 20995
    },
    {
      "entropy": 5.603206300735474,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.1167,
      "mean_token_accuracy": 0.2096370503306389,
      "num_tokens": 46871217.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.356511630008354,
      "eval_loss": 5.131128311157227,
      "eval_mean_token_accuracy": 0.21962888290933152,
      "eval_num_tokens": 46871217.0,
      "eval_runtime": 18.3072,
      "eval_samples_per_second": 1694.033,
      "eval_steps_per_second": 211.775,
      "step": 21000
    },
    {
      "entropy": 5.523681640625,
      "epoch": 2.359072274948054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.0558,
      "mean_token_accuracy": 0.22100480049848556,
      "num_tokens": 46881224.0,
      "step": 21005
    },
    {
      "entropy": 5.538973236083985,
      "epoch": 2.3596338518560116,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.0422,
      "mean_token_accuracy": 0.217678040266037,
      "num_tokens": 46892673.0,
      "step": 21010
    },
    {
      "entropy": 5.711062479019165,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.1123,
      "mean_token_accuracy": 0.2145504131913185,
      "num_tokens": 46902753.0,
      "step": 21015
    },
    {
      "entropy": 5.445335435867309,
      "epoch": 2.360757005671927,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044497958172935536,
      "loss": 4.9391,
      "mean_token_accuracy": 0.22050268352031707,
      "num_tokens": 46914093.0,
      "step": 21020
    },
    {
      "entropy": 5.501153755187988,
      "epoch": 2.361318582579884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.0313,
      "mean_token_accuracy": 0.221801994740963,
      "num_tokens": 46924946.0,
      "step": 21025
    },
    {
      "entropy": 5.572368478775024,
      "epoch": 2.361880159487842,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.0279,
      "mean_token_accuracy": 0.2114831730723381,
      "num_tokens": 46936756.0,
      "step": 21030
    },
    {
      "entropy": 5.659052801132202,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21154499799013138,
      "num_tokens": 46948552.0,
      "step": 21035
    },
    {
      "entropy": 5.638910579681396,
      "epoch": 2.363003313303757,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.0685,
      "mean_token_accuracy": 0.21179239600896835,
      "num_tokens": 46960926.0,
      "step": 21040
    },
    {
      "entropy": 5.595280313491822,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21454794257879256,
      "num_tokens": 46972657.0,
      "step": 21045
    },
    {
      "entropy": 5.563747501373291,
      "epoch": 2.364126467119672,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.0553,
      "mean_token_accuracy": 0.2223225474357605,
      "num_tokens": 46984585.0,
      "step": 21050
    },
    {
      "entropy": 5.511314725875854,
      "epoch": 2.3646880440276297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2128569945693016,
      "num_tokens": 46996052.0,
      "step": 21055
    },
    {
      "entropy": 5.619500541687012,
      "epoch": 2.365249620935587,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044476896854422785,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22417522370815277,
      "num_tokens": 47006185.0,
      "step": 21060
    },
    {
      "entropy": 5.51031460762024,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.0455,
      "mean_token_accuracy": 0.2177656039595604,
      "num_tokens": 47016647.0,
      "step": 21065
    },
    {
      "entropy": 5.587039375305176,
      "epoch": 2.366372774751502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21400221884250642,
      "num_tokens": 47026847.0,
      "step": 21070
    },
    {
      "entropy": 5.574076461791992,
      "epoch": 2.36693435165946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21229515969753265,
      "num_tokens": 47037948.0,
      "step": 21075
    },
    {
      "entropy": 5.574484634399414,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.0064,
      "mean_token_accuracy": 0.22095408141613007,
      "num_tokens": 47047751.0,
      "step": 21080
    },
    {
      "entropy": 5.580166673660278,
      "epoch": 2.3680575054753747,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21603707671165467,
      "num_tokens": 47059589.0,
      "step": 21085
    },
    {
      "entropy": 5.5431239128112795,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21586443930864335,
      "num_tokens": 47070060.0,
      "step": 21090
    },
    {
      "entropy": 5.624604797363281,
      "epoch": 2.36918065929129,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2157918021082878,
      "num_tokens": 47081419.0,
      "step": 21095
    },
    {
      "entropy": 5.568482255935669,
      "epoch": 2.3697422361992473,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20693252384662628,
      "num_tokens": 47093701.0,
      "step": 21100
    },
    {
      "entropy": 5.563736343383789,
      "epoch": 2.370303813107205,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2218781366944313,
      "num_tokens": 47104930.0,
      "step": 21105
    },
    {
      "entropy": 5.553672981262207,
      "epoch": 2.3708653900151626,
      "grad_norm": 1.0,
      "learning_rate": 0.0004445052156173796,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21273477226495743,
      "num_tokens": 47115109.0,
      "step": 21110
    },
    {
      "entropy": 5.514443302154541,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004444788106267161,
      "loss": 4.927,
      "mean_token_accuracy": 0.2207065612077713,
      "num_tokens": 47125386.0,
      "step": 21115
    },
    {
      "entropy": 5.538251304626465,
      "epoch": 2.3719885438310775,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.061,
      "mean_token_accuracy": 0.2117947593331337,
      "num_tokens": 47136066.0,
      "step": 21120
    },
    {
      "entropy": 5.568319797515869,
      "epoch": 2.372550120739035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444259844574167,
      "loss": 5.0909,
      "mean_token_accuracy": 0.21495731323957443,
      "num_tokens": 47147551.0,
      "step": 21125
    },
    {
      "entropy": 5.608590173721313,
      "epoch": 2.373111697646993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21623075008392334,
      "num_tokens": 47158798.0,
      "step": 21130
    },
    {
      "entropy": 5.646775817871093,
      "epoch": 2.3736732745549505,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004443731367097613,
      "loss": 5.099,
      "mean_token_accuracy": 0.21258010417222978,
      "num_tokens": 47169519.0,
      "step": 21135
    },
    {
      "entropy": 5.581028127670288,
      "epoch": 2.3742348514629077,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044434670474615334,
      "loss": 5.0445,
      "mean_token_accuracy": 0.2186293825507164,
      "num_tokens": 47180147.0,
      "step": 21140
    },
    {
      "entropy": 5.600242328643799,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.0935,
      "mean_token_accuracy": 0.2171688750386238,
      "num_tokens": 47191832.0,
      "step": 21145
    },
    {
      "entropy": 5.493726539611816,
      "epoch": 2.375358005278823,
      "grad_norm": 1.046875,
      "learning_rate": 0.000444293824643585,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22053854167461395,
      "num_tokens": 47201979.0,
      "step": 21150
    },
    {
      "entropy": 5.572039079666138,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.125,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.1229,
      "mean_token_accuracy": 0.21103024780750274,
      "num_tokens": 47212271.0,
      "step": 21155
    },
    {
      "entropy": 5.582833194732666,
      "epoch": 2.376481159094738,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004442409229794925,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21813535541296006,
      "num_tokens": 47223412.0,
      "step": 21160
    },
    {
      "entropy": 5.6280111312866214,
      "epoch": 2.3770427360026956,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.1384,
      "mean_token_accuracy": 0.20169369280338287,
      "num_tokens": 47235919.0,
      "step": 21165
    },
    {
      "entropy": 5.565730142593384,
      "epoch": 2.377604312910653,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004441879997606138,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21020394712686538,
      "num_tokens": 47246568.0,
      "step": 21170
    },
    {
      "entropy": 5.643782043457032,
      "epoch": 2.378165889818611,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.1198,
      "mean_token_accuracy": 0.2128550261259079,
      "num_tokens": 47259474.0,
      "step": 21175
    },
    {
      "entropy": 5.577867794036865,
      "epoch": 2.378727466726568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.017,
      "mean_token_accuracy": 0.21882927268743516,
      "num_tokens": 47270906.0,
      "step": 21180
    },
    {
      "entropy": 5.52909369468689,
      "epoch": 2.3792890436345258,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20161837637424468,
      "num_tokens": 47283341.0,
      "step": 21185
    },
    {
      "entropy": 5.637228727340698,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.047,
      "mean_token_accuracy": 0.21241812109947206,
      "num_tokens": 47294590.0,
      "step": 21190
    },
    {
      "entropy": 5.446432733535767,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004440555974554673,
      "loss": 4.8815,
      "mean_token_accuracy": 0.217696850001812,
      "num_tokens": 47305111.0,
      "step": 21195
    },
    {
      "entropy": 5.53251461982727,
      "epoch": 2.3809737743583983,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2162393346428871,
      "num_tokens": 47316145.0,
      "step": 21200
    },
    {
      "entropy": 5.558957242965699,
      "epoch": 2.381535351266356,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004440025988479349,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21577499955892562,
      "num_tokens": 47327085.0,
      "step": 21205
    },
    {
      "entropy": 5.6071906089782715,
      "epoch": 2.3820969281743136,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.1607,
      "mean_token_accuracy": 0.2092458873987198,
      "num_tokens": 47338308.0,
      "step": 21210
    },
    {
      "entropy": 5.635139083862304,
      "epoch": 2.382658505082271,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.1251,
      "mean_token_accuracy": 0.21251540631055832,
      "num_tokens": 47348717.0,
      "step": 21215
    },
    {
      "entropy": 5.573373794555664,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044392306058043453,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22313041985034943,
      "num_tokens": 47359804.0,
      "step": 21220
    },
    {
      "entropy": 5.560384798049927,
      "epoch": 2.383781658898186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2118856728076935,
      "num_tokens": 47370589.0,
      "step": 21225
    },
    {
      "entropy": 5.5313023090362545,
      "epoch": 2.384343235806144,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004438700081744819,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22019313424825668,
      "num_tokens": 47381586.0,
      "step": 21230
    },
    {
      "entropy": 5.489575576782227,
      "epoch": 2.384904812714101,
      "grad_norm": 0.875,
      "learning_rate": 0.00044384347390575337,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22742175906896592,
      "num_tokens": 47394064.0,
      "step": 21235
    },
    {
      "entropy": 5.532406568527222,
      "epoch": 2.3854663896220587,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21776185035705567,
      "num_tokens": 47405371.0,
      "step": 21240
    },
    {
      "entropy": 5.509053802490234,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21918126791715623,
      "num_tokens": 47416446.0,
      "step": 21245
    },
    {
      "entropy": 5.527863645553589,
      "epoch": 2.3865895434379736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004437638388466979,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2281848207116127,
      "num_tokens": 47426588.0,
      "step": 21250
    },
    {
      "entropy": 5.563670587539673,
      "epoch": 2.3871511203459312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20869171917438506,
      "num_tokens": 47438015.0,
      "step": 21255
    },
    {
      "entropy": 5.67417573928833,
      "epoch": 2.387712697253889,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.162,
      "mean_token_accuracy": 0.20388291627168656,
      "num_tokens": 47448953.0,
      "step": 21260
    },
    {
      "entropy": 5.631828117370605,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.0139,
      "mean_token_accuracy": 0.22232778519392013,
      "num_tokens": 47459669.0,
      "step": 21265
    },
    {
      "entropy": 5.6516917705535885,
      "epoch": 2.3888358510698042,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21675913482904435,
      "num_tokens": 47471018.0,
      "step": 21270
    },
    {
      "entropy": 5.62376823425293,
      "epoch": 2.3893974279777614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21429179906845092,
      "num_tokens": 47481613.0,
      "step": 21275
    },
    {
      "entropy": 5.522831821441651,
      "epoch": 2.389959004885719,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004436044236667603,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22070222645998,
      "num_tokens": 47491245.0,
      "step": 21280
    },
    {
      "entropy": 5.5057518005371096,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004435778356756618,
      "loss": 5.0046,
      "mean_token_accuracy": 0.21995053440332413,
      "num_tokens": 47502308.0,
      "step": 21285
    },
    {
      "entropy": 5.45496973991394,
      "epoch": 2.391082158701634,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044355124231697967,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2207445025444031,
      "num_tokens": 47512799.0,
      "step": 21290
    },
    {
      "entropy": 5.609679555892944,
      "epoch": 2.3916437356095916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.0622,
      "mean_token_accuracy": 0.22070784717798234,
      "num_tokens": 47524740.0,
      "step": 21295
    },
    {
      "entropy": 5.672084665298462,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.1126,
      "mean_token_accuracy": 0.21336732804775238,
      "num_tokens": 47535785.0,
      "step": 21300
    },
    {
      "entropy": 5.554426622390747,
      "epoch": 2.392766889425507,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044347143004389955,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22837549895048143,
      "num_tokens": 47545698.0,
      "step": 21305
    },
    {
      "entropy": 5.490638780593872,
      "epoch": 2.393328466333464,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.0657,
      "mean_token_accuracy": 0.22429481595754625,
      "num_tokens": 47559786.0,
      "step": 21310
    },
    {
      "entropy": 5.53627200126648,
      "epoch": 2.393890043241422,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.012,
      "mean_token_accuracy": 0.22180584222078323,
      "num_tokens": 47570267.0,
      "step": 21315
    },
    {
      "entropy": 5.672016239166259,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433915694930594,
      "loss": 5.1525,
      "mean_token_accuracy": 0.21067707538604735,
      "num_tokens": 47580462.0,
      "step": 21320
    },
    {
      "entropy": 5.550629758834839,
      "epoch": 2.395013197057337,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004433649385850099,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2179287001490593,
      "num_tokens": 47590651.0,
      "step": 21325
    },
    {
      "entropy": 5.658408355712891,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.174,
      "mean_token_accuracy": 0.2140586033463478,
      "num_tokens": 47602263.0,
      "step": 21330
    },
    {
      "entropy": 5.583672714233399,
      "epoch": 2.396136350873252,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004433116606873443,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2257632851600647,
      "num_tokens": 47613570.0,
      "step": 21335
    },
    {
      "entropy": 5.594074058532715,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.0854,
      "mean_token_accuracy": 0.21634993553161622,
      "num_tokens": 47624915.0,
      "step": 21340
    },
    {
      "entropy": 5.591269779205322,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432583613532446,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21544111967086793,
      "num_tokens": 47635974.0,
      "step": 21345
    },
    {
      "entropy": 5.566449689865112,
      "epoch": 2.3978210815971246,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004432317036496533,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2225796639919281,
      "num_tokens": 47647312.0,
      "step": 21350
    },
    {
      "entropy": 5.545644807815552,
      "epoch": 2.3983826585050823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044320504058949916,
      "loss": 4.918,
      "mean_token_accuracy": 0.22245418429374694,
      "num_tokens": 47659224.0,
      "step": 21355
    },
    {
      "entropy": 5.557574081420898,
      "epoch": 2.39894423541304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21603269279003143,
      "num_tokens": 47670720.0,
      "step": 21360
    },
    {
      "entropy": 5.534896755218506,
      "epoch": 2.3995058123209976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.0452,
      "mean_token_accuracy": 0.21488436609506606,
      "num_tokens": 47680646.0,
      "step": 21365
    },
    {
      "entropy": 5.634045743942261,
      "epoch": 2.400067389228955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21605815440416337,
      "num_tokens": 47692155.0,
      "step": 21370
    },
    {
      "entropy": 5.6861328125,
      "epoch": 2.4006289661369125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.1576,
      "mean_token_accuracy": 0.21131511330604552,
      "num_tokens": 47703662.0,
      "step": 21375
    },
    {
      "entropy": 5.562313175201416,
      "epoch": 2.40119054304487,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044307164497000813,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21777312010526656,
      "num_tokens": 47713617.0,
      "step": 21380
    },
    {
      "entropy": 5.566706848144531,
      "epoch": 2.4017521199528273,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044304494978831204,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21898352205753327,
      "num_tokens": 47725051.0,
      "step": 21385
    },
    {
      "entropy": 5.4853196144104,
      "epoch": 2.402313696860785,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004430182492559997,
      "loss": 4.9207,
      "mean_token_accuracy": 0.224828240275383,
      "num_tokens": 47735023.0,
      "step": 21390
    },
    {
      "entropy": 5.54026608467102,
      "epoch": 2.4028752737687427,
      "grad_norm": 0.953125,
      "learning_rate": 0.000442991543373921,
      "loss": 4.977,
      "mean_token_accuracy": 0.2202208921313286,
      "num_tokens": 47746252.0,
      "step": 21395
    },
    {
      "entropy": 5.572974014282226,
      "epoch": 2.4034368506767003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.0999,
      "mean_token_accuracy": 0.21111064106225969,
      "num_tokens": 47756980.0,
      "step": 21400
    },
    {
      "entropy": 5.577722549438477,
      "epoch": 2.4039984275846575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.051,
      "mean_token_accuracy": 0.22032827883958817,
      "num_tokens": 47767633.0,
      "step": 21405
    },
    {
      "entropy": 5.490738344192505,
      "epoch": 2.404560004492615,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004429113936375918,
      "loss": 4.984,
      "mean_token_accuracy": 0.21962652802467347,
      "num_tokens": 47779346.0,
      "step": 21410
    },
    {
      "entropy": 5.660192584991455,
      "epoch": 2.405121581400573,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.11,
      "mean_token_accuracy": 0.21778288930654527,
      "num_tokens": 47791039.0,
      "step": 21415
    },
    {
      "entropy": 5.6411350727081295,
      "epoch": 2.4056831583085305,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.1356,
      "mean_token_accuracy": 0.21318048983812332,
      "num_tokens": 47802096.0,
      "step": 21420
    },
    {
      "entropy": 5.583134746551513,
      "epoch": 2.4062447352164877,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.0305,
      "mean_token_accuracy": 0.2066798374056816,
      "num_tokens": 47814087.0,
      "step": 21425
    },
    {
      "entropy": 5.520400333404541,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004428044524773651,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2149854376912117,
      "num_tokens": 47824656.0,
      "step": 21430
    },
    {
      "entropy": 5.546704387664795,
      "epoch": 2.407367889032403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2269638568162918,
      "num_tokens": 47836045.0,
      "step": 21435
    },
    {
      "entropy": 5.575453662872315,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21632784903049468,
      "num_tokens": 47846494.0,
      "step": 21440
    },
    {
      "entropy": 5.616051292419433,
      "epoch": 2.408491042848318,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044272419050314833,
      "loss": 5.0925,
      "mean_token_accuracy": 0.21852462142705917,
      "num_tokens": 47858689.0,
      "step": 21445
    },
    {
      "entropy": 5.470279741287231,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044269742582980017,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22062879651784897,
      "num_tokens": 47870534.0,
      "step": 21450
    },
    {
      "entropy": 5.5604170799255375,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442670655816901,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2173978418111801,
      "num_tokens": 47880354.0,
      "step": 21455
    },
    {
      "entropy": 5.597535181045532,
      "epoch": 2.410175773572191,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21517990678548812,
      "num_tokens": 47892528.0,
      "step": 21460
    },
    {
      "entropy": 5.618278503417969,
      "epoch": 2.410737350480148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.0818,
      "mean_token_accuracy": 0.22260702401399612,
      "num_tokens": 47904653.0,
      "step": 21465
    },
    {
      "entropy": 5.628297996520996,
      "epoch": 2.411298927388106,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.1421,
      "mean_token_accuracy": 0.20936671644449234,
      "num_tokens": 47916452.0,
      "step": 21470
    },
    {
      "entropy": 5.614164590835571,
      "epoch": 2.4118605042960635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2148926690220833,
      "num_tokens": 47927746.0,
      "step": 21475
    },
    {
      "entropy": 5.5806042671203615,
      "epoch": 2.4124220812040207,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2114206463098526,
      "num_tokens": 47941057.0,
      "step": 21480
    },
    {
      "entropy": 5.551802110671997,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.0931,
      "mean_token_accuracy": 0.21322164982557296,
      "num_tokens": 47951912.0,
      "step": 21485
    },
    {
      "entropy": 5.5867386817932125,
      "epoch": 2.413545235019936,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.04,
      "mean_token_accuracy": 0.21603029519319533,
      "num_tokens": 47963483.0,
      "step": 21490
    },
    {
      "entropy": 5.658647155761718,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.1213,
      "mean_token_accuracy": 0.21130641549825668,
      "num_tokens": 47975047.0,
      "step": 21495
    },
    {
      "entropy": 5.614825010299683,
      "epoch": 2.414668388835851,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.078,
      "mean_token_accuracy": 0.20831698924303055,
      "num_tokens": 47986432.0,
      "step": 21500
    },
    {
      "entropy": 5.542427062988281,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044240266220020455,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2161945343017578,
      "num_tokens": 47997283.0,
      "step": 21505
    },
    {
      "entropy": 5.547086477279663,
      "epoch": 2.415791542651766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2136652573943138,
      "num_tokens": 48008353.0,
      "step": 21510
    },
    {
      "entropy": 5.561220741271972,
      "epoch": 2.416353119559724,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21254303008317948,
      "num_tokens": 48020065.0,
      "step": 21515
    },
    {
      "entropy": 5.5653430938720705,
      "epoch": 2.416914696467681,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21733295023441315,
      "num_tokens": 48030062.0,
      "step": 21520
    },
    {
      "entropy": 5.516132545471192,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004422953154610838,
      "loss": 5.0045,
      "mean_token_accuracy": 0.21048566997051238,
      "num_tokens": 48039639.0,
      "step": 21525
    },
    {
      "entropy": 5.556496572494507,
      "epoch": 2.4180378502835964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.0793,
      "mean_token_accuracy": 0.20929881930351257,
      "num_tokens": 48050428.0,
      "step": 21530
    },
    {
      "entropy": 5.598361587524414,
      "epoch": 2.418599427191554,
      "grad_norm": 1.0,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21500182449817656,
      "num_tokens": 48061214.0,
      "step": 21535
    },
    {
      "entropy": 5.4945845127105715,
      "epoch": 2.4191610040995113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044221474947290205,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2239317327737808,
      "num_tokens": 48071983.0,
      "step": 21540
    },
    {
      "entropy": 5.543382978439331,
      "epoch": 2.419722581007469,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004421878834940029,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2229834958910942,
      "num_tokens": 48083411.0,
      "step": 21545
    },
    {
      "entropy": 5.599159479141235,
      "epoch": 2.4202841579154266,
      "grad_norm": 1.0,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.048,
      "mean_token_accuracy": 0.2117968127131462,
      "num_tokens": 48094718.0,
      "step": 21550
    },
    {
      "entropy": 5.592083501815796,
      "epoch": 2.4208457348233843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21447183042764664,
      "num_tokens": 48106663.0,
      "step": 21555
    },
    {
      "entropy": 5.59105920791626,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.0338,
      "mean_token_accuracy": 0.21801367551088333,
      "num_tokens": 48117232.0,
      "step": 21560
    },
    {
      "entropy": 5.556071949005127,
      "epoch": 2.421968888639299,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.0581,
      "mean_token_accuracy": 0.2218802034854889,
      "num_tokens": 48128817.0,
      "step": 21565
    },
    {
      "entropy": 5.5663083553314205,
      "epoch": 2.422530465547257,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004420534737667157,
      "loss": 5.0075,
      "mean_token_accuracy": 0.22061389982700347,
      "num_tokens": 48140893.0,
      "step": 21570
    },
    {
      "entropy": 5.528827953338623,
      "epoch": 2.423092042455214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044202657586069153,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22064360082149506,
      "num_tokens": 48152661.0,
      "step": 21575
    },
    {
      "entropy": 5.593561267852783,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.0608,
      "mean_token_accuracy": 0.2170109674334526,
      "num_tokens": 48163342.0,
      "step": 21580
    },
    {
      "entropy": 5.55867075920105,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21451059281826018,
      "num_tokens": 48175321.0,
      "step": 21585
    },
    {
      "entropy": 5.599179649353028,
      "epoch": 2.424776773179087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2124024137854576,
      "num_tokens": 48186847.0,
      "step": 21590
    },
    {
      "entropy": 5.609414958953858,
      "epoch": 2.4253383500870442,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.1424,
      "mean_token_accuracy": 0.21877250224351882,
      "num_tokens": 48198840.0,
      "step": 21595
    },
    {
      "entropy": 5.634228181838989,
      "epoch": 2.425899926995002,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.1377,
      "mean_token_accuracy": 0.20519501864910125,
      "num_tokens": 48209948.0,
      "step": 21600
    },
    {
      "entropy": 5.535398340225219,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044186507677252345,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2201506495475769,
      "num_tokens": 48220125.0,
      "step": 21605
    },
    {
      "entropy": 5.573293209075928,
      "epoch": 2.427023080810917,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.022,
      "mean_token_accuracy": 0.21630526185035706,
      "num_tokens": 48230736.0,
      "step": 21610
    },
    {
      "entropy": 5.473100757598877,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004418112012296,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2250608339905739,
      "num_tokens": 48241387.0,
      "step": 21615
    },
    {
      "entropy": 5.582829570770263,
      "epoch": 2.428146234626832,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044178425549070934,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2189294844865799,
      "num_tokens": 48252279.0,
      "step": 21620
    },
    {
      "entropy": 5.641245985031128,
      "epoch": 2.4287078115347898,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.0485,
      "mean_token_accuracy": 0.2186858057975769,
      "num_tokens": 48264235.0,
      "step": 21625
    },
    {
      "entropy": 5.515635776519775,
      "epoch": 2.4292693884427474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044173034808236,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22286104559898376,
      "num_tokens": 48276194.0,
      "step": 21630
    },
    {
      "entropy": 5.6114058017730715,
      "epoch": 2.4298309653507046,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.0737,
      "mean_token_accuracy": 0.22069767266511917,
      "num_tokens": 48287129.0,
      "step": 21635
    },
    {
      "entropy": 5.528564071655273,
      "epoch": 2.4303925422586623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.0272,
      "mean_token_accuracy": 0.20962201207876205,
      "num_tokens": 48298850.0,
      "step": 21640
    },
    {
      "entropy": 5.629068088531494,
      "epoch": 2.43095411916662,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21317614763975143,
      "num_tokens": 48309968.0,
      "step": 21645
    },
    {
      "entropy": 5.587669754028321,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.007,
      "mean_token_accuracy": 0.21889949440956116,
      "num_tokens": 48320419.0,
      "step": 21650
    },
    {
      "entropy": 5.585038518905639,
      "epoch": 2.432077272982535,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044159548667323126,
      "loss": 5.0815,
      "mean_token_accuracy": 0.21158651411533355,
      "num_tokens": 48332611.0,
      "step": 21655
    },
    {
      "entropy": 5.551861715316773,
      "epoch": 2.4326388498904925,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.0544,
      "mean_token_accuracy": 0.22034801095724105,
      "num_tokens": 48344551.0,
      "step": 21660
    },
    {
      "entropy": 5.661687707901001,
      "epoch": 2.43320042679845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.0655,
      "mean_token_accuracy": 0.2153947427868843,
      "num_tokens": 48355142.0,
      "step": 21665
    },
    {
      "entropy": 5.545355749130249,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21683503240346907,
      "num_tokens": 48367342.0,
      "step": 21670
    },
    {
      "entropy": 5.628384304046631,
      "epoch": 2.434323580614365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.081,
      "mean_token_accuracy": 0.2121064245700836,
      "num_tokens": 48378608.0,
      "step": 21675
    },
    {
      "entropy": 5.609715461730957,
      "epoch": 2.4348851575223227,
      "grad_norm": 1.0,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21675150096416473,
      "num_tokens": 48390455.0,
      "step": 21680
    },
    {
      "entropy": 5.583488845825196,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.0239,
      "mean_token_accuracy": 0.21661816388368607,
      "num_tokens": 48401723.0,
      "step": 21685
    },
    {
      "entropy": 5.577321863174438,
      "epoch": 2.4360083113382376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21117826849222182,
      "num_tokens": 48414208.0,
      "step": 21690
    },
    {
      "entropy": 5.542520904541016,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2179997205734253,
      "num_tokens": 48425978.0,
      "step": 21695
    },
    {
      "entropy": 5.530568361282349,
      "epoch": 2.437131465154153,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004413524020277388,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2274817109107971,
      "num_tokens": 48435988.0,
      "step": 21700
    },
    {
      "entropy": 5.551613187789917,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004413253661275875,
      "loss": 5.026,
      "mean_token_accuracy": 0.2165103778243065,
      "num_tokens": 48447217.0,
      "step": 21705
    },
    {
      "entropy": 5.567695236206054,
      "epoch": 2.438254618970068,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044129832493157195,
      "loss": 5.0765,
      "mean_token_accuracy": 0.2137137159705162,
      "num_tokens": 48458597.0,
      "step": 21710
    },
    {
      "entropy": 5.5495100021362305,
      "epoch": 2.4388161958780255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044127127844055335,
      "loss": 5.0187,
      "mean_token_accuracy": 0.22059957683086395,
      "num_tokens": 48470601.0,
      "step": 21715
    },
    {
      "entropy": 5.565170478820801,
      "epoch": 2.439377772785983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044124422665539273,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21520838737487794,
      "num_tokens": 48481570.0,
      "step": 21720
    },
    {
      "entropy": 5.642107725143433,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.1644,
      "mean_token_accuracy": 0.20991120636463165,
      "num_tokens": 48492146.0,
      "step": 21725
    },
    {
      "entropy": 5.554776000976562,
      "epoch": 2.440500926601898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21089775562286378,
      "num_tokens": 48504644.0,
      "step": 21730
    },
    {
      "entropy": 5.500637578964233,
      "epoch": 2.4410625035098557,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21749120503664016,
      "num_tokens": 48516405.0,
      "step": 21735
    },
    {
      "entropy": 5.509037017822266,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.064,
      "mean_token_accuracy": 0.21374044120311736,
      "num_tokens": 48528759.0,
      "step": 21740
    },
    {
      "entropy": 5.689818286895752,
      "epoch": 2.442185657325771,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20790488868951798,
      "num_tokens": 48540731.0,
      "step": 21745
    },
    {
      "entropy": 5.614796257019043,
      "epoch": 2.442747234233728,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.0752,
      "mean_token_accuracy": 0.21953431218862535,
      "num_tokens": 48553642.0,
      "step": 21750
    },
    {
      "entropy": 5.645333623886108,
      "epoch": 2.443308811141686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.0918,
      "mean_token_accuracy": 0.21427350044250487,
      "num_tokens": 48564596.0,
      "step": 21755
    },
    {
      "entropy": 5.566315984725952,
      "epoch": 2.4438703880496435,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000441027621888392,
      "loss": 5.0851,
      "mean_token_accuracy": 0.21004440039396285,
      "num_tokens": 48574521.0,
      "step": 21760
    },
    {
      "entropy": 5.526814126968384,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044100052249473393,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22029173374176025,
      "num_tokens": 48586299.0,
      "step": 21765
    },
    {
      "entropy": 5.626562929153442,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2180091768503189,
      "num_tokens": 48598366.0,
      "step": 21770
    },
    {
      "entropy": 5.574664497375489,
      "epoch": 2.445555118773516,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21487296968698502,
      "num_tokens": 48609659.0,
      "step": 21775
    },
    {
      "entropy": 5.564121389389038,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044091919260408585,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22222990840673446,
      "num_tokens": 48621623.0,
      "step": 21780
    },
    {
      "entropy": 5.516990613937378,
      "epoch": 2.446678272589431,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004408920720735225,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2110305219888687,
      "num_tokens": 48632711.0,
      "step": 21785
    },
    {
      "entropy": 5.534310531616211,
      "epoch": 2.4472398494973886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044086494626089127,
      "loss": 5.0519,
      "mean_token_accuracy": 0.22054938226938248,
      "num_tokens": 48644518.0,
      "step": 21790
    },
    {
      "entropy": 5.573628711700439,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21454686373472215,
      "num_tokens": 48655261.0,
      "step": 21795
    },
    {
      "entropy": 5.535960912704468,
      "epoch": 2.448363003313304,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044081067879288,
      "loss": 5.0081,
      "mean_token_accuracy": 0.2251231476664543,
      "num_tokens": 48666701.0,
      "step": 21800
    },
    {
      "entropy": 5.479600811004639,
      "epoch": 2.448924580221261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407835371392279,
      "loss": 4.932,
      "mean_token_accuracy": 0.22092928886413574,
      "num_tokens": 48677919.0,
      "step": 21805
    },
    {
      "entropy": 5.534824228286743,
      "epoch": 2.449486157129219,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004407563902069636,
      "loss": 5.0142,
      "mean_token_accuracy": 0.22597133219242097,
      "num_tokens": 48688680.0,
      "step": 21810
    },
    {
      "entropy": 5.6254524230957035,
      "epoch": 2.4500477340371765,
      "grad_norm": 1.0,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21664523631334304,
      "num_tokens": 48701149.0,
      "step": 21815
    },
    {
      "entropy": 5.5314491271972654,
      "epoch": 2.450609310945134,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21720271110534667,
      "num_tokens": 48712236.0,
      "step": 21820
    },
    {
      "entropy": 5.4852988719940186,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004406749177471425,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22478351891040801,
      "num_tokens": 48723923.0,
      "step": 21825
    },
    {
      "entropy": 5.4944854259490965,
      "epoch": 2.451732464761049,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044064774970907516,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22012594789266587,
      "num_tokens": 48734925.0,
      "step": 21830
    },
    {
      "entropy": 5.591140270233154,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004406205763967193,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2223421111702919,
      "num_tokens": 48745758.0,
      "step": 21835
    },
    {
      "entropy": 5.504702043533325,
      "epoch": 2.4528556185769643,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004405933978109398,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21868227124214173,
      "num_tokens": 48756899.0,
      "step": 21840
    },
    {
      "entropy": 5.618100833892822,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2200506016612053,
      "num_tokens": 48766805.0,
      "step": 21845
    },
    {
      "entropy": 5.584695768356323,
      "epoch": 2.453978772392879,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044053902482257265,
      "loss": 5.0462,
      "mean_token_accuracy": 0.22308120727539063,
      "num_tokens": 48779360.0,
      "step": 21850
    },
    {
      "entropy": 5.526779842376709,
      "epoch": 2.454540349300837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000440511830421716,
      "loss": 5.019,
      "mean_token_accuracy": 0.21470218002796174,
      "num_tokens": 48789836.0,
      "step": 21855
    },
    {
      "entropy": 5.472415542602539,
      "epoch": 2.455101926208794,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21347327679395675,
      "num_tokens": 48801327.0,
      "step": 21860
    },
    {
      "entropy": 5.516214990615845,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044045742581098624,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22416805177927018,
      "num_tokens": 48812521.0,
      "step": 21865
    },
    {
      "entropy": 5.549032306671142,
      "epoch": 2.4562250800247094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004404302156028453,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21648556888103485,
      "num_tokens": 48824068.0,
      "step": 21870
    },
    {
      "entropy": 5.541382837295532,
      "epoch": 2.456786656932667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21719935089349746,
      "num_tokens": 48835183.0,
      "step": 21875
    },
    {
      "entropy": 5.55590410232544,
      "epoch": 2.4573482338406243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21651793867349625,
      "num_tokens": 48846616.0,
      "step": 21880
    },
    {
      "entropy": 5.534829711914062,
      "epoch": 2.457909810748582,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21416269689798356,
      "num_tokens": 48858159.0,
      "step": 21885
    },
    {
      "entropy": 5.516644716262817,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004403213221053258,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21950492560863494,
      "num_tokens": 48870363.0,
      "step": 21890
    },
    {
      "entropy": 5.586309719085693,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.088,
      "mean_token_accuracy": 0.21832150369882583,
      "num_tokens": 48882036.0,
      "step": 21895
    },
    {
      "entropy": 5.5996129512786865,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21438906341791153,
      "num_tokens": 48893741.0,
      "step": 21900
    },
    {
      "entropy": 5.645219993591309,
      "epoch": 2.460156118380412,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.1132,
      "mean_token_accuracy": 0.21576764881610871,
      "num_tokens": 48906260.0,
      "step": 21905
    },
    {
      "entropy": 5.648142528533936,
      "epoch": 2.46071769528837,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.0956,
      "mean_token_accuracy": 0.20847468674182892,
      "num_tokens": 48918078.0,
      "step": 21910
    },
    {
      "entropy": 5.590187978744507,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2149769589304924,
      "num_tokens": 48928999.0,
      "step": 21915
    },
    {
      "entropy": 5.508566999435425,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004401578239595501,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21427158415317535,
      "num_tokens": 48940587.0,
      "step": 21920
    },
    {
      "entropy": 5.582724475860596,
      "epoch": 2.4624024260122424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.1354,
      "mean_token_accuracy": 0.21512474715709687,
      "num_tokens": 48951632.0,
      "step": 21925
    },
    {
      "entropy": 5.673077821731567,
      "epoch": 2.4629640029202,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.1724,
      "mean_token_accuracy": 0.21167924106121064,
      "num_tokens": 48963578.0,
      "step": 21930
    },
    {
      "entropy": 5.59275712966919,
      "epoch": 2.4635255798281577,
      "grad_norm": 1.0,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21555898785591127,
      "num_tokens": 48974690.0,
      "step": 21935
    },
    {
      "entropy": 5.631569576263428,
      "epoch": 2.464087156736115,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21649783700704575,
      "num_tokens": 48985871.0,
      "step": 21940
    },
    {
      "entropy": 5.4917933464050295,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004400214308828384,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2246061623096466,
      "num_tokens": 48996475.0,
      "step": 21945
    },
    {
      "entropy": 5.632874584197998,
      "epoch": 2.46521031055203,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.1305,
      "mean_token_accuracy": 0.2143561914563179,
      "num_tokens": 49008900.0,
      "step": 21950
    },
    {
      "entropy": 5.476115608215332,
      "epoch": 2.4657718874599874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004399668368656481,
      "loss": 4.985,
      "mean_token_accuracy": 0.21711484640836715,
      "num_tokens": 49020583.0,
      "step": 21955
    },
    {
      "entropy": 5.64756145477295,
      "epoch": 2.466333464367945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.1098,
      "mean_token_accuracy": 0.2122824251651764,
      "num_tokens": 49031363.0,
      "step": 21960
    },
    {
      "entropy": 5.574328994750976,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21408655792474746,
      "num_tokens": 49043404.0,
      "step": 21965
    },
    {
      "entropy": 5.6341338634490965,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000439884906447476,
      "loss": 5.0792,
      "mean_token_accuracy": 0.2156294494867325,
      "num_tokens": 49054240.0,
      "step": 21970
    },
    {
      "entropy": 5.539603853225708,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.140625,
      "learning_rate": 0.000439857585806927,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22879186272621155,
      "num_tokens": 49064730.0,
      "step": 21975
    },
    {
      "entropy": 5.577368974685669,
      "epoch": 2.4685797719997753,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2165291875600815,
      "num_tokens": 49075218.0,
      "step": 21980
    },
    {
      "entropy": 5.528441858291626,
      "epoch": 2.469141348907733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22311071753501893,
      "num_tokens": 49086182.0,
      "step": 21985
    },
    {
      "entropy": 5.579283905029297,
      "epoch": 2.4697029258156906,
      "grad_norm": 0.875,
      "learning_rate": 0.00043977559239398353,
      "loss": 5.1278,
      "mean_token_accuracy": 0.2018406167626381,
      "num_tokens": 49099054.0,
      "step": 21990
    },
    {
      "entropy": 5.543419218063354,
      "epoch": 2.470264502723648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2209307536482811,
      "num_tokens": 49110234.0,
      "step": 21995
    },
    {
      "entropy": 5.515999031066895,
      "epoch": 2.4708260796316055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397209038846436,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21829309910535813,
      "num_tokens": 49121884.0,
      "step": 22000
    },
    {
      "entropy": 5.529776620864868,
      "epoch": 2.471387656539563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043969355176237207,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21821581721305847,
      "num_tokens": 49133491.0,
      "step": 22005
    },
    {
      "entropy": 5.497562265396118,
      "epoch": 2.471949233447521,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004396661943961941,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21200458109378814,
      "num_tokens": 49145895.0,
      "step": 22010
    },
    {
      "entropy": 5.65522198677063,
      "epoch": 2.472510810355478,
      "grad_norm": 1.125,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.0863,
      "mean_token_accuracy": 0.21225326508283615,
      "num_tokens": 49157840.0,
      "step": 22015
    },
    {
      "entropy": 5.580741500854492,
      "epoch": 2.4730723872634357,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.029,
      "mean_token_accuracy": 0.22168058454990386,
      "num_tokens": 49170391.0,
      "step": 22020
    },
    {
      "entropy": 5.562018156051636,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.046875,
      "learning_rate": 0.000439584090842932,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21814932078123092,
      "num_tokens": 49181721.0,
      "step": 22025
    },
    {
      "entropy": 5.592022323608399,
      "epoch": 2.474195541079351,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.042,
      "mean_token_accuracy": 0.21894780844449996,
      "num_tokens": 49192137.0,
      "step": 22030
    },
    {
      "entropy": 5.645582008361816,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.076,
      "mean_token_accuracy": 0.2210993379354477,
      "num_tokens": 49203260.0,
      "step": 22035
    },
    {
      "entropy": 5.533987426757813,
      "epoch": 2.475318694895266,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004395019401258794,
      "loss": 4.967,
      "mean_token_accuracy": 0.21953040957450867,
      "num_tokens": 49214799.0,
      "step": 22040
    },
    {
      "entropy": 5.557372379302978,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043947454607675464,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2235619306564331,
      "num_tokens": 49224595.0,
      "step": 22045
    },
    {
      "entropy": 5.577523946762085,
      "epoch": 2.476441848711181,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2200823277235031,
      "num_tokens": 49235190.0,
      "step": 22050
    },
    {
      "entropy": 5.517890596389771,
      "epoch": 2.4770034256191384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043941974226857763,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21959805488586426,
      "num_tokens": 49247103.0,
      "step": 22055
    },
    {
      "entropy": 5.66641526222229,
      "epoch": 2.477565002527096,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2110644832253456,
      "num_tokens": 49259874.0,
      "step": 22060
    },
    {
      "entropy": 5.456726932525635,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004393649175196472,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22478228360414504,
      "num_tokens": 49271059.0,
      "step": 22065
    },
    {
      "entropy": 5.511970710754395,
      "epoch": 2.478688156343011,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004393374972945815,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2232671171426773,
      "num_tokens": 49283142.0,
      "step": 22070
    },
    {
      "entropy": 5.594742059707642,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21906349956989288,
      "num_tokens": 49294022.0,
      "step": 22075
    },
    {
      "entropy": 5.5335160255432125,
      "epoch": 2.4798113101589263,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.0281,
      "mean_token_accuracy": 0.21883660107851027,
      "num_tokens": 49305849.0,
      "step": 22080
    },
    {
      "entropy": 5.563440418243408,
      "epoch": 2.480372887066884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043925520522745953,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21739322394132615,
      "num_tokens": 49316957.0,
      "step": 22085
    },
    {
      "entropy": 5.5804221630096436,
      "epoch": 2.480934463974841,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004392277640773553,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22525304853916167,
      "num_tokens": 49329040.0,
      "step": 22090
    },
    {
      "entropy": 5.554364204406738,
      "epoch": 2.481496040882799,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2201864406466484,
      "num_tokens": 49340037.0,
      "step": 22095
    },
    {
      "entropy": 5.510327768325806,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21222413331270218,
      "num_tokens": 49350059.0,
      "step": 22100
    },
    {
      "entropy": 5.519271945953369,
      "epoch": 2.482619194698714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043914540925608426,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2205967590212822,
      "num_tokens": 49360880.0,
      "step": 22105
    },
    {
      "entropy": 5.6320600509643555,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004391179471949213,
      "loss": 4.995,
      "mean_token_accuracy": 0.21839267909526824,
      "num_tokens": 49370725.0,
      "step": 22110
    },
    {
      "entropy": 5.588770389556885,
      "epoch": 2.483742348514629,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.0139,
      "mean_token_accuracy": 0.22032131254673004,
      "num_tokens": 49380991.0,
      "step": 22115
    },
    {
      "entropy": 5.501628780364991,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043906300739673265,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22091714441776275,
      "num_tokens": 49390802.0,
      "step": 22120
    },
    {
      "entropy": 5.561256361007691,
      "epoch": 2.4848655023305444,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043903552966145624,
      "loss": 5.0224,
      "mean_token_accuracy": 0.22130022794008256,
      "num_tokens": 49401507.0,
      "step": 22125
    },
    {
      "entropy": 5.600428581237793,
      "epoch": 2.4854270792385016,
      "grad_norm": 1.03125,
      "learning_rate": 0.000439008046703225,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20870541036128998,
      "num_tokens": 49412022.0,
      "step": 22130
    },
    {
      "entropy": 5.581590700149536,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438980558522914,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2231229856610298,
      "num_tokens": 49423294.0,
      "step": 22135
    },
    {
      "entropy": 5.534031534194947,
      "epoch": 2.486550233054417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004389530651213983,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21621880531311036,
      "num_tokens": 49433915.0,
      "step": 22140
    },
    {
      "entropy": 5.463319206237793,
      "epoch": 2.487111809962374,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21395231336355208,
      "num_tokens": 49445122.0,
      "step": 22145
    },
    {
      "entropy": 5.667027997970581,
      "epoch": 2.487673386870332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.1697,
      "mean_token_accuracy": 0.21674797236919402,
      "num_tokens": 49456276.0,
      "step": 22150
    },
    {
      "entropy": 5.597683429718018,
      "epoch": 2.4882349637782895,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21379594057798385,
      "num_tokens": 49467346.0,
      "step": 22155
    },
    {
      "entropy": 5.55587682723999,
      "epoch": 2.488796540686247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004388430393208005,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22803387194871902,
      "num_tokens": 49477635.0,
      "step": 22160
    },
    {
      "entropy": 5.54143705368042,
      "epoch": 2.4893581175942043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21625258922576904,
      "num_tokens": 49491060.0,
      "step": 22165
    },
    {
      "entropy": 5.607154369354248,
      "epoch": 2.489919694502162,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22447689920663833,
      "num_tokens": 49502551.0,
      "step": 22170
    },
    {
      "entropy": 5.556871604919434,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004387604651906152,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21721475720405578,
      "num_tokens": 49514505.0,
      "step": 22175
    },
    {
      "entropy": 5.481386995315551,
      "epoch": 2.4910428483180773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004387329300509913,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21796065121889113,
      "num_tokens": 49526301.0,
      "step": 22180
    },
    {
      "entropy": 5.502859020233155,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043870538969804733,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22226382195949554,
      "num_tokens": 49536546.0,
      "step": 22185
    },
    {
      "entropy": 5.6154650211334225,
      "epoch": 2.492166002133992,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21344098299741746,
      "num_tokens": 49548046.0,
      "step": 22190
    },
    {
      "entropy": 5.53171010017395,
      "epoch": 2.49272757904195,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00043865029335570725,
      "loss": 4.9756,
      "mean_token_accuracy": 0.21708943843841552,
      "num_tokens": 49560001.0,
      "step": 22195
    },
    {
      "entropy": 5.536324501037598,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004386227373680655,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2259240999817848,
      "num_tokens": 49571111.0,
      "step": 22200
    },
    {
      "entropy": 5.50006890296936,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004385951761706122,
      "loss": 4.967,
      "mean_token_accuracy": 0.22134612649679183,
      "num_tokens": 49582888.0,
      "step": 22205
    },
    {
      "entropy": 5.497315454483032,
      "epoch": 2.4944123097658224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043856760976422506,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21994670629501342,
      "num_tokens": 49593710.0,
      "step": 22210
    },
    {
      "entropy": 5.606324291229248,
      "epoch": 2.49497388667378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.0436,
      "mean_token_accuracy": 0.217649382352829,
      "num_tokens": 49604263.0,
      "step": 22215
    },
    {
      "entropy": 5.666529655456543,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.1167,
      "mean_token_accuracy": 0.21253093630075454,
      "num_tokens": 49615152.0,
      "step": 22220
    },
    {
      "entropy": 5.642738914489746,
      "epoch": 2.496097040489695,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20781790614128112,
      "num_tokens": 49627778.0,
      "step": 22225
    },
    {
      "entropy": 5.602905750274658,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.125,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2169175326824188,
      "num_tokens": 49637701.0,
      "step": 22230
    },
    {
      "entropy": 5.540108251571655,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004384296996290076,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22570626586675643,
      "num_tokens": 49647680.0,
      "step": 22235
    },
    {
      "entropy": 5.614991950988769,
      "epoch": 2.4977817712135675,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.1413,
      "mean_token_accuracy": 0.21307619363069535,
      "num_tokens": 49659419.0,
      "step": 22240
    },
    {
      "entropy": 5.577481269836426,
      "epoch": 2.498343348121525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.0332,
      "mean_token_accuracy": 0.22378612011671067,
      "num_tokens": 49671131.0,
      "step": 22245
    },
    {
      "entropy": 5.572822523117066,
      "epoch": 2.498904925029483,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043834689109687077,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22004782557487487,
      "num_tokens": 49681748.0,
      "step": 22250
    },
    {
      "entropy": 5.651333808898926,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.0294,
      "mean_token_accuracy": 0.22409933507442475,
      "num_tokens": 49693321.0,
      "step": 22255
    },
    {
      "entropy": 5.565788269042969,
      "epoch": 2.500028078845398,
      "grad_norm": 1.078125,
      "learning_rate": 0.000438291659402173,
      "loss": 5.0471,
      "mean_token_accuracy": 0.22087272852659226,
      "num_tokens": 49704419.0,
      "step": 22260
    },
    {
      "entropy": 5.576944065093994,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21740758717060088,
      "num_tokens": 49715692.0,
      "step": 22265
    },
    {
      "entropy": 5.643521976470947,
      "epoch": 2.501151232661313,
      "grad_norm": 1.25,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21380020529031754,
      "num_tokens": 49726324.0,
      "step": 22270
    },
    {
      "entropy": 5.488749647140503,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043820877286780224,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22135601788759232,
      "num_tokens": 49737069.0,
      "step": 22275
    },
    {
      "entropy": 5.5412516593933105,
      "epoch": 2.502274386477228,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.0027,
      "mean_token_accuracy": 0.22287641763687133,
      "num_tokens": 49748507.0,
      "step": 22280
    },
    {
      "entropy": 5.590618896484375,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004381534891936024,
      "loss": 4.991,
      "mean_token_accuracy": 0.21585899442434311,
      "num_tokens": 49759017.0,
      "step": 22285
    },
    {
      "entropy": 5.615526723861694,
      "epoch": 2.503397540293143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21813387125730516,
      "num_tokens": 49771353.0,
      "step": 22290
    },
    {
      "entropy": 5.663575887680054,
      "epoch": 2.5039591172011004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.1717,
      "mean_token_accuracy": 0.20551099479198456,
      "num_tokens": 49782337.0,
      "step": 22295
    },
    {
      "entropy": 5.600068855285644,
      "epoch": 2.504520694109058,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043807052472297626,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22347350418567657,
      "num_tokens": 49794212.0,
      "step": 22300
    },
    {
      "entropy": 5.503786563873291,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004380428595138024,
      "loss": 5.0016,
      "mean_token_accuracy": 0.22484295815229416,
      "num_tokens": 49804602.0,
      "step": 22305
    },
    {
      "entropy": 5.5869239330291744,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.1446,
      "mean_token_accuracy": 0.20636141747236253,
      "num_tokens": 49816277.0,
      "step": 22310
    },
    {
      "entropy": 5.601789712905884,
      "epoch": 2.506205424832931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043798751352229227,
      "loss": 5.0221,
      "mean_token_accuracy": 0.22358338087797164,
      "num_tokens": 49827851.0,
      "step": 22315
    },
    {
      "entropy": 5.511046981811523,
      "epoch": 2.5067670017408883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21214187145233154,
      "num_tokens": 49840492.0,
      "step": 22320
    },
    {
      "entropy": 5.461394119262695,
      "epoch": 2.507328578648846,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043793214677243967,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22560703456401826,
      "num_tokens": 49851427.0,
      "step": 22325
    },
    {
      "entropy": 5.549127864837646,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379044556153384,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21855543106794356,
      "num_tokens": 49862883.0,
      "step": 22330
    },
    {
      "entropy": 5.5366401195526125,
      "epoch": 2.508451732464761,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043787675927129625,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21636113077402114,
      "num_tokens": 49874120.0,
      "step": 22335
    },
    {
      "entropy": 5.64756817817688,
      "epoch": 2.5090133093727185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043784905774119483,
      "loss": 5.049,
      "mean_token_accuracy": 0.22350627630949021,
      "num_tokens": 49885961.0,
      "step": 22340
    },
    {
      "entropy": 5.488274717330933,
      "epoch": 2.509574886280676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043782135102591626,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22132514417171478,
      "num_tokens": 49896821.0,
      "step": 22345
    },
    {
      "entropy": 5.430840063095093,
      "epoch": 2.510136463188634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004377936391263427,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2180612027645111,
      "num_tokens": 49908148.0,
      "step": 22350
    },
    {
      "entropy": 5.577600336074829,
      "epoch": 2.5106980400965915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.0242,
      "mean_token_accuracy": 0.22057224810123444,
      "num_tokens": 49919487.0,
      "step": 22355
    },
    {
      "entropy": 5.530239009857178,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004377381997778403,
      "loss": 4.998,
      "mean_token_accuracy": 0.22135665267705917,
      "num_tokens": 49930722.0,
      "step": 22360
    },
    {
      "entropy": 5.6271885395050045,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20893642902374268,
      "num_tokens": 49941917.0,
      "step": 22365
    },
    {
      "entropy": 5.618588495254516,
      "epoch": 2.512382770820464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.0592,
      "mean_token_accuracy": 0.21162324547767639,
      "num_tokens": 49952681.0,
      "step": 22370
    },
    {
      "entropy": 5.60569543838501,
      "epoch": 2.5129443477284212,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21946488469839096,
      "num_tokens": 49965542.0,
      "step": 22375
    },
    {
      "entropy": 5.586624002456665,
      "epoch": 2.513505924636379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.1065,
      "mean_token_accuracy": 0.21076854318380356,
      "num_tokens": 49977506.0,
      "step": 22380
    },
    {
      "entropy": 5.563293695449829,
      "epoch": 2.5140675015443366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004375995107432077,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21394298523664473,
      "num_tokens": 49988674.0,
      "step": 22385
    },
    {
      "entropy": 5.567740201950073,
      "epoch": 2.5146290784522938,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21503067314624785,
      "num_tokens": 50001618.0,
      "step": 22390
    },
    {
      "entropy": 5.557821035385132,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004375439988825508,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21913586109876632,
      "num_tokens": 50012556.0,
      "step": 22395
    },
    {
      "entropy": 5.583634519577027,
      "epoch": 2.515752232268209,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004375162351885845,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2204224333167076,
      "num_tokens": 50023399.0,
      "step": 22400
    },
    {
      "entropy": 5.50512318611145,
      "epoch": 2.5163138091761668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043748846632003823,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2242409586906433,
      "num_tokens": 50035209.0,
      "step": 22405
    },
    {
      "entropy": 5.500016927719116,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.0354,
      "mean_token_accuracy": 0.213275770843029,
      "num_tokens": 50046383.0,
      "step": 22410
    },
    {
      "entropy": 5.532997751235962,
      "epoch": 2.5174369629920816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.0448,
      "mean_token_accuracy": 0.22128871083259583,
      "num_tokens": 50058785.0,
      "step": 22415
    },
    {
      "entropy": 5.637943267822266,
      "epoch": 2.5179985399000393,
      "grad_norm": 1.0,
      "learning_rate": 0.00043740512867576224,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21750791668891906,
      "num_tokens": 50069629.0,
      "step": 22420
    },
    {
      "entropy": 5.598317003250122,
      "epoch": 2.518560116807997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004373773391177395,
      "loss": 4.991,
      "mean_token_accuracy": 0.21792084723711014,
      "num_tokens": 50080923.0,
      "step": 22425
    },
    {
      "entropy": 5.544667434692383,
      "epoch": 2.519121693715954,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004373495443895593,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2196334034204483,
      "num_tokens": 50092133.0,
      "step": 22430
    },
    {
      "entropy": 5.476304817199707,
      "epoch": 2.519683270623912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004373217444921066,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22069301158189775,
      "num_tokens": 50104001.0,
      "step": 22435
    },
    {
      "entropy": 5.564997339248658,
      "epoch": 2.5202448475318695,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.0495,
      "mean_token_accuracy": 0.22318397760391234,
      "num_tokens": 50115908.0,
      "step": 22440
    },
    {
      "entropy": 5.611440420150757,
      "epoch": 2.520806424439827,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043726612919292476,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21328575611114503,
      "num_tokens": 50127380.0,
      "step": 22445
    },
    {
      "entropy": 5.632675504684448,
      "epoch": 2.521368001347785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004372383137929663,
      "loss": 4.96,
      "mean_token_accuracy": 0.22377159893512727,
      "num_tokens": 50138458.0,
      "step": 22450
    },
    {
      "entropy": 5.604140520095825,
      "epoch": 2.521929578255742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.1239,
      "mean_token_accuracy": 0.2125033989548683,
      "num_tokens": 50148814.0,
      "step": 22455
    },
    {
      "entropy": 5.648616981506348,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.1263,
      "mean_token_accuracy": 0.20774720162153243,
      "num_tokens": 50159233.0,
      "step": 22460
    },
    {
      "entropy": 5.532350015640259,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21927342116832732,
      "num_tokens": 50170752.0,
      "step": 22465
    },
    {
      "entropy": 5.522514772415161,
      "epoch": 2.5236143089796146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004371270005446833,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2168848603963852,
      "num_tokens": 50183759.0,
      "step": 22470
    },
    {
      "entropy": 5.566628932952881,
      "epoch": 2.5241758858875722,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004370991593249305,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2195198193192482,
      "num_tokens": 50194091.0,
      "step": 22475
    },
    {
      "entropy": 5.611344909667968,
      "epoch": 2.52473746279553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22107229083776475,
      "num_tokens": 50204591.0,
      "step": 22480
    },
    {
      "entropy": 5.5307446956634525,
      "epoch": 2.525299039703487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004370434614024113,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22079370319843292,
      "num_tokens": 50215809.0,
      "step": 22485
    },
    {
      "entropy": 5.544706773757935,
      "epoch": 2.525860616611445,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.07,
      "mean_token_accuracy": 0.21397980004549028,
      "num_tokens": 50227038.0,
      "step": 22490
    },
    {
      "entropy": 5.5876542091369625,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004369877428417856,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2230173885822296,
      "num_tokens": 50237001.0,
      "step": 22495
    },
    {
      "entropy": 5.569621276855469,
      "epoch": 2.52698377042736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004369598758244003,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22372982054948806,
      "num_tokens": 50248389.0,
      "step": 22500
    },
    {
      "entropy": 5.51845874786377,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043693200365014985,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22385300397872926,
      "num_tokens": 50259676.0,
      "step": 22505
    },
    {
      "entropy": 5.643263721466065,
      "epoch": 2.528106924243275,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.1147,
      "mean_token_accuracy": 0.21736001819372178,
      "num_tokens": 50270707.0,
      "step": 22510
    },
    {
      "entropy": 5.554719305038452,
      "epoch": 2.5286685011512327,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043687624383460306,
      "loss": 4.8803,
      "mean_token_accuracy": 0.23037693947553634,
      "num_tokens": 50282187.0,
      "step": 22515
    },
    {
      "entropy": 5.572443532943725,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.0554,
      "mean_token_accuracy": 0.22256208062171937,
      "num_tokens": 50293672.0,
      "step": 22520
    },
    {
      "entropy": 5.512158346176148,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.0,
      "learning_rate": 0.0004368204634022469,
      "loss": 4.9051,
      "mean_token_accuracy": 0.21990718990564345,
      "num_tokens": 50305235.0,
      "step": 22525
    },
    {
      "entropy": 5.583636665344239,
      "epoch": 2.530353231875105,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.07,
      "mean_token_accuracy": 0.21971380114555358,
      "num_tokens": 50316178.0,
      "step": 22530
    },
    {
      "entropy": 5.526670360565186,
      "epoch": 2.530914808783063,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004367646623601856,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22527377009391786,
      "num_tokens": 50326713.0,
      "step": 22535
    },
    {
      "entropy": 5.587937211990356,
      "epoch": 2.5314763856910205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.1487,
      "mean_token_accuracy": 0.21358381062746049,
      "num_tokens": 50338648.0,
      "step": 22540
    },
    {
      "entropy": 5.59843077659607,
      "epoch": 2.532037962598978,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004367088407155261,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22221237123012544,
      "num_tokens": 50350134.0,
      "step": 22545
    },
    {
      "entropy": 5.526543998718262,
      "epoch": 2.5325995395069354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004366809221694437,
      "loss": 4.9485,
      "mean_token_accuracy": 0.2282531291246414,
      "num_tokens": 50361101.0,
      "step": 22550
    },
    {
      "entropy": 5.4444053173065186,
      "epoch": 2.533161116414893,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043665299847537805,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21608720421791078,
      "num_tokens": 50372376.0,
      "step": 22555
    },
    {
      "entropy": 5.545334720611573,
      "epoch": 2.5337226933228507,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043662506963421815,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21991725116968155,
      "num_tokens": 50383494.0,
      "step": 22560
    },
    {
      "entropy": 5.618294954299927,
      "epoch": 2.534284270230808,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21738839745521546,
      "num_tokens": 50394720.0,
      "step": 22565
    },
    {
      "entropy": 5.637126064300537,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21412317752838134,
      "num_tokens": 50407109.0,
      "step": 22570
    },
    {
      "entropy": 5.6488104343414305,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.09375,
      "learning_rate": 0.000436541252237067,
      "loss": 5.1599,
      "mean_token_accuracy": 0.21271709054708482,
      "num_tokens": 50418664.0,
      "step": 22575
    },
    {
      "entropy": 5.522941398620605,
      "epoch": 2.5359690009546805,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004365133028164248,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2166971892118454,
      "num_tokens": 50431069.0,
      "step": 22580
    },
    {
      "entropy": 5.564639949798584,
      "epoch": 2.536530577862638,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364853482531363,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22059048116207122,
      "num_tokens": 50441484.0,
      "step": 22585
    },
    {
      "entropy": 5.52066388130188,
      "epoch": 2.537092154770596,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043645738854809177,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2231410712003708,
      "num_tokens": 50452594.0,
      "step": 22590
    },
    {
      "entropy": 5.630251932144165,
      "epoch": 2.5376537316785535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21055519133806228,
      "num_tokens": 50463342.0,
      "step": 22595
    },
    {
      "entropy": 5.625582122802735,
      "epoch": 2.538215308586511,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20996002554893495,
      "num_tokens": 50475164.0,
      "step": 22600
    },
    {
      "entropy": 5.605982446670533,
      "epoch": 2.5387768854944683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.092,
      "mean_token_accuracy": 0.2129226267337799,
      "num_tokens": 50486793.0,
      "step": 22605
    },
    {
      "entropy": 5.507909727096558,
      "epoch": 2.539338462402426,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21978907734155656,
      "num_tokens": 50499202.0,
      "step": 22610
    },
    {
      "entropy": 5.687625074386597,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.1183,
      "mean_token_accuracy": 0.21065786927938462,
      "num_tokens": 50511801.0,
      "step": 22615
    },
    {
      "entropy": 5.540392923355102,
      "epoch": 2.540461616218341,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043628952239081113,
      "loss": 4.966,
      "mean_token_accuracy": 0.2114495262503624,
      "num_tokens": 50522918.0,
      "step": 22620
    },
    {
      "entropy": 5.528519630432129,
      "epoch": 2.5410231931262985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043626152671840984,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22285551577806473,
      "num_tokens": 50534027.0,
      "step": 22625
    },
    {
      "entropy": 5.5656932353973385,
      "epoch": 2.541584770034256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000436233525911379,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22190548181533815,
      "num_tokens": 50543814.0,
      "step": 22630
    },
    {
      "entropy": 5.5245973587036135,
      "epoch": 2.542146346942214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004362055199706101,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22198621481657027,
      "num_tokens": 50553818.0,
      "step": 22635
    },
    {
      "entropy": 5.608940267562867,
      "epoch": 2.5427079238501715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.1161,
      "mean_token_accuracy": 0.2087908446788788,
      "num_tokens": 50564526.0,
      "step": 22640
    },
    {
      "entropy": 5.53328218460083,
      "epoch": 2.5432695007581287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043614949269142546,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2241331845521927,
      "num_tokens": 50576410.0,
      "step": 22645
    },
    {
      "entropy": 5.550402355194092,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043612147135479356,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22157868295907973,
      "num_tokens": 50586378.0,
      "step": 22650
    },
    {
      "entropy": 5.563230371475219,
      "epoch": 2.544392654574044,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.0366,
      "mean_token_accuracy": 0.22183768153190614,
      "num_tokens": 50597157.0,
      "step": 22655
    },
    {
      "entropy": 5.533241701126099,
      "epoch": 2.5449542314820013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004360654132919118,
      "loss": 4.973,
      "mean_token_accuracy": 0.22308151721954345,
      "num_tokens": 50607390.0,
      "step": 22660
    },
    {
      "entropy": 5.495513153076172,
      "epoch": 2.545515808389959,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004360373765674467,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22769471108913422,
      "num_tokens": 50618397.0,
      "step": 22665
    },
    {
      "entropy": 5.580721282958985,
      "epoch": 2.5460773852979166,
      "grad_norm": 0.890625,
      "learning_rate": 0.00043600933471548927,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20588489770889282,
      "num_tokens": 50631186.0,
      "step": 22670
    },
    {
      "entropy": 5.591237020492554,
      "epoch": 2.546638962205874,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004359812877369321,
      "loss": 5.0547,
      "mean_token_accuracy": 0.21413103491067886,
      "num_tokens": 50642761.0,
      "step": 22675
    },
    {
      "entropy": 5.51923942565918,
      "epoch": 2.5472005391138315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043595323563266815,
      "loss": 4.8815,
      "mean_token_accuracy": 0.23016863018274308,
      "num_tokens": 50653374.0,
      "step": 22680
    },
    {
      "entropy": 5.486351728439331,
      "epoch": 2.547762116021789,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043592517840359093,
      "loss": 5.0023,
      "mean_token_accuracy": 0.22065573930740356,
      "num_tokens": 50664331.0,
      "step": 22685
    },
    {
      "entropy": 5.601027107238769,
      "epoch": 2.548323692929747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004358971160505936,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21540087312459946,
      "num_tokens": 50675011.0,
      "step": 22690
    },
    {
      "entropy": 5.500874853134155,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043586904857456963,
      "loss": 4.8954,
      "mean_token_accuracy": 0.23109455704689025,
      "num_tokens": 50685139.0,
      "step": 22695
    },
    {
      "entropy": 5.564606714248657,
      "epoch": 2.5494468467456617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043584097597641275,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22105177789926528,
      "num_tokens": 50696610.0,
      "step": 22700
    },
    {
      "entropy": 5.521002292633057,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043581289825701683,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22059139907360076,
      "num_tokens": 50708662.0,
      "step": 22705
    },
    {
      "entropy": 5.4664140224456785,
      "epoch": 2.550570000561577,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004357848154172759,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22761558890342712,
      "num_tokens": 50719457.0,
      "step": 22710
    },
    {
      "entropy": 5.558020210266113,
      "epoch": 2.5511315774695342,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043575672745808404,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22210265696048737,
      "num_tokens": 50730038.0,
      "step": 22715
    },
    {
      "entropy": 5.607824087142944,
      "epoch": 2.551693154377492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.019,
      "mean_token_accuracy": 0.22215860784053804,
      "num_tokens": 50742327.0,
      "step": 22720
    },
    {
      "entropy": 5.606868076324463,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.0792,
      "mean_token_accuracy": 0.21630948334932326,
      "num_tokens": 50753056.0,
      "step": 22725
    },
    {
      "entropy": 5.536335515975952,
      "epoch": 2.552816308193407,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.0574,
      "mean_token_accuracy": 0.21457332819700242,
      "num_tokens": 50763961.0,
      "step": 22730
    },
    {
      "entropy": 5.5477691173553465,
      "epoch": 2.553377885101365,
      "grad_norm": 1.125,
      "learning_rate": 0.00043564432444469686,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22827388495206832,
      "num_tokens": 50774978.0,
      "step": 22735
    },
    {
      "entropy": 5.589822387695312,
      "epoch": 2.553939462009322,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.0438,
      "mean_token_accuracy": 0.2158685028553009,
      "num_tokens": 50786198.0,
      "step": 22740
    },
    {
      "entropy": 5.526922273635864,
      "epoch": 2.5545010389172798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004355880922445584,
      "loss": 4.9985,
      "mean_token_accuracy": 0.2138555184006691,
      "num_tokens": 50797685.0,
      "step": 22745
    },
    {
      "entropy": 5.4268121242523195,
      "epoch": 2.5550626158252374,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043555996847426077,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22677556425333023,
      "num_tokens": 50808518.0,
      "step": 22750
    },
    {
      "entropy": 5.551320219039917,
      "epoch": 2.5556241927331946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.0268,
      "mean_token_accuracy": 0.21556806862354277,
      "num_tokens": 50819241.0,
      "step": 22755
    },
    {
      "entropy": 5.648899078369141,
      "epoch": 2.5561857696411523,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20835838317871094,
      "num_tokens": 50830334.0,
      "step": 22760
    },
    {
      "entropy": 5.680402946472168,
      "epoch": 2.55674734654911,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004354755664932006,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21948817670345305,
      "num_tokens": 50840738.0,
      "step": 22765
    },
    {
      "entropy": 5.646964693069458,
      "epoch": 2.557308923457067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.1442,
      "mean_token_accuracy": 0.21152518540620804,
      "num_tokens": 50853397.0,
      "step": 22770
    },
    {
      "entropy": 5.64355616569519,
      "epoch": 2.557870500365025,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.0251,
      "mean_token_accuracy": 0.2178378462791443,
      "num_tokens": 50865786.0,
      "step": 22775
    },
    {
      "entropy": 5.5809845447540285,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.0481,
      "mean_token_accuracy": 0.21832258254289627,
      "num_tokens": 50876029.0,
      "step": 22780
    },
    {
      "entropy": 5.525030517578125,
      "epoch": 2.55899365418094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043536295898817767,
      "loss": 4.9712,
      "mean_token_accuracy": 0.22169970273971557,
      "num_tokens": 50886941.0,
      "step": 22785
    },
    {
      "entropy": 5.5865295886993405,
      "epoch": 2.559555231088898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004353347943446336,
      "loss": 4.9833,
      "mean_token_accuracy": 0.2203478917479515,
      "num_tokens": 50898231.0,
      "step": 22790
    },
    {
      "entropy": 5.596128177642822,
      "epoch": 2.560116807996855,
      "grad_norm": 1.125,
      "learning_rate": 0.00043530662459596744,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21507314294576646,
      "num_tokens": 50908282.0,
      "step": 22795
    },
    {
      "entropy": 5.541433525085449,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004352784497430762,
      "loss": 4.9355,
      "mean_token_accuracy": 0.23001914620399475,
      "num_tokens": 50918462.0,
      "step": 22800
    },
    {
      "entropy": 5.533170986175537,
      "epoch": 2.5612399618127704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2196831926703453,
      "num_tokens": 50929160.0,
      "step": 22805
    },
    {
      "entropy": 5.640112733840942,
      "epoch": 2.5618015387207276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.1886,
      "mean_token_accuracy": 0.20999986231327056,
      "num_tokens": 50943361.0,
      "step": 22810
    },
    {
      "entropy": 5.6669244289398195,
      "epoch": 2.5623631156286852,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.1096,
      "mean_token_accuracy": 0.2122982144355774,
      "num_tokens": 50954515.0,
      "step": 22815
    },
    {
      "entropy": 5.666681814193725,
      "epoch": 2.562924692536643,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.1055,
      "mean_token_accuracy": 0.2100799098610878,
      "num_tokens": 50965237.0,
      "step": 22820
    },
    {
      "entropy": 5.531550598144531,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351374989466477,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2262813463807106,
      "num_tokens": 50976406.0,
      "step": 22825
    },
    {
      "entropy": 5.439106369018555,
      "epoch": 2.5640478463525582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043510929348725065,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22172590345144272,
      "num_tokens": 50987214.0,
      "step": 22830
    },
    {
      "entropy": 5.556692790985108,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043508108292991156,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21933719515800476,
      "num_tokens": 50997668.0,
      "step": 22835
    },
    {
      "entropy": 5.600296497344971,
      "epoch": 2.565171000168473,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004350528672755287,
      "loss": 5.0106,
      "mean_token_accuracy": 0.22115788906812667,
      "num_tokens": 51008358.0,
      "step": 22840
    },
    {
      "entropy": 5.5975266456604,
      "epoch": 2.5657325770764308,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004350246465250005,
      "loss": 5.002,
      "mean_token_accuracy": 0.2213655188679695,
      "num_tokens": 51019410.0,
      "step": 22845
    },
    {
      "entropy": 5.681488084793091,
      "epoch": 2.566294153984388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043499642067922544,
      "loss": 5.169,
      "mean_token_accuracy": 0.2086511805653572,
      "num_tokens": 51032307.0,
      "step": 22850
    },
    {
      "entropy": 5.703001260757446,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.0449,
      "mean_token_accuracy": 0.21163269132375717,
      "num_tokens": 51043176.0,
      "step": 22855
    },
    {
      "entropy": 5.599759578704834,
      "epoch": 2.5674173078003033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043493995370553,
      "loss": 5.0565,
      "mean_token_accuracy": 0.211974661052227,
      "num_tokens": 51053977.0,
      "step": 22860
    },
    {
      "entropy": 5.552986669540405,
      "epoch": 2.5679788847082605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21619749665260315,
      "num_tokens": 51064801.0,
      "step": 22865
    },
    {
      "entropy": 5.592008209228515,
      "epoch": 2.568540461616218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043488346636163405,
      "loss": 4.9787,
      "mean_token_accuracy": 0.228873211145401,
      "num_tokens": 51075400.0,
      "step": 22870
    },
    {
      "entropy": 5.612654733657837,
      "epoch": 2.569102038524176,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20877302139997483,
      "num_tokens": 51087077.0,
      "step": 22875
    },
    {
      "entropy": 5.511624002456665,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043482695865473207,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21470762193202972,
      "num_tokens": 51099798.0,
      "step": 22880
    },
    {
      "entropy": 5.67081503868103,
      "epoch": 2.570225192340091,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.1458,
      "mean_token_accuracy": 0.21029067188501357,
      "num_tokens": 51111880.0,
      "step": 22885
    },
    {
      "entropy": 5.599918079376221,
      "epoch": 2.5707867692480484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.032,
      "mean_token_accuracy": 0.21171327978372573,
      "num_tokens": 51122358.0,
      "step": 22890
    },
    {
      "entropy": 5.550173664093018,
      "epoch": 2.571348346156006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.0719,
      "mean_token_accuracy": 0.220253287255764,
      "num_tokens": 51132775.0,
      "step": 22895
    },
    {
      "entropy": 5.569885969161987,
      "epoch": 2.5719099230639637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.0924,
      "mean_token_accuracy": 0.21865515112876893,
      "num_tokens": 51144974.0,
      "step": 22900
    },
    {
      "entropy": 5.586109209060669,
      "epoch": 2.572471499971921,
      "grad_norm": 1.015625,
      "learning_rate": 0.000434685600346561,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2285795569419861,
      "num_tokens": 51154700.0,
      "step": 22905
    },
    {
      "entropy": 5.626238536834717,
      "epoch": 2.5730330768798786,
      "grad_norm": 1.0,
      "learning_rate": 0.000434657313427971,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21202978789806365,
      "num_tokens": 51166910.0,
      "step": 22910
    },
    {
      "entropy": 5.527895069122314,
      "epoch": 2.5735946537878363,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004346290214258301,
      "loss": 4.963,
      "mean_token_accuracy": 0.21968026608228683,
      "num_tokens": 51177620.0,
      "step": 22915
    },
    {
      "entropy": 5.610811710357666,
      "epoch": 2.574156230695794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043460072434103934,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22617247700691223,
      "num_tokens": 51187894.0,
      "step": 22920
    },
    {
      "entropy": 5.554599189758301,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004345724221744996,
      "loss": 5.0563,
      "mean_token_accuracy": 0.2136721581220627,
      "num_tokens": 51199625.0,
      "step": 22925
    },
    {
      "entropy": 5.505504560470581,
      "epoch": 2.575279384511709,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043454411492711197,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21914729177951814,
      "num_tokens": 51211932.0,
      "step": 22930
    },
    {
      "entropy": 5.566793537139892,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.072,
      "mean_token_accuracy": 0.21639672070741653,
      "num_tokens": 51223196.0,
      "step": 22935
    },
    {
      "entropy": 5.623670768737793,
      "epoch": 2.576402538327624,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21516597419977188,
      "num_tokens": 51234246.0,
      "step": 22940
    },
    {
      "entropy": 5.596416807174682,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043445916270887626,
      "loss": 5.0002,
      "mean_token_accuracy": 0.217895071208477,
      "num_tokens": 51245056.0,
      "step": 22945
    },
    {
      "entropy": 5.5926331043243405,
      "epoch": 2.577525692143539,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2175916999578476,
      "num_tokens": 51256513.0,
      "step": 22950
    },
    {
      "entropy": 5.496473169326782,
      "epoch": 2.5780872690514967,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043440250250900854,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22021920382976531,
      "num_tokens": 51267609.0,
      "step": 22955
    },
    {
      "entropy": 5.6372575759887695,
      "epoch": 2.578648845959454,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.1277,
      "mean_token_accuracy": 0.2133200138807297,
      "num_tokens": 51277940.0,
      "step": 22960
    },
    {
      "entropy": 5.698448371887207,
      "epoch": 2.5792104228674115,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.115,
      "mean_token_accuracy": 0.2240264296531677,
      "num_tokens": 51288853.0,
      "step": 22965
    },
    {
      "entropy": 5.583156585693359,
      "epoch": 2.579771999775369,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.1355,
      "mean_token_accuracy": 0.21139006018638612,
      "num_tokens": 51299197.0,
      "step": 22970
    },
    {
      "entropy": 5.577366542816162,
      "epoch": 2.580333576683327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.1281,
      "mean_token_accuracy": 0.2098861366510391,
      "num_tokens": 51310821.0,
      "step": 22975
    },
    {
      "entropy": 5.7079774856567385,
      "epoch": 2.5808951535912845,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043426076320497557,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20675030946731568,
      "num_tokens": 51321805.0,
      "step": 22980
    },
    {
      "entropy": 5.510569047927857,
      "epoch": 2.5814567304992417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043423240012778437,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22075916528701783,
      "num_tokens": 51332647.0,
      "step": 22985
    },
    {
      "entropy": 5.451830244064331,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004342040319805717,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2221942126750946,
      "num_tokens": 51344715.0,
      "step": 22990
    },
    {
      "entropy": 5.592878007888794,
      "epoch": 2.582579884315157,
      "grad_norm": 1.046875,
      "learning_rate": 0.000434175658764241,
      "loss": 5.1201,
      "mean_token_accuracy": 0.21760886162519455,
      "num_tokens": 51355481.0,
      "step": 22995
    },
    {
      "entropy": 5.5855035305023195,
      "epoch": 2.5831414612231143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043414728047969553,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22976448982954026,
      "num_tokens": 51367098.0,
      "step": 23000
    },
    {
      "entropy": 5.5904542922973635,
      "epoch": 2.583703038131072,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.1357,
      "mean_token_accuracy": 0.2107250437140465,
      "num_tokens": 51379948.0,
      "step": 23005
    },
    {
      "entropy": 5.538628625869751,
      "epoch": 2.5842646150390296,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043409050870957505,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22268450260162354,
      "num_tokens": 51391675.0,
      "step": 23010
    },
    {
      "entropy": 5.502814054489136,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340621152258075,
      "loss": 4.92,
      "mean_token_accuracy": 0.2220867857336998,
      "num_tokens": 51401809.0,
      "step": 23015
    },
    {
      "entropy": 5.551659822463989,
      "epoch": 2.585387768854945,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21870362311601638,
      "num_tokens": 51413184.0,
      "step": 23020
    },
    {
      "entropy": 5.586262321472168,
      "epoch": 2.585949345762902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2117658242583275,
      "num_tokens": 51424751.0,
      "step": 23025
    },
    {
      "entropy": 5.514336252212525,
      "epoch": 2.58651092267086,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004339769043905255,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2214018240571022,
      "num_tokens": 51436550.0,
      "step": 23030
    },
    {
      "entropy": 5.597496795654297,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.022,
      "mean_token_accuracy": 0.2167671501636505,
      "num_tokens": 51447691.0,
      "step": 23035
    },
    {
      "entropy": 5.561586236953735,
      "epoch": 2.5876340764867747,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043392007185606533,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21641522645950317,
      "num_tokens": 51459715.0,
      "step": 23040
    },
    {
      "entropy": 5.550105619430542,
      "epoch": 2.5881956533947323,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004338916479982678,
      "loss": 4.9448,
      "mean_token_accuracy": 0.23015495836734773,
      "num_tokens": 51470541.0,
      "step": 23045
    },
    {
      "entropy": 5.554199314117431,
      "epoch": 2.58875723030269,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043386321908129856,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22763636708259583,
      "num_tokens": 51480724.0,
      "step": 23050
    },
    {
      "entropy": 5.681247043609619,
      "epoch": 2.589318807210647,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.1949,
      "mean_token_accuracy": 0.20930216908454896,
      "num_tokens": 51492204.0,
      "step": 23055
    },
    {
      "entropy": 5.591548442840576,
      "epoch": 2.589880384118605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.0518,
      "mean_token_accuracy": 0.20921705663204193,
      "num_tokens": 51503354.0,
      "step": 23060
    },
    {
      "entropy": 5.633863925933838,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2222941815853119,
      "num_tokens": 51513563.0,
      "step": 23065
    },
    {
      "entropy": 5.607774925231934,
      "epoch": 2.59100353793452,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21669528335332872,
      "num_tokens": 51524996.0,
      "step": 23070
    },
    {
      "entropy": 5.492296552658081,
      "epoch": 2.591565114842478,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004337209986405639,
      "loss": 4.9315,
      "mean_token_accuracy": 0.22678934335708617,
      "num_tokens": 51536822.0,
      "step": 23075
    },
    {
      "entropy": 5.488516569137573,
      "epoch": 2.592126691750435,
      "grad_norm": 1.125,
      "learning_rate": 0.000433692539387579,
      "loss": 4.9878,
      "mean_token_accuracy": 0.22353372126817703,
      "num_tokens": 51547545.0,
      "step": 23080
    },
    {
      "entropy": 5.573237371444702,
      "epoch": 2.5926882686583927,
      "grad_norm": 1.0,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.0588,
      "mean_token_accuracy": 0.21916783154010772,
      "num_tokens": 51560721.0,
      "step": 23085
    },
    {
      "entropy": 5.5374853134155275,
      "epoch": 2.5932498455663504,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043363560572401953,
      "loss": 4.9016,
      "mean_token_accuracy": 0.2240646243095398,
      "num_tokens": 51573226.0,
      "step": 23090
    },
    {
      "entropy": 5.580333852767945,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2188219204545021,
      "num_tokens": 51584338.0,
      "step": 23095
    },
    {
      "entropy": 5.5571019649505615,
      "epoch": 2.5943729993822653,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.0156,
      "mean_token_accuracy": 0.22155146896839142,
      "num_tokens": 51593979.0,
      "step": 23100
    },
    {
      "entropy": 5.5311380386352536,
      "epoch": 2.594934576290223,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004335501673483033,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22291422337293626,
      "num_tokens": 51604713.0,
      "step": 23105
    },
    {
      "entropy": 5.574566793441773,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043352167779192423,
      "loss": 5.0661,
      "mean_token_accuracy": 0.22262163162231446,
      "num_tokens": 51616034.0,
      "step": 23110
    },
    {
      "entropy": 5.47004804611206,
      "epoch": 2.5960577301061383,
      "grad_norm": 1.0,
      "learning_rate": 0.00043349318318815363,
      "loss": 4.9137,
      "mean_token_accuracy": 0.21750346571207047,
      "num_tokens": 51627551.0,
      "step": 23115
    },
    {
      "entropy": 5.562101793289185,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043346468353789853,
      "loss": 5.0125,
      "mean_token_accuracy": 0.2204153761267662,
      "num_tokens": 51639625.0,
      "step": 23120
    },
    {
      "entropy": 5.563265609741211,
      "epoch": 2.597180883922053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004334361788420664,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22809721380472184,
      "num_tokens": 51649918.0,
      "step": 23125
    },
    {
      "entropy": 5.525343370437622,
      "epoch": 2.597742460830011,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043340766910156497,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2303990289568901,
      "num_tokens": 51659527.0,
      "step": 23130
    },
    {
      "entropy": 5.5487383842468265,
      "epoch": 2.598304037737968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.0961,
      "mean_token_accuracy": 0.2114325374364853,
      "num_tokens": 51670704.0,
      "step": 23135
    },
    {
      "entropy": 5.536376476287842,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043335063449018515,
      "loss": 4.9991,
      "mean_token_accuracy": 0.2184857666492462,
      "num_tokens": 51681606.0,
      "step": 23140
    },
    {
      "entropy": 5.632254886627197,
      "epoch": 2.5994271915538834,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.0869,
      "mean_token_accuracy": 0.21845919340848924,
      "num_tokens": 51692592.0,
      "step": 23145
    },
    {
      "entropy": 5.58868145942688,
      "epoch": 2.599988768461841,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21466144174337387,
      "num_tokens": 51704990.0,
      "step": 23150
    },
    {
      "entropy": 5.566945934295655,
      "epoch": 2.6005503453697982,
      "grad_norm": 1.125,
      "learning_rate": 0.00043326504476079425,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21800480782985687,
      "num_tokens": 51715934.0,
      "step": 23155
    },
    {
      "entropy": 5.563785982131958,
      "epoch": 2.601111922277756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.1227,
      "mean_token_accuracy": 0.21141742169857025,
      "num_tokens": 51727314.0,
      "step": 23160
    },
    {
      "entropy": 5.588688755035401,
      "epoch": 2.6016734991857136,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004332079597435842,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22884813100099563,
      "num_tokens": 51738322.0,
      "step": 23165
    },
    {
      "entropy": 5.4961926460266115,
      "epoch": 2.602235076093671,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004331794096784206,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22447630017995834,
      "num_tokens": 51749221.0,
      "step": 23170
    },
    {
      "entropy": 5.49473557472229,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004331508545767633,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21312787234783173,
      "num_tokens": 51760568.0,
      "step": 23175
    },
    {
      "entropy": 5.561031198501587,
      "epoch": 2.603358229909586,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.0547,
      "mean_token_accuracy": 0.21892421394586564,
      "num_tokens": 51772652.0,
      "step": 23180
    },
    {
      "entropy": 5.607164812088013,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.0193,
      "mean_token_accuracy": 0.22160398513078688,
      "num_tokens": 51783739.0,
      "step": 23185
    },
    {
      "entropy": 5.546056175231934,
      "epoch": 2.604481383725501,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2174316018819809,
      "num_tokens": 51795521.0,
      "step": 23190
    },
    {
      "entropy": 5.510391283035278,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004330365838233827,
      "loss": 4.8743,
      "mean_token_accuracy": 0.224968883395195,
      "num_tokens": 51805041.0,
      "step": 23195
    },
    {
      "entropy": 5.579789829254151,
      "epoch": 2.6056045375414163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.0921,
      "mean_token_accuracy": 0.21877775341272354,
      "num_tokens": 51815157.0,
      "step": 23200
    },
    {
      "entropy": 5.626439428329467,
      "epoch": 2.606166114449374,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2208950325846672,
      "num_tokens": 51826760.0,
      "step": 23205
    },
    {
      "entropy": 5.7102015018463135,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20894674360752105,
      "num_tokens": 51838018.0,
      "step": 23210
    },
    {
      "entropy": 5.657157611846924,
      "epoch": 2.607289268265289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.092,
      "mean_token_accuracy": 0.21201003938913346,
      "num_tokens": 51848217.0,
      "step": 23215
    },
    {
      "entropy": 5.6005247116088865,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043289363216969973,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2224353477358818,
      "num_tokens": 51857537.0,
      "step": 23220
    },
    {
      "entropy": 5.5479389190673825,
      "epoch": 2.608412422081204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043286502675313757,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22170309573411942,
      "num_tokens": 51867961.0,
      "step": 23225
    },
    {
      "entropy": 5.531094980239868,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.1137,
      "mean_token_accuracy": 0.21121343225240707,
      "num_tokens": 51879868.0,
      "step": 23230
    },
    {
      "entropy": 5.592844533920288,
      "epoch": 2.609535575897119,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.0403,
      "mean_token_accuracy": 0.21018508225679397,
      "num_tokens": 51890849.0,
      "step": 23235
    },
    {
      "entropy": 5.58262825012207,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000432779180348193,
      "loss": 5.0032,
      "mean_token_accuracy": 0.220571132004261,
      "num_tokens": 51901193.0,
      "step": 23240
    },
    {
      "entropy": 5.598833131790161,
      "epoch": 2.6106587297130344,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043275055483116254,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21914700865745546,
      "num_tokens": 51912551.0,
      "step": 23245
    },
    {
      "entropy": 5.6759504795074465,
      "epoch": 2.6112203066209916,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.2076,
      "mean_token_accuracy": 0.20028226673603058,
      "num_tokens": 51923970.0,
      "step": 23250
    },
    {
      "entropy": 5.4627057075500485,
      "epoch": 2.6117818835289492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004326932887294967,
      "loss": 4.901,
      "mean_token_accuracy": 0.22524838745594025,
      "num_tokens": 51936667.0,
      "step": 23255
    },
    {
      "entropy": 5.546140432357788,
      "epoch": 2.612343460436907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004326646481466848,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2214517518877983,
      "num_tokens": 51947479.0,
      "step": 23260
    },
    {
      "entropy": 5.533769226074218,
      "epoch": 2.6129050373448646,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.0765,
      "mean_token_accuracy": 0.22183000296354294,
      "num_tokens": 51958669.0,
      "step": 23265
    },
    {
      "entropy": 5.5081915855407715,
      "epoch": 2.613466614252822,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043260735192166217,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22501513063907624,
      "num_tokens": 51969273.0,
      "step": 23270
    },
    {
      "entropy": 5.592966938018799,
      "epoch": 2.6140281911607794,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000432578696281276,
      "loss": 5.076,
      "mean_token_accuracy": 0.22122202813625336,
      "num_tokens": 51981220.0,
      "step": 23275
    },
    {
      "entropy": 5.5884363651275635,
      "epoch": 2.614589768068737,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432550035623523,
      "loss": 5.0859,
      "mean_token_accuracy": 0.21749777793884278,
      "num_tokens": 51992768.0,
      "step": 23280
    },
    {
      "entropy": 5.5614711284637455,
      "epoch": 2.6151513449766943,
      "grad_norm": 1.0,
      "learning_rate": 0.0004325213699493157,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21338960528373718,
      "num_tokens": 52002754.0,
      "step": 23285
    },
    {
      "entropy": 5.603413009643555,
      "epoch": 2.615712921884652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004324926992595669,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21750446259975434,
      "num_tokens": 52014503.0,
      "step": 23290
    },
    {
      "entropy": 5.585260629653931,
      "epoch": 2.6162744987926096,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21793373823165893,
      "num_tokens": 52026151.0,
      "step": 23295
    },
    {
      "entropy": 5.544194316864013,
      "epoch": 2.6168360757005673,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004324353428370965,
      "loss": 5.0195,
      "mean_token_accuracy": 0.2133221596479416,
      "num_tokens": 52037704.0,
      "step": 23300
    },
    {
      "entropy": 5.541483402252197,
      "epoch": 2.617397652608525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2215502917766571,
      "num_tokens": 52049326.0,
      "step": 23305
    },
    {
      "entropy": 5.635787296295166,
      "epoch": 2.617959229516482,
      "grad_norm": 1.125,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2204291984438896,
      "num_tokens": 52060030.0,
      "step": 23310
    },
    {
      "entropy": 5.666735982894897,
      "epoch": 2.61852080642444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000432349270609657,
      "loss": 5.1627,
      "mean_token_accuracy": 0.21223157197237014,
      "num_tokens": 52071744.0,
      "step": 23315
    },
    {
      "entropy": 5.630251789093018,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043232056984583527,
      "loss": 5.0029,
      "mean_token_accuracy": 0.2179658368229866,
      "num_tokens": 52082443.0,
      "step": 23320
    },
    {
      "entropy": 5.5396815776824955,
      "epoch": 2.6196439602403547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004322918640728656,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2193633422255516,
      "num_tokens": 52094211.0,
      "step": 23325
    },
    {
      "entropy": 5.502017831802368,
      "epoch": 2.6202055371483124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043226315329166203,
      "loss": 4.94,
      "mean_token_accuracy": 0.22275352478027344,
      "num_tokens": 52104802.0,
      "step": 23330
    },
    {
      "entropy": 5.61159119606018,
      "epoch": 2.62076711405627,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043223443750313873,
      "loss": 5.0661,
      "mean_token_accuracy": 0.21167808026075363,
      "num_tokens": 52115459.0,
      "step": 23335
    },
    {
      "entropy": 5.555730867385864,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21840057522058487,
      "num_tokens": 52125572.0,
      "step": 23340
    },
    {
      "entropy": 5.530693674087525,
      "epoch": 2.621890267872185,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.002,
      "mean_token_accuracy": 0.21787100434303283,
      "num_tokens": 52137405.0,
      "step": 23345
    },
    {
      "entropy": 5.604212379455566,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.1054,
      "mean_token_accuracy": 0.2097350239753723,
      "num_tokens": 52148131.0,
      "step": 23350
    },
    {
      "entropy": 5.587569522857666,
      "epoch": 2.6230134216881003,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004321195242941361,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21802699714899063,
      "num_tokens": 52159361.0,
      "step": 23355
    },
    {
      "entropy": 5.561152362823487,
      "epoch": 2.623574998596058,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043209078348273175,
      "loss": 5.0082,
      "mean_token_accuracy": 0.21006454825401305,
      "num_tokens": 52170521.0,
      "step": 23360
    },
    {
      "entropy": 5.622185897827149,
      "epoch": 2.624136575504015,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.0963,
      "mean_token_accuracy": 0.21495486944913864,
      "num_tokens": 52182069.0,
      "step": 23365
    },
    {
      "entropy": 5.486348247528076,
      "epoch": 2.624698152411973,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043203328685534396,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22593102008104324,
      "num_tokens": 52191867.0,
      "step": 23370
    },
    {
      "entropy": 5.648968505859375,
      "epoch": 2.6252597293199305,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.1894,
      "mean_token_accuracy": 0.2022934928536415,
      "num_tokens": 52204511.0,
      "step": 23375
    },
    {
      "entropy": 5.585955953598022,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043197577022795327,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22495020776987076,
      "num_tokens": 52214276.0,
      "step": 23380
    },
    {
      "entropy": 5.551560306549073,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22370319664478303,
      "num_tokens": 52225388.0,
      "step": 23385
    },
    {
      "entropy": 5.537782382965088,
      "epoch": 2.626944460043803,
      "grad_norm": 1.0,
      "learning_rate": 0.0004319182336078852,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2261159524321556,
      "num_tokens": 52235300.0,
      "step": 23390
    },
    {
      "entropy": 5.628648090362549,
      "epoch": 2.6275060369517607,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431889457802887,
      "loss": 5.1622,
      "mean_token_accuracy": 0.2085132747888565,
      "num_tokens": 52248392.0,
      "step": 23395
    },
    {
      "entropy": 5.6271782398223875,
      "epoch": 2.6280676138597183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21263641268014907,
      "num_tokens": 52259839.0,
      "step": 23400
    },
    {
      "entropy": 5.567991256713867,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2112468436360359,
      "num_tokens": 52271195.0,
      "step": 23405
    },
    {
      "entropy": 5.523331212997436,
      "epoch": 2.629190767675633,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.0029,
      "mean_token_accuracy": 0.22219393402338028,
      "num_tokens": 52282445.0,
      "step": 23410
    },
    {
      "entropy": 5.6567402362823485,
      "epoch": 2.629752344583591,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043177430463784685,
      "loss": 5.018,
      "mean_token_accuracy": 0.21835651844739914,
      "num_tokens": 52294144.0,
      "step": 23415
    },
    {
      "entropy": 5.56174054145813,
      "epoch": 2.630313921491548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043174550386490816,
      "loss": 5.0215,
      "mean_token_accuracy": 0.22025103271007537,
      "num_tokens": 52304510.0,
      "step": 23420
    },
    {
      "entropy": 5.507012748718262,
      "epoch": 2.6308754983995057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004317166981011319,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22272937446832658,
      "num_tokens": 52314507.0,
      "step": 23425
    },
    {
      "entropy": 5.492795610427857,
      "epoch": 2.6314370753074634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043168788734743516,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22755500674247742,
      "num_tokens": 52325104.0,
      "step": 23430
    },
    {
      "entropy": 5.6484067916870115,
      "epoch": 2.631998652215421,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.0704,
      "mean_token_accuracy": 0.20812637954950333,
      "num_tokens": 52336037.0,
      "step": 23435
    },
    {
      "entropy": 5.5775909423828125,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004316302508739499,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22170355618000032,
      "num_tokens": 52346901.0,
      "step": 23440
    },
    {
      "entropy": 5.552368831634522,
      "epoch": 2.633121806031336,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22527975738048553,
      "num_tokens": 52359159.0,
      "step": 23445
    },
    {
      "entropy": 5.512101697921753,
      "epoch": 2.6336833829392936,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043157259445179304,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22305854856967927,
      "num_tokens": 52369274.0,
      "step": 23450
    },
    {
      "entropy": 5.48590669631958,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004315437587622574,
      "loss": 4.9475,
      "mean_token_accuracy": 0.22396017909049987,
      "num_tokens": 52381360.0,
      "step": 23455
    },
    {
      "entropy": 5.565387821197509,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.125,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.0207,
      "mean_token_accuracy": 0.2139856904745102,
      "num_tokens": 52391550.0,
      "step": 23460
    },
    {
      "entropy": 5.539685010910034,
      "epoch": 2.635368113663166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043148607243086255,
      "loss": 4.976,
      "mean_token_accuracy": 0.22129542976617814,
      "num_tokens": 52401339.0,
      "step": 23465
    },
    {
      "entropy": 5.60643196105957,
      "epoch": 2.635929690571124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.0381,
      "mean_token_accuracy": 0.21454173922538758,
      "num_tokens": 52412062.0,
      "step": 23470
    },
    {
      "entropy": 5.5999805450439455,
      "epoch": 2.636491267479081,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004314283661691587,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2174910917878151,
      "num_tokens": 52422524.0,
      "step": 23475
    },
    {
      "entropy": 5.6260021209716795,
      "epoch": 2.6370528443870387,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.0781,
      "mean_token_accuracy": 0.21686032861471177,
      "num_tokens": 52433422.0,
      "step": 23480
    },
    {
      "entropy": 5.616447687149048,
      "epoch": 2.6376144212949963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21339434683322905,
      "num_tokens": 52443879.0,
      "step": 23485
    },
    {
      "entropy": 5.618897819519043,
      "epoch": 2.638175998202954,
      "grad_norm": 1.0,
      "learning_rate": 0.000431341769423352,
      "loss": 4.9851,
      "mean_token_accuracy": 0.21758433878421785,
      "num_tokens": 52455610.0,
      "step": 23490
    },
    {
      "entropy": 5.637785720825195,
      "epoch": 2.6387375751109117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.0907,
      "mean_token_accuracy": 0.217830128967762,
      "num_tokens": 52467392.0,
      "step": 23495
    },
    {
      "entropy": 5.593752098083496,
      "epoch": 2.639299152018869,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043128401336803616,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21870697736740113,
      "num_tokens": 52478812.0,
      "step": 23500
    },
    {
      "entropy": 5.588660335540771,
      "epoch": 2.6398607289268265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.058,
      "mean_token_accuracy": 0.21964890211820604,
      "num_tokens": 52488837.0,
      "step": 23505
    },
    {
      "entropy": 5.602921056747436,
      "epoch": 2.640422305834784,
      "grad_norm": 1.0,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.1006,
      "mean_token_accuracy": 0.21471676230430603,
      "num_tokens": 52501656.0,
      "step": 23510
    },
    {
      "entropy": 5.5091112613677975,
      "epoch": 2.6409838827427414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043119734196628507,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2182311609387398,
      "num_tokens": 52512280.0,
      "step": 23515
    },
    {
      "entropy": 5.554042530059815,
      "epoch": 2.641545459650699,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043116844155104025,
      "loss": 5.0185,
      "mean_token_accuracy": 0.22249879837036132,
      "num_tokens": 52523664.0,
      "step": 23520
    },
    {
      "entropy": 5.577472829818726,
      "epoch": 2.6421070365586568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.066,
      "mean_token_accuracy": 0.21412751823663712,
      "num_tokens": 52534913.0,
      "step": 23525
    },
    {
      "entropy": 5.509101724624633,
      "epoch": 2.6426686134666144,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004311106258040795,
      "loss": 4.972,
      "mean_token_accuracy": 0.22061587423086165,
      "num_tokens": 52545576.0,
      "step": 23530
    },
    {
      "entropy": 5.614801120758057,
      "epoch": 2.643230190374572,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.0526,
      "mean_token_accuracy": 0.2147383898496628,
      "num_tokens": 52556136.0,
      "step": 23535
    },
    {
      "entropy": 5.603466653823853,
      "epoch": 2.6437917672825293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004310527901746275,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2150006487965584,
      "num_tokens": 52567425.0,
      "step": 23540
    },
    {
      "entropy": 5.537900114059449,
      "epoch": 2.644353344190487,
      "grad_norm": 1.0625,
      "learning_rate": 0.000431023864906269,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21992027014493942,
      "num_tokens": 52578040.0,
      "step": 23545
    },
    {
      "entropy": 5.542844533920288,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043099493467005015,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22596735060214995,
      "num_tokens": 52588261.0,
      "step": 23550
    },
    {
      "entropy": 5.526273393630982,
      "epoch": 2.645476498006402,
      "grad_norm": 1.03125,
      "learning_rate": 0.000430965999466892,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2201795145869255,
      "num_tokens": 52599522.0,
      "step": 23555
    },
    {
      "entropy": 5.581976795196534,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.1282,
      "mean_token_accuracy": 0.21172450333833695,
      "num_tokens": 52610244.0,
      "step": 23560
    },
    {
      "entropy": 5.658301448822021,
      "epoch": 2.646599651822317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.0879,
      "mean_token_accuracy": 0.22061984091997147,
      "num_tokens": 52621745.0,
      "step": 23565
    },
    {
      "entropy": 5.636106252670288,
      "epoch": 2.6471612287302744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.0546,
      "mean_token_accuracy": 0.215291428565979,
      "num_tokens": 52633089.0,
      "step": 23570
    },
    {
      "entropy": 5.604184865951538,
      "epoch": 2.647722805638232,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21700284779071807,
      "num_tokens": 52644911.0,
      "step": 23575
    },
    {
      "entropy": 5.4774346351623535,
      "epoch": 2.6482843825461897,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043082124897926463,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2226342648267746,
      "num_tokens": 52656958.0,
      "step": 23580
    },
    {
      "entropy": 5.622688961029053,
      "epoch": 2.6488459594541474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.0879,
      "mean_token_accuracy": 0.21495932787656785,
      "num_tokens": 52668988.0,
      "step": 23585
    },
    {
      "entropy": 5.560962438583374,
      "epoch": 2.649407536362105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21815653294324874,
      "num_tokens": 52679948.0,
      "step": 23590
    },
    {
      "entropy": 5.59611439704895,
      "epoch": 2.6499691132700622,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004307343391424055,
      "loss": 4.977,
      "mean_token_accuracy": 0.2243204891681671,
      "num_tokens": 52690249.0,
      "step": 23595
    },
    {
      "entropy": 5.603466558456421,
      "epoch": 2.65053069017802,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.0816,
      "mean_token_accuracy": 0.2208031326532364,
      "num_tokens": 52702766.0,
      "step": 23600
    },
    {
      "entropy": 5.61129674911499,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.0514,
      "mean_token_accuracy": 0.2191538095474243,
      "num_tokens": 52714252.0,
      "step": 23605
    },
    {
      "entropy": 5.626386737823486,
      "epoch": 2.651653843993935,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.1018,
      "mean_token_accuracy": 0.21655078679323198,
      "num_tokens": 52724748.0,
      "step": 23610
    },
    {
      "entropy": 5.50154185295105,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.0687,
      "mean_token_accuracy": 0.21189037561416627,
      "num_tokens": 52735636.0,
      "step": 23615
    },
    {
      "entropy": 5.664398241043091,
      "epoch": 2.65277699780985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20696815699338914,
      "num_tokens": 52747812.0,
      "step": 23620
    },
    {
      "entropy": 5.569286632537842,
      "epoch": 2.6533385747178078,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004305603856102808,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22736284434795379,
      "num_tokens": 52759322.0,
      "step": 23625
    },
    {
      "entropy": 5.559090566635132,
      "epoch": 2.6539001516257654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004305313760136799,
      "loss": 4.971,
      "mean_token_accuracy": 0.22244565188884735,
      "num_tokens": 52770238.0,
      "step": 23630
    },
    {
      "entropy": 5.5765400409698485,
      "epoch": 2.6544617285337226,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21763380318880082,
      "num_tokens": 52781243.0,
      "step": 23635
    },
    {
      "entropy": 5.557396507263183,
      "epoch": 2.6550233054416803,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004304733419648638,
      "loss": 4.9704,
      "mean_token_accuracy": 0.214686843752861,
      "num_tokens": 52791341.0,
      "step": 23640
    },
    {
      "entropy": 5.610580158233643,
      "epoch": 2.655584882349638,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.0767,
      "mean_token_accuracy": 0.2168288841843605,
      "num_tokens": 52803322.0,
      "step": 23645
    },
    {
      "entropy": 5.481103134155274,
      "epoch": 2.656146459257595,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004304152881147219,
      "loss": 4.867,
      "mean_token_accuracy": 0.2271009564399719,
      "num_tokens": 52813561.0,
      "step": 23650
    },
    {
      "entropy": 5.567537927627564,
      "epoch": 2.656708036165553,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.0816,
      "mean_token_accuracy": 0.2179915115237236,
      "num_tokens": 52824150.0,
      "step": 23655
    },
    {
      "entropy": 5.465771007537842,
      "epoch": 2.6572696130735105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000430357214470648,
      "loss": 4.885,
      "mean_token_accuracy": 0.2289474219083786,
      "num_tokens": 52835013.0,
      "step": 23660
    },
    {
      "entropy": 5.592477464675904,
      "epoch": 2.6578311899814677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.0985,
      "mean_token_accuracy": 0.21799327284097672,
      "num_tokens": 52846682.0,
      "step": 23665
    },
    {
      "entropy": 5.518330764770508,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043029912104003827,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21043328940868378,
      "num_tokens": 52857561.0,
      "step": 23670
    },
    {
      "entropy": 5.608591318130493,
      "epoch": 2.658954343797383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043027006690709457,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2192274436354637,
      "num_tokens": 52868833.0,
      "step": 23675
    },
    {
      "entropy": 5.659629869461059,
      "epoch": 2.6595159207053407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004302410078302918,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21466712951660155,
      "num_tokens": 52879725.0,
      "step": 23680
    },
    {
      "entropy": 5.628242683410645,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043021194381055505,
      "loss": 5.115,
      "mean_token_accuracy": 0.20828332751989365,
      "num_tokens": 52890585.0,
      "step": 23685
    },
    {
      "entropy": 5.6309120655059814,
      "epoch": 2.6606390745212556,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20822112113237382,
      "num_tokens": 52902816.0,
      "step": 23690
    },
    {
      "entropy": 5.682051324844361,
      "epoch": 2.6612006514292132,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.1589,
      "mean_token_accuracy": 0.21635544747114183,
      "num_tokens": 52914523.0,
      "step": 23695
    },
    {
      "entropy": 5.624916505813599,
      "epoch": 2.661762228337171,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.0718,
      "mean_token_accuracy": 0.21726115196943283,
      "num_tokens": 52924953.0,
      "step": 23700
    },
    {
      "entropy": 5.483605241775512,
      "epoch": 2.662323805245128,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043009563832077977,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22247274667024614,
      "num_tokens": 52935435.0,
      "step": 23705
    },
    {
      "entropy": 5.553572177886963,
      "epoch": 2.662885382153086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.0438,
      "mean_token_accuracy": 0.22275198698043824,
      "num_tokens": 52946969.0,
      "step": 23710
    },
    {
      "entropy": 5.5648609638214115,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.0665,
      "mean_token_accuracy": 0.2154475450515747,
      "num_tokens": 52958634.0,
      "step": 23715
    },
    {
      "entropy": 5.560431480407715,
      "epoch": 2.664008535969001,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004300083573480031,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22854844629764556,
      "num_tokens": 52968412.0,
      "step": 23720
    },
    {
      "entropy": 5.481838607788086,
      "epoch": 2.6645701128769588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042997925381812313,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21472525000572204,
      "num_tokens": 52978653.0,
      "step": 23725
    },
    {
      "entropy": 5.5405731201171875,
      "epoch": 2.665131689784916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042995014535364363,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22489678412675856,
      "num_tokens": 52988518.0,
      "step": 23730
    },
    {
      "entropy": 5.545613098144531,
      "epoch": 2.6656932666928737,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004299210319554914,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2212355062365532,
      "num_tokens": 52999271.0,
      "step": 23735
    },
    {
      "entropy": 5.563249349594116,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.0568,
      "mean_token_accuracy": 0.2259846657514572,
      "num_tokens": 53010530.0,
      "step": 23740
    },
    {
      "entropy": 5.562306594848633,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21958933770656586,
      "num_tokens": 53021631.0,
      "step": 23745
    },
    {
      "entropy": 5.573923540115357,
      "epoch": 2.667377997416746,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042983366216826907,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22140460461378098,
      "num_tokens": 53033763.0,
      "step": 23750
    },
    {
      "entropy": 5.529876661300659,
      "epoch": 2.667939574324704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2172642767429352,
      "num_tokens": 53045839.0,
      "step": 23755
    },
    {
      "entropy": 5.582233095169068,
      "epoch": 2.668501151232661,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21412856876850128,
      "num_tokens": 53057295.0,
      "step": 23760
    },
    {
      "entropy": 5.704377365112305,
      "epoch": 2.6690627281406187,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.2497,
      "mean_token_accuracy": 0.2026568681001663,
      "num_tokens": 53068595.0,
      "step": 23765
    },
    {
      "entropy": 5.561754035949707,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042971710010347594,
      "loss": 4.9884,
      "mean_token_accuracy": 0.22342385947704316,
      "num_tokens": 53080170.0,
      "step": 23770
    },
    {
      "entropy": 5.5572364807128904,
      "epoch": 2.670185881956534,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042968794726932634,
      "loss": 4.9731,
      "mean_token_accuracy": 0.22074873000383377,
      "num_tokens": 53090894.0,
      "step": 23775
    },
    {
      "entropy": 5.606405305862427,
      "epoch": 2.6707474588644917,
      "grad_norm": 1.0,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.1154,
      "mean_token_accuracy": 0.21794503480195998,
      "num_tokens": 53103732.0,
      "step": 23780
    },
    {
      "entropy": 5.633679056167603,
      "epoch": 2.671309035772449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.0284,
      "mean_token_accuracy": 0.22087301313877106,
      "num_tokens": 53115237.0,
      "step": 23785
    },
    {
      "entropy": 5.5736119747161865,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042960045921864627,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21591022610664368,
      "num_tokens": 53125638.0,
      "step": 23790
    },
    {
      "entropy": 5.5172892093658445,
      "epoch": 2.6724321895883643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004295712866887709,
      "loss": 4.967,
      "mean_token_accuracy": 0.22454656511545182,
      "num_tokens": 53136153.0,
      "step": 23795
    },
    {
      "entropy": 5.543233871459961,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.0557,
      "mean_token_accuracy": 0.21552320718765258,
      "num_tokens": 53146562.0,
      "step": 23800
    },
    {
      "entropy": 5.692691469192505,
      "epoch": 2.673555343404279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004295129268651198,
      "loss": 5.1425,
      "mean_token_accuracy": 0.21330359429121018,
      "num_tokens": 53158460.0,
      "step": 23805
    },
    {
      "entropy": 5.563840103149414,
      "epoch": 2.674116920312237,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004294837395732025,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22257184684276582,
      "num_tokens": 53170576.0,
      "step": 23810
    },
    {
      "entropy": 5.546058416366577,
      "epoch": 2.6746784972201945,
      "grad_norm": 1.0,
      "learning_rate": 0.00042945454736246293,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22442166209220887,
      "num_tokens": 53181658.0,
      "step": 23815
    },
    {
      "entropy": 5.630431175231934,
      "epoch": 2.675240074128152,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.0729,
      "mean_token_accuracy": 0.21162078827619552,
      "num_tokens": 53193983.0,
      "step": 23820
    },
    {
      "entropy": 5.5821239948272705,
      "epoch": 2.6758016510361093,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042939614818823547,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21498725265264512,
      "num_tokens": 53205603.0,
      "step": 23825
    },
    {
      "entropy": 5.511491489410401,
      "epoch": 2.676363227944067,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004293669412266069,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22205933183431625,
      "num_tokens": 53216295.0,
      "step": 23830
    },
    {
      "entropy": 5.569905233383179,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004293377293498751,
      "loss": 4.9878,
      "mean_token_accuracy": 0.22368449866771697,
      "num_tokens": 53227055.0,
      "step": 23835
    },
    {
      "entropy": 5.584910154342651,
      "epoch": 2.677486381759982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042930851255897014,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2190754920244217,
      "num_tokens": 53239050.0,
      "step": 23840
    },
    {
      "entropy": 5.519415950775146,
      "epoch": 2.6780479586679395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004292792908548223,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22384258657693862,
      "num_tokens": 53249551.0,
      "step": 23845
    },
    {
      "entropy": 5.595851087570191,
      "epoch": 2.678609535575897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042925006423836197,
      "loss": 5.091,
      "mean_token_accuracy": 0.21674914211034774,
      "num_tokens": 53261645.0,
      "step": 23850
    },
    {
      "entropy": 5.647343492507934,
      "epoch": 2.6791711124838544,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.0893,
      "mean_token_accuracy": 0.21500888764858245,
      "num_tokens": 53272601.0,
      "step": 23855
    },
    {
      "entropy": 5.628533458709716,
      "epoch": 2.679732689391812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.0518,
      "mean_token_accuracy": 0.22167158126831055,
      "num_tokens": 53283190.0,
      "step": 23860
    },
    {
      "entropy": 5.5783556461334225,
      "epoch": 2.6802942662997697,
      "grad_norm": 1.0,
      "learning_rate": 0.0004291623549244129,
      "loss": 4.9901,
      "mean_token_accuracy": 0.22182126194238663,
      "num_tokens": 53294374.0,
      "step": 23865
    },
    {
      "entropy": 5.602520942687988,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.1324,
      "mean_token_accuracy": 0.21421256959438323,
      "num_tokens": 53305609.0,
      "step": 23870
    },
    {
      "entropy": 5.467453479766846,
      "epoch": 2.681417420115685,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042910385750394964,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2182304695248604,
      "num_tokens": 53316251.0,
      "step": 23875
    },
    {
      "entropy": 5.545586681365966,
      "epoch": 2.6819789970236423,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004290746014331623,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22117021977901458,
      "num_tokens": 53326798.0,
      "step": 23880
    },
    {
      "entropy": 5.62458815574646,
      "epoch": 2.6825405739316,
      "grad_norm": 1.0,
      "learning_rate": 0.00042904534045658,
      "loss": 5.0682,
      "mean_token_accuracy": 0.2084565818309784,
      "num_tokens": 53337586.0,
      "step": 23885
    },
    {
      "entropy": 5.601114320755005,
      "epoch": 2.6831021508395576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21873192936182023,
      "num_tokens": 53348517.0,
      "step": 23890
    },
    {
      "entropy": 5.57600507736206,
      "epoch": 2.683663727747515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004289868037897571,
      "loss": 5.0201,
      "mean_token_accuracy": 0.22332592457532882,
      "num_tokens": 53358937.0,
      "step": 23895
    },
    {
      "entropy": 5.56232533454895,
      "epoch": 2.6842253046554725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004289575281013802,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22454277724027633,
      "num_tokens": 53370007.0,
      "step": 23900
    },
    {
      "entropy": 5.503711986541748,
      "epoch": 2.68478688156343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000428928247510936,
      "loss": 4.8967,
      "mean_token_accuracy": 0.22211164385080337,
      "num_tokens": 53380913.0,
      "step": 23905
    },
    {
      "entropy": 5.512813377380371,
      "epoch": 2.685348458471388,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.014,
      "mean_token_accuracy": 0.216260227560997,
      "num_tokens": 53392526.0,
      "step": 23910
    },
    {
      "entropy": 5.586438989639282,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.0,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21622420251369476,
      "num_tokens": 53404522.0,
      "step": 23915
    },
    {
      "entropy": 5.5827264308929445,
      "epoch": 2.6864716122873027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042884037633652276,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2219641000032425,
      "num_tokens": 53415156.0,
      "step": 23920
    },
    {
      "entropy": 5.5664890766143795,
      "epoch": 2.6870331891952604,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042881107614713355,
      "loss": 5.0318,
      "mean_token_accuracy": 0.211286760866642,
      "num_tokens": 53426297.0,
      "step": 23925
    },
    {
      "entropy": 5.677676486968994,
      "epoch": 2.687594766103218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042878177106034,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20926543623208999,
      "num_tokens": 53438415.0,
      "step": 23930
    },
    {
      "entropy": 5.565650844573975,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.0812,
      "mean_token_accuracy": 0.2105254724621773,
      "num_tokens": 53450103.0,
      "step": 23935
    },
    {
      "entropy": 5.5605048656463625,
      "epoch": 2.688717919919133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042872314619827234,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2233229711651802,
      "num_tokens": 53461440.0,
      "step": 23940
    },
    {
      "entropy": 5.681878662109375,
      "epoch": 2.6892794968270906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000428693826424865,
      "loss": 5.138,
      "mean_token_accuracy": 0.20836156457662583,
      "num_tokens": 53472369.0,
      "step": 23945
    },
    {
      "entropy": 5.586555671691895,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004286645017577866,
      "loss": 5.0595,
      "mean_token_accuracy": 0.2236035630106926,
      "num_tokens": 53483739.0,
      "step": 23950
    },
    {
      "entropy": 5.638882398605347,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.0593,
      "mean_token_accuracy": 0.2186876878142357,
      "num_tokens": 53494131.0,
      "step": 23955
    },
    {
      "entropy": 5.579121255874634,
      "epoch": 2.690964227550963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.0861,
      "mean_token_accuracy": 0.213421094417572,
      "num_tokens": 53504923.0,
      "step": 23960
    },
    {
      "entropy": 5.642094469070434,
      "epoch": 2.6915258044589208,
      "grad_norm": 1.0,
      "learning_rate": 0.000428576498403863,
      "loss": 5.0732,
      "mean_token_accuracy": 0.21310838758945466,
      "num_tokens": 53516660.0,
      "step": 23965
    },
    {
      "entropy": 5.544324159622192,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.078125,
      "learning_rate": 0.000428547154171439,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2330232635140419,
      "num_tokens": 53527617.0,
      "step": 23970
    },
    {
      "entropy": 5.532059669494629,
      "epoch": 2.6926489582748356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21380311995744705,
      "num_tokens": 53537830.0,
      "step": 23975
    },
    {
      "entropy": 5.6660964012146,
      "epoch": 2.6932105351827933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.1263,
      "mean_token_accuracy": 0.21245850771665573,
      "num_tokens": 53549981.0,
      "step": 23980
    },
    {
      "entropy": 5.651842212677002,
      "epoch": 2.693772112090751,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004284590921438997,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21777883619070054,
      "num_tokens": 53559941.0,
      "step": 23985
    },
    {
      "entropy": 5.518470621109008,
      "epoch": 2.694333688998708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042842972836107954,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2193844050168991,
      "num_tokens": 53571611.0,
      "step": 23990
    },
    {
      "entropy": 5.608825159072876,
      "epoch": 2.694895265906666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004284003596929973,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21489308923482894,
      "num_tokens": 53582425.0,
      "step": 23995
    },
    {
      "entropy": 5.6208337306976315,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042837098614058815,
      "loss": 5.0906,
      "mean_token_accuracy": 0.2115202397108078,
      "num_tokens": 53593576.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.342446600127103,
      "eval_loss": 5.110117435455322,
      "eval_mean_token_accuracy": 0.2218813785679229,
      "eval_num_tokens": 53593576.0,
      "eval_runtime": 18.4666,
      "eval_samples_per_second": 1679.41,
      "eval_steps_per_second": 209.947,
      "step": 24000
    },
    {
      "entropy": 5.578591442108154,
      "epoch": 2.696018419722581,
      "grad_norm": 1.0,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21616641581058502,
      "num_tokens": 53603768.0,
      "step": 24005
    },
    {
      "entropy": 5.565996408462524,
      "epoch": 2.696579996630539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042831222438653016,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21934868395328522,
      "num_tokens": 53614464.0,
      "step": 24010
    },
    {
      "entropy": 5.636425447463989,
      "epoch": 2.697141573538496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.1063,
      "mean_token_accuracy": 0.21210873872041702,
      "num_tokens": 53625641.0,
      "step": 24015
    },
    {
      "entropy": 5.593984460830688,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2196569934487343,
      "num_tokens": 53636741.0,
      "step": 24020
    },
    {
      "entropy": 5.617830419540406,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.0002,
      "mean_token_accuracy": 0.21937325745820999,
      "num_tokens": 53647295.0,
      "step": 24025
    },
    {
      "entropy": 5.5866231441497805,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000428194642307653,
      "loss": 5.0257,
      "mean_token_accuracy": 0.22537248730659484,
      "num_tokens": 53658371.0,
      "step": 24030
    },
    {
      "entropy": 5.557814121246338,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21778634190559387,
      "num_tokens": 53669534.0,
      "step": 24035
    },
    {
      "entropy": 5.607058334350586,
      "epoch": 2.699949458078284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042813582199780913,
      "loss": 5.0685,
      "mean_token_accuracy": 0.21807554960250855,
      "num_tokens": 53681247.0,
      "step": 24040
    },
    {
      "entropy": 5.641543436050415,
      "epoch": 2.700511034986241,
      "grad_norm": 1.0,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.1005,
      "mean_token_accuracy": 0.20517665892839432,
      "num_tokens": 53692197.0,
      "step": 24045
    },
    {
      "entropy": 5.563797569274902,
      "epoch": 2.701072611894199,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042807698218434964,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2175533354282379,
      "num_tokens": 53703734.0,
      "step": 24050
    },
    {
      "entropy": 5.577855825424194,
      "epoch": 2.7016341888021564,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.0216,
      "mean_token_accuracy": 0.2254293069243431,
      "num_tokens": 53714888.0,
      "step": 24055
    },
    {
      "entropy": 5.551554918289185,
      "epoch": 2.702195765710114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042801812287476836,
      "loss": 4.9815,
      "mean_token_accuracy": 0.21350869685411453,
      "num_tokens": 53726231.0,
      "step": 24060
    },
    {
      "entropy": 5.5840002536773685,
      "epoch": 2.7027573426180718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21419646590948105,
      "num_tokens": 53737777.0,
      "step": 24065
    },
    {
      "entropy": 5.6163428783416744,
      "epoch": 2.703318919526029,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21501292735338212,
      "num_tokens": 53748277.0,
      "step": 24070
    },
    {
      "entropy": 5.56364336013794,
      "epoch": 2.7038804964339866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042792979737156717,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2204321324825287,
      "num_tokens": 53759205.0,
      "step": 24075
    },
    {
      "entropy": 5.561681890487671,
      "epoch": 2.7044420733419443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21088260412216187,
      "num_tokens": 53770984.0,
      "step": 24080
    },
    {
      "entropy": 5.49401707649231,
      "epoch": 2.7050036502499015,
      "grad_norm": 1.0,
      "learning_rate": 0.0004278708893544837,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22452450096607207,
      "num_tokens": 53782092.0,
      "step": 24085
    },
    {
      "entropy": 5.599517440795898,
      "epoch": 2.705565227157859,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21581290662288666,
      "num_tokens": 53794136.0,
      "step": 24090
    },
    {
      "entropy": 5.632866239547729,
      "epoch": 2.706126804065817,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21195352971553802,
      "num_tokens": 53804612.0,
      "step": 24095
    },
    {
      "entropy": 5.533676481246948,
      "epoch": 2.7066883809737745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21537099182605743,
      "num_tokens": 53815850.0,
      "step": 24100
    },
    {
      "entropy": 5.5518879890441895,
      "epoch": 2.707249957881732,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004277530149182014,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2195115178823471,
      "num_tokens": 53827050.0,
      "step": 24105
    },
    {
      "entropy": 5.630576181411743,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.0728,
      "mean_token_accuracy": 0.21360431462526322,
      "num_tokens": 53837697.0,
      "step": 24110
    },
    {
      "entropy": 5.584818983078003,
      "epoch": 2.708373111697647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042769404851401524,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2214365139603615,
      "num_tokens": 53849278.0,
      "step": 24115
    },
    {
      "entropy": 5.570425128936767,
      "epoch": 2.7089346886056047,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21602800041437148,
      "num_tokens": 53861951.0,
      "step": 24120
    },
    {
      "entropy": 5.635323810577392,
      "epoch": 2.709496265513562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042763506266247834,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22704201489686965,
      "num_tokens": 53873647.0,
      "step": 24125
    },
    {
      "entropy": 5.522173309326172,
      "epoch": 2.7100578424215196,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004276055624463009,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21510957628488542,
      "num_tokens": 53885063.0,
      "step": 24130
    },
    {
      "entropy": 5.540197515487671,
      "epoch": 2.7106194193294773,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042757605737110324,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2175866574048996,
      "num_tokens": 53896955.0,
      "step": 24135
    },
    {
      "entropy": 5.4998557567596436,
      "epoch": 2.7111809962374345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004275465474378246,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22656962126493455,
      "num_tokens": 53907862.0,
      "step": 24140
    },
    {
      "entropy": 5.614484262466431,
      "epoch": 2.711742573145392,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2203718051314354,
      "num_tokens": 53918918.0,
      "step": 24145
    },
    {
      "entropy": 5.563062000274658,
      "epoch": 2.71230415005335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21659125983715058,
      "num_tokens": 53930488.0,
      "step": 24150
    },
    {
      "entropy": 5.640937185287475,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004274579884989008,
      "loss": 5.059,
      "mean_token_accuracy": 0.21764473617076874,
      "num_tokens": 53942314.0,
      "step": 24155
    },
    {
      "entropy": 5.589636516571045,
      "epoch": 2.713427303869265,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042742845914269645,
      "loss": 5.0693,
      "mean_token_accuracy": 0.22693693190813063,
      "num_tokens": 53951915.0,
      "step": 24160
    },
    {
      "entropy": 5.544435501098633,
      "epoch": 2.7139888807772223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.0639,
      "mean_token_accuracy": 0.22284871488809585,
      "num_tokens": 53962815.0,
      "step": 24165
    },
    {
      "entropy": 5.559187889099121,
      "epoch": 2.71455045768518,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042736938587108447,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2174252137541771,
      "num_tokens": 53972740.0,
      "step": 24170
    },
    {
      "entropy": 5.633758544921875,
      "epoch": 2.7151120345931377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21622220873832704,
      "num_tokens": 53983632.0,
      "step": 24175
    },
    {
      "entropy": 5.636434841156006,
      "epoch": 2.715673611501095,
      "grad_norm": 1.0,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.0956,
      "mean_token_accuracy": 0.21320678144693375,
      "num_tokens": 53994085.0,
      "step": 24180
    },
    {
      "entropy": 5.591640949249268,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000427280739579766,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22000332027673722,
      "num_tokens": 54004106.0,
      "step": 24185
    },
    {
      "entropy": 5.520744132995605,
      "epoch": 2.71679676531701,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042725118111738305,
      "loss": 4.8959,
      "mean_token_accuracy": 0.22919558584690095,
      "num_tokens": 54014582.0,
      "step": 24190
    },
    {
      "entropy": 5.585865545272827,
      "epoch": 2.717358342224968,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21891430169343948,
      "num_tokens": 54025626.0,
      "step": 24195
    },
    {
      "entropy": 5.584996843338013,
      "epoch": 2.7179199191329255,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.067,
      "mean_token_accuracy": 0.21932579427957535,
      "num_tokens": 54036974.0,
      "step": 24200
    },
    {
      "entropy": 5.604928874969483,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.0576,
      "mean_token_accuracy": 0.2150803193449974,
      "num_tokens": 54048404.0,
      "step": 24205
    },
    {
      "entropy": 5.605258417129517,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21845546513795852,
      "num_tokens": 54059210.0,
      "step": 24210
    },
    {
      "entropy": 5.651535511016846,
      "epoch": 2.719604649856798,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.0755,
      "mean_token_accuracy": 0.22015629708766937,
      "num_tokens": 54070406.0,
      "step": 24215
    },
    {
      "entropy": 5.534593296051026,
      "epoch": 2.7201662267647553,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042707372857356615,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21685273945331573,
      "num_tokens": 54081497.0,
      "step": 24220
    },
    {
      "entropy": 5.472709846496582,
      "epoch": 2.720727803672713,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004270441361967994,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2331560343503952,
      "num_tokens": 54090934.0,
      "step": 24225
    },
    {
      "entropy": 5.422910928726196,
      "epoch": 2.7212893805806706,
      "grad_norm": 1.0,
      "learning_rate": 0.0004270145389788883,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2298227071762085,
      "num_tokens": 54102011.0,
      "step": 24230
    },
    {
      "entropy": 5.603714942932129,
      "epoch": 2.721850957488628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21626375317573548,
      "num_tokens": 54113066.0,
      "step": 24235
    },
    {
      "entropy": 5.5028722286224365,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004269553300234031,
      "loss": 4.9186,
      "mean_token_accuracy": 0.22253288179636002,
      "num_tokens": 54124291.0,
      "step": 24240
    },
    {
      "entropy": 5.633699941635132,
      "epoch": 2.722974111304543,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042692571828771423,
      "loss": 5.139,
      "mean_token_accuracy": 0.2172923728823662,
      "num_tokens": 54136147.0,
      "step": 24245
    },
    {
      "entropy": 5.613784551620483,
      "epoch": 2.723535688212501,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042689610171465147,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21826988309621811,
      "num_tokens": 54147953.0,
      "step": 24250
    },
    {
      "entropy": 5.5645959854125975,
      "epoch": 2.7240972651204585,
      "grad_norm": 1.046875,
      "learning_rate": 0.000426866480305158,
      "loss": 4.9912,
      "mean_token_accuracy": 0.22237271219491958,
      "num_tokens": 54159498.0,
      "step": 24255
    },
    {
      "entropy": 5.597626495361328,
      "epoch": 2.7246588420284157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21751420348882675,
      "num_tokens": 54170380.0,
      "step": 24260
    },
    {
      "entropy": 5.539787721633911,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.125,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.0852,
      "mean_token_accuracy": 0.21261837631464003,
      "num_tokens": 54181105.0,
      "step": 24265
    },
    {
      "entropy": 5.687046766281128,
      "epoch": 2.725781995844331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.0831,
      "mean_token_accuracy": 0.21189189702272415,
      "num_tokens": 54192142.0,
      "step": 24270
    },
    {
      "entropy": 5.552733087539673,
      "epoch": 2.7263435727522882,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004267479463217415,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21863804161548614,
      "num_tokens": 54202716.0,
      "step": 24275
    },
    {
      "entropy": 5.499549818038941,
      "epoch": 2.726905149660246,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042671830074424453,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22593102157115935,
      "num_tokens": 54213035.0,
      "step": 24280
    },
    {
      "entropy": 5.55492057800293,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000426688650335978,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2156070962548256,
      "num_tokens": 54223999.0,
      "step": 24285
    },
    {
      "entropy": 5.554051256179809,
      "epoch": 2.728028303476161,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042665899509788593,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2174388900399208,
      "num_tokens": 54234393.0,
      "step": 24290
    },
    {
      "entropy": 5.522165536880493,
      "epoch": 2.728589880384119,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004266293350309128,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22192768454551698,
      "num_tokens": 54244787.0,
      "step": 24295
    },
    {
      "entropy": 5.586071920394898,
      "epoch": 2.729151457292076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20879469960927963,
      "num_tokens": 54256033.0,
      "step": 24300
    },
    {
      "entropy": 5.60486102104187,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042657000041410026,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22465144842863083,
      "num_tokens": 54266796.0,
      "step": 24305
    },
    {
      "entropy": 5.657816457748413,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.0925,
      "mean_token_accuracy": 0.21339477002620696,
      "num_tokens": 54279049.0,
      "step": 24310
    },
    {
      "entropy": 5.589087009429932,
      "epoch": 2.7308361880159486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21298945546150208,
      "num_tokens": 54290544.0,
      "step": 24315
    },
    {
      "entropy": 5.615546751022339,
      "epoch": 2.7313977649239063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21779558807611465,
      "num_tokens": 54300905.0,
      "step": 24320
    },
    {
      "entropy": 5.524023103713989,
      "epoch": 2.731959341831864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042645127327546405,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22054869532585145,
      "num_tokens": 54313019.0,
      "step": 24325
    },
    {
      "entropy": 5.6477635383605955,
      "epoch": 2.732520918739821,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.08,
      "mean_token_accuracy": 0.2155701383948326,
      "num_tokens": 54324034.0,
      "step": 24330
    },
    {
      "entropy": 5.532745742797852,
      "epoch": 2.733082495647779,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004263918807687616,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21787755340337753,
      "num_tokens": 54335335.0,
      "step": 24335
    },
    {
      "entropy": 5.611420392990112,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004263621772843733,
      "loss": 5.0852,
      "mean_token_accuracy": 0.21833053827285767,
      "num_tokens": 54345659.0,
      "step": 24340
    },
    {
      "entropy": 5.625114154815674,
      "epoch": 2.734205649463694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.088,
      "mean_token_accuracy": 0.2146235540509224,
      "num_tokens": 54357959.0,
      "step": 24345
    },
    {
      "entropy": 5.608568954467773,
      "epoch": 2.734767226371652,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.0202,
      "mean_token_accuracy": 0.2181793987751007,
      "num_tokens": 54368779.0,
      "step": 24350
    },
    {
      "entropy": 5.633490180969238,
      "epoch": 2.735328803279609,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21381573677062987,
      "num_tokens": 54381726.0,
      "step": 24355
    },
    {
      "entropy": 5.621864461898804,
      "epoch": 2.7358903801875667,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2204273134469986,
      "num_tokens": 54393201.0,
      "step": 24360
    },
    {
      "entropy": 5.5663979053497314,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042621358758989584,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2282334089279175,
      "num_tokens": 54404178.0,
      "step": 24365
    },
    {
      "entropy": 5.484686708450317,
      "epoch": 2.7370135340034816,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004261838552031169,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22508466690778733,
      "num_tokens": 54414963.0,
      "step": 24370
    },
    {
      "entropy": 5.506991815567017,
      "epoch": 2.7375751109114392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042615411800258523,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23317905962467195,
      "num_tokens": 54425333.0,
      "step": 24375
    },
    {
      "entropy": 5.566774034500122,
      "epoch": 2.738136687819397,
      "grad_norm": 1.125,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.1231,
      "mean_token_accuracy": 0.20919182002544404,
      "num_tokens": 54436326.0,
      "step": 24380
    },
    {
      "entropy": 5.585274744033813,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2181594803929329,
      "num_tokens": 54446664.0,
      "step": 24385
    },
    {
      "entropy": 5.5991452693939205,
      "epoch": 2.739259841635312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22092458605766296,
      "num_tokens": 54457135.0,
      "step": 24390
    },
    {
      "entropy": 5.459041118621826,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042603512108187224,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2166807994246483,
      "num_tokens": 54468248.0,
      "step": 24395
    },
    {
      "entropy": 5.544882535934448,
      "epoch": 2.740382995451227,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042600535982678347,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2275804027915001,
      "num_tokens": 54478676.0,
      "step": 24400
    },
    {
      "entropy": 5.501358413696289,
      "epoch": 2.7409445723591848,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042597559376362544,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21614535599946977,
      "num_tokens": 54490342.0,
      "step": 24405
    },
    {
      "entropy": 5.5374962329864506,
      "epoch": 2.741506149267142,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004259458228933459,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22334169894456862,
      "num_tokens": 54501543.0,
      "step": 24410
    },
    {
      "entropy": 5.552964353561402,
      "epoch": 2.7420677261750996,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042591604721689297,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22198403030633926,
      "num_tokens": 54511777.0,
      "step": 24415
    },
    {
      "entropy": 5.601544332504273,
      "epoch": 2.7426293030830573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004258862667352145,
      "loss": 5.0161,
      "mean_token_accuracy": 0.22219249457120896,
      "num_tokens": 54523255.0,
      "step": 24420
    },
    {
      "entropy": 5.555226135253906,
      "epoch": 2.7431908799910145,
      "grad_norm": 1.0,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21505021452903747,
      "num_tokens": 54534158.0,
      "step": 24425
    },
    {
      "entropy": 5.670927953720093,
      "epoch": 2.743752456898972,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000425826691359974,
      "loss": 5.1016,
      "mean_token_accuracy": 0.21137073636054993,
      "num_tokens": 54545735.0,
      "step": 24430
    },
    {
      "entropy": 5.661331653594971,
      "epoch": 2.74431403380693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.0483,
      "mean_token_accuracy": 0.22167424410581588,
      "num_tokens": 54557112.0,
      "step": 24435
    },
    {
      "entropy": 5.6491217613220215,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.0851,
      "mean_token_accuracy": 0.21636408865451812,
      "num_tokens": 54568600.0,
      "step": 24440
    },
    {
      "entropy": 5.579397249221802,
      "epoch": 2.745437187622845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000425737292281633,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2213313549757004,
      "num_tokens": 54579008.0,
      "step": 24445
    },
    {
      "entropy": 5.5083106517791744,
      "epoch": 2.7459987645308024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21607676446437835,
      "num_tokens": 54590306.0,
      "step": 24450
    },
    {
      "entropy": 5.604772520065308,
      "epoch": 2.74656034143876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21486395895481109,
      "num_tokens": 54600999.0,
      "step": 24455
    },
    {
      "entropy": 5.634220933914184,
      "epoch": 2.7471219183467177,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2182108223438263,
      "num_tokens": 54612977.0,
      "step": 24460
    },
    {
      "entropy": 5.541520309448242,
      "epoch": 2.747683495254675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004256180263214687,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2277971997857094,
      "num_tokens": 54623597.0,
      "step": 24465
    },
    {
      "entropy": 5.4914634227752686,
      "epoch": 2.7482450721626326,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042558819783971295,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22908782213926315,
      "num_tokens": 54634618.0,
      "step": 24470
    },
    {
      "entropy": 5.58705325126648,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22058812379837037,
      "num_tokens": 54645291.0,
      "step": 24475
    },
    {
      "entropy": 5.597646236419678,
      "epoch": 2.749368225978548,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.0766,
      "mean_token_accuracy": 0.2165396884083748,
      "num_tokens": 54657294.0,
      "step": 24480
    },
    {
      "entropy": 5.613482475280762,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042549868362952556,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2138974502682686,
      "num_tokens": 54668627.0,
      "step": 24485
    },
    {
      "entropy": 5.604241752624512,
      "epoch": 2.750491379794463,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042546883597432327,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21873728781938553,
      "num_tokens": 54679763.0,
      "step": 24490
    },
    {
      "entropy": 5.603281879425049,
      "epoch": 2.7510529567024204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042543898352813485,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21953382194042206,
      "num_tokens": 54690476.0,
      "step": 24495
    },
    {
      "entropy": 5.604862642288208,
      "epoch": 2.751614533610378,
      "grad_norm": 1.03125,
      "learning_rate": 0.000425409126291911,
      "loss": 5.0456,
      "mean_token_accuracy": 0.22607291340827942,
      "num_tokens": 54701383.0,
      "step": 24500
    },
    {
      "entropy": 5.575025224685669,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004253792642666021,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2227158948779106,
      "num_tokens": 54712268.0,
      "step": 24505
    },
    {
      "entropy": 5.577183437347412,
      "epoch": 2.752737687426293,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042534939745315914,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22146208584308624,
      "num_tokens": 54722834.0,
      "step": 24510
    },
    {
      "entropy": 5.567540216445923,
      "epoch": 2.7532992643342507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000425319525852533,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22590145468711853,
      "num_tokens": 54734628.0,
      "step": 24515
    },
    {
      "entropy": 5.445689249038696,
      "epoch": 2.753860841242208,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004252896494656749,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23006256371736528,
      "num_tokens": 54745927.0,
      "step": 24520
    },
    {
      "entropy": 5.465902280807495,
      "epoch": 2.7544224181501655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004252597682935362,
      "loss": 4.905,
      "mean_token_accuracy": 0.22349003553390503,
      "num_tokens": 54757822.0,
      "step": 24525
    },
    {
      "entropy": 5.521550416946411,
      "epoch": 2.754983995058123,
      "grad_norm": 1.046875,
      "learning_rate": 0.000425229882337068,
      "loss": 4.9845,
      "mean_token_accuracy": 0.22758277952671052,
      "num_tokens": 54768964.0,
      "step": 24530
    },
    {
      "entropy": 5.5607585430145265,
      "epoch": 2.755545571966081,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042519999159722216,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2236445039510727,
      "num_tokens": 54779471.0,
      "step": 24535
    },
    {
      "entropy": 5.520379686355591,
      "epoch": 2.7561071488740385,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042517009607495035,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22570154815912247,
      "num_tokens": 54790426.0,
      "step": 24540
    },
    {
      "entropy": 5.503724670410156,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.027,
      "mean_token_accuracy": 0.22025859951972962,
      "num_tokens": 54801765.0,
      "step": 24545
    },
    {
      "entropy": 5.595951652526855,
      "epoch": 2.7572303026899534,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004251102906869364,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2127838060259819,
      "num_tokens": 54814272.0,
      "step": 24550
    },
    {
      "entropy": 5.5780298709869385,
      "epoch": 2.757791879597911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.0389,
      "mean_token_accuracy": 0.22074505388736726,
      "num_tokens": 54825610.0,
      "step": 24555
    },
    {
      "entropy": 5.657857656478882,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.0508,
      "mean_token_accuracy": 0.2220241829752922,
      "num_tokens": 54835546.0,
      "step": 24560
    },
    {
      "entropy": 5.603320789337158,
      "epoch": 2.758915033413826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2179194614291191,
      "num_tokens": 54845653.0,
      "step": 24565
    },
    {
      "entropy": 5.548583316802978,
      "epoch": 2.7594766103217836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042499062256368987,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22676272243261336,
      "num_tokens": 54855514.0,
      "step": 24570
    },
    {
      "entropy": 5.520393371582031,
      "epoch": 2.7600381872297413,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.0215,
      "mean_token_accuracy": 0.22037476152181626,
      "num_tokens": 54866131.0,
      "step": 24575
    },
    {
      "entropy": 5.516569375991821,
      "epoch": 2.760599764137699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004249307598436984,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23035652786493302,
      "num_tokens": 54876842.0,
      "step": 24580
    },
    {
      "entropy": 5.601043176651001,
      "epoch": 2.761161341045656,
      "grad_norm": 1.125,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.1086,
      "mean_token_accuracy": 0.21139676123857498,
      "num_tokens": 54888036.0,
      "step": 24585
    },
    {
      "entropy": 5.576989793777466,
      "epoch": 2.761722917953614,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042487087802829275,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2205380380153656,
      "num_tokens": 54898501.0,
      "step": 24590
    },
    {
      "entropy": 5.573254346847534,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004248409299621931,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2150132715702057,
      "num_tokens": 54909054.0,
      "step": 24595
    },
    {
      "entropy": 5.597199487686157,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004248109771251,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21306642144918442,
      "num_tokens": 54919572.0,
      "step": 24600
    },
    {
      "entropy": 5.658470201492309,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21735866069793702,
      "num_tokens": 54930929.0,
      "step": 24605
    },
    {
      "entropy": 5.510829591751099,
      "epoch": 2.763969225585444,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004247510571417487,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22310763895511626,
      "num_tokens": 54941317.0,
      "step": 24610
    },
    {
      "entropy": 5.559792613983154,
      "epoch": 2.764530802493401,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004247210899973985,
      "loss": 4.96,
      "mean_token_accuracy": 0.22705279141664506,
      "num_tokens": 54951203.0,
      "step": 24615
    },
    {
      "entropy": 5.619540452957153,
      "epoch": 2.765092379401359,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2193213537335396,
      "num_tokens": 54961485.0,
      "step": 24620
    },
    {
      "entropy": 5.554107141494751,
      "epoch": 2.7656539563093165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21936425864696502,
      "num_tokens": 54973152.0,
      "step": 24625
    },
    {
      "entropy": 5.55381498336792,
      "epoch": 2.766215533217274,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042463115996509984,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2162017270922661,
      "num_tokens": 54984863.0,
      "step": 24630
    },
    {
      "entropy": 5.599859714508057,
      "epoch": 2.766777110125232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.0067,
      "mean_token_accuracy": 0.214700648188591,
      "num_tokens": 54997382.0,
      "step": 24635
    },
    {
      "entropy": 5.642671728134156,
      "epoch": 2.767338687033189,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21463145315647125,
      "num_tokens": 55007686.0,
      "step": 24640
    },
    {
      "entropy": 5.591425848007202,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.0936,
      "mean_token_accuracy": 0.21315090656280516,
      "num_tokens": 55018776.0,
      "step": 24645
    },
    {
      "entropy": 5.5687745094299315,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245111865594273,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21574576944112778,
      "num_tokens": 55031566.0,
      "step": 24650
    },
    {
      "entropy": 5.550516271591187,
      "epoch": 2.7690234177570616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042448118130438583,
      "loss": 4.9063,
      "mean_token_accuracy": 0.23276772797107698,
      "num_tokens": 55043321.0,
      "step": 24655
    },
    {
      "entropy": 5.551924848556519,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042445117128980553,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21769571155309678,
      "num_tokens": 55052655.0,
      "step": 24660
    },
    {
      "entropy": 5.544489097595215,
      "epoch": 2.770146571572977,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.0492,
      "mean_token_accuracy": 0.2125790536403656,
      "num_tokens": 55063317.0,
      "step": 24665
    },
    {
      "entropy": 5.498974084854126,
      "epoch": 2.7707081484809346,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042439113698585103,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2245929479598999,
      "num_tokens": 55074916.0,
      "step": 24670
    },
    {
      "entropy": 5.544792890548706,
      "epoch": 2.7712697253888923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004243611126983882,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21823373883962632,
      "num_tokens": 55086456.0,
      "step": 24675
    },
    {
      "entropy": 5.684802722930908,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.0,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.2011,
      "mean_token_accuracy": 0.20449414402246474,
      "num_tokens": 55097268.0,
      "step": 24680
    },
    {
      "entropy": 5.597358179092407,
      "epoch": 2.772392879204807,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042430104985727155,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22295750826597213,
      "num_tokens": 55107451.0,
      "step": 24685
    },
    {
      "entropy": 5.553316879272461,
      "epoch": 2.772954456112765,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042427101130553,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22348614782094955,
      "num_tokens": 55117781.0,
      "step": 24690
    },
    {
      "entropy": 5.609001207351684,
      "epoch": 2.773516033020722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21368999630212784,
      "num_tokens": 55129519.0,
      "step": 24695
    },
    {
      "entropy": 5.483769464492798,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004242109199444629,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21724362224340438,
      "num_tokens": 55138964.0,
      "step": 24700
    },
    {
      "entropy": 5.603969860076904,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21583260893821715,
      "num_tokens": 55149953.0,
      "step": 24705
    },
    {
      "entropy": 5.624562978744507,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21555821001529693,
      "num_tokens": 55162208.0,
      "step": 24710
    },
    {
      "entropy": 5.570136690139771,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004241207472732531,
      "loss": 5.003,
      "mean_token_accuracy": 0.21367755830287932,
      "num_tokens": 55174014.0,
      "step": 24715
    },
    {
      "entropy": 5.6291767120361325,
      "epoch": 2.77632391756051,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.1058,
      "mean_token_accuracy": 0.2073919504880905,
      "num_tokens": 55185173.0,
      "step": 24720
    },
    {
      "entropy": 5.670839071273804,
      "epoch": 2.7768854944684676,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.0708,
      "mean_token_accuracy": 0.2202285036444664,
      "num_tokens": 55198404.0,
      "step": 24725
    },
    {
      "entropy": 5.529473876953125,
      "epoch": 2.777447071376425,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042403053186948237,
      "loss": 4.9969,
      "mean_token_accuracy": 0.21314286142587663,
      "num_tokens": 55209126.0,
      "step": 24730
    },
    {
      "entropy": 5.565711927413941,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004240004505765688,
      "loss": 5.0024,
      "mean_token_accuracy": 0.22023069113492966,
      "num_tokens": 55220311.0,
      "step": 24735
    },
    {
      "entropy": 5.561052417755127,
      "epoch": 2.77857022519234,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004239703645394231,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22402596920728685,
      "num_tokens": 55230525.0,
      "step": 24740
    },
    {
      "entropy": 5.635844612121582,
      "epoch": 2.7791318021002978,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.1332,
      "mean_token_accuracy": 0.2155425801873207,
      "num_tokens": 55242084.0,
      "step": 24745
    },
    {
      "entropy": 5.563123559951782,
      "epoch": 2.779693379008255,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042391017823626716,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2237740248441696,
      "num_tokens": 55253677.0,
      "step": 24750
    },
    {
      "entropy": 5.515515947341919,
      "epoch": 2.7802549559162126,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004238800779721734,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22986041158437728,
      "num_tokens": 55265118.0,
      "step": 24755
    },
    {
      "entropy": 5.492420864105225,
      "epoch": 2.7808165328241703,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042384997296768007,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22270534932613373,
      "num_tokens": 55277119.0,
      "step": 24760
    },
    {
      "entropy": 5.594024324417115,
      "epoch": 2.781378109732128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000423819863223746,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22795796245336533,
      "num_tokens": 55287828.0,
      "step": 24765
    },
    {
      "entropy": 5.663049697875977,
      "epoch": 2.7819396866400856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.0927,
      "mean_token_accuracy": 0.2163582220673561,
      "num_tokens": 55299115.0,
      "step": 24770
    },
    {
      "entropy": 5.542862129211426,
      "epoch": 2.782501263548043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004237596295213901,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21508273780345916,
      "num_tokens": 55310532.0,
      "step": 24775
    },
    {
      "entropy": 5.595054388046265,
      "epoch": 2.7830628404560005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004237295055648862,
      "loss": 5.0802,
      "mean_token_accuracy": 0.21247405111789702,
      "num_tokens": 55321127.0,
      "step": 24780
    },
    {
      "entropy": 5.63966851234436,
      "epoch": 2.783624417363958,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22515588104724885,
      "num_tokens": 55331791.0,
      "step": 24785
    },
    {
      "entropy": 5.61956820487976,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.0779,
      "mean_token_accuracy": 0.2136064201593399,
      "num_tokens": 55342724.0,
      "step": 24790
    },
    {
      "entropy": 5.530638980865478,
      "epoch": 2.784747571179873,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236391052855811,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21746279746294023,
      "num_tokens": 55354471.0,
      "step": 24795
    },
    {
      "entropy": 5.585718870162964,
      "epoch": 2.7853091480878307,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042360896239241306,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2180505856871605,
      "num_tokens": 55366273.0,
      "step": 24800
    },
    {
      "entropy": 5.539503479003907,
      "epoch": 2.785870724995788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042357881476747807,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22063181102275847,
      "num_tokens": 55377384.0,
      "step": 24805
    },
    {
      "entropy": 5.54762749671936,
      "epoch": 2.7864323019037456,
      "grad_norm": 1.140625,
      "learning_rate": 0.000423548662411736,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21748968511819838,
      "num_tokens": 55387348.0,
      "step": 24810
    },
    {
      "entropy": 5.5318480968475345,
      "epoch": 2.7869938788117032,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004235185053261469,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22986901551485062,
      "num_tokens": 55398484.0,
      "step": 24815
    },
    {
      "entropy": 5.554336738586426,
      "epoch": 2.787555455719661,
      "grad_norm": 1.09375,
      "learning_rate": 0.000423488343511671,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2256719723343849,
      "num_tokens": 55409347.0,
      "step": 24820
    },
    {
      "entropy": 5.522925615310669,
      "epoch": 2.7881170326276186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004234581769692686,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21806009709835053,
      "num_tokens": 55421048.0,
      "step": 24825
    },
    {
      "entropy": 5.541890287399292,
      "epoch": 2.788678609535576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004234280056999003,
      "loss": 4.957,
      "mean_token_accuracy": 0.22618326395750046,
      "num_tokens": 55431189.0,
      "step": 24830
    },
    {
      "entropy": 5.6283636569976805,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004233978297045267,
      "loss": 5.0299,
      "mean_token_accuracy": 0.215708689391613,
      "num_tokens": 55442695.0,
      "step": 24835
    },
    {
      "entropy": 5.647146368026734,
      "epoch": 2.789801763351491,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.0396,
      "mean_token_accuracy": 0.2175389975309372,
      "num_tokens": 55454954.0,
      "step": 24840
    },
    {
      "entropy": 5.586181402206421,
      "epoch": 2.7903633402594483,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.0788,
      "mean_token_accuracy": 0.22108782529830934,
      "num_tokens": 55466270.0,
      "step": 24845
    },
    {
      "entropy": 5.658647727966309,
      "epoch": 2.790924917167406,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.0519,
      "mean_token_accuracy": 0.2215530678629875,
      "num_tokens": 55476925.0,
      "step": 24850
    },
    {
      "entropy": 5.5436708450317385,
      "epoch": 2.7914864940753636,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004232770784821984,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21980566531419754,
      "num_tokens": 55489705.0,
      "step": 24855
    },
    {
      "entropy": 5.6390434265136715,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.109,
      "mean_token_accuracy": 0.2110452651977539,
      "num_tokens": 55501362.0,
      "step": 24860
    },
    {
      "entropy": 5.610900497436523,
      "epoch": 2.792609647891279,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21651371121406554,
      "num_tokens": 55512754.0,
      "step": 24865
    },
    {
      "entropy": 5.577921962738037,
      "epoch": 2.793171224799236,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042318646548949177,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21803237944841386,
      "num_tokens": 55524057.0,
      "step": 24870
    },
    {
      "entropy": 5.586559247970581,
      "epoch": 2.793732801707194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042315625172067794,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2216718852519989,
      "num_tokens": 55534768.0,
      "step": 24875
    },
    {
      "entropy": 5.668817615509033,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.0833,
      "mean_token_accuracy": 0.2151902586221695,
      "num_tokens": 55547135.0,
      "step": 24880
    },
    {
      "entropy": 5.611378288269043,
      "epoch": 2.7948559555231087,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21589729338884353,
      "num_tokens": 55559638.0,
      "step": 24885
    },
    {
      "entropy": 5.58187198638916,
      "epoch": 2.7954175324310664,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.0084,
      "mean_token_accuracy": 0.22547236680984498,
      "num_tokens": 55570985.0,
      "step": 24890
    },
    {
      "entropy": 5.5269547462463375,
      "epoch": 2.795979109339024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004230353494815203,
      "loss": 4.983,
      "mean_token_accuracy": 0.224268402159214,
      "num_tokens": 55581701.0,
      "step": 24895
    },
    {
      "entropy": 5.570846080780029,
      "epoch": 2.7965406862469813,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004230051121355672,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22094811052083968,
      "num_tokens": 55592891.0,
      "step": 24900
    },
    {
      "entropy": 5.4976616382598875,
      "epoch": 2.797102263154939,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.0046,
      "mean_token_accuracy": 0.22170836329460145,
      "num_tokens": 55603762.0,
      "step": 24905
    },
    {
      "entropy": 5.619992208480835,
      "epoch": 2.7976638400628966,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.0381,
      "mean_token_accuracy": 0.218446484208107,
      "num_tokens": 55614450.0,
      "step": 24910
    },
    {
      "entropy": 5.5790304183959964,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004229143718263186,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22151371836662292,
      "num_tokens": 55625240.0,
      "step": 24915
    },
    {
      "entropy": 5.51662974357605,
      "epoch": 2.798786993878812,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004228841156359829,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2320418491959572,
      "num_tokens": 55635646.0,
      "step": 24920
    },
    {
      "entropy": 5.476048994064331,
      "epoch": 2.799348570786769,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004228538547369596,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2252239853143692,
      "num_tokens": 55646669.0,
      "step": 24925
    },
    {
      "entropy": 5.682319116592407,
      "epoch": 2.799910147694727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.2209,
      "mean_token_accuracy": 0.2087971583008766,
      "num_tokens": 55658792.0,
      "step": 24930
    },
    {
      "entropy": 5.626200437545776,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21821920424699784,
      "num_tokens": 55670037.0,
      "step": 24935
    },
    {
      "entropy": 5.586581754684448,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004227630437974,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21317656189203263,
      "num_tokens": 55680405.0,
      "step": 24940
    },
    {
      "entropy": 5.59699444770813,
      "epoch": 2.8015948784185993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.008,
      "mean_token_accuracy": 0.21670155376195907,
      "num_tokens": 55692199.0,
      "step": 24945
    },
    {
      "entropy": 5.54582633972168,
      "epoch": 2.802156455326557,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042270247964525755,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22421081960201264,
      "num_tokens": 55702654.0,
      "step": 24950
    },
    {
      "entropy": 5.563938951492309,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22460066080093383,
      "num_tokens": 55712710.0,
      "step": 24955
    },
    {
      "entropy": 5.594060707092285,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.0149,
      "mean_token_accuracy": 0.217619888484478,
      "num_tokens": 55723429.0,
      "step": 24960
    },
    {
      "entropy": 5.6005395412445065,
      "epoch": 2.8038411860504295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.0466,
      "mean_token_accuracy": 0.22034031748771668,
      "num_tokens": 55735113.0,
      "step": 24965
    },
    {
      "entropy": 5.461372327804566,
      "epoch": 2.804402762958387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000422581294913838,
      "loss": 4.8741,
      "mean_token_accuracy": 0.22668529599905013,
      "num_tokens": 55745857.0,
      "step": 24970
    },
    {
      "entropy": 5.570568323135376,
      "epoch": 2.804964339866345,
      "grad_norm": 1.0625,
      "learning_rate": 0.000422550986980957,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21985904425382613,
      "num_tokens": 55756856.0,
      "step": 24975
    },
    {
      "entropy": 5.647183513641357,
      "epoch": 2.805525916774302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21230387091636657,
      "num_tokens": 55768827.0,
      "step": 24980
    },
    {
      "entropy": 5.577859401702881,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004224903570219181,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2181962862610817,
      "num_tokens": 55780360.0,
      "step": 24985
    },
    {
      "entropy": 5.5767889499664305,
      "epoch": 2.8066490705902174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042246003499769077,
      "loss": 5.0089,
      "mean_token_accuracy": 0.219021038711071,
      "num_tokens": 55791874.0,
      "step": 24990
    },
    {
      "entropy": 5.526770734786988,
      "epoch": 2.8072106474981746,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042242970827827874,
      "loss": 4.9954,
      "mean_token_accuracy": 0.22230184078216553,
      "num_tokens": 55802689.0,
      "step": 24995
    },
    {
      "entropy": 5.586924648284912,
      "epoch": 2.8077722244061323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004223993768646476,
      "loss": 5.0092,
      "mean_token_accuracy": 0.22043342739343644,
      "num_tokens": 55813945.0,
      "step": 25000
    },
    {
      "entropy": 5.542327976226806,
      "epoch": 2.80833380131409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004223690407577631,
      "loss": 4.9777,
      "mean_token_accuracy": 0.21736541986465455,
      "num_tokens": 55826403.0,
      "step": 25005
    },
    {
      "entropy": 5.51151180267334,
      "epoch": 2.8088953782220476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042233869995859127,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22149670422077178,
      "num_tokens": 55836858.0,
      "step": 25010
    },
    {
      "entropy": 5.614374780654908,
      "epoch": 2.8094569551300053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042230835446809803,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2211397409439087,
      "num_tokens": 55847664.0,
      "step": 25015
    },
    {
      "entropy": 5.5472979068756105,
      "epoch": 2.8100185320379625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222780042872496,
      "loss": 4.9783,
      "mean_token_accuracy": 0.22156820297241211,
      "num_tokens": 55858333.0,
      "step": 25020
    },
    {
      "entropy": 5.536615991592408,
      "epoch": 2.81058010894592,
      "grad_norm": 1.0,
      "learning_rate": 0.0004222476494170125,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22500264197587966,
      "num_tokens": 55868819.0,
      "step": 25025
    },
    {
      "entropy": 5.550322532653809,
      "epoch": 2.811141685853878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004222172898583531,
      "loss": 4.961,
      "mean_token_accuracy": 0.21872843205928802,
      "num_tokens": 55879133.0,
      "step": 25030
    },
    {
      "entropy": 5.500312185287475,
      "epoch": 2.811703262761835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004221869256122381,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22244384735822678,
      "num_tokens": 55890362.0,
      "step": 25035
    },
    {
      "entropy": 5.594393014907837,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.125,
      "learning_rate": 0.0004221565566796343,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21950483322143555,
      "num_tokens": 55899920.0,
      "step": 25040
    },
    {
      "entropy": 5.616783332824707,
      "epoch": 2.8128264165777503,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004221261830615086,
      "loss": 5.0172,
      "mean_token_accuracy": 0.221028171479702,
      "num_tokens": 55911598.0,
      "step": 25045
    },
    {
      "entropy": 5.4534807205200195,
      "epoch": 2.813387993485708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004220958047588282,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2281452625989914,
      "num_tokens": 55922514.0,
      "step": 25050
    },
    {
      "entropy": 5.5493556499481205,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004220654217725603,
      "loss": 4.9453,
      "mean_token_accuracy": 0.23105451911687852,
      "num_tokens": 55932796.0,
      "step": 25055
    },
    {
      "entropy": 5.5772490978240965,
      "epoch": 2.814511147301623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.0404,
      "mean_token_accuracy": 0.22109757512807846,
      "num_tokens": 55944337.0,
      "step": 25060
    },
    {
      "entropy": 5.6823913097381595,
      "epoch": 2.8150727242095805,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000422004641753132,
      "loss": 5.1343,
      "mean_token_accuracy": 0.214174222946167,
      "num_tokens": 55955871.0,
      "step": 25065
    },
    {
      "entropy": 5.654292011260987,
      "epoch": 2.815634301117538,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21053965240716935,
      "num_tokens": 55966091.0,
      "step": 25070
    },
    {
      "entropy": 5.591201734542847,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042194384301096457,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22415365427732467,
      "num_tokens": 55976755.0,
      "step": 25075
    },
    {
      "entropy": 5.568748044967651,
      "epoch": 2.816757454933453,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042191343662127337,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21950575858354568,
      "num_tokens": 55987579.0,
      "step": 25080
    },
    {
      "entropy": 5.613466024398804,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042188302555380144,
      "loss": 4.995,
      "mean_token_accuracy": 0.21878103017807007,
      "num_tokens": 55997027.0,
      "step": 25085
    },
    {
      "entropy": 5.566482019424439,
      "epoch": 2.817880608749368,
      "grad_norm": 1.03125,
      "learning_rate": 0.000421852609809517,
      "loss": 5.013,
      "mean_token_accuracy": 0.21883176416158676,
      "num_tokens": 56008928.0,
      "step": 25090
    },
    {
      "entropy": 5.544407320022583,
      "epoch": 2.8184421856573256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004218221893893884,
      "loss": 5.006,
      "mean_token_accuracy": 0.22116678655147554,
      "num_tokens": 56021313.0,
      "step": 25095
    },
    {
      "entropy": 5.609735918045044,
      "epoch": 2.8190037625652833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042179176429438454,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2230122610926628,
      "num_tokens": 56032002.0,
      "step": 25100
    },
    {
      "entropy": 5.58848729133606,
      "epoch": 2.819565339473241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004217613345254739,
      "loss": 4.9564,
      "mean_token_accuracy": 0.23111863434314728,
      "num_tokens": 56042905.0,
      "step": 25105
    },
    {
      "entropy": 5.571083927154541,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042173090008362535,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21709786504507064,
      "num_tokens": 56052507.0,
      "step": 25110
    },
    {
      "entropy": 5.5085145950317385,
      "epoch": 2.820688493289156,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042170046096980807,
      "loss": 5.052,
      "mean_token_accuracy": 0.2162548452615738,
      "num_tokens": 56065250.0,
      "step": 25115
    },
    {
      "entropy": 5.545064973831177,
      "epoch": 2.8212500701971135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042167001718499124,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22955594062805176,
      "num_tokens": 56075858.0,
      "step": 25120
    },
    {
      "entropy": 5.575212717056274,
      "epoch": 2.821811647105071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21703319549560546,
      "num_tokens": 56086647.0,
      "step": 25125
    },
    {
      "entropy": 5.671165418624878,
      "epoch": 2.8223732240130284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.1854,
      "mean_token_accuracy": 0.20540368407964707,
      "num_tokens": 56098124.0,
      "step": 25130
    },
    {
      "entropy": 5.587487459182739,
      "epoch": 2.822934800920986,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004215786578142375,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21753343641757966,
      "num_tokens": 56108505.0,
      "step": 25135
    },
    {
      "entropy": 5.572417831420898,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042154819535511733,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2209621787071228,
      "num_tokens": 56118423.0,
      "step": 25140
    },
    {
      "entropy": 5.604850625991821,
      "epoch": 2.8240579547369014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.0387,
      "mean_token_accuracy": 0.22199587225914003,
      "num_tokens": 56130346.0,
      "step": 25145
    },
    {
      "entropy": 5.572243881225586,
      "epoch": 2.824619531644859,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.0568,
      "mean_token_accuracy": 0.22458325326442719,
      "num_tokens": 56141044.0,
      "step": 25150
    },
    {
      "entropy": 5.519549703598022,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004214567799847289,
      "loss": 4.9447,
      "mean_token_accuracy": 0.22679221332073213,
      "num_tokens": 56151786.0,
      "step": 25155
    },
    {
      "entropy": 5.571550416946411,
      "epoch": 2.825742685460774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004214262988668242,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21448703855276108,
      "num_tokens": 56164196.0,
      "step": 25160
    },
    {
      "entropy": 5.582658815383911,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.0354,
      "mean_token_accuracy": 0.22042931914329528,
      "num_tokens": 56176259.0,
      "step": 25165
    },
    {
      "entropy": 5.581323337554932,
      "epoch": 2.8268658392766888,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2111596718430519,
      "num_tokens": 56186617.0,
      "step": 25170
    },
    {
      "entropy": 5.480284786224365,
      "epoch": 2.8274274161846464,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042133482754337236,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22193263918161393,
      "num_tokens": 56197609.0,
      "step": 25175
    },
    {
      "entropy": 5.4970368385314945,
      "epoch": 2.827988993092604,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042130432778221186,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23197292685508727,
      "num_tokens": 56207490.0,
      "step": 25180
    },
    {
      "entropy": 5.5481523990631105,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042127382336266467,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2154833883047104,
      "num_tokens": 56218167.0,
      "step": 25185
    },
    {
      "entropy": 5.509481048583984,
      "epoch": 2.829112146908519,
      "grad_norm": 1.125,
      "learning_rate": 0.0004212433142857024,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2252012759447098,
      "num_tokens": 56229199.0,
      "step": 25190
    },
    {
      "entropy": 5.595104789733886,
      "epoch": 2.8296737238164766,
      "grad_norm": 1.0,
      "learning_rate": 0.00042121280055229624,
      "loss": 5.1193,
      "mean_token_accuracy": 0.21491046249866486,
      "num_tokens": 56240165.0,
      "step": 25195
    },
    {
      "entropy": 5.6850790023803714,
      "epoch": 2.8302353007244343,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004211822821634177,
      "loss": 5.1374,
      "mean_token_accuracy": 0.2145019456744194,
      "num_tokens": 56252447.0,
      "step": 25200
    },
    {
      "entropy": 5.631262683868409,
      "epoch": 2.830796877632392,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.0095,
      "mean_token_accuracy": 0.22407207041978836,
      "num_tokens": 56263116.0,
      "step": 25205
    },
    {
      "entropy": 5.580379390716553,
      "epoch": 2.831358454540349,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.0586,
      "mean_token_accuracy": 0.21293819695711136,
      "num_tokens": 56276257.0,
      "step": 25210
    },
    {
      "entropy": 5.519699668884277,
      "epoch": 2.831920031448307,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042109069907366646,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22100067287683486,
      "num_tokens": 56287865.0,
      "step": 25215
    },
    {
      "entropy": 5.605134963989258,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042106016207261754,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2270639047026634,
      "num_tokens": 56298122.0,
      "step": 25220
    },
    {
      "entropy": 5.64921612739563,
      "epoch": 2.8330431852642217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.0982,
      "mean_token_accuracy": 0.2178875043988228,
      "num_tokens": 56309455.0,
      "step": 25225
    },
    {
      "entropy": 5.619197082519531,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.041,
      "mean_token_accuracy": 0.2202734813094139,
      "num_tokens": 56319804.0,
      "step": 25230
    },
    {
      "entropy": 5.593238401412964,
      "epoch": 2.834166339080137,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042096852316968756,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22795910835266114,
      "num_tokens": 56330136.0,
      "step": 25235
    },
    {
      "entropy": 5.691605138778686,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.1693,
      "mean_token_accuracy": 0.21114054024219514,
      "num_tokens": 56341861.0,
      "step": 25240
    },
    {
      "entropy": 5.5180230140686035,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2146826833486557,
      "num_tokens": 56352016.0,
      "step": 25245
    },
    {
      "entropy": 5.486512565612793,
      "epoch": 2.8358510698040096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004208768424375089,
      "loss": 4.928,
      "mean_token_accuracy": 0.22150842398405074,
      "num_tokens": 56363740.0,
      "step": 25250
    },
    {
      "entropy": 5.602108430862427,
      "epoch": 2.8364126467119672,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042084627290260124,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22699320316314697,
      "num_tokens": 56375080.0,
      "step": 25255
    },
    {
      "entropy": 5.595633554458618,
      "epoch": 2.836974223619925,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004208156987238918,
      "loss": 5.0823,
      "mean_token_accuracy": 0.21420987248420714,
      "num_tokens": 56385858.0,
      "step": 25260
    },
    {
      "entropy": 5.660498714447021,
      "epoch": 2.837535800527882,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.0438,
      "mean_token_accuracy": 0.21765101850032806,
      "num_tokens": 56394423.0,
      "step": 25265
    },
    {
      "entropy": 5.555856704711914,
      "epoch": 2.83809737743584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21691131442785264,
      "num_tokens": 56405843.0,
      "step": 25270
    },
    {
      "entropy": 5.520572090148926,
      "epoch": 2.8386589543437974,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042072394833468765,
      "loss": 4.935,
      "mean_token_accuracy": 0.2209247201681137,
      "num_tokens": 56417757.0,
      "step": 25275
    },
    {
      "entropy": 5.589885473251343,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.046875,
      "learning_rate": 0.000420693355590507,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22375067323446274,
      "num_tokens": 56429937.0,
      "step": 25280
    },
    {
      "entropy": 5.565918016433716,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042066275820739333,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22702056616544725,
      "num_tokens": 56441114.0,
      "step": 25285
    },
    {
      "entropy": 5.52249174118042,
      "epoch": 2.84034368506767,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000420632156186321,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2166154831647873,
      "num_tokens": 56451839.0,
      "step": 25290
    },
    {
      "entropy": 5.6178265571594235,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2231721505522728,
      "num_tokens": 56463384.0,
      "step": 25295
    },
    {
      "entropy": 5.581218290328979,
      "epoch": 2.8414668388835853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042057093823419795,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21585605442523956,
      "num_tokens": 56474905.0,
      "step": 25300
    },
    {
      "entropy": 5.558044719696045,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042054032230509636,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2183586463332176,
      "num_tokens": 56485319.0,
      "step": 25305
    },
    {
      "entropy": 5.554645967483521,
      "epoch": 2.8425899926995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.0644,
      "mean_token_accuracy": 0.21631322503089906,
      "num_tokens": 56496064.0,
      "step": 25310
    },
    {
      "entropy": 5.556373214721679,
      "epoch": 2.843151569607458,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21797137409448625,
      "num_tokens": 56507167.0,
      "step": 25315
    },
    {
      "entropy": 5.680315208435059,
      "epoch": 2.843713146515415,
      "grad_norm": 1.125,
      "learning_rate": 0.0004204484467173297,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2209374949336052,
      "num_tokens": 56518400.0,
      "step": 25320
    },
    {
      "entropy": 5.669488763809204,
      "epoch": 2.8442747234233727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.1279,
      "mean_token_accuracy": 0.22028294950723648,
      "num_tokens": 56529630.0,
      "step": 25325
    },
    {
      "entropy": 5.65755443572998,
      "epoch": 2.8448363003313304,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21869464814662934,
      "num_tokens": 56539520.0,
      "step": 25330
    },
    {
      "entropy": 5.604018211364746,
      "epoch": 2.845397877239288,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004203565294493494,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22339892387390137,
      "num_tokens": 56550677.0,
      "step": 25335
    },
    {
      "entropy": 5.516450119018555,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.0618,
      "mean_token_accuracy": 0.21852022558450698,
      "num_tokens": 56561694.0,
      "step": 25340
    },
    {
      "entropy": 5.589140367507935,
      "epoch": 2.846521031055203,
      "grad_norm": 1.125,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21444917172193528,
      "num_tokens": 56572322.0,
      "step": 25345
    },
    {
      "entropy": 5.635671663284302,
      "epoch": 2.8470826079631606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.0264,
      "mean_token_accuracy": 0.2142741248011589,
      "num_tokens": 56583198.0,
      "step": 25350
    },
    {
      "entropy": 5.628100156784058,
      "epoch": 2.8476441848711183,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.0931,
      "mean_token_accuracy": 0.22161636501550674,
      "num_tokens": 56594732.0,
      "step": 25355
    },
    {
      "entropy": 5.5968011856079105,
      "epoch": 2.8482057617790755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000420203241451565,
      "loss": 5.0394,
      "mean_token_accuracy": 0.22027551382780075,
      "num_tokens": 56605907.0,
      "step": 25360
    },
    {
      "entropy": 5.644277286529541,
      "epoch": 2.848767338687033,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042017256997812016,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21813699156045913,
      "num_tokens": 56616841.0,
      "step": 25365
    },
    {
      "entropy": 5.613884544372558,
      "epoch": 2.849328915594991,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21441019624471663,
      "num_tokens": 56628044.0,
      "step": 25370
    },
    {
      "entropy": 5.600745725631714,
      "epoch": 2.849890492502948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21478781551122667,
      "num_tokens": 56638350.0,
      "step": 25375
    },
    {
      "entropy": 5.5860696792602536,
      "epoch": 2.8504520694109057,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004200805278275873,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2136698603630066,
      "num_tokens": 56649923.0,
      "step": 25380
    },
    {
      "entropy": 5.522228622436524,
      "epoch": 2.8510136463188633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004200498378705997,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22378322333097458,
      "num_tokens": 56661210.0,
      "step": 25385
    },
    {
      "entropy": 5.655221891403198,
      "epoch": 2.851575223226821,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004200191432951692,
      "loss": 5.1193,
      "mean_token_accuracy": 0.2094395011663437,
      "num_tokens": 56673106.0,
      "step": 25390
    },
    {
      "entropy": 5.563511753082276,
      "epoch": 2.8521368001347787,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2215463787317276,
      "num_tokens": 56684703.0,
      "step": 25395
    },
    {
      "entropy": 5.614510774612427,
      "epoch": 2.852698377042736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041995774029288807,
      "loss": 5.0024,
      "mean_token_accuracy": 0.22331321090459824,
      "num_tokens": 56696881.0,
      "step": 25400
    },
    {
      "entropy": 5.5443933486938475,
      "epoch": 2.8532599539506935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041992703186799273,
      "loss": 4.9861,
      "mean_token_accuracy": 0.22306904196739197,
      "num_tokens": 56707372.0,
      "step": 25405
    },
    {
      "entropy": 5.525978994369507,
      "epoch": 2.853821530858651,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041989631882856447,
      "loss": 4.92,
      "mean_token_accuracy": 0.22541695088148117,
      "num_tokens": 56717954.0,
      "step": 25410
    },
    {
      "entropy": 5.594223260879517,
      "epoch": 2.8543831077666084,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2189414158463478,
      "num_tokens": 56728798.0,
      "step": 25415
    },
    {
      "entropy": 5.561855316162109,
      "epoch": 2.854944684674566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004198348789100211,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21590952277183534,
      "num_tokens": 56740095.0,
      "step": 25420
    },
    {
      "entropy": 5.558118581771851,
      "epoch": 2.8555062615825237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041980415203286223,
      "loss": 4.9408,
      "mean_token_accuracy": 0.21523516178131102,
      "num_tokens": 56751002.0,
      "step": 25425
    },
    {
      "entropy": 5.564060926437378,
      "epoch": 2.8560678384904814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000419773420545083,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22377447485923768,
      "num_tokens": 56762244.0,
      "step": 25430
    },
    {
      "entropy": 5.615686130523682,
      "epoch": 2.856629415398439,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.023,
      "mean_token_accuracy": 0.2226770654320717,
      "num_tokens": 56772918.0,
      "step": 25435
    },
    {
      "entropy": 5.520317649841308,
      "epoch": 2.8571909923063963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004197119437415776,
      "loss": 4.9038,
      "mean_token_accuracy": 0.23154317140579223,
      "num_tokens": 56784069.0,
      "step": 25440
    },
    {
      "entropy": 5.506422519683838,
      "epoch": 2.857752569214354,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004196811984278089,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22556194961071013,
      "num_tokens": 56795187.0,
      "step": 25445
    },
    {
      "entropy": 5.5809722423553465,
      "epoch": 2.8583141461223116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004196504485073347,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2174427628517151,
      "num_tokens": 56806766.0,
      "step": 25450
    },
    {
      "entropy": 5.615195608139038,
      "epoch": 2.858875723030269,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21578108966350557,
      "num_tokens": 56817872.0,
      "step": 25455
    },
    {
      "entropy": 5.6641490936279295,
      "epoch": 2.8594372999382265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.111,
      "mean_token_accuracy": 0.2080081358551979,
      "num_tokens": 56828291.0,
      "step": 25460
    },
    {
      "entropy": 5.57077579498291,
      "epoch": 2.859998876846184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004195581711154709,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2233085125684738,
      "num_tokens": 56839165.0,
      "step": 25465
    },
    {
      "entropy": 5.527594470977784,
      "epoch": 2.8605604537541414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041952740277796717,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22602853029966355,
      "num_tokens": 56850754.0,
      "step": 25470
    },
    {
      "entropy": 5.511898851394653,
      "epoch": 2.861122030662099,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2193566530942917,
      "num_tokens": 56862798.0,
      "step": 25475
    },
    {
      "entropy": 5.51699800491333,
      "epoch": 2.8616836075700567,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004194658522985139,
      "loss": 4.95,
      "mean_token_accuracy": 0.22888269424438476,
      "num_tokens": 56873479.0,
      "step": 25480
    },
    {
      "entropy": 5.592377185821533,
      "epoch": 2.8622451844780143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004194350701585241,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22263701111078263,
      "num_tokens": 56883924.0,
      "step": 25485
    },
    {
      "entropy": 5.604985094070434,
      "epoch": 2.862806761385972,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.026,
      "mean_token_accuracy": 0.22422022074460984,
      "num_tokens": 56895622.0,
      "step": 25490
    },
    {
      "entropy": 5.537213659286499,
      "epoch": 2.8633683382939292,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004193734920829188,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21866356283426286,
      "num_tokens": 56906504.0,
      "step": 25495
    },
    {
      "entropy": 5.499357604980469,
      "epoch": 2.863929915201887,
      "grad_norm": 1.0625,
      "learning_rate": 0.000419342696149264,
      "loss": 4.9802,
      "mean_token_accuracy": 0.22114625573158264,
      "num_tokens": 56917228.0,
      "step": 25500
    },
    {
      "entropy": 5.590795183181763,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004193118956196818,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2217819333076477,
      "num_tokens": 56927981.0,
      "step": 25505
    },
    {
      "entropy": 5.573788261413574,
      "epoch": 2.8650530690178018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004192810904951528,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21901824921369553,
      "num_tokens": 56938552.0,
      "step": 25510
    },
    {
      "entropy": 5.563544893264771,
      "epoch": 2.8656146459257594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000419250280776658,
      "loss": 4.9192,
      "mean_token_accuracy": 0.23040911257267,
      "num_tokens": 56949902.0,
      "step": 25515
    },
    {
      "entropy": 5.629152250289917,
      "epoch": 2.866176222833717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.1074,
      "mean_token_accuracy": 0.2194615200161934,
      "num_tokens": 56961335.0,
      "step": 25520
    },
    {
      "entropy": 5.5360767364501955,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004191886475616947,
      "loss": 4.954,
      "mean_token_accuracy": 0.22819391787052154,
      "num_tokens": 56972542.0,
      "step": 25525
    },
    {
      "entropy": 5.545526170730591,
      "epoch": 2.8672993766496324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041915782406718875,
      "loss": 5.0332,
      "mean_token_accuracy": 0.22328097522258758,
      "num_tokens": 56983838.0,
      "step": 25530
    },
    {
      "entropy": 5.6138969421386715,
      "epoch": 2.8678609535575896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004191269959826418,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2242555633187294,
      "num_tokens": 56996125.0,
      "step": 25535
    },
    {
      "entropy": 5.599759674072265,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041909616330903544,
      "loss": 4.994,
      "mean_token_accuracy": 0.21808208376169205,
      "num_tokens": 57007599.0,
      "step": 25540
    },
    {
      "entropy": 5.551576280593872,
      "epoch": 2.868984107373505,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041906532604735127,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22213561534881593,
      "num_tokens": 57018102.0,
      "step": 25545
    },
    {
      "entropy": 5.581289005279541,
      "epoch": 2.869545684281462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.0882,
      "mean_token_accuracy": 0.21564992964267732,
      "num_tokens": 57029253.0,
      "step": 25550
    },
    {
      "entropy": 5.668344593048095,
      "epoch": 2.87010726118942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041900363776367747,
      "loss": 5.056,
      "mean_token_accuracy": 0.21355798989534377,
      "num_tokens": 57040485.0,
      "step": 25555
    },
    {
      "entropy": 5.512403249740601,
      "epoch": 2.8706688380973775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004189727867436519,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22931162565946578,
      "num_tokens": 57051340.0,
      "step": 25560
    },
    {
      "entropy": 5.452878856658936,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041894193113947703,
      "loss": 4.9502,
      "mean_token_accuracy": 0.216417133808136,
      "num_tokens": 57061726.0,
      "step": 25565
    },
    {
      "entropy": 5.57494854927063,
      "epoch": 2.8717919919132924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041891107095213523,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22031405866146087,
      "num_tokens": 57073367.0,
      "step": 25570
    },
    {
      "entropy": 5.643231964111328,
      "epoch": 2.87235356882125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.1005,
      "mean_token_accuracy": 0.22341191470623017,
      "num_tokens": 57085379.0,
      "step": 25575
    },
    {
      "entropy": 5.640823173522949,
      "epoch": 2.8729151457292077,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004188493368318814,
      "loss": 5.0335,
      "mean_token_accuracy": 0.22408192604780197,
      "num_tokens": 57096151.0,
      "step": 25580
    },
    {
      "entropy": 5.598472785949707,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000418818462900935,
      "loss": 5.0628,
      "mean_token_accuracy": 0.21280703097581863,
      "num_tokens": 57107378.0,
      "step": 25585
    },
    {
      "entropy": 5.5615967273712155,
      "epoch": 2.8740382995451226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21971744298934937,
      "num_tokens": 57119814.0,
      "step": 25590
    },
    {
      "entropy": 5.760375452041626,
      "epoch": 2.8745998764530802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.103,
      "mean_token_accuracy": 0.21882810741662978,
      "num_tokens": 57132070.0,
      "step": 25595
    },
    {
      "entropy": 5.700741529464722,
      "epoch": 2.875161453361038,
      "grad_norm": 1.1875,
      "learning_rate": 0.000418725813636615,
      "loss": 5.0998,
      "mean_token_accuracy": 0.21673182994127274,
      "num_tokens": 57143216.0,
      "step": 25600
    },
    {
      "entropy": 5.525060701370239,
      "epoch": 2.875723030268995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.0047,
      "mean_token_accuracy": 0.22576377540826797,
      "num_tokens": 57154485.0,
      "step": 25605
    },
    {
      "entropy": 5.610519647598267,
      "epoch": 2.876284607176953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2165905773639679,
      "num_tokens": 57164738.0,
      "step": 25610
    },
    {
      "entropy": 5.640727615356445,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.052,
      "mean_token_accuracy": 0.21466321498155594,
      "num_tokens": 57175790.0,
      "step": 25615
    },
    {
      "entropy": 5.578556680679322,
      "epoch": 2.877407760992868,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.0934,
      "mean_token_accuracy": 0.2170747399330139,
      "num_tokens": 57187201.0,
      "step": 25620
    },
    {
      "entropy": 5.565053701400757,
      "epoch": 2.8779693379008258,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041857130668348705,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22318926751613616,
      "num_tokens": 57197591.0,
      "step": 25625
    },
    {
      "entropy": 5.6514708518981935,
      "epoch": 2.878530914808783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21837160289287566,
      "num_tokens": 57210194.0,
      "step": 25630
    },
    {
      "entropy": 5.644469308853149,
      "epoch": 2.8790924917167406,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004185094718957848,
      "loss": 4.999,
      "mean_token_accuracy": 0.2174074500799179,
      "num_tokens": 57220752.0,
      "step": 25635
    },
    {
      "entropy": 5.565226221084595,
      "epoch": 2.8796540686246983,
      "grad_norm": 1.125,
      "learning_rate": 0.00041847854764734556,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22885596007108688,
      "num_tokens": 57233109.0,
      "step": 25640
    },
    {
      "entropy": 5.574466133117676,
      "epoch": 2.8802156455326555,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.1026,
      "mean_token_accuracy": 0.2143482208251953,
      "num_tokens": 57244548.0,
      "step": 25645
    },
    {
      "entropy": 5.563758754730225,
      "epoch": 2.880777222440613,
      "grad_norm": 1.125,
      "learning_rate": 0.00041841668544621396,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22564852982759476,
      "num_tokens": 57255328.0,
      "step": 25650
    },
    {
      "entropy": 5.602558755874634,
      "epoch": 2.881338799348571,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004183857474954915,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21747546046972274,
      "num_tokens": 57268890.0,
      "step": 25655
    },
    {
      "entropy": 5.6551464080810545,
      "epoch": 2.881900376256528,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21358465701341628,
      "num_tokens": 57278505.0,
      "step": 25660
    },
    {
      "entropy": 5.688299560546875,
      "epoch": 2.8824619531644857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2148282214999199,
      "num_tokens": 57289499.0,
      "step": 25665
    },
    {
      "entropy": 5.739356327056885,
      "epoch": 2.8830235300724434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.1971,
      "mean_token_accuracy": 0.21448376923799514,
      "num_tokens": 57300970.0,
      "step": 25670
    },
    {
      "entropy": 5.594494438171386,
      "epoch": 2.883585106980401,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041826195004787616,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22415292710065843,
      "num_tokens": 57311997.0,
      "step": 25675
    },
    {
      "entropy": 5.611565732955933,
      "epoch": 2.8841466838883587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.0637,
      "mean_token_accuracy": 0.2175603464245796,
      "num_tokens": 57323385.0,
      "step": 25680
    },
    {
      "entropy": 5.583552503585816,
      "epoch": 2.884708260796316,
      "grad_norm": 1.0,
      "learning_rate": 0.0004182000239510301,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21423603743314742,
      "num_tokens": 57335389.0,
      "step": 25685
    },
    {
      "entropy": 5.650975847244263,
      "epoch": 2.8852698377042736,
      "grad_norm": 1.125,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.0616,
      "mean_token_accuracy": 0.2180809870362282,
      "num_tokens": 57346843.0,
      "step": 25690
    },
    {
      "entropy": 5.6475342273712155,
      "epoch": 2.8858314146122312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21908630877733232,
      "num_tokens": 57358253.0,
      "step": 25695
    },
    {
      "entropy": 5.633585786819458,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004181071006116442,
      "loss": 5.1027,
      "mean_token_accuracy": 0.21224163323640824,
      "num_tokens": 57369204.0,
      "step": 25700
    },
    {
      "entropy": 5.584709167480469,
      "epoch": 2.886954568428146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041807611705069593,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2277815967798233,
      "num_tokens": 57380214.0,
      "step": 25705
    },
    {
      "entropy": 5.627397155761718,
      "epoch": 2.887516145336104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041804512893414865,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21769311875104905,
      "num_tokens": 57391848.0,
      "step": 25710
    },
    {
      "entropy": 5.702947759628296,
      "epoch": 2.8880777222440615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2151841476559639,
      "num_tokens": 57403026.0,
      "step": 25715
    },
    {
      "entropy": 5.586522722244263,
      "epoch": 2.888639299152019,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21862108558416365,
      "num_tokens": 57414223.0,
      "step": 25720
    },
    {
      "entropy": 5.560843181610108,
      "epoch": 2.8892008760599763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004179521372607798,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22406353950500488,
      "num_tokens": 57424661.0,
      "step": 25725
    },
    {
      "entropy": 5.550930547714233,
      "epoch": 2.889762452967934,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041792113093170447,
      "loss": 4.9665,
      "mean_token_accuracy": 0.22301300168037413,
      "num_tokens": 57435859.0,
      "step": 25730
    },
    {
      "entropy": 5.5747840881347654,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21414980739355088,
      "num_tokens": 57447784.0,
      "step": 25735
    },
    {
      "entropy": 5.589072036743164,
      "epoch": 2.890885606783849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004178591046225485,
      "loss": 5.0654,
      "mean_token_accuracy": 0.21422325521707536,
      "num_tokens": 57458881.0,
      "step": 25740
    },
    {
      "entropy": 5.572484302520752,
      "epoch": 2.8914471836918065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21537573039531707,
      "num_tokens": 57470287.0,
      "step": 25745
    },
    {
      "entropy": 5.663502836227417,
      "epoch": 2.892008760599764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.0837,
      "mean_token_accuracy": 0.21067904084920883,
      "num_tokens": 57481723.0,
      "step": 25750
    },
    {
      "entropy": 5.677995681762695,
      "epoch": 2.8925703375077214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041776603104611427,
      "loss": 5.07,
      "mean_token_accuracy": 0.217250294983387,
      "num_tokens": 57492804.0,
      "step": 25755
    },
    {
      "entropy": 5.537966632843018,
      "epoch": 2.893131914415679,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041773499742786736,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2224613457918167,
      "num_tokens": 57504148.0,
      "step": 25760
    },
    {
      "entropy": 5.566112089157104,
      "epoch": 2.8936934913236367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.0345,
      "mean_token_accuracy": 0.22237896174192429,
      "num_tokens": 57516026.0,
      "step": 25765
    },
    {
      "entropy": 5.602253532409668,
      "epoch": 2.8942550682315944,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21200381070375443,
      "num_tokens": 57526882.0,
      "step": 25770
    },
    {
      "entropy": 5.546275091171265,
      "epoch": 2.894816645139552,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041764186930865377,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2232278913259506,
      "num_tokens": 57537875.0,
      "step": 25775
    },
    {
      "entropy": 5.625234842300415,
      "epoch": 2.8953782220475093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004176108175173863,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21952898055315018,
      "num_tokens": 57548143.0,
      "step": 25780
    },
    {
      "entropy": 5.554040431976318,
      "epoch": 2.895939798955467,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004175797611853351,
      "loss": 5.003,
      "mean_token_accuracy": 0.2160586193203926,
      "num_tokens": 57558127.0,
      "step": 25785
    },
    {
      "entropy": 5.590300703048706,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000417548700313489,
      "loss": 5.0586,
      "mean_token_accuracy": 0.22761774510145188,
      "num_tokens": 57570028.0,
      "step": 25790
    },
    {
      "entropy": 5.614522314071655,
      "epoch": 2.897062952771382,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000417517634902837,
      "loss": 4.9692,
      "mean_token_accuracy": 0.21977844685316086,
      "num_tokens": 57581089.0,
      "step": 25795
    },
    {
      "entropy": 5.655324506759643,
      "epoch": 2.8976245296793395,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.1224,
      "mean_token_accuracy": 0.21721505075693132,
      "num_tokens": 57592555.0,
      "step": 25800
    },
    {
      "entropy": 5.571609354019165,
      "epoch": 2.898186106587297,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041745549046907205,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21958452463150024,
      "num_tokens": 57603492.0,
      "step": 25805
    },
    {
      "entropy": 5.523623752593994,
      "epoch": 2.898747683495255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041742441144793774,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21952830404043197,
      "num_tokens": 57613477.0,
      "step": 25810
    },
    {
      "entropy": 5.5999102115631105,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004173933278919549,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21374092400074005,
      "num_tokens": 57623703.0,
      "step": 25815
    },
    {
      "entropy": 5.656207466125489,
      "epoch": 2.8998708373111697,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004173622398021133,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22194262444972992,
      "num_tokens": 57637172.0,
      "step": 25820
    },
    {
      "entropy": 5.615019130706787,
      "epoch": 2.9004324142191273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041733114717940273,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2158161073923111,
      "num_tokens": 57648224.0,
      "step": 25825
    },
    {
      "entropy": 5.584861326217651,
      "epoch": 2.900993991127085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004173000500248133,
      "loss": 5.0348,
      "mean_token_accuracy": 0.23003637939691543,
      "num_tokens": 57659093.0,
      "step": 25830
    },
    {
      "entropy": 5.651944017410278,
      "epoch": 2.901555568035042,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.1056,
      "mean_token_accuracy": 0.21817418932914734,
      "num_tokens": 57671399.0,
      "step": 25835
    },
    {
      "entropy": 5.679384040832519,
      "epoch": 2.902117144943,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.0245,
      "mean_token_accuracy": 0.22144753932952882,
      "num_tokens": 57682150.0,
      "step": 25840
    },
    {
      "entropy": 5.590591859817505,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041720673137967356,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2178274616599083,
      "num_tokens": 57693232.0,
      "step": 25845
    },
    {
      "entropy": 5.563284587860108,
      "epoch": 2.9032402987589148,
      "grad_norm": 1.125,
      "learning_rate": 0.00041717561610747124,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21882018446922302,
      "num_tokens": 57703422.0,
      "step": 25850
    },
    {
      "entropy": 5.582779836654663,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.0147,
      "mean_token_accuracy": 0.21573485881090165,
      "num_tokens": 57714454.0,
      "step": 25855
    },
    {
      "entropy": 5.576024961471558,
      "epoch": 2.90436345257483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004171133719832774,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23014864921569825,
      "num_tokens": 57726243.0,
      "step": 25860
    },
    {
      "entropy": 5.5571897506713865,
      "epoch": 2.9049250294827877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041708224313326747,
      "loss": 4.9843,
      "mean_token_accuracy": 0.2169030115008354,
      "num_tokens": 57737882.0,
      "step": 25865
    },
    {
      "entropy": 5.6396819114685055,
      "epoch": 2.9054866063907454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417051109759304,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2249320238828659,
      "num_tokens": 57750205.0,
      "step": 25870
    },
    {
      "entropy": 5.566202545166016,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.125,
      "learning_rate": 0.000417019971862378,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22730337530374528,
      "num_tokens": 57761762.0,
      "step": 25875
    },
    {
      "entropy": 5.600724744796753,
      "epoch": 2.9066097602066603,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21449862867593766,
      "num_tokens": 57771280.0,
      "step": 25880
    },
    {
      "entropy": 5.613783359527588,
      "epoch": 2.907171337114618,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041695768250360466,
      "loss": 4.9797,
      "mean_token_accuracy": 0.21861250698566437,
      "num_tokens": 57784099.0,
      "step": 25885
    },
    {
      "entropy": 5.613382244110108,
      "epoch": 2.907732914022575,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2170623406767845,
      "num_tokens": 57795544.0,
      "step": 25890
    },
    {
      "entropy": 5.523683309555054,
      "epoch": 2.908294490930533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004168953750648809,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22913769632577896,
      "num_tokens": 57805963.0,
      "step": 25895
    },
    {
      "entropy": 5.620227909088134,
      "epoch": 2.9088560678384905,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2137061133980751,
      "num_tokens": 57815281.0,
      "step": 25900
    },
    {
      "entropy": 5.6322211742401125,
      "epoch": 2.909417644746448,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21436946243047714,
      "num_tokens": 57826172.0,
      "step": 25905
    },
    {
      "entropy": 5.634811592102051,
      "epoch": 2.909979221654406,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041680188002424755,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22140201032161713,
      "num_tokens": 57836617.0,
      "step": 25910
    },
    {
      "entropy": 5.592221117019653,
      "epoch": 2.910540798562363,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004167707059793262,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22624870240688325,
      "num_tokens": 57847636.0,
      "step": 25915
    },
    {
      "entropy": 5.64596905708313,
      "epoch": 2.9111023754703207,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041673952742037057,
      "loss": 5.1105,
      "mean_token_accuracy": 0.2161758065223694,
      "num_tokens": 57858816.0,
      "step": 25920
    },
    {
      "entropy": 5.629503536224365,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.0488,
      "mean_token_accuracy": 0.22242549657821656,
      "num_tokens": 57870749.0,
      "step": 25925
    },
    {
      "entropy": 5.5872907638549805,
      "epoch": 2.9122255292862356,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.0464,
      "mean_token_accuracy": 0.2194126844406128,
      "num_tokens": 57883454.0,
      "step": 25930
    },
    {
      "entropy": 5.5849597454071045,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004166459646692262,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2223423093557358,
      "num_tokens": 57894498.0,
      "step": 25935
    },
    {
      "entropy": 5.515616464614868,
      "epoch": 2.913348683102151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004166147680640623,
      "loss": 4.8521,
      "mean_token_accuracy": 0.23164899200201033,
      "num_tokens": 57905332.0,
      "step": 25940
    },
    {
      "entropy": 5.5360723495483395,
      "epoch": 2.913910260010108,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041658356694982936,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22640062719583512,
      "num_tokens": 57916456.0,
      "step": 25945
    },
    {
      "entropy": 5.561935234069824,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.0242,
      "mean_token_accuracy": 0.22651287764310837,
      "num_tokens": 57927615.0,
      "step": 25950
    },
    {
      "entropy": 5.53674349784851,
      "epoch": 2.9150334138260234,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004165211511981299,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2300539568066597,
      "num_tokens": 57937771.0,
      "step": 25955
    },
    {
      "entropy": 5.542462539672852,
      "epoch": 2.915594990733981,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004164899365626507,
      "loss": 4.885,
      "mean_token_accuracy": 0.22034025341272354,
      "num_tokens": 57948564.0,
      "step": 25960
    },
    {
      "entropy": 5.520370054244995,
      "epoch": 2.9161565676419388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004164587174220771,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2225578621029854,
      "num_tokens": 57959758.0,
      "step": 25965
    },
    {
      "entropy": 5.595324516296387,
      "epoch": 2.916718144549896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2197493627667427,
      "num_tokens": 57970988.0,
      "step": 25970
    },
    {
      "entropy": 5.570679378509522,
      "epoch": 2.9172797214578536,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041639626562962293,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2304811507463455,
      "num_tokens": 57981056.0,
      "step": 25975
    },
    {
      "entropy": 5.5430840969085695,
      "epoch": 2.9178412983658113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2269206762313843,
      "num_tokens": 57992092.0,
      "step": 25980
    },
    {
      "entropy": 5.62078890800476,
      "epoch": 2.9184028752737685,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000416333795828721,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22632600963115693,
      "num_tokens": 58003000.0,
      "step": 25985
    },
    {
      "entropy": 5.562895250320435,
      "epoch": 2.918964452181726,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004163025541775885,
      "loss": 4.9709,
      "mean_token_accuracy": 0.21766369491815568,
      "num_tokens": 58013932.0,
      "step": 25990
    },
    {
      "entropy": 5.666604042053223,
      "epoch": 2.919526029089684,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.0609,
      "mean_token_accuracy": 0.22104134261608124,
      "num_tokens": 58024481.0,
      "step": 25995
    },
    {
      "entropy": 5.61755862236023,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.0191,
      "mean_token_accuracy": 0.2214810222387314,
      "num_tokens": 58035753.0,
      "step": 26000
    },
    {
      "entropy": 5.704206132888794,
      "epoch": 2.920649182905599,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.1555,
      "mean_token_accuracy": 0.21011510640382766,
      "num_tokens": 58046599.0,
      "step": 26005
    },
    {
      "entropy": 5.6567765235900875,
      "epoch": 2.9212107598135564,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041617754259172665,
      "loss": 5.0803,
      "mean_token_accuracy": 0.21270370185375215,
      "num_tokens": 58058151.0,
      "step": 26010
    },
    {
      "entropy": 5.658368968963623,
      "epoch": 2.921772336721514,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.0342,
      "mean_token_accuracy": 0.221675506234169,
      "num_tokens": 58070116.0,
      "step": 26015
    },
    {
      "entropy": 5.620903491973877,
      "epoch": 2.9223339136294717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.0243,
      "mean_token_accuracy": 0.22097396701574326,
      "num_tokens": 58081089.0,
      "step": 26020
    },
    {
      "entropy": 5.501225852966309,
      "epoch": 2.922895490537429,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041608373669979696,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2262648269534111,
      "num_tokens": 58092033.0,
      "step": 26025
    },
    {
      "entropy": 5.570938062667847,
      "epoch": 2.9234570674453866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004160524590835042,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22078582197427749,
      "num_tokens": 58102864.0,
      "step": 26030
    },
    {
      "entropy": 5.534692192077637,
      "epoch": 2.9240186443533442,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041602117697604664,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20868182182312012,
      "num_tokens": 58114607.0,
      "step": 26035
    },
    {
      "entropy": 5.614441013336181,
      "epoch": 2.9245802212613015,
      "grad_norm": 1.0,
      "learning_rate": 0.00041598989037842,
      "loss": 5.0204,
      "mean_token_accuracy": 0.22077268809080125,
      "num_tokens": 58125935.0,
      "step": 26040
    },
    {
      "entropy": 5.597686243057251,
      "epoch": 2.925141798169259,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004159585992916206,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2237626940011978,
      "num_tokens": 58136439.0,
      "step": 26045
    },
    {
      "entropy": 5.618555736541748,
      "epoch": 2.925703375077217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2176521211862564,
      "num_tokens": 58147778.0,
      "step": 26050
    },
    {
      "entropy": 5.633947038650513,
      "epoch": 2.9262649519851744,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041589600365448884,
      "loss": 5.0184,
      "mean_token_accuracy": 0.22408173829317093,
      "num_tokens": 58159290.0,
      "step": 26055
    },
    {
      "entropy": 5.61400260925293,
      "epoch": 2.926826528893132,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21134112030267715,
      "num_tokens": 58169917.0,
      "step": 26060
    },
    {
      "entropy": 5.576641035079956,
      "epoch": 2.9273881058010893,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004158333900726235,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2250667005777359,
      "num_tokens": 58180905.0,
      "step": 26065
    },
    {
      "entropy": 5.602845525741577,
      "epoch": 2.927949682709047,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.0099,
      "mean_token_accuracy": 0.22072386145591735,
      "num_tokens": 58192312.0,
      "step": 26070
    },
    {
      "entropy": 5.540474367141724,
      "epoch": 2.9285112596170046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004157707585539994,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2309114456176758,
      "num_tokens": 58204739.0,
      "step": 26075
    },
    {
      "entropy": 5.574822330474854,
      "epoch": 2.929072836524962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041573943607089536,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22134900242090225,
      "num_tokens": 58215663.0,
      "step": 26080
    },
    {
      "entropy": 5.567423343658447,
      "epoch": 2.9296344134329195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004157081091065932,
      "loss": 4.9938,
      "mean_token_accuracy": 0.220917509496212,
      "num_tokens": 58226480.0,
      "step": 26085
    },
    {
      "entropy": 5.7500543117523195,
      "epoch": 2.930195990340877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.2484,
      "mean_token_accuracy": 0.20866832584142686,
      "num_tokens": 58238087.0,
      "step": 26090
    },
    {
      "entropy": 5.5660158634185795,
      "epoch": 2.930757567248835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000415645441738384,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21390675604343415,
      "num_tokens": 58248738.0,
      "step": 26095
    },
    {
      "entropy": 5.538493251800537,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041561410133647244,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2201481357216835,
      "num_tokens": 58260440.0,
      "step": 26100
    },
    {
      "entropy": 5.654761934280396,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.0564,
      "mean_token_accuracy": 0.21754082888364792,
      "num_tokens": 58272704.0,
      "step": 26105
    },
    {
      "entropy": 5.605885171890259,
      "epoch": 2.9324422979727074,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041555140710202484,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21658962219953537,
      "num_tokens": 58284254.0,
      "step": 26110
    },
    {
      "entropy": 5.632247400283814,
      "epoch": 2.933003874880665,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.1067,
      "mean_token_accuracy": 0.21685757637023925,
      "num_tokens": 58296472.0,
      "step": 26115
    },
    {
      "entropy": 5.545483160018921,
      "epoch": 2.9335654517886223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004154886949667321,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2296714186668396,
      "num_tokens": 58307734.0,
      "step": 26120
    },
    {
      "entropy": 5.599571084976196,
      "epoch": 2.93412702869658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041545733218876466,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21950629651546477,
      "num_tokens": 58318329.0,
      "step": 26125
    },
    {
      "entropy": 5.513032007217407,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004154259649385813,
      "loss": 4.9704,
      "mean_token_accuracy": 0.21915198266506195,
      "num_tokens": 58330036.0,
      "step": 26130
    },
    {
      "entropy": 5.58485517501831,
      "epoch": 2.935250182512495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041539459321718076,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22689759731292725,
      "num_tokens": 58340303.0,
      "step": 26135
    },
    {
      "entropy": 5.586071586608886,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004153632170255619,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23410147726535796,
      "num_tokens": 58352394.0,
      "step": 26140
    },
    {
      "entropy": 5.637967777252197,
      "epoch": 2.93637333632841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.1246,
      "mean_token_accuracy": 0.21274390667676926,
      "num_tokens": 58364605.0,
      "step": 26145
    },
    {
      "entropy": 5.594564151763916,
      "epoch": 2.936934913236368,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004153004512356656,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2276041477918625,
      "num_tokens": 58375514.0,
      "step": 26150
    },
    {
      "entropy": 5.609416675567627,
      "epoch": 2.9374964901443255,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004152690616393866,
      "loss": 4.993,
      "mean_token_accuracy": 0.22454345226287842,
      "num_tokens": 58387135.0,
      "step": 26155
    },
    {
      "entropy": 5.555465650558472,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041523766757688634,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21825236231088638,
      "num_tokens": 58399077.0,
      "step": 26160
    },
    {
      "entropy": 5.695095682144165,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.125,
      "mean_token_accuracy": 0.22403019815683364,
      "num_tokens": 58409897.0,
      "step": 26165
    },
    {
      "entropy": 5.519167804718018,
      "epoch": 2.939181220868198,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004151748660572203,
      "loss": 4.9301,
      "mean_token_accuracy": 0.23374995738267898,
      "num_tokens": 58421189.0,
      "step": 26170
    },
    {
      "entropy": 5.547763776779175,
      "epoch": 2.939742797776155,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004151434586020541,
      "loss": 5.0237,
      "mean_token_accuracy": 0.21700484156608582,
      "num_tokens": 58432258.0,
      "step": 26175
    },
    {
      "entropy": 5.650778293609619,
      "epoch": 2.940304374684113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.0573,
      "mean_token_accuracy": 0.2222208559513092,
      "num_tokens": 58444288.0,
      "step": 26180
    },
    {
      "entropy": 5.66866979598999,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041508063030605585,
      "loss": 5.0226,
      "mean_token_accuracy": 0.22542050033807753,
      "num_tokens": 58453367.0,
      "step": 26185
    },
    {
      "entropy": 5.637253284454346,
      "epoch": 2.941427528500028,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041504920946722413,
      "loss": 5.1542,
      "mean_token_accuracy": 0.2115066409111023,
      "num_tokens": 58464862.0,
      "step": 26190
    },
    {
      "entropy": 5.598409938812256,
      "epoch": 2.941989105407986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.007,
      "mean_token_accuracy": 0.22939672470092773,
      "num_tokens": 58476353.0,
      "step": 26195
    },
    {
      "entropy": 5.639093637466431,
      "epoch": 2.942550682315943,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21162650883197784,
      "num_tokens": 58488330.0,
      "step": 26200
    },
    {
      "entropy": 5.601886415481568,
      "epoch": 2.9431122592239007,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041495492019940467,
      "loss": 4.9954,
      "mean_token_accuracy": 0.22765822112560272,
      "num_tokens": 58500004.0,
      "step": 26205
    },
    {
      "entropy": 5.539389324188233,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004149234815296925,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21526535749435424,
      "num_tokens": 58511433.0,
      "step": 26210
    },
    {
      "entropy": 5.521959543228149,
      "epoch": 2.9442354130398156,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041489203840476243,
      "loss": 4.9359,
      "mean_token_accuracy": 0.21600623428821564,
      "num_tokens": 58522234.0,
      "step": 26215
    },
    {
      "entropy": 5.632538604736328,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.0225,
      "mean_token_accuracy": 0.22278033047914506,
      "num_tokens": 58534370.0,
      "step": 26220
    },
    {
      "entropy": 5.653337383270264,
      "epoch": 2.945358566855731,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004148291387932533,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22723967880010604,
      "num_tokens": 58545859.0,
      "step": 26225
    },
    {
      "entropy": 5.5322731018066404,
      "epoch": 2.9459201437636886,
      "grad_norm": 1.0,
      "learning_rate": 0.00041479768230867696,
      "loss": 4.9439,
      "mean_token_accuracy": 0.23416736871004104,
      "num_tokens": 58557159.0,
      "step": 26230
    },
    {
      "entropy": 5.578436803817749,
      "epoch": 2.946481720671646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041476622137288823,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2155196636915207,
      "num_tokens": 58568985.0,
      "step": 26235
    },
    {
      "entropy": 5.634815120697022,
      "epoch": 2.9470432975796035,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041473475598688886,
      "loss": 5.0095,
      "mean_token_accuracy": 0.22339210510253907,
      "num_tokens": 58579897.0,
      "step": 26240
    },
    {
      "entropy": 5.5571599960327145,
      "epoch": 2.947604874487561,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004147032861516805,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22709094882011413,
      "num_tokens": 58590042.0,
      "step": 26245
    },
    {
      "entropy": 5.581413221359253,
      "epoch": 2.948166451395519,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.0765,
      "mean_token_accuracy": 0.22158852517604827,
      "num_tokens": 58601645.0,
      "step": 26250
    },
    {
      "entropy": 5.583115911483764,
      "epoch": 2.948728028303476,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.0956,
      "mean_token_accuracy": 0.21628828793764115,
      "num_tokens": 58613881.0,
      "step": 26255
    },
    {
      "entropy": 5.59061369895935,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041460884996082363,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22673786133527757,
      "num_tokens": 58623430.0,
      "step": 26260
    },
    {
      "entropy": 5.549013423919678,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041457736233880166,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21436901539564132,
      "num_tokens": 58635289.0,
      "step": 26265
    },
    {
      "entropy": 5.495336961746216,
      "epoch": 2.9504127590273486,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004145458702725823,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2179991364479065,
      "num_tokens": 58646086.0,
      "step": 26270
    },
    {
      "entropy": 5.577433252334595,
      "epoch": 2.9509743359353062,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041451437376316825,
      "loss": 4.9771,
      "mean_token_accuracy": 0.21610312461853026,
      "num_tokens": 58657662.0,
      "step": 26275
    },
    {
      "entropy": 5.6374939441680905,
      "epoch": 2.951535912843264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004144828728115624,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22490467727184296,
      "num_tokens": 58668054.0,
      "step": 26280
    },
    {
      "entropy": 5.639890670776367,
      "epoch": 2.9520974897512215,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.0975,
      "mean_token_accuracy": 0.21883268803358077,
      "num_tokens": 58679285.0,
      "step": 26285
    },
    {
      "entropy": 5.606722784042359,
      "epoch": 2.952659066659179,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.0876,
      "mean_token_accuracy": 0.22413426041603088,
      "num_tokens": 58691878.0,
      "step": 26290
    },
    {
      "entropy": 5.5102152824401855,
      "epoch": 2.9532206435671364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004143883433136252,
      "loss": 4.942,
      "mean_token_accuracy": 0.22438009977340698,
      "num_tokens": 58702130.0,
      "step": 26295
    },
    {
      "entropy": 5.593381261825561,
      "epoch": 2.953782220475094,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004143568246032838,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22713501304388045,
      "num_tokens": 58712872.0,
      "step": 26300
    },
    {
      "entropy": 5.615232324600219,
      "epoch": 2.9543437973830518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.0643,
      "mean_token_accuracy": 0.21327606439590455,
      "num_tokens": 58724783.0,
      "step": 26305
    },
    {
      "entropy": 5.646614503860474,
      "epoch": 2.954905374291009,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041429377387207865,
      "loss": 5.0663,
      "mean_token_accuracy": 0.21575479507446288,
      "num_tokens": 58736473.0,
      "step": 26310
    },
    {
      "entropy": 5.646933269500733,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2159762278199196,
      "num_tokens": 58748223.0,
      "step": 26315
    },
    {
      "entropy": 5.535382986068726,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142307054002025,
      "loss": 4.9126,
      "mean_token_accuracy": 0.23071178793907166,
      "num_tokens": 58759876.0,
      "step": 26320
    },
    {
      "entropy": 5.603032445907592,
      "epoch": 2.956590105014882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.0375,
      "mean_token_accuracy": 0.22269703298807145,
      "num_tokens": 58771288.0,
      "step": 26325
    },
    {
      "entropy": 5.647957801818848,
      "epoch": 2.957151681922839,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041416761919568784,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21296254098415374,
      "num_tokens": 58781664.0,
      "step": 26330
    },
    {
      "entropy": 5.609287261962891,
      "epoch": 2.957713258830797,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21598078161478043,
      "num_tokens": 58792808.0,
      "step": 26335
    },
    {
      "entropy": 5.540486621856689,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004141045152665695,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22886265963315963,
      "num_tokens": 58804455.0,
      "step": 26340
    },
    {
      "entropy": 5.541367626190185,
      "epoch": 2.958836412646712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041407295665779533,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2272347092628479,
      "num_tokens": 58814718.0,
      "step": 26345
    },
    {
      "entropy": 5.609828853607178,
      "epoch": 2.9593979895546694,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.066,
      "mean_token_accuracy": 0.21754777431488037,
      "num_tokens": 58825649.0,
      "step": 26350
    },
    {
      "entropy": 5.623872089385986,
      "epoch": 2.959959566462627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21512219309806824,
      "num_tokens": 58836529.0,
      "step": 26355
    },
    {
      "entropy": 5.613033390045166,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.0229,
      "mean_token_accuracy": 0.21933559328317642,
      "num_tokens": 58847879.0,
      "step": 26360
    },
    {
      "entropy": 5.658273839950562,
      "epoch": 2.961082720278542,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004139466779513805,
      "loss": 5.1604,
      "mean_token_accuracy": 0.21273771822452545,
      "num_tokens": 58859283.0,
      "step": 26365
    },
    {
      "entropy": 5.638077354431152,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.0472,
      "mean_token_accuracy": 0.21325098723173141,
      "num_tokens": 58870529.0,
      "step": 26370
    },
    {
      "entropy": 5.629994964599609,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041388351204945534,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2274690881371498,
      "num_tokens": 58880828.0,
      "step": 26375
    },
    {
      "entropy": 5.689904117584229,
      "epoch": 2.962767451002415,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.0832,
      "mean_token_accuracy": 0.2250263974070549,
      "num_tokens": 58892121.0,
      "step": 26380
    },
    {
      "entropy": 5.601434421539307,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21580485254526138,
      "num_tokens": 58902768.0,
      "step": 26385
    },
    {
      "entropy": 5.65405387878418,
      "epoch": 2.9638906048183298,
      "grad_norm": 1.125,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21984906047582625,
      "num_tokens": 58913404.0,
      "step": 26390
    },
    {
      "entropy": 5.702129554748535,
      "epoch": 2.9644521817262874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21488552242517472,
      "num_tokens": 58925086.0,
      "step": 26395
    },
    {
      "entropy": 5.575828123092651,
      "epoch": 2.965013758634245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004137255199254112,
      "loss": 5.0039,
      "mean_token_accuracy": 0.22010749578475952,
      "num_tokens": 58935425.0,
      "step": 26400
    },
    {
      "entropy": 5.598635149002075,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.0226,
      "mean_token_accuracy": 0.21772139370441437,
      "num_tokens": 58945528.0,
      "step": 26405
    },
    {
      "entropy": 5.587639188766479,
      "epoch": 2.96613691245016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041366229214922926,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22252168357372284,
      "num_tokens": 58957008.0,
      "step": 26410
    },
    {
      "entropy": 5.630810499191284,
      "epoch": 2.9666984893581176,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004136306716380432,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21012120991945266,
      "num_tokens": 58968648.0,
      "step": 26415
    },
    {
      "entropy": 5.629441165924073,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041359904671280285,
      "loss": 5.0313,
      "mean_token_accuracy": 0.2175984039902687,
      "num_tokens": 58979794.0,
      "step": 26420
    },
    {
      "entropy": 5.576268720626831,
      "epoch": 2.9678216431740325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004135674173745151,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22458766996860505,
      "num_tokens": 58990534.0,
      "step": 26425
    },
    {
      "entropy": 5.564456367492676,
      "epoch": 2.96838322008199,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041353578362418695,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22139880955219268,
      "num_tokens": 59001717.0,
      "step": 26430
    },
    {
      "entropy": 5.552737092971801,
      "epoch": 2.968944796989948,
      "grad_norm": 1.125,
      "learning_rate": 0.0004135041454628258,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21807296574115753,
      "num_tokens": 59012118.0,
      "step": 26435
    },
    {
      "entropy": 5.671481895446777,
      "epoch": 2.9695063738979055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004134725028914389,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21620264947414397,
      "num_tokens": 59023940.0,
      "step": 26440
    },
    {
      "entropy": 5.643783235549927,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041344085591103383,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22160876542329788,
      "num_tokens": 59035147.0,
      "step": 26445
    },
    {
      "entropy": 5.506395721435547,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004134092045226182,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2245059922337532,
      "num_tokens": 59045531.0,
      "step": 26450
    },
    {
      "entropy": 5.626800155639648,
      "epoch": 2.971191104621778,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041337754872719984,
      "loss": 5.1116,
      "mean_token_accuracy": 0.22233861833810806,
      "num_tokens": 59056364.0,
      "step": 26455
    },
    {
      "entropy": 5.572789859771729,
      "epoch": 2.9717526815297353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004133458885257866,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2222903177142143,
      "num_tokens": 59066504.0,
      "step": 26460
    },
    {
      "entropy": 5.5498956680297855,
      "epoch": 2.972314258437693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041331422391938663,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22405615150928498,
      "num_tokens": 59077888.0,
      "step": 26465
    },
    {
      "entropy": 5.619155645370483,
      "epoch": 2.9728758353456506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004132825549090082,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22275225520133973,
      "num_tokens": 59088222.0,
      "step": 26470
    },
    {
      "entropy": 5.626445293426514,
      "epoch": 2.9734374122536082,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004132508814956595,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21573846638202668,
      "num_tokens": 59099302.0,
      "step": 26475
    },
    {
      "entropy": 5.599838781356811,
      "epoch": 2.973998989161566,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004132192036803491,
      "loss": 4.9951,
      "mean_token_accuracy": 0.2174055188894272,
      "num_tokens": 59109238.0,
      "step": 26480
    },
    {
      "entropy": 5.485777425765991,
      "epoch": 2.974560566069523,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041318752146408575,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22521258443593978,
      "num_tokens": 59119795.0,
      "step": 26485
    },
    {
      "entropy": 5.505061674118042,
      "epoch": 2.975122142977481,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004131558348478781,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2170192301273346,
      "num_tokens": 59129649.0,
      "step": 26490
    },
    {
      "entropy": 5.634202194213867,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.072,
      "mean_token_accuracy": 0.21550042927265167,
      "num_tokens": 59140802.0,
      "step": 26495
    },
    {
      "entropy": 5.644146823883057,
      "epoch": 2.9762452967933957,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041309244841966565,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2227318063378334,
      "num_tokens": 59151333.0,
      "step": 26500
    },
    {
      "entropy": 5.623249816894531,
      "epoch": 2.9768068737013533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21820364892482758,
      "num_tokens": 59163414.0,
      "step": 26505
    },
    {
      "entropy": 5.643855381011963,
      "epoch": 2.977368450609311,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041302904440378494,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2215748429298401,
      "num_tokens": 59173884.0,
      "step": 26510
    },
    {
      "entropy": 5.6350874423980715,
      "epoch": 2.9779300275172687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21306706219911575,
      "num_tokens": 59185263.0,
      "step": 26515
    },
    {
      "entropy": 5.6928027153015135,
      "epoch": 2.978491604425226,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041296562280831106,
      "loss": 5.1427,
      "mean_token_accuracy": 0.2064191848039627,
      "num_tokens": 59196279.0,
      "step": 26520
    },
    {
      "entropy": 5.489399480819702,
      "epoch": 2.9790531813331835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041293390542075095,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22855395674705506,
      "num_tokens": 59207537.0,
      "step": 26525
    },
    {
      "entropy": 5.539731168746949,
      "epoch": 2.979614758241141,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004129021836413217,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22517792582511903,
      "num_tokens": 59218583.0,
      "step": 26530
    },
    {
      "entropy": 5.540792942047119,
      "epoch": 2.980176335149099,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004128704574710334,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22394772469997407,
      "num_tokens": 59228638.0,
      "step": 26535
    },
    {
      "entropy": 5.6510340690612795,
      "epoch": 2.980737912057056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.1185,
      "mean_token_accuracy": 0.21636506021022797,
      "num_tokens": 59240574.0,
      "step": 26540
    },
    {
      "entropy": 5.588852739334106,
      "epoch": 2.9812994889650137,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041280699196192074,
      "loss": 4.944,
      "mean_token_accuracy": 0.222528837621212,
      "num_tokens": 59252986.0,
      "step": 26545
    },
    {
      "entropy": 5.597573518753052,
      "epoch": 2.9818610658729714,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.067,
      "mean_token_accuracy": 0.21743728518486022,
      "num_tokens": 59263034.0,
      "step": 26550
    },
    {
      "entropy": 5.620887708663941,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21177513301372528,
      "num_tokens": 59275170.0,
      "step": 26555
    },
    {
      "entropy": 5.622505950927734,
      "epoch": 2.9829842196888863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412711760792068,
      "loss": 5.0611,
      "mean_token_accuracy": 0.2190517947077751,
      "num_tokens": 59287168.0,
      "step": 26560
    },
    {
      "entropy": 5.64993052482605,
      "epoch": 2.983545796596844,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.0652,
      "mean_token_accuracy": 0.22665119022130967,
      "num_tokens": 59298043.0,
      "step": 26565
    },
    {
      "entropy": 5.544261360168457,
      "epoch": 2.9841073735048016,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041264825141983547,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22628054171800613,
      "num_tokens": 59310454.0,
      "step": 26570
    },
    {
      "entropy": 5.682544898986817,
      "epoch": 2.9846689504127593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21375297904014587,
      "num_tokens": 59323275.0,
      "step": 26575
    },
    {
      "entropy": 5.5441851139068605,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004125847245165084,
      "loss": 4.9703,
      "mean_token_accuracy": 0.22393755614757538,
      "num_tokens": 59334003.0,
      "step": 26580
    },
    {
      "entropy": 5.545044755935669,
      "epoch": 2.985792104228674,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004125529544932126,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22399765402078628,
      "num_tokens": 59345403.0,
      "step": 26585
    },
    {
      "entropy": 5.618722867965698,
      "epoch": 2.986353681136632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21585316509008406,
      "num_tokens": 59357457.0,
      "step": 26590
    },
    {
      "entropy": 5.576439332962036,
      "epoch": 2.986915258044589,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004124894013084145,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22701843678951264,
      "num_tokens": 59369059.0,
      "step": 26595
    },
    {
      "entropy": 5.589621829986572,
      "epoch": 2.9874768349525467,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041245761814893576,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22546342462301255,
      "num_tokens": 59380700.0,
      "step": 26600
    },
    {
      "entropy": 5.553256845474243,
      "epoch": 2.9880384118605043,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004124258306127531,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21469439417123795,
      "num_tokens": 59391504.0,
      "step": 26605
    },
    {
      "entropy": 5.636124992370606,
      "epoch": 2.988599988768462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.0534,
      "mean_token_accuracy": 0.2114922136068344,
      "num_tokens": 59402648.0,
      "step": 26610
    },
    {
      "entropy": 5.634245157241821,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004123622424143248,
      "loss": 5.0817,
      "mean_token_accuracy": 0.2146984562277794,
      "num_tokens": 59413842.0,
      "step": 26615
    },
    {
      "entropy": 5.643642997741699,
      "epoch": 2.989723142584377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041233044175410387,
      "loss": 5.1028,
      "mean_token_accuracy": 0.21884154081344603,
      "num_tokens": 59425630.0,
      "step": 26620
    },
    {
      "entropy": 5.582832956314087,
      "epoch": 2.9902847194923345,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041229863672122844,
      "loss": 4.9828,
      "mean_token_accuracy": 0.2234177440404892,
      "num_tokens": 59437140.0,
      "step": 26625
    },
    {
      "entropy": 5.559973049163818,
      "epoch": 2.990846296400292,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004122668273167111,
      "loss": 4.8951,
      "mean_token_accuracy": 0.2259802460670471,
      "num_tokens": 59447629.0,
      "step": 26630
    },
    {
      "entropy": 5.540464687347412,
      "epoch": 2.9914078733082494,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004122350135415648,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21124925911426545,
      "num_tokens": 59458469.0,
      "step": 26635
    },
    {
      "entropy": 5.5908775806427,
      "epoch": 2.991969450216207,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004122031953968024,
      "loss": 4.9844,
      "mean_token_accuracy": 0.216435444355011,
      "num_tokens": 59468864.0,
      "step": 26640
    },
    {
      "entropy": 5.557978200912475,
      "epoch": 2.9925310271241647,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041217137288343704,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2210207238793373,
      "num_tokens": 59480689.0,
      "step": 26645
    },
    {
      "entropy": 5.652043581008911,
      "epoch": 2.993092604032122,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000412139546002482,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21523143947124482,
      "num_tokens": 59491510.0,
      "step": 26650
    },
    {
      "entropy": 5.632281303405762,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.034,
      "mean_token_accuracy": 0.21296322345733643,
      "num_tokens": 59502838.0,
      "step": 26655
    },
    {
      "entropy": 5.654413843154908,
      "epoch": 2.9942157578480373,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004120758791418564,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22164530754089357,
      "num_tokens": 59514356.0,
      "step": 26660
    },
    {
      "entropy": 5.564588928222657,
      "epoch": 2.994777334755995,
      "grad_norm": 1.0,
      "learning_rate": 0.0004120440391642131,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21485511660575868,
      "num_tokens": 59525933.0,
      "step": 26665
    },
    {
      "entropy": 5.654409265518188,
      "epoch": 2.9953389116639526,
      "grad_norm": 1.125,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.0918,
      "mean_token_accuracy": 0.21768117547035218,
      "num_tokens": 59536856.0,
      "step": 26670
    },
    {
      "entropy": 5.6187457084655765,
      "epoch": 2.99590048857191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22241251766681672,
      "num_tokens": 59546023.0,
      "step": 26675
    },
    {
      "entropy": 5.594716501235962,
      "epoch": 2.9964620654798675,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004119484930541267,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22546919137239457,
      "num_tokens": 59557515.0,
      "step": 26680
    },
    {
      "entropy": 5.567170429229736,
      "epoch": 2.997023642387825,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21754275411367416,
      "num_tokens": 59570198.0,
      "step": 26685
    },
    {
      "entropy": 5.62050838470459,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21333523988723754,
      "num_tokens": 59580365.0,
      "step": 26690
    },
    {
      "entropy": 5.543890523910522,
      "epoch": 2.99814679620374,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004118529076996029,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22816642224788666,
      "num_tokens": 59591304.0,
      "step": 26695
    },
    {
      "entropy": 5.6284414768219,
      "epoch": 2.9987083731116977,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.0859,
      "mean_token_accuracy": 0.2127271920442581,
      "num_tokens": 59603005.0,
      "step": 26700
    },
    {
      "entropy": 5.561646175384522,
      "epoch": 2.9992699500196554,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041178916234098134,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2305796518921852,
      "num_tokens": 59614211.0,
      "step": 26705
    },
    {
      "entropy": 5.5651898860931395,
      "epoch": 2.999831526927613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004117572831280332,
      "loss": 4.995,
      "mean_token_accuracy": 0.21677072644233703,
      "num_tokens": 59624162.0,
      "step": 26710
    },
    {
      "entropy": 5.581919193267822,
      "epoch": 3.0003369461447744,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041172539956068,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21130081514517465,
      "num_tokens": 59633165.0,
      "step": 26715
    },
    {
      "entropy": 5.677504014968872,
      "epoch": 3.000898523052732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041169351163993705,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21618219912052156,
      "num_tokens": 59645130.0,
      "step": 26720
    },
    {
      "entropy": 5.540994548797608,
      "epoch": 3.0014600999606897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23702286928892136,
      "num_tokens": 59656826.0,
      "step": 26725
    },
    {
      "entropy": 5.605648326873779,
      "epoch": 3.002021676868647,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004116297227423435,
      "loss": 5.0617,
      "mean_token_accuracy": 0.22284125834703444,
      "num_tokens": 59667743.0,
      "step": 26730
    },
    {
      "entropy": 5.488157129287719,
      "epoch": 3.0025832537766046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23550435453653334,
      "num_tokens": 59679750.0,
      "step": 26735
    },
    {
      "entropy": 5.514609241485596,
      "epoch": 3.0031448306845623,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004115659164433765,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22490912228822707,
      "num_tokens": 59689113.0,
      "step": 26740
    },
    {
      "entropy": 5.540727758407593,
      "epoch": 3.00370640759252,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041153400677091745,
      "loss": 4.884,
      "mean_token_accuracy": 0.22390312254428862,
      "num_tokens": 59701282.0,
      "step": 26745
    },
    {
      "entropy": 5.5822405338287355,
      "epoch": 3.004267984500477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041150209275116266,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21711159497499466,
      "num_tokens": 59713536.0,
      "step": 26750
    },
    {
      "entropy": 5.66988205909729,
      "epoch": 3.004829561408435,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041147017438512834,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21340368837118148,
      "num_tokens": 59723818.0,
      "step": 26755
    },
    {
      "entropy": 5.655794239044189,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041143825167383073,
      "loss": 5.0446,
      "mean_token_accuracy": 0.2227813184261322,
      "num_tokens": 59735486.0,
      "step": 26760
    },
    {
      "entropy": 5.61421389579773,
      "epoch": 3.00595271522435,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004114063246182863,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22424156367778778,
      "num_tokens": 59746006.0,
      "step": 26765
    },
    {
      "entropy": 5.638077735900879,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.125,
      "learning_rate": 0.00041137439321951147,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2232664093375206,
      "num_tokens": 59756346.0,
      "step": 26770
    },
    {
      "entropy": 5.547239589691162,
      "epoch": 3.007075869040265,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041134245747852314,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2325556233525276,
      "num_tokens": 59768038.0,
      "step": 26775
    },
    {
      "entropy": 5.602739048004151,
      "epoch": 3.0076374459482227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004113105173963382,
      "loss": 4.9967,
      "mean_token_accuracy": 0.22056771516799928,
      "num_tokens": 59779211.0,
      "step": 26780
    },
    {
      "entropy": 5.576734209060669,
      "epoch": 3.0081990228561803,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004112785729739733,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22647871673107148,
      "num_tokens": 59790777.0,
      "step": 26785
    },
    {
      "entropy": 5.6158522605896,
      "epoch": 3.0087605997641376,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041124662421244593,
      "loss": 4.963,
      "mean_token_accuracy": 0.21684933602809905,
      "num_tokens": 59802030.0,
      "step": 26790
    },
    {
      "entropy": 5.595113134384155,
      "epoch": 3.0093221766720952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004112146711127732,
      "loss": 5.011,
      "mean_token_accuracy": 0.21827859580516815,
      "num_tokens": 59814616.0,
      "step": 26795
    },
    {
      "entropy": 5.549525308609009,
      "epoch": 3.009883753580053,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004111827136759725,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22685077488422395,
      "num_tokens": 59825797.0,
      "step": 26800
    },
    {
      "entropy": 5.53820309638977,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004111507519030613,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22748759090900422,
      "num_tokens": 59836699.0,
      "step": 26805
    },
    {
      "entropy": 5.635485696792602,
      "epoch": 3.0110069073959678,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22331581860780716,
      "num_tokens": 59848575.0,
      "step": 26810
    },
    {
      "entropy": 5.568087959289551,
      "epoch": 3.0115684843039254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004110868153529786,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2327573776245117,
      "num_tokens": 59860471.0,
      "step": 26815
    },
    {
      "entropy": 5.58874740600586,
      "epoch": 3.012130061211883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004110548405778428,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22745164036750792,
      "num_tokens": 59871759.0,
      "step": 26820
    },
    {
      "entropy": 5.649412775039673,
      "epoch": 3.0126916381198403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004110228614706681,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21968938857316972,
      "num_tokens": 59884375.0,
      "step": 26825
    },
    {
      "entropy": 5.625506448745727,
      "epoch": 3.013253215027798,
      "grad_norm": 1.125,
      "learning_rate": 0.00041099087803247276,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21857048869132994,
      "num_tokens": 59895606.0,
      "step": 26830
    },
    {
      "entropy": 5.609880256652832,
      "epoch": 3.0138147919357556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004109588902642751,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22371029555797578,
      "num_tokens": 59908090.0,
      "step": 26835
    },
    {
      "entropy": 5.591064977645874,
      "epoch": 3.0143763688437133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004109268981670936,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22125555127859114,
      "num_tokens": 59919218.0,
      "step": 26840
    },
    {
      "entropy": 5.638739347457886,
      "epoch": 3.0149379457516705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000410894901741947,
      "loss": 5.0967,
      "mean_token_accuracy": 0.22532893270254134,
      "num_tokens": 59931195.0,
      "step": 26845
    },
    {
      "entropy": 5.601947689056397,
      "epoch": 3.015499522659628,
      "grad_norm": 1.03125,
      "learning_rate": 0.000410862900989854,
      "loss": 4.9942,
      "mean_token_accuracy": 0.227974596619606,
      "num_tokens": 59943653.0,
      "step": 26850
    },
    {
      "entropy": 5.585467100143433,
      "epoch": 3.016061099567586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004108308959118336,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22072963714599608,
      "num_tokens": 59954677.0,
      "step": 26855
    },
    {
      "entropy": 5.620510578155518,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004107988865089047,
      "loss": 5.0278,
      "mean_token_accuracy": 0.22064756751060485,
      "num_tokens": 59966135.0,
      "step": 26860
    },
    {
      "entropy": 5.5398838996887205,
      "epoch": 3.0171842533835007,
      "grad_norm": 1.0,
      "learning_rate": 0.00041076687278208664,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23357374519109725,
      "num_tokens": 59977828.0,
      "step": 26865
    },
    {
      "entropy": 5.515426540374756,
      "epoch": 3.0177458302914584,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22919440865516663,
      "num_tokens": 59989226.0,
      "step": 26870
    },
    {
      "entropy": 5.632263422012329,
      "epoch": 3.018307407199416,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041070283236086024,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22977058887481688,
      "num_tokens": 60000461.0,
      "step": 26875
    },
    {
      "entropy": 5.645603656768799,
      "epoch": 3.0188689841073737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21765302866697311,
      "num_tokens": 60012364.0,
      "step": 26880
    },
    {
      "entropy": 5.608248519897461,
      "epoch": 3.019430561015331,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041063877465631084,
      "loss": 4.917,
      "mean_token_accuracy": 0.225404055416584,
      "num_tokens": 60022913.0,
      "step": 26885
    },
    {
      "entropy": 5.582176256179809,
      "epoch": 3.0199921379232886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004106067393253394,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22898890525102616,
      "num_tokens": 60034407.0,
      "step": 26890
    },
    {
      "entropy": 5.551155376434326,
      "epoch": 3.0205537148312462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004105746996765968,
      "loss": 4.974,
      "mean_token_accuracy": 0.21646855622529984,
      "num_tokens": 60044063.0,
      "step": 26895
    },
    {
      "entropy": 5.593520069122315,
      "epoch": 3.021115291739204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004105426557111031,
      "loss": 4.907,
      "mean_token_accuracy": 0.22843308597803116,
      "num_tokens": 60054432.0,
      "step": 26900
    },
    {
      "entropy": 5.632973575592041,
      "epoch": 3.021676868647161,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004105106074298787,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22432471811771393,
      "num_tokens": 60065876.0,
      "step": 26905
    },
    {
      "entropy": 5.590543937683106,
      "epoch": 3.022238445555119,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000410478554833944,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2217106431722641,
      "num_tokens": 60076108.0,
      "step": 26910
    },
    {
      "entropy": 5.5102338790893555,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041044649792431963,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22549874782562257,
      "num_tokens": 60087386.0,
      "step": 26915
    },
    {
      "entropy": 5.571774101257324,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.9369,
      "mean_token_accuracy": 0.23116425722837447,
      "num_tokens": 60099017.0,
      "step": 26920
    },
    {
      "entropy": 5.5828940868377686,
      "epoch": 3.0239231762789913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.9085,
      "mean_token_accuracy": 0.23068351447582244,
      "num_tokens": 60110676.0,
      "step": 26925
    },
    {
      "entropy": 5.507435750961304,
      "epoch": 3.024484753186949,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23166873604059218,
      "num_tokens": 60120807.0,
      "step": 26930
    },
    {
      "entropy": 5.614236974716187,
      "epoch": 3.0250463300949066,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041031822716934086,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22589677274227143,
      "num_tokens": 60132174.0,
      "step": 26935
    },
    {
      "entropy": 5.64521541595459,
      "epoch": 3.025607907002864,
      "grad_norm": 1.125,
      "learning_rate": 0.00041028614870658106,
      "loss": 5.0726,
      "mean_token_accuracy": 0.21649714857339858,
      "num_tokens": 60144311.0,
      "step": 26940
    },
    {
      "entropy": 5.6139678955078125,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004102540659362579,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2219152107834816,
      "num_tokens": 60155368.0,
      "step": 26945
    },
    {
      "entropy": 5.612851142883301,
      "epoch": 3.026731060818779,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041022197885939285,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2213252827525139,
      "num_tokens": 60167633.0,
      "step": 26950
    },
    {
      "entropy": 5.527492475509644,
      "epoch": 3.027292637726737,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041018988747700745,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2290205627679825,
      "num_tokens": 60180596.0,
      "step": 26955
    },
    {
      "entropy": 5.671173906326294,
      "epoch": 3.027854214634694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004101577917901236,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21988179683685302,
      "num_tokens": 60191811.0,
      "step": 26960
    },
    {
      "entropy": 5.684167289733887,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041012569179976326,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22580773383378983,
      "num_tokens": 60202575.0,
      "step": 26965
    },
    {
      "entropy": 5.573834466934204,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004100935875069485,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22461958974599838,
      "num_tokens": 60212541.0,
      "step": 26970
    },
    {
      "entropy": 5.601000595092773,
      "epoch": 3.029538945358567,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004100614789127015,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2293855994939804,
      "num_tokens": 60223018.0,
      "step": 26975
    },
    {
      "entropy": 5.632523775100708,
      "epoch": 3.0301005222665243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004100293660180447,
      "loss": 5.0468,
      "mean_token_accuracy": 0.22110558897256852,
      "num_tokens": 60234104.0,
      "step": 26980
    },
    {
      "entropy": 5.568659973144531,
      "epoch": 3.030662099174482,
      "grad_norm": 1.125,
      "learning_rate": 0.00040999724882400046,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2155134782195091,
      "num_tokens": 60245119.0,
      "step": 26985
    },
    {
      "entropy": 5.592618608474732,
      "epoch": 3.0312236760824396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004099651273315915,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22773853987455367,
      "num_tokens": 60255984.0,
      "step": 26990
    },
    {
      "entropy": 5.632629871368408,
      "epoch": 3.0317852529903973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004099330015418405,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2321978762745857,
      "num_tokens": 60266764.0,
      "step": 26995
    },
    {
      "entropy": 5.559029150009155,
      "epoch": 3.0323468298983545,
      "grad_norm": 1.125,
      "learning_rate": 0.00040990087145577044,
      "loss": 4.925,
      "mean_token_accuracy": 0.22614272981882094,
      "num_tokens": 60278284.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.314620394973617,
      "eval_loss": 5.091062545776367,
      "eval_mean_token_accuracy": 0.22515073032347788,
      "eval_num_tokens": 60278284.0,
      "eval_runtime": 18.4709,
      "eval_samples_per_second": 1679.018,
      "eval_steps_per_second": 209.898,
      "step": 27000
    },
    {
      "entropy": 5.522239685058594,
      "epoch": 3.032908406806312,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22710039615631103,
      "num_tokens": 60287993.0,
      "step": 27005
    },
    {
      "entropy": 5.645898914337158,
      "epoch": 3.03346998371427,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004098365983987652,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21729963421821594,
      "num_tokens": 60300040.0,
      "step": 27010
    },
    {
      "entropy": 5.660606956481933,
      "epoch": 3.034031560622227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040980445542987653,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21964868903160095,
      "num_tokens": 60311008.0,
      "step": 27015
    },
    {
      "entropy": 5.584987688064575,
      "epoch": 3.0345931375301847,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21739915162324905,
      "num_tokens": 60321071.0,
      "step": 27020
    },
    {
      "entropy": 5.571645927429199,
      "epoch": 3.0351547144381423,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040974015661644433,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21911347657442093,
      "num_tokens": 60333075.0,
      "step": 27025
    },
    {
      "entropy": 5.636417484283447,
      "epoch": 3.0357162913461,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004097080007739481,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21459150016307832,
      "num_tokens": 60343393.0,
      "step": 27030
    },
    {
      "entropy": 5.614227819442749,
      "epoch": 3.036277868254057,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040967584064229695,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21016294956207277,
      "num_tokens": 60353981.0,
      "step": 27035
    },
    {
      "entropy": 5.615854406356812,
      "epoch": 3.036839445162015,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040964367622251473,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22936490029096604,
      "num_tokens": 60364544.0,
      "step": 27040
    },
    {
      "entropy": 5.663313150405884,
      "epoch": 3.0374010220699725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004096115075156257,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2266458436846733,
      "num_tokens": 60376031.0,
      "step": 27045
    },
    {
      "entropy": 5.5908252716064455,
      "epoch": 3.03796259897793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.0205,
      "mean_token_accuracy": 0.21652786582708358,
      "num_tokens": 60387320.0,
      "step": 27050
    },
    {
      "entropy": 5.642997598648071,
      "epoch": 3.0385241758858874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004095471572446242,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22459330260753632,
      "num_tokens": 60397345.0,
      "step": 27055
    },
    {
      "entropy": 5.586333990097046,
      "epoch": 3.039085752793845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004095149756825606,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2188069060444832,
      "num_tokens": 60408194.0,
      "step": 27060
    },
    {
      "entropy": 5.647498703002929,
      "epoch": 3.0396473297018027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004094827898374881,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22076041996479034,
      "num_tokens": 60418928.0,
      "step": 27065
    },
    {
      "entropy": 5.562618732452393,
      "epoch": 3.0402089066097604,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004094505997104314,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22796293944120408,
      "num_tokens": 60428819.0,
      "step": 27070
    },
    {
      "entropy": 5.5007768154144285,
      "epoch": 3.0407704835177176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22949544489383697,
      "num_tokens": 60440575.0,
      "step": 27075
    },
    {
      "entropy": 5.59874997138977,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040938620661446526,
      "loss": 4.9981,
      "mean_token_accuracy": 0.215423084795475,
      "num_tokens": 60452977.0,
      "step": 27080
    },
    {
      "entropy": 5.691001081466675,
      "epoch": 3.041893637333633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004093540036476062,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21874061077833176,
      "num_tokens": 60463552.0,
      "step": 27085
    },
    {
      "entropy": 5.6838513851165775,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040932179640286356,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21682770997285844,
      "num_tokens": 60474674.0,
      "step": 27090
    },
    {
      "entropy": 5.568595457077026,
      "epoch": 3.043016791149548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004092895848812628,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23225746899843216,
      "num_tokens": 60486140.0,
      "step": 27095
    },
    {
      "entropy": 5.554031324386597,
      "epoch": 3.0435783680575055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040925736908382953,
      "loss": 4.9358,
      "mean_token_accuracy": 0.23075240552425386,
      "num_tokens": 60497386.0,
      "step": 27100
    },
    {
      "entropy": 5.6708756446838375,
      "epoch": 3.044139944965463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004092251490115896,
      "loss": 5.0643,
      "mean_token_accuracy": 0.2149008885025978,
      "num_tokens": 60508535.0,
      "step": 27105
    },
    {
      "entropy": 5.603385019302368,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040919292466556895,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21762676239013673,
      "num_tokens": 60520791.0,
      "step": 27110
    },
    {
      "entropy": 5.547401809692383,
      "epoch": 3.045263098781378,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004091606960467934,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22379705905914307,
      "num_tokens": 60533676.0,
      "step": 27115
    },
    {
      "entropy": 5.636317539215088,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040912846315628934,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21850074231624603,
      "num_tokens": 60545216.0,
      "step": 27120
    },
    {
      "entropy": 5.527074480056763,
      "epoch": 3.0463862525972933,
      "grad_norm": 1.046875,
      "learning_rate": 0.000409096225995083,
      "loss": 4.9171,
      "mean_token_accuracy": 0.2286738157272339,
      "num_tokens": 60554730.0,
      "step": 27125
    },
    {
      "entropy": 5.452763557434082,
      "epoch": 3.0469478295052506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2261198341846466,
      "num_tokens": 60565494.0,
      "step": 27130
    },
    {
      "entropy": 5.557615661621094,
      "epoch": 3.047509406413208,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040903173886466934,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22458071112632752,
      "num_tokens": 60576249.0,
      "step": 27135
    },
    {
      "entropy": 5.554509115219116,
      "epoch": 3.048070983321166,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22498320043087006,
      "num_tokens": 60587253.0,
      "step": 27140
    },
    {
      "entropy": 5.634591722488404,
      "epoch": 3.0486325602291235,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040896723466376557,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21414890587329866,
      "num_tokens": 60598158.0,
      "step": 27145
    },
    {
      "entropy": 5.5887864112854,
      "epoch": 3.0491941371370808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23488861322402954,
      "num_tokens": 60608812.0,
      "step": 27150
    },
    {
      "entropy": 5.601798343658447,
      "epoch": 3.0497557140450384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004089027134005871,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2345137283205986,
      "num_tokens": 60619264.0,
      "step": 27155
    },
    {
      "entropy": 5.481906747817993,
      "epoch": 3.050317290952996,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004088704463732127,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22772693634033203,
      "num_tokens": 60629688.0,
      "step": 27160
    },
    {
      "entropy": 5.564380121231079,
      "epoch": 3.0508788678609537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004088381750833514,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22298312783241273,
      "num_tokens": 60640790.0,
      "step": 27165
    },
    {
      "entropy": 5.61034779548645,
      "epoch": 3.051440444768911,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004088058995320307,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22790122479200364,
      "num_tokens": 60651952.0,
      "step": 27170
    },
    {
      "entropy": 5.645634555816651,
      "epoch": 3.0520020216768686,
      "grad_norm": 1.0,
      "learning_rate": 0.0004087736197202784,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2265968456864357,
      "num_tokens": 60665388.0,
      "step": 27175
    },
    {
      "entropy": 5.521217393875122,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040874133564912195,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22851123958826064,
      "num_tokens": 60676407.0,
      "step": 27180
    },
    {
      "entropy": 5.559431505203247,
      "epoch": 3.053125175492784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040870904731958965,
      "loss": 4.913,
      "mean_token_accuracy": 0.22336605340242385,
      "num_tokens": 60686796.0,
      "step": 27185
    },
    {
      "entropy": 5.611384868621826,
      "epoch": 3.053686752400741,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004086767547327094,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22170102149248122,
      "num_tokens": 60697770.0,
      "step": 27190
    },
    {
      "entropy": 5.5846490383148195,
      "epoch": 3.054248329308699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040864445788950935,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22385431379079818,
      "num_tokens": 60709059.0,
      "step": 27195
    },
    {
      "entropy": 5.580405139923096,
      "epoch": 3.0548099062166565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000408612156791018,
      "loss": 4.932,
      "mean_token_accuracy": 0.22518311738967894,
      "num_tokens": 60720228.0,
      "step": 27200
    },
    {
      "entropy": 5.645976448059082,
      "epoch": 3.0553714831246137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004085798514382637,
      "loss": 5.0472,
      "mean_token_accuracy": 0.22292101979255677,
      "num_tokens": 60730658.0,
      "step": 27205
    },
    {
      "entropy": 5.612660026550293,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040854754183227516,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22710839062929153,
      "num_tokens": 60741757.0,
      "step": 27210
    },
    {
      "entropy": 5.594130611419677,
      "epoch": 3.056494636940529,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004085152279740811,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22434774339199065,
      "num_tokens": 60755452.0,
      "step": 27215
    },
    {
      "entropy": 5.6375102519989015,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040848290986471037,
      "loss": 4.974,
      "mean_token_accuracy": 0.2237017959356308,
      "num_tokens": 60765715.0,
      "step": 27220
    },
    {
      "entropy": 5.606237411499023,
      "epoch": 3.057617790756444,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.0374,
      "mean_token_accuracy": 0.22248474955558778,
      "num_tokens": 60777021.0,
      "step": 27225
    },
    {
      "entropy": 5.52393069267273,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040841826089655525,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22294956892728807,
      "num_tokens": 60787309.0,
      "step": 27230
    },
    {
      "entropy": 5.593198204040528,
      "epoch": 3.0587409445723592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040838593003982926,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22070877254009247,
      "num_tokens": 60797566.0,
      "step": 27235
    },
    {
      "entropy": 5.636385345458985,
      "epoch": 3.059302521480317,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004083535949360436,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22455444782972336,
      "num_tokens": 60808213.0,
      "step": 27240
    },
    {
      "entropy": 5.621904420852661,
      "epoch": 3.059864098388274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004083212555862277,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2185733735561371,
      "num_tokens": 60820656.0,
      "step": 27245
    },
    {
      "entropy": 5.609159374237061,
      "epoch": 3.0604256752962318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004082889119914113,
      "loss": 5.0287,
      "mean_token_accuracy": 0.22517671436071396,
      "num_tokens": 60831572.0,
      "step": 27250
    },
    {
      "entropy": 5.53594560623169,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004082565641526242,
      "loss": 4.8734,
      "mean_token_accuracy": 0.23071824461221696,
      "num_tokens": 60842443.0,
      "step": 27255
    },
    {
      "entropy": 5.567267608642578,
      "epoch": 3.061548829112147,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040822421207089654,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23157826960086822,
      "num_tokens": 60853427.0,
      "step": 27260
    },
    {
      "entropy": 5.638148784637451,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004081918557472582,
      "loss": 5.0397,
      "mean_token_accuracy": 0.22421229332685472,
      "num_tokens": 60864517.0,
      "step": 27265
    },
    {
      "entropy": 5.660494709014893,
      "epoch": 3.062671982928062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004081594951827396,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22424403131008147,
      "num_tokens": 60874280.0,
      "step": 27270
    },
    {
      "entropy": 5.609122037887573,
      "epoch": 3.0632335598360196,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040812713037837096,
      "loss": 4.9906,
      "mean_token_accuracy": 0.2230392649769783,
      "num_tokens": 60885656.0,
      "step": 27275
    },
    {
      "entropy": 5.637769269943237,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.125,
      "learning_rate": 0.00040809476133518284,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22749855667352675,
      "num_tokens": 60896856.0,
      "step": 27280
    },
    {
      "entropy": 5.505795574188232,
      "epoch": 3.0643567136519345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040806238805420595,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2292202442884445,
      "num_tokens": 60907464.0,
      "step": 27285
    },
    {
      "entropy": 5.612604570388794,
      "epoch": 3.064918290559892,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000408030010536471,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21987420022487641,
      "num_tokens": 60918321.0,
      "step": 27290
    },
    {
      "entropy": 5.66195387840271,
      "epoch": 3.06547986746785,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040799762878300895,
      "loss": 5.022,
      "mean_token_accuracy": 0.21263538897037507,
      "num_tokens": 60928823.0,
      "step": 27295
    },
    {
      "entropy": 5.700636672973633,
      "epoch": 3.066041444375807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040796524279485084,
      "loss": 5.0828,
      "mean_token_accuracy": 0.22104556262493133,
      "num_tokens": 60940161.0,
      "step": 27300
    },
    {
      "entropy": 5.582065629959106,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040793285257302776,
      "loss": 5.0367,
      "mean_token_accuracy": 0.21822647601366044,
      "num_tokens": 60951968.0,
      "step": 27305
    },
    {
      "entropy": 5.6701267719268795,
      "epoch": 3.0671645981917224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000407900458118571,
      "loss": 5.0316,
      "mean_token_accuracy": 0.214189313352108,
      "num_tokens": 60963619.0,
      "step": 27310
    },
    {
      "entropy": 5.638307809829712,
      "epoch": 3.06772617509968,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004078680594325123,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2268623411655426,
      "num_tokens": 60974615.0,
      "step": 27315
    },
    {
      "entropy": 5.5946063041687015,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.125,
      "learning_rate": 0.000407835656515883,
      "loss": 5.047,
      "mean_token_accuracy": 0.21622041314840318,
      "num_tokens": 60986943.0,
      "step": 27320
    },
    {
      "entropy": 5.602978706359863,
      "epoch": 3.068849328915595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040780324936971486,
      "loss": 5.0595,
      "mean_token_accuracy": 0.22224546670913697,
      "num_tokens": 60998306.0,
      "step": 27325
    },
    {
      "entropy": 5.668218040466309,
      "epoch": 3.0694109058235526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004077708379950397,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2159038484096527,
      "num_tokens": 61010102.0,
      "step": 27330
    },
    {
      "entropy": 5.628308200836182,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004077384223928896,
      "loss": 4.9437,
      "mean_token_accuracy": 0.2274816334247589,
      "num_tokens": 61019474.0,
      "step": 27335
    },
    {
      "entropy": 5.55144772529602,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004077060025642967,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22961228489875793,
      "num_tokens": 61030401.0,
      "step": 27340
    },
    {
      "entropy": 5.682353258132935,
      "epoch": 3.071095636547425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.04,
      "mean_token_accuracy": 0.2210219234228134,
      "num_tokens": 61042139.0,
      "step": 27345
    },
    {
      "entropy": 5.558083868026733,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004076411502319114,
      "loss": 4.8998,
      "mean_token_accuracy": 0.2245197832584381,
      "num_tokens": 61052468.0,
      "step": 27350
    },
    {
      "entropy": 5.61196985244751,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040760871773018405,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22829356044530869,
      "num_tokens": 61063625.0,
      "step": 27355
    },
    {
      "entropy": 5.509925413131714,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004075762810061437,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2278460219502449,
      "num_tokens": 61074246.0,
      "step": 27360
    },
    {
      "entropy": 5.635882377624512,
      "epoch": 3.0733419441792553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000407543840060823,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21541392356157302,
      "num_tokens": 61086096.0,
      "step": 27365
    },
    {
      "entropy": 5.625319766998291,
      "epoch": 3.073903521087213,
      "grad_norm": 1.125,
      "learning_rate": 0.0004075113948952551,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22497875690460206,
      "num_tokens": 61097176.0,
      "step": 27370
    },
    {
      "entropy": 5.70286226272583,
      "epoch": 3.0744650979951706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004074789455104731,
      "loss": 5.0947,
      "mean_token_accuracy": 0.22301459759473802,
      "num_tokens": 61107865.0,
      "step": 27375
    },
    {
      "entropy": 5.640517950057983,
      "epoch": 3.075026674903128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040744649190750995,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22442771196365358,
      "num_tokens": 61118682.0,
      "step": 27380
    },
    {
      "entropy": 5.590420389175415,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.125,
      "learning_rate": 0.0004074140340873992,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2356923297047615,
      "num_tokens": 61130210.0,
      "step": 27385
    },
    {
      "entropy": 5.521203851699829,
      "epoch": 3.076149828719043,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004073815720511743,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21896070390939712,
      "num_tokens": 61141216.0,
      "step": 27390
    },
    {
      "entropy": 5.584208631515503,
      "epoch": 3.0767114056270004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040734910579986875,
      "loss": 5.0321,
      "mean_token_accuracy": 0.22358753979206086,
      "num_tokens": 61153233.0,
      "step": 27395
    },
    {
      "entropy": 5.607898950576782,
      "epoch": 3.077272982534958,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040731663533451624,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2276044249534607,
      "num_tokens": 61163044.0,
      "step": 27400
    },
    {
      "entropy": 5.544701814651489,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040728416065615087,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22299281656742095,
      "num_tokens": 61174595.0,
      "step": 27405
    },
    {
      "entropy": 5.523153638839721,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.949,
      "mean_token_accuracy": 0.2277911886572838,
      "num_tokens": 61185215.0,
      "step": 27410
    },
    {
      "entropy": 5.522256708145141,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040721919866451724,
      "loss": 4.8175,
      "mean_token_accuracy": 0.23813325613737107,
      "num_tokens": 61195504.0,
      "step": 27415
    },
    {
      "entropy": 5.4726499080657955,
      "epoch": 3.0795192901667883,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22731742560863494,
      "num_tokens": 61206040.0,
      "step": 27420
    },
    {
      "entropy": 5.6050714492797855,
      "epoch": 3.080080867074746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21884002387523652,
      "num_tokens": 61217797.0,
      "step": 27425
    },
    {
      "entropy": 5.684288740158081,
      "epoch": 3.0806424439827036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040712172410532397,
      "loss": 5.0217,
      "mean_token_accuracy": 0.22267551422119142,
      "num_tokens": 61230158.0,
      "step": 27430
    },
    {
      "entropy": 5.670296335220337,
      "epoch": 3.081204020890661,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004070892241705994,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22718664407730102,
      "num_tokens": 61241072.0,
      "step": 27435
    },
    {
      "entropy": 5.47069616317749,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23025847375392913,
      "num_tokens": 61251057.0,
      "step": 27440
    },
    {
      "entropy": 5.483143043518067,
      "epoch": 3.082327174706576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000407024211684869,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23259462416172028,
      "num_tokens": 61262729.0,
      "step": 27445
    },
    {
      "entropy": 5.564653635025024,
      "epoch": 3.082888751614534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040699169913593306,
      "loss": 4.938,
      "mean_token_accuracy": 0.23576701581478118,
      "num_tokens": 61274665.0,
      "step": 27450
    },
    {
      "entropy": 5.637094211578369,
      "epoch": 3.083450328522491,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004069591823843302,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22828344702720643,
      "num_tokens": 61285244.0,
      "step": 27455
    },
    {
      "entropy": 5.586458158493042,
      "epoch": 3.0840119054304487,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004069266614310959,
      "loss": 4.9782,
      "mean_token_accuracy": 0.2203328415751457,
      "num_tokens": 61295739.0,
      "step": 27460
    },
    {
      "entropy": 5.585461044311524,
      "epoch": 3.0845734823384063,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040689413627726544,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22495850622653962,
      "num_tokens": 61305964.0,
      "step": 27465
    },
    {
      "entropy": 5.678827381134033,
      "epoch": 3.085135059246364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2193343698978424,
      "num_tokens": 61317004.0,
      "step": 27470
    },
    {
      "entropy": 5.660269403457642,
      "epoch": 3.085696636154321,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040682907337195894,
      "loss": 5.0921,
      "mean_token_accuracy": 0.20821431279182434,
      "num_tokens": 61329014.0,
      "step": 27475
    },
    {
      "entropy": 5.558941316604614,
      "epoch": 3.086258213062279,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004067965356225545,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2207883045077324,
      "num_tokens": 61340522.0,
      "step": 27480
    },
    {
      "entropy": 5.6490129947662355,
      "epoch": 3.0868197899702365,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040676399367669733,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2249574437737465,
      "num_tokens": 61351595.0,
      "step": 27485
    },
    {
      "entropy": 5.647689723968506,
      "epoch": 3.0873813668781938,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004067314475354234,
      "loss": 5.0312,
      "mean_token_accuracy": 0.22579217702150345,
      "num_tokens": 61362432.0,
      "step": 27490
    },
    {
      "entropy": 5.565252208709717,
      "epoch": 3.0879429437861514,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040669889719976916,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22341400384902954,
      "num_tokens": 61372717.0,
      "step": 27495
    },
    {
      "entropy": 5.629382753372193,
      "epoch": 3.088504520694109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22222141474485396,
      "num_tokens": 61383150.0,
      "step": 27500
    },
    {
      "entropy": 5.624037647247315,
      "epoch": 3.0890660976020667,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040663378394946534,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20934944599866867,
      "num_tokens": 61395538.0,
      "step": 27505
    },
    {
      "entropy": 5.612083911895752,
      "epoch": 3.089627674510024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004066012210368889,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22463186085224152,
      "num_tokens": 61407523.0,
      "step": 27510
    },
    {
      "entropy": 5.558055639266968,
      "epoch": 3.0901892514179816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004065686539340786,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21954285949468613,
      "num_tokens": 61418680.0,
      "step": 27515
    },
    {
      "entropy": 5.582892227172851,
      "epoch": 3.0907508283259393,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040653608264207134,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2250683069229126,
      "num_tokens": 61430006.0,
      "step": 27520
    },
    {
      "entropy": 5.482368803024292,
      "epoch": 3.091312405233897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040650350716190417,
      "loss": 4.7982,
      "mean_token_accuracy": 0.23845161646604537,
      "num_tokens": 61440564.0,
      "step": 27525
    },
    {
      "entropy": 5.481808519363403,
      "epoch": 3.091873982141854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004064709274946143,
      "loss": 4.7867,
      "mean_token_accuracy": 0.22938708066940308,
      "num_tokens": 61451589.0,
      "step": 27530
    },
    {
      "entropy": 5.521551132202148,
      "epoch": 3.092435559049812,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004064383436412392,
      "loss": 4.8357,
      "mean_token_accuracy": 0.2326434627175331,
      "num_tokens": 61462177.0,
      "step": 27535
    },
    {
      "entropy": 5.626884508132934,
      "epoch": 3.0929971359577695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004064057556028162,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22440531104803085,
      "num_tokens": 61472626.0,
      "step": 27540
    },
    {
      "entropy": 5.536695146560669,
      "epoch": 3.093558712865727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000406373163380383,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2309451714158058,
      "num_tokens": 61483839.0,
      "step": 27545
    },
    {
      "entropy": 5.554372310638428,
      "epoch": 3.0941202897736844,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040634056697497726,
      "loss": 4.9659,
      "mean_token_accuracy": 0.233952397108078,
      "num_tokens": 61495099.0,
      "step": 27550
    },
    {
      "entropy": 5.581116151809693,
      "epoch": 3.094681866681642,
      "grad_norm": 1.0625,
      "learning_rate": 0.000406307966387637,
      "loss": 4.9679,
      "mean_token_accuracy": 0.22596677988767624,
      "num_tokens": 61506902.0,
      "step": 27555
    },
    {
      "entropy": 5.6106428623199465,
      "epoch": 3.0952434435895997,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040627536161940016,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2193318113684654,
      "num_tokens": 61519357.0,
      "step": 27560
    },
    {
      "entropy": 5.610545825958252,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004062427526713049,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23586168587207795,
      "num_tokens": 61529510.0,
      "step": 27565
    },
    {
      "entropy": 5.5096691131591795,
      "epoch": 3.0963665974055146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22895343154668807,
      "num_tokens": 61539679.0,
      "step": 27570
    },
    {
      "entropy": 5.5362104892730715,
      "epoch": 3.0969281743134722,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004061775222396922,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2313595712184906,
      "num_tokens": 61549488.0,
      "step": 27575
    },
    {
      "entropy": 5.532021474838257,
      "epoch": 3.09748975122143,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004061449007582519,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22595098316669465,
      "num_tokens": 61561167.0,
      "step": 27580
    },
    {
      "entropy": 5.592445802688599,
      "epoch": 3.098051328129387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000406112275101107,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22816782295703888,
      "num_tokens": 61572730.0,
      "step": 27585
    },
    {
      "entropy": 5.548894309997559,
      "epoch": 3.0986129050373448,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040607964526929655,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22972952723503112,
      "num_tokens": 61583846.0,
      "step": 27590
    },
    {
      "entropy": 5.580250072479248,
      "epoch": 3.0991744819453024,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040604701126385924,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2225916177034378,
      "num_tokens": 61594360.0,
      "step": 27595
    },
    {
      "entropy": 5.6082371234893795,
      "epoch": 3.09973605885326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004060143730858342,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2269076868891716,
      "num_tokens": 61606750.0,
      "step": 27600
    },
    {
      "entropy": 5.593834447860718,
      "epoch": 3.1002976357612173,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040598173073626086,
      "loss": 4.9973,
      "mean_token_accuracy": 0.21092354506254196,
      "num_tokens": 61618238.0,
      "step": 27605
    },
    {
      "entropy": 5.603138208389282,
      "epoch": 3.100859212669175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004059490842161783,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21837475746870041,
      "num_tokens": 61629568.0,
      "step": 27610
    },
    {
      "entropy": 5.648110008239746,
      "epoch": 3.1014207895771326,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040591643352662615,
      "loss": 5.0377,
      "mean_token_accuracy": 0.22413529455661774,
      "num_tokens": 61640103.0,
      "step": 27615
    },
    {
      "entropy": 5.5683952331542965,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000405883778668644,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23357743322849273,
      "num_tokens": 61651392.0,
      "step": 27620
    },
    {
      "entropy": 5.621907091140747,
      "epoch": 3.1025439433930475,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040585111964327157,
      "loss": 5.0728,
      "mean_token_accuracy": 0.21314149349927902,
      "num_tokens": 61664932.0,
      "step": 27625
    },
    {
      "entropy": 5.557618808746338,
      "epoch": 3.103105520301005,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22493624538183213,
      "num_tokens": 61675719.0,
      "step": 27630
    },
    {
      "entropy": 5.649971151351929,
      "epoch": 3.103667097208963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004057857890945155,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21867202818393708,
      "num_tokens": 61687298.0,
      "step": 27635
    },
    {
      "entropy": 5.667820453643799,
      "epoch": 3.1042286741169205,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040575311757321203,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21561678051948546,
      "num_tokens": 61697943.0,
      "step": 27640
    },
    {
      "entropy": 5.603138971328735,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040572044188867863,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22025690078735352,
      "num_tokens": 61708897.0,
      "step": 27645
    },
    {
      "entropy": 5.561166620254516,
      "epoch": 3.1053518279328354,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.0734,
      "mean_token_accuracy": 0.22548136115074158,
      "num_tokens": 61720409.0,
      "step": 27650
    },
    {
      "entropy": 5.662033271789551,
      "epoch": 3.105913404840793,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004056550780340835,
      "loss": 4.9872,
      "mean_token_accuracy": 0.222395221889019,
      "num_tokens": 61731378.0,
      "step": 27655
    },
    {
      "entropy": 5.617042779922485,
      "epoch": 3.1064749817487507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004056223898661031,
      "loss": 4.9014,
      "mean_token_accuracy": 0.21743318438529968,
      "num_tokens": 61742713.0,
      "step": 27660
    },
    {
      "entropy": 5.634508514404297,
      "epoch": 3.107036558656708,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040558969753905514,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22231070548295975,
      "num_tokens": 61753561.0,
      "step": 27665
    },
    {
      "entropy": 5.634906768798828,
      "epoch": 3.1075981355646656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004055570010539806,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21993272751569748,
      "num_tokens": 61764966.0,
      "step": 27670
    },
    {
      "entropy": 5.611012125015259,
      "epoch": 3.1081597124726232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004055243004119204,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21620290577411652,
      "num_tokens": 61776988.0,
      "step": 27675
    },
    {
      "entropy": 5.582904195785522,
      "epoch": 3.108721289380581,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040549159561391595,
      "loss": 5.0119,
      "mean_token_accuracy": 0.21883946806192398,
      "num_tokens": 61788327.0,
      "step": 27680
    },
    {
      "entropy": 5.530537891387939,
      "epoch": 3.109282866288538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004054588866610085,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23060596734285355,
      "num_tokens": 61799872.0,
      "step": 27685
    },
    {
      "entropy": 5.556320333480835,
      "epoch": 3.1098444431964958,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040542617355423947,
      "loss": 4.9467,
      "mean_token_accuracy": 0.223028963804245,
      "num_tokens": 61810736.0,
      "step": 27690
    },
    {
      "entropy": 5.636730194091797,
      "epoch": 3.1104060201044534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040539345629465043,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21846839934587478,
      "num_tokens": 61821236.0,
      "step": 27695
    },
    {
      "entropy": 5.695653867721558,
      "epoch": 3.1109675970124107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040536073488328325,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22218354493379594,
      "num_tokens": 61832282.0,
      "step": 27700
    },
    {
      "entropy": 5.723922395706177,
      "epoch": 3.1115291739203683,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.0519,
      "mean_token_accuracy": 0.21767011880874634,
      "num_tokens": 61842376.0,
      "step": 27705
    },
    {
      "entropy": 5.602731418609619,
      "epoch": 3.112090750828326,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004052952796093817,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22689368724822997,
      "num_tokens": 61852227.0,
      "step": 27710
    },
    {
      "entropy": 5.6472328186035154,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040526254574893154,
      "loss": 5.0192,
      "mean_token_accuracy": 0.22654732912778855,
      "num_tokens": 61863119.0,
      "step": 27715
    },
    {
      "entropy": 5.562690925598145,
      "epoch": 3.113213904644241,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040522980774087143,
      "loss": 4.919,
      "mean_token_accuracy": 0.22735822349786758,
      "num_tokens": 61874036.0,
      "step": 27720
    },
    {
      "entropy": 5.469445371627808,
      "epoch": 3.1137754815521985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004051970655862437,
      "loss": 4.8617,
      "mean_token_accuracy": 0.228839772939682,
      "num_tokens": 61884744.0,
      "step": 27725
    },
    {
      "entropy": 5.5563334941864015,
      "epoch": 3.114337058460156,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004051643192860908,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21851946115493776,
      "num_tokens": 61896866.0,
      "step": 27730
    },
    {
      "entropy": 5.59778242111206,
      "epoch": 3.114898635368114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040513156884145564,
      "loss": 4.8751,
      "mean_token_accuracy": 0.23127712309360504,
      "num_tokens": 61907384.0,
      "step": 27735
    },
    {
      "entropy": 5.509355688095093,
      "epoch": 3.115460212276071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004050988142533808,
      "loss": 4.7934,
      "mean_token_accuracy": 0.2303095802664757,
      "num_tokens": 61917719.0,
      "step": 27740
    },
    {
      "entropy": 5.557158279418945,
      "epoch": 3.1160217891840287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004050660555229093,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22337172478437423,
      "num_tokens": 61928159.0,
      "step": 27745
    },
    {
      "entropy": 5.553275156021118,
      "epoch": 3.1165833660919864,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004050332926510841,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22265890389680862,
      "num_tokens": 61938748.0,
      "step": 27750
    },
    {
      "entropy": 5.498833131790161,
      "epoch": 3.117144942999944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.8418,
      "mean_token_accuracy": 0.23673962354660033,
      "num_tokens": 61950429.0,
      "step": 27755
    },
    {
      "entropy": 5.570922183990478,
      "epoch": 3.1177065199079013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22349240630865097,
      "num_tokens": 61961156.0,
      "step": 27760
    },
    {
      "entropy": 5.5416100978851315,
      "epoch": 3.118268096815859,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040493497919791906,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22492013722658158,
      "num_tokens": 61972673.0,
      "step": 27765
    },
    {
      "entropy": 5.564424610137939,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040490219977111237,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2280188113451004,
      "num_tokens": 61984481.0,
      "step": 27770
    },
    {
      "entropy": 5.590832138061524,
      "epoch": 3.1193912506317742,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004048694162081693,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2260962978005409,
      "num_tokens": 61994919.0,
      "step": 27775
    },
    {
      "entropy": 5.575479364395141,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004048366285101336,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22449266910552979,
      "num_tokens": 62005303.0,
      "step": 27780
    },
    {
      "entropy": 5.515126037597656,
      "epoch": 3.120514404447689,
      "grad_norm": 1.125,
      "learning_rate": 0.0004048038366780494,
      "loss": 4.8055,
      "mean_token_accuracy": 0.230194653570652,
      "num_tokens": 62015104.0,
      "step": 27785
    },
    {
      "entropy": 5.530291414260864,
      "epoch": 3.121075981355647,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004047710407129606,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2186094790697098,
      "num_tokens": 62025577.0,
      "step": 27790
    },
    {
      "entropy": 5.711292362213134,
      "epoch": 3.121637558263604,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.099,
      "mean_token_accuracy": 0.21214851140975952,
      "num_tokens": 62037001.0,
      "step": 27795
    },
    {
      "entropy": 5.6339515209197994,
      "epoch": 3.1221991351715617,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004047054363879466,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21972793191671372,
      "num_tokens": 62049520.0,
      "step": 27800
    },
    {
      "entropy": 5.578485774993896,
      "epoch": 3.1227607120795193,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040467262803011025,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22616419047117234,
      "num_tokens": 62061650.0,
      "step": 27805
    },
    {
      "entropy": 5.60917706489563,
      "epoch": 3.123322288987477,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.053,
      "mean_token_accuracy": 0.2203711062669754,
      "num_tokens": 62074433.0,
      "step": 27810
    },
    {
      "entropy": 5.627907752990723,
      "epoch": 3.123883865895434,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040460699892900197,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22549874782562257,
      "num_tokens": 62084885.0,
      "step": 27815
    },
    {
      "entropy": 5.602635526657105,
      "epoch": 3.124445442803392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004045741781878198,
      "loss": 4.989,
      "mean_token_accuracy": 0.22144778072834015,
      "num_tokens": 62096847.0,
      "step": 27820
    },
    {
      "entropy": 5.613147306442261,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040454135332094543,
      "loss": 5.041,
      "mean_token_accuracy": 0.22053713500499725,
      "num_tokens": 62108131.0,
      "step": 27825
    },
    {
      "entropy": 5.588590288162232,
      "epoch": 3.125568596619307,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004045085243294242,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22828343361616135,
      "num_tokens": 62118450.0,
      "step": 27830
    },
    {
      "entropy": 5.585448265075684,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004044756912143013,
      "loss": 5.021,
      "mean_token_accuracy": 0.2196891039609909,
      "num_tokens": 62130702.0,
      "step": 27835
    },
    {
      "entropy": 5.6151337146759035,
      "epoch": 3.126691750435222,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040444285397662225,
      "loss": 5.0328,
      "mean_token_accuracy": 0.2198004260659218,
      "num_tokens": 62142863.0,
      "step": 27840
    },
    {
      "entropy": 5.578385639190674,
      "epoch": 3.1272533273431797,
      "grad_norm": 1.0,
      "learning_rate": 0.0004044100126174324,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21969542801380157,
      "num_tokens": 62153787.0,
      "step": 27845
    },
    {
      "entropy": 5.650721597671509,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040437716713777756,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2192936822772026,
      "num_tokens": 62166862.0,
      "step": 27850
    },
    {
      "entropy": 5.5795704364776615,
      "epoch": 3.1283764811590946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004043443175387036,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22483613938093186,
      "num_tokens": 62177431.0,
      "step": 27855
    },
    {
      "entropy": 5.539773607254029,
      "epoch": 3.1289380580670523,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040431146382125644,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2227855145931244,
      "num_tokens": 62189465.0,
      "step": 27860
    },
    {
      "entropy": 5.5759721279144285,
      "epoch": 3.12949963497501,
      "grad_norm": 1.15625,
      "learning_rate": 0.000404278605986482,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23152057826519012,
      "num_tokens": 62201327.0,
      "step": 27865
    },
    {
      "entropy": 5.579485511779785,
      "epoch": 3.1300612118829676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004042457440354267,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2209005042910576,
      "num_tokens": 62212689.0,
      "step": 27870
    },
    {
      "entropy": 5.532590007781982,
      "epoch": 3.130622788790925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004042128779691368,
      "loss": 4.9607,
      "mean_token_accuracy": 0.22387367784976958,
      "num_tokens": 62223771.0,
      "step": 27875
    },
    {
      "entropy": 5.63000111579895,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004041800077886587,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22577962130308152,
      "num_tokens": 62234137.0,
      "step": 27880
    },
    {
      "entropy": 5.609326028823853,
      "epoch": 3.13174594260684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040414713349503907,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2242940455675125,
      "num_tokens": 62244978.0,
      "step": 27885
    },
    {
      "entropy": 5.6305506229400635,
      "epoch": 3.1323075195147974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040411425508932457,
      "loss": 5.031,
      "mean_token_accuracy": 0.21738668233156205,
      "num_tokens": 62257364.0,
      "step": 27890
    },
    {
      "entropy": 5.588703870773315,
      "epoch": 3.132869096422755,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040408137257256225,
      "loss": 4.8316,
      "mean_token_accuracy": 0.22864541709423064,
      "num_tokens": 62268135.0,
      "step": 27895
    },
    {
      "entropy": 5.588724422454834,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004040484859457988,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22337744981050492,
      "num_tokens": 62279096.0,
      "step": 27900
    },
    {
      "entropy": 5.55109543800354,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040401559521008155,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22065655291080474,
      "num_tokens": 62289640.0,
      "step": 27905
    },
    {
      "entropy": 5.648727941513061,
      "epoch": 3.1345538271466276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040398270036645775,
      "loss": 5.0783,
      "mean_token_accuracy": 0.21489726454019548,
      "num_tokens": 62301085.0,
      "step": 27910
    },
    {
      "entropy": 5.50249137878418,
      "epoch": 3.135115404054585,
      "grad_norm": 1.125,
      "learning_rate": 0.0004039498014159748,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23171485513448714,
      "num_tokens": 62312667.0,
      "step": 27915
    },
    {
      "entropy": 5.621457624435425,
      "epoch": 3.135676980962543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040391689835968004,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22398776412010193,
      "num_tokens": 62324640.0,
      "step": 27920
    },
    {
      "entropy": 5.626853513717651,
      "epoch": 3.1362385578705005,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004038839911986213,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22224689573049544,
      "num_tokens": 62335940.0,
      "step": 27925
    },
    {
      "entropy": 5.654066276550293,
      "epoch": 3.1368001347784578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004038510799338462,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2257298484444618,
      "num_tokens": 62346745.0,
      "step": 27930
    },
    {
      "entropy": 5.641764116287232,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004038181645664029,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21686127185821533,
      "num_tokens": 62357694.0,
      "step": 27935
    },
    {
      "entropy": 5.6584662914276125,
      "epoch": 3.137923288594373,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.1101,
      "mean_token_accuracy": 0.21463717222213746,
      "num_tokens": 62370621.0,
      "step": 27940
    },
    {
      "entropy": 5.606350231170654,
      "epoch": 3.1384848655023307,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004037523215277033,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22476252168416977,
      "num_tokens": 62381248.0,
      "step": 27945
    },
    {
      "entropy": 5.68328857421875,
      "epoch": 3.139046442410288,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004037193938585436,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22111238837242125,
      "num_tokens": 62392102.0,
      "step": 27950
    },
    {
      "entropy": 5.652276802062988,
      "epoch": 3.1396080193182456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004036864620909085,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22874764651060103,
      "num_tokens": 62403254.0,
      "step": 27955
    },
    {
      "entropy": 5.56681718826294,
      "epoch": 3.1401695962262033,
      "grad_norm": 1.25,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23986366391181946,
      "num_tokens": 62413686.0,
      "step": 27960
    },
    {
      "entropy": 5.584179925918579,
      "epoch": 3.140731173134161,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040362058626440637,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2268734559416771,
      "num_tokens": 62424119.0,
      "step": 27965
    },
    {
      "entropy": 5.5717535495758055,
      "epoch": 3.141292750042118,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040358764220763685,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2278338700532913,
      "num_tokens": 62434900.0,
      "step": 27970
    },
    {
      "entropy": 5.594063377380371,
      "epoch": 3.141854326950076,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040355469405658687,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23251951932907106,
      "num_tokens": 62446766.0,
      "step": 27975
    },
    {
      "entropy": 5.581806325912476,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.125,
      "learning_rate": 0.0004035217418123056,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2214086875319481,
      "num_tokens": 62457863.0,
      "step": 27980
    },
    {
      "entropy": 5.6539843559265135,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040348878547584227,
      "loss": 5.0582,
      "mean_token_accuracy": 0.2168028473854065,
      "num_tokens": 62470380.0,
      "step": 27985
    },
    {
      "entropy": 5.560296821594238,
      "epoch": 3.1435390576739484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040345582504824624,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22321562618017196,
      "num_tokens": 62482135.0,
      "step": 27990
    },
    {
      "entropy": 5.726427888870239,
      "epoch": 3.144100634581906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004034228605305668,
      "loss": 5.1446,
      "mean_token_accuracy": 0.21394767612218857,
      "num_tokens": 62494409.0,
      "step": 27995
    },
    {
      "entropy": 5.695817327499389,
      "epoch": 3.1446622114898637,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040338989192385377,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21482565402984619,
      "num_tokens": 62506247.0,
      "step": 28000
    },
    {
      "entropy": 5.674962186813355,
      "epoch": 3.145223788397821,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004033569192291567,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2133306458592415,
      "num_tokens": 62517580.0,
      "step": 28005
    },
    {
      "entropy": 5.586198234558106,
      "epoch": 3.1457853653057786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004033239424475256,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2232923686504364,
      "num_tokens": 62529086.0,
      "step": 28010
    },
    {
      "entropy": 5.548358488082886,
      "epoch": 3.1463469422137362,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040329096158001033,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22403448969125747,
      "num_tokens": 62540381.0,
      "step": 28015
    },
    {
      "entropy": 5.612685012817383,
      "epoch": 3.146908519121694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004032579766276612,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21473678797483445,
      "num_tokens": 62551756.0,
      "step": 28020
    },
    {
      "entropy": 5.647135829925537,
      "epoch": 3.147470096029651,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040322498759152824,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21850218027830123,
      "num_tokens": 62563666.0,
      "step": 28025
    },
    {
      "entropy": 5.714922285079956,
      "epoch": 3.1480316729376088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.1206,
      "mean_token_accuracy": 0.21375250965356826,
      "num_tokens": 62576326.0,
      "step": 28030
    },
    {
      "entropy": 5.633888483047485,
      "epoch": 3.1485932498455664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004031589972721129,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2255189299583435,
      "num_tokens": 62586911.0,
      "step": 28035
    },
    {
      "entropy": 5.4982349395751955,
      "epoch": 3.149154826753524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004031259959909317,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22494369000196457,
      "num_tokens": 62596694.0,
      "step": 28040
    },
    {
      "entropy": 5.653518104553223,
      "epoch": 3.1497164036614813,
      "grad_norm": 1.140625,
      "learning_rate": 0.000403092990630169,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22567842602729798,
      "num_tokens": 62607576.0,
      "step": 28045
    },
    {
      "entropy": 5.569518566131592,
      "epoch": 3.150277980569439,
      "grad_norm": 1.125,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23064083606004715,
      "num_tokens": 62618015.0,
      "step": 28050
    },
    {
      "entropy": 5.532310628890992,
      "epoch": 3.1508395574773966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004030269676741032,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22184432595968245,
      "num_tokens": 62629317.0,
      "step": 28055
    },
    {
      "entropy": 5.586925411224366,
      "epoch": 3.1514011343853543,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004029939500809023,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21703411936759948,
      "num_tokens": 62640077.0,
      "step": 28060
    },
    {
      "entropy": 5.817638111114502,
      "epoch": 3.1519627112933115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004029609284123243,
      "loss": 5.1728,
      "mean_token_accuracy": 0.2153615579009056,
      "num_tokens": 62652399.0,
      "step": 28065
    },
    {
      "entropy": 5.630207443237305,
      "epoch": 3.152524288201269,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040292790266942074,
      "loss": 4.9208,
      "mean_token_accuracy": 0.23316072821617126,
      "num_tokens": 62663489.0,
      "step": 28070
    },
    {
      "entropy": 5.597611999511718,
      "epoch": 3.153085865109227,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040289487285324324,
      "loss": 5.0499,
      "mean_token_accuracy": 0.21627089977264405,
      "num_tokens": 62674975.0,
      "step": 28075
    },
    {
      "entropy": 5.754488897323609,
      "epoch": 3.153647442017184,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.1064,
      "mean_token_accuracy": 0.2103109285235405,
      "num_tokens": 62686608.0,
      "step": 28080
    },
    {
      "entropy": 5.623704671859741,
      "epoch": 3.1542090189251417,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004028288010052728,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21797872632741927,
      "num_tokens": 62697844.0,
      "step": 28085
    },
    {
      "entropy": 5.514297151565552,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004027957589755837,
      "loss": 4.895,
      "mean_token_accuracy": 0.2297250121831894,
      "num_tokens": 62709092.0,
      "step": 28090
    },
    {
      "entropy": 5.5182853698730465,
      "epoch": 3.155332172741057,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040276271287682813,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2254588022828102,
      "num_tokens": 62720047.0,
      "step": 28095
    },
    {
      "entropy": 5.646989202499389,
      "epoch": 3.1558937496490143,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040272966271005816,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22520441859960555,
      "num_tokens": 62729952.0,
      "step": 28100
    },
    {
      "entropy": 5.6846277713775635,
      "epoch": 3.156455326556972,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040269660847632626,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22060168534517288,
      "num_tokens": 62741961.0,
      "step": 28105
    },
    {
      "entropy": 5.581115293502807,
      "epoch": 3.1570169034649296,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040266355017668475,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21926377415657045,
      "num_tokens": 62753552.0,
      "step": 28110
    },
    {
      "entropy": 5.508254480361939,
      "epoch": 3.1575784803728872,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004026304878121864,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23185245394706727,
      "num_tokens": 62765239.0,
      "step": 28115
    },
    {
      "entropy": 5.512913846969605,
      "epoch": 3.1581400572808445,
      "grad_norm": 1.125,
      "learning_rate": 0.00040259742138388366,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2255934000015259,
      "num_tokens": 62776975.0,
      "step": 28120
    },
    {
      "entropy": 5.549148225784302,
      "epoch": 3.158701634188802,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040256435089282966,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22477833330631256,
      "num_tokens": 62787435.0,
      "step": 28125
    },
    {
      "entropy": 5.588636350631714,
      "epoch": 3.15926321109676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004025312763400772,
      "loss": 4.9546,
      "mean_token_accuracy": 0.2245561510324478,
      "num_tokens": 62799014.0,
      "step": 28130
    },
    {
      "entropy": 5.600396347045899,
      "epoch": 3.1598247880047174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004024981977266794,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21434666067361832,
      "num_tokens": 62810196.0,
      "step": 28135
    },
    {
      "entropy": 5.513755655288696,
      "epoch": 3.1603863649126747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004024651150536896,
      "loss": 4.867,
      "mean_token_accuracy": 0.22421334087848663,
      "num_tokens": 62821416.0,
      "step": 28140
    },
    {
      "entropy": 5.57053508758545,
      "epoch": 3.1609479418206323,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004024320283221611,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22632193714380264,
      "num_tokens": 62831736.0,
      "step": 28145
    },
    {
      "entropy": 5.567305421829223,
      "epoch": 3.16150951872859,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040239893753314735,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22721513509750366,
      "num_tokens": 62842017.0,
      "step": 28150
    },
    {
      "entropy": 5.65273871421814,
      "epoch": 3.1620710956365476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004023658426877021,
      "loss": 5.0872,
      "mean_token_accuracy": 0.21274038702249526,
      "num_tokens": 62853594.0,
      "step": 28155
    },
    {
      "entropy": 5.609517431259155,
      "epoch": 3.162632672544505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22717080712318422,
      "num_tokens": 62864831.0,
      "step": 28160
    },
    {
      "entropy": 5.703553867340088,
      "epoch": 3.1631942494524625,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.0593,
      "mean_token_accuracy": 0.22234822660684586,
      "num_tokens": 62877557.0,
      "step": 28165
    },
    {
      "entropy": 5.565579605102539,
      "epoch": 3.16375582636042,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040226653382331484,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22601305991411208,
      "num_tokens": 62889162.0,
      "step": 28170
    },
    {
      "entropy": 5.49726972579956,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000402233422762682,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22704395800828933,
      "num_tokens": 62899376.0,
      "step": 28175
    },
    {
      "entropy": 5.58323335647583,
      "epoch": 3.164878980176335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22608432024717331,
      "num_tokens": 62912448.0,
      "step": 28180
    },
    {
      "entropy": 5.592794132232666,
      "epoch": 3.1654405570842927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004021671884889862,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2238811194896698,
      "num_tokens": 62924021.0,
      "step": 28185
    },
    {
      "entropy": 5.641269683837891,
      "epoch": 3.1660021339922504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004021340652780321,
      "loss": 5.0243,
      "mean_token_accuracy": 0.22193113565444947,
      "num_tokens": 62935330.0,
      "step": 28190
    },
    {
      "entropy": 5.578600454330444,
      "epoch": 3.1665637109002076,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004021009380190801,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22852264940738679,
      "num_tokens": 62947211.0,
      "step": 28195
    },
    {
      "entropy": 5.715967750549316,
      "epoch": 3.1671252878081653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040206780671318495,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2197772428393364,
      "num_tokens": 62959012.0,
      "step": 28200
    },
    {
      "entropy": 5.647746229171753,
      "epoch": 3.167686864716123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004020346713614015,
      "loss": 5.0646,
      "mean_token_accuracy": 0.2211380809545517,
      "num_tokens": 62970608.0,
      "step": 28205
    },
    {
      "entropy": 5.6823042869567875,
      "epoch": 3.1682484416240806,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040200153196478486,
      "loss": 5.0515,
      "mean_token_accuracy": 0.22555961906909944,
      "num_tokens": 62982613.0,
      "step": 28210
    },
    {
      "entropy": 5.5096086978912355,
      "epoch": 3.168810018532038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004019683885243902,
      "loss": 4.8948,
      "mean_token_accuracy": 0.22356845885515214,
      "num_tokens": 62994394.0,
      "step": 28215
    },
    {
      "entropy": 5.608139896392823,
      "epoch": 3.1693715954399955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004019352410412728,
      "loss": 5.046,
      "mean_token_accuracy": 0.22121376395225525,
      "num_tokens": 63007017.0,
      "step": 28220
    },
    {
      "entropy": 5.645710897445679,
      "epoch": 3.169933172347953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040190208951648807,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2214498773217201,
      "num_tokens": 63018325.0,
      "step": 28225
    },
    {
      "entropy": 5.68251519203186,
      "epoch": 3.170494749255911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004018689339510916,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22514639645814896,
      "num_tokens": 63030694.0,
      "step": 28230
    },
    {
      "entropy": 5.530396032333374,
      "epoch": 3.171056326163868,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040183577434613904,
      "loss": 4.927,
      "mean_token_accuracy": 0.22445379793643952,
      "num_tokens": 63040889.0,
      "step": 28235
    },
    {
      "entropy": 5.4680568218231205,
      "epoch": 3.1716179030718257,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004018026107026863,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23373449891805648,
      "num_tokens": 63051385.0,
      "step": 28240
    },
    {
      "entropy": 5.526856565475464,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004017694430217892,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22108276188373566,
      "num_tokens": 63061989.0,
      "step": 28245
    },
    {
      "entropy": 5.669236946105957,
      "epoch": 3.172741056887741,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004017362713045039,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21660840809345244,
      "num_tokens": 63073031.0,
      "step": 28250
    },
    {
      "entropy": 5.588223648071289,
      "epoch": 3.173302633795698,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040170309555188655,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2256650745868683,
      "num_tokens": 63083674.0,
      "step": 28255
    },
    {
      "entropy": 5.538250207901001,
      "epoch": 3.173864210703656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004016699157649936,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2368538960814476,
      "num_tokens": 63094916.0,
      "step": 28260
    },
    {
      "entropy": 5.571185064315796,
      "epoch": 3.1744257876116135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21617870926856994,
      "num_tokens": 63106307.0,
      "step": 28265
    },
    {
      "entropy": 5.54916296005249,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040160354409260625,
      "loss": 4.8862,
      "mean_token_accuracy": 0.2241792395710945,
      "num_tokens": 63115871.0,
      "step": 28270
    },
    {
      "entropy": 5.4778868675231935,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004015703522092254,
      "loss": 4.8828,
      "mean_token_accuracy": 0.2305024191737175,
      "num_tokens": 63126282.0,
      "step": 28275
    },
    {
      "entropy": 5.5309157371521,
      "epoch": 3.176110518335486,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004015371562957954,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22547917515039445,
      "num_tokens": 63138061.0,
      "step": 28280
    },
    {
      "entropy": 5.640383100509643,
      "epoch": 3.1766720952434437,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004015039563533732,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22064743041992188,
      "num_tokens": 63148478.0,
      "step": 28285
    },
    {
      "entropy": 5.645572900772095,
      "epoch": 3.177233672151401,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004014707523830161,
      "loss": 4.9962,
      "mean_token_accuracy": 0.22998037040233613,
      "num_tokens": 63161205.0,
      "step": 28290
    },
    {
      "entropy": 5.592842388153076,
      "epoch": 3.1777952490593586,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004014375443857811,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2291691705584526,
      "num_tokens": 63173532.0,
      "step": 28295
    },
    {
      "entropy": 5.611170625686645,
      "epoch": 3.1783568259673163,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004014043323627257,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22191998809576036,
      "num_tokens": 63185254.0,
      "step": 28300
    },
    {
      "entropy": 5.647577905654908,
      "epoch": 3.178918402875274,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004013711163149072,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21572269946336747,
      "num_tokens": 63196481.0,
      "step": 28305
    },
    {
      "entropy": 5.575115823745728,
      "epoch": 3.179479979783231,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040133789624338353,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22845613211393356,
      "num_tokens": 63207411.0,
      "step": 28310
    },
    {
      "entropy": 5.595909357070923,
      "epoch": 3.180041556691189,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040130467214921216,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22292390614748,
      "num_tokens": 63219097.0,
      "step": 28315
    },
    {
      "entropy": 5.629835844039917,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000401271444033451,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21961677372455596,
      "num_tokens": 63230530.0,
      "step": 28320
    },
    {
      "entropy": 5.580039358139038,
      "epoch": 3.181164710507104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004012382118971582,
      "loss": 4.9642,
      "mean_token_accuracy": 0.226490581035614,
      "num_tokens": 63241918.0,
      "step": 28325
    },
    {
      "entropy": 5.675161838531494,
      "epoch": 3.1817262874150614,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004012049757413917,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21632503867149352,
      "num_tokens": 63253821.0,
      "step": 28330
    },
    {
      "entropy": 5.648536825180054,
      "epoch": 3.182287864323019,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040117173556720976,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22554912120103837,
      "num_tokens": 63265463.0,
      "step": 28335
    },
    {
      "entropy": 5.525294637680053,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040113849137567094,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22703124284744264,
      "num_tokens": 63276010.0,
      "step": 28340
    },
    {
      "entropy": 5.5307801246643065,
      "epoch": 3.1834110181389343,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004011052431678336,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22321630418300628,
      "num_tokens": 63287179.0,
      "step": 28345
    },
    {
      "entropy": 5.583430385589599,
      "epoch": 3.1839725950468916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004010719909447564,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22702300250530244,
      "num_tokens": 63298453.0,
      "step": 28350
    },
    {
      "entropy": 5.5841882705688475,
      "epoch": 3.1845341719548492,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004010387347074981,
      "loss": 4.9369,
      "mean_token_accuracy": 0.220402692258358,
      "num_tokens": 63310130.0,
      "step": 28355
    },
    {
      "entropy": 5.592842769622803,
      "epoch": 3.185095748862807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040100547445711765,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22334382981061934,
      "num_tokens": 63319809.0,
      "step": 28360
    },
    {
      "entropy": 5.601833009719849,
      "epoch": 3.185657325770764,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040097221019467393,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2222991779446602,
      "num_tokens": 63331263.0,
      "step": 28365
    },
    {
      "entropy": 5.595993614196777,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.9609,
      "mean_token_accuracy": 0.222407828271389,
      "num_tokens": 63341983.0,
      "step": 28370
    },
    {
      "entropy": 5.692892694473267,
      "epoch": 3.1867804795866794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040090566963783375,
      "loss": 4.9743,
      "mean_token_accuracy": 0.22852126508951187,
      "num_tokens": 63352856.0,
      "step": 28375
    },
    {
      "entropy": 5.598694229125977,
      "epoch": 3.187342056494637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040087239334555603,
      "loss": 5.0178,
      "mean_token_accuracy": 0.22259896397590637,
      "num_tokens": 63364893.0,
      "step": 28380
    },
    {
      "entropy": 5.606858015060425,
      "epoch": 3.1879036334025943,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040083911304545234,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22884502559900283,
      "num_tokens": 63375222.0,
      "step": 28385
    },
    {
      "entropy": 5.49627914428711,
      "epoch": 3.188465210310552,
      "grad_norm": 1.0,
      "learning_rate": 0.00040080582873858254,
      "loss": 4.9262,
      "mean_token_accuracy": 0.21856527030467987,
      "num_tokens": 63387056.0,
      "step": 28390
    },
    {
      "entropy": 5.618712759017944,
      "epoch": 3.1890267872185096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004007725404260064,
      "loss": 5.0312,
      "mean_token_accuracy": 0.22088707685470582,
      "num_tokens": 63397475.0,
      "step": 28395
    },
    {
      "entropy": 5.59579610824585,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004007392481087837,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22524986118078233,
      "num_tokens": 63407920.0,
      "step": 28400
    },
    {
      "entropy": 5.514806699752808,
      "epoch": 3.1901499410344245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004007059517879747,
      "loss": 4.883,
      "mean_token_accuracy": 0.2333171010017395,
      "num_tokens": 63418537.0,
      "step": 28405
    },
    {
      "entropy": 5.475877523422241,
      "epoch": 3.190711517942382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004006726514646393,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22656189054250717,
      "num_tokens": 63429844.0,
      "step": 28410
    },
    {
      "entropy": 5.608374404907226,
      "epoch": 3.19127309485034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040063934713983803,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22443969398736954,
      "num_tokens": 63441417.0,
      "step": 28415
    },
    {
      "entropy": 5.536208438873291,
      "epoch": 3.1918346717582975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22067860513925552,
      "num_tokens": 63451772.0,
      "step": 28420
    },
    {
      "entropy": 5.549397659301758,
      "epoch": 3.1923962486662547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040057272649007944,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22724031060934066,
      "num_tokens": 63462683.0,
      "step": 28425
    },
    {
      "entropy": 5.6528455257415775,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040053941016724326,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21743520945310593,
      "num_tokens": 63473627.0,
      "step": 28430
    },
    {
      "entropy": 5.638227033615112,
      "epoch": 3.19351940248217,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.0501,
      "mean_token_accuracy": 0.22136472463607787,
      "num_tokens": 63485554.0,
      "step": 28435
    },
    {
      "entropy": 5.681614637374878,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004004727655309614,
      "loss": 5.0195,
      "mean_token_accuracy": 0.2250698372721672,
      "num_tokens": 63495881.0,
      "step": 28440
    },
    {
      "entropy": 5.655694389343262,
      "epoch": 3.194642556298085,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040043943721963766,
      "loss": 4.9583,
      "mean_token_accuracy": 0.21817548871040343,
      "num_tokens": 63506639.0,
      "step": 28445
    },
    {
      "entropy": 5.612939453125,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040040610491427357,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21423243582248688,
      "num_tokens": 63518482.0,
      "step": 28450
    },
    {
      "entropy": 5.499284887313843,
      "epoch": 3.1957657101140002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004003727686159305,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22716365456581117,
      "num_tokens": 63529019.0,
      "step": 28455
    },
    {
      "entropy": 5.598705720901489,
      "epoch": 3.1963272870219575,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040033942832566984,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22512402087450029,
      "num_tokens": 63540330.0,
      "step": 28460
    },
    {
      "entropy": 5.542538976669311,
      "epoch": 3.196888863929915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004003060840445531,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2184634104371071,
      "num_tokens": 63552689.0,
      "step": 28465
    },
    {
      "entropy": 5.5746564865112305,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.125,
      "learning_rate": 0.0004002727357736422,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22719901353120803,
      "num_tokens": 63562954.0,
      "step": 28470
    },
    {
      "entropy": 5.649770784378052,
      "epoch": 3.1980120177458304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040023938351399865,
      "loss": 5.0177,
      "mean_token_accuracy": 0.22075794637203217,
      "num_tokens": 63574697.0,
      "step": 28475
    },
    {
      "entropy": 5.62939224243164,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004002060272666846,
      "loss": 4.999,
      "mean_token_accuracy": 0.23243386447429656,
      "num_tokens": 63585060.0,
      "step": 28480
    },
    {
      "entropy": 5.595090579986572,
      "epoch": 3.1991351715617453,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004001726670327621,
      "loss": 4.9644,
      "mean_token_accuracy": 0.23038899451494216,
      "num_tokens": 63596749.0,
      "step": 28485
    },
    {
      "entropy": 5.541469383239746,
      "epoch": 3.199696748469703,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004001393028132933,
      "loss": 4.8351,
      "mean_token_accuracy": 0.23246824890375137,
      "num_tokens": 63607629.0,
      "step": 28490
    },
    {
      "entropy": 5.545659732818604,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040010593460934065,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2275586649775505,
      "num_tokens": 63617354.0,
      "step": 28495
    },
    {
      "entropy": 5.551672792434692,
      "epoch": 3.200819902285618,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040007256242196644,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22281697690486907,
      "num_tokens": 63629320.0,
      "step": 28500
    },
    {
      "entropy": 5.645938682556152,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040003918625223335,
      "loss": 4.9905,
      "mean_token_accuracy": 0.2186792701482773,
      "num_tokens": 63640453.0,
      "step": 28505
    },
    {
      "entropy": 5.582763671875,
      "epoch": 3.201943056101533,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040000580610120406,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2290517657995224,
      "num_tokens": 63652484.0,
      "step": 28510
    },
    {
      "entropy": 5.557972049713134,
      "epoch": 3.202504633009491,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003999724219699415,
      "loss": 4.9929,
      "mean_token_accuracy": 0.23150503486394883,
      "num_tokens": 63663084.0,
      "step": 28515
    },
    {
      "entropy": 5.535643243789673,
      "epoch": 3.203066209917448,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003999390338595085,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22430554479360582,
      "num_tokens": 63674023.0,
      "step": 28520
    },
    {
      "entropy": 5.647040367126465,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039990564177096817,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22746275812387468,
      "num_tokens": 63685370.0,
      "step": 28525
    },
    {
      "entropy": 5.623907423019409,
      "epoch": 3.2041893637333634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.0395,
      "mean_token_accuracy": 0.2224223092198372,
      "num_tokens": 63696550.0,
      "step": 28530
    },
    {
      "entropy": 5.723749113082886,
      "epoch": 3.204750940641321,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.0815,
      "mean_token_accuracy": 0.22133966386318207,
      "num_tokens": 63708040.0,
      "step": 28535
    },
    {
      "entropy": 5.6280927658081055,
      "epoch": 3.2053125175492783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003998054416473363,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2188633307814598,
      "num_tokens": 63719704.0,
      "step": 28540
    },
    {
      "entropy": 5.589123821258545,
      "epoch": 3.205874094457236,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039977203365700026,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22122121155261992,
      "num_tokens": 63730111.0,
      "step": 28545
    },
    {
      "entropy": 5.6268377780914305,
      "epoch": 3.2064356713651936,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003997386216938743,
      "loss": 4.973,
      "mean_token_accuracy": 0.23212677836418152,
      "num_tokens": 63741054.0,
      "step": 28550
    },
    {
      "entropy": 5.637464189529419,
      "epoch": 3.206997248273151,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039970520575902216,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22215241491794585,
      "num_tokens": 63752598.0,
      "step": 28555
    },
    {
      "entropy": 5.596385383605957,
      "epoch": 3.2075588251811085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000399671785853508,
      "loss": 5.0448,
      "mean_token_accuracy": 0.22921195179224013,
      "num_tokens": 63763927.0,
      "step": 28560
    },
    {
      "entropy": 5.598347234725952,
      "epoch": 3.208120402089066,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039963836197839584,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22475967407226563,
      "num_tokens": 63775072.0,
      "step": 28565
    },
    {
      "entropy": 5.554984426498413,
      "epoch": 3.208681978997024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039960493413474977,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22510649710893632,
      "num_tokens": 63786044.0,
      "step": 28570
    },
    {
      "entropy": 5.561098051071167,
      "epoch": 3.209243555904981,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039957150232363433,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22479264736175536,
      "num_tokens": 63797889.0,
      "step": 28575
    },
    {
      "entropy": 5.672915744781494,
      "epoch": 3.2098051328129387,
      "grad_norm": 1.15625,
      "learning_rate": 0.000399538066546114,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21616403311491011,
      "num_tokens": 63809567.0,
      "step": 28580
    },
    {
      "entropy": 5.668558740615845,
      "epoch": 3.2103667097208963,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039950462680325336,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22252431958913804,
      "num_tokens": 63819617.0,
      "step": 28585
    },
    {
      "entropy": 5.680795955657959,
      "epoch": 3.210928286628854,
      "grad_norm": 1.125,
      "learning_rate": 0.000399471183096117,
      "loss": 5.1401,
      "mean_token_accuracy": 0.21411874741315842,
      "num_tokens": 63831535.0,
      "step": 28590
    },
    {
      "entropy": 5.677439260482788,
      "epoch": 3.211489863536811,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003994377354257698,
      "loss": 5.0547,
      "mean_token_accuracy": 0.22520504593849183,
      "num_tokens": 63842629.0,
      "step": 28595
    },
    {
      "entropy": 5.692846059799194,
      "epoch": 3.212051440444769,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039940428379327703,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21888138800859452,
      "num_tokens": 63853677.0,
      "step": 28600
    },
    {
      "entropy": 5.652558517456055,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003993708281997036,
      "loss": 5.0245,
      "mean_token_accuracy": 0.2169983059167862,
      "num_tokens": 63865536.0,
      "step": 28605
    },
    {
      "entropy": 5.61738600730896,
      "epoch": 3.213174594260684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003993373686461147,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21813882440328597,
      "num_tokens": 63875849.0,
      "step": 28610
    },
    {
      "entropy": 5.646253156661987,
      "epoch": 3.2137361711686414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003993039051335758,
      "loss": 5.0195,
      "mean_token_accuracy": 0.22597746551036835,
      "num_tokens": 63887222.0,
      "step": 28615
    },
    {
      "entropy": 5.594878768920898,
      "epoch": 3.214297748076599,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003992704376631523,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22728149890899657,
      "num_tokens": 63897969.0,
      "step": 28620
    },
    {
      "entropy": 5.722101926803589,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003992369662359099,
      "loss": 5.0948,
      "mean_token_accuracy": 0.2245547130703926,
      "num_tokens": 63908913.0,
      "step": 28625
    },
    {
      "entropy": 5.634805154800415,
      "epoch": 3.2154209018925144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2191460222005844,
      "num_tokens": 63919269.0,
      "step": 28630
    },
    {
      "entropy": 5.5699716091156,
      "epoch": 3.2159824788004716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039917001151523144,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22149965316057205,
      "num_tokens": 63930624.0,
      "step": 28635
    },
    {
      "entropy": 5.572513771057129,
      "epoch": 3.2165440557084293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003991365282239272,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22260877639055252,
      "num_tokens": 63941571.0,
      "step": 28640
    },
    {
      "entropy": 5.6370776176452635,
      "epoch": 3.217105632616387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039910304098006775,
      "loss": 5.04,
      "mean_token_accuracy": 0.22261199802160264,
      "num_tokens": 63953221.0,
      "step": 28645
    },
    {
      "entropy": 5.63666729927063,
      "epoch": 3.217667209524344,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003990695497847193,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22600871026515962,
      "num_tokens": 63965076.0,
      "step": 28650
    },
    {
      "entropy": 5.615154075622558,
      "epoch": 3.218228786432302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003990360546389484,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23532200753688812,
      "num_tokens": 63976601.0,
      "step": 28655
    },
    {
      "entropy": 5.559586477279663,
      "epoch": 3.2187903633402595,
      "grad_norm": 1.125,
      "learning_rate": 0.00039900255554382126,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2268896296620369,
      "num_tokens": 63987757.0,
      "step": 28660
    },
    {
      "entropy": 5.616945600509643,
      "epoch": 3.219351940248217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039896905250040473,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21981529742479325,
      "num_tokens": 64000104.0,
      "step": 28665
    },
    {
      "entropy": 5.603150653839111,
      "epoch": 3.2199135171561744,
      "grad_norm": 1.125,
      "learning_rate": 0.00039893554550976545,
      "loss": 4.8911,
      "mean_token_accuracy": 0.234066042304039,
      "num_tokens": 64010773.0,
      "step": 28670
    },
    {
      "entropy": 5.564967060089112,
      "epoch": 3.220475094064132,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003989020345729705,
      "loss": 4.9104,
      "mean_token_accuracy": 0.225721774995327,
      "num_tokens": 64022269.0,
      "step": 28675
    },
    {
      "entropy": 5.578052949905396,
      "epoch": 3.2210366709720897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039886851969108653,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2256040617823601,
      "num_tokens": 64033622.0,
      "step": 28680
    },
    {
      "entropy": 5.557599782943726,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039883500086518086,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22575750648975373,
      "num_tokens": 64044243.0,
      "step": 28685
    },
    {
      "entropy": 5.670924234390259,
      "epoch": 3.2221598247880046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039880147809632084,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22276675701141357,
      "num_tokens": 64054452.0,
      "step": 28690
    },
    {
      "entropy": 5.619150161743164,
      "epoch": 3.222721401695962,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039876795138557363,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22552392333745958,
      "num_tokens": 64065743.0,
      "step": 28695
    },
    {
      "entropy": 5.6658824443817135,
      "epoch": 3.22328297860392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003987344207340069,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22398579865694046,
      "num_tokens": 64077359.0,
      "step": 28700
    },
    {
      "entropy": 5.5664574146270756,
      "epoch": 3.2238445555118775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039870088614268824,
      "loss": 4.918,
      "mean_token_accuracy": 0.2325995534658432,
      "num_tokens": 64088281.0,
      "step": 28705
    },
    {
      "entropy": 5.563473176956177,
      "epoch": 3.2244061324198348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039866734761268534,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22170877307653428,
      "num_tokens": 64099600.0,
      "step": 28710
    },
    {
      "entropy": 5.668042612075806,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039863380514506614,
      "loss": 4.9492,
      "mean_token_accuracy": 0.2288786679506302,
      "num_tokens": 64110381.0,
      "step": 28715
    },
    {
      "entropy": 5.625917196273804,
      "epoch": 3.22552928623575,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039860025874089865,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2206554889678955,
      "num_tokens": 64120878.0,
      "step": 28720
    },
    {
      "entropy": 5.5008951187133786,
      "epoch": 3.2260908631437077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000398566708401251,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2380290761590004,
      "num_tokens": 64131427.0,
      "step": 28725
    },
    {
      "entropy": 5.6074951171875,
      "epoch": 3.226652440051665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039853315412719144,
      "loss": 5.0248,
      "mean_token_accuracy": 0.222533118724823,
      "num_tokens": 64142833.0,
      "step": 28730
    },
    {
      "entropy": 5.535354995727539,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003984995959197883,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22509381622076036,
      "num_tokens": 64152405.0,
      "step": 28735
    },
    {
      "entropy": 5.6260666847229,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22199005037546157,
      "num_tokens": 64164398.0,
      "step": 28740
    },
    {
      "entropy": 5.621505928039551,
      "epoch": 3.2283371707755375,
      "grad_norm": 1.125,
      "learning_rate": 0.00039843246770922557,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22048428803682327,
      "num_tokens": 64175313.0,
      "step": 28745
    },
    {
      "entropy": 5.494534063339233,
      "epoch": 3.228898747683495,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003983988977082033,
      "loss": 4.861,
      "mean_token_accuracy": 0.22561048865318298,
      "num_tokens": 64185866.0,
      "step": 28750
    },
    {
      "entropy": 5.578336906433106,
      "epoch": 3.229460324591453,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039836532377811237,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2276625007390976,
      "num_tokens": 64197126.0,
      "step": 28755
    },
    {
      "entropy": 5.529683446884155,
      "epoch": 3.2300219014994105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003983317459200217,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22240368127822877,
      "num_tokens": 64209271.0,
      "step": 28760
    },
    {
      "entropy": 5.663451719284057,
      "epoch": 3.2305834784073677,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039829816413500035,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2216580927371979,
      "num_tokens": 64221200.0,
      "step": 28765
    },
    {
      "entropy": 5.654840660095215,
      "epoch": 3.2311450553153254,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003982645784241176,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21755270212888717,
      "num_tokens": 64230835.0,
      "step": 28770
    },
    {
      "entropy": 5.623922967910767,
      "epoch": 3.231706632223283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003982309887884429,
      "loss": 4.9355,
      "mean_token_accuracy": 0.23253943175077438,
      "num_tokens": 64242243.0,
      "step": 28775
    },
    {
      "entropy": 5.613471937179566,
      "epoch": 3.2322682091312407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22777685672044753,
      "num_tokens": 64253362.0,
      "step": 28780
    },
    {
      "entropy": 5.5891608715057375,
      "epoch": 3.232829786039198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039816379774699573,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2259465679526329,
      "num_tokens": 64264462.0,
      "step": 28785
    },
    {
      "entropy": 5.554599952697754,
      "epoch": 3.2333913629471556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039813019634336264,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2218666583299637,
      "num_tokens": 64275153.0,
      "step": 28790
    },
    {
      "entropy": 5.571752738952637,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003980965910192163,
      "loss": 4.9158,
      "mean_token_accuracy": 0.2332244709134102,
      "num_tokens": 64288082.0,
      "step": 28795
    },
    {
      "entropy": 5.574436902999878,
      "epoch": 3.234514516763071,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039806298177562675,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2252626284956932,
      "num_tokens": 64298872.0,
      "step": 28800
    },
    {
      "entropy": 5.627319192886352,
      "epoch": 3.235076093671028,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039802936861366416,
      "loss": 5.0719,
      "mean_token_accuracy": 0.2140505850315094,
      "num_tokens": 64310245.0,
      "step": 28805
    },
    {
      "entropy": 5.66608304977417,
      "epoch": 3.2356376705789858,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003979957515343988,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2224906101822853,
      "num_tokens": 64323681.0,
      "step": 28810
    },
    {
      "entropy": 5.583838701248169,
      "epoch": 3.2361992474869434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000397962130538901,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22457601875066757,
      "num_tokens": 64335265.0,
      "step": 28815
    },
    {
      "entropy": 5.680289030075073,
      "epoch": 3.236760824394901,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039792850562824124,
      "loss": 5.0356,
      "mean_token_accuracy": 0.22150188833475112,
      "num_tokens": 64346279.0,
      "step": 28820
    },
    {
      "entropy": 5.635941886901856,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003978948768034902,
      "loss": 4.964,
      "mean_token_accuracy": 0.23079219162464143,
      "num_tokens": 64357511.0,
      "step": 28825
    },
    {
      "entropy": 5.596211338043213,
      "epoch": 3.237883978210816,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039786124406571854,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2177782714366913,
      "num_tokens": 64367863.0,
      "step": 28830
    },
    {
      "entropy": 5.544863891601563,
      "epoch": 3.2384455551187736,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039782760741599734,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22308127880096434,
      "num_tokens": 64379746.0,
      "step": 28835
    },
    {
      "entropy": 5.604982948303222,
      "epoch": 3.239007132026731,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003977939668553973,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22891133278608322,
      "num_tokens": 64390370.0,
      "step": 28840
    },
    {
      "entropy": 5.6408263683319095,
      "epoch": 3.2395687089346885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039776032238498995,
      "loss": 5.0702,
      "mean_token_accuracy": 0.21778807193040847,
      "num_tokens": 64402620.0,
      "step": 28845
    },
    {
      "entropy": 5.620437383651733,
      "epoch": 3.240130285842646,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003977266740058462,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2280335932970047,
      "num_tokens": 64412969.0,
      "step": 28850
    },
    {
      "entropy": 5.610666131973266,
      "epoch": 3.240691862750604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003976930217190376,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21933689564466477,
      "num_tokens": 64423679.0,
      "step": 28855
    },
    {
      "entropy": 5.709005069732666,
      "epoch": 3.241253439658561,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003976593655256356,
      "loss": 5.0045,
      "mean_token_accuracy": 0.225109700858593,
      "num_tokens": 64436124.0,
      "step": 28860
    },
    {
      "entropy": 5.584344482421875,
      "epoch": 3.2418150165665187,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22356816828250886,
      "num_tokens": 64446494.0,
      "step": 28865
    },
    {
      "entropy": 5.562240219116211,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.109375,
      "learning_rate": 0.000397592041423338,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2278032749891281,
      "num_tokens": 64456152.0,
      "step": 28870
    },
    {
      "entropy": 5.660274648666382,
      "epoch": 3.242938170382434,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039755837351658605,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22307189106941222,
      "num_tokens": 64468004.0,
      "step": 28875
    },
    {
      "entropy": 5.618701457977295,
      "epoch": 3.2434997472903913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003975247017075279,
      "loss": 4.952,
      "mean_token_accuracy": 0.2292359948158264,
      "num_tokens": 64478506.0,
      "step": 28880
    },
    {
      "entropy": 5.535727691650391,
      "epoch": 3.244061324198349,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003974910259972358,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22485654652118683,
      "num_tokens": 64489379.0,
      "step": 28885
    },
    {
      "entropy": 5.621009778976441,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003974573463867819,
      "loss": 5.0149,
      "mean_token_accuracy": 0.22636639028787614,
      "num_tokens": 64499500.0,
      "step": 28890
    },
    {
      "entropy": 5.547055006027222,
      "epoch": 3.2451844780142642,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039742366287723855,
      "loss": 4.916,
      "mean_token_accuracy": 0.23254459202289582,
      "num_tokens": 64510869.0,
      "step": 28895
    },
    {
      "entropy": 5.55799298286438,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039738997546967834,
      "loss": 4.9936,
      "mean_token_accuracy": 0.21728908717632295,
      "num_tokens": 64521581.0,
      "step": 28900
    },
    {
      "entropy": 5.653681039810181,
      "epoch": 3.246307631830179,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039735628416517394,
      "loss": 5.0458,
      "mean_token_accuracy": 0.21907103508710862,
      "num_tokens": 64532436.0,
      "step": 28905
    },
    {
      "entropy": 5.680411863327026,
      "epoch": 3.246869208738137,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039732258896479786,
      "loss": 5.009,
      "mean_token_accuracy": 0.22143014371395112,
      "num_tokens": 64543546.0,
      "step": 28910
    },
    {
      "entropy": 5.729063558578491,
      "epoch": 3.2474307856460944,
      "grad_norm": 1.125,
      "learning_rate": 0.0003972888898696231,
      "loss": 5.1047,
      "mean_token_accuracy": 0.216753052175045,
      "num_tokens": 64554964.0,
      "step": 28915
    },
    {
      "entropy": 5.637884569168091,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039725518688072276,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22022752612829208,
      "num_tokens": 64564082.0,
      "step": 28920
    },
    {
      "entropy": 5.625524044036865,
      "epoch": 3.2485539394620093,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039722147999916976,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21797842979431153,
      "num_tokens": 64575377.0,
      "step": 28925
    },
    {
      "entropy": 5.65581169128418,
      "epoch": 3.249115516369967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003971877692260375,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2235351487994194,
      "num_tokens": 64587686.0,
      "step": 28930
    },
    {
      "entropy": 5.590783786773682,
      "epoch": 3.249677093277924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039715405456239924,
      "loss": 4.907,
      "mean_token_accuracy": 0.21993665099143983,
      "num_tokens": 64598501.0,
      "step": 28935
    },
    {
      "entropy": 5.570322751998901,
      "epoch": 3.250238670185882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003971203360093286,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2241857185959816,
      "num_tokens": 64609341.0,
      "step": 28940
    },
    {
      "entropy": 5.668070077896118,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039708661356789896,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21986350268125535,
      "num_tokens": 64621613.0,
      "step": 28945
    },
    {
      "entropy": 5.5079131603240965,
      "epoch": 3.251361824001797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039705288723918415,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21936972290277482,
      "num_tokens": 64633358.0,
      "step": 28950
    },
    {
      "entropy": 5.568751859664917,
      "epoch": 3.2519234009097544,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22413573712110518,
      "num_tokens": 64644799.0,
      "step": 28955
    },
    {
      "entropy": 5.574782133102417,
      "epoch": 3.252484977817712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000396985422924195,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22964214235544206,
      "num_tokens": 64657028.0,
      "step": 28960
    },
    {
      "entropy": 5.715937519073487,
      "epoch": 3.2530465547256697,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039695168494006857,
      "loss": 5.0516,
      "mean_token_accuracy": 0.22329056411981582,
      "num_tokens": 64668572.0,
      "step": 28965
    },
    {
      "entropy": 5.540576314926147,
      "epoch": 3.2536081316336274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003969179430729531,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22625081539154052,
      "num_tokens": 64679168.0,
      "step": 28970
    },
    {
      "entropy": 5.691466760635376,
      "epoch": 3.2541697085415846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003968841973239231,
      "loss": 5.0258,
      "mean_token_accuracy": 0.22450696676969528,
      "num_tokens": 64691571.0,
      "step": 28975
    },
    {
      "entropy": 5.693182468414307,
      "epoch": 3.2547312854495423,
      "grad_norm": 0.984375,
      "learning_rate": 0.000396850447694053,
      "loss": 5.0493,
      "mean_token_accuracy": 0.22156324833631516,
      "num_tokens": 64703659.0,
      "step": 28980
    },
    {
      "entropy": 5.610098600387573,
      "epoch": 3.2552928623575,
      "grad_norm": 1.796875,
      "learning_rate": 0.0003968166941844174,
      "loss": 4.8985,
      "mean_token_accuracy": 0.23464993834495546,
      "num_tokens": 64715121.0,
      "step": 28985
    },
    {
      "entropy": 5.578380346298218,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.171875,
      "learning_rate": 0.000396782936796091,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2218610018491745,
      "num_tokens": 64726289.0,
      "step": 28990
    },
    {
      "entropy": 5.60251784324646,
      "epoch": 3.256416016173415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003967491755301487,
      "loss": 4.9379,
      "mean_token_accuracy": 0.23008400201797485,
      "num_tokens": 64737378.0,
      "step": 28995
    },
    {
      "entropy": 5.709953641891479,
      "epoch": 3.2569775930813725,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.1408,
      "mean_token_accuracy": 0.2108800709247589,
      "num_tokens": 64749203.0,
      "step": 29000
    },
    {
      "entropy": 5.568360805511475,
      "epoch": 3.25753916998933,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039668164136971615,
      "loss": 4.884,
      "mean_token_accuracy": 0.23128418773412704,
      "num_tokens": 64759514.0,
      "step": 29005
    },
    {
      "entropy": 5.583649969100952,
      "epoch": 3.258100746897288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003966478684773763,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2246372163295746,
      "num_tokens": 64769370.0,
      "step": 29010
    },
    {
      "entropy": 5.572261810302734,
      "epoch": 3.258662323805245,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039661409171172117,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22287596762180328,
      "num_tokens": 64779930.0,
      "step": 29015
    },
    {
      "entropy": 5.595095014572143,
      "epoch": 3.2592239007132027,
      "grad_norm": 1.0,
      "learning_rate": 0.00039658031107382615,
      "loss": 5.009,
      "mean_token_accuracy": 0.21806578785181047,
      "num_tokens": 64791726.0,
      "step": 29020
    },
    {
      "entropy": 5.545407676696778,
      "epoch": 3.2597854776211603,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039654652656476695,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2353230223059654,
      "num_tokens": 64802778.0,
      "step": 29025
    },
    {
      "entropy": 5.588475131988526,
      "epoch": 3.2603470545291175,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003965127381856191,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22143194526433946,
      "num_tokens": 64814664.0,
      "step": 29030
    },
    {
      "entropy": 5.602697229385376,
      "epoch": 3.260908631437075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039647894593745865,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2250217542052269,
      "num_tokens": 64825775.0,
      "step": 29035
    },
    {
      "entropy": 5.679543924331665,
      "epoch": 3.261470208345033,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039644514982136136,
      "loss": 5.0827,
      "mean_token_accuracy": 0.22122997492551805,
      "num_tokens": 64837376.0,
      "step": 29040
    },
    {
      "entropy": 5.66120171546936,
      "epoch": 3.2620317852529905,
      "grad_norm": 1.125,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22315388917922974,
      "num_tokens": 64847941.0,
      "step": 29045
    },
    {
      "entropy": 5.678922700881958,
      "epoch": 3.2625933621609478,
      "grad_norm": 1.0625,
      "learning_rate": 0.000396377545989661,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2191832795739174,
      "num_tokens": 64858793.0,
      "step": 29050
    },
    {
      "entropy": 5.626545429229736,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039634373827621047,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22564042806625367,
      "num_tokens": 64870233.0,
      "step": 29055
    },
    {
      "entropy": 5.593213415145874,
      "epoch": 3.263716515976863,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23442210704088212,
      "num_tokens": 64881875.0,
      "step": 29060
    },
    {
      "entropy": 5.6038435935974125,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039627611125949086,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22837890088558196,
      "num_tokens": 64892837.0,
      "step": 29065
    },
    {
      "entropy": 5.748413705825806,
      "epoch": 3.264839669792778,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2185278058052063,
      "num_tokens": 64904773.0,
      "step": 29070
    },
    {
      "entropy": 5.602965402603149,
      "epoch": 3.2654012467007356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039620846879685767,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22119901478290557,
      "num_tokens": 64916596.0,
      "step": 29075
    },
    {
      "entropy": 5.567519235610962,
      "epoch": 3.2659628236086933,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039617464177601564,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22482199668884278,
      "num_tokens": 64927736.0,
      "step": 29080
    },
    {
      "entropy": 5.66715784072876,
      "epoch": 3.266524400516651,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.0534,
      "mean_token_accuracy": 0.2169657677412033,
      "num_tokens": 64940426.0,
      "step": 29085
    },
    {
      "entropy": 5.605153656005859,
      "epoch": 3.267085977424608,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039610697616066604,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22389972805976868,
      "num_tokens": 64951034.0,
      "step": 29090
    },
    {
      "entropy": 5.71485481262207,
      "epoch": 3.267647554332566,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039607313756831296,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21867750436067582,
      "num_tokens": 64962900.0,
      "step": 29095
    },
    {
      "entropy": 5.5827555656433105,
      "epoch": 3.2682091312405235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2341007947921753,
      "num_tokens": 64974820.0,
      "step": 29100
    },
    {
      "entropy": 5.512251663208008,
      "epoch": 3.268770708148481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2249526709318161,
      "num_tokens": 64985964.0,
      "step": 29105
    },
    {
      "entropy": 5.608211994171143,
      "epoch": 3.2693322850564384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039597159866546996,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2227803036570549,
      "num_tokens": 64996772.0,
      "step": 29110
    },
    {
      "entropy": 5.580973052978516,
      "epoch": 3.269893861964396,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003959377446595201,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22808169573545456,
      "num_tokens": 65007296.0,
      "step": 29115
    },
    {
      "entropy": 5.547392511367798,
      "epoch": 3.2704554388723537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039590388680286543,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2325539067387581,
      "num_tokens": 65018384.0,
      "step": 29120
    },
    {
      "entropy": 5.540770435333252,
      "epoch": 3.271017015780311,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003958700250965842,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22917550057172775,
      "num_tokens": 65029042.0,
      "step": 29125
    },
    {
      "entropy": 5.621155452728272,
      "epoch": 3.2715785926882686,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003958361595417544,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22887336909770967,
      "num_tokens": 65040792.0,
      "step": 29130
    },
    {
      "entropy": 5.620250940322876,
      "epoch": 3.2721401695962262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003958022901394545,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2325655147433281,
      "num_tokens": 65051702.0,
      "step": 29135
    },
    {
      "entropy": 5.564271354675293,
      "epoch": 3.272701746504184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003957684168907628,
      "loss": 4.9263,
      "mean_token_accuracy": 0.22834542542695999,
      "num_tokens": 65063288.0,
      "step": 29140
    },
    {
      "entropy": 5.668840742111206,
      "epoch": 3.273263323412141,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039573453979675793,
      "loss": 5.0792,
      "mean_token_accuracy": 0.208476322889328,
      "num_tokens": 65074913.0,
      "step": 29145
    },
    {
      "entropy": 5.6129457473754885,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039570065885851844,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2164423570036888,
      "num_tokens": 65085327.0,
      "step": 29150
    },
    {
      "entropy": 5.616092300415039,
      "epoch": 3.2743864772280564,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003956667740771231,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22371978163719178,
      "num_tokens": 65095709.0,
      "step": 29155
    },
    {
      "entropy": 5.680869102478027,
      "epoch": 3.274948054136014,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003956328854536509,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22179834693670272,
      "num_tokens": 65106786.0,
      "step": 29160
    },
    {
      "entropy": 5.597148466110229,
      "epoch": 3.2755096310439713,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039559899298918094,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22083580046892165,
      "num_tokens": 65117506.0,
      "step": 29165
    },
    {
      "entropy": 5.476471614837647,
      "epoch": 3.276071207951929,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23840109407901763,
      "num_tokens": 65127830.0,
      "step": 29170
    },
    {
      "entropy": 5.60612735748291,
      "epoch": 3.2766327848598866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000395531196541564,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22516121119260787,
      "num_tokens": 65139562.0,
      "step": 29175
    },
    {
      "entropy": 5.621824312210083,
      "epoch": 3.2771943617678443,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039549729256057583,
      "loss": 5.0074,
      "mean_token_accuracy": 0.22522930651903153,
      "num_tokens": 65150799.0,
      "step": 29180
    },
    {
      "entropy": 5.557878255844116,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000395463384742907,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2206523284316063,
      "num_tokens": 65161458.0,
      "step": 29185
    },
    {
      "entropy": 5.578874015808106,
      "epoch": 3.278317515583759,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039542947308963733,
      "loss": 4.9372,
      "mean_token_accuracy": 0.22932585924863816,
      "num_tokens": 65172098.0,
      "step": 29190
    },
    {
      "entropy": 5.646677112579345,
      "epoch": 3.278879092491717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003953955576018466,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22250466346740722,
      "num_tokens": 65183668.0,
      "step": 29195
    },
    {
      "entropy": 5.601449823379516,
      "epoch": 3.2794406693996745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003953616382806145,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2290105015039444,
      "num_tokens": 65193920.0,
      "step": 29200
    },
    {
      "entropy": 5.655771493911743,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039532771512702124,
      "loss": 5.025,
      "mean_token_accuracy": 0.21662343591451644,
      "num_tokens": 65203397.0,
      "step": 29205
    },
    {
      "entropy": 5.573693561553955,
      "epoch": 3.2805638232155894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003952937881421468,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22273130863904952,
      "num_tokens": 65213821.0,
      "step": 29210
    },
    {
      "entropy": 5.535756254196167,
      "epoch": 3.281125400123547,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22628660947084428,
      "num_tokens": 65224034.0,
      "step": 29215
    },
    {
      "entropy": 5.5211631774902346,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003952259226828759,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22212125957012177,
      "num_tokens": 65234879.0,
      "step": 29220
    },
    {
      "entropy": 5.612458324432373,
      "epoch": 3.282248553939462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039519198421064005,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23135270327329635,
      "num_tokens": 65246530.0,
      "step": 29225
    },
    {
      "entropy": 5.54842734336853,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039515804191144496,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2268802836537361,
      "num_tokens": 65258102.0,
      "step": 29230
    },
    {
      "entropy": 5.567752075195313,
      "epoch": 3.2833717077553772,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039512409578637114,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2398371711373329,
      "num_tokens": 65269396.0,
      "step": 29235
    },
    {
      "entropy": 5.552043628692627,
      "epoch": 3.2839332846633345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039509014583649947,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2311846435070038,
      "num_tokens": 65280004.0,
      "step": 29240
    },
    {
      "entropy": 5.610094928741455,
      "epoch": 3.284494861571292,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039505619206291106,
      "loss": 4.966,
      "mean_token_accuracy": 0.22929433286190032,
      "num_tokens": 65291564.0,
      "step": 29245
    },
    {
      "entropy": 5.556966352462768,
      "epoch": 3.2850564384792498,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.8744,
      "mean_token_accuracy": 0.23181488662958144,
      "num_tokens": 65303032.0,
      "step": 29250
    },
    {
      "entropy": 5.515826320648193,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003949882730489082,
      "loss": 4.8427,
      "mean_token_accuracy": 0.22802288830280304,
      "num_tokens": 65313658.0,
      "step": 29255
    },
    {
      "entropy": 5.700947093963623,
      "epoch": 3.2861795922951647,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039495430781065645,
      "loss": 5.0337,
      "mean_token_accuracy": 0.22538705468177794,
      "num_tokens": 65324593.0,
      "step": 29260
    },
    {
      "entropy": 5.680521440505982,
      "epoch": 3.2867411692031223,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000394920338753013,
      "loss": 4.9846,
      "mean_token_accuracy": 0.23106090277433394,
      "num_tokens": 65335059.0,
      "step": 29265
    },
    {
      "entropy": 5.702870321273804,
      "epoch": 3.28730274611108,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003948863658770594,
      "loss": 5.0908,
      "mean_token_accuracy": 0.22158855497837066,
      "num_tokens": 65347082.0,
      "step": 29270
    },
    {
      "entropy": 5.513784742355346,
      "epoch": 3.2878643230190376,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22503457665443422,
      "num_tokens": 65358644.0,
      "step": 29275
    },
    {
      "entropy": 5.603819179534912,
      "epoch": 3.288425899926995,
      "grad_norm": 1.125,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.922,
      "mean_token_accuracy": 0.2299182116985321,
      "num_tokens": 65369652.0,
      "step": 29280
    },
    {
      "entropy": 5.618800926208496,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.9231,
      "mean_token_accuracy": 0.23179646134376525,
      "num_tokens": 65380196.0,
      "step": 29285
    },
    {
      "entropy": 5.538757944107056,
      "epoch": 3.28954905374291,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039475043621178005,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22716948091983796,
      "num_tokens": 65391135.0,
      "step": 29290
    },
    {
      "entropy": 5.66074914932251,
      "epoch": 3.290110630650868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003947164442605041,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21987611800432205,
      "num_tokens": 65402330.0,
      "step": 29295
    },
    {
      "entropy": 5.660170841217041,
      "epoch": 3.290672207558825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003946824484974101,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2238352581858635,
      "num_tokens": 65414111.0,
      "step": 29300
    },
    {
      "entropy": 5.617268896102905,
      "epoch": 3.2912337844667827,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003946484489235805,
      "loss": 4.9773,
      "mean_token_accuracy": 0.223811312019825,
      "num_tokens": 65424760.0,
      "step": 29305
    },
    {
      "entropy": 5.539198780059815,
      "epoch": 3.2917953613747404,
      "grad_norm": 1.125,
      "learning_rate": 0.00039461444554009787,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2302793487906456,
      "num_tokens": 65435213.0,
      "step": 29310
    },
    {
      "entropy": 5.524311780929565,
      "epoch": 3.2923569382826976,
      "grad_norm": 1.0625,
      "learning_rate": 0.000394580438348045,
      "loss": 4.884,
      "mean_token_accuracy": 0.2180895358324051,
      "num_tokens": 65446141.0,
      "step": 29315
    },
    {
      "entropy": 5.640000724792481,
      "epoch": 3.2929185151906553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039454642734850447,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22849621772766113,
      "num_tokens": 65457634.0,
      "step": 29320
    },
    {
      "entropy": 5.596838474273682,
      "epoch": 3.293480092098613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003945124125425594,
      "loss": 5.008,
      "mean_token_accuracy": 0.22478293627500534,
      "num_tokens": 65468457.0,
      "step": 29325
    },
    {
      "entropy": 5.655100584030151,
      "epoch": 3.2940416690065706,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003944783939312928,
      "loss": 4.9726,
      "mean_token_accuracy": 0.2232600674033165,
      "num_tokens": 65479938.0,
      "step": 29330
    },
    {
      "entropy": 5.6469110488891605,
      "epoch": 3.294603245914528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003944443715157877,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21842904686927794,
      "num_tokens": 65492030.0,
      "step": 29335
    },
    {
      "entropy": 5.679401493072509,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003944103452971275,
      "loss": 5.003,
      "mean_token_accuracy": 0.22627753019332886,
      "num_tokens": 65503385.0,
      "step": 29340
    },
    {
      "entropy": 5.535217428207398,
      "epoch": 3.295726399730443,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.8906,
      "mean_token_accuracy": 0.22694095820188523,
      "num_tokens": 65514027.0,
      "step": 29345
    },
    {
      "entropy": 5.543468046188354,
      "epoch": 3.296287976638401,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039434228145467554,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22214825749397277,
      "num_tokens": 65525653.0,
      "step": 29350
    },
    {
      "entropy": 5.532548475265503,
      "epoch": 3.296849553546358,
      "grad_norm": 1.109375,
      "learning_rate": 0.000394308243833051,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2227436661720276,
      "num_tokens": 65536040.0,
      "step": 29355
    },
    {
      "entropy": 5.71178035736084,
      "epoch": 3.2974111304543157,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039427420241260557,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21805694252252578,
      "num_tokens": 65547505.0,
      "step": 29360
    },
    {
      "entropy": 5.679691505432129,
      "epoch": 3.2979727073622733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.9494,
      "mean_token_accuracy": 0.23311132937669754,
      "num_tokens": 65557565.0,
      "step": 29365
    },
    {
      "entropy": 5.569593238830566,
      "epoch": 3.298534284270231,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039420610817958836,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22286006361246108,
      "num_tokens": 65568729.0,
      "step": 29370
    },
    {
      "entropy": 5.6255536556243895,
      "epoch": 3.299095861178188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003941720553691846,
      "loss": 5.0497,
      "mean_token_accuracy": 0.22369274944067002,
      "num_tokens": 65580664.0,
      "step": 29375
    },
    {
      "entropy": 5.526137733459473,
      "epoch": 3.299657438086146,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003941379987642963,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23252714574337005,
      "num_tokens": 65591348.0,
      "step": 29380
    },
    {
      "entropy": 5.604962635040283,
      "epoch": 3.3002190149941035,
      "grad_norm": 1.15625,
      "learning_rate": 0.000394103938366008,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22575028538703917,
      "num_tokens": 65603553.0,
      "step": 29385
    },
    {
      "entropy": 5.587457799911499,
      "epoch": 3.300780591902061,
      "grad_norm": 1.125,
      "learning_rate": 0.0003940698741754041,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2316314086318016,
      "num_tokens": 65613903.0,
      "step": 29390
    },
    {
      "entropy": 5.606266450881958,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003940358061935692,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2260923430323601,
      "num_tokens": 65624862.0,
      "step": 29395
    },
    {
      "entropy": 5.549845266342163,
      "epoch": 3.301903745717976,
      "grad_norm": 1.125,
      "learning_rate": 0.0003940017344215881,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2244821310043335,
      "num_tokens": 65636828.0,
      "step": 29400
    },
    {
      "entropy": 5.622585391998291,
      "epoch": 3.3024653226259337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039396765886054555,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22432016879320144,
      "num_tokens": 65648251.0,
      "step": 29405
    },
    {
      "entropy": 5.684137153625488,
      "epoch": 3.303026899533891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039393357951152667,
      "loss": 5.1081,
      "mean_token_accuracy": 0.21863993406295776,
      "num_tokens": 65659715.0,
      "step": 29410
    },
    {
      "entropy": 5.638957214355469,
      "epoch": 3.3035884764418486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039389949637561646,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22060005217790604,
      "num_tokens": 65670844.0,
      "step": 29415
    },
    {
      "entropy": 5.532516813278198,
      "epoch": 3.3041500533498063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039386540945390017,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23008323311805726,
      "num_tokens": 65682404.0,
      "step": 29420
    },
    {
      "entropy": 5.5967834949493405,
      "epoch": 3.304711630257764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003938313187474631,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22379263043403624,
      "num_tokens": 65692826.0,
      "step": 29425
    },
    {
      "entropy": 5.616573429107666,
      "epoch": 3.3052732071657216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003937972242573908,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22712322920560837,
      "num_tokens": 65704139.0,
      "step": 29430
    },
    {
      "entropy": 5.5529036045074465,
      "epoch": 3.305834784073679,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003937631259847688,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22671705931425096,
      "num_tokens": 65716282.0,
      "step": 29435
    },
    {
      "entropy": 5.64962363243103,
      "epoch": 3.3063963609816365,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21727515161037445,
      "num_tokens": 65728086.0,
      "step": 29440
    },
    {
      "entropy": 5.6755164623260494,
      "epoch": 3.306957937889594,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039369491809621867,
      "loss": 5.0322,
      "mean_token_accuracy": 0.22613831013441085,
      "num_tokens": 65740584.0,
      "step": 29445
    },
    {
      "entropy": 5.596152591705322,
      "epoch": 3.3075195147975514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003936608084824623,
      "loss": 4.9982,
      "mean_token_accuracy": 0.2238476201891899,
      "num_tokens": 65751479.0,
      "step": 29450
    },
    {
      "entropy": 5.56687183380127,
      "epoch": 3.308081091705509,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039362669509049975,
      "loss": 4.9983,
      "mean_token_accuracy": 0.22575914412736892,
      "num_tokens": 65762073.0,
      "step": 29455
    },
    {
      "entropy": 5.540249872207641,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039359257792141717,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21964372247457503,
      "num_tokens": 65772808.0,
      "step": 29460
    },
    {
      "entropy": 5.586643838882447,
      "epoch": 3.3092042455214243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000393558456976301,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22664046585559844,
      "num_tokens": 65783507.0,
      "step": 29465
    },
    {
      "entropy": 5.628276109695435,
      "epoch": 3.3097658224293816,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003935243322562375,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22871482372283936,
      "num_tokens": 65794250.0,
      "step": 29470
    },
    {
      "entropy": 5.578682804107666,
      "epoch": 3.310327399337339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003934902037623133,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22142910361289977,
      "num_tokens": 65805487.0,
      "step": 29475
    },
    {
      "entropy": 5.625634479522705,
      "epoch": 3.310888976245297,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039345607149561513,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2185932144522667,
      "num_tokens": 65818264.0,
      "step": 29480
    },
    {
      "entropy": 5.719651079177856,
      "epoch": 3.3114505531532545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.0904,
      "mean_token_accuracy": 0.22662899345159532,
      "num_tokens": 65829686.0,
      "step": 29485
    },
    {
      "entropy": 5.585727691650391,
      "epoch": 3.3120121300612118,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003933877956482438,
      "loss": 4.8839,
      "mean_token_accuracy": 0.23620502203702926,
      "num_tokens": 65840800.0,
      "step": 29490
    },
    {
      "entropy": 5.61941294670105,
      "epoch": 3.3125737069691694,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039335365206974465,
      "loss": 4.9697,
      "mean_token_accuracy": 0.2290483981370926,
      "num_tokens": 65853634.0,
      "step": 29495
    },
    {
      "entropy": 5.589557313919068,
      "epoch": 3.313135283877127,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003933195047228193,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22021453976631164,
      "num_tokens": 65864754.0,
      "step": 29500
    },
    {
      "entropy": 5.55167179107666,
      "epoch": 3.3136968607850843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039328535360855503,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22082629203796386,
      "num_tokens": 65876609.0,
      "step": 29505
    },
    {
      "entropy": 5.574027395248413,
      "epoch": 3.314258437693042,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039325119872803924,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21986743211746215,
      "num_tokens": 65887583.0,
      "step": 29510
    },
    {
      "entropy": 5.622020149230957,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23562580943107606,
      "num_tokens": 65898131.0,
      "step": 29515
    },
    {
      "entropy": 5.566421842575073,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22519881427288055,
      "num_tokens": 65909353.0,
      "step": 29520
    },
    {
      "entropy": 5.601705837249756,
      "epoch": 3.315943168416915,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039314871149985834,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22314474284648894,
      "num_tokens": 65920709.0,
      "step": 29525
    },
    {
      "entropy": 5.636108684539795,
      "epoch": 3.316504745324872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003931145415652126,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22749040722846986,
      "num_tokens": 65932231.0,
      "step": 29530
    },
    {
      "entropy": 5.723119735717773,
      "epoch": 3.31706632223283,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2162722408771515,
      "num_tokens": 65942629.0,
      "step": 29535
    },
    {
      "entropy": 5.638902759552002,
      "epoch": 3.3176278991407875,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003930461904145709,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22951164841651917,
      "num_tokens": 65954385.0,
      "step": 29540
    },
    {
      "entropy": 5.592551946640015,
      "epoch": 3.3181894760487447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003930120092007512,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22887906730175017,
      "num_tokens": 65966283.0,
      "step": 29545
    },
    {
      "entropy": 5.666283369064331,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003929778242293834,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21589374244213105,
      "num_tokens": 65978594.0,
      "step": 29550
    },
    {
      "entropy": 5.553373670578003,
      "epoch": 3.31931262986466,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22189778238534927,
      "num_tokens": 65989398.0,
      "step": 29555
    },
    {
      "entropy": 5.572187900543213,
      "epoch": 3.3198742067726177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039290944301835723,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22488241046667098,
      "num_tokens": 66000545.0,
      "step": 29560
    },
    {
      "entropy": 5.5813665866851805,
      "epoch": 3.320435783680575,
      "grad_norm": 1.125,
      "learning_rate": 0.00039287524678087617,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2265148639678955,
      "num_tokens": 66011638.0,
      "step": 29565
    },
    {
      "entropy": 5.55621976852417,
      "epoch": 3.3209973605885326,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039284104679020154,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22367535531520844,
      "num_tokens": 66023284.0,
      "step": 29570
    },
    {
      "entropy": 5.54575252532959,
      "epoch": 3.3215589374964902,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21942208409309388,
      "num_tokens": 66034099.0,
      "step": 29575
    },
    {
      "entropy": 5.578746366500854,
      "epoch": 3.322120514404448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003927726355536275,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22676158249378203,
      "num_tokens": 66043885.0,
      "step": 29580
    },
    {
      "entropy": 5.643530607223511,
      "epoch": 3.322682091312405,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039273842430990634,
      "loss": 5.0337,
      "mean_token_accuracy": 0.21319985538721084,
      "num_tokens": 66054190.0,
      "step": 29585
    },
    {
      "entropy": 5.609216165542603,
      "epoch": 3.3232436682203628,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003927042093173481,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22630761563777924,
      "num_tokens": 66065254.0,
      "step": 29590
    },
    {
      "entropy": 5.63163275718689,
      "epoch": 3.3238052451283204,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003926699905770421,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21926576644182205,
      "num_tokens": 66076582.0,
      "step": 29595
    },
    {
      "entropy": 5.675109434127807,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000392635768090078,
      "loss": 5.0128,
      "mean_token_accuracy": 0.2225568637251854,
      "num_tokens": 66086410.0,
      "step": 29600
    },
    {
      "entropy": 5.639864826202393,
      "epoch": 3.3249283989442353,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039260154185754545,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21384839117527008,
      "num_tokens": 66098393.0,
      "step": 29605
    },
    {
      "entropy": 5.651441144943237,
      "epoch": 3.325489975852193,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003925673118805341,
      "loss": 4.9892,
      "mean_token_accuracy": 0.22520021498203277,
      "num_tokens": 66108077.0,
      "step": 29610
    },
    {
      "entropy": 5.479496192932129,
      "epoch": 3.3260515527601506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000392533078160134,
      "loss": 4.783,
      "mean_token_accuracy": 0.2394810825586319,
      "num_tokens": 66118008.0,
      "step": 29615
    },
    {
      "entropy": 5.590894937515259,
      "epoch": 3.3266131296681083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039249884069743514,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21898385286331176,
      "num_tokens": 66128697.0,
      "step": 29620
    },
    {
      "entropy": 5.652091789245605,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003924645994935276,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22295572310686113,
      "num_tokens": 66139800.0,
      "step": 29625
    },
    {
      "entropy": 5.624443817138672,
      "epoch": 3.327736283484023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003924303545495017,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23206108212471008,
      "num_tokens": 66151202.0,
      "step": 29630
    },
    {
      "entropy": 5.6243949890136715,
      "epoch": 3.328297860391981,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039239610586644765,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21996790915727615,
      "num_tokens": 66162159.0,
      "step": 29635
    },
    {
      "entropy": 5.612666654586792,
      "epoch": 3.328859437299938,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003923618534454561,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22944057136774063,
      "num_tokens": 66174492.0,
      "step": 29640
    },
    {
      "entropy": 5.570739984512329,
      "epoch": 3.3294210142078957,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003923275972876176,
      "loss": 5.0074,
      "mean_token_accuracy": 0.22675951719284057,
      "num_tokens": 66186617.0,
      "step": 29645
    },
    {
      "entropy": 5.579316663742065,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23117612302303314,
      "num_tokens": 66198001.0,
      "step": 29650
    },
    {
      "entropy": 5.536938762664795,
      "epoch": 3.330544168023811,
      "grad_norm": 1.078125,
      "learning_rate": 0.000392259073765763,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2277321085333824,
      "num_tokens": 66208017.0,
      "step": 29655
    },
    {
      "entropy": 5.532782649993896,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039222480640392854,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22235383987426757,
      "num_tokens": 66219798.0,
      "step": 29660
    },
    {
      "entropy": 5.5980195045471195,
      "epoch": 3.331667321839726,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039219053530961087,
      "loss": 4.9431,
      "mean_token_accuracy": 0.23159387409687043,
      "num_tokens": 66230662.0,
      "step": 29665
    },
    {
      "entropy": 5.605710506439209,
      "epoch": 3.3322288987476836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2304295301437378,
      "num_tokens": 66240787.0,
      "step": 29670
    },
    {
      "entropy": 5.556810140609741,
      "epoch": 3.3327904756556412,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003921219819278905,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2280251070857048,
      "num_tokens": 66250937.0,
      "step": 29675
    },
    {
      "entropy": 5.665018463134766,
      "epoch": 3.3333520525635985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003920876996426706,
      "loss": 5.063,
      "mean_token_accuracy": 0.21299825310707093,
      "num_tokens": 66261322.0,
      "step": 29680
    },
    {
      "entropy": 5.593554735183716,
      "epoch": 3.333913629471556,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22554341852664947,
      "num_tokens": 66271307.0,
      "step": 29685
    },
    {
      "entropy": 5.5325901985168455,
      "epoch": 3.334475206379514,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23113306909799575,
      "num_tokens": 66283067.0,
      "step": 29690
    },
    {
      "entropy": 5.589841032028199,
      "epoch": 3.335036783287471,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003919848304226711,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22608632594347,
      "num_tokens": 66295853.0,
      "step": 29695
    },
    {
      "entropy": 5.627658128738403,
      "epoch": 3.3355983601954287,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003919505332315307,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23193719387054443,
      "num_tokens": 66307573.0,
      "step": 29700
    },
    {
      "entropy": 5.572369527816773,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039191623231663985,
      "loss": 4.9285,
      "mean_token_accuracy": 0.23076995760202407,
      "num_tokens": 66318100.0,
      "step": 29705
    },
    {
      "entropy": 5.633770608901978,
      "epoch": 3.336721514011344,
      "grad_norm": 1.171875,
      "learning_rate": 0.000391881927679091,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2213249236345291,
      "num_tokens": 66328918.0,
      "step": 29710
    },
    {
      "entropy": 5.5817207336425785,
      "epoch": 3.3372830909193016,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039184761931997607,
      "loss": 5.018,
      "mean_token_accuracy": 0.2190604731440544,
      "num_tokens": 66341124.0,
      "step": 29715
    },
    {
      "entropy": 5.573544979095459,
      "epoch": 3.337844667827259,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003918133072403878,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22146647721529006,
      "num_tokens": 66351033.0,
      "step": 29720
    },
    {
      "entropy": 5.600718355178833,
      "epoch": 3.3384062447352165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039177899144141835,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22711846083402634,
      "num_tokens": 66363338.0,
      "step": 29725
    },
    {
      "entropy": 5.576024150848388,
      "epoch": 3.338967821643174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003917446719241606,
      "loss": 4.9187,
      "mean_token_accuracy": 0.2256780296564102,
      "num_tokens": 66373716.0,
      "step": 29730
    },
    {
      "entropy": 5.62104287147522,
      "epoch": 3.3395293985511314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.9248,
      "mean_token_accuracy": 0.2295919880270958,
      "num_tokens": 66384448.0,
      "step": 29735
    },
    {
      "entropy": 5.5219744682312015,
      "epoch": 3.340090975459089,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039167602173915104,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2284557804465294,
      "num_tokens": 66395662.0,
      "step": 29740
    },
    {
      "entropy": 5.521811389923096,
      "epoch": 3.3406525523670467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22805907130241393,
      "num_tokens": 66407856.0,
      "step": 29745
    },
    {
      "entropy": 5.6853415966033936,
      "epoch": 3.3412141292750044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039160735669410246,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21378647536039352,
      "num_tokens": 66419324.0,
      "step": 29750
    },
    {
      "entropy": 5.559719753265381,
      "epoch": 3.3417757061829616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003915730186017964,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22605901956558228,
      "num_tokens": 66430720.0,
      "step": 29755
    },
    {
      "entropy": 5.5866370677948,
      "epoch": 3.3423372830909193,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039153867679776023,
      "loss": 4.969,
      "mean_token_accuracy": 0.2161650165915489,
      "num_tokens": 66441736.0,
      "step": 29760
    },
    {
      "entropy": 5.637329626083374,
      "epoch": 3.342898859998877,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003915043312830873,
      "loss": 5.0129,
      "mean_token_accuracy": 0.2280655786395073,
      "num_tokens": 66451851.0,
      "step": 29765
    },
    {
      "entropy": 5.719700908660888,
      "epoch": 3.3434604369068346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003914699820588714,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21477592289447783,
      "num_tokens": 66463105.0,
      "step": 29770
    },
    {
      "entropy": 5.668391704559326,
      "epoch": 3.344022013814792,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039143562912620603,
      "loss": 5.0688,
      "mean_token_accuracy": 0.22007047086954118,
      "num_tokens": 66474613.0,
      "step": 29775
    },
    {
      "entropy": 5.591215753555298,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.109375,
      "learning_rate": 0.000391401272486185,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21816185265779495,
      "num_tokens": 66486783.0,
      "step": 29780
    },
    {
      "entropy": 5.717396688461304,
      "epoch": 3.345145167630707,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.0524,
      "mean_token_accuracy": 0.22173926383256912,
      "num_tokens": 66498581.0,
      "step": 29785
    },
    {
      "entropy": 5.555296850204468,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039133254808845216,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2336464747786522,
      "num_tokens": 66509480.0,
      "step": 29790
    },
    {
      "entropy": 5.5545419216156,
      "epoch": 3.346268321446622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039129818033292843,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2246441051363945,
      "num_tokens": 66521831.0,
      "step": 29795
    },
    {
      "entropy": 5.633495759963989,
      "epoch": 3.3468298983545797,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003912638088744256,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22612600028514862,
      "num_tokens": 66533900.0,
      "step": 29800
    },
    {
      "entropy": 5.75235948562622,
      "epoch": 3.3473914752625373,
      "grad_norm": 1.21875,
      "learning_rate": 0.000391229433714038,
      "loss": 5.0868,
      "mean_token_accuracy": 0.21828896403312684,
      "num_tokens": 66543848.0,
      "step": 29805
    },
    {
      "entropy": 5.640008354187012,
      "epoch": 3.347953052170495,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22232812494039536,
      "num_tokens": 66554113.0,
      "step": 29810
    },
    {
      "entropy": 5.652220964431763,
      "epoch": 3.348514629078452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003911606722919866,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22316179424524307,
      "num_tokens": 66565202.0,
      "step": 29815
    },
    {
      "entropy": 5.651106119155884,
      "epoch": 3.34907620598641,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003911262860325123,
      "loss": 5.046,
      "mean_token_accuracy": 0.21576206535100936,
      "num_tokens": 66575930.0,
      "step": 29820
    },
    {
      "entropy": 5.6542895317077635,
      "epoch": 3.3496377828943675,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003910918960755319,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2267905056476593,
      "num_tokens": 66585422.0,
      "step": 29825
    },
    {
      "entropy": 5.644791746139527,
      "epoch": 3.3501993598023248,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003910575024221406,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22077556550502778,
      "num_tokens": 66597395.0,
      "step": 29830
    },
    {
      "entropy": 5.6065749168396,
      "epoch": 3.3507609367102824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039102310507343335,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22566002160310744,
      "num_tokens": 66608464.0,
      "step": 29835
    },
    {
      "entropy": 5.64999566078186,
      "epoch": 3.35132251361824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003909887040305055,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2208642154932022,
      "num_tokens": 66619930.0,
      "step": 29840
    },
    {
      "entropy": 5.5566801071167,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003909542992944523,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22478957772254943,
      "num_tokens": 66632833.0,
      "step": 29845
    },
    {
      "entropy": 5.598654747009277,
      "epoch": 3.352445667434155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003909198908663692,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21686347872018813,
      "num_tokens": 66643641.0,
      "step": 29850
    },
    {
      "entropy": 5.691706800460816,
      "epoch": 3.3530072443421126,
      "grad_norm": 1.125,
      "learning_rate": 0.00039088547874735187,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21782829165458678,
      "num_tokens": 66653863.0,
      "step": 29855
    },
    {
      "entropy": 5.545014381408691,
      "epoch": 3.3535688212500703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039085106293849594,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22786420583724976,
      "num_tokens": 66665885.0,
      "step": 29860
    },
    {
      "entropy": 5.661197519302368,
      "epoch": 3.354130398158028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039081664344089724,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22007269710302352,
      "num_tokens": 66676584.0,
      "step": 29865
    },
    {
      "entropy": 5.541337680816651,
      "epoch": 3.354691975065985,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2318682223558426,
      "num_tokens": 66687882.0,
      "step": 29870
    },
    {
      "entropy": 5.606650924682617,
      "epoch": 3.355253551973943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003907477933838553,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22139347195625306,
      "num_tokens": 66699665.0,
      "step": 29875
    },
    {
      "entropy": 5.642121124267578,
      "epoch": 3.3558151288819005,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21755595952272416,
      "num_tokens": 66711455.0,
      "step": 29880
    },
    {
      "entropy": 5.590040636062622,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039067892858499496,
      "loss": 4.8508,
      "mean_token_accuracy": 0.2357852652668953,
      "num_tokens": 66722369.0,
      "step": 29885
    },
    {
      "entropy": 5.582774686813354,
      "epoch": 3.3569382826978154,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003906444906601237,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22407312989234923,
      "num_tokens": 66733200.0,
      "step": 29890
    },
    {
      "entropy": 5.555273056030273,
      "epoch": 3.357499859605773,
      "grad_norm": 1.0625,
      "learning_rate": 0.000390610049053087,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23069697469472886,
      "num_tokens": 66744650.0,
      "step": 29895
    },
    {
      "entropy": 5.6173571109771725,
      "epoch": 3.3580614365137307,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039057560376498147,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22193157076835632,
      "num_tokens": 66755712.0,
      "step": 29900
    },
    {
      "entropy": 5.679641675949097,
      "epoch": 3.3586230134216883,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003905411547969038,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22572795897722245,
      "num_tokens": 66765855.0,
      "step": 29905
    },
    {
      "entropy": 5.651850080490112,
      "epoch": 3.3591845903296456,
      "grad_norm": 1.125,
      "learning_rate": 0.0003905067021499511,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2292482554912567,
      "num_tokens": 66776870.0,
      "step": 29910
    },
    {
      "entropy": 5.57082166671753,
      "epoch": 3.359746167237603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.9367,
      "mean_token_accuracy": 0.23118110448122026,
      "num_tokens": 66787226.0,
      "step": 29915
    },
    {
      "entropy": 5.560002851486206,
      "epoch": 3.360307744145561,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003904377858238081,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22744033336639405,
      "num_tokens": 66797352.0,
      "step": 29920
    },
    {
      "entropy": 5.657568502426147,
      "epoch": 3.360869321053518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003904033221468122,
      "loss": 5.0592,
      "mean_token_accuracy": 0.2203360080718994,
      "num_tokens": 66809227.0,
      "step": 29925
    },
    {
      "entropy": 5.675014686584473,
      "epoch": 3.3614308979614758,
      "grad_norm": 1.125,
      "learning_rate": 0.0003903688547953297,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2204924464225769,
      "num_tokens": 66821196.0,
      "step": 29930
    },
    {
      "entropy": 5.647099494934082,
      "epoch": 3.3619924748694334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003903343837704582,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22358233332633973,
      "num_tokens": 66831958.0,
      "step": 29935
    },
    {
      "entropy": 5.616717290878296,
      "epoch": 3.362554051777391,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003902999090732951,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22546710968017578,
      "num_tokens": 66843727.0,
      "step": 29940
    },
    {
      "entropy": 5.557184362411499,
      "epoch": 3.3631156286853483,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039026543070493823,
      "loss": 4.9384,
      "mean_token_accuracy": 0.23007319271564483,
      "num_tokens": 66854756.0,
      "step": 29945
    },
    {
      "entropy": 5.509380578994751,
      "epoch": 3.363677205593306,
      "grad_norm": 1.125,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.9157,
      "mean_token_accuracy": 0.23089403063058853,
      "num_tokens": 66865615.0,
      "step": 29950
    },
    {
      "entropy": 5.588259744644165,
      "epoch": 3.3642387825012636,
      "grad_norm": 1.0,
      "learning_rate": 0.0003901964629590344,
      "loss": 4.9728,
      "mean_token_accuracy": 0.217125041782856,
      "num_tokens": 66877262.0,
      "step": 29955
    },
    {
      "entropy": 5.7316468238830565,
      "epoch": 3.3648003594092213,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003901619735836833,
      "loss": 5.0491,
      "mean_token_accuracy": 0.21789505779743196,
      "num_tokens": 66889405.0,
      "step": 29960
    },
    {
      "entropy": 5.614517593383789,
      "epoch": 3.3653619363171785,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039012748054153044,
      "loss": 4.92,
      "mean_token_accuracy": 0.23164958655834197,
      "num_tokens": 66900013.0,
      "step": 29965
    },
    {
      "entropy": 5.5538990020751955,
      "epoch": 3.365923513225136,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003900929838336738,
      "loss": 4.8297,
      "mean_token_accuracy": 0.2325373187661171,
      "num_tokens": 66909731.0,
      "step": 29970
    },
    {
      "entropy": 5.533984804153443,
      "epoch": 3.366485090133094,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.8807,
      "mean_token_accuracy": 0.23249143064022065,
      "num_tokens": 66919731.0,
      "step": 29975
    },
    {
      "entropy": 5.584724998474121,
      "epoch": 3.367046667041051,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003900239794252434,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22184532284736633,
      "num_tokens": 66930169.0,
      "step": 29980
    },
    {
      "entropy": 5.586294031143188,
      "epoch": 3.3676082439490087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003899894717268667,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22738590091466904,
      "num_tokens": 66942010.0,
      "step": 29985
    },
    {
      "entropy": 5.628681945800781,
      "epoch": 3.3681698208569664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003899549603671807,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2295805498957634,
      "num_tokens": 66953972.0,
      "step": 29990
    },
    {
      "entropy": 5.622872304916382,
      "epoch": 3.368731397764924,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038992044534728404,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22343397587537767,
      "num_tokens": 66965548.0,
      "step": 29995
    },
    {
      "entropy": 5.620169544219971,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038988592666827594,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2245585724711418,
      "num_tokens": 66976414.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.36813664356396,
      "eval_loss": 5.083132743835449,
      "eval_mean_token_accuracy": 0.2259759949364089,
      "eval_num_tokens": 66976414.0,
      "eval_runtime": 18.3792,
      "eval_samples_per_second": 1687.399,
      "eval_steps_per_second": 210.945,
      "step": 30000
    },
    {
      "entropy": 5.498836183547974,
      "epoch": 3.369854551580839,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003898514043312553,
      "loss": 4.9196,
      "mean_token_accuracy": 0.23293859809637069,
      "num_tokens": 66987392.0,
      "step": 30005
    },
    {
      "entropy": 5.727926588058471,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038981687833732146,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21703556478023528,
      "num_tokens": 67000444.0,
      "step": 30010
    },
    {
      "entropy": 5.637356615066528,
      "epoch": 3.3709777053967542,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003897823486875737,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22360758483409882,
      "num_tokens": 67012124.0,
      "step": 30015
    },
    {
      "entropy": 5.717983293533325,
      "epoch": 3.3715392823047114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003897478153831114,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2206479251384735,
      "num_tokens": 67024179.0,
      "step": 30020
    },
    {
      "entropy": 5.582675981521606,
      "epoch": 3.372100859212669,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22978525310754777,
      "num_tokens": 67035876.0,
      "step": 30025
    },
    {
      "entropy": 5.512537288665771,
      "epoch": 3.3726624361206268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038967873781444167,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23018353879451753,
      "num_tokens": 67046499.0,
      "step": 30030
    },
    {
      "entropy": 5.597685050964356,
      "epoch": 3.3732240130285844,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038964419355243365,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22459489852190018,
      "num_tokens": 67058357.0,
      "step": 30035
    },
    {
      "entropy": 5.653625917434693,
      "epoch": 3.3737855899365417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003896096456401101,
      "loss": 5.0484,
      "mean_token_accuracy": 0.2165927395224571,
      "num_tokens": 67069753.0,
      "step": 30040
    },
    {
      "entropy": 5.597239971160889,
      "epoch": 3.3743471668444993,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22614361047744752,
      "num_tokens": 67080185.0,
      "step": 30045
    },
    {
      "entropy": 5.612949991226197,
      "epoch": 3.374908743752457,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22654871195554732,
      "num_tokens": 67091382.0,
      "step": 30050
    },
    {
      "entropy": 5.688648700714111,
      "epoch": 3.3754703206604146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003895059800122466,
      "loss": 5.0978,
      "mean_token_accuracy": 0.20925647169351577,
      "num_tokens": 67102952.0,
      "step": 30055
    },
    {
      "entropy": 5.577458143234253,
      "epoch": 3.376031897568372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000389471417509662,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22214341312646865,
      "num_tokens": 67113498.0,
      "step": 30060
    },
    {
      "entropy": 5.576776027679443,
      "epoch": 3.3765934744763295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000389436851362263,
      "loss": 4.9715,
      "mean_token_accuracy": 0.22255683094263076,
      "num_tokens": 67124591.0,
      "step": 30065
    },
    {
      "entropy": 5.610620498657227,
      "epoch": 3.377155051384287,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038940228157115036,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2223120868206024,
      "num_tokens": 67135627.0,
      "step": 30070
    },
    {
      "entropy": 5.59518985748291,
      "epoch": 3.3777166282922444,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038936770813742473,
      "loss": 4.8726,
      "mean_token_accuracy": 0.2318180739879608,
      "num_tokens": 67146084.0,
      "step": 30075
    },
    {
      "entropy": 5.571487236022949,
      "epoch": 3.378278205200202,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038933313106218694,
      "loss": 4.9382,
      "mean_token_accuracy": 0.23071243315935136,
      "num_tokens": 67157749.0,
      "step": 30080
    },
    {
      "entropy": 5.61850323677063,
      "epoch": 3.3788397821081597,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038929855034653776,
      "loss": 5.0489,
      "mean_token_accuracy": 0.22089750468730926,
      "num_tokens": 67169386.0,
      "step": 30085
    },
    {
      "entropy": 5.588671016693115,
      "epoch": 3.3794013590161174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038926396599157855,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2270330548286438,
      "num_tokens": 67181436.0,
      "step": 30090
    },
    {
      "entropy": 5.6323097229003904,
      "epoch": 3.379962935924075,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003892293779984103,
      "loss": 4.9685,
      "mean_token_accuracy": 0.23121491968631744,
      "num_tokens": 67192576.0,
      "step": 30095
    },
    {
      "entropy": 5.685299968719482,
      "epoch": 3.3805245128320323,
      "grad_norm": 1.0,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.05,
      "mean_token_accuracy": 0.22415960878133773,
      "num_tokens": 67205083.0,
      "step": 30100
    },
    {
      "entropy": 5.463965511322021,
      "epoch": 3.38108608973999,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22642669677734376,
      "num_tokens": 67216499.0,
      "step": 30105
    },
    {
      "entropy": 5.5205284595489506,
      "epoch": 3.3816476666479476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23224846124649048,
      "num_tokens": 67228390.0,
      "step": 30110
    },
    {
      "entropy": 5.645029592514038,
      "epoch": 3.382209243555905,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000389090989665675,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22554770857095718,
      "num_tokens": 67239655.0,
      "step": 30115
    },
    {
      "entropy": 5.640701866149902,
      "epoch": 3.3827708204638625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003890563834979835,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21888767778873444,
      "num_tokens": 67251712.0,
      "step": 30120
    },
    {
      "entropy": 5.620458650588989,
      "epoch": 3.38333239737182,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003890217736986926,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22880298048257827,
      "num_tokens": 67262494.0,
      "step": 30125
    },
    {
      "entropy": 5.574230098724366,
      "epoch": 3.383893974279778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038898716026890414,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2308744952082634,
      "num_tokens": 67273179.0,
      "step": 30130
    },
    {
      "entropy": 5.599629497528076,
      "epoch": 3.384455551187735,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003889525432097203,
      "loss": 4.9805,
      "mean_token_accuracy": 0.22810672968626022,
      "num_tokens": 67284031.0,
      "step": 30135
    },
    {
      "entropy": 5.6661937713623045,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003889179225222432,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21490170806646347,
      "num_tokens": 67295728.0,
      "step": 30140
    },
    {
      "entropy": 5.643882322311401,
      "epoch": 3.3855787050036503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038888329820757536,
      "loss": 5.025,
      "mean_token_accuracy": 0.21895310133695603,
      "num_tokens": 67306880.0,
      "step": 30145
    },
    {
      "entropy": 5.591914987564087,
      "epoch": 3.386140281911608,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22725603729486465,
      "num_tokens": 67317849.0,
      "step": 30150
    },
    {
      "entropy": 5.5356730937957765,
      "epoch": 3.386701858819565,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.9235,
      "mean_token_accuracy": 0.2311479091644287,
      "num_tokens": 67329421.0,
      "step": 30155
    },
    {
      "entropy": 5.629998350143433,
      "epoch": 3.387263435727523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038877940351145186,
      "loss": 5.0122,
      "mean_token_accuracy": 0.22255397886037825,
      "num_tokens": 67340072.0,
      "step": 30160
    },
    {
      "entropy": 5.58991756439209,
      "epoch": 3.3878250126354805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22975624948740006,
      "num_tokens": 67351687.0,
      "step": 30165
    },
    {
      "entropy": 5.577630996704102,
      "epoch": 3.3883865895434377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22290034145116805,
      "num_tokens": 67362978.0,
      "step": 30170
    },
    {
      "entropy": 5.558002185821533,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000388675476210306,
      "loss": 4.919,
      "mean_token_accuracy": 0.23568300306797027,
      "num_tokens": 67374142.0,
      "step": 30175
    },
    {
      "entropy": 5.610469818115234,
      "epoch": 3.389509743359353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038864082653617744,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2174864813685417,
      "num_tokens": 67385196.0,
      "step": 30180
    },
    {
      "entropy": 5.570521879196167,
      "epoch": 3.3900713202673107,
      "grad_norm": 1.125,
      "learning_rate": 0.00038860617324368085,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2285354569554329,
      "num_tokens": 67395830.0,
      "step": 30185
    },
    {
      "entropy": 5.5695160865783695,
      "epoch": 3.3906328971752684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003885715163339196,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22069849222898483,
      "num_tokens": 67405950.0,
      "step": 30190
    },
    {
      "entropy": 5.608520889282227,
      "epoch": 3.3911944740832256,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22797985523939132,
      "num_tokens": 67417993.0,
      "step": 30195
    },
    {
      "entropy": 5.6568577766418455,
      "epoch": 3.3917560509911833,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.945,
      "mean_token_accuracy": 0.23161717057228087,
      "num_tokens": 67429222.0,
      "step": 30200
    },
    {
      "entropy": 5.6494324684143065,
      "epoch": 3.392317627899141,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003884675239120838,
      "loss": 5.0621,
      "mean_token_accuracy": 0.2171331062912941,
      "num_tokens": 67439870.0,
      "step": 30205
    },
    {
      "entropy": 5.566991138458252,
      "epoch": 3.392879204807098,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038843285254430005,
      "loss": 4.833,
      "mean_token_accuracy": 0.23095864206552505,
      "num_tokens": 67450491.0,
      "step": 30210
    },
    {
      "entropy": 5.62115216255188,
      "epoch": 3.393440781715056,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003883981775647705,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22271132320165635,
      "num_tokens": 67461432.0,
      "step": 30215
    },
    {
      "entropy": 5.6057322978973385,
      "epoch": 3.3940023586230135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003883634989745988,
      "loss": 4.9703,
      "mean_token_accuracy": 0.2258584350347519,
      "num_tokens": 67472614.0,
      "step": 30220
    },
    {
      "entropy": 5.602308511734009,
      "epoch": 3.394563935530971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038832881677488944,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22383649349212648,
      "num_tokens": 67484383.0,
      "step": 30225
    },
    {
      "entropy": 5.627634954452515,
      "epoch": 3.3951255124389284,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21782958060503005,
      "num_tokens": 67495697.0,
      "step": 30230
    },
    {
      "entropy": 5.580782794952393,
      "epoch": 3.395687089346886,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003882594415512746,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22612252533435823,
      "num_tokens": 67507937.0,
      "step": 30235
    },
    {
      "entropy": 5.586582374572754,
      "epoch": 3.3962486662548437,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22762888818979263,
      "num_tokens": 67519649.0,
      "step": 30240
    },
    {
      "entropy": 5.6986894607543945,
      "epoch": 3.3968102431628013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003881900519027618,
      "loss": 5.0574,
      "mean_token_accuracy": 0.22408982813358308,
      "num_tokens": 67532301.0,
      "step": 30245
    },
    {
      "entropy": 5.611316204071045,
      "epoch": 3.3973718200707586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003881553516719303,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22715854793787002,
      "num_tokens": 67545106.0,
      "step": 30250
    },
    {
      "entropy": 5.642910385131836,
      "epoch": 3.397933396978716,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038812064783818847,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21850500404834747,
      "num_tokens": 67556188.0,
      "step": 30255
    },
    {
      "entropy": 5.576532220840454,
      "epoch": 3.398494973886674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22897903025150299,
      "num_tokens": 67567707.0,
      "step": 30260
    },
    {
      "entropy": 5.596336030960083,
      "epoch": 3.399056550794631,
      "grad_norm": 1.125,
      "learning_rate": 0.000388051229366394,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2198603242635727,
      "num_tokens": 67578334.0,
      "step": 30265
    },
    {
      "entropy": 5.506949234008789,
      "epoch": 3.3996181277025888,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038801651473055167,
      "loss": 4.7819,
      "mean_token_accuracy": 0.22902728766202926,
      "num_tokens": 67588896.0,
      "step": 30270
    },
    {
      "entropy": 5.588464879989624,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003879817964962197,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22058474272489548,
      "num_tokens": 67599032.0,
      "step": 30275
    },
    {
      "entropy": 5.618516874313355,
      "epoch": 3.400741281518504,
      "grad_norm": 1.0,
      "learning_rate": 0.0003879470746645035,
      "loss": 5.0394,
      "mean_token_accuracy": 0.21719000190496446,
      "num_tokens": 67610095.0,
      "step": 30280
    },
    {
      "entropy": 5.656797599792481,
      "epoch": 3.4013028584264617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038791234923650866,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21699742525815963,
      "num_tokens": 67622039.0,
      "step": 30285
    },
    {
      "entropy": 5.6708485126495365,
      "epoch": 3.401864435334419,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038787762021334077,
      "loss": 5.0117,
      "mean_token_accuracy": 0.22028757780790328,
      "num_tokens": 67634123.0,
      "step": 30290
    },
    {
      "entropy": 5.6263471126556395,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.125,
      "learning_rate": 0.0003878428875961057,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22373634278774263,
      "num_tokens": 67644696.0,
      "step": 30295
    },
    {
      "entropy": 5.651935958862305,
      "epoch": 3.4029875891503343,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038780815138590935,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2173176497220993,
      "num_tokens": 67656193.0,
      "step": 30300
    },
    {
      "entropy": 5.535907888412476,
      "epoch": 3.4035491660582915,
      "grad_norm": 1.125,
      "learning_rate": 0.0003877734115838577,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22892231196165086,
      "num_tokens": 67668261.0,
      "step": 30305
    },
    {
      "entropy": 5.635741901397705,
      "epoch": 3.404110742966249,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038773866819105684,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22026900351047515,
      "num_tokens": 67678264.0,
      "step": 30310
    },
    {
      "entropy": 5.620917701721192,
      "epoch": 3.404672319874207,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003877039212086131,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22660502046346664,
      "num_tokens": 67688455.0,
      "step": 30315
    },
    {
      "entropy": 5.545678186416626,
      "epoch": 3.4052338967821645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22927458882331847,
      "num_tokens": 67700970.0,
      "step": 30320
    },
    {
      "entropy": 5.625824689865112,
      "epoch": 3.4057954736901217,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038763441647922246,
      "loss": 5.004,
      "mean_token_accuracy": 0.22628458887338637,
      "num_tokens": 67711835.0,
      "step": 30325
    },
    {
      "entropy": 5.561767673492431,
      "epoch": 3.4063570505980794,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003875996587344885,
      "loss": 4.9431,
      "mean_token_accuracy": 0.22443973124027253,
      "num_tokens": 67722915.0,
      "step": 30330
    },
    {
      "entropy": 5.673721122741699,
      "epoch": 3.406918627506037,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038756489740453785,
      "loss": 4.9966,
      "mean_token_accuracy": 0.22753774523735046,
      "num_tokens": 67734301.0,
      "step": 30335
    },
    {
      "entropy": 5.65439281463623,
      "epoch": 3.4074802044139947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038753013249047706,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22496710121631622,
      "num_tokens": 67745763.0,
      "step": 30340
    },
    {
      "entropy": 5.608177328109742,
      "epoch": 3.408041781321952,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003874953639934134,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22925713062286376,
      "num_tokens": 67757241.0,
      "step": 30345
    },
    {
      "entropy": 5.600006532669068,
      "epoch": 3.4086033582299096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003874605919144536,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22481861859560012,
      "num_tokens": 67768231.0,
      "step": 30350
    },
    {
      "entropy": 5.6430778980255125,
      "epoch": 3.4091649351378672,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.0791,
      "mean_token_accuracy": 0.21761218160390855,
      "num_tokens": 67780735.0,
      "step": 30355
    },
    {
      "entropy": 5.605105638504028,
      "epoch": 3.4097265120458244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003873910370152747,
      "loss": 4.8907,
      "mean_token_accuracy": 0.235588838160038,
      "num_tokens": 67791817.0,
      "step": 30360
    },
    {
      "entropy": 5.643160009384156,
      "epoch": 3.410288088953782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003873562541972702,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2285029411315918,
      "num_tokens": 67803591.0,
      "step": 30365
    },
    {
      "entropy": 5.603289937973022,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003873214678017991,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22237349301576614,
      "num_tokens": 67813250.0,
      "step": 30370
    },
    {
      "entropy": 5.572981929779052,
      "epoch": 3.4114112427696974,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003872866778299688,
      "loss": 4.9553,
      "mean_token_accuracy": 0.2233840212225914,
      "num_tokens": 67825042.0,
      "step": 30375
    },
    {
      "entropy": 5.655691003799438,
      "epoch": 3.411972819677655,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003872518842828872,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21964495331048967,
      "num_tokens": 67837535.0,
      "step": 30380
    },
    {
      "entropy": 5.708159923553467,
      "epoch": 3.4125343965856123,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038721708716166203,
      "loss": 5.0546,
      "mean_token_accuracy": 0.22701074779033661,
      "num_tokens": 67848950.0,
      "step": 30385
    },
    {
      "entropy": 5.6243674755096436,
      "epoch": 3.41309597349357,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003871822864674013,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21576156169176103,
      "num_tokens": 67861488.0,
      "step": 30390
    },
    {
      "entropy": 5.625985431671142,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038714748220121305,
      "loss": 5.0744,
      "mean_token_accuracy": 0.2212109684944153,
      "num_tokens": 67872627.0,
      "step": 30395
    },
    {
      "entropy": 5.587953042984009,
      "epoch": 3.414219127309485,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003871126743642055,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2197524592280388,
      "num_tokens": 67884033.0,
      "step": 30400
    },
    {
      "entropy": 5.607099485397339,
      "epoch": 3.4147807042174425,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038707786295748693,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21708672046661376,
      "num_tokens": 67895696.0,
      "step": 30405
    },
    {
      "entropy": 5.652241277694702,
      "epoch": 3.4153422811254,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.953,
      "mean_token_accuracy": 0.23263088017702102,
      "num_tokens": 67906328.0,
      "step": 30410
    },
    {
      "entropy": 5.5763074398040775,
      "epoch": 3.415903858033358,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2321538120508194,
      "num_tokens": 67917258.0,
      "step": 30415
    },
    {
      "entropy": 5.559299659729004,
      "epoch": 3.416465434941315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003869734073301495,
      "loss": 4.9076,
      "mean_token_accuracy": 0.2224826529622078,
      "num_tokens": 67927431.0,
      "step": 30420
    },
    {
      "entropy": 5.7054667472839355,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038693858165567195,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22721890211105347,
      "num_tokens": 67938801.0,
      "step": 30425
    },
    {
      "entropy": 5.595062160491944,
      "epoch": 3.4175885887572304,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22907227277755737,
      "num_tokens": 67949690.0,
      "step": 30430
    },
    {
      "entropy": 5.5349634170532225,
      "epoch": 3.418150165665188,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038686891961532227,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22979704588651656,
      "num_tokens": 67959542.0,
      "step": 30435
    },
    {
      "entropy": 5.561603116989136,
      "epoch": 3.4187117425731453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.8605,
      "mean_token_accuracy": 0.2257370740175247,
      "num_tokens": 67969759.0,
      "step": 30440
    },
    {
      "entropy": 5.5440000057220455,
      "epoch": 3.419273319481103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038679924332717363,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22839921861886978,
      "num_tokens": 67982302.0,
      "step": 30445
    },
    {
      "entropy": 5.621159267425537,
      "epoch": 3.4198348963890606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038676439984294776,
      "loss": 5.0321,
      "mean_token_accuracy": 0.2203314021229744,
      "num_tokens": 67994452.0,
      "step": 30450
    },
    {
      "entropy": 5.661916780471802,
      "epoch": 3.420396473297018,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038672955280009996,
      "loss": 4.9944,
      "mean_token_accuracy": 0.2217008352279663,
      "num_tokens": 68006445.0,
      "step": 30455
    },
    {
      "entropy": 5.666984939575196,
      "epoch": 3.4209580502049755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038669470219973996,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21985424309968948,
      "num_tokens": 68017720.0,
      "step": 30460
    },
    {
      "entropy": 5.643949317932129,
      "epoch": 3.421519627112933,
      "grad_norm": 1.125,
      "learning_rate": 0.00038665984804297734,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2209124445915222,
      "num_tokens": 68028199.0,
      "step": 30465
    },
    {
      "entropy": 5.62737717628479,
      "epoch": 3.4220812040208908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003866249903309218,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21649700552225112,
      "num_tokens": 68040171.0,
      "step": 30470
    },
    {
      "entropy": 5.640940999984741,
      "epoch": 3.4226427809288484,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003865901290646833,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22276835292577743,
      "num_tokens": 68050484.0,
      "step": 30475
    },
    {
      "entropy": 5.65801968574524,
      "epoch": 3.4232043578368057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038655526424537175,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2247977674007416,
      "num_tokens": 68062110.0,
      "step": 30480
    },
    {
      "entropy": 5.675104665756225,
      "epoch": 3.4237659347447633,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038652039587409734,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22148724347352983,
      "num_tokens": 68074890.0,
      "step": 30485
    },
    {
      "entropy": 5.597316455841065,
      "epoch": 3.424327511652721,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003864855239519703,
      "loss": 4.9804,
      "mean_token_accuracy": 0.21429742723703385,
      "num_tokens": 68084909.0,
      "step": 30490
    },
    {
      "entropy": 5.4986487865448,
      "epoch": 3.424889088560678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038645064848010093,
      "loss": 4.8644,
      "mean_token_accuracy": 0.228127221763134,
      "num_tokens": 68097114.0,
      "step": 30495
    },
    {
      "entropy": 5.642004346847534,
      "epoch": 3.425450665468636,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038641576945959964,
      "loss": 4.991,
      "mean_token_accuracy": 0.23050685077905655,
      "num_tokens": 68108377.0,
      "step": 30500
    },
    {
      "entropy": 5.610368633270264,
      "epoch": 3.4260122423765935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038638088689157713,
      "loss": 4.98,
      "mean_token_accuracy": 0.22784813940525056,
      "num_tokens": 68118859.0,
      "step": 30505
    },
    {
      "entropy": 5.601176309585571,
      "epoch": 3.426573819284551,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038634600077714395,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2310940593481064,
      "num_tokens": 68130151.0,
      "step": 30510
    },
    {
      "entropy": 5.549965906143188,
      "epoch": 3.4271353961925084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003863111111174109,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22808461487293244,
      "num_tokens": 68142514.0,
      "step": 30515
    },
    {
      "entropy": 5.641172075271607,
      "epoch": 3.427696973100466,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038627621791348886,
      "loss": 5.0584,
      "mean_token_accuracy": 0.21964176893234252,
      "num_tokens": 68153751.0,
      "step": 30520
    },
    {
      "entropy": 5.674189567565918,
      "epoch": 3.4282585500084237,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038624132116648897,
      "loss": 4.959,
      "mean_token_accuracy": 0.2288801223039627,
      "num_tokens": 68164885.0,
      "step": 30525
    },
    {
      "entropy": 5.573997974395752,
      "epoch": 3.4288201269163814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003862064208775223,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23156129866838454,
      "num_tokens": 68176320.0,
      "step": 30530
    },
    {
      "entropy": 5.587771463394165,
      "epoch": 3.4293817038243386,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038617151704769994,
      "loss": 4.9366,
      "mean_token_accuracy": 0.23092004358768464,
      "num_tokens": 68186665.0,
      "step": 30535
    },
    {
      "entropy": 5.620091962814331,
      "epoch": 3.4299432807322963,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003861366096781334,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21372112184762954,
      "num_tokens": 68198011.0,
      "step": 30540
    },
    {
      "entropy": 5.686958599090576,
      "epoch": 3.430504857640254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003861016987699342,
      "loss": 5.0833,
      "mean_token_accuracy": 0.2197772517800331,
      "num_tokens": 68209659.0,
      "step": 30545
    },
    {
      "entropy": 5.622744035720825,
      "epoch": 3.431066434548211,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23235464245080947,
      "num_tokens": 68220424.0,
      "step": 30550
    },
    {
      "entropy": 5.572650051116943,
      "epoch": 3.431628011456169,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.931,
      "mean_token_accuracy": 0.22637303173542023,
      "num_tokens": 68231413.0,
      "step": 30555
    },
    {
      "entropy": 5.630416774749756,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038599694482465606,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23513485193252565,
      "num_tokens": 68242435.0,
      "step": 30560
    },
    {
      "entropy": 5.634328842163086,
      "epoch": 3.432751165272084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038596201977304263,
      "loss": 4.9488,
      "mean_token_accuracy": 0.23241870999336242,
      "num_tokens": 68254462.0,
      "step": 30565
    },
    {
      "entropy": 5.611524772644043,
      "epoch": 3.433312742180042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038592709118835547,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21784374117851257,
      "num_tokens": 68265842.0,
      "step": 30570
    },
    {
      "entropy": 5.592187690734863,
      "epoch": 3.433874319087999,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038589215907170664,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22957080602645874,
      "num_tokens": 68277117.0,
      "step": 30575
    },
    {
      "entropy": 5.654007482528686,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003858572234242086,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2286091610789299,
      "num_tokens": 68287869.0,
      "step": 30580
    },
    {
      "entropy": 5.646498823165894,
      "epoch": 3.4349974729039143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22375863790512085,
      "num_tokens": 68300711.0,
      "step": 30585
    },
    {
      "entropy": 5.652654981613159,
      "epoch": 3.4355590498118715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003857873415411138,
      "loss": 5.0686,
      "mean_token_accuracy": 0.22467149645090104,
      "num_tokens": 68311997.0,
      "step": 30590
    },
    {
      "entropy": 5.597251462936401,
      "epoch": 3.436120626719829,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003857523953077423,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2245718315243721,
      "num_tokens": 68322129.0,
      "step": 30595
    },
    {
      "entropy": 5.536134672164917,
      "epoch": 3.436682203627787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038571744554797156,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2320508047938347,
      "num_tokens": 68334202.0,
      "step": 30600
    },
    {
      "entropy": 5.578438949584961,
      "epoch": 3.4372437805357445,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038568249226291433,
      "loss": 4.9473,
      "mean_token_accuracy": 0.23296740353107454,
      "num_tokens": 68345370.0,
      "step": 30605
    },
    {
      "entropy": 5.556228494644165,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.125,
      "learning_rate": 0.00038564753545368375,
      "loss": 4.833,
      "mean_token_accuracy": 0.2293636366724968,
      "num_tokens": 68355703.0,
      "step": 30610
    },
    {
      "entropy": 5.584860992431641,
      "epoch": 3.4383669343516594,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038561257512139264,
      "loss": 4.9913,
      "mean_token_accuracy": 0.21566554307937622,
      "num_tokens": 68367412.0,
      "step": 30615
    },
    {
      "entropy": 5.595297384262085,
      "epoch": 3.438928511259617,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038557761126715426,
      "loss": 4.952,
      "mean_token_accuracy": 0.22787563502788544,
      "num_tokens": 68378074.0,
      "step": 30620
    },
    {
      "entropy": 5.641434144973755,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003855426438920818,
      "loss": 4.9963,
      "mean_token_accuracy": 0.21986093670129775,
      "num_tokens": 68388730.0,
      "step": 30625
    },
    {
      "entropy": 5.562958097457885,
      "epoch": 3.440051665075532,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22801231741905212,
      "num_tokens": 68399172.0,
      "step": 30630
    },
    {
      "entropy": 5.624584579467774,
      "epoch": 3.4406132419834896,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038547269858388853,
      "loss": 4.9956,
      "mean_token_accuracy": 0.23046808391809465,
      "num_tokens": 68411704.0,
      "step": 30635
    },
    {
      "entropy": 5.586614513397217,
      "epoch": 3.4411748188914473,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003854377206529947,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2212286338210106,
      "num_tokens": 68422774.0,
      "step": 30640
    },
    {
      "entropy": 5.59922308921814,
      "epoch": 3.4417363957994045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000385402739205721,
      "loss": 4.8704,
      "mean_token_accuracy": 0.2336407944560051,
      "num_tokens": 68432964.0,
      "step": 30645
    },
    {
      "entropy": 5.60124626159668,
      "epoch": 3.442297972707362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003853677542431813,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21939849555492402,
      "num_tokens": 68445875.0,
      "step": 30650
    },
    {
      "entropy": 5.646076965332031,
      "epoch": 3.44285954961532,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003853327657664895,
      "loss": 5.0313,
      "mean_token_accuracy": 0.220718415081501,
      "num_tokens": 68457972.0,
      "step": 30655
    },
    {
      "entropy": 5.571281766891479,
      "epoch": 3.4434211265232775,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003852977737767596,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23019070029258729,
      "num_tokens": 68468825.0,
      "step": 30660
    },
    {
      "entropy": 5.534804010391236,
      "epoch": 3.443982703431235,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003852627782751059,
      "loss": 4.8845,
      "mean_token_accuracy": 0.2328508198261261,
      "num_tokens": 68479427.0,
      "step": 30665
    },
    {
      "entropy": 5.559336805343628,
      "epoch": 3.4445442803391924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.889,
      "mean_token_accuracy": 0.22634125947952272,
      "num_tokens": 68490542.0,
      "step": 30670
    },
    {
      "entropy": 5.52205982208252,
      "epoch": 3.44510585724715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003851927767404839,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23038580268621445,
      "num_tokens": 68501396.0,
      "step": 30675
    },
    {
      "entropy": 5.687872219085693,
      "epoch": 3.4456674341551077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003851577707097445,
      "loss": 5.0951,
      "mean_token_accuracy": 0.2107059672474861,
      "num_tokens": 68512710.0,
      "step": 30680
    },
    {
      "entropy": 5.607181167602539,
      "epoch": 3.446229011063065,
      "grad_norm": 1.046875,
      "learning_rate": 0.000385122761171539,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21522696763277055,
      "num_tokens": 68524289.0,
      "step": 30685
    },
    {
      "entropy": 5.613995885848999,
      "epoch": 3.4467905879710226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038508774812698207,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2214533269405365,
      "num_tokens": 68536783.0,
      "step": 30690
    },
    {
      "entropy": 5.6208408832550045,
      "epoch": 3.44735216487898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003850527315771885,
      "loss": 5.0064,
      "mean_token_accuracy": 0.22231868654489517,
      "num_tokens": 68548113.0,
      "step": 30695
    },
    {
      "entropy": 5.588720703125,
      "epoch": 3.447913741786938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038501771152327333,
      "loss": 4.8811,
      "mean_token_accuracy": 0.22698419094085692,
      "num_tokens": 68560234.0,
      "step": 30700
    },
    {
      "entropy": 5.591107130050659,
      "epoch": 3.448475318694895,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038498268796635155,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23398738354444504,
      "num_tokens": 68571712.0,
      "step": 30705
    },
    {
      "entropy": 5.546265029907227,
      "epoch": 3.4490368956028528,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22711391150951385,
      "num_tokens": 68583166.0,
      "step": 30710
    },
    {
      "entropy": 5.523808908462525,
      "epoch": 3.4495984725108104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.8279,
      "mean_token_accuracy": 0.2428829029202461,
      "num_tokens": 68594026.0,
      "step": 30715
    },
    {
      "entropy": 5.55025281906128,
      "epoch": 3.450160049418768,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003848775962886987,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2227582797408104,
      "num_tokens": 68604046.0,
      "step": 30720
    },
    {
      "entropy": 5.532541131973266,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003848425587309032,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2367151603102684,
      "num_tokens": 68615575.0,
      "step": 30725
    },
    {
      "entropy": 5.639980983734131,
      "epoch": 3.451283203234683,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038480751767567807,
      "loss": 5.0445,
      "mean_token_accuracy": 0.2230700209736824,
      "num_tokens": 68626805.0,
      "step": 30730
    },
    {
      "entropy": 5.727596569061279,
      "epoch": 3.4518447801426406,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000384772473124139,
      "loss": 5.0443,
      "mean_token_accuracy": 0.22200241833925247,
      "num_tokens": 68638802.0,
      "step": 30735
    },
    {
      "entropy": 5.633684349060059,
      "epoch": 3.452406357050598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003847374250774018,
      "loss": 5.0484,
      "mean_token_accuracy": 0.22055359482765197,
      "num_tokens": 68651137.0,
      "step": 30740
    },
    {
      "entropy": 5.53149266242981,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038470237353658237,
      "loss": 4.893,
      "mean_token_accuracy": 0.23582312762737273,
      "num_tokens": 68661268.0,
      "step": 30745
    },
    {
      "entropy": 5.505885887145996,
      "epoch": 3.453529510866513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038466731850279683,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22172735184431075,
      "num_tokens": 68672334.0,
      "step": 30750
    },
    {
      "entropy": 5.612859725952148,
      "epoch": 3.454091087774471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038463225997716137,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22528911083936692,
      "num_tokens": 68683765.0,
      "step": 30755
    },
    {
      "entropy": 5.763868808746338,
      "epoch": 3.4546526646824285,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038459719796079227,
      "loss": 5.0882,
      "mean_token_accuracy": 0.2203528493642807,
      "num_tokens": 68694969.0,
      "step": 30760
    },
    {
      "entropy": 5.672326612472534,
      "epoch": 3.4552142415903857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003845621324548059,
      "loss": 5.035,
      "mean_token_accuracy": 0.22192320227622986,
      "num_tokens": 68708586.0,
      "step": 30765
    },
    {
      "entropy": 5.6415123462677,
      "epoch": 3.4557758184983434,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003845270634603186,
      "loss": 4.9828,
      "mean_token_accuracy": 0.224620957672596,
      "num_tokens": 68719649.0,
      "step": 30770
    },
    {
      "entropy": 5.644384098052979,
      "epoch": 3.456337395406301,
      "grad_norm": 1.125,
      "learning_rate": 0.00038449199097844724,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2218916818499565,
      "num_tokens": 68729491.0,
      "step": 30775
    },
    {
      "entropy": 5.584581327438355,
      "epoch": 3.4568989723142582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003844569150103084,
      "loss": 4.926,
      "mean_token_accuracy": 0.22644747495651246,
      "num_tokens": 68741087.0,
      "step": 30780
    },
    {
      "entropy": 5.583010959625244,
      "epoch": 3.457460549222216,
      "grad_norm": 1.25,
      "learning_rate": 0.0003844218355570189,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23217882513999938,
      "num_tokens": 68752699.0,
      "step": 30785
    },
    {
      "entropy": 5.60115237236023,
      "epoch": 3.4580221261301736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23090978413820268,
      "num_tokens": 68763066.0,
      "step": 30790
    },
    {
      "entropy": 5.603287696838379,
      "epoch": 3.4585837030381312,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003843516661994559,
      "loss": 4.916,
      "mean_token_accuracy": 0.22179577499628067,
      "num_tokens": 68773878.0,
      "step": 30795
    },
    {
      "entropy": 5.5610016822814945,
      "epoch": 3.4591452799460884,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003843165762974167,
      "loss": 4.969,
      "mean_token_accuracy": 0.22652481347322465,
      "num_tokens": 68783864.0,
      "step": 30800
    },
    {
      "entropy": 5.570725202560425,
      "epoch": 3.459706856854046,
      "grad_norm": 1.125,
      "learning_rate": 0.00038428148291469515,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23039771467447281,
      "num_tokens": 68794280.0,
      "step": 30805
    },
    {
      "entropy": 5.682920694351196,
      "epoch": 3.4602684337620038,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000384246386052409,
      "loss": 5.0473,
      "mean_token_accuracy": 0.22190349996089936,
      "num_tokens": 68806125.0,
      "step": 30810
    },
    {
      "entropy": 5.635946559906006,
      "epoch": 3.4608300106699614,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003842112857116754,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2276395469903946,
      "num_tokens": 68816742.0,
      "step": 30815
    },
    {
      "entropy": 5.540281343460083,
      "epoch": 3.4613915875779186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003841761818936122,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22429234832525252,
      "num_tokens": 68828504.0,
      "step": 30820
    },
    {
      "entropy": 5.6991472244262695,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000384141074599337,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2171107292175293,
      "num_tokens": 68839993.0,
      "step": 30825
    },
    {
      "entropy": 5.545357275009155,
      "epoch": 3.462514741393834,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23282638490200042,
      "num_tokens": 68850395.0,
      "step": 30830
    },
    {
      "entropy": 5.566565370559692,
      "epoch": 3.463076318301791,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003840708495866221,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22498460859060287,
      "num_tokens": 68860621.0,
      "step": 30835
    },
    {
      "entropy": 5.573178052902222,
      "epoch": 3.463637895209749,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003840357318704183,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2211657866835594,
      "num_tokens": 68872838.0,
      "step": 30840
    },
    {
      "entropy": 5.602518749237061,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003840006106824746,
      "loss": 4.9273,
      "mean_token_accuracy": 0.23054139614105223,
      "num_tokens": 68883372.0,
      "step": 30845
    },
    {
      "entropy": 5.623771381378174,
      "epoch": 3.464761049025664,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003839654860239091,
      "loss": 4.9477,
      "mean_token_accuracy": 0.230891589820385,
      "num_tokens": 68894148.0,
      "step": 30850
    },
    {
      "entropy": 5.637095403671265,
      "epoch": 3.465322625933622,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003839303578958404,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2289642795920372,
      "num_tokens": 68904381.0,
      "step": 30855
    },
    {
      "entropy": 5.620836496353149,
      "epoch": 3.465884202841579,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038389522629938676,
      "loss": 5.0476,
      "mean_token_accuracy": 0.21868545114994048,
      "num_tokens": 68916354.0,
      "step": 30860
    },
    {
      "entropy": 5.585093688964844,
      "epoch": 3.4664457797495367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22926416099071503,
      "num_tokens": 68927551.0,
      "step": 30865
    },
    {
      "entropy": 5.691178703308106,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003838249527057995,
      "loss": 5.0431,
      "mean_token_accuracy": 0.22172333300113678,
      "num_tokens": 68939528.0,
      "step": 30870
    },
    {
      "entropy": 5.591048049926758,
      "epoch": 3.4675689335654516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038378981071090346,
      "loss": 4.8877,
      "mean_token_accuracy": 0.2297346994280815,
      "num_tokens": 68951634.0,
      "step": 30875
    },
    {
      "entropy": 5.616899394989014,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.9238,
      "mean_token_accuracy": 0.2344073623418808,
      "num_tokens": 68962259.0,
      "step": 30880
    },
    {
      "entropy": 5.603940486907959,
      "epoch": 3.468692087381367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038371951633050106,
      "loss": 5.0337,
      "mean_token_accuracy": 0.22616843730211258,
      "num_tokens": 68973185.0,
      "step": 30885
    },
    {
      "entropy": 5.671930932998658,
      "epoch": 3.4692536642893246,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038368436394723294,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21892376393079757,
      "num_tokens": 68984052.0,
      "step": 30890
    },
    {
      "entropy": 5.6203296184539795,
      "epoch": 3.469815241197282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038364920810341255,
      "loss": 5.0248,
      "mean_token_accuracy": 0.21819281429052353,
      "num_tokens": 68993730.0,
      "step": 30895
    },
    {
      "entropy": 5.675148296356201,
      "epoch": 3.4703768181052395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003836140488001592,
      "loss": 5.0029,
      "mean_token_accuracy": 0.22315287590026855,
      "num_tokens": 69006229.0,
      "step": 30900
    },
    {
      "entropy": 5.618323707580567,
      "epoch": 3.470938395013197,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038357888603859236,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2209508314728737,
      "num_tokens": 69018518.0,
      "step": 30905
    },
    {
      "entropy": 5.646619844436645,
      "epoch": 3.471499971921155,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038354371981983175,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21438881307840346,
      "num_tokens": 69031113.0,
      "step": 30910
    },
    {
      "entropy": 5.707032966613769,
      "epoch": 3.472061548829112,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003835085501449969,
      "loss": 5.1108,
      "mean_token_accuracy": 0.21590147018432618,
      "num_tokens": 69042038.0,
      "step": 30915
    },
    {
      "entropy": 5.535485649108887,
      "epoch": 3.4726231257370697,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038347337701520785,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23805686831474304,
      "num_tokens": 69053675.0,
      "step": 30920
    },
    {
      "entropy": 5.601852226257324,
      "epoch": 3.4731847026450273,
      "grad_norm": 1.125,
      "learning_rate": 0.00038343820043158436,
      "loss": 5.0161,
      "mean_token_accuracy": 0.22149735391139985,
      "num_tokens": 69064475.0,
      "step": 30925
    },
    {
      "entropy": 5.577740859985352,
      "epoch": 3.4737462795529845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003834030203952465,
      "loss": 4.955,
      "mean_token_accuracy": 0.2246198296546936,
      "num_tokens": 69076069.0,
      "step": 30930
    },
    {
      "entropy": 5.64802918434143,
      "epoch": 3.474307856460942,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003833678369073144,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2233104959130287,
      "num_tokens": 69087144.0,
      "step": 30935
    },
    {
      "entropy": 5.587260913848877,
      "epoch": 3.4748694333689,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22656745761632918,
      "num_tokens": 69098035.0,
      "step": 30940
    },
    {
      "entropy": 5.612680149078369,
      "epoch": 3.4754310102768575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003832974595811488,
      "loss": 4.9835,
      "mean_token_accuracy": 0.23262157738208772,
      "num_tokens": 69111722.0,
      "step": 30945
    },
    {
      "entropy": 5.628200721740723,
      "epoch": 3.475992587184815,
      "grad_norm": 1.109375,
      "learning_rate": 0.000383262265745156,
      "loss": 4.9012,
      "mean_token_accuracy": 0.2328459084033966,
      "num_tokens": 69122228.0,
      "step": 30950
    },
    {
      "entropy": 5.616391324996949,
      "epoch": 3.4765541640927724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038322706846205073,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22811250388622284,
      "num_tokens": 69134150.0,
      "step": 30955
    },
    {
      "entropy": 5.523842430114746,
      "epoch": 3.47711574100073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038319186773295364,
      "loss": 4.988,
      "mean_token_accuracy": 0.21889585852622986,
      "num_tokens": 69145757.0,
      "step": 30960
    },
    {
      "entropy": 5.677860736846924,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003831566635589856,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2274955078959465,
      "num_tokens": 69157545.0,
      "step": 30965
    },
    {
      "entropy": 5.623992538452148,
      "epoch": 3.478238894816645,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003831214559412674,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2221482902765274,
      "num_tokens": 69169070.0,
      "step": 30970
    },
    {
      "entropy": 5.5855793952941895,
      "epoch": 3.4788004717246026,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038308624488092005,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22870742827653884,
      "num_tokens": 69178903.0,
      "step": 30975
    },
    {
      "entropy": 5.52427191734314,
      "epoch": 3.4793620486325603,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003830510303790647,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2276802733540535,
      "num_tokens": 69190489.0,
      "step": 30980
    },
    {
      "entropy": 5.536361265182495,
      "epoch": 3.479923625540518,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038301581243682263,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22407276779413224,
      "num_tokens": 69200574.0,
      "step": 30985
    },
    {
      "entropy": 5.556664943695068,
      "epoch": 3.480485202448475,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.8813,
      "mean_token_accuracy": 0.23485807776451112,
      "num_tokens": 69210597.0,
      "step": 30990
    },
    {
      "entropy": 5.655705690383911,
      "epoch": 3.481046779356433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038294536623566407,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22046328634023665,
      "num_tokens": 69221392.0,
      "step": 30995
    },
    {
      "entropy": 5.504465341567993,
      "epoch": 3.4816083562643905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003829101379789905,
      "loss": 4.8805,
      "mean_token_accuracy": 0.21872581541538239,
      "num_tokens": 69232825.0,
      "step": 31000
    },
    {
      "entropy": 5.592964744567871,
      "epoch": 3.482169933172348,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.9579,
      "mean_token_accuracy": 0.22676152884960174,
      "num_tokens": 69243989.0,
      "step": 31005
    },
    {
      "entropy": 5.649304485321045,
      "epoch": 3.4827315100803053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003828396711590631,
      "loss": 5.029,
      "mean_token_accuracy": 0.21906811445951463,
      "num_tokens": 69255828.0,
      "step": 31010
    },
    {
      "entropy": 5.615948629379273,
      "epoch": 3.483293086988263,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038280443259805307,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22184234857559204,
      "num_tokens": 69266496.0,
      "step": 31015
    },
    {
      "entropy": 5.592926645278931,
      "epoch": 3.4838546638962207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003827691906045081,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21771771013736724,
      "num_tokens": 69277884.0,
      "step": 31020
    },
    {
      "entropy": 5.687379932403564,
      "epoch": 3.484416240804178,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003827339451795503,
      "loss": 5.1193,
      "mean_token_accuracy": 0.21911515891551972,
      "num_tokens": 69290660.0,
      "step": 31025
    },
    {
      "entropy": 5.725242328643799,
      "epoch": 3.4849778177121356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003826986963243018,
      "loss": 5.0866,
      "mean_token_accuracy": 0.2312595158815384,
      "num_tokens": 69302310.0,
      "step": 31030
    },
    {
      "entropy": 5.641945886611938,
      "epoch": 3.485539394620093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003826634440398852,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22722719460725785,
      "num_tokens": 69313804.0,
      "step": 31035
    },
    {
      "entropy": 5.575401163101196,
      "epoch": 3.486100971528051,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22185549587011338,
      "num_tokens": 69325372.0,
      "step": 31040
    },
    {
      "entropy": 5.578647232055664,
      "epoch": 3.4866625484360085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038259292918803696,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2193128615617752,
      "num_tokens": 69335835.0,
      "step": 31045
    },
    {
      "entropy": 5.576835346221924,
      "epoch": 3.4872241253439658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003825576666228506,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22088015377521514,
      "num_tokens": 69346895.0,
      "step": 31050
    },
    {
      "entropy": 5.587823057174683,
      "epoch": 3.4877857022519234,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003825224006329864,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22548305690288545,
      "num_tokens": 69357649.0,
      "step": 31055
    },
    {
      "entropy": 5.650987005233764,
      "epoch": 3.488347279159881,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003824871312195673,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22424115240573883,
      "num_tokens": 69367973.0,
      "step": 31060
    },
    {
      "entropy": 5.653545904159546,
      "epoch": 3.4889088560678383,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003824518583837163,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23377825170755387,
      "num_tokens": 69379411.0,
      "step": 31065
    },
    {
      "entropy": 5.660772275924683,
      "epoch": 3.489470432975796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003824165821265564,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2284015715122223,
      "num_tokens": 69391261.0,
      "step": 31070
    },
    {
      "entropy": 5.548551893234253,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22847407460212707,
      "num_tokens": 69401712.0,
      "step": 31075
    },
    {
      "entropy": 5.4873895168304445,
      "epoch": 3.4905935867917113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000382346019352803,
      "loss": 4.8828,
      "mean_token_accuracy": 0.2414736732840538,
      "num_tokens": 69412612.0,
      "step": 31080
    },
    {
      "entropy": 5.633964014053345,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003823107328384562,
      "loss": 5.0147,
      "mean_token_accuracy": 0.22693875133991243,
      "num_tokens": 69424042.0,
      "step": 31085
    },
    {
      "entropy": 5.630054140090943,
      "epoch": 3.491716740607626,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003822754429072942,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22821305990219115,
      "num_tokens": 69434667.0,
      "step": 31090
    },
    {
      "entropy": 5.667054510116577,
      "epoch": 3.492278317515584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003822401495604403,
      "loss": 5.0584,
      "mean_token_accuracy": 0.22180386036634445,
      "num_tokens": 69447015.0,
      "step": 31095
    },
    {
      "entropy": 5.557612133026123,
      "epoch": 3.4928398944235415,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038220485279901855,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22073886096477507,
      "num_tokens": 69457585.0,
      "step": 31100
    },
    {
      "entropy": 5.595785188674927,
      "epoch": 3.4934014713314987,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038216955262415273,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21562528908252715,
      "num_tokens": 69467751.0,
      "step": 31105
    },
    {
      "entropy": 5.682295513153076,
      "epoch": 3.4939630482394564,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038213424903696673,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21974599957466126,
      "num_tokens": 69478898.0,
      "step": 31110
    },
    {
      "entropy": 5.663791227340698,
      "epoch": 3.494524625147414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038209894203858477,
      "loss": 5.0127,
      "mean_token_accuracy": 0.2188984602689743,
      "num_tokens": 69488367.0,
      "step": 31115
    },
    {
      "entropy": 5.5961815357208256,
      "epoch": 3.4950862020553712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038206363163013093,
      "loss": 4.9911,
      "mean_token_accuracy": 0.22627557665109635,
      "num_tokens": 69500446.0,
      "step": 31120
    },
    {
      "entropy": 5.625974035263061,
      "epoch": 3.495647778963329,
      "grad_norm": 1.0,
      "learning_rate": 0.00038202831781272944,
      "loss": 5.0106,
      "mean_token_accuracy": 0.22420797497034073,
      "num_tokens": 69512383.0,
      "step": 31125
    },
    {
      "entropy": 5.597306776046753,
      "epoch": 3.4962093558712866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2286100134253502,
      "num_tokens": 69523238.0,
      "step": 31130
    },
    {
      "entropy": 5.744542074203491,
      "epoch": 3.4967709327792442,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038195767995558174,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21973237842321397,
      "num_tokens": 69535196.0,
      "step": 31135
    },
    {
      "entropy": 5.638227891921997,
      "epoch": 3.497332509687202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22537333965301515,
      "num_tokens": 69546896.0,
      "step": 31140
    },
    {
      "entropy": 5.561940908432007,
      "epoch": 3.497894086595159,
      "grad_norm": 1.09375,
      "learning_rate": 0.000381887028476138,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22727818191051483,
      "num_tokens": 69558230.0,
      "step": 31145
    },
    {
      "entropy": 5.515419435501099,
      "epoch": 3.4984556635031168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038185169763086703,
      "loss": 4.9339,
      "mean_token_accuracy": 0.23084830790758132,
      "num_tokens": 69568891.0,
      "step": 31150
    },
    {
      "entropy": 5.620729064941406,
      "epoch": 3.4990172404110744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003818163633833966,
      "loss": 5.0201,
      "mean_token_accuracy": 0.22264616340398788,
      "num_tokens": 69579988.0,
      "step": 31155
    },
    {
      "entropy": 5.626387691497802,
      "epoch": 3.4995788173190316,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22579900920391083,
      "num_tokens": 69590714.0,
      "step": 31160
    },
    {
      "entropy": 5.564715480804443,
      "epoch": 3.5001403942269893,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003817456846863575,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23260865956544877,
      "num_tokens": 69601755.0,
      "step": 31165
    },
    {
      "entropy": 5.542750549316406,
      "epoch": 3.500701971134947,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22923919260501863,
      "num_tokens": 69612862.0,
      "step": 31170
    },
    {
      "entropy": 5.694656467437744,
      "epoch": 3.5012635480429046,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003816749923940226,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2226084813475609,
      "num_tokens": 69624746.0,
      "step": 31175
    },
    {
      "entropy": 5.5727287292480465,
      "epoch": 3.5018251249508623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038163964115243276,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23264790326356888,
      "num_tokens": 69634477.0,
      "step": 31180
    },
    {
      "entropy": 5.596529626846314,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038160428651539535,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2196561276912689,
      "num_tokens": 69645913.0,
      "step": 31185
    },
    {
      "entropy": 5.582387113571167,
      "epoch": 3.502948278766777,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038156892848403606,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2244902342557907,
      "num_tokens": 69657727.0,
      "step": 31190
    },
    {
      "entropy": 5.503789186477661,
      "epoch": 3.503509855674735,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003815335670594808,
      "loss": 4.8187,
      "mean_token_accuracy": 0.25251021683216096,
      "num_tokens": 69668649.0,
      "step": 31195
    },
    {
      "entropy": 5.603407096862793,
      "epoch": 3.504071432582692,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038149820224285546,
      "loss": 5.027,
      "mean_token_accuracy": 0.2208692952990532,
      "num_tokens": 69680286.0,
      "step": 31200
    },
    {
      "entropy": 5.602105045318604,
      "epoch": 3.5046330094906497,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000381462834035286,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22843061685562133,
      "num_tokens": 69689801.0,
      "step": 31205
    },
    {
      "entropy": 5.626051425933838,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.954,
      "mean_token_accuracy": 0.23021032065153121,
      "num_tokens": 69700898.0,
      "step": 31210
    },
    {
      "entropy": 5.589344120025634,
      "epoch": 3.5057561633065646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.9055,
      "mean_token_accuracy": 0.23139309138059616,
      "num_tokens": 69711969.0,
      "step": 31215
    },
    {
      "entropy": 5.593577337265015,
      "epoch": 3.5063177402145222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003813567090781752,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22582054734230042,
      "num_tokens": 69723458.0,
      "step": 31220
    },
    {
      "entropy": 5.547695970535278,
      "epoch": 3.50687931712248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000381321327318092,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23070870637893676,
      "num_tokens": 69735527.0,
      "step": 31225
    },
    {
      "entropy": 5.658506202697754,
      "epoch": 3.5074408940304376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003812859421726964,
      "loss": 5.039,
      "mean_token_accuracy": 0.22916138917207718,
      "num_tokens": 69748829.0,
      "step": 31230
    },
    {
      "entropy": 5.637313747406006,
      "epoch": 3.5080024709383952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038125055364311526,
      "loss": 4.851,
      "mean_token_accuracy": 0.22989099472761154,
      "num_tokens": 69758970.0,
      "step": 31235
    },
    {
      "entropy": 5.629860830307007,
      "epoch": 3.5085640478463525,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003812151617304753,
      "loss": 5.017,
      "mean_token_accuracy": 0.2227175384759903,
      "num_tokens": 69770082.0,
      "step": 31240
    },
    {
      "entropy": 5.581304597854614,
      "epoch": 3.50912562475431,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22376695573329924,
      "num_tokens": 69781386.0,
      "step": 31245
    },
    {
      "entropy": 5.651660633087158,
      "epoch": 3.5096872016622678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038114436776052644,
      "loss": 4.9969,
      "mean_token_accuracy": 0.2219899296760559,
      "num_tokens": 69792571.0,
      "step": 31250
    },
    {
      "entropy": 5.659713315963745,
      "epoch": 3.510248778570225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003811089657054717,
      "loss": 5.0138,
      "mean_token_accuracy": 0.22699073404073716,
      "num_tokens": 69804737.0,
      "step": 31255
    },
    {
      "entropy": 5.636247205734253,
      "epoch": 3.5108103554781827,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2279593288898468,
      "num_tokens": 69816584.0,
      "step": 31260
    },
    {
      "entropy": 5.532318067550659,
      "epoch": 3.5113719323861403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003810381514608378,
      "loss": 4.8242,
      "mean_token_accuracy": 0.24673908501863479,
      "num_tokens": 69828657.0,
      "step": 31265
    },
    {
      "entropy": 5.6084624290466305,
      "epoch": 3.511933509294098,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003810027392735133,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2278391480445862,
      "num_tokens": 69839780.0,
      "step": 31270
    },
    {
      "entropy": 5.597680139541626,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003809673237110204,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22566191703081132,
      "num_tokens": 69850452.0,
      "step": 31275
    },
    {
      "entropy": 5.609313440322876,
      "epoch": 3.513056663110013,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003809319047744869,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22245176881551743,
      "num_tokens": 69861840.0,
      "step": 31280
    },
    {
      "entropy": 5.676811742782593,
      "epoch": 3.5136182400179705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2295298308134079,
      "num_tokens": 69872653.0,
      "step": 31285
    },
    {
      "entropy": 5.60264835357666,
      "epoch": 3.514179816925928,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038086105678380896,
      "loss": 4.999,
      "mean_token_accuracy": 0.22100569754838945,
      "num_tokens": 69883381.0,
      "step": 31290
    },
    {
      "entropy": 5.70157904624939,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038082562773192026,
      "loss": 5.06,
      "mean_token_accuracy": 0.22065893113613128,
      "num_tokens": 69894732.0,
      "step": 31295
    },
    {
      "entropy": 5.710716342926025,
      "epoch": 3.515302970741843,
      "grad_norm": 1.125,
      "learning_rate": 0.00038079019531050257,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22541242688894272,
      "num_tokens": 69905645.0,
      "step": 31300
    },
    {
      "entropy": 5.683207225799561,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038075475952068405,
      "loss": 4.988,
      "mean_token_accuracy": 0.21811397969722748,
      "num_tokens": 69917176.0,
      "step": 31305
    },
    {
      "entropy": 5.61809983253479,
      "epoch": 3.516426124557758,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000380719320363593,
      "loss": 5.0609,
      "mean_token_accuracy": 0.226804219186306,
      "num_tokens": 69928923.0,
      "step": 31310
    },
    {
      "entropy": 5.557211923599243,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038068387784035774,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21887889206409455,
      "num_tokens": 69938981.0,
      "step": 31315
    },
    {
      "entropy": 5.589187669754028,
      "epoch": 3.5175492783736733,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003806484319521069,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22526101917028427,
      "num_tokens": 69949265.0,
      "step": 31320
    },
    {
      "entropy": 5.535823106765747,
      "epoch": 3.518110855281631,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038061298269996893,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23370217233896257,
      "num_tokens": 69961399.0,
      "step": 31325
    },
    {
      "entropy": 5.608912944793701,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.957,
      "mean_token_accuracy": 0.22106441110372543,
      "num_tokens": 69971601.0,
      "step": 31330
    },
    {
      "entropy": 5.5562280178070065,
      "epoch": 3.519234009097546,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003805420741085469,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2150966003537178,
      "num_tokens": 69982375.0,
      "step": 31335
    },
    {
      "entropy": 5.600587558746338,
      "epoch": 3.5197955860055035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003805066147715205,
      "loss": 4.9522,
      "mean_token_accuracy": 0.23036173731088638,
      "num_tokens": 69994787.0,
      "step": 31340
    },
    {
      "entropy": 5.722999858856201,
      "epoch": 3.520357162913461,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003804711520751226,
      "loss": 5.1396,
      "mean_token_accuracy": 0.2152253806591034,
      "num_tokens": 70006077.0,
      "step": 31345
    },
    {
      "entropy": 5.632815265655518,
      "epoch": 3.5209187398214183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038043568602048227,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22259599417448045,
      "num_tokens": 70016655.0,
      "step": 31350
    },
    {
      "entropy": 5.613462305068969,
      "epoch": 3.521480316729376,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003804002166087289,
      "loss": 4.8867,
      "mean_token_accuracy": 0.22586303502321242,
      "num_tokens": 70026857.0,
      "step": 31355
    },
    {
      "entropy": 5.590369606018067,
      "epoch": 3.5220418936373337,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.9382,
      "mean_token_accuracy": 0.22736468017101288,
      "num_tokens": 70038311.0,
      "step": 31360
    },
    {
      "entropy": 5.582656145095825,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.125,
      "learning_rate": 0.0003803292677184002,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22841569632291794,
      "num_tokens": 70048613.0,
      "step": 31365
    },
    {
      "entropy": 5.698272848129273,
      "epoch": 3.523165047453249,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038029378824208394,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22869995534420012,
      "num_tokens": 70058686.0,
      "step": 31370
    },
    {
      "entropy": 5.663999319076538,
      "epoch": 3.523726624361206,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003802583054131727,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22415636628866195,
      "num_tokens": 70068845.0,
      "step": 31375
    },
    {
      "entropy": 5.531022310256958,
      "epoch": 3.524288201269164,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003802228192327961,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22126260697841643,
      "num_tokens": 70080223.0,
      "step": 31380
    },
    {
      "entropy": 5.615438175201416,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038018732970208414,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21958295106887818,
      "num_tokens": 70091729.0,
      "step": 31385
    },
    {
      "entropy": 5.599906778335571,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003801518368221668,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2195892482995987,
      "num_tokens": 70103172.0,
      "step": 31390
    },
    {
      "entropy": 5.655094718933105,
      "epoch": 3.5259729319930364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003801163405941743,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21849222630262374,
      "num_tokens": 70114303.0,
      "step": 31395
    },
    {
      "entropy": 5.573291778564453,
      "epoch": 3.526534508900994,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038008084101923665,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22513729482889175,
      "num_tokens": 70126802.0,
      "step": 31400
    },
    {
      "entropy": 5.706686544418335,
      "epoch": 3.5270960858089513,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038004533809848435,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22818758487701415,
      "num_tokens": 70138768.0,
      "step": 31405
    },
    {
      "entropy": 5.5518388748168945,
      "epoch": 3.527657662716909,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038000983183304765,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22707962095737458,
      "num_tokens": 70148411.0,
      "step": 31410
    },
    {
      "entropy": 5.499326992034912,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003799743222240572,
      "loss": 4.913,
      "mean_token_accuracy": 0.22991109937429427,
      "num_tokens": 70159788.0,
      "step": 31415
    },
    {
      "entropy": 5.603577184677124,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003799388092726436,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22422004640102386,
      "num_tokens": 70169707.0,
      "step": 31420
    },
    {
      "entropy": 5.5953291893005375,
      "epoch": 3.529342393440782,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037990329297993763,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2283357337117195,
      "num_tokens": 70180958.0,
      "step": 31425
    },
    {
      "entropy": 5.543913269042969,
      "epoch": 3.529903970348739,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.916,
      "mean_token_accuracy": 0.22122438997030258,
      "num_tokens": 70192848.0,
      "step": 31430
    },
    {
      "entropy": 5.5620276927948,
      "epoch": 3.530465547256697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037983225037517207,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22521841078996657,
      "num_tokens": 70204034.0,
      "step": 31435
    },
    {
      "entropy": 5.677851629257202,
      "epoch": 3.5310271241646545,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037979672406537445,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22261106520891188,
      "num_tokens": 70214274.0,
      "step": 31440
    },
    {
      "entropy": 5.668095684051513,
      "epoch": 3.5315887010726117,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003797611944188086,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22275160253047943,
      "num_tokens": 70223890.0,
      "step": 31445
    },
    {
      "entropy": 5.655187129974365,
      "epoch": 3.5321502779805694,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037972566143660566,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22041944563388824,
      "num_tokens": 70234529.0,
      "step": 31450
    },
    {
      "entropy": 5.53455924987793,
      "epoch": 3.532711854888527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003796901251198969,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22813289910554885,
      "num_tokens": 70246900.0,
      "step": 31455
    },
    {
      "entropy": 5.695550203323364,
      "epoch": 3.5332734317964847,
      "grad_norm": 1.171875,
      "learning_rate": 0.000379654585469814,
      "loss": 5.0934,
      "mean_token_accuracy": 0.21934499591588974,
      "num_tokens": 70257942.0,
      "step": 31460
    },
    {
      "entropy": 5.680350494384766,
      "epoch": 3.5338350087044423,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003796190424874886,
      "loss": 5.0326,
      "mean_token_accuracy": 0.22442620992660522,
      "num_tokens": 70270297.0,
      "step": 31465
    },
    {
      "entropy": 5.655364561080932,
      "epoch": 3.5343965856123996,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003795834961740523,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2182549551129341,
      "num_tokens": 70281542.0,
      "step": 31470
    },
    {
      "entropy": 5.616425895690918,
      "epoch": 3.534958162520357,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037954794653063695,
      "loss": 4.9481,
      "mean_token_accuracy": 0.220279261469841,
      "num_tokens": 70290957.0,
      "step": 31475
    },
    {
      "entropy": 5.63186354637146,
      "epoch": 3.535519739428315,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003795123935583743,
      "loss": 4.983,
      "mean_token_accuracy": 0.23349702954292298,
      "num_tokens": 70302796.0,
      "step": 31480
    },
    {
      "entropy": 5.576093912124634,
      "epoch": 3.536081316336272,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003794768372583966,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2348925292491913,
      "num_tokens": 70313709.0,
      "step": 31485
    },
    {
      "entropy": 5.624137020111084,
      "epoch": 3.5366428932442298,
      "grad_norm": 1.125,
      "learning_rate": 0.00037944127763183587,
      "loss": 5.0047,
      "mean_token_accuracy": 0.224055677652359,
      "num_tokens": 70324651.0,
      "step": 31490
    },
    {
      "entropy": 5.704397106170655,
      "epoch": 3.5372044701521874,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037940571467982436,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21710876226425171,
      "num_tokens": 70335810.0,
      "step": 31495
    },
    {
      "entropy": 5.603557348251343,
      "epoch": 3.5377660470601446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.914,
      "mean_token_accuracy": 0.23150250613689421,
      "num_tokens": 70347171.0,
      "step": 31500
    },
    {
      "entropy": 5.583105087280273,
      "epoch": 3.5383276239681023,
      "grad_norm": 1.0,
      "learning_rate": 0.00037933457880397843,
      "loss": 4.861,
      "mean_token_accuracy": 0.2343127653002739,
      "num_tokens": 70359122.0,
      "step": 31505
    },
    {
      "entropy": 5.463804149627686,
      "epoch": 3.53888920087606,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22881662845611572,
      "num_tokens": 70369691.0,
      "step": 31510
    },
    {
      "entropy": 5.586231422424317,
      "epoch": 3.5394507777840176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003792634296399187,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22469241619110109,
      "num_tokens": 70379602.0,
      "step": 31515
    },
    {
      "entropy": 5.630769538879394,
      "epoch": 3.5400123546919753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037922785007764043,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2182932198047638,
      "num_tokens": 70391819.0,
      "step": 31520
    },
    {
      "entropy": 5.701958847045899,
      "epoch": 3.5405739315999325,
      "grad_norm": 1.171875,
      "learning_rate": 0.000379192267196707,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2243003949522972,
      "num_tokens": 70402078.0,
      "step": 31525
    },
    {
      "entropy": 5.555619144439698,
      "epoch": 3.54113550850789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037915668099825135,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22388084083795548,
      "num_tokens": 70413829.0,
      "step": 31530
    },
    {
      "entropy": 5.458155918121338,
      "epoch": 3.541697085415848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037912109148340663,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2247252956032753,
      "num_tokens": 70426221.0,
      "step": 31535
    },
    {
      "entropy": 5.697107172012329,
      "epoch": 3.542258662323805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037908549865330596,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21692131459712982,
      "num_tokens": 70437215.0,
      "step": 31540
    },
    {
      "entropy": 5.639275550842285,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22930872440338135,
      "num_tokens": 70448918.0,
      "step": 31545
    },
    {
      "entropy": 5.599945878982544,
      "epoch": 3.5433818161397204,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037901430305187,
      "loss": 4.9991,
      "mean_token_accuracy": 0.22276908457279204,
      "num_tokens": 70460713.0,
      "step": 31550
    },
    {
      "entropy": 5.579898834228516,
      "epoch": 3.543943393047678,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22252333611249925,
      "num_tokens": 70471460.0,
      "step": 31555
    },
    {
      "entropy": 5.603148508071899,
      "epoch": 3.5445049699556357,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003789430942030112,
      "loss": 4.978,
      "mean_token_accuracy": 0.22467345446348191,
      "num_tokens": 70482359.0,
      "step": 31560
    },
    {
      "entropy": 5.666013097763061,
      "epoch": 3.545066546863593,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037890748481363216,
      "loss": 5.0273,
      "mean_token_accuracy": 0.22005094587802887,
      "num_tokens": 70493272.0,
      "step": 31565
    },
    {
      "entropy": 5.609143400192261,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003788718721157987,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22742524296045302,
      "num_tokens": 70504118.0,
      "step": 31570
    },
    {
      "entropy": 5.681338214874268,
      "epoch": 3.5461897006795082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003788362561106445,
      "loss": 5.0725,
      "mean_token_accuracy": 0.2187308743596077,
      "num_tokens": 70515379.0,
      "step": 31575
    },
    {
      "entropy": 5.643668842315674,
      "epoch": 3.5467512775874654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037880063679930355,
      "loss": 4.9738,
      "mean_token_accuracy": 0.2265568420290947,
      "num_tokens": 70526047.0,
      "step": 31580
    },
    {
      "entropy": 5.643024587631226,
      "epoch": 3.547312854495423,
      "grad_norm": 1.125,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.9649,
      "mean_token_accuracy": 0.23068373203277587,
      "num_tokens": 70537614.0,
      "step": 31585
    },
    {
      "entropy": 5.602196836471558,
      "epoch": 3.5478744314033808,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003787293882625984,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22613541334867476,
      "num_tokens": 70548048.0,
      "step": 31590
    },
    {
      "entropy": 5.564011669158935,
      "epoch": 3.548436008311338,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.8676,
      "mean_token_accuracy": 0.234664124250412,
      "num_tokens": 70558668.0,
      "step": 31595
    },
    {
      "entropy": 5.61557092666626,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003786581265147575,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21754308491945268,
      "num_tokens": 70570044.0,
      "step": 31600
    },
    {
      "entropy": 5.608210849761963,
      "epoch": 3.5495591621272533,
      "grad_norm": 1.125,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.8673,
      "mean_token_accuracy": 0.23167943060398102,
      "num_tokens": 70580857.0,
      "step": 31605
    },
    {
      "entropy": 5.552534294128418,
      "epoch": 3.550120739035211,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000378586851564857,
      "loss": 4.8391,
      "mean_token_accuracy": 0.2335154339671135,
      "num_tokens": 70592476.0,
      "step": 31610
    },
    {
      "entropy": 5.5732933521270756,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000378551209141971,
      "loss": 4.922,
      "mean_token_accuracy": 0.23038065284490586,
      "num_tokens": 70603614.0,
      "step": 31615
    },
    {
      "entropy": 5.637519979476929,
      "epoch": 3.551243892851126,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003785155634219744,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21290383934974672,
      "num_tokens": 70615239.0,
      "step": 31620
    },
    {
      "entropy": 5.653652667999268,
      "epoch": 3.5518054697590835,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003784799144060021,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2246955707669258,
      "num_tokens": 70625785.0,
      "step": 31625
    },
    {
      "entropy": 5.663790655136109,
      "epoch": 3.552367046667041,
      "grad_norm": 1.125,
      "learning_rate": 0.00037844426209518916,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22274963557720184,
      "num_tokens": 70636942.0,
      "step": 31630
    },
    {
      "entropy": 5.571205472946167,
      "epoch": 3.5529286235749984,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037840860649067084,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2275235667824745,
      "num_tokens": 70648106.0,
      "step": 31635
    },
    {
      "entropy": 5.585953092575073,
      "epoch": 3.553490200482956,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037837294759358235,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22491390854120255,
      "num_tokens": 70658850.0,
      "step": 31640
    },
    {
      "entropy": 5.735801696777344,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037833728540505914,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21594427824020385,
      "num_tokens": 70671060.0,
      "step": 31645
    },
    {
      "entropy": 5.576700830459595,
      "epoch": 3.5546133542988714,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003783016199262367,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23682533502578734,
      "num_tokens": 70682186.0,
      "step": 31650
    },
    {
      "entropy": 5.592959880828857,
      "epoch": 3.555174931206829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003782659511582505,
      "loss": 4.934,
      "mean_token_accuracy": 0.22254671305418014,
      "num_tokens": 70692622.0,
      "step": 31655
    },
    {
      "entropy": 5.529252529144287,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003782302791022365,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2330821320414543,
      "num_tokens": 70703293.0,
      "step": 31660
    },
    {
      "entropy": 5.510266590118408,
      "epoch": 3.556298085022744,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037819460375933034,
      "loss": 4.8498,
      "mean_token_accuracy": 0.233161760866642,
      "num_tokens": 70714961.0,
      "step": 31665
    },
    {
      "entropy": 5.674548721313476,
      "epoch": 3.5568596619307016,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003781589251306679,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2219880238175392,
      "num_tokens": 70727030.0,
      "step": 31670
    },
    {
      "entropy": 5.671895360946655,
      "epoch": 3.557421238838659,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003781232432173853,
      "loss": 4.941,
      "mean_token_accuracy": 0.22403909415006637,
      "num_tokens": 70738331.0,
      "step": 31675
    },
    {
      "entropy": 5.629755020141602,
      "epoch": 3.5579828157466165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037808755802061874,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21832553595304488,
      "num_tokens": 70748347.0,
      "step": 31680
    },
    {
      "entropy": 5.647200393676758,
      "epoch": 3.558544392654574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003780518695415043,
      "loss": 4.9875,
      "mean_token_accuracy": 0.23210833966732025,
      "num_tokens": 70759905.0,
      "step": 31685
    },
    {
      "entropy": 5.747242021560669,
      "epoch": 3.5591059695625313,
      "grad_norm": 1.125,
      "learning_rate": 0.00037801617778117835,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2233905702829361,
      "num_tokens": 70770538.0,
      "step": 31690
    },
    {
      "entropy": 5.6275880336761475,
      "epoch": 3.559667546470489,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037798048274077735,
      "loss": 5.0308,
      "mean_token_accuracy": 0.22350818961858748,
      "num_tokens": 70781881.0,
      "step": 31695
    },
    {
      "entropy": 5.517162752151489,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22986642867326737,
      "num_tokens": 70792299.0,
      "step": 31700
    },
    {
      "entropy": 5.6115865230560305,
      "epoch": 3.5607907002864043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037790908282429647,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2255768358707428,
      "num_tokens": 70802470.0,
      "step": 31705
    },
    {
      "entropy": 5.6694659233093265,
      "epoch": 3.561352277194362,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037787337795049,
      "loss": 5.0482,
      "mean_token_accuracy": 0.22447729408740996,
      "num_tokens": 70813502.0,
      "step": 31710
    },
    {
      "entropy": 5.6248291015625,
      "epoch": 3.561913854102319,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037783766980115525,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22986246049404144,
      "num_tokens": 70823917.0,
      "step": 31715
    },
    {
      "entropy": 5.646764087677002,
      "epoch": 3.562475431010277,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037780195837742927,
      "loss": 5.055,
      "mean_token_accuracy": 0.21806555837392808,
      "num_tokens": 70834503.0,
      "step": 31720
    },
    {
      "entropy": 5.632489395141602,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.9234,
      "mean_token_accuracy": 0.22751325815916063,
      "num_tokens": 70844954.0,
      "step": 31725
    },
    {
      "entropy": 5.667081356048584,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037773052571135157,
      "loss": 5.0681,
      "mean_token_accuracy": 0.21936519593000411,
      "num_tokens": 70856142.0,
      "step": 31730
    },
    {
      "entropy": 5.615232753753662,
      "epoch": 3.5641601617341494,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003776948044712745,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22542186677455903,
      "num_tokens": 70866880.0,
      "step": 31735
    },
    {
      "entropy": 5.617778015136719,
      "epoch": 3.564721738642107,
      "grad_norm": 1.234375,
      "learning_rate": 0.000377659079961355,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23312271237373353,
      "num_tokens": 70876836.0,
      "step": 31740
    },
    {
      "entropy": 5.596471118927002,
      "epoch": 3.5652833155500647,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037762335218273057,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22692760080099106,
      "num_tokens": 70889430.0,
      "step": 31745
    },
    {
      "entropy": 5.560275268554688,
      "epoch": 3.5658448924580224,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003775876211365388,
      "loss": 4.867,
      "mean_token_accuracy": 0.23141802698373795,
      "num_tokens": 70899291.0,
      "step": 31750
    },
    {
      "entropy": 5.6284722805023195,
      "epoch": 3.5664064693659796,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003775518868239174,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22342922240495683,
      "num_tokens": 70910596.0,
      "step": 31755
    },
    {
      "entropy": 5.647970485687256,
      "epoch": 3.5669680462739373,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037751614924600417,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2313631385564804,
      "num_tokens": 70921969.0,
      "step": 31760
    },
    {
      "entropy": 5.618114566802978,
      "epoch": 3.567529623181895,
      "grad_norm": 1.140625,
      "learning_rate": 0.000377480408403937,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22113393992185593,
      "num_tokens": 70933372.0,
      "step": 31765
    },
    {
      "entropy": 5.695416069030761,
      "epoch": 3.568091200089852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.0805,
      "mean_token_accuracy": 0.21857646256685256,
      "num_tokens": 70944989.0,
      "step": 31770
    },
    {
      "entropy": 5.606934213638306,
      "epoch": 3.56865277699781,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037740891693189286,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21884279549121857,
      "num_tokens": 70957516.0,
      "step": 31775
    },
    {
      "entropy": 5.682017993927002,
      "epoch": 3.5692143539057675,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003773731663041922,
      "loss": 5.0393,
      "mean_token_accuracy": 0.21971554905176163,
      "num_tokens": 70969485.0,
      "step": 31780
    },
    {
      "entropy": 5.625828838348388,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037733741241689024,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2266034871339798,
      "num_tokens": 70980335.0,
      "step": 31785
    },
    {
      "entropy": 5.556629657745361,
      "epoch": 3.5703375077216823,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003773016552711254,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2282062739133835,
      "num_tokens": 70991870.0,
      "step": 31790
    },
    {
      "entropy": 5.6298002243042,
      "epoch": 3.57089908462964,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037726589486803617,
      "loss": 5.0664,
      "mean_token_accuracy": 0.22207688838243483,
      "num_tokens": 71003906.0,
      "step": 31795
    },
    {
      "entropy": 5.57889313697815,
      "epoch": 3.5714606615375977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037723013120876113,
      "loss": 4.932,
      "mean_token_accuracy": 0.22229187190532684,
      "num_tokens": 71015280.0,
      "step": 31800
    },
    {
      "entropy": 5.663444757461548,
      "epoch": 3.5720222384455553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037719436429443915,
      "loss": 4.9933,
      "mean_token_accuracy": 0.22944018691778184,
      "num_tokens": 71026299.0,
      "step": 31805
    },
    {
      "entropy": 5.711491298675537,
      "epoch": 3.5725838153535125,
      "grad_norm": 1.125,
      "learning_rate": 0.00037715859412620895,
      "loss": 5.0406,
      "mean_token_accuracy": 0.22367714643478392,
      "num_tokens": 71037704.0,
      "step": 31810
    },
    {
      "entropy": 5.624472427368164,
      "epoch": 3.57314539226147,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003771228207052095,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22842941284179688,
      "num_tokens": 71048359.0,
      "step": 31815
    },
    {
      "entropy": 5.589843845367431,
      "epoch": 3.573706969169428,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037708704403257974,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22712651491165162,
      "num_tokens": 71059723.0,
      "step": 31820
    },
    {
      "entropy": 5.541415214538574,
      "epoch": 3.574268546077385,
      "grad_norm": 1.078125,
      "learning_rate": 0.000377051264109459,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23508215695619583,
      "num_tokens": 71071437.0,
      "step": 31825
    },
    {
      "entropy": 5.55882773399353,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.8263,
      "mean_token_accuracy": 0.24070803076028824,
      "num_tokens": 71082544.0,
      "step": 31830
    },
    {
      "entropy": 5.599774551391602,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037697969451630135,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22746757715940474,
      "num_tokens": 71092429.0,
      "step": 31835
    },
    {
      "entropy": 5.564616441726685,
      "epoch": 3.575953276801258,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037694390484854323,
      "loss": 4.9695,
      "mean_token_accuracy": 0.2225315809249878,
      "num_tokens": 71103154.0,
      "step": 31840
    },
    {
      "entropy": 5.58641791343689,
      "epoch": 3.5765148537092157,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.8993,
      "mean_token_accuracy": 0.23438135981559755,
      "num_tokens": 71112921.0,
      "step": 31845
    },
    {
      "entropy": 5.641908311843872,
      "epoch": 3.577076430617173,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037687231577636626,
      "loss": 4.9604,
      "mean_token_accuracy": 0.23080647438764573,
      "num_tokens": 71123281.0,
      "step": 31850
    },
    {
      "entropy": 5.676341867446899,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.085,
      "mean_token_accuracy": 0.2168935775756836,
      "num_tokens": 71134358.0,
      "step": 31855
    },
    {
      "entropy": 5.664652252197266,
      "epoch": 3.5781995844330883,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037680071372957327,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22028670758008956,
      "num_tokens": 71146805.0,
      "step": 31860
    },
    {
      "entropy": 5.616459560394287,
      "epoch": 3.5787611613410455,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037676490784354543,
      "loss": 4.9418,
      "mean_token_accuracy": 0.2229050874710083,
      "num_tokens": 71157388.0,
      "step": 31865
    },
    {
      "entropy": 5.477285480499267,
      "epoch": 3.579322738249003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003767290987172835,
      "loss": 4.7871,
      "mean_token_accuracy": 0.23536480516195296,
      "num_tokens": 71168601.0,
      "step": 31870
    },
    {
      "entropy": 5.590088510513306,
      "epoch": 3.579884315156961,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037669328635192757,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22727326899766923,
      "num_tokens": 71180796.0,
      "step": 31875
    },
    {
      "entropy": 5.5983551979064945,
      "epoch": 3.580445892064918,
      "grad_norm": 1.171875,
      "learning_rate": 0.000376657470748618,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2308433473110199,
      "num_tokens": 71191686.0,
      "step": 31880
    },
    {
      "entropy": 5.554427003860473,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.7916,
      "mean_token_accuracy": 0.23404247760772706,
      "num_tokens": 71202454.0,
      "step": 31885
    },
    {
      "entropy": 5.632414102554321,
      "epoch": 3.5815690458808334,
      "grad_norm": 1.125,
      "learning_rate": 0.0003765858298326994,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2345498636364937,
      "num_tokens": 71213729.0,
      "step": 31890
    },
    {
      "entropy": 5.7281430721282955,
      "epoch": 3.582130622788791,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037655000452237154,
      "loss": 5.1107,
      "mean_token_accuracy": 0.22076607793569564,
      "num_tokens": 71226810.0,
      "step": 31895
    },
    {
      "entropy": 5.577607583999634,
      "epoch": 3.5826921996967487,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003765141759786521,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2339457869529724,
      "num_tokens": 71237233.0,
      "step": 31900
    },
    {
      "entropy": 5.624127006530761,
      "epoch": 3.583253776604706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037647834420268195,
      "loss": 4.9438,
      "mean_token_accuracy": 0.236861552298069,
      "num_tokens": 71248678.0,
      "step": 31905
    },
    {
      "entropy": 5.493608188629151,
      "epoch": 3.5838153535126636,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23382886201143266,
      "num_tokens": 71259098.0,
      "step": 31910
    },
    {
      "entropy": 5.558871793746948,
      "epoch": 3.584376930420621,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037640667095855307,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23506172150373458,
      "num_tokens": 71270082.0,
      "step": 31915
    },
    {
      "entropy": 5.6563468933105465,
      "epoch": 3.5849385073285784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037637082949267654,
      "loss": 4.9703,
      "mean_token_accuracy": 0.23690439462661744,
      "num_tokens": 71280975.0,
      "step": 31920
    },
    {
      "entropy": 5.607619524002075,
      "epoch": 3.585500084236536,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003763349847991134,
      "loss": 4.9634,
      "mean_token_accuracy": 0.23303506523370743,
      "num_tokens": 71292552.0,
      "step": 31925
    },
    {
      "entropy": 5.662751913070679,
      "epoch": 3.5860616611444938,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000376299136879005,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21415218114852905,
      "num_tokens": 71304653.0,
      "step": 31930
    },
    {
      "entropy": 5.626469421386719,
      "epoch": 3.5866232380524514,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037626328573349296,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22446006238460542,
      "num_tokens": 71316734.0,
      "step": 31935
    },
    {
      "entropy": 5.735386943817138,
      "epoch": 3.587184814960409,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003762274313637185,
      "loss": 5.1197,
      "mean_token_accuracy": 0.21663598716259003,
      "num_tokens": 71328228.0,
      "step": 31940
    },
    {
      "entropy": 5.635874176025391,
      "epoch": 3.5877463918683663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003761915737708234,
      "loss": 4.9509,
      "mean_token_accuracy": 0.22667859643697738,
      "num_tokens": 71339921.0,
      "step": 31945
    },
    {
      "entropy": 5.651909065246582,
      "epoch": 3.588307968776324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037615571295594936,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22358210533857345,
      "num_tokens": 71350995.0,
      "step": 31950
    },
    {
      "entropy": 5.571035146713257,
      "epoch": 3.5888695456842816,
      "grad_norm": 1.125,
      "learning_rate": 0.0003761198489202382,
      "loss": 4.9808,
      "mean_token_accuracy": 0.22682126760482788,
      "num_tokens": 71362546.0,
      "step": 31955
    },
    {
      "entropy": 5.647260046005249,
      "epoch": 3.589431122592239,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003760839816648317,
      "loss": 5.0364,
      "mean_token_accuracy": 0.22539767920970916,
      "num_tokens": 71373431.0,
      "step": 31960
    },
    {
      "entropy": 5.658452939987183,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037604811119087213,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2341421827673912,
      "num_tokens": 71384488.0,
      "step": 31965
    },
    {
      "entropy": 5.624486064910888,
      "epoch": 3.590554276408154,
      "grad_norm": 1.125,
      "learning_rate": 0.0003760122374995015,
      "loss": 4.972,
      "mean_token_accuracy": 0.22718350291252137,
      "num_tokens": 71396558.0,
      "step": 31970
    },
    {
      "entropy": 5.55250654220581,
      "epoch": 3.5911158533161114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22830573618412017,
      "num_tokens": 71407107.0,
      "step": 31975
    },
    {
      "entropy": 5.691866207122803,
      "epoch": 3.591677430224069,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037594048046909616,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22291654497385024,
      "num_tokens": 71419801.0,
      "step": 31980
    },
    {
      "entropy": 5.725202941894532,
      "epoch": 3.5922390071320267,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003759045971323462,
      "loss": 5.0915,
      "mean_token_accuracy": 0.21790891140699387,
      "num_tokens": 71431740.0,
      "step": 31985
    },
    {
      "entropy": 5.64716067314148,
      "epoch": 3.5928005840399844,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037586871058275476,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22599525451660157,
      "num_tokens": 71443514.0,
      "step": 31990
    },
    {
      "entropy": 5.608718967437744,
      "epoch": 3.593362160947942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037583282082146434,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22601809054613115,
      "num_tokens": 71454240.0,
      "step": 31995
    },
    {
      "entropy": 5.577605581283569,
      "epoch": 3.5939237378558992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22579507678747177,
      "num_tokens": 71464707.0,
      "step": 32000
    },
    {
      "entropy": 5.65604453086853,
      "epoch": 3.594485314763857,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003757610316683581,
      "loss": 4.9843,
      "mean_token_accuracy": 0.220093435049057,
      "num_tokens": 71475931.0,
      "step": 32005
    },
    {
      "entropy": 5.66040186882019,
      "epoch": 3.5950468916718146,
      "grad_norm": 1.125,
      "learning_rate": 0.00037572513227882807,
      "loss": 4.9579,
      "mean_token_accuracy": 0.21935688555240632,
      "num_tokens": 71487851.0,
      "step": 32010
    },
    {
      "entropy": 5.654375171661377,
      "epoch": 3.595608468579772,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037568922968217065,
      "loss": 4.9951,
      "mean_token_accuracy": 0.23127286732196808,
      "num_tokens": 71498144.0,
      "step": 32015
    },
    {
      "entropy": 5.6076868057250975,
      "epoch": 3.5961700454877294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037565332387952914,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22439606636762618,
      "num_tokens": 71508074.0,
      "step": 32020
    },
    {
      "entropy": 5.639175891876221,
      "epoch": 3.596731622395687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2264718309044838,
      "num_tokens": 71519609.0,
      "step": 32025
    },
    {
      "entropy": 5.56922550201416,
      "epoch": 3.5972931993036448,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003755815026608669,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23565355092287063,
      "num_tokens": 71530256.0,
      "step": 32030
    },
    {
      "entropy": 5.6861754894256595,
      "epoch": 3.5978547762116024,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037554558724713305,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22376068383455278,
      "num_tokens": 71540383.0,
      "step": 32035
    },
    {
      "entropy": 5.642518949508667,
      "epoch": 3.5984163531195597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003755096686319886,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22309790104627608,
      "num_tokens": 71551739.0,
      "step": 32040
    },
    {
      "entropy": 5.661593723297119,
      "epoch": 3.5989779300275173,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9765,
      "mean_token_accuracy": 0.23120597302913665,
      "num_tokens": 71562429.0,
      "step": 32045
    },
    {
      "entropy": 5.670395040512085,
      "epoch": 3.599539506935475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037543782180204303,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21878775656223298,
      "num_tokens": 71573179.0,
      "step": 32050
    },
    {
      "entropy": 5.634996509552002,
      "epoch": 3.600101083843432,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037540189358952955,
      "loss": 5.0272,
      "mean_token_accuracy": 0.2188015341758728,
      "num_tokens": 71584223.0,
      "step": 32055
    },
    {
      "entropy": 5.5930150032043455,
      "epoch": 3.60066266075139,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003753659621801808,
      "loss": 4.972,
      "mean_token_accuracy": 0.2286396637558937,
      "num_tokens": 71595166.0,
      "step": 32060
    },
    {
      "entropy": 5.615708255767823,
      "epoch": 3.6012242376593475,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037533002757514095,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22022706270217896,
      "num_tokens": 71606290.0,
      "step": 32065
    },
    {
      "entropy": 5.6385204792022705,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.203125,
      "learning_rate": 0.000375294089775554,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2287387102842331,
      "num_tokens": 71617545.0,
      "step": 32070
    },
    {
      "entropy": 5.658947896957398,
      "epoch": 3.6023473914752624,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003752581487825644,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2310576096177101,
      "num_tokens": 71627949.0,
      "step": 32075
    },
    {
      "entropy": 5.665899562835693,
      "epoch": 3.60290896838322,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037522220459731644,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22499061673879622,
      "num_tokens": 71639017.0,
      "step": 32080
    },
    {
      "entropy": 5.543808460235596,
      "epoch": 3.6034705452911777,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23193786889314652,
      "num_tokens": 71649189.0,
      "step": 32085
    },
    {
      "entropy": 5.529028081893921,
      "epoch": 3.6040321221991354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22883875519037247,
      "num_tokens": 71660083.0,
      "step": 32090
    },
    {
      "entropy": 5.558794021606445,
      "epoch": 3.6045936991070926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037511435289946775,
      "loss": 4.89,
      "mean_token_accuracy": 0.23111376017332078,
      "num_tokens": 71670874.0,
      "step": 32095
    },
    {
      "entropy": 5.6834397315979,
      "epoch": 3.6051552760150503,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003750783959566322,
      "loss": 4.9935,
      "mean_token_accuracy": 0.22449166774749757,
      "num_tokens": 71681479.0,
      "step": 32100
    },
    {
      "entropy": 5.615257692337036,
      "epoch": 3.605716852923008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2315642461180687,
      "num_tokens": 71693865.0,
      "step": 32105
    },
    {
      "entropy": 5.587757205963134,
      "epoch": 3.606278429830965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.8849,
      "mean_token_accuracy": 0.230779267847538,
      "num_tokens": 71704513.0,
      "step": 32110
    },
    {
      "entropy": 5.563805770874024,
      "epoch": 3.606840006738923,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037497050601349605,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22132747173309325,
      "num_tokens": 71715444.0,
      "step": 32115
    },
    {
      "entropy": 5.600963163375854,
      "epoch": 3.6074015836468805,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.891,
      "mean_token_accuracy": 0.23026208877563475,
      "num_tokens": 71725452.0,
      "step": 32120
    },
    {
      "entropy": 5.629352378845215,
      "epoch": 3.607963160554838,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.9857,
      "mean_token_accuracy": 0.221768319606781,
      "num_tokens": 71736536.0,
      "step": 32125
    },
    {
      "entropy": 5.575925350189209,
      "epoch": 3.608524737462796,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003748625874224632,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22598017007112503,
      "num_tokens": 71748420.0,
      "step": 32130
    },
    {
      "entropy": 5.598881578445434,
      "epoch": 3.609086314370753,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037482660819793134,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2329712212085724,
      "num_tokens": 71758779.0,
      "step": 32135
    },
    {
      "entropy": 5.7285058975219725,
      "epoch": 3.6096478912787107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003747906257948814,
      "loss": 5.1392,
      "mean_token_accuracy": 0.21476680934429168,
      "num_tokens": 71771042.0,
      "step": 32140
    },
    {
      "entropy": 5.651338815689087,
      "epoch": 3.6102094681866683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003747546402144594,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23215834498405458,
      "num_tokens": 71782234.0,
      "step": 32145
    },
    {
      "entropy": 5.6185314655303955,
      "epoch": 3.6107710450946255,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.9082,
      "mean_token_accuracy": 0.23065560013055803,
      "num_tokens": 71793115.0,
      "step": 32150
    },
    {
      "entropy": 5.544287109375,
      "epoch": 3.611332622002583,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037468265952608196,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23684015572071077,
      "num_tokens": 71804466.0,
      "step": 32155
    },
    {
      "entropy": 5.632847261428833,
      "epoch": 3.611894198910541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003746466644204183,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21474687606096268,
      "num_tokens": 71815248.0,
      "step": 32160
    },
    {
      "entropy": 5.679193592071533,
      "epoch": 3.612455775818498,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003746106661419663,
      "loss": 4.9911,
      "mean_token_accuracy": 0.23439489603042601,
      "num_tokens": 71825816.0,
      "step": 32165
    },
    {
      "entropy": 5.623431587219239,
      "epoch": 3.6130173527264557,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000374574664691872,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22387421429157256,
      "num_tokens": 71837918.0,
      "step": 32170
    },
    {
      "entropy": 5.637952280044556,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037453866007128176,
      "loss": 5.0495,
      "mean_token_accuracy": 0.22424838095903396,
      "num_tokens": 71848736.0,
      "step": 32175
    },
    {
      "entropy": 5.564182996749878,
      "epoch": 3.614140506542371,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037450265228134203,
      "loss": 4.8764,
      "mean_token_accuracy": 0.23441838771104812,
      "num_tokens": 71859903.0,
      "step": 32180
    },
    {
      "entropy": 5.6084144592285154,
      "epoch": 3.6147020834503287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037446664132319925,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22006642669439316,
      "num_tokens": 71871656.0,
      "step": 32185
    },
    {
      "entropy": 5.657885456085205,
      "epoch": 3.615263660358286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000374430627198,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22140689045190812,
      "num_tokens": 71883282.0,
      "step": 32190
    },
    {
      "entropy": 5.647553157806397,
      "epoch": 3.6158252372662436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22415488064289094,
      "num_tokens": 71894473.0,
      "step": 32195
    },
    {
      "entropy": 5.638541126251221,
      "epoch": 3.6163868141742013,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003743585894510192,
      "loss": 5.0746,
      "mean_token_accuracy": 0.21466537564992905,
      "num_tokens": 71905555.0,
      "step": 32200
    },
    {
      "entropy": 5.5487652778625485,
      "epoch": 3.6169483910821585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037432256583153127,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2194278657436371,
      "num_tokens": 71916599.0,
      "step": 32205
    },
    {
      "entropy": 5.657393074035644,
      "epoch": 3.617509967990116,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037428653904957433,
      "loss": 4.9475,
      "mean_token_accuracy": 0.22550550997257232,
      "num_tokens": 71926173.0,
      "step": 32210
    },
    {
      "entropy": 5.674569416046142,
      "epoch": 3.618071544898074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003742505091062955,
      "loss": 5.0369,
      "mean_token_accuracy": 0.2244032084941864,
      "num_tokens": 71937389.0,
      "step": 32215
    },
    {
      "entropy": 5.6286702156066895,
      "epoch": 3.6186331218060315,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003742144760028419,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22477958649396895,
      "num_tokens": 71949216.0,
      "step": 32220
    },
    {
      "entropy": 5.6799070835113525,
      "epoch": 3.619194698713989,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000374178439740361,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21652624756097794,
      "num_tokens": 71959861.0,
      "step": 32225
    },
    {
      "entropy": 5.626048469543457,
      "epoch": 3.6197562756219464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003741424003200001,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22252779304981232,
      "num_tokens": 71970043.0,
      "step": 32230
    },
    {
      "entropy": 5.633416795730591,
      "epoch": 3.620317852529904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037410635774290667,
      "loss": 4.947,
      "mean_token_accuracy": 0.23424825370311736,
      "num_tokens": 71981096.0,
      "step": 32235
    },
    {
      "entropy": 5.630755472183227,
      "epoch": 3.6208794294378617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2273010179400444,
      "num_tokens": 71992759.0,
      "step": 32240
    },
    {
      "entropy": 5.619220018386841,
      "epoch": 3.621441006345819,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000374034263123113,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22678032964468003,
      "num_tokens": 72003187.0,
      "step": 32245
    },
    {
      "entropy": 5.744490289688111,
      "epoch": 3.6220025832537766,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037399821108270823,
      "loss": 5.1276,
      "mean_token_accuracy": 0.21593839824199676,
      "num_tokens": 72013712.0,
      "step": 32250
    },
    {
      "entropy": 5.720804834365845,
      "epoch": 3.622564160161734,
      "grad_norm": 1.125,
      "learning_rate": 0.000373962155890162,
      "loss": 5.0891,
      "mean_token_accuracy": 0.21821607351303102,
      "num_tokens": 72024693.0,
      "step": 32255
    },
    {
      "entropy": 5.633784675598145,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003739260975466223,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23241652697324752,
      "num_tokens": 72036578.0,
      "step": 32260
    },
    {
      "entropy": 5.6779951572418215,
      "epoch": 3.623687313977649,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037389003605323737,
      "loss": 5.0477,
      "mean_token_accuracy": 0.22544585764408112,
      "num_tokens": 72047574.0,
      "step": 32265
    },
    {
      "entropy": 5.659405612945557,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.9555,
      "mean_token_accuracy": 0.23084328770637513,
      "num_tokens": 72059998.0,
      "step": 32270
    },
    {
      "entropy": 5.671789026260376,
      "epoch": 3.6248104677935644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037381790362152453,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21959096491336821,
      "num_tokens": 72070753.0,
      "step": 32275
    },
    {
      "entropy": 5.589607238769531,
      "epoch": 3.625372044701522,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.9008,
      "mean_token_accuracy": 0.229271999001503,
      "num_tokens": 72081718.0,
      "step": 32280
    },
    {
      "entropy": 5.595297050476074,
      "epoch": 3.6259336216094793,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2277024507522583,
      "num_tokens": 72092403.0,
      "step": 32285
    },
    {
      "entropy": 5.58658127784729,
      "epoch": 3.626495198517437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003737096813788238,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22990573793649674,
      "num_tokens": 72103711.0,
      "step": 32290
    },
    {
      "entropy": 5.681488561630249,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003736736010104829,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21983443647623063,
      "num_tokens": 72115095.0,
      "step": 32295
    },
    {
      "entropy": 5.695921802520752,
      "epoch": 3.627618352333352,
      "grad_norm": 1.125,
      "learning_rate": 0.0003736375175003363,
      "loss": 5.0307,
      "mean_token_accuracy": 0.22102911323308944,
      "num_tokens": 72126559.0,
      "step": 32300
    },
    {
      "entropy": 5.617884826660156,
      "epoch": 3.6281799292413095,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037360143084953285,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22404062300920485,
      "num_tokens": 72137752.0,
      "step": 32305
    },
    {
      "entropy": 5.533916282653808,
      "epoch": 3.628741506149267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22901957631111144,
      "num_tokens": 72148888.0,
      "step": 32310
    },
    {
      "entropy": 5.6205638408660885,
      "epoch": 3.629303083057225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2291433185338974,
      "num_tokens": 72159607.0,
      "step": 32315
    },
    {
      "entropy": 5.675597381591797,
      "epoch": 3.6298646599651825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003734931520646724,
      "loss": 5.0079,
      "mean_token_accuracy": 0.22556714117527008,
      "num_tokens": 72171204.0,
      "step": 32320
    },
    {
      "entropy": 5.593688344955444,
      "epoch": 3.6304262368731397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037345705286273285,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2271861508488655,
      "num_tokens": 72182125.0,
      "step": 32325
    },
    {
      "entropy": 5.639527034759522,
      "epoch": 3.6309878137810974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22232643961906434,
      "num_tokens": 72193093.0,
      "step": 32330
    },
    {
      "entropy": 5.669413232803345,
      "epoch": 3.631549390689055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037338484505527103,
      "loss": 5.0344,
      "mean_token_accuracy": 0.21599610894918442,
      "num_tokens": 72204235.0,
      "step": 32335
    },
    {
      "entropy": 5.602661514282227,
      "epoch": 3.6321109675970122,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037334873645204785,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2214491680264473,
      "num_tokens": 72215160.0,
      "step": 32340
    },
    {
      "entropy": 5.616254615783691,
      "epoch": 3.63267254450497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037331262471736285,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2264119029045105,
      "num_tokens": 72227517.0,
      "step": 32345
    },
    {
      "entropy": 5.714451503753662,
      "epoch": 3.6332341214129276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037327650985236564,
      "loss": 5.0399,
      "mean_token_accuracy": 0.2225249871611595,
      "num_tokens": 72238616.0,
      "step": 32350
    },
    {
      "entropy": 5.7022645473480225,
      "epoch": 3.633795698320885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003732403918582063,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22471022009849548,
      "num_tokens": 72250252.0,
      "step": 32355
    },
    {
      "entropy": 5.6914379596710205,
      "epoch": 3.6343572752288424,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037320427073603475,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2167406663298607,
      "num_tokens": 72261197.0,
      "step": 32360
    },
    {
      "entropy": 5.539922571182251,
      "epoch": 3.6349188521368,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.926,
      "mean_token_accuracy": 0.2297421231865883,
      "num_tokens": 72273838.0,
      "step": 32365
    },
    {
      "entropy": 5.603180837631226,
      "epoch": 3.6354804290447578,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003731320191122557,
      "loss": 4.9463,
      "mean_token_accuracy": 0.23481651097536088,
      "num_tokens": 72285047.0,
      "step": 32370
    },
    {
      "entropy": 5.642091941833496,
      "epoch": 3.6360420059527154,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003730958886129486,
      "loss": 5.0346,
      "mean_token_accuracy": 0.22384515404701233,
      "num_tokens": 72295883.0,
      "step": 32375
    },
    {
      "entropy": 5.676258897781372,
      "epoch": 3.6366035828606726,
      "grad_norm": 1.125,
      "learning_rate": 0.00037305975499023046,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22925477027893065,
      "num_tokens": 72307586.0,
      "step": 32380
    },
    {
      "entropy": 5.581606149673462,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003730236182452517,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22693742513656617,
      "num_tokens": 72318501.0,
      "step": 32385
    },
    {
      "entropy": 5.616404104232788,
      "epoch": 3.637726736676588,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003729874783791629,
      "loss": 4.981,
      "mean_token_accuracy": 0.2238684505224228,
      "num_tokens": 72330178.0,
      "step": 32390
    },
    {
      "entropy": 5.683768606185913,
      "epoch": 3.638288313584545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003729513353931148,
      "loss": 5.0607,
      "mean_token_accuracy": 0.223661932349205,
      "num_tokens": 72340304.0,
      "step": 32395
    },
    {
      "entropy": 5.675357913970947,
      "epoch": 3.638849890492503,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003729151892882581,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22869589179754257,
      "num_tokens": 72350353.0,
      "step": 32400
    },
    {
      "entropy": 5.621267557144165,
      "epoch": 3.6394114674004605,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003728790400657439,
      "loss": 4.9304,
      "mean_token_accuracy": 0.23512963354587554,
      "num_tokens": 72361293.0,
      "step": 32405
    },
    {
      "entropy": 5.565197944641113,
      "epoch": 3.639973044308418,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.928,
      "mean_token_accuracy": 0.23160567581653596,
      "num_tokens": 72372013.0,
      "step": 32410
    },
    {
      "entropy": 5.575791931152343,
      "epoch": 3.640534621216376,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037280673227234687,
      "loss": 5.0055,
      "mean_token_accuracy": 0.21743751615285872,
      "num_tokens": 72382611.0,
      "step": 32415
    },
    {
      "entropy": 5.647897958755493,
      "epoch": 3.641096198124333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2291641801595688,
      "num_tokens": 72393701.0,
      "step": 32420
    },
    {
      "entropy": 5.783651542663574,
      "epoch": 3.6416577750322907,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003727344120221326,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21645863205194474,
      "num_tokens": 72404168.0,
      "step": 32425
    },
    {
      "entropy": 5.680489349365234,
      "epoch": 3.6422193519402484,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037269824722859743,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2227896362543106,
      "num_tokens": 72415184.0,
      "step": 32430
    },
    {
      "entropy": 5.613320159912109,
      "epoch": 3.6427809288482056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003726620793243121,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22827332019805907,
      "num_tokens": 72426189.0,
      "step": 32435
    },
    {
      "entropy": 5.61771411895752,
      "epoch": 3.6433425057561633,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003726259083104284,
      "loss": 5.0043,
      "mean_token_accuracy": 0.22680219858884812,
      "num_tokens": 72436741.0,
      "step": 32440
    },
    {
      "entropy": 5.700105381011963,
      "epoch": 3.643904082664121,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037258973418809784,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2290955364704132,
      "num_tokens": 72448610.0,
      "step": 32445
    },
    {
      "entropy": 5.63028244972229,
      "epoch": 3.644465659572078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003725535569584723,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22818560600280763,
      "num_tokens": 72460789.0,
      "step": 32450
    },
    {
      "entropy": 5.579542827606201,
      "epoch": 3.645027236480036,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2209724724292755,
      "num_tokens": 72471065.0,
      "step": 32455
    },
    {
      "entropy": 5.59852032661438,
      "epoch": 3.6455888133879935,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000372481193181944,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23046256005764007,
      "num_tokens": 72481788.0,
      "step": 32460
    },
    {
      "entropy": 5.6232503890991214,
      "epoch": 3.646150390295951,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23059139102697374,
      "num_tokens": 72493840.0,
      "step": 32465
    },
    {
      "entropy": 5.590166330337524,
      "epoch": 3.6467119672039088,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003724088169900597,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22173532396554946,
      "num_tokens": 72505002.0,
      "step": 32470
    },
    {
      "entropy": 5.565439605712891,
      "epoch": 3.647273544111866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037237262424123975,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22809031158685683,
      "num_tokens": 72515917.0,
      "step": 32475
    },
    {
      "entropy": 5.666340112686157,
      "epoch": 3.6478351210198237,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037233642839203763,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22627322375774384,
      "num_tokens": 72527500.0,
      "step": 32480
    },
    {
      "entropy": 5.531005573272705,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003723002294436059,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23148231953382492,
      "num_tokens": 72538402.0,
      "step": 32485
    },
    {
      "entropy": 5.597081899642944,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22196704894304276,
      "num_tokens": 72550260.0,
      "step": 32490
    },
    {
      "entropy": 5.736956834793091,
      "epoch": 3.649519851743696,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003722278222536639,
      "loss": 5.0311,
      "mean_token_accuracy": 0.22172872871160507,
      "num_tokens": 72562335.0,
      "step": 32495
    },
    {
      "entropy": 5.633713674545288,
      "epoch": 3.650081428651654,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037219161401445914,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22692375779151916,
      "num_tokens": 72572620.0,
      "step": 32500
    },
    {
      "entropy": 5.54709906578064,
      "epoch": 3.6506430055596115,
      "grad_norm": 1.0,
      "learning_rate": 0.00037215540268063573,
      "loss": 5.0274,
      "mean_token_accuracy": 0.22289025485515596,
      "num_tokens": 72584487.0,
      "step": 32505
    },
    {
      "entropy": 5.620947170257568,
      "epoch": 3.651204582467569,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003721191882533466,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21871782541275026,
      "num_tokens": 72596191.0,
      "step": 32510
    },
    {
      "entropy": 5.6736565113067625,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003720829707337449,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22380343228578567,
      "num_tokens": 72605901.0,
      "step": 32515
    },
    {
      "entropy": 5.650775623321533,
      "epoch": 3.652327736283484,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003720467501229836,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22407910078763962,
      "num_tokens": 72617231.0,
      "step": 32520
    },
    {
      "entropy": 5.719034910202026,
      "epoch": 3.6528893131914417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003720105264222163,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22251069992780687,
      "num_tokens": 72628777.0,
      "step": 32525
    },
    {
      "entropy": 5.659135198593139,
      "epoch": 3.653450890099399,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037197429963259615,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22739089280366898,
      "num_tokens": 72639670.0,
      "step": 32530
    },
    {
      "entropy": 5.5817389488220215,
      "epoch": 3.6540124670073566,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003719380697552767,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23295606970787047,
      "num_tokens": 72651010.0,
      "step": 32535
    },
    {
      "entropy": 5.762537622451783,
      "epoch": 3.6545740439153143,
      "grad_norm": 1.125,
      "learning_rate": 0.0003719018367914115,
      "loss": 5.083,
      "mean_token_accuracy": 0.21491123288869857,
      "num_tokens": 72662900.0,
      "step": 32540
    },
    {
      "entropy": 5.6630126953125,
      "epoch": 3.6551356208232715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003718656007421542,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2236749529838562,
      "num_tokens": 72674348.0,
      "step": 32545
    },
    {
      "entropy": 5.633159875869751,
      "epoch": 3.655697197731229,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037182936160865866,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23431225419044494,
      "num_tokens": 72685771.0,
      "step": 32550
    },
    {
      "entropy": 5.594312524795532,
      "epoch": 3.656258774639187,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23133568465709686,
      "num_tokens": 72697121.0,
      "step": 32555
    },
    {
      "entropy": 5.549131202697754,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22950998544692994,
      "num_tokens": 72707245.0,
      "step": 32560
    },
    {
      "entropy": 5.584798908233642,
      "epoch": 3.657381928455102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23329731971025466,
      "num_tokens": 72718328.0,
      "step": 32565
    },
    {
      "entropy": 5.7160858631134035,
      "epoch": 3.6579435053630593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003716843742553724,
      "loss": 5.0385,
      "mean_token_accuracy": 0.2217957928776741,
      "num_tokens": 72729192.0,
      "step": 32570
    },
    {
      "entropy": 5.610609722137451,
      "epoch": 3.658505082271017,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003716481197179953,
      "loss": 5.0,
      "mean_token_accuracy": 0.22190637439489364,
      "num_tokens": 72740213.0,
      "step": 32575
    },
    {
      "entropy": 5.544430208206177,
      "epoch": 3.6590666591789747,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037161186210330464,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22879549860954285,
      "num_tokens": 72752477.0,
      "step": 32580
    },
    {
      "entropy": 5.696344947814941,
      "epoch": 3.659628236086932,
      "grad_norm": 1.125,
      "learning_rate": 0.00037157560141245485,
      "loss": 4.9357,
      "mean_token_accuracy": 0.23291114419698716,
      "num_tokens": 72763926.0,
      "step": 32585
    },
    {
      "entropy": 5.645985126495361,
      "epoch": 3.6601898129948895,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037153933764660034,
      "loss": 4.991,
      "mean_token_accuracy": 0.22380518019199372,
      "num_tokens": 72774922.0,
      "step": 32590
    },
    {
      "entropy": 5.591260004043579,
      "epoch": 3.660751389902847,
      "grad_norm": 1.125,
      "learning_rate": 0.000371503070806896,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2244374394416809,
      "num_tokens": 72786362.0,
      "step": 32595
    },
    {
      "entropy": 5.604332447052002,
      "epoch": 3.661312966810805,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2282548278570175,
      "num_tokens": 72798337.0,
      "step": 32600
    },
    {
      "entropy": 5.648778581619263,
      "epoch": 3.6618745437187625,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037143052791055654,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23463111668825148,
      "num_tokens": 72808923.0,
      "step": 32605
    },
    {
      "entropy": 5.566091251373291,
      "epoch": 3.6624361206267197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22958754301071166,
      "num_tokens": 72820148.0,
      "step": 32610
    },
    {
      "entropy": 5.595463752746582,
      "epoch": 3.6629976975346774,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003713579727326754,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22608145624399184,
      "num_tokens": 72830412.0,
      "step": 32615
    },
    {
      "entropy": 5.621809482574463,
      "epoch": 3.663559274442635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.892,
      "mean_token_accuracy": 0.2319553256034851,
      "num_tokens": 72840857.0,
      "step": 32620
    },
    {
      "entropy": 5.59148850440979,
      "epoch": 3.6641208513505923,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22692949026823045,
      "num_tokens": 72852269.0,
      "step": 32625
    },
    {
      "entropy": 5.616663551330566,
      "epoch": 3.66468242825855,
      "grad_norm": 1.125,
      "learning_rate": 0.000371249116958178,
      "loss": 5.0043,
      "mean_token_accuracy": 0.22489778846502304,
      "num_tokens": 72863536.0,
      "step": 32630
    },
    {
      "entropy": 5.573177433013916,
      "epoch": 3.6652440051665076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.963,
      "mean_token_accuracy": 0.22791206389665603,
      "num_tokens": 72873835.0,
      "step": 32635
    },
    {
      "entropy": 5.6507963180542,
      "epoch": 3.665805582074465,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003711765311168748,
      "loss": 4.8719,
      "mean_token_accuracy": 0.22276733964681625,
      "num_tokens": 72883916.0,
      "step": 32640
    },
    {
      "entropy": 5.640057516098023,
      "epoch": 3.6663671589824225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037114023360219824,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2192799136042595,
      "num_tokens": 72895928.0,
      "step": 32645
    },
    {
      "entropy": 5.560006904602051,
      "epoch": 3.66692873589038,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037110393302637937,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22601151317358018,
      "num_tokens": 72908138.0,
      "step": 32650
    },
    {
      "entropy": 5.662873792648315,
      "epoch": 3.667490312798338,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037106762939057404,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22054456174373627,
      "num_tokens": 72918326.0,
      "step": 32655
    },
    {
      "entropy": 5.6778768539428714,
      "epoch": 3.6680518897062955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003710313226959381,
      "loss": 5.025,
      "mean_token_accuracy": 0.22625663131475449,
      "num_tokens": 72929140.0,
      "step": 32660
    },
    {
      "entropy": 5.644418144226075,
      "epoch": 3.6686134666142527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003709950129436276,
      "loss": 4.9824,
      "mean_token_accuracy": 0.23289711624383927,
      "num_tokens": 72941355.0,
      "step": 32665
    },
    {
      "entropy": 5.6564170837402346,
      "epoch": 3.6691750435222104,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003709587001347987,
      "loss": 5.0001,
      "mean_token_accuracy": 0.22235267907381057,
      "num_tokens": 72951841.0,
      "step": 32670
    },
    {
      "entropy": 5.569346618652344,
      "epoch": 3.669736620430168,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037092238427060756,
      "loss": 4.924,
      "mean_token_accuracy": 0.2237014353275299,
      "num_tokens": 72963843.0,
      "step": 32675
    },
    {
      "entropy": 5.654981851577759,
      "epoch": 3.6702981973381252,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003708860653522104,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22554971426725387,
      "num_tokens": 72973709.0,
      "step": 32680
    },
    {
      "entropy": 5.564440631866455,
      "epoch": 3.670859774246083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.7928,
      "mean_token_accuracy": 0.24218248426914216,
      "num_tokens": 72984658.0,
      "step": 32685
    },
    {
      "entropy": 5.535476350784302,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037081341835742433,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22502538710832595,
      "num_tokens": 72995745.0,
      "step": 32690
    },
    {
      "entropy": 5.587911748886109,
      "epoch": 3.671982928061998,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003707770902833483,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22476495653390885,
      "num_tokens": 73007619.0,
      "step": 32695
    },
    {
      "entropy": 5.622140598297119,
      "epoch": 3.672544504969956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22597305327653885,
      "num_tokens": 73017779.0,
      "step": 32700
    },
    {
      "entropy": 5.653342199325562,
      "epoch": 3.673106081877913,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000370704424987614,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2271291881799698,
      "num_tokens": 73028246.0,
      "step": 32705
    },
    {
      "entropy": 5.633440971374512,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003706680877682694,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21716709285974503,
      "num_tokens": 73038896.0,
      "step": 32710
    },
    {
      "entropy": 5.679209280014038,
      "epoch": 3.6742292356938284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037063174750281573,
      "loss": 5.0511,
      "mean_token_accuracy": 0.21585436761379242,
      "num_tokens": 73050960.0,
      "step": 32715
    },
    {
      "entropy": 5.693519926071167,
      "epoch": 3.6747908126017856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.9469,
      "mean_token_accuracy": 0.23750385642051697,
      "num_tokens": 73061537.0,
      "step": 32720
    },
    {
      "entropy": 5.685467863082886,
      "epoch": 3.6753523895097433,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037055905783821,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21861578077077864,
      "num_tokens": 73072357.0,
      "step": 32725
    },
    {
      "entropy": 5.583578586578369,
      "epoch": 3.675913966417701,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037052270844137216,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2308937504887581,
      "num_tokens": 73083527.0,
      "step": 32730
    },
    {
      "entropy": 5.576510620117188,
      "epoch": 3.676475543325658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.886,
      "mean_token_accuracy": 0.22586217075586318,
      "num_tokens": 73095068.0,
      "step": 32735
    },
    {
      "entropy": 5.528926849365234,
      "epoch": 3.677037120233616,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003704500005244139,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23275781869888307,
      "num_tokens": 73106318.0,
      "step": 32740
    },
    {
      "entropy": 5.612875843048096,
      "epoch": 3.6775986971415735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037041364200660835,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22662498205900192,
      "num_tokens": 73118433.0,
      "step": 32745
    },
    {
      "entropy": 5.652720975875854,
      "epoch": 3.678160274049531,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037037728045079553,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2189541682600975,
      "num_tokens": 73129887.0,
      "step": 32750
    },
    {
      "entropy": 5.577433109283447,
      "epoch": 3.678721850957489,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000370340915858133,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22969317734241484,
      "num_tokens": 73141529.0,
      "step": 32755
    },
    {
      "entropy": 5.605078458786011,
      "epoch": 3.679283427865446,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003703045482297788,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22309874892234802,
      "num_tokens": 73151317.0,
      "step": 32760
    },
    {
      "entropy": 5.50536789894104,
      "epoch": 3.6798450047734037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2310761883854866,
      "num_tokens": 73162382.0,
      "step": 32765
    },
    {
      "entropy": 5.6725365161895756,
      "epoch": 3.6804065816813614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003702318038706272,
      "loss": 5.0517,
      "mean_token_accuracy": 0.22495098412036896,
      "num_tokens": 73173142.0,
      "step": 32770
    },
    {
      "entropy": 5.595385599136352,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22339274883270263,
      "num_tokens": 73185074.0,
      "step": 32775
    },
    {
      "entropy": 5.704447841644287,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037015904738260533,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2303239658474922,
      "num_tokens": 73195806.0,
      "step": 32780
    },
    {
      "entropy": 5.7439038276672365,
      "epoch": 3.682091312405234,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003701226645931637,
      "loss": 5.1089,
      "mean_token_accuracy": 0.2102091833949089,
      "num_tokens": 73207213.0,
      "step": 32785
    },
    {
      "entropy": 5.634860229492188,
      "epoch": 3.6826528893131916,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003700862787749796,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22626604288816451,
      "num_tokens": 73219715.0,
      "step": 32790
    },
    {
      "entropy": 5.60251898765564,
      "epoch": 3.6832144662211492,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003700498899292115,
      "loss": 4.9177,
      "mean_token_accuracy": 0.228821761906147,
      "num_tokens": 73229868.0,
      "step": 32795
    },
    {
      "entropy": 5.636297512054443,
      "epoch": 3.6837760431291064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.9128,
      "mean_token_accuracy": 0.224118410050869,
      "num_tokens": 73240347.0,
      "step": 32800
    },
    {
      "entropy": 5.654324913024903,
      "epoch": 3.684337620037064,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003699771031595579,
      "loss": 4.9542,
      "mean_token_accuracy": 0.23108764737844467,
      "num_tokens": 73250165.0,
      "step": 32805
    },
    {
      "entropy": 5.716234636306763,
      "epoch": 3.6848991969450218,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036994070523799005,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22806993126869202,
      "num_tokens": 73261320.0,
      "step": 32810
    },
    {
      "entropy": 5.637402963638306,
      "epoch": 3.685460773852979,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036990430429347334,
      "loss": 5.0211,
      "mean_token_accuracy": 0.22209927439689636,
      "num_tokens": 73272342.0,
      "step": 32815
    },
    {
      "entropy": 5.663855934143067,
      "epoch": 3.6860223507609367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036986790032716673,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22913243621587753,
      "num_tokens": 73283790.0,
      "step": 32820
    },
    {
      "entropy": 5.5629082202911375,
      "epoch": 3.6865839276688943,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22284430563449859,
      "num_tokens": 73295869.0,
      "step": 32825
    },
    {
      "entropy": 5.5368396759033205,
      "epoch": 3.6871455045768515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003697950833338204,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21993789970874786,
      "num_tokens": 73307518.0,
      "step": 32830
    },
    {
      "entropy": 5.623408412933349,
      "epoch": 3.687707081484809,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036975867030909933,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22553968876600267,
      "num_tokens": 73318419.0,
      "step": 32835
    },
    {
      "entropy": 5.634183168411255,
      "epoch": 3.688268658392767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22500164806842804,
      "num_tokens": 73328094.0,
      "step": 32840
    },
    {
      "entropy": 5.627250051498413,
      "epoch": 3.6888302353007245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.9457,
      "mean_token_accuracy": 0.23439213037490844,
      "num_tokens": 73339156.0,
      "step": 32845
    },
    {
      "entropy": 5.60182375907898,
      "epoch": 3.689391812208682,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036964941313665715,
      "loss": 4.8501,
      "mean_token_accuracy": 0.24023815244436264,
      "num_tokens": 73350740.0,
      "step": 32850
    },
    {
      "entropy": 5.671402883529663,
      "epoch": 3.6899533891166394,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003696129880502823,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2281578630208969,
      "num_tokens": 73360745.0,
      "step": 32855
    },
    {
      "entropy": 5.643281602859497,
      "epoch": 3.690514966024597,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000369576559951393,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22479527592658996,
      "num_tokens": 73372580.0,
      "step": 32860
    },
    {
      "entropy": 5.648338365554809,
      "epoch": 3.6910765429325547,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036954012884114946,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22593629062175752,
      "num_tokens": 73383970.0,
      "step": 32865
    },
    {
      "entropy": 5.622895431518555,
      "epoch": 3.691638119840512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003695036947207115,
      "loss": 4.9126,
      "mean_token_accuracy": 0.23315507769584656,
      "num_tokens": 73394472.0,
      "step": 32870
    },
    {
      "entropy": 5.604364967346191,
      "epoch": 3.6921996967484696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036946725759123936,
      "loss": 5.0025,
      "mean_token_accuracy": 0.22293138056993483,
      "num_tokens": 73405415.0,
      "step": 32875
    },
    {
      "entropy": 5.671274995803833,
      "epoch": 3.6927612736564273,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003694308174538929,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22492717504501342,
      "num_tokens": 73416583.0,
      "step": 32880
    },
    {
      "entropy": 5.712176132202148,
      "epoch": 3.693322850564385,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036939437430983274,
      "loss": 5.0464,
      "mean_token_accuracy": 0.22287210524082185,
      "num_tokens": 73427408.0,
      "step": 32885
    },
    {
      "entropy": 5.588544797897339,
      "epoch": 3.6938844274723426,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.9174,
      "mean_token_accuracy": 0.2375206768512726,
      "num_tokens": 73438509.0,
      "step": 32890
    },
    {
      "entropy": 5.603142166137696,
      "epoch": 3.6944460043803,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.9289,
      "mean_token_accuracy": 0.23122102469205857,
      "num_tokens": 73449197.0,
      "step": 32895
    },
    {
      "entropy": 5.65603346824646,
      "epoch": 3.6950075812882575,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036928502684897314,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2196334034204483,
      "num_tokens": 73460777.0,
      "step": 32900
    },
    {
      "entropy": 5.733854055404663,
      "epoch": 3.695569158196215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036924857168966216,
      "loss": 4.9946,
      "mean_token_accuracy": 0.22207329720258712,
      "num_tokens": 73471747.0,
      "step": 32905
    },
    {
      "entropy": 5.523287010192871,
      "epoch": 3.6961307351041723,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003692121135294401,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23051729053258896,
      "num_tokens": 73481258.0,
      "step": 32910
    },
    {
      "entropy": 5.560393238067627,
      "epoch": 3.69669231201213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22564314305782318,
      "num_tokens": 73492073.0,
      "step": 32915
    },
    {
      "entropy": 5.568282699584961,
      "epoch": 3.6972538889200877,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22064192742109298,
      "num_tokens": 73502813.0,
      "step": 32920
    },
    {
      "entropy": 5.669391107559204,
      "epoch": 3.697815465828045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036910272105491625,
      "loss": 4.981,
      "mean_token_accuracy": 0.23291026204824447,
      "num_tokens": 73514417.0,
      "step": 32925
    },
    {
      "entropy": 5.671319627761841,
      "epoch": 3.6983770427360025,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036906625090265925,
      "loss": 4.934,
      "mean_token_accuracy": 0.2244684100151062,
      "num_tokens": 73525636.0,
      "step": 32930
    },
    {
      "entropy": 5.565916585922241,
      "epoch": 3.69893861964396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22886175513267518,
      "num_tokens": 73536557.0,
      "step": 32935
    },
    {
      "entropy": 5.568959426879883,
      "epoch": 3.699500196551918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003689933016139888,
      "loss": 4.856,
      "mean_token_accuracy": 0.2388333261013031,
      "num_tokens": 73547910.0,
      "step": 32940
    },
    {
      "entropy": 5.53167495727539,
      "epoch": 3.7000617734598755,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2336424320936203,
      "num_tokens": 73557583.0,
      "step": 32945
    },
    {
      "entropy": 5.59385929107666,
      "epoch": 3.7006233503678327,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003689203403541857,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2212565064430237,
      "num_tokens": 73567300.0,
      "step": 32950
    },
    {
      "entropy": 5.674770641326904,
      "epoch": 3.7011849272757904,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036888385523801325,
      "loss": 5.0265,
      "mean_token_accuracy": 0.2187275171279907,
      "num_tokens": 73577387.0,
      "step": 32955
    },
    {
      "entropy": 5.723411989212036,
      "epoch": 3.701746504183748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036884736713254247,
      "loss": 5.0177,
      "mean_token_accuracy": 0.22313751578330993,
      "num_tokens": 73588340.0,
      "step": 32960
    },
    {
      "entropy": 5.647705459594727,
      "epoch": 3.7023080810917053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036881087603893515,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22730784714221955,
      "num_tokens": 73599650.0,
      "step": 32965
    },
    {
      "entropy": 5.715647459030151,
      "epoch": 3.702869657999663,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036877438195835307,
      "loss": 5.0544,
      "mean_token_accuracy": 0.22700998485088347,
      "num_tokens": 73611279.0,
      "step": 32970
    },
    {
      "entropy": 5.554759550094604,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2245277687907219,
      "num_tokens": 73623279.0,
      "step": 32975
    },
    {
      "entropy": 5.6211460590362545,
      "epoch": 3.7039928118155783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036870138484091297,
      "loss": 5.0342,
      "mean_token_accuracy": 0.22009627819061278,
      "num_tokens": 73633756.0,
      "step": 32980
    },
    {
      "entropy": 5.734121561050415,
      "epoch": 3.704554388723536,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036866488180637926,
      "loss": 5.0732,
      "mean_token_accuracy": 0.21610168218612671,
      "num_tokens": 73646033.0,
      "step": 32985
    },
    {
      "entropy": 5.624131441116333,
      "epoch": 3.705115965631493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003686283757895193,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22154430598020552,
      "num_tokens": 73655833.0,
      "step": 32990
    },
    {
      "entropy": 5.690882825851441,
      "epoch": 3.705677542539451,
      "grad_norm": 1.125,
      "learning_rate": 0.00036859186679149555,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21927231401205063,
      "num_tokens": 73667755.0,
      "step": 32995
    },
    {
      "entropy": 5.6824303150177,
      "epoch": 3.7062391194474085,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003685553548134705,
      "loss": 5.0715,
      "mean_token_accuracy": 0.22573737055063248,
      "num_tokens": 73680537.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.3784090971436305,
      "eval_loss": 5.072376251220703,
      "eval_mean_token_accuracy": 0.22859097435985887,
      "eval_num_tokens": 73680537.0,
      "eval_runtime": 18.4843,
      "eval_samples_per_second": 1677.799,
      "eval_steps_per_second": 209.745,
      "step": 33000
    },
    {
      "entropy": 5.631691741943359,
      "epoch": 3.7068006963553657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23526250720024108,
      "num_tokens": 73692446.0,
      "step": 33005
    },
    {
      "entropy": 5.603858518600464,
      "epoch": 3.7073622732633233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003684823219220666,
      "loss": 4.9914,
      "mean_token_accuracy": 0.22010553479194642,
      "num_tokens": 73704059.0,
      "step": 33010
    },
    {
      "entropy": 5.590863561630249,
      "epoch": 3.707923850171281,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003684458010110133,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22634463906288146,
      "num_tokens": 73715234.0,
      "step": 33015
    },
    {
      "entropy": 5.687485790252685,
      "epoch": 3.7084854270792382,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003684092771246094,
      "loss": 4.9925,
      "mean_token_accuracy": 0.2209886223077774,
      "num_tokens": 73726930.0,
      "step": 33020
    },
    {
      "entropy": 5.665072679519653,
      "epoch": 3.709047003987196,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003683727502640178,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22634626626968385,
      "num_tokens": 73738107.0,
      "step": 33025
    },
    {
      "entropy": 5.673057317733765,
      "epoch": 3.7096085808951536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.0815,
      "mean_token_accuracy": 0.21855539977550506,
      "num_tokens": 73749165.0,
      "step": 33030
    },
    {
      "entropy": 5.6276407718658445,
      "epoch": 3.710170157803111,
      "grad_norm": 1.125,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.9269,
      "mean_token_accuracy": 0.23233648985624314,
      "num_tokens": 73760921.0,
      "step": 33035
    },
    {
      "entropy": 5.625083637237549,
      "epoch": 3.710731734711069,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003682631518487483,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22879309505224227,
      "num_tokens": 73772045.0,
      "step": 33040
    },
    {
      "entropy": 5.6245198249816895,
      "epoch": 3.711293311619026,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003682266131030375,
      "loss": 4.9063,
      "mean_token_accuracy": 0.23458320200443267,
      "num_tokens": 73782698.0,
      "step": 33045
    },
    {
      "entropy": 5.542952537536621,
      "epoch": 3.7118548885269838,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036819007138895525,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2293461576104164,
      "num_tokens": 73793758.0,
      "step": 33050
    },
    {
      "entropy": 5.635536623001099,
      "epoch": 3.7124164654349414,
      "grad_norm": 1.171875,
      "learning_rate": 0.000368153526707665,
      "loss": 5.0021,
      "mean_token_accuracy": 0.22026467621326445,
      "num_tokens": 73805479.0,
      "step": 33055
    },
    {
      "entropy": 5.628945636749267,
      "epoch": 3.7129780423428986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003681169790603303,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22020086646080017,
      "num_tokens": 73817361.0,
      "step": 33060
    },
    {
      "entropy": 5.5966307640075685,
      "epoch": 3.7135396192508563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036808042844811484,
      "loss": 4.9897,
      "mean_token_accuracy": 0.22329289764165877,
      "num_tokens": 73829651.0,
      "step": 33065
    },
    {
      "entropy": 5.570048713684082,
      "epoch": 3.714101196158814,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22906951159238814,
      "num_tokens": 73839608.0,
      "step": 33070
    },
    {
      "entropy": 5.617272901535034,
      "epoch": 3.7146627730667716,
      "grad_norm": 1.109375,
      "learning_rate": 0.000368007318333697,
      "loss": 4.966,
      "mean_token_accuracy": 0.22839832603931426,
      "num_tokens": 73850663.0,
      "step": 33075
    },
    {
      "entropy": 5.672142457962036,
      "epoch": 3.7152243499747293,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003679707588338225,
      "loss": 4.9937,
      "mean_token_accuracy": 0.227647265791893,
      "num_tokens": 73862441.0,
      "step": 33080
    },
    {
      "entropy": 5.654837942123413,
      "epoch": 3.7157859268826865,
      "grad_norm": 1.21875,
      "learning_rate": 0.000367934196373723,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2320241943001747,
      "num_tokens": 73873549.0,
      "step": 33085
    },
    {
      "entropy": 5.596856784820557,
      "epoch": 3.716347503790644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036789763095456264,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2268900603055954,
      "num_tokens": 73884872.0,
      "step": 33090
    },
    {
      "entropy": 5.677890920639038,
      "epoch": 3.716909080698602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003678610625775056,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20989345610141755,
      "num_tokens": 73896137.0,
      "step": 33095
    },
    {
      "entropy": 5.56166443824768,
      "epoch": 3.717470657606559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036782449124371643,
      "loss": 4.8669,
      "mean_token_accuracy": 0.22368674874305725,
      "num_tokens": 73906135.0,
      "step": 33100
    },
    {
      "entropy": 5.694778108596802,
      "epoch": 3.7180322345145167,
      "grad_norm": 1.0,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2260030373930931,
      "num_tokens": 73916856.0,
      "step": 33105
    },
    {
      "entropy": 5.521934366226196,
      "epoch": 3.7185938114224744,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036775133971059904,
      "loss": 4.8899,
      "mean_token_accuracy": 0.23564945608377458,
      "num_tokens": 73928268.0,
      "step": 33110
    },
    {
      "entropy": 5.636529731750488,
      "epoch": 3.7191553883304316,
      "grad_norm": 1.125,
      "learning_rate": 0.0003677147595136001,
      "loss": 5.0229,
      "mean_token_accuracy": 0.22882362306118012,
      "num_tokens": 73940327.0,
      "step": 33115
    },
    {
      "entropy": 5.727239418029785,
      "epoch": 3.7197169652383892,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036767817636452723,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2280044972896576,
      "num_tokens": 73951488.0,
      "step": 33120
    },
    {
      "entropy": 5.579854154586792,
      "epoch": 3.720278542146347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036764159026454534,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22693801969289779,
      "num_tokens": 73963479.0,
      "step": 33125
    },
    {
      "entropy": 5.521704196929932,
      "epoch": 3.7208401190543046,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003676050012148193,
      "loss": 4.8373,
      "mean_token_accuracy": 0.230353844165802,
      "num_tokens": 73974878.0,
      "step": 33130
    },
    {
      "entropy": 5.592448902130127,
      "epoch": 3.7214016959622622,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000367568409216514,
      "loss": 4.9799,
      "mean_token_accuracy": 0.22668465375900268,
      "num_tokens": 73985983.0,
      "step": 33135
    },
    {
      "entropy": 5.616083192825317,
      "epoch": 3.7219632728702194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003675318142707947,
      "loss": 5.0024,
      "mean_token_accuracy": 0.22042674869298934,
      "num_tokens": 73997141.0,
      "step": 33140
    },
    {
      "entropy": 5.6835895538330075,
      "epoch": 3.722524849778177,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22929033041000366,
      "num_tokens": 74008262.0,
      "step": 33145
    },
    {
      "entropy": 5.659400415420532,
      "epoch": 3.7230864266861348,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036745861554177486,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2268736854195595,
      "num_tokens": 74020465.0,
      "step": 33150
    },
    {
      "entropy": 5.642080974578858,
      "epoch": 3.723648003594092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036742201176080496,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22980567216873168,
      "num_tokens": 74031100.0,
      "step": 33155
    },
    {
      "entropy": 5.661765384674072,
      "epoch": 3.7242095805020496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003673854050370824,
      "loss": 5.0802,
      "mean_token_accuracy": 0.21660980135202407,
      "num_tokens": 74043313.0,
      "step": 33160
    },
    {
      "entropy": 5.594183683395386,
      "epoch": 3.7247711574100073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2256613239645958,
      "num_tokens": 74054540.0,
      "step": 33165
    },
    {
      "entropy": 5.695078325271607,
      "epoch": 3.725332734317965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003673121827660415,
      "loss": 5.1241,
      "mean_token_accuracy": 0.2196938067674637,
      "num_tokens": 74065636.0,
      "step": 33170
    },
    {
      "entropy": 5.605242300033569,
      "epoch": 3.7258943112259226,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2278485804796219,
      "num_tokens": 74076516.0,
      "step": 33175
    },
    {
      "entropy": 5.666334629058838,
      "epoch": 3.72645588813388,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036723894873797786,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2247696578502655,
      "num_tokens": 74087481.0,
      "step": 33180
    },
    {
      "entropy": 5.715510702133178,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036720232731797726,
      "loss": 5.0145,
      "mean_token_accuracy": 0.223515060544014,
      "num_tokens": 74098577.0,
      "step": 33185
    },
    {
      "entropy": 5.710046720504761,
      "epoch": 3.727579041949795,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036716570296221873,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21255189925432205,
      "num_tokens": 74108813.0,
      "step": 33190
    },
    {
      "entropy": 5.600740051269531,
      "epoch": 3.7281406188577524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22186410129070283,
      "num_tokens": 74118937.0,
      "step": 33195
    },
    {
      "entropy": 5.601558351516724,
      "epoch": 3.72870219576571,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003670924454480928,
      "loss": 4.9807,
      "mean_token_accuracy": 0.22355486899614335,
      "num_tokens": 74129879.0,
      "step": 33200
    },
    {
      "entropy": 5.602246522903442,
      "epoch": 3.7292637726736677,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2291163206100464,
      "num_tokens": 74140455.0,
      "step": 33205
    },
    {
      "entropy": 5.653490018844605,
      "epoch": 3.729825349581625,
      "grad_norm": 1.125,
      "learning_rate": 0.00036701917620493026,
      "loss": 5.0146,
      "mean_token_accuracy": 0.22486889660358428,
      "num_tokens": 74152508.0,
      "step": 33210
    },
    {
      "entropy": 5.678843212127686,
      "epoch": 3.7303869264895826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003669825371878764,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2349158078432083,
      "num_tokens": 74163700.0,
      "step": 33215
    },
    {
      "entropy": 5.59092526435852,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036694589524206275,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22716458588838578,
      "num_tokens": 74174715.0,
      "step": 33220
    },
    {
      "entropy": 5.650402641296386,
      "epoch": 3.731510080305498,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003669092503686562,
      "loss": 5.0012,
      "mean_token_accuracy": 0.23013211488723756,
      "num_tokens": 74186679.0,
      "step": 33225
    },
    {
      "entropy": 5.667631196975708,
      "epoch": 3.7320716572134556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2284498080611229,
      "num_tokens": 74196382.0,
      "step": 33230
    },
    {
      "entropy": 5.666950273513794,
      "epoch": 3.732633234121413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036683595184373147,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22569028586149215,
      "num_tokens": 74207587.0,
      "step": 33235
    },
    {
      "entropy": 5.598141050338745,
      "epoch": 3.7331948110293705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.943,
      "mean_token_accuracy": 0.22658098489046097,
      "num_tokens": 74218826.0,
      "step": 33240
    },
    {
      "entropy": 5.7237263202667235,
      "epoch": 3.733756387937328,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036676264162243744,
      "loss": 5.0543,
      "mean_token_accuracy": 0.22202853560447694,
      "num_tokens": 74230450.0,
      "step": 33245
    },
    {
      "entropy": 5.684976863861084,
      "epoch": 3.7343179648452853,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003667259821285698,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21876973360776902,
      "num_tokens": 74241438.0,
      "step": 33250
    },
    {
      "entropy": 5.694694948196411,
      "epoch": 3.734879541753243,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003666893197141112,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21931712925434113,
      "num_tokens": 74252460.0,
      "step": 33255
    },
    {
      "entropy": 5.651597118377685,
      "epoch": 3.7354411186612007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003666526543802291,
      "loss": 4.9337,
      "mean_token_accuracy": 0.23123737126588823,
      "num_tokens": 74262984.0,
      "step": 33260
    },
    {
      "entropy": 5.633333826065064,
      "epoch": 3.7360026955691583,
      "grad_norm": 1.09375,
      "learning_rate": 0.000366615986128091,
      "loss": 5.0739,
      "mean_token_accuracy": 0.21804485470056534,
      "num_tokens": 74274368.0,
      "step": 33265
    },
    {
      "entropy": 5.586025857925415,
      "epoch": 3.736564272477116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003665793149588644,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2332453966140747,
      "num_tokens": 74286394.0,
      "step": 33270
    },
    {
      "entropy": 5.57075686454773,
      "epoch": 3.737125849385073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22990977317094802,
      "num_tokens": 74296803.0,
      "step": 33275
    },
    {
      "entropy": 5.658049774169922,
      "epoch": 3.737687426293031,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036650596387381615,
      "loss": 4.9829,
      "mean_token_accuracy": 0.23030478060245513,
      "num_tokens": 74306471.0,
      "step": 33280
    },
    {
      "entropy": 5.645783758163452,
      "epoch": 3.7382490032009885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036646928396033,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23234597891569136,
      "num_tokens": 74318875.0,
      "step": 33285
    },
    {
      "entropy": 5.66106104850769,
      "epoch": 3.7388105801089457,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036643260113442634,
      "loss": 4.992,
      "mean_token_accuracy": 0.2229016274213791,
      "num_tokens": 74328464.0,
      "step": 33290
    },
    {
      "entropy": 5.6911052703857425,
      "epoch": 3.7393721570169034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003663959153972732,
      "loss": 4.9673,
      "mean_token_accuracy": 0.23234632462263108,
      "num_tokens": 74339462.0,
      "step": 33295
    },
    {
      "entropy": 5.685894584655761,
      "epoch": 3.739933733924861,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003663592267500388,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22474826127290726,
      "num_tokens": 74350667.0,
      "step": 33300
    },
    {
      "entropy": 5.612080192565918,
      "epoch": 3.7404953108328187,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003663225351938911,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22646359205245972,
      "num_tokens": 74361734.0,
      "step": 33305
    },
    {
      "entropy": 5.626529598236084,
      "epoch": 3.741056887740776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003662858407299984,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21572073698043823,
      "num_tokens": 74373465.0,
      "step": 33310
    },
    {
      "entropy": 5.632554388046264,
      "epoch": 3.7416184646487336,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036624914335952924,
      "loss": 4.9485,
      "mean_token_accuracy": 0.22214386612176895,
      "num_tokens": 74384853.0,
      "step": 33315
    },
    {
      "entropy": 5.580294370651245,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036621244308365195,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22262656688690186,
      "num_tokens": 74394356.0,
      "step": 33320
    },
    {
      "entropy": 5.666336536407471,
      "epoch": 3.742741618464649,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036617573990353503,
      "loss": 5.036,
      "mean_token_accuracy": 0.2175791531801224,
      "num_tokens": 74406721.0,
      "step": 33325
    },
    {
      "entropy": 5.64500060081482,
      "epoch": 3.743303195372606,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003661390338203472,
      "loss": 4.9887,
      "mean_token_accuracy": 0.222313991189003,
      "num_tokens": 74418217.0,
      "step": 33330
    },
    {
      "entropy": 5.595300006866455,
      "epoch": 3.743864772280564,
      "grad_norm": 1.125,
      "learning_rate": 0.0003661023248352572,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22968244850635527,
      "num_tokens": 74428644.0,
      "step": 33335
    },
    {
      "entropy": 5.571133804321289,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23081299066543579,
      "num_tokens": 74439575.0,
      "step": 33340
    },
    {
      "entropy": 5.661177015304565,
      "epoch": 3.7449879260964787,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036602889816404613,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2258031725883484,
      "num_tokens": 74450191.0,
      "step": 33345
    },
    {
      "entropy": 5.681750202178955,
      "epoch": 3.7455495030044363,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036599218048026294,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22801845669746398,
      "num_tokens": 74460816.0,
      "step": 33350
    },
    {
      "entropy": 5.6600728034973145,
      "epoch": 3.746111079912394,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036595545989925345,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2338688477873802,
      "num_tokens": 74470999.0,
      "step": 33355
    },
    {
      "entropy": 5.605210542678833,
      "epoch": 3.7466726568203517,
      "grad_norm": 1.125,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.8012,
      "mean_token_accuracy": 0.2415744036436081,
      "num_tokens": 74480884.0,
      "step": 33360
    },
    {
      "entropy": 5.558702898025513,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.125,
      "learning_rate": 0.00036588201005023223,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22493193447589874,
      "num_tokens": 74492500.0,
      "step": 33365
    },
    {
      "entropy": 5.591831493377685,
      "epoch": 3.7477958106362665,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003658452807845593,
      "loss": 4.9316,
      "mean_token_accuracy": 0.23056563585996628,
      "num_tokens": 74504635.0,
      "step": 33370
    },
    {
      "entropy": 5.673161315917969,
      "epoch": 3.748357387544224,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036580854862633747,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2197078600525856,
      "num_tokens": 74515743.0,
      "step": 33375
    },
    {
      "entropy": 5.5942955493927,
      "epoch": 3.748918964452182,
      "grad_norm": 1.125,
      "learning_rate": 0.0003657718135767362,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22089087665081025,
      "num_tokens": 74528089.0,
      "step": 33380
    },
    {
      "entropy": 5.706080532073974,
      "epoch": 3.749480541360139,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003657350756369251,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22851069271564484,
      "num_tokens": 74538543.0,
      "step": 33385
    },
    {
      "entropy": 5.740759372711182,
      "epoch": 3.7500421182680967,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036569833480807403,
      "loss": 5.0429,
      "mean_token_accuracy": 0.22525286227464675,
      "num_tokens": 74549825.0,
      "step": 33390
    },
    {
      "entropy": 5.582098865509034,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003656615910913528,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22838485836982728,
      "num_tokens": 74560304.0,
      "step": 33395
    },
    {
      "entropy": 5.522641515731811,
      "epoch": 3.751165272084012,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036562484448793135,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23058775663375855,
      "num_tokens": 74571549.0,
      "step": 33400
    },
    {
      "entropy": 5.688288164138794,
      "epoch": 3.7517268489919693,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003655880949989798,
      "loss": 5.0147,
      "mean_token_accuracy": 0.22302559316158294,
      "num_tokens": 74581865.0,
      "step": 33405
    },
    {
      "entropy": 5.623798370361328,
      "epoch": 3.752288425899927,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036555134262566806,
      "loss": 4.8436,
      "mean_token_accuracy": 0.22597986906766893,
      "num_tokens": 74592079.0,
      "step": 33410
    },
    {
      "entropy": 5.608408975601196,
      "epoch": 3.7528500028078846,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003655145873691665,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22567484825849532,
      "num_tokens": 74603058.0,
      "step": 33415
    },
    {
      "entropy": 5.636031770706177,
      "epoch": 3.7534115797158423,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036547782923064537,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22420779317617417,
      "num_tokens": 74614008.0,
      "step": 33420
    },
    {
      "entropy": 5.743435049057007,
      "epoch": 3.7539731566237995,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036544106821127505,
      "loss": 5.064,
      "mean_token_accuracy": 0.21418398320674897,
      "num_tokens": 74625146.0,
      "step": 33425
    },
    {
      "entropy": 5.705688571929931,
      "epoch": 3.754534733531757,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000365404304312226,
      "loss": 5.0606,
      "mean_token_accuracy": 0.22429585158824922,
      "num_tokens": 74636578.0,
      "step": 33430
    },
    {
      "entropy": 5.616029596328735,
      "epoch": 3.755096310439715,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036536753753466897,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22165464460849763,
      "num_tokens": 74648551.0,
      "step": 33435
    },
    {
      "entropy": 5.616789388656616,
      "epoch": 3.755657887347672,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003653307678797744,
      "loss": 4.833,
      "mean_token_accuracy": 0.23674579560756684,
      "num_tokens": 74658712.0,
      "step": 33440
    },
    {
      "entropy": 5.60693736076355,
      "epoch": 3.7562194642556297,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003652939953487131,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2262820705771446,
      "num_tokens": 74670196.0,
      "step": 33445
    },
    {
      "entropy": 5.684357786178589,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036525721994265606,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2242514669895172,
      "num_tokens": 74680577.0,
      "step": 33450
    },
    {
      "entropy": 5.667964458465576,
      "epoch": 3.757342618071545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23148188441991807,
      "num_tokens": 74692054.0,
      "step": 33455
    },
    {
      "entropy": 5.623154211044311,
      "epoch": 3.7579041949795027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22452137768268585,
      "num_tokens": 74703324.0,
      "step": 33460
    },
    {
      "entropy": 5.573706102371216,
      "epoch": 3.75846577188746,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.925,
      "mean_token_accuracy": 0.23048687875270843,
      "num_tokens": 74713028.0,
      "step": 33465
    },
    {
      "entropy": 5.5716901302337645,
      "epoch": 3.7590273487954176,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036511008959188966,
      "loss": 4.925,
      "mean_token_accuracy": 0.2217716947197914,
      "num_tokens": 74723964.0,
      "step": 33470
    },
    {
      "entropy": 5.6384728908538815,
      "epoch": 3.759588925703375,
      "grad_norm": 1.125,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2276859238743782,
      "num_tokens": 74734928.0,
      "step": 33475
    },
    {
      "entropy": 5.734301900863647,
      "epoch": 3.7601505026113324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003650365071969803,
      "loss": 5.0889,
      "mean_token_accuracy": 0.22043924927711486,
      "num_tokens": 74747198.0,
      "step": 33480
    },
    {
      "entropy": 5.672165632247925,
      "epoch": 3.76071207951929,
      "grad_norm": 1.25,
      "learning_rate": 0.0003649997116987439,
      "loss": 4.9762,
      "mean_token_accuracy": 0.23305270969867706,
      "num_tokens": 74758180.0,
      "step": 33485
    },
    {
      "entropy": 5.653761386871338,
      "epoch": 3.7612736564272478,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003649629133348818,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22656406760215758,
      "num_tokens": 74769227.0,
      "step": 33490
    },
    {
      "entropy": 5.625983953475952,
      "epoch": 3.7618352333352054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2317298337817192,
      "num_tokens": 74780347.0,
      "step": 33495
    },
    {
      "entropy": 5.598511123657227,
      "epoch": 3.7623968102431626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2255079999566078,
      "num_tokens": 74791902.0,
      "step": 33500
    },
    {
      "entropy": 5.605039978027344,
      "epoch": 3.7629583871511203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036485250106125844,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22473534792661667,
      "num_tokens": 74803530.0,
      "step": 33505
    },
    {
      "entropy": 5.635183429718017,
      "epoch": 3.763519964059078,
      "grad_norm": 1.21875,
      "learning_rate": 0.000364815691246611,
      "loss": 4.9386,
      "mean_token_accuracy": 0.23302847146987915,
      "num_tokens": 74815128.0,
      "step": 33510
    },
    {
      "entropy": 5.718488454818726,
      "epoch": 3.7640815409670356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036477887857219726,
      "loss": 5.1135,
      "mean_token_accuracy": 0.22440395206212999,
      "num_tokens": 74827729.0,
      "step": 33515
    },
    {
      "entropy": 5.577252149581909,
      "epoch": 3.764643117874993,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036474206303918914,
      "loss": 4.91,
      "mean_token_accuracy": 0.2278133064508438,
      "num_tokens": 74837922.0,
      "step": 33520
    },
    {
      "entropy": 5.631310653686524,
      "epoch": 3.7652046947829505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000364705244648759,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22971495389938354,
      "num_tokens": 74848239.0,
      "step": 33525
    },
    {
      "entropy": 5.6818687438964846,
      "epoch": 3.765766271690908,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003646684234020789,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22547849714756013,
      "num_tokens": 74860454.0,
      "step": 33530
    },
    {
      "entropy": 5.590918111801147,
      "epoch": 3.7663278485988654,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22432397603988646,
      "num_tokens": 74871849.0,
      "step": 33535
    },
    {
      "entropy": 5.6630809783935545,
      "epoch": 3.766889425506823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036459477234465935,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21670203506946564,
      "num_tokens": 74883877.0,
      "step": 33540
    },
    {
      "entropy": 5.841873502731323,
      "epoch": 3.7674510024147807,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036455794253626486,
      "loss": 5.2001,
      "mean_token_accuracy": 0.21795394271612167,
      "num_tokens": 74897218.0,
      "step": 33545
    },
    {
      "entropy": 5.715346336364746,
      "epoch": 3.7680125793227384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036452110987631066,
      "loss": 4.9762,
      "mean_token_accuracy": 0.22343896627426146,
      "num_tokens": 74908708.0,
      "step": 33550
    },
    {
      "entropy": 5.651488876342773,
      "epoch": 3.768574156230696,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.923,
      "mean_token_accuracy": 0.2335190460085869,
      "num_tokens": 74919664.0,
      "step": 33555
    },
    {
      "entropy": 5.583813285827636,
      "epoch": 3.7691357331386532,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.9584,
      "mean_token_accuracy": 0.23020050376653672,
      "num_tokens": 74930386.0,
      "step": 33560
    },
    {
      "entropy": 5.582110357284546,
      "epoch": 3.769697310046611,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22379569709300995,
      "num_tokens": 74942835.0,
      "step": 33565
    },
    {
      "entropy": 5.681372022628784,
      "epoch": 3.7702588869545686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036437375074435446,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2220294326543808,
      "num_tokens": 74954332.0,
      "step": 33570
    },
    {
      "entropy": 5.7078697204589846,
      "epoch": 3.770820463862526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003643369038441955,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22257156819105148,
      "num_tokens": 74965530.0,
      "step": 33575
    },
    {
      "entropy": 5.635831069946289,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036430005409951476,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22371526062488556,
      "num_tokens": 74976319.0,
      "step": 33580
    },
    {
      "entropy": 5.641131973266601,
      "epoch": 3.771943617678441,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036426320151148575,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22222513407468797,
      "num_tokens": 74987948.0,
      "step": 33585
    },
    {
      "entropy": 5.641069936752319,
      "epoch": 3.7725051945863988,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036422634608128184,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22940164506435395,
      "num_tokens": 74998989.0,
      "step": 33590
    },
    {
      "entropy": 5.639626932144165,
      "epoch": 3.773066771494356,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036418948781007636,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22756045311689377,
      "num_tokens": 75010510.0,
      "step": 33595
    },
    {
      "entropy": 5.6510241508483885,
      "epoch": 3.7736283484023136,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036415262669904305,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21850184351205826,
      "num_tokens": 75020708.0,
      "step": 33600
    },
    {
      "entropy": 5.672294569015503,
      "epoch": 3.7741899253102713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036411576274935546,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22748733460903167,
      "num_tokens": 75031468.0,
      "step": 33605
    },
    {
      "entropy": 5.5780304908752445,
      "epoch": 3.774751502218229,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.8671,
      "mean_token_accuracy": 0.23415984064340592,
      "num_tokens": 75042170.0,
      "step": 33610
    },
    {
      "entropy": 5.582749271392823,
      "epoch": 3.775313079126186,
      "grad_norm": 1.140625,
      "learning_rate": 0.000364042026338713,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2286226361989975,
      "num_tokens": 75053916.0,
      "step": 33615
    },
    {
      "entropy": 5.6280676364898685,
      "epoch": 3.775874656034144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036400515388010574,
      "loss": 4.9191,
      "mean_token_accuracy": 0.235379159450531,
      "num_tokens": 75065026.0,
      "step": 33620
    },
    {
      "entropy": 5.5831419944763185,
      "epoch": 3.7764362329421015,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003639682785875399,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22333890795707703,
      "num_tokens": 75075617.0,
      "step": 33625
    },
    {
      "entropy": 5.663079166412354,
      "epoch": 3.7769978098500587,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036393140046218957,
      "loss": 4.9703,
      "mean_token_accuracy": 0.2203453302383423,
      "num_tokens": 75085931.0,
      "step": 33630
    },
    {
      "entropy": 5.712485504150391,
      "epoch": 3.7775593867580164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036389451950522895,
      "loss": 4.9873,
      "mean_token_accuracy": 0.23152564018964766,
      "num_tokens": 75096404.0,
      "step": 33635
    },
    {
      "entropy": 5.723748445510864,
      "epoch": 3.778120963665974,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036385763571783236,
      "loss": 5.0616,
      "mean_token_accuracy": 0.22499655932188034,
      "num_tokens": 75108300.0,
      "step": 33640
    },
    {
      "entropy": 5.663790512084961,
      "epoch": 3.7786825405739317,
      "grad_norm": 1.125,
      "learning_rate": 0.0003638207491011742,
      "loss": 5.0048,
      "mean_token_accuracy": 0.22460048794746398,
      "num_tokens": 75119799.0,
      "step": 33645
    },
    {
      "entropy": 5.592073726654053,
      "epoch": 3.7792441174818894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22830481976270675,
      "num_tokens": 75130803.0,
      "step": 33650
    },
    {
      "entropy": 5.6639965057373045,
      "epoch": 3.7798056943898466,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036374696738477105,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23108194321393966,
      "num_tokens": 75140621.0,
      "step": 33655
    },
    {
      "entropy": 5.641145992279053,
      "epoch": 3.7803672712978043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003637100722873754,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21867793202400207,
      "num_tokens": 75152446.0,
      "step": 33660
    },
    {
      "entropy": 5.687719774246216,
      "epoch": 3.780928848205762,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22355817407369613,
      "num_tokens": 75162936.0,
      "step": 33665
    },
    {
      "entropy": 5.552319288253784,
      "epoch": 3.781490425113719,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.9104,
      "mean_token_accuracy": 0.23029358237981795,
      "num_tokens": 75174184.0,
      "step": 33670
    },
    {
      "entropy": 5.485078144073486,
      "epoch": 3.782052002021677,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.8617,
      "mean_token_accuracy": 0.23626603037118912,
      "num_tokens": 75185609.0,
      "step": 33675
    },
    {
      "entropy": 5.7207958698272705,
      "epoch": 3.7826135789296345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036356246366391054,
      "loss": 5.0469,
      "mean_token_accuracy": 0.22147736698389053,
      "num_tokens": 75196098.0,
      "step": 33680
    },
    {
      "entropy": 5.754191493988037,
      "epoch": 3.783175155837592,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036352555445544864,
      "loss": 5.0345,
      "mean_token_accuracy": 0.22461846023797988,
      "num_tokens": 75206411.0,
      "step": 33685
    },
    {
      "entropy": 5.617139673233032,
      "epoch": 3.78373673274555,
      "grad_norm": 1.125,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22717685401439666,
      "num_tokens": 75217029.0,
      "step": 33690
    },
    {
      "entropy": 5.522969579696655,
      "epoch": 3.784298309653507,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22271322906017305,
      "num_tokens": 75228124.0,
      "step": 33695
    },
    {
      "entropy": 5.545496559143066,
      "epoch": 3.7848598865614647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003634148099226364,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2187255635857582,
      "num_tokens": 75239776.0,
      "step": 33700
    },
    {
      "entropy": 5.59564356803894,
      "epoch": 3.7854214634694223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22780342251062394,
      "num_tokens": 75251354.0,
      "step": 33705
    },
    {
      "entropy": 5.68123083114624,
      "epoch": 3.7859830403773795,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003633409661563267,
      "loss": 5.006,
      "mean_token_accuracy": 0.21771450936794282,
      "num_tokens": 75263167.0,
      "step": 33710
    },
    {
      "entropy": 5.597994804382324,
      "epoch": 3.786544617285337,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.9207,
      "mean_token_accuracy": 0.23408613502979278,
      "num_tokens": 75274200.0,
      "step": 33715
    },
    {
      "entropy": 5.651874685287476,
      "epoch": 3.787106194193295,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003632671111387746,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2232692778110504,
      "num_tokens": 75287060.0,
      "step": 33720
    },
    {
      "entropy": 5.735655355453491,
      "epoch": 3.787667771101252,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036323017941372203,
      "loss": 5.0163,
      "mean_token_accuracy": 0.23096109181642532,
      "num_tokens": 75297862.0,
      "step": 33725
    },
    {
      "entropy": 5.699450969696045,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003631932448793864,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22092121839523315,
      "num_tokens": 75308844.0,
      "step": 33730
    },
    {
      "entropy": 5.69988522529602,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003631563075369436,
      "loss": 5.0716,
      "mean_token_accuracy": 0.21638896614313125,
      "num_tokens": 75321035.0,
      "step": 33735
    },
    {
      "entropy": 5.584645318984985,
      "epoch": 3.789352501825125,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22926487922668456,
      "num_tokens": 75331407.0,
      "step": 33740
    },
    {
      "entropy": 5.6060545444488525,
      "epoch": 3.7899140787330827,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2279842585325241,
      "num_tokens": 75343082.0,
      "step": 33745
    },
    {
      "entropy": 5.624531555175781,
      "epoch": 3.79047565564104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.9065,
      "mean_token_accuracy": 0.232708403468132,
      "num_tokens": 75353345.0,
      "step": 33750
    },
    {
      "entropy": 5.693700742721558,
      "epoch": 3.7910372325489976,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003630085301096245,
      "loss": 5.0686,
      "mean_token_accuracy": 0.22483919858932494,
      "num_tokens": 75364039.0,
      "step": 33755
    },
    {
      "entropy": 5.718462991714477,
      "epoch": 3.7915988094569553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036297157874428933,
      "loss": 4.9827,
      "mean_token_accuracy": 0.22805007100105285,
      "num_tokens": 75374166.0,
      "step": 33760
    },
    {
      "entropy": 5.653206157684326,
      "epoch": 3.7921603863649125,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036293462457790505,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22355058640241623,
      "num_tokens": 75384792.0,
      "step": 33765
    },
    {
      "entropy": 5.65735969543457,
      "epoch": 3.79272196327287,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036289766761164814,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21863144040107726,
      "num_tokens": 75395478.0,
      "step": 33770
    },
    {
      "entropy": 5.652249956130982,
      "epoch": 3.793283540180828,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036286070784669543,
      "loss": 5.0161,
      "mean_token_accuracy": 0.223700113594532,
      "num_tokens": 75406293.0,
      "step": 33775
    },
    {
      "entropy": 5.6881303787231445,
      "epoch": 3.7938451170887855,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036282374528422373,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2341945067048073,
      "num_tokens": 75417588.0,
      "step": 33780
    },
    {
      "entropy": 5.582632398605346,
      "epoch": 3.794406693996743,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2274443179368973,
      "num_tokens": 75428651.0,
      "step": 33785
    },
    {
      "entropy": 5.684594058990479,
      "epoch": 3.7949682709047003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000362749811771431,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2241087883710861,
      "num_tokens": 75438808.0,
      "step": 33790
    },
    {
      "entropy": 5.63442587852478,
      "epoch": 3.795529847812658,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22389754354953767,
      "num_tokens": 75449326.0,
      "step": 33795
    },
    {
      "entropy": 5.623478364944458,
      "epoch": 3.7960914247206157,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003626758670826861,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2225932076573372,
      "num_tokens": 75460499.0,
      "step": 33800
    },
    {
      "entropy": 5.599629592895508,
      "epoch": 3.796653001628573,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003626388905502747,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2286992073059082,
      "num_tokens": 75471031.0,
      "step": 33805
    },
    {
      "entropy": 5.632548570632935,
      "epoch": 3.7972145785365305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000362601911227407,
      "loss": 4.9785,
      "mean_token_accuracy": 0.22424534112215042,
      "num_tokens": 75483608.0,
      "step": 33810
    },
    {
      "entropy": 5.696910381317139,
      "epoch": 3.797776155444488,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003625649291152604,
      "loss": 4.9955,
      "mean_token_accuracy": 0.23328544944524765,
      "num_tokens": 75495221.0,
      "step": 33815
    },
    {
      "entropy": 5.6120655059814455,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036252794421501257,
      "loss": 4.911,
      "mean_token_accuracy": 0.23255457431077958,
      "num_tokens": 75507245.0,
      "step": 33820
    },
    {
      "entropy": 5.7289183139801025,
      "epoch": 3.798899309260403,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036249095652784096,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21710752993822097,
      "num_tokens": 75518525.0,
      "step": 33825
    },
    {
      "entropy": 5.663200759887696,
      "epoch": 3.7994608861683608,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036245396605492344,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21936965733766556,
      "num_tokens": 75529674.0,
      "step": 33830
    },
    {
      "entropy": 5.5640363693237305,
      "epoch": 3.8000224630763184,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23420652598142624,
      "num_tokens": 75539596.0,
      "step": 33835
    },
    {
      "entropy": 5.554714345932007,
      "epoch": 3.800584039984276,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036237997675656167,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22812998592853545,
      "num_tokens": 75551164.0,
      "step": 33840
    },
    {
      "entropy": 5.669698762893677,
      "epoch": 3.8011456168922333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003623429779334732,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2202707603573799,
      "num_tokens": 75563758.0,
      "step": 33845
    },
    {
      "entropy": 5.744740056991577,
      "epoch": 3.801707193800191,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036230597632935053,
      "loss": 5.0828,
      "mean_token_accuracy": 0.2253193661570549,
      "num_tokens": 75574332.0,
      "step": 33850
    },
    {
      "entropy": 5.682096815109253,
      "epoch": 3.8022687707081486,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003622689719453717,
      "loss": 5.068,
      "mean_token_accuracy": 0.21802436113357543,
      "num_tokens": 75586229.0,
      "step": 33855
    },
    {
      "entropy": 5.673987579345703,
      "epoch": 3.802830347616106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036223196478271495,
      "loss": 4.99,
      "mean_token_accuracy": 0.22717526853084563,
      "num_tokens": 75597396.0,
      "step": 33860
    },
    {
      "entropy": 5.670917320251465,
      "epoch": 3.8033919245240635,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003621949548425585,
      "loss": 5.033,
      "mean_token_accuracy": 0.22046541273593903,
      "num_tokens": 75609171.0,
      "step": 33865
    },
    {
      "entropy": 5.723601579666138,
      "epoch": 3.803953501432021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000362157942126081,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2286200076341629,
      "num_tokens": 75619855.0,
      "step": 33870
    },
    {
      "entropy": 5.759358072280884,
      "epoch": 3.804515078339979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003621209266344608,
      "loss": 5.0544,
      "mean_token_accuracy": 0.221026448905468,
      "num_tokens": 75631810.0,
      "step": 33875
    },
    {
      "entropy": 5.60663685798645,
      "epoch": 3.8050766552479365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003620839083688765,
      "loss": 4.9815,
      "mean_token_accuracy": 0.228364397585392,
      "num_tokens": 75644894.0,
      "step": 33880
    },
    {
      "entropy": 5.6446404457092285,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22928918302059173,
      "num_tokens": 75657357.0,
      "step": 33885
    },
    {
      "entropy": 5.651206159591675,
      "epoch": 3.8061998090638514,
      "grad_norm": 1.125,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2275436267256737,
      "num_tokens": 75667357.0,
      "step": 33890
    },
    {
      "entropy": 5.618384218215942,
      "epoch": 3.806761385971809,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2271783947944641,
      "num_tokens": 75678758.0,
      "step": 33895
    },
    {
      "entropy": 5.69348554611206,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22802987545728684,
      "num_tokens": 75688392.0,
      "step": 33900
    },
    {
      "entropy": 5.6391904830932615,
      "epoch": 3.807884539787724,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036189877547275023,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22927262187004088,
      "num_tokens": 75699839.0,
      "step": 33905
    },
    {
      "entropy": 5.682568454742432,
      "epoch": 3.8084461166956816,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.9457,
      "mean_token_accuracy": 0.23073735237121581,
      "num_tokens": 75711877.0,
      "step": 33910
    },
    {
      "entropy": 5.589724206924439,
      "epoch": 3.8090076936036388,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22467228770256042,
      "num_tokens": 75721646.0,
      "step": 33915
    },
    {
      "entropy": 5.6617112159729,
      "epoch": 3.8095692705115964,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003617876625229543,
      "loss": 5.0705,
      "mean_token_accuracy": 0.21886153221130372,
      "num_tokens": 75733291.0,
      "step": 33920
    },
    {
      "entropy": 5.647152137756348,
      "epoch": 3.810130847419554,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.9596,
      "mean_token_accuracy": 0.23599462509155272,
      "num_tokens": 75744475.0,
      "step": 33925
    },
    {
      "entropy": 5.597776126861572,
      "epoch": 3.8106924243275118,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.951,
      "mean_token_accuracy": 0.22562574297189714,
      "num_tokens": 75755098.0,
      "step": 33930
    },
    {
      "entropy": 5.688869667053223,
      "epoch": 3.8112540012354694,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.9714,
      "mean_token_accuracy": 0.23103374242782593,
      "num_tokens": 75767405.0,
      "step": 33935
    },
    {
      "entropy": 5.6486146450042725,
      "epoch": 3.8118155781434266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036163947324179186,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21409243047237397,
      "num_tokens": 75778614.0,
      "step": 33940
    },
    {
      "entropy": 5.62540864944458,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.9224,
      "mean_token_accuracy": 0.23291261196136476,
      "num_tokens": 75789087.0,
      "step": 33945
    },
    {
      "entropy": 5.625563716888427,
      "epoch": 3.812938731959342,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2251404732465744,
      "num_tokens": 75799431.0,
      "step": 33950
    },
    {
      "entropy": 5.60964789390564,
      "epoch": 3.813500308867299,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036152830231113013,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22703890055418013,
      "num_tokens": 75810149.0,
      "step": 33955
    },
    {
      "entropy": 5.640748834609985,
      "epoch": 3.814061885775257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003614912398224802,
      "loss": 4.986,
      "mean_token_accuracy": 0.22285518497228624,
      "num_tokens": 75820072.0,
      "step": 33960
    },
    {
      "entropy": 5.6502344608306885,
      "epoch": 3.8146234626832145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003614541745799156,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21905476748943328,
      "num_tokens": 75831017.0,
      "step": 33965
    },
    {
      "entropy": 5.626031827926636,
      "epoch": 3.815185039591172,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003614171065846166,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22218097001314163,
      "num_tokens": 75842606.0,
      "step": 33970
    },
    {
      "entropy": 5.621058940887451,
      "epoch": 3.81574661649913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003613800358377635,
      "loss": 4.8926,
      "mean_token_accuracy": 0.23138351291418074,
      "num_tokens": 75853820.0,
      "step": 33975
    },
    {
      "entropy": 5.60271987915039,
      "epoch": 3.816308193407087,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2299947112798691,
      "num_tokens": 75864959.0,
      "step": 33980
    },
    {
      "entropy": 5.621663045883179,
      "epoch": 3.8168697703150447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036130588609411597,
      "loss": 4.882,
      "mean_token_accuracy": 0.23172407299280168,
      "num_tokens": 75877247.0,
      "step": 33985
    },
    {
      "entropy": 5.712012672424317,
      "epoch": 3.8174313472230024,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036126880709968264,
      "loss": 5.0786,
      "mean_token_accuracy": 0.21360162198543547,
      "num_tokens": 75888654.0,
      "step": 33990
    },
    {
      "entropy": 5.663818216323852,
      "epoch": 3.8179929241309596,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22351295202970506,
      "num_tokens": 75899953.0,
      "step": 33995
    },
    {
      "entropy": 5.6659125804901125,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003611946408714998,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2317900225520134,
      "num_tokens": 75911174.0,
      "step": 34000
    },
    {
      "entropy": 5.589650297164917,
      "epoch": 3.819116077946875,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.9443,
      "mean_token_accuracy": 0.2366092935204506,
      "num_tokens": 75922221.0,
      "step": 34005
    },
    {
      "entropy": 5.528977155685425,
      "epoch": 3.819677654854832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000361120463665434,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22304649651050568,
      "num_tokens": 75932630.0,
      "step": 34010
    },
    {
      "entropy": 5.607732677459717,
      "epoch": 3.82023923176279,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003610833709486471,
      "loss": 4.93,
      "mean_token_accuracy": 0.22752389013767244,
      "num_tokens": 75944490.0,
      "step": 34015
    },
    {
      "entropy": 5.6118604183197025,
      "epoch": 3.8208008086707475,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.9246,
      "mean_token_accuracy": 0.23030331134796142,
      "num_tokens": 75956608.0,
      "step": 34020
    },
    {
      "entropy": 5.66412992477417,
      "epoch": 3.821362385578705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036100917729347073,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2217874765396118,
      "num_tokens": 75967516.0,
      "step": 34025
    },
    {
      "entropy": 5.593828105926514,
      "epoch": 3.8219239624866628,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036097207635744363,
      "loss": 4.8696,
      "mean_token_accuracy": 0.22951493114233018,
      "num_tokens": 75978329.0,
      "step": 34030
    },
    {
      "entropy": 5.720518445968628,
      "epoch": 3.82248553939462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23173032253980635,
      "num_tokens": 75990170.0,
      "step": 34035
    },
    {
      "entropy": 5.621156120300293,
      "epoch": 3.8230471163025777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000360897866274418,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22858864068984985,
      "num_tokens": 76002368.0,
      "step": 34040
    },
    {
      "entropy": 5.584942436218261,
      "epoch": 3.8236086932105353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.9325,
      "mean_token_accuracy": 0.228273506462574,
      "num_tokens": 76012867.0,
      "step": 34045
    },
    {
      "entropy": 5.609594631195068,
      "epoch": 3.8241702701184925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036082364525130684,
      "loss": 5.0161,
      "mean_token_accuracy": 0.22223753929138185,
      "num_tokens": 76023900.0,
      "step": 34050
    },
    {
      "entropy": 5.609399604797363,
      "epoch": 3.82473184702645,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.8933,
      "mean_token_accuracy": 0.23461530804634095,
      "num_tokens": 76034296.0,
      "step": 34055
    },
    {
      "entropy": 5.692914390563965,
      "epoch": 3.825293423934408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036074941329756327,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22874194383621216,
      "num_tokens": 76045355.0,
      "step": 34060
    },
    {
      "entropy": 5.653396415710449,
      "epoch": 3.8258550008423655,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.8793,
      "mean_token_accuracy": 0.2363809660077095,
      "num_tokens": 76056158.0,
      "step": 34065
    },
    {
      "entropy": 5.61739764213562,
      "epoch": 3.826416577750323,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036067517042264147,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22597333937883377,
      "num_tokens": 76066930.0,
      "step": 34070
    },
    {
      "entropy": 5.644573926925659,
      "epoch": 3.8269781546582804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22755246013402938,
      "num_tokens": 76078791.0,
      "step": 34075
    },
    {
      "entropy": 5.694556045532226,
      "epoch": 3.827539731566238,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003606009166359972,
      "loss": 5.0271,
      "mean_token_accuracy": 0.22635923475027084,
      "num_tokens": 76090356.0,
      "step": 34080
    },
    {
      "entropy": 5.558886861801147,
      "epoch": 3.8281013084741957,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036056378565373434,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23403114676475525,
      "num_tokens": 76101137.0,
      "step": 34085
    },
    {
      "entropy": 5.528753709793091,
      "epoch": 3.828662885382153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.8201,
      "mean_token_accuracy": 0.2318944126367569,
      "num_tokens": 76111661.0,
      "step": 34090
    },
    {
      "entropy": 5.693347311019897,
      "epoch": 3.8292244622901106,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003604895155172386,
      "loss": 4.9701,
      "mean_token_accuracy": 0.22803875654935837,
      "num_tokens": 76122670.0,
      "step": 34095
    },
    {
      "entropy": 5.698940086364746,
      "epoch": 3.8297860391980683,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003604523763653706,
      "loss": 5.0713,
      "mean_token_accuracy": 0.22152069509029387,
      "num_tokens": 76133899.0,
      "step": 34100
    },
    {
      "entropy": 5.662249279022217,
      "epoch": 3.8303476161060255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036041523449266573,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2178792119026184,
      "num_tokens": 76144865.0,
      "step": 34105
    },
    {
      "entropy": 5.714614772796631,
      "epoch": 3.830909193013983,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036037808990030667,
      "loss": 5.0011,
      "mean_token_accuracy": 0.22563935965299606,
      "num_tokens": 76157392.0,
      "step": 34110
    },
    {
      "entropy": 5.706194114685059,
      "epoch": 3.831470769921941,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003603409425894761,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2236991450190544,
      "num_tokens": 76168339.0,
      "step": 34115
    },
    {
      "entropy": 5.579552793502808,
      "epoch": 3.8320323468298985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22942325174808503,
      "num_tokens": 76179377.0,
      "step": 34120
    },
    {
      "entropy": 5.614682674407959,
      "epoch": 3.832593923737856,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.95,
      "mean_token_accuracy": 0.22273012697696687,
      "num_tokens": 76190030.0,
      "step": 34125
    },
    {
      "entropy": 5.547447061538696,
      "epoch": 3.8331555006458133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22597415894269943,
      "num_tokens": 76201339.0,
      "step": 34130
    },
    {
      "entropy": 5.676809024810791,
      "epoch": 3.833717077553771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000360192326185096,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2291543796658516,
      "num_tokens": 76211760.0,
      "step": 34135
    },
    {
      "entropy": 5.682628870010376,
      "epoch": 3.8342786544617287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003601551652996514,
      "loss": 5.0173,
      "mean_token_accuracy": 0.22506078481674194,
      "num_tokens": 76222573.0,
      "step": 34140
    },
    {
      "entropy": 5.687285995483398,
      "epoch": 3.834840231369686,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003601180017028334,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21658241897821426,
      "num_tokens": 76233547.0,
      "step": 34145
    },
    {
      "entropy": 5.613495969772339,
      "epoch": 3.8354018082776435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003600808353958252,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22320879101753235,
      "num_tokens": 76246067.0,
      "step": 34150
    },
    {
      "entropy": 5.559151363372803,
      "epoch": 3.835963385185601,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003600436663798102,
      "loss": 4.8363,
      "mean_token_accuracy": 0.238016377389431,
      "num_tokens": 76257698.0,
      "step": 34155
    },
    {
      "entropy": 5.5502406597137455,
      "epoch": 3.836524962093559,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23324674367904663,
      "num_tokens": 76268533.0,
      "step": 34160
    },
    {
      "entropy": 5.564681816101074,
      "epoch": 3.8370865390015165,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23581327497959137,
      "num_tokens": 76278529.0,
      "step": 34165
    },
    {
      "entropy": 5.5838212966918945,
      "epoch": 3.8376481159094737,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2212705969810486,
      "num_tokens": 76289461.0,
      "step": 34170
    },
    {
      "entropy": 5.674845266342163,
      "epoch": 3.8382096928174314,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000359894963249353,
      "loss": 4.9466,
      "mean_token_accuracy": 0.23030066788196563,
      "num_tokens": 76300641.0,
      "step": 34175
    },
    {
      "entropy": 5.667232799530029,
      "epoch": 3.838771269725389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035985778070605785,
      "loss": 5.0439,
      "mean_token_accuracy": 0.21916007548570632,
      "num_tokens": 76310918.0,
      "step": 34180
    },
    {
      "entropy": 5.5853475570678714,
      "epoch": 3.8393328466333463,
      "grad_norm": 1.125,
      "learning_rate": 0.0003598205954608581,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23398132175207137,
      "num_tokens": 76323037.0,
      "step": 34185
    },
    {
      "entropy": 5.652505731582641,
      "epoch": 3.839894423541304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035978340751493767,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22367725521326065,
      "num_tokens": 76334252.0,
      "step": 34190
    },
    {
      "entropy": 5.6791728973388675,
      "epoch": 3.8404560004492616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035974621686948063,
      "loss": 5.0293,
      "mean_token_accuracy": 0.22460603415966035,
      "num_tokens": 76344928.0,
      "step": 34195
    },
    {
      "entropy": 5.674833726882935,
      "epoch": 3.841017577357219,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003597090235256712,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21757633984088898,
      "num_tokens": 76356529.0,
      "step": 34200
    },
    {
      "entropy": 5.6199716091156,
      "epoch": 3.8415791542651765,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22206085175275803,
      "num_tokens": 76366842.0,
      "step": 34205
    },
    {
      "entropy": 5.676772403717041,
      "epoch": 3.842140731173134,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003596346287477322,
      "loss": 4.9991,
      "mean_token_accuracy": 0.22076408118009566,
      "num_tokens": 76377109.0,
      "step": 34210
    },
    {
      "entropy": 5.631798839569091,
      "epoch": 3.842702308081092,
      "grad_norm": 1.125,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22121315747499465,
      "num_tokens": 76387621.0,
      "step": 34215
    },
    {
      "entropy": 5.674119758605957,
      "epoch": 3.8432638849890495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003595602231905957,
      "loss": 5.062,
      "mean_token_accuracy": 0.2156602069735527,
      "num_tokens": 76398537.0,
      "step": 34220
    },
    {
      "entropy": 5.654905986785889,
      "epoch": 3.8438254618970067,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.9425,
      "mean_token_accuracy": 0.23019365817308426,
      "num_tokens": 76410192.0,
      "step": 34225
    },
    {
      "entropy": 5.625110483169555,
      "epoch": 3.8443870388049644,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035948580686373803,
      "loss": 4.941,
      "mean_token_accuracy": 0.23117994964122773,
      "num_tokens": 76420024.0,
      "step": 34230
    },
    {
      "entropy": 5.643015146255493,
      "epoch": 3.844948615712922,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035944859466462554,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2336903139948845,
      "num_tokens": 76430833.0,
      "step": 34235
    },
    {
      "entropy": 5.60926308631897,
      "epoch": 3.8455101926208792,
      "grad_norm": 1.15625,
      "learning_rate": 0.000359411379776637,
      "loss": 4.9551,
      "mean_token_accuracy": 0.22463239580392838,
      "num_tokens": 76442800.0,
      "step": 34240
    },
    {
      "entropy": 5.6823282718658445,
      "epoch": 3.846071769528837,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035937416220095745,
      "loss": 4.9463,
      "mean_token_accuracy": 0.23056726902723312,
      "num_tokens": 76454988.0,
      "step": 34245
    },
    {
      "entropy": 5.750157833099365,
      "epoch": 3.8466333464367946,
      "grad_norm": 1.25,
      "learning_rate": 0.0003593369419387718,
      "loss": 5.0487,
      "mean_token_accuracy": 0.21820524483919143,
      "num_tokens": 76466523.0,
      "step": 34250
    },
    {
      "entropy": 5.737688875198364,
      "epoch": 3.847194923344752,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003592997189912652,
      "loss": 5.08,
      "mean_token_accuracy": 0.21154401302337647,
      "num_tokens": 76478691.0,
      "step": 34255
    },
    {
      "entropy": 5.676547002792359,
      "epoch": 3.84775650025271,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035926249335962296,
      "loss": 5.0551,
      "mean_token_accuracy": 0.22262391000986098,
      "num_tokens": 76490544.0,
      "step": 34260
    },
    {
      "entropy": 5.636931419372559,
      "epoch": 3.848318077160667,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.876,
      "mean_token_accuracy": 0.22966126054525376,
      "num_tokens": 76500269.0,
      "step": 34265
    },
    {
      "entropy": 5.563238334655762,
      "epoch": 3.8488796540686248,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003591880340486723,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22964812219142913,
      "num_tokens": 76511458.0,
      "step": 34270
    },
    {
      "entropy": 5.615138816833496,
      "epoch": 3.8494412309765824,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003591508003717347,
      "loss": 5.0206,
      "mean_token_accuracy": 0.22187732607126237,
      "num_tokens": 76521722.0,
      "step": 34275
    },
    {
      "entropy": 5.702325487136841,
      "epoch": 3.8500028078845396,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035911356401540306,
      "loss": 5.0533,
      "mean_token_accuracy": 0.23010772615671157,
      "num_tokens": 76532935.0,
      "step": 34280
    },
    {
      "entropy": 5.612896490097046,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22285665422677994,
      "num_tokens": 76543545.0,
      "step": 34285
    },
    {
      "entropy": 5.596207141876221,
      "epoch": 3.851125961700455,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2335416704416275,
      "num_tokens": 76553621.0,
      "step": 34290
    },
    {
      "entropy": 5.664820814132691,
      "epoch": 3.851687538608412,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035900183888190005,
      "loss": 4.9685,
      "mean_token_accuracy": 0.23303519189357758,
      "num_tokens": 76563154.0,
      "step": 34295
    },
    {
      "entropy": 5.67323260307312,
      "epoch": 3.85224911551637,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22219994962215422,
      "num_tokens": 76573930.0,
      "step": 34300
    },
    {
      "entropy": 5.625096654891967,
      "epoch": 3.8528106924243275,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035892734208433264,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22689452022314072,
      "num_tokens": 76585216.0,
      "step": 34305
    },
    {
      "entropy": 5.640145349502563,
      "epoch": 3.853372269332285,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035889008967653725,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2291739970445633,
      "num_tokens": 76596479.0,
      "step": 34310
    },
    {
      "entropy": 5.693407106399536,
      "epoch": 3.853933846240243,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003588528345976489,
      "loss": 5.0579,
      "mean_token_accuracy": 0.22097597569227218,
      "num_tokens": 76607123.0,
      "step": 34315
    },
    {
      "entropy": 5.733864974975586,
      "epoch": 3.8544954231482,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003588155768488537,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22865079641342162,
      "num_tokens": 76616898.0,
      "step": 34320
    },
    {
      "entropy": 5.631343221664428,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035877831643133803,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23587502241134645,
      "num_tokens": 76627528.0,
      "step": 34325
    },
    {
      "entropy": 5.633316469192505,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003587410533462883,
      "loss": 4.881,
      "mean_token_accuracy": 0.22430422753095627,
      "num_tokens": 76639173.0,
      "step": 34330
    },
    {
      "entropy": 5.634465837478638,
      "epoch": 3.8561801538720726,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035870378759489104,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2242377743124962,
      "num_tokens": 76648970.0,
      "step": 34335
    },
    {
      "entropy": 5.602747583389283,
      "epoch": 3.8567417307800302,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22511939257383345,
      "num_tokens": 76659101.0,
      "step": 34340
    },
    {
      "entropy": 5.652119731903076,
      "epoch": 3.857303307687988,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003586292480977998,
      "loss": 4.9535,
      "mean_token_accuracy": 0.22818142175674438,
      "num_tokens": 76670087.0,
      "step": 34345
    },
    {
      "entropy": 5.708567810058594,
      "epoch": 3.8578648845959456,
      "grad_norm": 1.125,
      "learning_rate": 0.00035859197435447927,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2293691173195839,
      "num_tokens": 76681116.0,
      "step": 34350
    },
    {
      "entropy": 5.618736219406128,
      "epoch": 3.8584264615039032,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035855469794955787,
      "loss": 4.948,
      "mean_token_accuracy": 0.2228906646370888,
      "num_tokens": 76692050.0,
      "step": 34355
    },
    {
      "entropy": 5.751301193237305,
      "epoch": 3.8589880384118604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035851741888422254,
      "loss": 5.0377,
      "mean_token_accuracy": 0.22140361815690995,
      "num_tokens": 76704338.0,
      "step": 34360
    },
    {
      "entropy": 5.608965110778809,
      "epoch": 3.859549615319818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003584801371596601,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23143182098865508,
      "num_tokens": 76715300.0,
      "step": 34365
    },
    {
      "entropy": 5.678367185592651,
      "epoch": 3.8601111922277758,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003584428527770578,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2245469331741333,
      "num_tokens": 76726654.0,
      "step": 34370
    },
    {
      "entropy": 5.537823247909546,
      "epoch": 3.860672769135733,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035840556573760266,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2342199131846428,
      "num_tokens": 76737087.0,
      "step": 34375
    },
    {
      "entropy": 5.584610319137573,
      "epoch": 3.8612343460436906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2284644663333893,
      "num_tokens": 76748006.0,
      "step": 34380
    },
    {
      "entropy": 5.615546226501465,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035833098369288303,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2426811784505844,
      "num_tokens": 76758063.0,
      "step": 34385
    },
    {
      "entropy": 5.613029718399048,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23107589930295944,
      "num_tokens": 76769080.0,
      "step": 34390
    },
    {
      "entropy": 5.6224353313446045,
      "epoch": 3.862919076767563,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035825639103500005,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22645397037267684,
      "num_tokens": 76780743.0,
      "step": 34395
    },
    {
      "entropy": 5.625202703475952,
      "epoch": 3.863480653675521,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035821909072909113,
      "loss": 4.9431,
      "mean_token_accuracy": 0.23056639432907106,
      "num_tokens": 76792277.0,
      "step": 34400
    },
    {
      "entropy": 5.621227550506592,
      "epoch": 3.8640422305834785,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2273109510540962,
      "num_tokens": 76802979.0,
      "step": 34405
    },
    {
      "entropy": 5.571267414093017,
      "epoch": 3.864603807491436,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.896,
      "mean_token_accuracy": 0.22913137078285217,
      "num_tokens": 76815016.0,
      "step": 34410
    },
    {
      "entropy": 5.560669803619385,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23779640346765518,
      "num_tokens": 76826360.0,
      "step": 34415
    },
    {
      "entropy": 5.649004888534546,
      "epoch": 3.865726961307351,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035806986302005205,
      "loss": 5.0096,
      "mean_token_accuracy": 0.22907016426324844,
      "num_tokens": 76838876.0,
      "step": 34420
    },
    {
      "entropy": 5.575813579559326,
      "epoch": 3.8662885382153087,
      "grad_norm": 1.125,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.893,
      "mean_token_accuracy": 0.2321113407611847,
      "num_tokens": 76849156.0,
      "step": 34425
    },
    {
      "entropy": 5.6442484855651855,
      "epoch": 3.866850115123266,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035799523329092093,
      "loss": 5.0577,
      "mean_token_accuracy": 0.22036772221326828,
      "num_tokens": 76859683.0,
      "step": 34430
    },
    {
      "entropy": 5.597307920455933,
      "epoch": 3.8674116920312236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035795791446186073,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23655631244182587,
      "num_tokens": 76870509.0,
      "step": 34435
    },
    {
      "entropy": 5.683844041824341,
      "epoch": 3.8679732689391813,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003579205929913882,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22762128561735154,
      "num_tokens": 76882181.0,
      "step": 34440
    },
    {
      "entropy": 5.66847414970398,
      "epoch": 3.868534845847139,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003578832688806919,
      "loss": 5.0359,
      "mean_token_accuracy": 0.22158266901969909,
      "num_tokens": 76894709.0,
      "step": 34445
    },
    {
      "entropy": 5.604011058807373,
      "epoch": 3.8690964227550966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003578459421309602,
      "loss": 4.8954,
      "mean_token_accuracy": 0.23298301696777343,
      "num_tokens": 76905974.0,
      "step": 34450
    },
    {
      "entropy": 5.731389093399048,
      "epoch": 3.869657999663054,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003578086127433816,
      "loss": 5.0592,
      "mean_token_accuracy": 0.22718816846609116,
      "num_tokens": 76918009.0,
      "step": 34455
    },
    {
      "entropy": 5.678001928329468,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035777128071914454,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2289826288819313,
      "num_tokens": 76928750.0,
      "step": 34460
    },
    {
      "entropy": 5.684007072448731,
      "epoch": 3.870781153478969,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003577339460594378,
      "loss": 5.0678,
      "mean_token_accuracy": 0.22906058877706528,
      "num_tokens": 76939989.0,
      "step": 34465
    },
    {
      "entropy": 5.521080732345581,
      "epoch": 3.8713427303869263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003576966087654502,
      "loss": 4.8742,
      "mean_token_accuracy": 0.2339620903134346,
      "num_tokens": 76950026.0,
      "step": 34470
    },
    {
      "entropy": 5.6976922988891605,
      "epoch": 3.871904307294884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035765926883837054,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21997570544481276,
      "num_tokens": 76961198.0,
      "step": 34475
    },
    {
      "entropy": 5.635962867736817,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003576219262793877,
      "loss": 4.938,
      "mean_token_accuracy": 0.23339524120092392,
      "num_tokens": 76973012.0,
      "step": 34480
    },
    {
      "entropy": 5.652032041549683,
      "epoch": 3.873027461110799,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.945,
      "mean_token_accuracy": 0.22968221753835677,
      "num_tokens": 76983613.0,
      "step": 34485
    },
    {
      "entropy": 5.705576467514038,
      "epoch": 3.8735890380187565,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003575472332704684,
      "loss": 5.0893,
      "mean_token_accuracy": 0.2157877579331398,
      "num_tokens": 76995471.0,
      "step": 34490
    },
    {
      "entropy": 5.5868675231933596,
      "epoch": 3.874150614926714,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035750988282291037,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23320302218198777,
      "num_tokens": 77005777.0,
      "step": 34495
    },
    {
      "entropy": 5.665107488632202,
      "epoch": 3.874712191834672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035747252974820565,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21998475044965743,
      "num_tokens": 77017099.0,
      "step": 34500
    },
    {
      "entropy": 5.627462148666382,
      "epoch": 3.8752737687426295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035743517404754354,
      "loss": 5.0204,
      "mean_token_accuracy": 0.22293082475662232,
      "num_tokens": 77028350.0,
      "step": 34505
    },
    {
      "entropy": 5.615108489990234,
      "epoch": 3.8758353456505867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003573978157221136,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23548561334609985,
      "num_tokens": 77039355.0,
      "step": 34510
    },
    {
      "entropy": 5.727294206619263,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035736045477310517,
      "loss": 5.0597,
      "mean_token_accuracy": 0.2199416622519493,
      "num_tokens": 77049966.0,
      "step": 34515
    },
    {
      "entropy": 5.691061210632324,
      "epoch": 3.876958499466502,
      "grad_norm": 1.125,
      "learning_rate": 0.00035732309120170786,
      "loss": 4.932,
      "mean_token_accuracy": 0.226275497674942,
      "num_tokens": 77061102.0,
      "step": 34520
    },
    {
      "entropy": 5.614276695251465,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035728572500911155,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2240409567952156,
      "num_tokens": 77071516.0,
      "step": 34525
    },
    {
      "entropy": 5.619024801254272,
      "epoch": 3.878081653282417,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035724835619650573,
      "loss": 5.0369,
      "mean_token_accuracy": 0.22174447029829025,
      "num_tokens": 77081924.0,
      "step": 34530
    },
    {
      "entropy": 5.660076379776001,
      "epoch": 3.8786432301903746,
      "grad_norm": 1.125,
      "learning_rate": 0.0003572109847650804,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22510516196489333,
      "num_tokens": 77093064.0,
      "step": 34535
    },
    {
      "entropy": 5.668350696563721,
      "epoch": 3.8792048070983323,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035717361071602546,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22587309181690216,
      "num_tokens": 77104428.0,
      "step": 34540
    },
    {
      "entropy": 5.639728975296021,
      "epoch": 3.87976638400629,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2376953035593033,
      "num_tokens": 77114949.0,
      "step": 34545
    },
    {
      "entropy": 5.704399633407593,
      "epoch": 3.880327960914247,
      "grad_norm": 1.125,
      "learning_rate": 0.00035709885476978666,
      "loss": 5.0427,
      "mean_token_accuracy": 0.22749326080083848,
      "num_tokens": 77125178.0,
      "step": 34550
    },
    {
      "entropy": 5.565169906616211,
      "epoch": 3.880889537822205,
      "grad_norm": 1.125,
      "learning_rate": 0.00035706147287498303,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23527367711067199,
      "num_tokens": 77135967.0,
      "step": 34555
    },
    {
      "entropy": 5.670198774337768,
      "epoch": 3.8814511147301625,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035702408836731043,
      "loss": 5.0292,
      "mean_token_accuracy": 0.23366452157497405,
      "num_tokens": 77148052.0,
      "step": 34560
    },
    {
      "entropy": 5.683129549026489,
      "epoch": 3.8820126916381197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000356986701247959,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22875525057315826,
      "num_tokens": 77159008.0,
      "step": 34565
    },
    {
      "entropy": 5.624520015716553,
      "epoch": 3.8825742685460773,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22401080131530762,
      "num_tokens": 77169846.0,
      "step": 34570
    },
    {
      "entropy": 5.609382295608521,
      "epoch": 3.883135845454035,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003569119191789815,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22398761063814163,
      "num_tokens": 77182881.0,
      "step": 34575
    },
    {
      "entropy": 5.641481542587281,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035687452423173655,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22423758506774902,
      "num_tokens": 77193852.0,
      "step": 34580
    },
    {
      "entropy": 5.688366031646728,
      "epoch": 3.88425899926995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000356837126677575,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22452111840248107,
      "num_tokens": 77205141.0,
      "step": 34585
    },
    {
      "entropy": 5.722354459762573,
      "epoch": 3.8848205761779075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003567997265176876,
      "loss": 5.0282,
      "mean_token_accuracy": 0.2164707750082016,
      "num_tokens": 77217521.0,
      "step": 34590
    },
    {
      "entropy": 5.739170503616333,
      "epoch": 3.885382153085865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035676232375326513,
      "loss": 5.038,
      "mean_token_accuracy": 0.22232122719287872,
      "num_tokens": 77227468.0,
      "step": 34595
    },
    {
      "entropy": 5.578084421157837,
      "epoch": 3.885943729993823,
      "grad_norm": 1.125,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22471892535686494,
      "num_tokens": 77237908.0,
      "step": 34600
    },
    {
      "entropy": 5.558370542526245,
      "epoch": 3.88650530690178,
      "grad_norm": 1.171875,
      "learning_rate": 0.000356687510415579,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2330235794186592,
      "num_tokens": 77248621.0,
      "step": 34605
    },
    {
      "entropy": 5.672929334640503,
      "epoch": 3.8870668838097377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003566500998446973,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21950124502182006,
      "num_tokens": 77260261.0,
      "step": 34610
    },
    {
      "entropy": 5.650392150878906,
      "epoch": 3.8876284607176954,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035661268667404485,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22530285120010377,
      "num_tokens": 77270985.0,
      "step": 34615
    },
    {
      "entropy": 5.659456062316894,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003565752709048127,
      "loss": 4.9889,
      "mean_token_accuracy": 0.23151240795850753,
      "num_tokens": 77281876.0,
      "step": 34620
    },
    {
      "entropy": 5.577308464050293,
      "epoch": 3.8887516145336103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003565378525381924,
      "loss": 4.875,
      "mean_token_accuracy": 0.22832536101341247,
      "num_tokens": 77293141.0,
      "step": 34625
    },
    {
      "entropy": 5.638823366165161,
      "epoch": 3.889313191441568,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22956283390522003,
      "num_tokens": 77305723.0,
      "step": 34630
    },
    {
      "entropy": 5.562638521194458,
      "epoch": 3.8898747683495256,
      "grad_norm": 1.125,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2353903904557228,
      "num_tokens": 77317736.0,
      "step": 34635
    },
    {
      "entropy": 5.602745246887207,
      "epoch": 3.8904363452574833,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.9594,
      "mean_token_accuracy": 0.22286449670791625,
      "num_tokens": 77328293.0,
      "step": 34640
    },
    {
      "entropy": 5.658864831924438,
      "epoch": 3.8909979221654405,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003563881531216579,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21817657649517058,
      "num_tokens": 77339371.0,
      "step": 34645
    },
    {
      "entropy": 5.68715648651123,
      "epoch": 3.891559499073398,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003563507217859691,
      "loss": 5.0211,
      "mean_token_accuracy": 0.21797109991312028,
      "num_tokens": 77350234.0,
      "step": 34650
    },
    {
      "entropy": 5.695639610290527,
      "epoch": 3.892121075981356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003563132878600419,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22716296911239625,
      "num_tokens": 77361144.0,
      "step": 34655
    },
    {
      "entropy": 5.65096640586853,
      "epoch": 3.892682652889313,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003562758513450678,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22468167692422866,
      "num_tokens": 77372090.0,
      "step": 34660
    },
    {
      "entropy": 5.705678129196167,
      "epoch": 3.8932442297972707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003562384122422393,
      "loss": 5.064,
      "mean_token_accuracy": 0.22158280164003372,
      "num_tokens": 77382974.0,
      "step": 34665
    },
    {
      "entropy": 5.686750316619873,
      "epoch": 3.8938058067052284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003562009705527481,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22451305836439134,
      "num_tokens": 77395971.0,
      "step": 34670
    },
    {
      "entropy": 5.673870515823364,
      "epoch": 3.8943673836131856,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035616352627778665,
      "loss": 4.9763,
      "mean_token_accuracy": 0.2207277625799179,
      "num_tokens": 77405776.0,
      "step": 34675
    },
    {
      "entropy": 5.67974886894226,
      "epoch": 3.8949289605211432,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035612607941854695,
      "loss": 4.9954,
      "mean_token_accuracy": 0.22430122196674346,
      "num_tokens": 77417668.0,
      "step": 34680
    },
    {
      "entropy": 5.66153130531311,
      "epoch": 3.895490537429101,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2253752037882805,
      "num_tokens": 77428581.0,
      "step": 34685
    },
    {
      "entropy": 5.695534515380859,
      "epoch": 3.8960521143370586,
      "grad_norm": 1.125,
      "learning_rate": 0.0003560511779520025,
      "loss": 5.0431,
      "mean_token_accuracy": 0.22623122334480286,
      "num_tokens": 77441587.0,
      "step": 34690
    },
    {
      "entropy": 5.672238159179687,
      "epoch": 3.896613691245016,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003560137233470826,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22756811529397963,
      "num_tokens": 77452579.0,
      "step": 34695
    },
    {
      "entropy": 5.613932991027832,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22636728733778,
      "num_tokens": 77464465.0,
      "step": 34700
    },
    {
      "entropy": 5.600366830825806,
      "epoch": 3.897736845060931,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22009779065847396,
      "num_tokens": 77474944.0,
      "step": 34705
    },
    {
      "entropy": 5.571808099746704,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.8399,
      "mean_token_accuracy": 0.23388955146074294,
      "num_tokens": 77485389.0,
      "step": 34710
    },
    {
      "entropy": 5.650898551940918,
      "epoch": 3.898859998876846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2274793118238449,
      "num_tokens": 77496834.0,
      "step": 34715
    },
    {
      "entropy": 5.568704843521118,
      "epoch": 3.8994215757848036,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2284489631652832,
      "num_tokens": 77508123.0,
      "step": 34720
    },
    {
      "entropy": 5.61333532333374,
      "epoch": 3.8999831526927613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003557889415896332,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22489914298057556,
      "num_tokens": 77519599.0,
      "step": 34725
    },
    {
      "entropy": 5.597734022140503,
      "epoch": 3.900544729600719,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035575146895320327,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2294714108109474,
      "num_tokens": 77532241.0,
      "step": 34730
    },
    {
      "entropy": 5.7228857517242435,
      "epoch": 3.9011063065086766,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003557139937456152,
      "loss": 5.0189,
      "mean_token_accuracy": 0.22619763761758804,
      "num_tokens": 77542702.0,
      "step": 34735
    },
    {
      "entropy": 5.632202625274658,
      "epoch": 3.901667883416634,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035567651596806236,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22393217980861663,
      "num_tokens": 77554235.0,
      "step": 34740
    },
    {
      "entropy": 5.618323421478271,
      "epoch": 3.9022294603245915,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035563903562173786,
      "loss": 4.9636,
      "mean_token_accuracy": 0.23177284598350525,
      "num_tokens": 77565704.0,
      "step": 34745
    },
    {
      "entropy": 5.58668475151062,
      "epoch": 3.902791037232549,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22522929161787034,
      "num_tokens": 77576524.0,
      "step": 34750
    },
    {
      "entropy": 5.638580083847046,
      "epoch": 3.9033526141405064,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.9288,
      "mean_token_accuracy": 0.23365162760019303,
      "num_tokens": 77587292.0,
      "step": 34755
    },
    {
      "entropy": 5.6489002227783205,
      "epoch": 3.903914191048464,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003555265791820695,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21825221329927444,
      "num_tokens": 77597958.0,
      "step": 34760
    },
    {
      "entropy": 5.704643440246582,
      "epoch": 3.9044757679564217,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003554890885725935,
      "loss": 5.0481,
      "mean_token_accuracy": 0.22391968220472336,
      "num_tokens": 77609264.0,
      "step": 34765
    },
    {
      "entropy": 5.7505964756011965,
      "epoch": 3.905037344864379,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003554515954003137,
      "loss": 5.0461,
      "mean_token_accuracy": 0.22607079297304153,
      "num_tokens": 77621224.0,
      "step": 34770
    },
    {
      "entropy": 5.661201858520508,
      "epoch": 3.9055989217723366,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003554140996664239,
      "loss": 5.0214,
      "mean_token_accuracy": 0.2219918817281723,
      "num_tokens": 77632009.0,
      "step": 34775
    },
    {
      "entropy": 5.626842832565307,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.171875,
      "learning_rate": 0.000355376601372118,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22537139505147935,
      "num_tokens": 77643094.0,
      "step": 34780
    },
    {
      "entropy": 5.6496326446533205,
      "epoch": 3.906722075588252,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003553391005185899,
      "loss": 4.9886,
      "mean_token_accuracy": 0.23433544784784316,
      "num_tokens": 77654049.0,
      "step": 34785
    },
    {
      "entropy": 5.619156551361084,
      "epoch": 3.9072836524962096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.919,
      "mean_token_accuracy": 0.23082916140556337,
      "num_tokens": 77664820.0,
      "step": 34790
    },
    {
      "entropy": 5.744994735717773,
      "epoch": 3.907845229404167,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003552640911386436,
      "loss": 5.0755,
      "mean_token_accuracy": 0.22085265517234803,
      "num_tokens": 77675107.0,
      "step": 34795
    },
    {
      "entropy": 5.64300127029419,
      "epoch": 3.9084068063121244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035522658261461366,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2229372262954712,
      "num_tokens": 77685487.0,
      "step": 34800
    },
    {
      "entropy": 5.639738321304321,
      "epoch": 3.908968383220082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2247209295630455,
      "num_tokens": 77697391.0,
      "step": 34805
    },
    {
      "entropy": 5.6749461650848385,
      "epoch": 3.9095299601280393,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2265079364180565,
      "num_tokens": 77707270.0,
      "step": 34810
    },
    {
      "entropy": 5.531836271286011,
      "epoch": 3.910091537035997,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23479119092226028,
      "num_tokens": 77717164.0,
      "step": 34815
    },
    {
      "entropy": 5.7000970363616945,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003550765229859828,
      "loss": 5.0359,
      "mean_token_accuracy": 0.225142639875412,
      "num_tokens": 77728142.0,
      "step": 34820
    },
    {
      "entropy": 5.685300827026367,
      "epoch": 3.9112146908519123,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003550390017016698,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2335679978132248,
      "num_tokens": 77740162.0,
      "step": 34825
    },
    {
      "entropy": 5.647401142120361,
      "epoch": 3.91177626775987,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035500147786888387,
      "loss": 4.917,
      "mean_token_accuracy": 0.23925721943378447,
      "num_tokens": 77751574.0,
      "step": 34830
    },
    {
      "entropy": 5.690142965316772,
      "epoch": 3.912337844667827,
      "grad_norm": 1.25,
      "learning_rate": 0.00035496395148881994,
      "loss": 5.0237,
      "mean_token_accuracy": 0.22619060873985292,
      "num_tokens": 77762919.0,
      "step": 34835
    },
    {
      "entropy": 5.6410496711730955,
      "epoch": 3.912899421575785,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.961,
      "mean_token_accuracy": 0.22562432587146758,
      "num_tokens": 77774578.0,
      "step": 34840
    },
    {
      "entropy": 5.690762901306153,
      "epoch": 3.9134609984837425,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035488889109163725,
      "loss": 5.0029,
      "mean_token_accuracy": 0.22155731916427612,
      "num_tokens": 77784872.0,
      "step": 34845
    },
    {
      "entropy": 5.605316257476806,
      "epoch": 3.9140225753916997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22862707823514938,
      "num_tokens": 77796159.0,
      "step": 34850
    },
    {
      "entropy": 5.568557214736939,
      "epoch": 3.9145841522996574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23184582740068435,
      "num_tokens": 77805779.0,
      "step": 34855
    },
    {
      "entropy": 5.678985548019409,
      "epoch": 3.915145729207615,
      "grad_norm": 1.125,
      "learning_rate": 0.0003547762814211517,
      "loss": 5.0066,
      "mean_token_accuracy": 0.22828080505132675,
      "num_tokens": 77818102.0,
      "step": 34860
    },
    {
      "entropy": 5.67267107963562,
      "epoch": 3.9157073061155723,
      "grad_norm": 1.078125,
      "learning_rate": 0.000354738739782514,
      "loss": 5.0203,
      "mean_token_accuracy": 0.2173686519265175,
      "num_tokens": 77830107.0,
      "step": 34865
    },
    {
      "entropy": 5.6725654125213625,
      "epoch": 3.91626888302353,
      "grad_norm": 1.28125,
      "learning_rate": 0.000354701195604964,
      "loss": 5.0234,
      "mean_token_accuracy": 0.22307849079370498,
      "num_tokens": 77841388.0,
      "step": 34870
    },
    {
      "entropy": 5.643773365020752,
      "epoch": 3.9168304599314876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003546636488896969,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2274055913090706,
      "num_tokens": 77853418.0,
      "step": 34875
    },
    {
      "entropy": 5.706916809082031,
      "epoch": 3.9173920368394453,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003546260996379083,
      "loss": 4.9854,
      "mean_token_accuracy": 0.22463717460632324,
      "num_tokens": 77865019.0,
      "step": 34880
    },
    {
      "entropy": 5.676601505279541,
      "epoch": 3.917953613747403,
      "grad_norm": 1.25,
      "learning_rate": 0.000354588547850794,
      "loss": 4.93,
      "mean_token_accuracy": 0.2274512082338333,
      "num_tokens": 77876546.0,
      "step": 34885
    },
    {
      "entropy": 5.71756386756897,
      "epoch": 3.91851519065536,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003545509935295493,
      "loss": 5.0233,
      "mean_token_accuracy": 0.23046770989894866,
      "num_tokens": 77888081.0,
      "step": 34890
    },
    {
      "entropy": 5.540466833114624,
      "epoch": 3.919076767563318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003545134366753701,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23357495218515395,
      "num_tokens": 77900106.0,
      "step": 34895
    },
    {
      "entropy": 5.6112854957580565,
      "epoch": 3.9196383444712755,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2309060961008072,
      "num_tokens": 77910452.0,
      "step": 34900
    },
    {
      "entropy": 5.6970374584198,
      "epoch": 3.9201999213792327,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035443831537299177,
      "loss": 5.0208,
      "mean_token_accuracy": 0.22531363368034363,
      "num_tokens": 77921826.0,
      "step": 34905
    },
    {
      "entropy": 5.77866940498352,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003544007509271844,
      "loss": 5.0613,
      "mean_token_accuracy": 0.22439448833465575,
      "num_tokens": 77933909.0,
      "step": 34910
    },
    {
      "entropy": 5.766520690917969,
      "epoch": 3.921323075195148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035436318395322636,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2212739408016205,
      "num_tokens": 77945232.0,
      "step": 34915
    },
    {
      "entropy": 5.643659830093384,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003543256144523137,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22486903816461562,
      "num_tokens": 77957305.0,
      "step": 34920
    },
    {
      "entropy": 5.589197397232056,
      "epoch": 3.9224462290110633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23661915510892867,
      "num_tokens": 77968036.0,
      "step": 34925
    },
    {
      "entropy": 5.6196229457855225,
      "epoch": 3.9230078059190205,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003542504678744099,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22866115868091583,
      "num_tokens": 77978826.0,
      "step": 34930
    },
    {
      "entropy": 5.623037576675415,
      "epoch": 3.923569382826978,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2254600927233696,
      "num_tokens": 77989820.0,
      "step": 34935
    },
    {
      "entropy": 5.64576940536499,
      "epoch": 3.924130959734936,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22426143437623977,
      "num_tokens": 78000114.0,
      "step": 34940
    },
    {
      "entropy": 5.6020091533660885,
      "epoch": 3.924692536642893,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.9241,
      "mean_token_accuracy": 0.23342568129301072,
      "num_tokens": 78010680.0,
      "step": 34945
    },
    {
      "entropy": 5.628191995620727,
      "epoch": 3.9252541135508507,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22826392501592635,
      "num_tokens": 78022172.0,
      "step": 34950
    },
    {
      "entropy": 5.629863357543945,
      "epoch": 3.9258156904588084,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2184725359082222,
      "num_tokens": 78033535.0,
      "step": 34955
    },
    {
      "entropy": 5.657886695861817,
      "epoch": 3.9263772673667656,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003540249676182131,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22113814651966096,
      "num_tokens": 78044555.0,
      "step": 34960
    },
    {
      "entropy": 5.648629188537598,
      "epoch": 3.9269388442747233,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035398737542854936,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23492586463689805,
      "num_tokens": 78056078.0,
      "step": 34965
    },
    {
      "entropy": 5.650190639495849,
      "epoch": 3.927500421182681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22581399083137513,
      "num_tokens": 78068324.0,
      "step": 34970
    },
    {
      "entropy": 5.608314657211304,
      "epoch": 3.9280619980906386,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035391218350545287,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22395882457494737,
      "num_tokens": 78080013.0,
      "step": 34975
    },
    {
      "entropy": 5.6381196022033695,
      "epoch": 3.9286235749985963,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035387458377441434,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2276500552892685,
      "num_tokens": 78091494.0,
      "step": 34980
    },
    {
      "entropy": 5.659497499465942,
      "epoch": 3.9291851519065535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22452893406152724,
      "num_tokens": 78102037.0,
      "step": 34985
    },
    {
      "entropy": 5.637392044067383,
      "epoch": 3.929746728814511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035379937677934245,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2224537655711174,
      "num_tokens": 78112651.0,
      "step": 34990
    },
    {
      "entropy": 5.673631715774536,
      "epoch": 3.930308305722469,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003537617695177037,
      "loss": 5.017,
      "mean_token_accuracy": 0.22335184514522552,
      "num_tokens": 78123273.0,
      "step": 34995
    },
    {
      "entropy": 5.659987735748291,
      "epoch": 3.930869882630426,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035372415974825974,
      "loss": 5.0189,
      "mean_token_accuracy": 0.22154339253902436,
      "num_tokens": 78133868.0,
      "step": 35000
    },
    {
      "entropy": 5.761922550201416,
      "epoch": 3.9314314595383837,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035368654747220795,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21975772380828856,
      "num_tokens": 78145001.0,
      "step": 35005
    },
    {
      "entropy": 5.740580654144287,
      "epoch": 3.9319930364463413,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035364893269074606,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21776216924190522,
      "num_tokens": 78155699.0,
      "step": 35010
    },
    {
      "entropy": 5.613238716125489,
      "epoch": 3.932554613354299,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035361131540507157,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23024097532033921,
      "num_tokens": 78165941.0,
      "step": 35015
    },
    {
      "entropy": 5.612911033630371,
      "epoch": 3.9331161902622567,
      "grad_norm": 1.125,
      "learning_rate": 0.0003535736956163823,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2309500679373741,
      "num_tokens": 78177049.0,
      "step": 35020
    },
    {
      "entropy": 5.664494132995605,
      "epoch": 3.933677767170214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035353607332587613,
      "loss": 5.0618,
      "mean_token_accuracy": 0.22294115275144577,
      "num_tokens": 78188834.0,
      "step": 35025
    },
    {
      "entropy": 5.727819108963013,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003534984485347509,
      "loss": 4.9967,
      "mean_token_accuracy": 0.22912617325782775,
      "num_tokens": 78199847.0,
      "step": 35030
    },
    {
      "entropy": 5.551779842376709,
      "epoch": 3.934800920986129,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003534608212442046,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22995148301124574,
      "num_tokens": 78210747.0,
      "step": 35035
    },
    {
      "entropy": 5.6443463325500485,
      "epoch": 3.9353624978940864,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035342319145543533,
      "loss": 5.003,
      "mean_token_accuracy": 0.2292477160692215,
      "num_tokens": 78221711.0,
      "step": 35040
    },
    {
      "entropy": 5.766300249099731,
      "epoch": 3.935924074802044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035338555916964126,
      "loss": 5.1169,
      "mean_token_accuracy": 0.21329459100961684,
      "num_tokens": 78233957.0,
      "step": 35045
    },
    {
      "entropy": 5.619800853729248,
      "epoch": 3.9364856517100018,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035334792438802056,
      "loss": 4.9986,
      "mean_token_accuracy": 0.22497449815273285,
      "num_tokens": 78245403.0,
      "step": 35050
    },
    {
      "entropy": 5.692008113861084,
      "epoch": 3.937047228617959,
      "grad_norm": 1.125,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.948,
      "mean_token_accuracy": 0.22239017337560654,
      "num_tokens": 78256062.0,
      "step": 35055
    },
    {
      "entropy": 5.666768360137939,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003532726473420927,
      "loss": 4.9886,
      "mean_token_accuracy": 0.225938980281353,
      "num_tokens": 78265895.0,
      "step": 35060
    },
    {
      "entropy": 5.667288112640381,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22139344364404678,
      "num_tokens": 78276752.0,
      "step": 35065
    },
    {
      "entropy": 5.669629812240601,
      "epoch": 3.938731959341832,
      "grad_norm": 1.25,
      "learning_rate": 0.00035319736032723905,
      "loss": 4.9817,
      "mean_token_accuracy": 0.2339001104235649,
      "num_tokens": 78288357.0,
      "step": 35070
    },
    {
      "entropy": 5.677090978622436,
      "epoch": 3.9392935362497896,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003531597130844616,
      "loss": 4.927,
      "mean_token_accuracy": 0.23194187134504318,
      "num_tokens": 78299558.0,
      "step": 35075
    },
    {
      "entropy": 5.762809944152832,
      "epoch": 3.939855113157747,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035312206335304843,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22564319223165513,
      "num_tokens": 78311443.0,
      "step": 35080
    },
    {
      "entropy": 5.724372386932373,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.25,
      "learning_rate": 0.00035308441113419855,
      "loss": 5.0585,
      "mean_token_accuracy": 0.22287312000989914,
      "num_tokens": 78322575.0,
      "step": 35085
    },
    {
      "entropy": 5.661104488372803,
      "epoch": 3.940978266973662,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035304675642911074,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22475543916225432,
      "num_tokens": 78333583.0,
      "step": 35090
    },
    {
      "entropy": 5.640227794647217,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003530090992389838,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22855928391218186,
      "num_tokens": 78343339.0,
      "step": 35095
    },
    {
      "entropy": 5.707261085510254,
      "epoch": 3.942101420789577,
      "grad_norm": 1.1875,
      "learning_rate": 0.000352971439565017,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22655223459005355,
      "num_tokens": 78354613.0,
      "step": 35100
    },
    {
      "entropy": 5.62253999710083,
      "epoch": 3.9426629976975347,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22602506428956987,
      "num_tokens": 78365105.0,
      "step": 35105
    },
    {
      "entropy": 5.5766541957855225,
      "epoch": 3.9432245746054924,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035289611277036,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2401183396577835,
      "num_tokens": 78375311.0,
      "step": 35110
    },
    {
      "entropy": 5.600499963760376,
      "epoch": 3.94378615151345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.8942,
      "mean_token_accuracy": 0.23248492926359177,
      "num_tokens": 78386653.0,
      "step": 35115
    },
    {
      "entropy": 5.531075286865234,
      "epoch": 3.9443477284214072,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035282077605473317,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22547387182712555,
      "num_tokens": 78397296.0,
      "step": 35120
    },
    {
      "entropy": 5.651582336425781,
      "epoch": 3.944909305329365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003527831039795545,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23283239156007768,
      "num_tokens": 78408825.0,
      "step": 35125
    },
    {
      "entropy": 5.608019018173218,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23291198462247847,
      "num_tokens": 78419895.0,
      "step": 35130
    },
    {
      "entropy": 5.608262157440185,
      "epoch": 3.94603245914528,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2335692748427391,
      "num_tokens": 78430111.0,
      "step": 35135
    },
    {
      "entropy": 5.615688753128052,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035267007289895176,
      "loss": 4.9992,
      "mean_token_accuracy": 0.222836235165596,
      "num_tokens": 78441778.0,
      "step": 35140
    },
    {
      "entropy": 5.647122859954834,
      "epoch": 3.947155612961195,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22926730960607528,
      "num_tokens": 78451930.0,
      "step": 35145
    },
    {
      "entropy": 5.596216249465942,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22956764549016953,
      "num_tokens": 78463002.0,
      "step": 35150
    },
    {
      "entropy": 5.565324068069458,
      "epoch": 3.94827876677711,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.953,
      "mean_token_accuracy": 0.22337931841611863,
      "num_tokens": 78473573.0,
      "step": 35155
    },
    {
      "entropy": 5.650953722000122,
      "epoch": 3.9488403436850676,
      "grad_norm": 1.125,
      "learning_rate": 0.00035251933017444814,
      "loss": 4.974,
      "mean_token_accuracy": 0.22542529106140136,
      "num_tokens": 78485336.0,
      "step": 35160
    },
    {
      "entropy": 5.679855060577393,
      "epoch": 3.9494019205930253,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2290257468819618,
      "num_tokens": 78497801.0,
      "step": 35165
    },
    {
      "entropy": 5.614393806457519,
      "epoch": 3.949963497500983,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.9607,
      "mean_token_accuracy": 0.22752908617258072,
      "num_tokens": 78509288.0,
      "step": 35170
    },
    {
      "entropy": 5.631801509857178,
      "epoch": 3.95052507440894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003524062472102927,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2324563220143318,
      "num_tokens": 78519871.0,
      "step": 35175
    },
    {
      "entropy": 5.607153415679932,
      "epoch": 3.951086651316898,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003523685479580174,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2231569617986679,
      "num_tokens": 78530287.0,
      "step": 35180
    },
    {
      "entropy": 5.650895500183106,
      "epoch": 3.9516482282248555,
      "grad_norm": 1.125,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.953,
      "mean_token_accuracy": 0.226356802880764,
      "num_tokens": 78542341.0,
      "step": 35185
    },
    {
      "entropy": 5.595199012756348,
      "epoch": 3.9522098051328127,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.8422,
      "mean_token_accuracy": 0.23371480107307435,
      "num_tokens": 78552533.0,
      "step": 35190
    },
    {
      "entropy": 5.666037464141846,
      "epoch": 3.9527713820407704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22826727628707885,
      "num_tokens": 78564007.0,
      "step": 35195
    },
    {
      "entropy": 5.617849683761596,
      "epoch": 3.953332958948728,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035221772632647503,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2224250003695488,
      "num_tokens": 78574015.0,
      "step": 35200
    },
    {
      "entropy": 5.669429922103882,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035218001476898194,
      "loss": 5.0296,
      "mean_token_accuracy": 0.22737761288881303,
      "num_tokens": 78586230.0,
      "step": 35205
    },
    {
      "entropy": 5.646058702468872,
      "epoch": 3.9544561127646434,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.9783,
      "mean_token_accuracy": 0.22968566715717315,
      "num_tokens": 78596819.0,
      "step": 35210
    },
    {
      "entropy": 5.625739622116089,
      "epoch": 3.9550176896726006,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003521045842828718,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22982620000839232,
      "num_tokens": 78608788.0,
      "step": 35215
    },
    {
      "entropy": 5.646988201141357,
      "epoch": 3.9555792665805583,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22439119070768357,
      "num_tokens": 78619124.0,
      "step": 35220
    },
    {
      "entropy": 5.753645133972168,
      "epoch": 3.956140843488516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003520291439766024,
      "loss": 5.047,
      "mean_token_accuracy": 0.22623891681432723,
      "num_tokens": 78630799.0,
      "step": 35225
    },
    {
      "entropy": 5.589961290359497,
      "epoch": 3.956702420396473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23381823003292085,
      "num_tokens": 78641664.0,
      "step": 35230
    },
    {
      "entropy": 5.686951446533203,
      "epoch": 3.957263997304431,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035195369385978165,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2324879825115204,
      "num_tokens": 78652025.0,
      "step": 35235
    },
    {
      "entropy": 5.65976243019104,
      "epoch": 3.9578255742123885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035191596512541747,
      "loss": 4.9414,
      "mean_token_accuracy": 0.23900072127580643,
      "num_tokens": 78663539.0,
      "step": 35240
    },
    {
      "entropy": 5.6216450214385985,
      "epoch": 3.9583871511203457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003518782339420192,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21972978860139847,
      "num_tokens": 78674398.0,
      "step": 35245
    },
    {
      "entropy": 5.641387176513672,
      "epoch": 3.9589487280283033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035184050031078815,
      "loss": 5.009,
      "mean_token_accuracy": 0.21750928461551666,
      "num_tokens": 78685945.0,
      "step": 35250
    },
    {
      "entropy": 5.616379261016846,
      "epoch": 3.959510304936261,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2183590814471245,
      "num_tokens": 78696664.0,
      "step": 35255
    },
    {
      "entropy": 5.635143232345581,
      "epoch": 3.9600718818442187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.8459,
      "mean_token_accuracy": 0.23802008628845214,
      "num_tokens": 78706812.0,
      "step": 35260
    },
    {
      "entropy": 5.5832366943359375,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2336641177535057,
      "num_tokens": 78716657.0,
      "step": 35265
    },
    {
      "entropy": 5.642794275283814,
      "epoch": 3.9611950356601335,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.9884,
      "mean_token_accuracy": 0.22143436819314957,
      "num_tokens": 78728313.0,
      "step": 35270
    },
    {
      "entropy": 5.646801900863648,
      "epoch": 3.961756612568091,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003516517954791944,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2336377039551735,
      "num_tokens": 78739888.0,
      "step": 35275
    },
    {
      "entropy": 5.644894599914551,
      "epoch": 3.962318189476049,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2276619553565979,
      "num_tokens": 78750624.0,
      "step": 35280
    },
    {
      "entropy": 5.646116113662719,
      "epoch": 3.962879766384006,
      "grad_norm": 1.125,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2294930949807167,
      "num_tokens": 78762506.0,
      "step": 35285
    },
    {
      "entropy": 5.674573230743408,
      "epoch": 3.9634413432919637,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003515385432831499,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22446807473897934,
      "num_tokens": 78772565.0,
      "step": 35290
    },
    {
      "entropy": 5.614506959915161,
      "epoch": 3.9640029201999214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.8741,
      "mean_token_accuracy": 0.22958765923976898,
      "num_tokens": 78784107.0,
      "step": 35295
    },
    {
      "entropy": 5.665911245346069,
      "epoch": 3.964564497107879,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003514630296320323,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22444697767496108,
      "num_tokens": 78796182.0,
      "step": 35300
    },
    {
      "entropy": 5.690092515945435,
      "epoch": 3.9651260740158367,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003514252691539536,
      "loss": 4.9941,
      "mean_token_accuracy": 0.22996919006109237,
      "num_tokens": 78806620.0,
      "step": 35305
    },
    {
      "entropy": 5.647673177719116,
      "epoch": 3.965687650923794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.8869,
      "mean_token_accuracy": 0.24151475131511688,
      "num_tokens": 78817034.0,
      "step": 35310
    },
    {
      "entropy": 5.68063645362854,
      "epoch": 3.9662492278317516,
      "grad_norm": 1.484375,
      "learning_rate": 0.000351349740898768,
      "loss": 4.9415,
      "mean_token_accuracy": 0.23134747445583342,
      "num_tokens": 78828974.0,
      "step": 35315
    },
    {
      "entropy": 5.663855648040771,
      "epoch": 3.9668108047397093,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003513119731240659,
      "loss": 4.9757,
      "mean_token_accuracy": 0.2245747983455658,
      "num_tokens": 78840514.0,
      "step": 35320
    },
    {
      "entropy": 5.629500961303711,
      "epoch": 3.9673723816476665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035127420291956104,
      "loss": 4.986,
      "mean_token_accuracy": 0.23033688962459564,
      "num_tokens": 78850839.0,
      "step": 35325
    },
    {
      "entropy": 5.658786725997925,
      "epoch": 3.967933958555624,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003512364302864559,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22983096688985824,
      "num_tokens": 78862931.0,
      "step": 35330
    },
    {
      "entropy": 5.667521286010742,
      "epoch": 3.968495535463582,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22601879835128785,
      "num_tokens": 78872324.0,
      "step": 35335
    },
    {
      "entropy": 5.579931545257568,
      "epoch": 3.969057112371539,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22562773674726486,
      "num_tokens": 78884075.0,
      "step": 35340
    },
    {
      "entropy": 5.526426887512207,
      "epoch": 3.9696186892794967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.863,
      "mean_token_accuracy": 0.23188286125659943,
      "num_tokens": 78894804.0,
      "step": 35345
    },
    {
      "entropy": 5.681999778747558,
      "epoch": 3.9701802661874543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035108531549208846,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2202178046107292,
      "num_tokens": 78906069.0,
      "step": 35350
    },
    {
      "entropy": 5.808830261230469,
      "epoch": 3.970741843095412,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035104753073402433,
      "loss": 5.0919,
      "mean_token_accuracy": 0.2198920726776123,
      "num_tokens": 78918490.0,
      "step": 35355
    },
    {
      "entropy": 5.704706001281738,
      "epoch": 3.9713034200033697,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003510097435545774,
      "loss": 4.987,
      "mean_token_accuracy": 0.2238558977842331,
      "num_tokens": 78929790.0,
      "step": 35360
    },
    {
      "entropy": 5.640917158126831,
      "epoch": 3.971864996911327,
      "grad_norm": 1.125,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.9476,
      "mean_token_accuracy": 0.23115588426589967,
      "num_tokens": 78941600.0,
      "step": 35365
    },
    {
      "entropy": 5.710745334625244,
      "epoch": 3.9724265738192845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003509341619363474,
      "loss": 5.0797,
      "mean_token_accuracy": 0.22394702285528184,
      "num_tokens": 78953764.0,
      "step": 35370
    },
    {
      "entropy": 5.693164348602295,
      "epoch": 3.972988150727242,
      "grad_norm": 1.125,
      "learning_rate": 0.00035089636749997095,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22614773362874985,
      "num_tokens": 78964781.0,
      "step": 35375
    },
    {
      "entropy": 5.74279637336731,
      "epoch": 3.9735497276351994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035085857064702487,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2241686388850212,
      "num_tokens": 78976325.0,
      "step": 35380
    },
    {
      "entropy": 5.627242755889893,
      "epoch": 3.974111304543157,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003508207713787123,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23464736789464952,
      "num_tokens": 78987174.0,
      "step": 35385
    },
    {
      "entropy": 5.652206659317017,
      "epoch": 3.9746728814511147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.9434,
      "mean_token_accuracy": 0.229256671667099,
      "num_tokens": 78999195.0,
      "step": 35390
    },
    {
      "entropy": 5.563788938522339,
      "epoch": 3.9752344583590724,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.805,
      "mean_token_accuracy": 0.2347317725419998,
      "num_tokens": 79010981.0,
      "step": 35395
    },
    {
      "entropy": 5.638928365707398,
      "epoch": 3.97579603526703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22971960455179213,
      "num_tokens": 79022059.0,
      "step": 35400
    },
    {
      "entropy": 5.767231512069702,
      "epoch": 3.9763576121749873,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035066955017587135,
      "loss": 5.0856,
      "mean_token_accuracy": 0.22902196347713472,
      "num_tokens": 79033261.0,
      "step": 35405
    },
    {
      "entropy": 5.628084421157837,
      "epoch": 3.976919189082945,
      "grad_norm": 1.21875,
      "learning_rate": 0.000350631738848782,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22427176982164382,
      "num_tokens": 79044518.0,
      "step": 35410
    },
    {
      "entropy": 5.65206561088562,
      "epoch": 3.9774807659909026,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003505939251135488,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22250538915395737,
      "num_tokens": 79055202.0,
      "step": 35415
    },
    {
      "entropy": 5.710760164260864,
      "epoch": 3.97804234289886,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003505561089713756,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22386627048254013,
      "num_tokens": 79066448.0,
      "step": 35420
    },
    {
      "entropy": 5.672876834869385,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003505182904234667,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2296256825327873,
      "num_tokens": 79077745.0,
      "step": 35425
    },
    {
      "entropy": 5.740192174911499,
      "epoch": 3.979165496714775,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003504804694710261,
      "loss": 5.0453,
      "mean_token_accuracy": 0.21713448762893678,
      "num_tokens": 79088011.0,
      "step": 35430
    },
    {
      "entropy": 5.775643301010132,
      "epoch": 3.9797270736227324,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003504426461152581,
      "loss": 5.0671,
      "mean_token_accuracy": 0.2128016769886017,
      "num_tokens": 79098718.0,
      "step": 35435
    },
    {
      "entropy": 5.662784576416016,
      "epoch": 3.98028865053069,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22170129418373108,
      "num_tokens": 79108216.0,
      "step": 35440
    },
    {
      "entropy": 5.637986850738526,
      "epoch": 3.9808502274386477,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22388456612825394,
      "num_tokens": 79119709.0,
      "step": 35445
    },
    {
      "entropy": 5.705166673660278,
      "epoch": 3.9814118043466054,
      "grad_norm": 1.15625,
      "learning_rate": 0.000350329161640033,
      "loss": 5.0396,
      "mean_token_accuracy": 0.22701950073242189,
      "num_tokens": 79131824.0,
      "step": 35450
    },
    {
      "entropy": 5.661404943466186,
      "epoch": 3.981973381254563,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035029132868299914,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21900811344385146,
      "num_tokens": 79142855.0,
      "step": 35455
    },
    {
      "entropy": 5.696152019500732,
      "epoch": 3.9825349581625202,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035025349332866017,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21638162881135942,
      "num_tokens": 79153759.0,
      "step": 35460
    },
    {
      "entropy": 5.6866453170776365,
      "epoch": 3.983096535070478,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035021565557822085,
      "loss": 4.9499,
      "mean_token_accuracy": 0.23079103231430054,
      "num_tokens": 79163586.0,
      "step": 35465
    },
    {
      "entropy": 5.723671817779541,
      "epoch": 3.9836581119784356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003501778154328858,
      "loss": 5.0373,
      "mean_token_accuracy": 0.21622067391872407,
      "num_tokens": 79175067.0,
      "step": 35470
    },
    {
      "entropy": 5.556923580169678,
      "epoch": 3.9842196888863928,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.8546,
      "mean_token_accuracy": 0.23320851475000381,
      "num_tokens": 79186039.0,
      "step": 35475
    },
    {
      "entropy": 5.673986434936523,
      "epoch": 3.9847812657943504,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035010212796234827,
      "loss": 5.0039,
      "mean_token_accuracy": 0.2230786547064781,
      "num_tokens": 79198571.0,
      "step": 35480
    },
    {
      "entropy": 5.553048324584961,
      "epoch": 3.985342842702308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.8805,
      "mean_token_accuracy": 0.23154521286487578,
      "num_tokens": 79210253.0,
      "step": 35485
    },
    {
      "entropy": 5.6039674282073975,
      "epoch": 3.9859044196102658,
      "grad_norm": 1.125,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22889718264341355,
      "num_tokens": 79221611.0,
      "step": 35490
    },
    {
      "entropy": 5.691542482376098,
      "epoch": 3.9864659965182234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034998857882494823,
      "loss": 5.0577,
      "mean_token_accuracy": 0.22186990529298783,
      "num_tokens": 79233270.0,
      "step": 35495
    },
    {
      "entropy": 5.673780059814453,
      "epoch": 3.9870275734261806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034995072433554377,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21655985116958618,
      "num_tokens": 79245338.0,
      "step": 35500
    },
    {
      "entropy": 5.740801620483398,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003499128674596792,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22188702821731568,
      "num_tokens": 79255558.0,
      "step": 35505
    },
    {
      "entropy": 5.760221719741821,
      "epoch": 3.988150727242096,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034987500819855976,
      "loss": 5.0398,
      "mean_token_accuracy": 0.23006787300109863,
      "num_tokens": 79266075.0,
      "step": 35510
    },
    {
      "entropy": 5.729678678512573,
      "epoch": 3.988712304150053,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034983714655339115,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2242028146982193,
      "num_tokens": 79277667.0,
      "step": 35515
    },
    {
      "entropy": 5.627741813659668,
      "epoch": 3.989273881058011,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003497992825253787,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22878700494766235,
      "num_tokens": 79288755.0,
      "step": 35520
    },
    {
      "entropy": 5.6460041999816895,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.296875,
      "learning_rate": 0.00034976141611572817,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22819837629795076,
      "num_tokens": 79299048.0,
      "step": 35525
    },
    {
      "entropy": 5.607081604003906,
      "epoch": 3.9903970348739257,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22828857153654097,
      "num_tokens": 79310847.0,
      "step": 35530
    },
    {
      "entropy": 5.696435260772705,
      "epoch": 3.9909586117818834,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034968567615633516,
      "loss": 5.0039,
      "mean_token_accuracy": 0.22275372445583344,
      "num_tokens": 79322496.0,
      "step": 35535
    },
    {
      "entropy": 5.674264287948608,
      "epoch": 3.991520188689841,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034964780260900445,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22396553307771683,
      "num_tokens": 79334216.0,
      "step": 35540
    },
    {
      "entropy": 5.580070209503174,
      "epoch": 3.9920817655977987,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23496217727661134,
      "num_tokens": 79344320.0,
      "step": 35545
    },
    {
      "entropy": 5.689203405380249,
      "epoch": 3.9926433425057564,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034957204838510406,
      "loss": 5.1242,
      "mean_token_accuracy": 0.21611237972974778,
      "num_tokens": 79355775.0,
      "step": 35550
    },
    {
      "entropy": 5.634100008010864,
      "epoch": 3.9932049194137136,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.8725,
      "mean_token_accuracy": 0.23061866462230682,
      "num_tokens": 79366866.0,
      "step": 35555
    },
    {
      "entropy": 5.642268657684326,
      "epoch": 3.9937664963216712,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000349496284663592,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2149496778845787,
      "num_tokens": 79378599.0,
      "step": 35560
    },
    {
      "entropy": 5.616348648071289,
      "epoch": 3.994328073229629,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.8683,
      "mean_token_accuracy": 0.2317390412092209,
      "num_tokens": 79389283.0,
      "step": 35565
    },
    {
      "entropy": 5.679325675964355,
      "epoch": 3.994889650137586,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003494205114541176,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21899312138557434,
      "num_tokens": 79400858.0,
      "step": 35570
    },
    {
      "entropy": 5.7094801902771,
      "epoch": 3.995451227045544,
      "grad_norm": 1.359375,
      "learning_rate": 0.00034938262129441026,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2181319147348404,
      "num_tokens": 79411122.0,
      "step": 35575
    },
    {
      "entropy": 5.595687770843506,
      "epoch": 3.9960128039535014,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.845,
      "mean_token_accuracy": 0.23444899469614028,
      "num_tokens": 79420153.0,
      "step": 35580
    },
    {
      "entropy": 5.7296521186828615,
      "epoch": 3.996574380861459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003493068338710877,
      "loss": 5.0871,
      "mean_token_accuracy": 0.2249401643872261,
      "num_tokens": 79432964.0,
      "step": 35585
    },
    {
      "entropy": 5.654682302474976,
      "epoch": 3.9971359577694168,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003492689366098855,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2301451027393341,
      "num_tokens": 79445191.0,
      "step": 35590
    },
    {
      "entropy": 5.664460849761963,
      "epoch": 3.997697534677374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.9305,
      "mean_token_accuracy": 0.23067076802253722,
      "num_tokens": 79455553.0,
      "step": 35595
    },
    {
      "entropy": 5.683468151092529,
      "epoch": 3.9982591115853316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003491931349944325,
      "loss": 5.0084,
      "mean_token_accuracy": 0.22189596593379973,
      "num_tokens": 79466781.0,
      "step": 35600
    },
    {
      "entropy": 5.596692705154419,
      "epoch": 3.9988206884932893,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034915523064259533,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2315227508544922,
      "num_tokens": 79478656.0,
      "step": 35605
    },
    {
      "entropy": 5.610975503921509,
      "epoch": 3.9993822654012465,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22631843239068986,
      "num_tokens": 79489763.0,
      "step": 35610
    },
    {
      "entropy": 5.671654272079468,
      "epoch": 3.999943842309204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000349079414856734,
      "loss": 5.0583,
      "mean_token_accuracy": 0.22405458837747574,
      "num_tokens": 79501828.0,
      "step": 35615
    },
    {
      "entropy": 5.7000807126363116,
      "epoch": 4.000449261526366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22106686731179556,
      "num_tokens": 79510316.0,
      "step": 35620
    },
    {
      "entropy": 5.604257011413575,
      "epoch": 4.001010838434324,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.883,
      "mean_token_accuracy": 0.2304907888174057,
      "num_tokens": 79521555.0,
      "step": 35625
    },
    {
      "entropy": 5.7420121192932125,
      "epoch": 4.001572415342281,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034896567349058027,
      "loss": 5.0158,
      "mean_token_accuracy": 0.2266194239258766,
      "num_tokens": 79533147.0,
      "step": 35630
    },
    {
      "entropy": 5.727225828170776,
      "epoch": 4.002133992250239,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034892775499006126,
      "loss": 5.0382,
      "mean_token_accuracy": 0.22355887293815613,
      "num_tokens": 79544695.0,
      "step": 35635
    },
    {
      "entropy": 5.569227981567383,
      "epoch": 4.002695569158196,
      "grad_norm": 1.125,
      "learning_rate": 0.0003488898341356538,
      "loss": 4.7978,
      "mean_token_accuracy": 0.24077791720628738,
      "num_tokens": 79555740.0,
      "step": 35640
    },
    {
      "entropy": 5.667155408859253,
      "epoch": 4.0032571460661535,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22823058366775512,
      "num_tokens": 79565648.0,
      "step": 35645
    },
    {
      "entropy": 5.723361539840698,
      "epoch": 4.003818722974112,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.9722,
      "mean_token_accuracy": 0.22938107699155807,
      "num_tokens": 79577107.0,
      "step": 35650
    },
    {
      "entropy": 5.648545789718628,
      "epoch": 4.004380299882069,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2307181403040886,
      "num_tokens": 79586689.0,
      "step": 35655
    },
    {
      "entropy": 5.703628969192505,
      "epoch": 4.004941876790026,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003487381272032897,
      "loss": 5.0002,
      "mean_token_accuracy": 0.22494050860404968,
      "num_tokens": 79597517.0,
      "step": 35660
    },
    {
      "entropy": 5.6912394046783445,
      "epoch": 4.005503453697984,
      "grad_norm": 1.0625,
      "learning_rate": 0.000348700194597553,
      "loss": 4.948,
      "mean_token_accuracy": 0.22564575374126433,
      "num_tokens": 79609516.0,
      "step": 35665
    },
    {
      "entropy": 5.609646940231324,
      "epoch": 4.006065030605941,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23271773755550385,
      "num_tokens": 79620706.0,
      "step": 35670
    },
    {
      "entropy": 5.622392225265503,
      "epoch": 4.006626607513899,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003486243223473592,
      "loss": 4.8384,
      "mean_token_accuracy": 0.22739255875349046,
      "num_tokens": 79630897.0,
      "step": 35675
    },
    {
      "entropy": 5.62032904624939,
      "epoch": 4.007188184421857,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.92,
      "mean_token_accuracy": 0.22915639132261276,
      "num_tokens": 79642426.0,
      "step": 35680
    },
    {
      "entropy": 5.669781112670899,
      "epoch": 4.007749761329814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2276816889643669,
      "num_tokens": 79653491.0,
      "step": 35685
    },
    {
      "entropy": 5.600366163253784,
      "epoch": 4.008311338237772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.8509,
      "mean_token_accuracy": 0.2329068049788475,
      "num_tokens": 79665329.0,
      "step": 35690
    },
    {
      "entropy": 5.640780687332153,
      "epoch": 4.008872915145729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22984424531459807,
      "num_tokens": 79676552.0,
      "step": 35695
    },
    {
      "entropy": 5.6005395412445065,
      "epoch": 4.009434492053686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034843460071904407,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23180220276117325,
      "num_tokens": 79687359.0,
      "step": 35700
    },
    {
      "entropy": 5.628088712692261,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.838,
      "mean_token_accuracy": 0.23301016688346862,
      "num_tokens": 79698728.0,
      "step": 35705
    },
    {
      "entropy": 5.682158899307251,
      "epoch": 4.010557645869602,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003483586956919572,
      "loss": 5.0341,
      "mean_token_accuracy": 0.22768660634756088,
      "num_tokens": 79709018.0,
      "step": 35710
    },
    {
      "entropy": 5.672723770141602,
      "epoch": 4.011119222777559,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22661390751600266,
      "num_tokens": 79720166.0,
      "step": 35715
    },
    {
      "entropy": 5.644338512420655,
      "epoch": 4.011680799685517,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003482827813217932,
      "loss": 4.7897,
      "mean_token_accuracy": 0.24143872410058975,
      "num_tokens": 79730467.0,
      "step": 35720
    },
    {
      "entropy": 5.687226247787476,
      "epoch": 4.012242376593474,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.9554,
      "mean_token_accuracy": 0.23261012732982636,
      "num_tokens": 79740570.0,
      "step": 35725
    },
    {
      "entropy": 5.63517165184021,
      "epoch": 4.012803953501432,
      "grad_norm": 1.125,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22902173548936844,
      "num_tokens": 79751856.0,
      "step": 35730
    },
    {
      "entropy": 5.651562166213989,
      "epoch": 4.01336553040939,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.916,
      "mean_token_accuracy": 0.23849166482686995,
      "num_tokens": 79763267.0,
      "step": 35735
    },
    {
      "entropy": 5.704994106292725,
      "epoch": 4.013927107317347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003481309245909088,
      "loss": 5.026,
      "mean_token_accuracy": 0.2286354273557663,
      "num_tokens": 79775202.0,
      "step": 35740
    },
    {
      "entropy": 5.655553150177002,
      "epoch": 4.014488684225305,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22058544903993607,
      "num_tokens": 79787087.0,
      "step": 35745
    },
    {
      "entropy": 5.739774942398071,
      "epoch": 4.015050261133262,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003480549822495293,
      "loss": 5.0159,
      "mean_token_accuracy": 0.22267971485853194,
      "num_tokens": 79799014.0,
      "step": 35750
    },
    {
      "entropy": 5.65931282043457,
      "epoch": 4.015611838041219,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.8728,
      "mean_token_accuracy": 0.236114501953125,
      "num_tokens": 79810455.0,
      "step": 35755
    },
    {
      "entropy": 5.708392000198364,
      "epoch": 4.0161734149491775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23157034516334535,
      "num_tokens": 79821614.0,
      "step": 35760
    },
    {
      "entropy": 5.670370292663574,
      "epoch": 4.016734991857135,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003479410512947407,
      "loss": 5.0585,
      "mean_token_accuracy": 0.22349075376987457,
      "num_tokens": 79832418.0,
      "step": 35765
    },
    {
      "entropy": 5.702157068252563,
      "epoch": 4.017296568765093,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00034790306966325606,
      "loss": 5.0306,
      "mean_token_accuracy": 0.22623196244239807,
      "num_tokens": 79844234.0,
      "step": 35770
    },
    {
      "entropy": 5.617024755477905,
      "epoch": 4.01785814567305,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23480382412672043,
      "num_tokens": 79855284.0,
      "step": 35775
    },
    {
      "entropy": 5.619846820831299,
      "epoch": 4.018419722581007,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.8589,
      "mean_token_accuracy": 0.22767732441425323,
      "num_tokens": 79866169.0,
      "step": 35780
    },
    {
      "entropy": 5.699887895584107,
      "epoch": 4.018981299488965,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034778911084606793,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22334081530570984,
      "num_tokens": 79878044.0,
      "step": 35785
    },
    {
      "entropy": 5.605615663528442,
      "epoch": 4.0195428763969225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.8673,
      "mean_token_accuracy": 0.23628222197294235,
      "num_tokens": 79888859.0,
      "step": 35790
    },
    {
      "entropy": 5.695027637481689,
      "epoch": 4.02010445330488,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23404659032821656,
      "num_tokens": 79899886.0,
      "step": 35795
    },
    {
      "entropy": 5.684042978286743,
      "epoch": 4.020666030212838,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.9715,
      "mean_token_accuracy": 0.23565018475055693,
      "num_tokens": 79911752.0,
      "step": 35800
    },
    {
      "entropy": 5.6244913101196286,
      "epoch": 4.021227607120795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2239330530166626,
      "num_tokens": 79923076.0,
      "step": 35805
    },
    {
      "entropy": 5.694273805618286,
      "epoch": 4.021789184028753,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23016718477010728,
      "num_tokens": 79933866.0,
      "step": 35810
    },
    {
      "entropy": 5.672938060760498,
      "epoch": 4.02235076093671,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.9252,
      "mean_token_accuracy": 0.23624117970466613,
      "num_tokens": 79946046.0,
      "step": 35815
    },
    {
      "entropy": 5.636408090591431,
      "epoch": 4.022912337844668,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.8857,
      "mean_token_accuracy": 0.23540912717580795,
      "num_tokens": 79957887.0,
      "step": 35820
    },
    {
      "entropy": 5.57770938873291,
      "epoch": 4.023473914752626,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2341275230050087,
      "num_tokens": 79968101.0,
      "step": 35825
    },
    {
      "entropy": 5.625013780593872,
      "epoch": 4.024035491660583,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22596494257450103,
      "num_tokens": 79978176.0,
      "step": 35830
    },
    {
      "entropy": 5.652029228210449,
      "epoch": 4.02459706856854,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.9044,
      "mean_token_accuracy": 0.23173956871032714,
      "num_tokens": 79990324.0,
      "step": 35835
    },
    {
      "entropy": 5.670257949829102,
      "epoch": 4.025158645476498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.9425,
      "mean_token_accuracy": 0.23047096878290177,
      "num_tokens": 80001898.0,
      "step": 35840
    },
    {
      "entropy": 5.608722686767578,
      "epoch": 4.0257202223844555,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003473330673169572,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22525845766067504,
      "num_tokens": 80012072.0,
      "step": 35845
    },
    {
      "entropy": 5.615287351608276,
      "epoch": 4.026281799292413,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22669098228216172,
      "num_tokens": 80022718.0,
      "step": 35850
    },
    {
      "entropy": 5.751804971694947,
      "epoch": 4.026843376200371,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22649255245923997,
      "num_tokens": 80033899.0,
      "step": 35855
    },
    {
      "entropy": 5.764189577102661,
      "epoch": 4.027404953108328,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22726648151874543,
      "num_tokens": 80045398.0,
      "step": 35860
    },
    {
      "entropy": 5.709136867523194,
      "epoch": 4.027966530016286,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003471809790043667,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2293703868985176,
      "num_tokens": 80055186.0,
      "step": 35865
    },
    {
      "entropy": 5.644249057769775,
      "epoch": 4.028528106924243,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.9573,
      "mean_token_accuracy": 0.23329757004976273,
      "num_tokens": 80065545.0,
      "step": 35870
    },
    {
      "entropy": 5.653292083740235,
      "epoch": 4.029089683832201,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034710492105360653,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2311354547739029,
      "num_tokens": 80077451.0,
      "step": 35875
    },
    {
      "entropy": 5.60709023475647,
      "epoch": 4.029651260740159,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23539225310087203,
      "num_tokens": 80088215.0,
      "step": 35880
    },
    {
      "entropy": 5.5677347660064695,
      "epoch": 4.030212837648116,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.8004,
      "mean_token_accuracy": 0.23597032576799393,
      "num_tokens": 80098393.0,
      "step": 35885
    },
    {
      "entropy": 5.652753210067749,
      "epoch": 4.030774414556073,
      "grad_norm": 1.140625,
      "learning_rate": 0.000346990816911629,
      "loss": 4.914,
      "mean_token_accuracy": 0.22395094037055968,
      "num_tokens": 80110183.0,
      "step": 35890
    },
    {
      "entropy": 5.588706684112549,
      "epoch": 4.031335991464031,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.8973,
      "mean_token_accuracy": 0.229658043384552,
      "num_tokens": 80120954.0,
      "step": 35895
    },
    {
      "entropy": 5.589990043640137,
      "epoch": 4.031897568371988,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23498144298791884,
      "num_tokens": 80131607.0,
      "step": 35900
    },
    {
      "entropy": 5.576420831680298,
      "epoch": 4.032459145279946,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.882,
      "mean_token_accuracy": 0.23124945908784866,
      "num_tokens": 80141700.0,
      "step": 35905
    },
    {
      "entropy": 5.630169439315796,
      "epoch": 4.033020722187904,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.9063,
      "mean_token_accuracy": 0.23158090710639953,
      "num_tokens": 80151776.0,
      "step": 35910
    },
    {
      "entropy": 5.674283695220947,
      "epoch": 4.033582299095861,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22656586170196533,
      "num_tokens": 80162838.0,
      "step": 35915
    },
    {
      "entropy": 5.65856146812439,
      "epoch": 4.034143876003819,
      "grad_norm": 1.125,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22559571713209153,
      "num_tokens": 80174104.0,
      "step": 35920
    },
    {
      "entropy": 5.658313369750976,
      "epoch": 4.034705452911776,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034672449374266575,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22278406471014023,
      "num_tokens": 80185015.0,
      "step": 35925
    },
    {
      "entropy": 5.674750757217407,
      "epoch": 4.0352670298197335,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23600404113531112,
      "num_tokens": 80196893.0,
      "step": 35930
    },
    {
      "entropy": 5.726228141784668,
      "epoch": 4.035828606727692,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.9698,
      "mean_token_accuracy": 0.23054434061050416,
      "num_tokens": 80209910.0,
      "step": 35935
    },
    {
      "entropy": 5.5982996940612795,
      "epoch": 4.036390183635649,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22841207683086395,
      "num_tokens": 80219780.0,
      "step": 35940
    },
    {
      "entropy": 5.676177740097046,
      "epoch": 4.036951760543606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003465722588058497,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22479236572980882,
      "num_tokens": 80231316.0,
      "step": 35945
    },
    {
      "entropy": 5.6963001728057865,
      "epoch": 4.037513337451564,
      "grad_norm": 1.25,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22584600895643234,
      "num_tokens": 80242622.0,
      "step": 35950
    },
    {
      "entropy": 5.7320136547088625,
      "epoch": 4.038074914359521,
      "grad_norm": 1.25,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.9562,
      "mean_token_accuracy": 0.23380132168531417,
      "num_tokens": 80255357.0,
      "step": 35955
    },
    {
      "entropy": 5.704930353164673,
      "epoch": 4.0386364912674795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2266574263572693,
      "num_tokens": 80267430.0,
      "step": 35960
    },
    {
      "entropy": 5.632170486450195,
      "epoch": 4.039198068175437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.8907,
      "mean_token_accuracy": 0.23173805177211762,
      "num_tokens": 80277820.0,
      "step": 35965
    },
    {
      "entropy": 5.731638956069946,
      "epoch": 4.039759645083394,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034638191386893115,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2179819718003273,
      "num_tokens": 80289559.0,
      "step": 35970
    },
    {
      "entropy": 5.659202766418457,
      "epoch": 4.040321221991352,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034634383805821504,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2229696422815323,
      "num_tokens": 80300452.0,
      "step": 35975
    },
    {
      "entropy": 5.676018095016479,
      "epoch": 4.040882798899309,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22834801077842712,
      "num_tokens": 80311403.0,
      "step": 35980
    },
    {
      "entropy": 5.662664699554443,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000346267679623149,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23174329102039337,
      "num_tokens": 80323245.0,
      "step": 35985
    },
    {
      "entropy": 5.6373086929321286,
      "epoch": 4.042005952715225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2251288563013077,
      "num_tokens": 80333666.0,
      "step": 35990
    },
    {
      "entropy": 5.688879060745239,
      "epoch": 4.042567529623182,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23670517802238464,
      "num_tokens": 80345291.0,
      "step": 35995
    },
    {
      "entropy": 5.607341384887695,
      "epoch": 4.04312910653114,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23364525586366652,
      "num_tokens": 80355612.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.3869820013776835,
      "eval_loss": 5.065199375152588,
      "eval_mean_token_accuracy": 0.2304669190721857,
      "eval_num_tokens": 80355612.0,
      "eval_runtime": 18.268,
      "eval_samples_per_second": 1697.669,
      "eval_steps_per_second": 212.229,
      "step": 36000
    },
    {
      "entropy": 5.590196752548218,
      "epoch": 4.043690683439097,
      "grad_norm": 1.15625,
      "learning_rate": 0.000346115335532431,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23283191472291948,
      "num_tokens": 80365952.0,
      "step": 36005
    },
    {
      "entropy": 5.6591204643249515,
      "epoch": 4.044252260347054,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034607724384587006,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22825287133455277,
      "num_tokens": 80376962.0,
      "step": 36010
    },
    {
      "entropy": 5.744942855834961,
      "epoch": 4.044813837255012,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034603914989618197,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21866713017225264,
      "num_tokens": 80388671.0,
      "step": 36015
    },
    {
      "entropy": 5.735807800292969,
      "epoch": 4.04537541416297,
      "grad_norm": 1.25,
      "learning_rate": 0.0003460010536845799,
      "loss": 4.957,
      "mean_token_accuracy": 0.23126793801784515,
      "num_tokens": 80399215.0,
      "step": 36020
    },
    {
      "entropy": 5.66960301399231,
      "epoch": 4.045936991070927,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003459629552122767,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22636230289936066,
      "num_tokens": 80410590.0,
      "step": 36025
    },
    {
      "entropy": 5.629859495162964,
      "epoch": 4.046498567978885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003459248544804855,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2379998415708542,
      "num_tokens": 80422085.0,
      "step": 36030
    },
    {
      "entropy": 5.579884719848633,
      "epoch": 4.047060144886842,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003458867514904194,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23552334755659105,
      "num_tokens": 80434389.0,
      "step": 36035
    },
    {
      "entropy": 5.645457172393799,
      "epoch": 4.047621721794799,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003458486462432916,
      "loss": 4.9425,
      "mean_token_accuracy": 0.23415064066648483,
      "num_tokens": 80445073.0,
      "step": 36040
    },
    {
      "entropy": 5.6494221687316895,
      "epoch": 4.0481832987027575,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034581053874031554,
      "loss": 4.8691,
      "mean_token_accuracy": 0.2362838178873062,
      "num_tokens": 80455206.0,
      "step": 36045
    },
    {
      "entropy": 5.641750717163086,
      "epoch": 4.048744875610715,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003457724289827044,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2348868116736412,
      "num_tokens": 80467707.0,
      "step": 36050
    },
    {
      "entropy": 5.6485360622406,
      "epoch": 4.049306452518673,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003457343169716716,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23087506592273713,
      "num_tokens": 80478493.0,
      "step": 36055
    },
    {
      "entropy": 5.611560583114624,
      "epoch": 4.04986802942663,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034569620270843087,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22765483409166337,
      "num_tokens": 80490057.0,
      "step": 36060
    },
    {
      "entropy": 5.6537707328796385,
      "epoch": 4.050429606334587,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034565808619419554,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22802636921405792,
      "num_tokens": 80500590.0,
      "step": 36065
    },
    {
      "entropy": 5.647326850891114,
      "epoch": 4.050991183242545,
      "grad_norm": 1.25,
      "learning_rate": 0.0003456199674301793,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2311888501048088,
      "num_tokens": 80510836.0,
      "step": 36070
    },
    {
      "entropy": 5.6292534351348875,
      "epoch": 4.051552760150503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000345581846417596,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2265482947230339,
      "num_tokens": 80521770.0,
      "step": 36075
    },
    {
      "entropy": 5.730333375930786,
      "epoch": 4.05211433705846,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034554372315765927,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2360702484846115,
      "num_tokens": 80532331.0,
      "step": 36080
    },
    {
      "entropy": 5.746664905548096,
      "epoch": 4.052675913966418,
      "grad_norm": 1.125,
      "learning_rate": 0.0003455055976515831,
      "loss": 4.912,
      "mean_token_accuracy": 0.22898183315992354,
      "num_tokens": 80544858.0,
      "step": 36085
    },
    {
      "entropy": 5.695457839965821,
      "epoch": 4.053237490874375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034546746990058125,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22210728675127028,
      "num_tokens": 80556493.0,
      "step": 36090
    },
    {
      "entropy": 5.679467487335205,
      "epoch": 4.053799067782332,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034542933990586795,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22822870910167695,
      "num_tokens": 80567416.0,
      "step": 36095
    },
    {
      "entropy": 5.657695150375366,
      "epoch": 4.0543606446902904,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034539120766865713,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22726806104183198,
      "num_tokens": 80578175.0,
      "step": 36100
    },
    {
      "entropy": 5.631509160995483,
      "epoch": 4.054922221598248,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003453530731901628,
      "loss": 4.8512,
      "mean_token_accuracy": 0.23446832597255707,
      "num_tokens": 80588755.0,
      "step": 36105
    },
    {
      "entropy": 5.66676893234253,
      "epoch": 4.055483798506206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003453149364715995,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22928957790136337,
      "num_tokens": 80599734.0,
      "step": 36110
    },
    {
      "entropy": 5.624989986419678,
      "epoch": 4.056045375414163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034527679751418126,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23941211253404618,
      "num_tokens": 80610765.0,
      "step": 36115
    },
    {
      "entropy": 5.722071266174316,
      "epoch": 4.05660695232212,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003452386563191226,
      "loss": 4.9732,
      "mean_token_accuracy": 0.23175185173749924,
      "num_tokens": 80622680.0,
      "step": 36120
    },
    {
      "entropy": 5.710647630691528,
      "epoch": 4.057168529230078,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034520051288763784,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2250156283378601,
      "num_tokens": 80633115.0,
      "step": 36125
    },
    {
      "entropy": 5.673319101333618,
      "epoch": 4.0577301061380355,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003451623672209415,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22996209263801576,
      "num_tokens": 80643750.0,
      "step": 36130
    },
    {
      "entropy": 5.668999481201172,
      "epoch": 4.058291683045993,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034512421932024825,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22619499564170836,
      "num_tokens": 80654392.0,
      "step": 36135
    },
    {
      "entropy": 5.596457529067993,
      "epoch": 4.058853259953951,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003450860691867726,
      "loss": 4.9121,
      "mean_token_accuracy": 0.23076938837766647,
      "num_tokens": 80665509.0,
      "step": 36140
    },
    {
      "entropy": 5.631325387954712,
      "epoch": 4.059414836861908,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034504791682172934,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2317704513669014,
      "num_tokens": 80676895.0,
      "step": 36145
    },
    {
      "entropy": 5.694951057434082,
      "epoch": 4.059976413769866,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003450097622263333,
      "loss": 4.9217,
      "mean_token_accuracy": 0.23010152876377105,
      "num_tokens": 80688321.0,
      "step": 36150
    },
    {
      "entropy": 5.621122550964356,
      "epoch": 4.060537990677823,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034497160540179924,
      "loss": 4.817,
      "mean_token_accuracy": 0.24346933215856553,
      "num_tokens": 80699534.0,
      "step": 36155
    },
    {
      "entropy": 5.522291946411133,
      "epoch": 4.061099567585781,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034493344634934205,
      "loss": 4.8367,
      "mean_token_accuracy": 0.24181941598653794,
      "num_tokens": 80710999.0,
      "step": 36160
    },
    {
      "entropy": 5.586549091339111,
      "epoch": 4.061661144493739,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003448952850701769,
      "loss": 4.9071,
      "mean_token_accuracy": 0.231584532558918,
      "num_tokens": 80722943.0,
      "step": 36165
    },
    {
      "entropy": 5.633302354812622,
      "epoch": 4.062222721401696,
      "grad_norm": 1.25,
      "learning_rate": 0.00034485712156551873,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2234898701310158,
      "num_tokens": 80733142.0,
      "step": 36170
    },
    {
      "entropy": 5.625020933151245,
      "epoch": 4.062784298309653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034481895583658267,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2269438922405243,
      "num_tokens": 80745228.0,
      "step": 36175
    },
    {
      "entropy": 5.680726671218872,
      "epoch": 4.063345875217611,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034478078788458406,
      "loss": 4.9436,
      "mean_token_accuracy": 0.23229171633720397,
      "num_tokens": 80756866.0,
      "step": 36180
    },
    {
      "entropy": 5.605130004882812,
      "epoch": 4.0639074521255685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000344742617710738,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2511804893612862,
      "num_tokens": 80767541.0,
      "step": 36185
    },
    {
      "entropy": 5.615648365020752,
      "epoch": 4.064469029033527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034470444531626003,
      "loss": 4.918,
      "mean_token_accuracy": 0.23032878041267396,
      "num_tokens": 80779577.0,
      "step": 36190
    },
    {
      "entropy": 5.695377445220947,
      "epoch": 4.065030605941484,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034466627070236545,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2324230670928955,
      "num_tokens": 80790569.0,
      "step": 36195
    },
    {
      "entropy": 5.755552434921265,
      "epoch": 4.065592182849441,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003446280938702698,
      "loss": 5.035,
      "mean_token_accuracy": 0.22064341753721237,
      "num_tokens": 80800887.0,
      "step": 36200
    },
    {
      "entropy": 5.702422571182251,
      "epoch": 4.066153759757399,
      "grad_norm": 1.125,
      "learning_rate": 0.00034458991482118865,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22676954716444014,
      "num_tokens": 80811967.0,
      "step": 36205
    },
    {
      "entropy": 5.618140602111817,
      "epoch": 4.066715336665356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003445517335563376,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23278744220733644,
      "num_tokens": 80824770.0,
      "step": 36210
    },
    {
      "entropy": 5.609582710266113,
      "epoch": 4.067276913573314,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034451355007693237,
      "loss": 4.935,
      "mean_token_accuracy": 0.23221534341573716,
      "num_tokens": 80834913.0,
      "step": 36215
    },
    {
      "entropy": 5.645982360839843,
      "epoch": 4.067838490481272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003444753643841888,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22470242232084275,
      "num_tokens": 80846555.0,
      "step": 36220
    },
    {
      "entropy": 5.73993444442749,
      "epoch": 4.068400067389229,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034443717647932265,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21822172850370408,
      "num_tokens": 80857774.0,
      "step": 36225
    },
    {
      "entropy": 5.778255271911621,
      "epoch": 4.068961644297186,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003443989863635498,
      "loss": 5.0084,
      "mean_token_accuracy": 0.22369984537363052,
      "num_tokens": 80868618.0,
      "step": 36230
    },
    {
      "entropy": 5.67216796875,
      "epoch": 4.069523221205144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034436079403808647,
      "loss": 5.0219,
      "mean_token_accuracy": 0.22057662904262543,
      "num_tokens": 80880319.0,
      "step": 36235
    },
    {
      "entropy": 5.67352442741394,
      "epoch": 4.070084798113101,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034432259950414846,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22778881192207337,
      "num_tokens": 80891194.0,
      "step": 36240
    },
    {
      "entropy": 5.5992552757263185,
      "epoch": 4.0706463750210595,
      "grad_norm": 1.140625,
      "learning_rate": 0.000344284402762952,
      "loss": 4.8067,
      "mean_token_accuracy": 0.22886781841516496,
      "num_tokens": 80902322.0,
      "step": 36245
    },
    {
      "entropy": 5.725928688049317,
      "epoch": 4.071207951929017,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003442462038157132,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22325926423072814,
      "num_tokens": 80913324.0,
      "step": 36250
    },
    {
      "entropy": 5.584973621368408,
      "epoch": 4.071769528836974,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003442080026636485,
      "loss": 4.8542,
      "mean_token_accuracy": 0.2326980337500572,
      "num_tokens": 80923360.0,
      "step": 36255
    },
    {
      "entropy": 5.69178614616394,
      "epoch": 4.072331105744932,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003441697993079742,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22249678373336793,
      "num_tokens": 80935359.0,
      "step": 36260
    },
    {
      "entropy": 5.625059032440186,
      "epoch": 4.072892682652889,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003441315937499066,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23238444626331328,
      "num_tokens": 80945885.0,
      "step": 36265
    },
    {
      "entropy": 5.705935382843018,
      "epoch": 4.0734542595608465,
      "grad_norm": 1.328125,
      "learning_rate": 0.00034409338599066224,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2299763187766075,
      "num_tokens": 80955700.0,
      "step": 36270
    },
    {
      "entropy": 5.630369520187378,
      "epoch": 4.074015836468805,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003440551760314577,
      "loss": 4.9157,
      "mean_token_accuracy": 0.23645513206720353,
      "num_tokens": 80967130.0,
      "step": 36275
    },
    {
      "entropy": 5.643083620071411,
      "epoch": 4.074577413376762,
      "grad_norm": 1.125,
      "learning_rate": 0.00034401696387350954,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22980746179819106,
      "num_tokens": 80979677.0,
      "step": 36280
    },
    {
      "entropy": 5.678291130065918,
      "epoch": 4.075138990284719,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003439787495180345,
      "loss": 4.9398,
      "mean_token_accuracy": 0.23016615211963654,
      "num_tokens": 80990563.0,
      "step": 36285
    },
    {
      "entropy": 5.628010988235474,
      "epoch": 4.075700567192677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034394053296624936,
      "loss": 4.9189,
      "mean_token_accuracy": 0.23589287400245668,
      "num_tokens": 81001731.0,
      "step": 36290
    },
    {
      "entropy": 5.684636783599854,
      "epoch": 4.076262144100634,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003439023142193709,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22589996010065078,
      "num_tokens": 81013072.0,
      "step": 36295
    },
    {
      "entropy": 5.6545875549316404,
      "epoch": 4.0768237210085925,
      "grad_norm": 1.140625,
      "learning_rate": 0.000343864093278616,
      "loss": 4.9455,
      "mean_token_accuracy": 0.23185234069824218,
      "num_tokens": 81023628.0,
      "step": 36300
    },
    {
      "entropy": 5.599822092056274,
      "epoch": 4.07738529791655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003438258701452017,
      "loss": 4.8262,
      "mean_token_accuracy": 0.2395503729581833,
      "num_tokens": 81034856.0,
      "step": 36305
    },
    {
      "entropy": 5.690893077850342,
      "epoch": 4.077946874824507,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003437876448203449,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2327300563454628,
      "num_tokens": 81046998.0,
      "step": 36310
    },
    {
      "entropy": 5.65068507194519,
      "epoch": 4.078508451732465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003437494173052629,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22844240069389343,
      "num_tokens": 81058292.0,
      "step": 36315
    },
    {
      "entropy": 5.676546716690064,
      "epoch": 4.079070028640422,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003437111876011728,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23444679230451584,
      "num_tokens": 81069907.0,
      "step": 36320
    },
    {
      "entropy": 5.661868095397949,
      "epoch": 4.0796316055483794,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034367295570929176,
      "loss": 4.9856,
      "mean_token_accuracy": 0.220047202706337,
      "num_tokens": 81080700.0,
      "step": 36325
    },
    {
      "entropy": 5.635608148574829,
      "epoch": 4.0801931824563376,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003436347216308373,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23117526471614838,
      "num_tokens": 81092484.0,
      "step": 36330
    },
    {
      "entropy": 5.673785400390625,
      "epoch": 4.080754759364295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003435964853670265,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2333633244037628,
      "num_tokens": 81103992.0,
      "step": 36335
    },
    {
      "entropy": 5.690815830230713,
      "epoch": 4.081316336272253,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034355824691907714,
      "loss": 4.9469,
      "mean_token_accuracy": 0.225727841258049,
      "num_tokens": 81114995.0,
      "step": 36340
    },
    {
      "entropy": 5.561531019210816,
      "epoch": 4.08187791318021,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034352000628820655,
      "loss": 4.79,
      "mean_token_accuracy": 0.23911346942186357,
      "num_tokens": 81125627.0,
      "step": 36345
    },
    {
      "entropy": 5.676815176010132,
      "epoch": 4.082439490088167,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003434817634756324,
      "loss": 5.0362,
      "mean_token_accuracy": 0.22099366933107376,
      "num_tokens": 81136957.0,
      "step": 36350
    },
    {
      "entropy": 5.638783121109009,
      "epoch": 4.083001066996125,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003434435184825724,
      "loss": 4.92,
      "mean_token_accuracy": 0.23185926228761672,
      "num_tokens": 81148840.0,
      "step": 36355
    },
    {
      "entropy": 5.721991300582886,
      "epoch": 4.083562643904083,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034340527131024415,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22764165848493575,
      "num_tokens": 81160385.0,
      "step": 36360
    },
    {
      "entropy": 5.636494302749634,
      "epoch": 4.08412422081204,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003433670219598656,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23194045722484588,
      "num_tokens": 81171334.0,
      "step": 36365
    },
    {
      "entropy": 5.621437835693359,
      "epoch": 4.084685797719998,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003433287704326545,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2309906393289566,
      "num_tokens": 81182547.0,
      "step": 36370
    },
    {
      "entropy": 5.656683397293091,
      "epoch": 4.085247374627955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003432905167298288,
      "loss": 4.889,
      "mean_token_accuracy": 0.22896115332841874,
      "num_tokens": 81193427.0,
      "step": 36375
    },
    {
      "entropy": 5.6744715690612795,
      "epoch": 4.085808951535913,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034325226085260666,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23523192554712297,
      "num_tokens": 81204522.0,
      "step": 36380
    },
    {
      "entropy": 5.690072107315063,
      "epoch": 4.0863705284438705,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000343214002802206,
      "loss": 5.0166,
      "mean_token_accuracy": 0.22908492237329484,
      "num_tokens": 81215209.0,
      "step": 36385
    },
    {
      "entropy": 5.667820739746094,
      "epoch": 4.086932105351828,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034317574257984514,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22620604783296586,
      "num_tokens": 81226083.0,
      "step": 36390
    },
    {
      "entropy": 5.621369504928589,
      "epoch": 4.087493682259786,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000343137480186742,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2334452375769615,
      "num_tokens": 81236531.0,
      "step": 36395
    },
    {
      "entropy": 5.610297584533692,
      "epoch": 4.088055259167743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034309921562411527,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22230205237865447,
      "num_tokens": 81246856.0,
      "step": 36400
    },
    {
      "entropy": 5.697844791412353,
      "epoch": 4.0886168360757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034306094889318295,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21929504126310348,
      "num_tokens": 81259352.0,
      "step": 36405
    },
    {
      "entropy": 5.744544219970703,
      "epoch": 4.089178412983658,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003430226799951637,
      "loss": 5.021,
      "mean_token_accuracy": 0.23009985089302062,
      "num_tokens": 81270020.0,
      "step": 36410
    },
    {
      "entropy": 5.654482412338257,
      "epoch": 4.089739989891616,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034298440893127584,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22715982645750046,
      "num_tokens": 81281625.0,
      "step": 36415
    },
    {
      "entropy": 5.619283771514892,
      "epoch": 4.090301566799573,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034294613570273817,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2395695522427559,
      "num_tokens": 81292340.0,
      "step": 36420
    },
    {
      "entropy": 5.612215232849121,
      "epoch": 4.090863143707531,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034290786031076917,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23287078887224197,
      "num_tokens": 81304494.0,
      "step": 36425
    },
    {
      "entropy": 5.628943157196045,
      "epoch": 4.091424720615488,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003428695827565875,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2357201650738716,
      "num_tokens": 81315365.0,
      "step": 36430
    },
    {
      "entropy": 5.685174989700317,
      "epoch": 4.091986297523446,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034283130304141204,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23249864280223848,
      "num_tokens": 81326970.0,
      "step": 36435
    },
    {
      "entropy": 5.648203325271607,
      "epoch": 4.092547874431403,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003427930211664615,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23477537035942078,
      "num_tokens": 81337117.0,
      "step": 36440
    },
    {
      "entropy": 5.665441274642944,
      "epoch": 4.093109451339361,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000342754737132955,
      "loss": 4.979,
      "mean_token_accuracy": 0.22675028145313264,
      "num_tokens": 81347984.0,
      "step": 36445
    },
    {
      "entropy": 5.694481325149536,
      "epoch": 4.093671028247319,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034271645094211125,
      "loss": 4.9421,
      "mean_token_accuracy": 0.2349219724535942,
      "num_tokens": 81358685.0,
      "step": 36450
    },
    {
      "entropy": 5.682196378707886,
      "epoch": 4.094232605155276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003426781625951495,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2251468226313591,
      "num_tokens": 81368724.0,
      "step": 36455
    },
    {
      "entropy": 5.724859762191772,
      "epoch": 4.094794182063233,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034263987209328886,
      "loss": 5.0349,
      "mean_token_accuracy": 0.23383666723966598,
      "num_tokens": 81380570.0,
      "step": 36460
    },
    {
      "entropy": 5.675860261917114,
      "epoch": 4.095355758971191,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003426015794377483,
      "loss": 4.9606,
      "mean_token_accuracy": 0.2312458872795105,
      "num_tokens": 81391606.0,
      "step": 36465
    },
    {
      "entropy": 5.694782972335815,
      "epoch": 4.0959173358791485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003425632846297474,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22378539741039277,
      "num_tokens": 81402439.0,
      "step": 36470
    },
    {
      "entropy": 5.651801109313965,
      "epoch": 4.096478912787107,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034252498767050525,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22799098640680313,
      "num_tokens": 81413500.0,
      "step": 36475
    },
    {
      "entropy": 5.5711266040802006,
      "epoch": 4.097040489695064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003424866885612413,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23152434974908828,
      "num_tokens": 81424656.0,
      "step": 36480
    },
    {
      "entropy": 5.759660768508911,
      "epoch": 4.097602066603021,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000342448387303175,
      "loss": 5.0147,
      "mean_token_accuracy": 0.2283300668001175,
      "num_tokens": 81436597.0,
      "step": 36485
    },
    {
      "entropy": 5.725331735610962,
      "epoch": 4.098163643510979,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034241008389752594,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22555960863828659,
      "num_tokens": 81448955.0,
      "step": 36490
    },
    {
      "entropy": 5.646128988265991,
      "epoch": 4.098725220418936,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0003423717783455136,
      "loss": 4.9153,
      "mean_token_accuracy": 0.24076973050832748,
      "num_tokens": 81460445.0,
      "step": 36495
    },
    {
      "entropy": 5.683642864227295,
      "epoch": 4.099286797326894,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034233347064835775,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22714272886514664,
      "num_tokens": 81472103.0,
      "step": 36500
    },
    {
      "entropy": 5.622395181655884,
      "epoch": 4.099848374234852,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003422951608072781,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22963182628154755,
      "num_tokens": 81483051.0,
      "step": 36505
    },
    {
      "entropy": 5.5805971145629885,
      "epoch": 4.100409951142809,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003422568488234944,
      "loss": 4.8696,
      "mean_token_accuracy": 0.23012986779212952,
      "num_tokens": 81493755.0,
      "step": 36510
    },
    {
      "entropy": 5.5900365829467775,
      "epoch": 4.100971528050766,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034221853469822654,
      "loss": 4.8366,
      "mean_token_accuracy": 0.2379036158323288,
      "num_tokens": 81505738.0,
      "step": 36515
    },
    {
      "entropy": 5.704603815078736,
      "epoch": 4.101533104958724,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034218021843269444,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22349274456501006,
      "num_tokens": 81516410.0,
      "step": 36520
    },
    {
      "entropy": 5.651790904998779,
      "epoch": 4.1020946818666815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003421419000281182,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22370974123477935,
      "num_tokens": 81527516.0,
      "step": 36525
    },
    {
      "entropy": 5.696272325515747,
      "epoch": 4.10265625877464,
      "grad_norm": 1.125,
      "learning_rate": 0.0003421035794857178,
      "loss": 4.968,
      "mean_token_accuracy": 0.22764220386743544,
      "num_tokens": 81538226.0,
      "step": 36530
    },
    {
      "entropy": 5.706035900115967,
      "epoch": 4.103217835682597,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003420652568067134,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22627484649419785,
      "num_tokens": 81549699.0,
      "step": 36535
    },
    {
      "entropy": 5.754118299484253,
      "epoch": 4.103779412590554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034202693199232516,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21749091595411302,
      "num_tokens": 81560404.0,
      "step": 36540
    },
    {
      "entropy": 5.734842824935913,
      "epoch": 4.104340989498512,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034198860504377343,
      "loss": 4.966,
      "mean_token_accuracy": 0.22779139280319213,
      "num_tokens": 81571758.0,
      "step": 36545
    },
    {
      "entropy": 5.638869094848633,
      "epoch": 4.104902566406469,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034195027596227857,
      "loss": 4.828,
      "mean_token_accuracy": 0.23408588171005248,
      "num_tokens": 81583555.0,
      "step": 36550
    },
    {
      "entropy": 5.588170433044434,
      "epoch": 4.1054641433144266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003419119447490609,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23820296674966812,
      "num_tokens": 81594901.0,
      "step": 36555
    },
    {
      "entropy": 5.667775392532349,
      "epoch": 4.106025720222385,
      "grad_norm": 1.109375,
      "learning_rate": 0.000341873611405341,
      "loss": 4.8882,
      "mean_token_accuracy": 0.23614549040794372,
      "num_tokens": 81606229.0,
      "step": 36560
    },
    {
      "entropy": 5.781834411621094,
      "epoch": 4.106587297130342,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003418352759323394,
      "loss": 4.9868,
      "mean_token_accuracy": 0.23517215251922607,
      "num_tokens": 81616543.0,
      "step": 36565
    },
    {
      "entropy": 5.669215250015259,
      "epoch": 4.1071488740383,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034179693833127667,
      "loss": 4.968,
      "mean_token_accuracy": 0.23125911355018616,
      "num_tokens": 81627517.0,
      "step": 36570
    },
    {
      "entropy": 5.6545249938964846,
      "epoch": 4.107710450946257,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034175859860337347,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22890535145998,
      "num_tokens": 81637235.0,
      "step": 36575
    },
    {
      "entropy": 5.640963077545166,
      "epoch": 4.108272027854214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034172025674985076,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22559230625629426,
      "num_tokens": 81648539.0,
      "step": 36580
    },
    {
      "entropy": 5.623393297195435,
      "epoch": 4.1088336047621725,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034168191277192904,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22939818054437638,
      "num_tokens": 81659981.0,
      "step": 36585
    },
    {
      "entropy": 5.6929789066314695,
      "epoch": 4.10939518167013,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034164356667082933,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23124457001686097,
      "num_tokens": 81669602.0,
      "step": 36590
    },
    {
      "entropy": 5.636481285095215,
      "epoch": 4.109956758578087,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003416052184477727,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22802553623914718,
      "num_tokens": 81679995.0,
      "step": 36595
    },
    {
      "entropy": 5.619531774520874,
      "epoch": 4.110518335486045,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003415668681039801,
      "loss": 4.883,
      "mean_token_accuracy": 0.23314385414123534,
      "num_tokens": 81691295.0,
      "step": 36600
    },
    {
      "entropy": 5.647545003890992,
      "epoch": 4.111079912394002,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034152851564067254,
      "loss": 4.929,
      "mean_token_accuracy": 0.2306102767586708,
      "num_tokens": 81700966.0,
      "step": 36605
    },
    {
      "entropy": 5.714365434646607,
      "epoch": 4.1116414893019595,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003414901610590712,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2233348160982132,
      "num_tokens": 81713469.0,
      "step": 36610
    },
    {
      "entropy": 5.700283527374268,
      "epoch": 4.112203066209918,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003414518043603974,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22983406633138656,
      "num_tokens": 81724748.0,
      "step": 36615
    },
    {
      "entropy": 5.7114380359649655,
      "epoch": 4.112764643117875,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034141344554587243,
      "loss": 5.0291,
      "mean_token_accuracy": 0.22162622213363647,
      "num_tokens": 81735694.0,
      "step": 36620
    },
    {
      "entropy": 5.745974397659301,
      "epoch": 4.113326220025833,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003413750846167175,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22529985755681992,
      "num_tokens": 81748251.0,
      "step": 36625
    },
    {
      "entropy": 5.609404039382935,
      "epoch": 4.11388779693379,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003413367215741542,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23372037708759308,
      "num_tokens": 81760750.0,
      "step": 36630
    },
    {
      "entropy": 5.68262734413147,
      "epoch": 4.114449373841747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034129835641940394,
      "loss": 5.0141,
      "mean_token_accuracy": 0.22635464519262313,
      "num_tokens": 81772100.0,
      "step": 36635
    },
    {
      "entropy": 5.706469440460205,
      "epoch": 4.1150109507497055,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034125998915368827,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22416517287492752,
      "num_tokens": 81783795.0,
      "step": 36640
    },
    {
      "entropy": 5.65731725692749,
      "epoch": 4.115572527657663,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034122161977822885,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2297600120306015,
      "num_tokens": 81794713.0,
      "step": 36645
    },
    {
      "entropy": 5.69622015953064,
      "epoch": 4.11613410456562,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003411832482942474,
      "loss": 4.9392,
      "mean_token_accuracy": 0.2294745311141014,
      "num_tokens": 81806493.0,
      "step": 36650
    },
    {
      "entropy": 5.638850498199463,
      "epoch": 4.116695681473578,
      "grad_norm": 1.25,
      "learning_rate": 0.00034114487470296575,
      "loss": 4.8105,
      "mean_token_accuracy": 0.23626838326454164,
      "num_tokens": 81818232.0,
      "step": 36655
    },
    {
      "entropy": 5.573049926757813,
      "epoch": 4.117257258381535,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034110649900560554,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2311757445335388,
      "num_tokens": 81829439.0,
      "step": 36660
    },
    {
      "entropy": 5.652594089508057,
      "epoch": 4.117818835289493,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003410681212033887,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22912263870239258,
      "num_tokens": 81839585.0,
      "step": 36665
    },
    {
      "entropy": 5.66446852684021,
      "epoch": 4.1183804121974505,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034102974129753726,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2289142295718193,
      "num_tokens": 81850648.0,
      "step": 36670
    },
    {
      "entropy": 5.640189123153687,
      "epoch": 4.118941989105408,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003409913592892733,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22708747088909148,
      "num_tokens": 81862597.0,
      "step": 36675
    },
    {
      "entropy": 5.712638521194458,
      "epoch": 4.119503566013366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003409529751798189,
      "loss": 4.9671,
      "mean_token_accuracy": 0.23237769305706024,
      "num_tokens": 81874540.0,
      "step": 36680
    },
    {
      "entropy": 5.631936120986938,
      "epoch": 4.120065142921323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003409145889703961,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23807260245084763,
      "num_tokens": 81884205.0,
      "step": 36685
    },
    {
      "entropy": 5.578629875183106,
      "epoch": 4.12062671982928,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034087620066222733,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23151097148656846,
      "num_tokens": 81895552.0,
      "step": 36690
    },
    {
      "entropy": 5.739275741577148,
      "epoch": 4.121188296737238,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003408378102565347,
      "loss": 5.0458,
      "mean_token_accuracy": 0.220717716217041,
      "num_tokens": 81908169.0,
      "step": 36695
    },
    {
      "entropy": 5.656075525283813,
      "epoch": 4.121749873645196,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003407994177545407,
      "loss": 4.9304,
      "mean_token_accuracy": 0.23143512904644012,
      "num_tokens": 81919465.0,
      "step": 36700
    },
    {
      "entropy": 5.665368223190308,
      "epoch": 4.122311450553153,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003407610231574677,
      "loss": 5.0126,
      "mean_token_accuracy": 0.2222020909190178,
      "num_tokens": 81930558.0,
      "step": 36705
    },
    {
      "entropy": 5.697026491165161,
      "epoch": 4.122873027461111,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003407226264665382,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22990752607584,
      "num_tokens": 81941325.0,
      "step": 36710
    },
    {
      "entropy": 5.636926126480103,
      "epoch": 4.123434604369068,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034068422768297486,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2318975418806076,
      "num_tokens": 81953225.0,
      "step": 36715
    },
    {
      "entropy": 5.6604651927948,
      "epoch": 4.123996181277026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003406458268080001,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22819886803627015,
      "num_tokens": 81964365.0,
      "step": 36720
    },
    {
      "entropy": 5.645818758010864,
      "epoch": 4.1245577581849835,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034060742384283695,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2265789330005646,
      "num_tokens": 81974648.0,
      "step": 36725
    },
    {
      "entropy": 5.644466733932495,
      "epoch": 4.125119335092941,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003405690187887078,
      "loss": 4.8806,
      "mean_token_accuracy": 0.237032550573349,
      "num_tokens": 81985502.0,
      "step": 36730
    },
    {
      "entropy": 5.740030097961426,
      "epoch": 4.125680912000899,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034053061164683574,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22573346197605132,
      "num_tokens": 81998279.0,
      "step": 36735
    },
    {
      "entropy": 5.626086568832397,
      "epoch": 4.126242488908856,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034049220241844357,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2312358498573303,
      "num_tokens": 82009911.0,
      "step": 36740
    },
    {
      "entropy": 5.568450450897217,
      "epoch": 4.126804065816813,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003404537911047543,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2389209046959877,
      "num_tokens": 82020219.0,
      "step": 36745
    },
    {
      "entropy": 5.5839752674102785,
      "epoch": 4.127365642724771,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034041537770699095,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2367979884147644,
      "num_tokens": 82030997.0,
      "step": 36750
    },
    {
      "entropy": 5.692557144165039,
      "epoch": 4.127927219632729,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003403769622263765,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23352580964565278,
      "num_tokens": 82043384.0,
      "step": 36755
    },
    {
      "entropy": 5.654639196395874,
      "epoch": 4.128488796540687,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034033854466413425,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23407539427280427,
      "num_tokens": 82053346.0,
      "step": 36760
    },
    {
      "entropy": 5.615247106552124,
      "epoch": 4.129050373448644,
      "grad_norm": 1.125,
      "learning_rate": 0.0003403001250214875,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23757008761167525,
      "num_tokens": 82063628.0,
      "step": 36765
    },
    {
      "entropy": 5.62428617477417,
      "epoch": 4.129611950356601,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003402617032996593,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23960004597902299,
      "num_tokens": 82074097.0,
      "step": 36770
    },
    {
      "entropy": 5.637108421325683,
      "epoch": 4.130173527264559,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003402232794998733,
      "loss": 4.9854,
      "mean_token_accuracy": 0.22278853356838227,
      "num_tokens": 82085159.0,
      "step": 36775
    },
    {
      "entropy": 5.6797408103942875,
      "epoch": 4.130735104172516,
      "grad_norm": 1.25,
      "learning_rate": 0.00034018485362335266,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22649972885847092,
      "num_tokens": 82096299.0,
      "step": 36780
    },
    {
      "entropy": 5.61149959564209,
      "epoch": 4.131296681080474,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003401464256713211,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22796035706996917,
      "num_tokens": 82107111.0,
      "step": 36785
    },
    {
      "entropy": 5.624709749221802,
      "epoch": 4.131858257988432,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034010799564500197,
      "loss": 4.9249,
      "mean_token_accuracy": 0.23236822187900544,
      "num_tokens": 82117889.0,
      "step": 36790
    },
    {
      "entropy": 5.623511505126953,
      "epoch": 4.132419834896389,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003400695635456191,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23098205029964447,
      "num_tokens": 82128380.0,
      "step": 36795
    },
    {
      "entropy": 5.618681383132935,
      "epoch": 4.132981411804346,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003400311293743961,
      "loss": 4.8204,
      "mean_token_accuracy": 0.24192424416542052,
      "num_tokens": 82138368.0,
      "step": 36800
    },
    {
      "entropy": 5.696034479141235,
      "epoch": 4.133542988712304,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003399926931325567,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2290026515722275,
      "num_tokens": 82148449.0,
      "step": 36805
    },
    {
      "entropy": 5.546083974838257,
      "epoch": 4.1341045656202615,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033995425482132464,
      "loss": 4.8398,
      "mean_token_accuracy": 0.23359708487987518,
      "num_tokens": 82160937.0,
      "step": 36810
    },
    {
      "entropy": 5.6258679866790775,
      "epoch": 4.13466614252822,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033991581444192405,
      "loss": 4.848,
      "mean_token_accuracy": 0.23437905460596084,
      "num_tokens": 82171546.0,
      "step": 36815
    },
    {
      "entropy": 5.6689207553863525,
      "epoch": 4.135227719436177,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003398773719955787,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23728092312812804,
      "num_tokens": 82182613.0,
      "step": 36820
    },
    {
      "entropy": 5.5405356884002686,
      "epoch": 4.135789296344134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003398389274835126,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23718023300170898,
      "num_tokens": 82193757.0,
      "step": 36825
    },
    {
      "entropy": 5.619712591171265,
      "epoch": 4.136350873252092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033980048090694997,
      "loss": 4.95,
      "mean_token_accuracy": 0.22148610651493073,
      "num_tokens": 82205601.0,
      "step": 36830
    },
    {
      "entropy": 5.7000432968139645,
      "epoch": 4.136912450160049,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003397620322671149,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22999887764453888,
      "num_tokens": 82217022.0,
      "step": 36835
    },
    {
      "entropy": 5.643433904647827,
      "epoch": 4.137474027068007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033972358156523163,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23503183722496032,
      "num_tokens": 82227682.0,
      "step": 36840
    },
    {
      "entropy": 5.591345596313476,
      "epoch": 4.138035603975965,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003396851288025243,
      "loss": 4.8634,
      "mean_token_accuracy": 0.23105525970458984,
      "num_tokens": 82238754.0,
      "step": 36845
    },
    {
      "entropy": 5.723630905151367,
      "epoch": 4.138597180883922,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003396466739802175,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22757933884859086,
      "num_tokens": 82249536.0,
      "step": 36850
    },
    {
      "entropy": 5.656334257125854,
      "epoch": 4.13915875779188,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003396082170995355,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22855059653520585,
      "num_tokens": 82260374.0,
      "step": 36855
    },
    {
      "entropy": 5.66479811668396,
      "epoch": 4.139720334699837,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003395697581617028,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23054119646549226,
      "num_tokens": 82271801.0,
      "step": 36860
    },
    {
      "entropy": 5.706098747253418,
      "epoch": 4.1402819116077945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033953129716794396,
      "loss": 5.0474,
      "mean_token_accuracy": 0.2261705294251442,
      "num_tokens": 82283149.0,
      "step": 36865
    },
    {
      "entropy": 5.6687578678131105,
      "epoch": 4.140843488515753,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003394928341194836,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2296896681189537,
      "num_tokens": 82293003.0,
      "step": 36870
    },
    {
      "entropy": 5.679989290237427,
      "epoch": 4.14140506542371,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003394543690175464,
      "loss": 4.9302,
      "mean_token_accuracy": 0.23326730132102966,
      "num_tokens": 82303734.0,
      "step": 36875
    },
    {
      "entropy": 5.601885795593262,
      "epoch": 4.141966642331667,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033941590186335703,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23045816719532014,
      "num_tokens": 82314569.0,
      "step": 36880
    },
    {
      "entropy": 5.660868978500366,
      "epoch": 4.142528219239625,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003393774326581404,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22924479991197586,
      "num_tokens": 82325174.0,
      "step": 36885
    },
    {
      "entropy": 5.6992409229278564,
      "epoch": 4.143089796147582,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033933896140312135,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22879980653524398,
      "num_tokens": 82337448.0,
      "step": 36890
    },
    {
      "entropy": 5.717817878723144,
      "epoch": 4.1436513730555395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033930048809952484,
      "loss": 4.989,
      "mean_token_accuracy": 0.22893106788396836,
      "num_tokens": 82348705.0,
      "step": 36895
    },
    {
      "entropy": 5.735385274887085,
      "epoch": 4.144212949963498,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033926201274857575,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22934221476316452,
      "num_tokens": 82360439.0,
      "step": 36900
    },
    {
      "entropy": 5.643145370483398,
      "epoch": 4.144774526871455,
      "grad_norm": 1.25,
      "learning_rate": 0.00033922353535149934,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2311052903532982,
      "num_tokens": 82371346.0,
      "step": 36905
    },
    {
      "entropy": 5.587526893615722,
      "epoch": 4.145336103779413,
      "grad_norm": 1.125,
      "learning_rate": 0.0003391850559095207,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2362853169441223,
      "num_tokens": 82382221.0,
      "step": 36910
    },
    {
      "entropy": 5.674576473236084,
      "epoch": 4.14589768068737,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003391465744238649,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2207997500896454,
      "num_tokens": 82392290.0,
      "step": 36915
    },
    {
      "entropy": 5.730078792572021,
      "epoch": 4.146459257595327,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003391080908957573,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22116219252347946,
      "num_tokens": 82403513.0,
      "step": 36920
    },
    {
      "entropy": 5.644337368011475,
      "epoch": 4.1470208345032855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033906960532642327,
      "loss": 4.8459,
      "mean_token_accuracy": 0.24286959171295167,
      "num_tokens": 82414537.0,
      "step": 36925
    },
    {
      "entropy": 5.615932607650757,
      "epoch": 4.147582411411243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003390311177170881,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2357925519347191,
      "num_tokens": 82425330.0,
      "step": 36930
    },
    {
      "entropy": 5.5215953350067135,
      "epoch": 4.1481439883192,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003389926280689773,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23420846611261367,
      "num_tokens": 82435703.0,
      "step": 36935
    },
    {
      "entropy": 5.673090600967408,
      "epoch": 4.148705565227158,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033895413638331646,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2284681722521782,
      "num_tokens": 82447302.0,
      "step": 36940
    },
    {
      "entropy": 5.682822370529175,
      "epoch": 4.149267142135115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003389156426613311,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22003765851259233,
      "num_tokens": 82458379.0,
      "step": 36945
    },
    {
      "entropy": 5.6642899990081785,
      "epoch": 4.149828719043073,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033887714690424687,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22947726547718048,
      "num_tokens": 82469728.0,
      "step": 36950
    },
    {
      "entropy": 5.677286100387573,
      "epoch": 4.150390295951031,
      "grad_norm": 1.125,
      "learning_rate": 0.00033883864911328954,
      "loss": 5.019,
      "mean_token_accuracy": 0.2224253997206688,
      "num_tokens": 82482216.0,
      "step": 36955
    },
    {
      "entropy": 5.7236878871917725,
      "epoch": 4.150951872858988,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033880014928968483,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22826855331659318,
      "num_tokens": 82494511.0,
      "step": 36960
    },
    {
      "entropy": 5.68039984703064,
      "epoch": 4.151513449766946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033876164743465865,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2324000746011734,
      "num_tokens": 82505982.0,
      "step": 36965
    },
    {
      "entropy": 5.633678007125854,
      "epoch": 4.152075026674903,
      "grad_norm": 1.25,
      "learning_rate": 0.0003387231435494368,
      "loss": 4.8734,
      "mean_token_accuracy": 0.23125271499156952,
      "num_tokens": 82516682.0,
      "step": 36970
    },
    {
      "entropy": 5.637824535369873,
      "epoch": 4.15263660358286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033868463763524545,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22435077279806137,
      "num_tokens": 82527742.0,
      "step": 36975
    },
    {
      "entropy": 5.62707896232605,
      "epoch": 4.1531981804908185,
      "grad_norm": 1.125,
      "learning_rate": 0.00033864612969331046,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23091549277305604,
      "num_tokens": 82538130.0,
      "step": 36980
    },
    {
      "entropy": 5.635627412796021,
      "epoch": 4.153759757398776,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003386076197248581,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2321009859442711,
      "num_tokens": 82549695.0,
      "step": 36985
    },
    {
      "entropy": 5.699470520019531,
      "epoch": 4.154321334306733,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003385691077311144,
      "loss": 4.9638,
      "mean_token_accuracy": 0.2246597021818161,
      "num_tokens": 82560778.0,
      "step": 36990
    },
    {
      "entropy": 5.609732675552368,
      "epoch": 4.154882911214691,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033853059371330565,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22865843623876572,
      "num_tokens": 82572021.0,
      "step": 36995
    },
    {
      "entropy": 5.732578754425049,
      "epoch": 4.155444488122648,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003384920776726582,
      "loss": 4.9537,
      "mean_token_accuracy": 0.23213366121053697,
      "num_tokens": 82583173.0,
      "step": 37000
    },
    {
      "entropy": 5.642125272750855,
      "epoch": 4.156006065030606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033845355961039823,
      "loss": 4.8757,
      "mean_token_accuracy": 0.232870851457119,
      "num_tokens": 82594279.0,
      "step": 37005
    },
    {
      "entropy": 5.691335248947143,
      "epoch": 4.1565676419385635,
      "grad_norm": 1.25,
      "learning_rate": 0.00033841503952775255,
      "loss": 5.0296,
      "mean_token_accuracy": 0.22540131956338882,
      "num_tokens": 82605038.0,
      "step": 37010
    },
    {
      "entropy": 5.718512058258057,
      "epoch": 4.157129218846521,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033837651742594734,
      "loss": 5.0413,
      "mean_token_accuracy": 0.2277311861515045,
      "num_tokens": 82616627.0,
      "step": 37015
    },
    {
      "entropy": 5.637061023712159,
      "epoch": 4.157690795754479,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033833799330620917,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22955915033817292,
      "num_tokens": 82628057.0,
      "step": 37020
    },
    {
      "entropy": 5.642458820343018,
      "epoch": 4.158252372662436,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003382994671697647,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22658882886171341,
      "num_tokens": 82639890.0,
      "step": 37025
    },
    {
      "entropy": 5.5861937522888185,
      "epoch": 4.158813949570393,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033826093901784065,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22816205769777298,
      "num_tokens": 82651770.0,
      "step": 37030
    },
    {
      "entropy": 5.650781202316284,
      "epoch": 4.159375526478351,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033822240885166384,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2315462812781334,
      "num_tokens": 82662159.0,
      "step": 37035
    },
    {
      "entropy": 5.688238286972046,
      "epoch": 4.159937103386309,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033818387667246095,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23182308673858643,
      "num_tokens": 82674348.0,
      "step": 37040
    },
    {
      "entropy": 5.665853261947632,
      "epoch": 4.160498680294267,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000338145342481459,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23697919845581056,
      "num_tokens": 82685674.0,
      "step": 37045
    },
    {
      "entropy": 5.741336441040039,
      "epoch": 4.161060257202224,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003381068062798848,
      "loss": 5.0448,
      "mean_token_accuracy": 0.2156429409980774,
      "num_tokens": 82696992.0,
      "step": 37050
    },
    {
      "entropy": 5.78253173828125,
      "epoch": 4.161621834110181,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003380682680689654,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2218104511499405,
      "num_tokens": 82708741.0,
      "step": 37055
    },
    {
      "entropy": 5.65472445487976,
      "epoch": 4.162183411018139,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033802972784992784,
      "loss": 4.9339,
      "mean_token_accuracy": 0.23012663871049882,
      "num_tokens": 82720687.0,
      "step": 37060
    },
    {
      "entropy": 5.60554986000061,
      "epoch": 4.1627449879260965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003379911856239994,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2267501100897789,
      "num_tokens": 82731863.0,
      "step": 37065
    },
    {
      "entropy": 5.630327796936035,
      "epoch": 4.163306564834054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003379526413924071,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23045322597026824,
      "num_tokens": 82742931.0,
      "step": 37070
    },
    {
      "entropy": 5.6748441696167,
      "epoch": 4.163868141742012,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003379140951563784,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22907516360282898,
      "num_tokens": 82754244.0,
      "step": 37075
    },
    {
      "entropy": 5.625123596191406,
      "epoch": 4.164429718649969,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033787554691714033,
      "loss": 4.895,
      "mean_token_accuracy": 0.22885477244853974,
      "num_tokens": 82764929.0,
      "step": 37080
    },
    {
      "entropy": 5.612340879440308,
      "epoch": 4.164991295557926,
      "grad_norm": 1.125,
      "learning_rate": 0.00033783699667592066,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2303437754511833,
      "num_tokens": 82775767.0,
      "step": 37085
    },
    {
      "entropy": 5.6064530372619625,
      "epoch": 4.165552872465884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033779844443394656,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2300922229886055,
      "num_tokens": 82786902.0,
      "step": 37090
    },
    {
      "entropy": 5.755582237243653,
      "epoch": 4.166114449373842,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033775989019244556,
      "loss": 4.993,
      "mean_token_accuracy": 0.22435547709465026,
      "num_tokens": 82797677.0,
      "step": 37095
    },
    {
      "entropy": 5.743398904800415,
      "epoch": 4.1666760262818,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033772133395264535,
      "loss": 4.917,
      "mean_token_accuracy": 0.22503797113895416,
      "num_tokens": 82807246.0,
      "step": 37100
    },
    {
      "entropy": 5.684249019622802,
      "epoch": 4.167237603189757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033768277571577364,
      "loss": 4.9399,
      "mean_token_accuracy": 0.23185885101556777,
      "num_tokens": 82818983.0,
      "step": 37105
    },
    {
      "entropy": 5.592262506484985,
      "epoch": 4.167799180097714,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033764421548305795,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23462702929973603,
      "num_tokens": 82829340.0,
      "step": 37110
    },
    {
      "entropy": 5.639827013015747,
      "epoch": 4.168360757005672,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033760565325572614,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22168837785720824,
      "num_tokens": 82840741.0,
      "step": 37115
    },
    {
      "entropy": 5.7042662620544435,
      "epoch": 4.168922333913629,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003375670890350061,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22670411318540573,
      "num_tokens": 82851284.0,
      "step": 37120
    },
    {
      "entropy": 5.51874623298645,
      "epoch": 4.169483910821587,
      "grad_norm": 1.125,
      "learning_rate": 0.00033752852282212557,
      "loss": 4.7417,
      "mean_token_accuracy": 0.2378714755177498,
      "num_tokens": 82861856.0,
      "step": 37125
    },
    {
      "entropy": 5.696953535079956,
      "epoch": 4.170045487729545,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003374899546183128,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22107143998146056,
      "num_tokens": 82873354.0,
      "step": 37130
    },
    {
      "entropy": 5.6591400623321535,
      "epoch": 4.170607064637502,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033745138442479544,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22705059498548508,
      "num_tokens": 82883891.0,
      "step": 37135
    },
    {
      "entropy": 5.655903100967407,
      "epoch": 4.17116864154546,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003374128122428019,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23759907931089402,
      "num_tokens": 82893685.0,
      "step": 37140
    },
    {
      "entropy": 5.605355167388916,
      "epoch": 4.171730218453417,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003373742380735602,
      "loss": 4.92,
      "mean_token_accuracy": 0.23999615758657455,
      "num_tokens": 82904947.0,
      "step": 37145
    },
    {
      "entropy": 5.645442390441895,
      "epoch": 4.1722917953613745,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003373356619182985,
      "loss": 5.0213,
      "mean_token_accuracy": 0.22264742106199265,
      "num_tokens": 82915368.0,
      "step": 37150
    },
    {
      "entropy": 5.675550079345703,
      "epoch": 4.172853372269333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003372970837782453,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22644060105085373,
      "num_tokens": 82926594.0,
      "step": 37155
    },
    {
      "entropy": 5.65231032371521,
      "epoch": 4.17341494917729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003372585036546287,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23624562621116638,
      "num_tokens": 82937699.0,
      "step": 37160
    },
    {
      "entropy": 5.680290651321411,
      "epoch": 4.173976526085247,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003372199215486772,
      "loss": 4.9935,
      "mean_token_accuracy": 0.226241897046566,
      "num_tokens": 82948822.0,
      "step": 37165
    },
    {
      "entropy": 5.667576599121094,
      "epoch": 4.174538102993205,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003371813374616192,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22585275918245315,
      "num_tokens": 82959597.0,
      "step": 37170
    },
    {
      "entropy": 5.678720426559448,
      "epoch": 4.175099679901162,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033714275139468336,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2299666330218315,
      "num_tokens": 82970740.0,
      "step": 37175
    },
    {
      "entropy": 5.699298334121704,
      "epoch": 4.17566125680912,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003371041633490982,
      "loss": 5.0134,
      "mean_token_accuracy": 0.2237880378961563,
      "num_tokens": 82982923.0,
      "step": 37180
    },
    {
      "entropy": 5.5747801780700685,
      "epoch": 4.176222833717078,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003370655733260924,
      "loss": 4.8716,
      "mean_token_accuracy": 0.22874038964509963,
      "num_tokens": 82993271.0,
      "step": 37185
    },
    {
      "entropy": 5.677834415435791,
      "epoch": 4.176784410625035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033702698132689467,
      "loss": 4.9071,
      "mean_token_accuracy": 0.23533493727445604,
      "num_tokens": 83004688.0,
      "step": 37190
    },
    {
      "entropy": 5.645793056488037,
      "epoch": 4.177345987532993,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003369883873527338,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22316526174545287,
      "num_tokens": 83016310.0,
      "step": 37195
    },
    {
      "entropy": 5.7350828647613525,
      "epoch": 4.17790756444095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033694979140483865,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2206283241510391,
      "num_tokens": 83027327.0,
      "step": 37200
    },
    {
      "entropy": 5.662732315063477,
      "epoch": 4.1784691413489075,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003369111934844381,
      "loss": 4.9465,
      "mean_token_accuracy": 0.22720438092947007,
      "num_tokens": 83038493.0,
      "step": 37205
    },
    {
      "entropy": 5.6317812442779545,
      "epoch": 4.179030718256866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033687259359276115,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2299230620265007,
      "num_tokens": 83050082.0,
      "step": 37210
    },
    {
      "entropy": 5.638767337799072,
      "epoch": 4.179592295164823,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003368339917310368,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23284061551094054,
      "num_tokens": 83061234.0,
      "step": 37215
    },
    {
      "entropy": 5.6451211929321286,
      "epoch": 4.18015387207278,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003367953879004942,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22757349908351898,
      "num_tokens": 83071454.0,
      "step": 37220
    },
    {
      "entropy": 5.669976091384887,
      "epoch": 4.180715448980738,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003367567821023625,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2276309013366699,
      "num_tokens": 83083204.0,
      "step": 37225
    },
    {
      "entropy": 5.629638242721557,
      "epoch": 4.181277025888695,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003367181743378708,
      "loss": 4.8903,
      "mean_token_accuracy": 0.23186191767454148,
      "num_tokens": 83094654.0,
      "step": 37230
    },
    {
      "entropy": 5.709295797348022,
      "epoch": 4.181838602796653,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033667956460824866,
      "loss": 5.0001,
      "mean_token_accuracy": 0.22227130383253096,
      "num_tokens": 83105539.0,
      "step": 37235
    },
    {
      "entropy": 5.6730352401733395,
      "epoch": 4.182400179704611,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003366409529147251,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23100529611110687,
      "num_tokens": 83115741.0,
      "step": 37240
    },
    {
      "entropy": 5.671301984786988,
      "epoch": 4.182961756612568,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033660233925852977,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23451288938522338,
      "num_tokens": 83126739.0,
      "step": 37245
    },
    {
      "entropy": 5.6025653839111325,
      "epoch": 4.183523333520526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033656372364089206,
      "loss": 4.82,
      "mean_token_accuracy": 0.24012753516435623,
      "num_tokens": 83138821.0,
      "step": 37250
    },
    {
      "entropy": 5.673244524002075,
      "epoch": 4.184084910428483,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033652510606304155,
      "loss": 4.9989,
      "mean_token_accuracy": 0.22291918843984604,
      "num_tokens": 83149757.0,
      "step": 37255
    },
    {
      "entropy": 5.640285921096802,
      "epoch": 4.18464648733644,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033648648652620766,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22529419362545014,
      "num_tokens": 83161627.0,
      "step": 37260
    },
    {
      "entropy": 5.630374908447266,
      "epoch": 4.1852080642443985,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033644786503162036,
      "loss": 4.9024,
      "mean_token_accuracy": 0.23174309581518174,
      "num_tokens": 83172903.0,
      "step": 37265
    },
    {
      "entropy": 5.564214468002319,
      "epoch": 4.185769641152356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003364092415805091,
      "loss": 4.816,
      "mean_token_accuracy": 0.23570147901773453,
      "num_tokens": 83184142.0,
      "step": 37270
    },
    {
      "entropy": 5.703928709030151,
      "epoch": 4.186331218060313,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033637061617410377,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2248835563659668,
      "num_tokens": 83195206.0,
      "step": 37275
    },
    {
      "entropy": 5.683653211593628,
      "epoch": 4.186892794968271,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003363319888136343,
      "loss": 4.9155,
      "mean_token_accuracy": 0.23537100702524186,
      "num_tokens": 83207579.0,
      "step": 37280
    },
    {
      "entropy": 5.705076360702515,
      "epoch": 4.187454371876228,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003362933595003305,
      "loss": 5.0032,
      "mean_token_accuracy": 0.22310433387756348,
      "num_tokens": 83219964.0,
      "step": 37285
    },
    {
      "entropy": 5.617296171188355,
      "epoch": 4.188015948784186,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003362547282354223,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22959368526935578,
      "num_tokens": 83231241.0,
      "step": 37290
    },
    {
      "entropy": 5.618032836914063,
      "epoch": 4.188577525692144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003362160950201398,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23827614933252333,
      "num_tokens": 83241784.0,
      "step": 37295
    },
    {
      "entropy": 5.613654232025146,
      "epoch": 4.189139102600101,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003361774598557132,
      "loss": 4.8567,
      "mean_token_accuracy": 0.22952408194541932,
      "num_tokens": 83252815.0,
      "step": 37300
    },
    {
      "entropy": 5.63432092666626,
      "epoch": 4.189700679508059,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033613882274337244,
      "loss": 4.9343,
      "mean_token_accuracy": 0.23475790321826934,
      "num_tokens": 83263938.0,
      "step": 37305
    },
    {
      "entropy": 5.705732440948486,
      "epoch": 4.190262256416016,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003361001836843479,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22093250155448912,
      "num_tokens": 83274746.0,
      "step": 37310
    },
    {
      "entropy": 5.749871253967285,
      "epoch": 4.190823833323973,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003360615426798697,
      "loss": 5.0874,
      "mean_token_accuracy": 0.21723815500736238,
      "num_tokens": 83285707.0,
      "step": 37315
    },
    {
      "entropy": 5.751135969161988,
      "epoch": 4.1913854102319315,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003360228997311685,
      "loss": 4.9638,
      "mean_token_accuracy": 0.23377616554498673,
      "num_tokens": 83297175.0,
      "step": 37320
    },
    {
      "entropy": 5.654204034805298,
      "epoch": 4.191946987139889,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003359842548394744,
      "loss": 4.8331,
      "mean_token_accuracy": 0.24068362563848494,
      "num_tokens": 83307731.0,
      "step": 37325
    },
    {
      "entropy": 5.6682662010192875,
      "epoch": 4.192508564047847,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003359456080060179,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23040866404771804,
      "num_tokens": 83318698.0,
      "step": 37330
    },
    {
      "entropy": 5.579011392593384,
      "epoch": 4.193070140955804,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003359069592320297,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23311724960803987,
      "num_tokens": 83330085.0,
      "step": 37335
    },
    {
      "entropy": 5.71684308052063,
      "epoch": 4.193631717863761,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003358683085187403,
      "loss": 4.975,
      "mean_token_accuracy": 0.2247040629386902,
      "num_tokens": 83340606.0,
      "step": 37340
    },
    {
      "entropy": 5.612220764160156,
      "epoch": 4.194193294771719,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033582965586738037,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22983310967683793,
      "num_tokens": 83352177.0,
      "step": 37345
    },
    {
      "entropy": 5.698770570755005,
      "epoch": 4.1947548716796765,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003357910012791804,
      "loss": 5.0354,
      "mean_token_accuracy": 0.22003313452005385,
      "num_tokens": 83362474.0,
      "step": 37350
    },
    {
      "entropy": 5.694744300842285,
      "epoch": 4.195316448587634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003357523447553716,
      "loss": 4.9832,
      "mean_token_accuracy": 0.2278537318110466,
      "num_tokens": 83373187.0,
      "step": 37355
    },
    {
      "entropy": 5.645866632461548,
      "epoch": 4.195878025495592,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033571368629718457,
      "loss": 4.874,
      "mean_token_accuracy": 0.23204565346240996,
      "num_tokens": 83385110.0,
      "step": 37360
    },
    {
      "entropy": 5.688356256484985,
      "epoch": 4.196439602403549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033567502590585007,
      "loss": 4.9632,
      "mean_token_accuracy": 0.23347969502210617,
      "num_tokens": 83396692.0,
      "step": 37365
    },
    {
      "entropy": 5.684838342666626,
      "epoch": 4.197001179311506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003356363635825994,
      "loss": 4.8784,
      "mean_token_accuracy": 0.23175875693559647,
      "num_tokens": 83408433.0,
      "step": 37370
    },
    {
      "entropy": 5.7116625785827635,
      "epoch": 4.197562756219464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003355976993286633,
      "loss": 4.9653,
      "mean_token_accuracy": 0.23377851098775865,
      "num_tokens": 83419169.0,
      "step": 37375
    },
    {
      "entropy": 5.671844482421875,
      "epoch": 4.198124333127422,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003355590331452729,
      "loss": 4.9214,
      "mean_token_accuracy": 0.23918306678533555,
      "num_tokens": 83430247.0,
      "step": 37380
    },
    {
      "entropy": 5.636804866790771,
      "epoch": 4.19868591003538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033552036503365945,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23840180188417434,
      "num_tokens": 83442294.0,
      "step": 37385
    },
    {
      "entropy": 5.631360197067261,
      "epoch": 4.199247486943337,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003354816949950541,
      "loss": 4.888,
      "mean_token_accuracy": 0.23832783550024034,
      "num_tokens": 83453388.0,
      "step": 37390
    },
    {
      "entropy": 5.568445825576783,
      "epoch": 4.199809063851294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033544302303068814,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2408038333058357,
      "num_tokens": 83465151.0,
      "step": 37395
    },
    {
      "entropy": 5.665221786499023,
      "epoch": 4.200370640759252,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003354043491417929,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2297303631901741,
      "num_tokens": 83474731.0,
      "step": 37400
    },
    {
      "entropy": 5.630221176147461,
      "epoch": 4.2009322176672095,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033536567332959966,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2316799595952034,
      "num_tokens": 83486456.0,
      "step": 37405
    },
    {
      "entropy": 5.663875341415405,
      "epoch": 4.201493794575167,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033532699559534,
      "loss": 4.898,
      "mean_token_accuracy": 0.23154804110527039,
      "num_tokens": 83496503.0,
      "step": 37410
    },
    {
      "entropy": 5.7489063262939455,
      "epoch": 4.202055371483125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033528831594024553,
      "loss": 4.998,
      "mean_token_accuracy": 0.22965807616710662,
      "num_tokens": 83508198.0,
      "step": 37415
    },
    {
      "entropy": 5.682371950149536,
      "epoch": 4.202616948391082,
      "grad_norm": 1.125,
      "learning_rate": 0.00033524963436554753,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22976473569869996,
      "num_tokens": 83520255.0,
      "step": 37420
    },
    {
      "entropy": 5.671053552627564,
      "epoch": 4.20317852529904,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033521095087247794,
      "loss": 4.9831,
      "mean_token_accuracy": 0.22402899861335754,
      "num_tokens": 83532083.0,
      "step": 37425
    },
    {
      "entropy": 5.721648550033569,
      "epoch": 4.203740102206997,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003351722654622683,
      "loss": 5.0042,
      "mean_token_accuracy": 0.22800311893224717,
      "num_tokens": 83542891.0,
      "step": 37430
    },
    {
      "entropy": 5.659850168228149,
      "epoch": 4.204301679114955,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003351335781361503,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23512688726186753,
      "num_tokens": 83554238.0,
      "step": 37435
    },
    {
      "entropy": 5.676511764526367,
      "epoch": 4.204863256022913,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033509488889535595,
      "loss": 4.9798,
      "mean_token_accuracy": 0.22144967466592788,
      "num_tokens": 83565681.0,
      "step": 37440
    },
    {
      "entropy": 5.565623569488525,
      "epoch": 4.20542483293087,
      "grad_norm": 1.125,
      "learning_rate": 0.00033505619774111696,
      "loss": 4.7934,
      "mean_token_accuracy": 0.2408516749739647,
      "num_tokens": 83576844.0,
      "step": 37445
    },
    {
      "entropy": 5.642740917205811,
      "epoch": 4.205986409838827,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033501750467466535,
      "loss": 4.912,
      "mean_token_accuracy": 0.23806908577680588,
      "num_tokens": 83589501.0,
      "step": 37450
    },
    {
      "entropy": 5.69083948135376,
      "epoch": 4.206547986746785,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003349788096972331,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2370864138007164,
      "num_tokens": 83600305.0,
      "step": 37455
    },
    {
      "entropy": 5.703097152709961,
      "epoch": 4.207109563654742,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003349401128100523,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22106315493583678,
      "num_tokens": 83611908.0,
      "step": 37460
    },
    {
      "entropy": 5.637328433990478,
      "epoch": 4.2076711405627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033490141401435516,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22861524373292924,
      "num_tokens": 83624156.0,
      "step": 37465
    },
    {
      "entropy": 5.624855327606201,
      "epoch": 4.208232717470658,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003348627133113736,
      "loss": 4.8473,
      "mean_token_accuracy": 0.24010405391454698,
      "num_tokens": 83634655.0,
      "step": 37470
    },
    {
      "entropy": 5.602944183349609,
      "epoch": 4.208794294378615,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033482401070234006,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23659285753965378,
      "num_tokens": 83645015.0,
      "step": 37475
    },
    {
      "entropy": 5.628920793533325,
      "epoch": 4.209355871286573,
      "grad_norm": 1.078125,
      "learning_rate": 0.000334785306188487,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23525004088878632,
      "num_tokens": 83656540.0,
      "step": 37480
    },
    {
      "entropy": 5.648653984069824,
      "epoch": 4.20991744819453,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003347465997710464,
      "loss": 4.908,
      "mean_token_accuracy": 0.23661597967147827,
      "num_tokens": 83666798.0,
      "step": 37485
    },
    {
      "entropy": 5.606933546066284,
      "epoch": 4.2104790251024875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000334707891451251,
      "loss": 4.892,
      "mean_token_accuracy": 0.22975942194461824,
      "num_tokens": 83677058.0,
      "step": 37490
    },
    {
      "entropy": 5.578820276260376,
      "epoch": 4.211040602010446,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003346691812303331,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23564749509096145,
      "num_tokens": 83688126.0,
      "step": 37495
    },
    {
      "entropy": 5.685010099411011,
      "epoch": 4.211602178918403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003346304691095254,
      "loss": 5.0444,
      "mean_token_accuracy": 0.22047360241413116,
      "num_tokens": 83700298.0,
      "step": 37500
    },
    {
      "entropy": 5.624043989181518,
      "epoch": 4.21216375582636,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003345917550900604,
      "loss": 4.8463,
      "mean_token_accuracy": 0.23158847838640212,
      "num_tokens": 83711037.0,
      "step": 37505
    },
    {
      "entropy": 5.665063333511353,
      "epoch": 4.212725332734318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033455303917317075,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23116595149040223,
      "num_tokens": 83722306.0,
      "step": 37510
    },
    {
      "entropy": 5.675876808166504,
      "epoch": 4.213286909642275,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033451432136008937,
      "loss": 4.941,
      "mean_token_accuracy": 0.23530891090631484,
      "num_tokens": 83733306.0,
      "step": 37515
    },
    {
      "entropy": 5.561833906173706,
      "epoch": 4.2138484865502335,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003344756016520488,
      "loss": 4.8808,
      "mean_token_accuracy": 0.23195406943559646,
      "num_tokens": 83745475.0,
      "step": 37520
    },
    {
      "entropy": 5.64740514755249,
      "epoch": 4.214410063458191,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033443688005028213,
      "loss": 4.8614,
      "mean_token_accuracy": 0.2309349372982979,
      "num_tokens": 83756992.0,
      "step": 37525
    },
    {
      "entropy": 5.64736385345459,
      "epoch": 4.214971640366148,
      "grad_norm": 1.140625,
      "learning_rate": 0.000334398156556022,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22654524445533752,
      "num_tokens": 83768881.0,
      "step": 37530
    },
    {
      "entropy": 5.767061567306518,
      "epoch": 4.215533217274106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003343594311705016,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22967997193336487,
      "num_tokens": 83781144.0,
      "step": 37535
    },
    {
      "entropy": 5.730119943618774,
      "epoch": 4.216094794182063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033432070389495386,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2198358118534088,
      "num_tokens": 83792449.0,
      "step": 37540
    },
    {
      "entropy": 5.652020740509033,
      "epoch": 4.2166563710900204,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003342819747306119,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23490617126226426,
      "num_tokens": 83802889.0,
      "step": 37545
    },
    {
      "entropy": 5.704814004898071,
      "epoch": 4.2172179479979786,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033424324367870896,
      "loss": 4.981,
      "mean_token_accuracy": 0.22387195080518724,
      "num_tokens": 83813527.0,
      "step": 37550
    },
    {
      "entropy": 5.5747167587280275,
      "epoch": 4.217779524905936,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033420451074047807,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23749386072158812,
      "num_tokens": 83824963.0,
      "step": 37555
    },
    {
      "entropy": 5.6766805171966555,
      "epoch": 4.218341101813893,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003341657759171526,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2285766139626503,
      "num_tokens": 83836188.0,
      "step": 37560
    },
    {
      "entropy": 5.720235013961792,
      "epoch": 4.218902678721851,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003341270392099659,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23345244228839873,
      "num_tokens": 83848297.0,
      "step": 37565
    },
    {
      "entropy": 5.624458599090576,
      "epoch": 4.219464255629808,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033408830062015135,
      "loss": 4.8696,
      "mean_token_accuracy": 0.2209872454404831,
      "num_tokens": 83859074.0,
      "step": 37570
    },
    {
      "entropy": 5.594960308074951,
      "epoch": 4.220025832537766,
      "grad_norm": 1.125,
      "learning_rate": 0.00033404956014894237,
      "loss": 4.8948,
      "mean_token_accuracy": 0.2320300370454788,
      "num_tokens": 83869921.0,
      "step": 37575
    },
    {
      "entropy": 5.659229946136475,
      "epoch": 4.220587409445724,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033401081779757254,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22852270901203156,
      "num_tokens": 83881994.0,
      "step": 37580
    },
    {
      "entropy": 5.661713552474976,
      "epoch": 4.221148986353681,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003339720735672753,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22264990210533142,
      "num_tokens": 83893530.0,
      "step": 37585
    },
    {
      "entropy": 5.651681661605835,
      "epoch": 4.221710563261639,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033393332745928453,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2326386332511902,
      "num_tokens": 83904929.0,
      "step": 37590
    },
    {
      "entropy": 5.605773878097534,
      "epoch": 4.222272140169596,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033389457947483357,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23500357270240785,
      "num_tokens": 83915888.0,
      "step": 37595
    },
    {
      "entropy": 5.626063108444214,
      "epoch": 4.222833717077553,
      "grad_norm": 1.125,
      "learning_rate": 0.00033385582961515645,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22982733696699142,
      "num_tokens": 83927481.0,
      "step": 37600
    },
    {
      "entropy": 5.7299456119537355,
      "epoch": 4.2233952939855115,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033381707788148683,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2299892023205757,
      "num_tokens": 83938644.0,
      "step": 37605
    },
    {
      "entropy": 5.629663038253784,
      "epoch": 4.223956870893469,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033377832427505864,
      "loss": 4.8613,
      "mean_token_accuracy": 0.24248053133487701,
      "num_tokens": 83949165.0,
      "step": 37610
    },
    {
      "entropy": 5.662967348098755,
      "epoch": 4.224518447801427,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033373956879710593,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2261957362294197,
      "num_tokens": 83961112.0,
      "step": 37615
    },
    {
      "entropy": 5.668846559524536,
      "epoch": 4.225080024709384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033370081144886246,
      "loss": 4.9531,
      "mean_token_accuracy": 0.23566002696752547,
      "num_tokens": 83972451.0,
      "step": 37620
    },
    {
      "entropy": 5.559815168380737,
      "epoch": 4.225641601617341,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033366205223156237,
      "loss": 4.8239,
      "mean_token_accuracy": 0.234810009598732,
      "num_tokens": 83984082.0,
      "step": 37625
    },
    {
      "entropy": 5.724247550964355,
      "epoch": 4.226203178525299,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033362329114643987,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21136383563280106,
      "num_tokens": 83995853.0,
      "step": 37630
    },
    {
      "entropy": 5.71464581489563,
      "epoch": 4.226764755433257,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033358452819472895,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2262819692492485,
      "num_tokens": 84006440.0,
      "step": 37635
    },
    {
      "entropy": 5.782872438430786,
      "epoch": 4.227326332341214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033354576337766396,
      "loss": 5.0683,
      "mean_token_accuracy": 0.22323616296052934,
      "num_tokens": 84018968.0,
      "step": 37640
    },
    {
      "entropy": 5.675022315979004,
      "epoch": 4.227887909249172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003335069966964792,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23431886732578278,
      "num_tokens": 84030607.0,
      "step": 37645
    },
    {
      "entropy": 5.626776552200317,
      "epoch": 4.228449486157129,
      "grad_norm": 1.09375,
      "learning_rate": 0.000333468228152409,
      "loss": 4.8954,
      "mean_token_accuracy": 0.23087312430143356,
      "num_tokens": 84042598.0,
      "step": 37650
    },
    {
      "entropy": 5.572483253479004,
      "epoch": 4.229011063065086,
      "grad_norm": 1.25,
      "learning_rate": 0.00033342945774668766,
      "loss": 4.8328,
      "mean_token_accuracy": 0.23270253837108612,
      "num_tokens": 84052708.0,
      "step": 37655
    },
    {
      "entropy": 5.675103282928466,
      "epoch": 4.2295726399730444,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033339068548054974,
      "loss": 5.0082,
      "mean_token_accuracy": 0.231100133061409,
      "num_tokens": 84063573.0,
      "step": 37660
    },
    {
      "entropy": 5.763943386077881,
      "epoch": 4.230134216881002,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033335191135522974,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22878436893224716,
      "num_tokens": 84074132.0,
      "step": 37665
    },
    {
      "entropy": 5.5587340831756595,
      "epoch": 4.23069579378896,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033331313537196235,
      "loss": 4.8378,
      "mean_token_accuracy": 0.25550861805677416,
      "num_tokens": 84086424.0,
      "step": 37670
    },
    {
      "entropy": 5.68973822593689,
      "epoch": 4.231257370696917,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000333274357531982,
      "loss": 4.9659,
      "mean_token_accuracy": 0.2288963109254837,
      "num_tokens": 84098413.0,
      "step": 37675
    },
    {
      "entropy": 5.514787292480468,
      "epoch": 4.231818947604874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003332355778365235,
      "loss": 4.8896,
      "mean_token_accuracy": 0.23030814826488494,
      "num_tokens": 84109156.0,
      "step": 37680
    },
    {
      "entropy": 5.6633916854858395,
      "epoch": 4.232380524512832,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003331967962868217,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2345364049077034,
      "num_tokens": 84120677.0,
      "step": 37685
    },
    {
      "entropy": 5.665642309188843,
      "epoch": 4.2329421014207895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003331580128841112,
      "loss": 4.9314,
      "mean_token_accuracy": 0.23022504150867462,
      "num_tokens": 84132607.0,
      "step": 37690
    },
    {
      "entropy": 5.69099907875061,
      "epoch": 4.233503678328747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033311922762962716,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2284603461623192,
      "num_tokens": 84143958.0,
      "step": 37695
    },
    {
      "entropy": 5.78001127243042,
      "epoch": 4.234065255236705,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003330804405246043,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22599360048770906,
      "num_tokens": 84155154.0,
      "step": 37700
    },
    {
      "entropy": 5.68619933128357,
      "epoch": 4.234626832144662,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003330416515702777,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22421970069408417,
      "num_tokens": 84166100.0,
      "step": 37705
    },
    {
      "entropy": 5.704460382461548,
      "epoch": 4.23518840905262,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033300286076788237,
      "loss": 4.9851,
      "mean_token_accuracy": 0.23176732808351516,
      "num_tokens": 84177249.0,
      "step": 37710
    },
    {
      "entropy": 5.719858884811401,
      "epoch": 4.235749985960577,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003329640681186535,
      "loss": 4.936,
      "mean_token_accuracy": 0.22732725292444228,
      "num_tokens": 84187922.0,
      "step": 37715
    },
    {
      "entropy": 5.6761702537536625,
      "epoch": 4.236311562868535,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033292527362382626,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2343997836112976,
      "num_tokens": 84198929.0,
      "step": 37720
    },
    {
      "entropy": 5.564704895019531,
      "epoch": 4.236873139776493,
      "grad_norm": 1.25,
      "learning_rate": 0.00033288647728463576,
      "loss": 4.9335,
      "mean_token_accuracy": 0.23008327186107635,
      "num_tokens": 84209503.0,
      "step": 37725
    },
    {
      "entropy": 5.642581367492676,
      "epoch": 4.23743471668445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033284767910231745,
      "loss": 5.0048,
      "mean_token_accuracy": 0.21978932321071626,
      "num_tokens": 84220667.0,
      "step": 37730
    },
    {
      "entropy": 5.6967857837677,
      "epoch": 4.237996293592407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003328088790781065,
      "loss": 4.933,
      "mean_token_accuracy": 0.23126581609249114,
      "num_tokens": 84232072.0,
      "step": 37735
    },
    {
      "entropy": 5.732508277893066,
      "epoch": 4.238557870500365,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033277007721323854,
      "loss": 4.9798,
      "mean_token_accuracy": 0.23001950085163117,
      "num_tokens": 84243765.0,
      "step": 37740
    },
    {
      "entropy": 5.7159764766693115,
      "epoch": 4.2391194474083225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033273127350894884,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22583615481853486,
      "num_tokens": 84254869.0,
      "step": 37745
    },
    {
      "entropy": 5.709999513626099,
      "epoch": 4.23968102431628,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00033269246796647296,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2264400228857994,
      "num_tokens": 84264992.0,
      "step": 37750
    },
    {
      "entropy": 5.617316102981567,
      "epoch": 4.240242601224238,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003326536605870466,
      "loss": 4.9092,
      "mean_token_accuracy": 0.23394135683774947,
      "num_tokens": 84275648.0,
      "step": 37755
    },
    {
      "entropy": 5.738589525222778,
      "epoch": 4.240804178132195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033261485137190533,
      "loss": 4.9246,
      "mean_token_accuracy": 0.2303572937846184,
      "num_tokens": 84287217.0,
      "step": 37760
    },
    {
      "entropy": 5.666767120361328,
      "epoch": 4.241365755040153,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033257604032228485,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21593877375125886,
      "num_tokens": 84298764.0,
      "step": 37765
    },
    {
      "entropy": 5.716465950012207,
      "epoch": 4.24192733194811,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033253722743942086,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22242688983678818,
      "num_tokens": 84310016.0,
      "step": 37770
    },
    {
      "entropy": 5.645199489593506,
      "epoch": 4.2424889088560676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003324984127245492,
      "loss": 4.9464,
      "mean_token_accuracy": 0.23211029917001724,
      "num_tokens": 84321836.0,
      "step": 37775
    },
    {
      "entropy": 5.599290990829468,
      "epoch": 4.243050485764026,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003324595961789058,
      "loss": 4.857,
      "mean_token_accuracy": 0.2285323902964592,
      "num_tokens": 84332434.0,
      "step": 37780
    },
    {
      "entropy": 5.577896785736084,
      "epoch": 4.243612062671983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033242077780372664,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23542834669351578,
      "num_tokens": 84343997.0,
      "step": 37785
    },
    {
      "entropy": 5.63525800704956,
      "epoch": 4.24417363957994,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033238195760024766,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23689993619918823,
      "num_tokens": 84354157.0,
      "step": 37790
    },
    {
      "entropy": 5.590666913986206,
      "epoch": 4.244735216487898,
      "grad_norm": 1.125,
      "learning_rate": 0.00033234313556970485,
      "loss": 4.8288,
      "mean_token_accuracy": 0.23852373659610748,
      "num_tokens": 84365888.0,
      "step": 37795
    },
    {
      "entropy": 5.659765815734863,
      "epoch": 4.245296793395855,
      "grad_norm": 1.328125,
      "learning_rate": 0.00033230431171333435,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23361786454916,
      "num_tokens": 84376121.0,
      "step": 37800
    },
    {
      "entropy": 5.588488531112671,
      "epoch": 4.2458583703038135,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003322654860323723,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23607395142316817,
      "num_tokens": 84387318.0,
      "step": 37805
    },
    {
      "entropy": 5.666105175018311,
      "epoch": 4.246419947211771,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033222665852805514,
      "loss": 4.937,
      "mean_token_accuracy": 0.22337328046560287,
      "num_tokens": 84398952.0,
      "step": 37810
    },
    {
      "entropy": 5.6811727523803714,
      "epoch": 4.246981524119728,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00033218782920161884,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22422133833169938,
      "num_tokens": 84410010.0,
      "step": 37815
    },
    {
      "entropy": 5.664622020721436,
      "epoch": 4.247543101027686,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033214899805429984,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23434912860393525,
      "num_tokens": 84421879.0,
      "step": 37820
    },
    {
      "entropy": 5.744507265090943,
      "epoch": 4.248104677935643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033211016508733463,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2198643609881401,
      "num_tokens": 84432811.0,
      "step": 37825
    },
    {
      "entropy": 5.692774629592895,
      "epoch": 4.2486662548436005,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033207133030195964,
      "loss": 4.9748,
      "mean_token_accuracy": 0.23356997966766357,
      "num_tokens": 84444330.0,
      "step": 37830
    },
    {
      "entropy": 5.668860578536988,
      "epoch": 4.249227831751559,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003320324936994113,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2343224748969078,
      "num_tokens": 84455827.0,
      "step": 37835
    },
    {
      "entropy": 5.630180740356446,
      "epoch": 4.249789408659516,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033199365528092625,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23302194476127625,
      "num_tokens": 84465608.0,
      "step": 37840
    },
    {
      "entropy": 5.661642217636109,
      "epoch": 4.250350985567474,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033195481504774113,
      "loss": 4.8539,
      "mean_token_accuracy": 0.2340153932571411,
      "num_tokens": 84476142.0,
      "step": 37845
    },
    {
      "entropy": 5.69264407157898,
      "epoch": 4.250912562475431,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033191597300109263,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22754786610603334,
      "num_tokens": 84486496.0,
      "step": 37850
    },
    {
      "entropy": 5.639788198471069,
      "epoch": 4.251474139383388,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033187712914221744,
      "loss": 4.9067,
      "mean_token_accuracy": 0.23629194051027297,
      "num_tokens": 84497529.0,
      "step": 37855
    },
    {
      "entropy": 5.598657464981079,
      "epoch": 4.2520357162913465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003318382834723524,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23453774452209472,
      "num_tokens": 84508016.0,
      "step": 37860
    },
    {
      "entropy": 5.736274671554566,
      "epoch": 4.252597293199304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033179943599273445,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2237883761525154,
      "num_tokens": 84519787.0,
      "step": 37865
    },
    {
      "entropy": 5.643628168106079,
      "epoch": 4.253158870107261,
      "grad_norm": 1.25,
      "learning_rate": 0.0003317605867046003,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22720746994018554,
      "num_tokens": 84530908.0,
      "step": 37870
    },
    {
      "entropy": 5.772639560699463,
      "epoch": 4.253720447015219,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033172173560918716,
      "loss": 5.0647,
      "mean_token_accuracy": 0.21911345273256302,
      "num_tokens": 84543502.0,
      "step": 37875
    },
    {
      "entropy": 5.615488719940186,
      "epoch": 4.254282023923176,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033168288270773197,
      "loss": 4.8928,
      "mean_token_accuracy": 0.23451053500175476,
      "num_tokens": 84554500.0,
      "step": 37880
    },
    {
      "entropy": 5.721595096588135,
      "epoch": 4.254843600831133,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033164402800147177,
      "loss": 5.0749,
      "mean_token_accuracy": 0.2257968157529831,
      "num_tokens": 84566704.0,
      "step": 37885
    },
    {
      "entropy": 5.6654096126556395,
      "epoch": 4.2554051777390915,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003316051714916436,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23263275623321533,
      "num_tokens": 84578264.0,
      "step": 37890
    },
    {
      "entropy": 5.612163305282593,
      "epoch": 4.255966754647049,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033156631317948505,
      "loss": 4.7271,
      "mean_token_accuracy": 0.2456175833940506,
      "num_tokens": 84588513.0,
      "step": 37895
    },
    {
      "entropy": 5.678362607955933,
      "epoch": 4.256528331555007,
      "grad_norm": 1.15625,
      "learning_rate": 0.000331527453066233,
      "loss": 4.8939,
      "mean_token_accuracy": 0.24118649661540986,
      "num_tokens": 84600384.0,
      "step": 37900
    },
    {
      "entropy": 5.712940406799317,
      "epoch": 4.257089908462964,
      "grad_norm": 1.265625,
      "learning_rate": 0.000331488591153125,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22608315497636794,
      "num_tokens": 84610827.0,
      "step": 37905
    },
    {
      "entropy": 5.65475435256958,
      "epoch": 4.257651485370921,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003314497274413983,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23943812847137452,
      "num_tokens": 84621130.0,
      "step": 37910
    },
    {
      "entropy": 5.675967836380005,
      "epoch": 4.258213062278879,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033141086193229043,
      "loss": 4.9579,
      "mean_token_accuracy": 0.23274551779031755,
      "num_tokens": 84633428.0,
      "step": 37915
    },
    {
      "entropy": 5.638808059692383,
      "epoch": 4.258774639186837,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003313719946270388,
      "loss": 4.9305,
      "mean_token_accuracy": 0.23201637715101242,
      "num_tokens": 84644519.0,
      "step": 37920
    },
    {
      "entropy": 5.722440242767334,
      "epoch": 4.259336216094794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033133312552688094,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2270657017827034,
      "num_tokens": 84654752.0,
      "step": 37925
    },
    {
      "entropy": 5.6750922203063965,
      "epoch": 4.259897793002752,
      "grad_norm": 1.125,
      "learning_rate": 0.00033129425463305465,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2395712122321129,
      "num_tokens": 84666384.0,
      "step": 37930
    },
    {
      "entropy": 5.767443609237671,
      "epoch": 4.260459369910709,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003312553819467974,
      "loss": 5.035,
      "mean_token_accuracy": 0.223727186024189,
      "num_tokens": 84676630.0,
      "step": 37935
    },
    {
      "entropy": 5.584244680404663,
      "epoch": 4.261020946818666,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033121650746934693,
      "loss": 4.9018,
      "mean_token_accuracy": 0.2349336102604866,
      "num_tokens": 84689407.0,
      "step": 37940
    },
    {
      "entropy": 5.680332469940185,
      "epoch": 4.2615825237266245,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003311776312019412,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22634439021348954,
      "num_tokens": 84699237.0,
      "step": 37945
    },
    {
      "entropy": 5.6781689643859865,
      "epoch": 4.262144100634582,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003311387531458177,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2319628432393074,
      "num_tokens": 84710074.0,
      "step": 37950
    },
    {
      "entropy": 5.6758664608001705,
      "epoch": 4.26270567754254,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003310998733022146,
      "loss": 5.0756,
      "mean_token_accuracy": 0.21470967233181,
      "num_tokens": 84721315.0,
      "step": 37955
    },
    {
      "entropy": 5.736408948898315,
      "epoch": 4.263267254450497,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003310609916723698,
      "loss": 4.984,
      "mean_token_accuracy": 0.23390817642211914,
      "num_tokens": 84732654.0,
      "step": 37960
    },
    {
      "entropy": 5.73418116569519,
      "epoch": 4.263828831358454,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003310221082575213,
      "loss": 5.0026,
      "mean_token_accuracy": 0.23417560756206512,
      "num_tokens": 84745568.0,
      "step": 37965
    },
    {
      "entropy": 5.707739591598511,
      "epoch": 4.264390408266412,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033098322305890715,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22296725064516068,
      "num_tokens": 84756443.0,
      "step": 37970
    },
    {
      "entropy": 5.70841875076294,
      "epoch": 4.26495198517437,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033094433607776545,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2354825273156166,
      "num_tokens": 84766789.0,
      "step": 37975
    },
    {
      "entropy": 5.706614446640015,
      "epoch": 4.265513562082327,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033090544731533434,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21941515803337097,
      "num_tokens": 84778186.0,
      "step": 37980
    },
    {
      "entropy": 5.688936758041382,
      "epoch": 4.266075138990285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003308665567728522,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22632597684860228,
      "num_tokens": 84788431.0,
      "step": 37985
    },
    {
      "entropy": 5.729051494598389,
      "epoch": 4.266636715898242,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033082766445155715,
      "loss": 5.0279,
      "mean_token_accuracy": 0.22891363650560378,
      "num_tokens": 84799881.0,
      "step": 37990
    },
    {
      "entropy": 5.696711921691895,
      "epoch": 4.2671982928062,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033078877035268766,
      "loss": 4.9716,
      "mean_token_accuracy": 0.2290883705019951,
      "num_tokens": 84810710.0,
      "step": 37995
    },
    {
      "entropy": 5.75107364654541,
      "epoch": 4.267759869714157,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033074987447748205,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23760987371206282,
      "num_tokens": 84821932.0,
      "step": 38000
    },
    {
      "entropy": 5.619548225402832,
      "epoch": 4.268321446622115,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033071097682717883,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23744158148765565,
      "num_tokens": 84832662.0,
      "step": 38005
    },
    {
      "entropy": 5.713272666931152,
      "epoch": 4.268883023530073,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003306720774030165,
      "loss": 4.9426,
      "mean_token_accuracy": 0.23155046701431276,
      "num_tokens": 84843199.0,
      "step": 38010
    },
    {
      "entropy": 5.652693510055542,
      "epoch": 4.26944460043803,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003306331762062336,
      "loss": 4.908,
      "mean_token_accuracy": 0.23266184329986572,
      "num_tokens": 84853367.0,
      "step": 38015
    },
    {
      "entropy": 5.64333028793335,
      "epoch": 4.270006177345987,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003305942732380689,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2340444192290306,
      "num_tokens": 84863527.0,
      "step": 38020
    },
    {
      "entropy": 5.576052188873291,
      "epoch": 4.270567754253945,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003305553684997609,
      "loss": 4.872,
      "mean_token_accuracy": 0.2325233906507492,
      "num_tokens": 84874929.0,
      "step": 38025
    },
    {
      "entropy": 5.658908367156982,
      "epoch": 4.2711293311619025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033051646199254843,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23439836204051973,
      "num_tokens": 84885182.0,
      "step": 38030
    },
    {
      "entropy": 5.718637180328369,
      "epoch": 4.271690908069861,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003304775537176702,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22533922046422958,
      "num_tokens": 84896115.0,
      "step": 38035
    },
    {
      "entropy": 5.656443738937378,
      "epoch": 4.272252484977818,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003304386436763652,
      "loss": 4.959,
      "mean_token_accuracy": 0.22455570697784424,
      "num_tokens": 84908174.0,
      "step": 38040
    },
    {
      "entropy": 5.691889333724975,
      "epoch": 4.272814061885775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003303997318698723,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22152685970067978,
      "num_tokens": 84920893.0,
      "step": 38045
    },
    {
      "entropy": 5.713045644760132,
      "epoch": 4.273375638793733,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003303608182994304,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23089690655469894,
      "num_tokens": 84931860.0,
      "step": 38050
    },
    {
      "entropy": 5.636694431304932,
      "epoch": 4.27393721570169,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003303219029662787,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2439257323741913,
      "num_tokens": 84942533.0,
      "step": 38055
    },
    {
      "entropy": 5.5484740257263185,
      "epoch": 4.274498792609648,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033028298587165606,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23241391479969026,
      "num_tokens": 84954018.0,
      "step": 38060
    },
    {
      "entropy": 5.701983547210693,
      "epoch": 4.275060369517606,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003302440670168017,
      "loss": 5.0654,
      "mean_token_accuracy": 0.22285820543766022,
      "num_tokens": 84966150.0,
      "step": 38065
    },
    {
      "entropy": 5.723220348358154,
      "epoch": 4.275621946425563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003302051464029548,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22483697384595872,
      "num_tokens": 84977668.0,
      "step": 38070
    },
    {
      "entropy": 5.651337146759033,
      "epoch": 4.27618352333352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033016622403135473,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22906267940998076,
      "num_tokens": 84988735.0,
      "step": 38075
    },
    {
      "entropy": 5.6185204029083256,
      "epoch": 4.276745100241478,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033012729990324063,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22334937155246734,
      "num_tokens": 84999573.0,
      "step": 38080
    },
    {
      "entropy": 5.594700670242309,
      "epoch": 4.2773066771494355,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003300883740198519,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22835638523101806,
      "num_tokens": 85010582.0,
      "step": 38085
    },
    {
      "entropy": 5.619273710250854,
      "epoch": 4.277868254057394,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033004944638242805,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23428082168102266,
      "num_tokens": 85021563.0,
      "step": 38090
    },
    {
      "entropy": 5.772944784164428,
      "epoch": 4.278429830965351,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003300105169922084,
      "loss": 5.1545,
      "mean_token_accuracy": 0.2157556638121605,
      "num_tokens": 85033231.0,
      "step": 38095
    },
    {
      "entropy": 5.621572542190552,
      "epoch": 4.278991407873308,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003299715858504326,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23986939042806626,
      "num_tokens": 85044110.0,
      "step": 38100
    },
    {
      "entropy": 5.61566686630249,
      "epoch": 4.279552984781266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032993265295834013,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23262737244367598,
      "num_tokens": 85055425.0,
      "step": 38105
    },
    {
      "entropy": 5.686057043075562,
      "epoch": 4.280114561689223,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003298937183171707,
      "loss": 4.9876,
      "mean_token_accuracy": 0.23010520339012147,
      "num_tokens": 85067358.0,
      "step": 38110
    },
    {
      "entropy": 5.742514801025391,
      "epoch": 4.2806761385971805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000329854781928164,
      "loss": 4.9617,
      "mean_token_accuracy": 0.23286068588495254,
      "num_tokens": 85078967.0,
      "step": 38115
    },
    {
      "entropy": 5.67026834487915,
      "epoch": 4.281237715505139,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003298158437925598,
      "loss": 4.8637,
      "mean_token_accuracy": 0.23255991637706758,
      "num_tokens": 85088515.0,
      "step": 38120
    },
    {
      "entropy": 5.623492002487183,
      "epoch": 4.281799292413096,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003297769039115978,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23290664553642274,
      "num_tokens": 85099535.0,
      "step": 38125
    },
    {
      "entropy": 5.778953552246094,
      "epoch": 4.282360869321053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032973796228651797,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21478398144245148,
      "num_tokens": 85110818.0,
      "step": 38130
    },
    {
      "entropy": 5.656431531906128,
      "epoch": 4.282922446229011,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032969901891856027,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22423724830150604,
      "num_tokens": 85120362.0,
      "step": 38135
    },
    {
      "entropy": 5.637246036529541,
      "epoch": 4.283484023136968,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032966007380896443,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2307762771844864,
      "num_tokens": 85131358.0,
      "step": 38140
    },
    {
      "entropy": 5.721139335632325,
      "epoch": 4.2840456000449265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003296211269589708,
      "loss": 4.9428,
      "mean_token_accuracy": 0.23006663024425505,
      "num_tokens": 85143395.0,
      "step": 38145
    },
    {
      "entropy": 5.670250511169433,
      "epoch": 4.284607176952884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003295821783698192,
      "loss": 4.8766,
      "mean_token_accuracy": 0.24118876755237578,
      "num_tokens": 85155654.0,
      "step": 38150
    },
    {
      "entropy": 5.64733362197876,
      "epoch": 4.285168753860841,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032954322804274995,
      "loss": 5.0295,
      "mean_token_accuracy": 0.2232024848461151,
      "num_tokens": 85168285.0,
      "step": 38155
    },
    {
      "entropy": 5.685760593414306,
      "epoch": 4.285730330768799,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032950427597900315,
      "loss": 5.0105,
      "mean_token_accuracy": 0.22273268252611161,
      "num_tokens": 85180058.0,
      "step": 38160
    },
    {
      "entropy": 5.655829811096192,
      "epoch": 4.286291907676756,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000329465322179819,
      "loss": 4.9792,
      "mean_token_accuracy": 0.22978472262620925,
      "num_tokens": 85190990.0,
      "step": 38165
    },
    {
      "entropy": 5.7767304420471195,
      "epoch": 4.2868534845847135,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032942636664643794,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22890933901071547,
      "num_tokens": 85201136.0,
      "step": 38170
    },
    {
      "entropy": 5.712374019622803,
      "epoch": 4.287415061492672,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032938740938010024,
      "loss": 4.9732,
      "mean_token_accuracy": 0.23075773417949677,
      "num_tokens": 85212424.0,
      "step": 38175
    },
    {
      "entropy": 5.736089563369751,
      "epoch": 4.287976638400629,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003293484503820463,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22589612007141113,
      "num_tokens": 85223767.0,
      "step": 38180
    },
    {
      "entropy": 5.74754638671875,
      "epoch": 4.288538215308587,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032930948965351664,
      "loss": 4.9989,
      "mean_token_accuracy": 0.23113124072551727,
      "num_tokens": 85234374.0,
      "step": 38185
    },
    {
      "entropy": 5.64695782661438,
      "epoch": 4.289099792216544,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003292705271957518,
      "loss": 4.8868,
      "mean_token_accuracy": 0.23705639094114303,
      "num_tokens": 85245322.0,
      "step": 38190
    },
    {
      "entropy": 5.7683430194854735,
      "epoch": 4.289661369124501,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032923156300999223,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2231900453567505,
      "num_tokens": 85257114.0,
      "step": 38195
    },
    {
      "entropy": 5.731420564651489,
      "epoch": 4.2902229460324595,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032919259709747874,
      "loss": 4.933,
      "mean_token_accuracy": 0.22911721765995025,
      "num_tokens": 85267564.0,
      "step": 38200
    },
    {
      "entropy": 5.637549781799317,
      "epoch": 4.290784522940417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032915362945945195,
      "loss": 4.8852,
      "mean_token_accuracy": 0.23504430055618286,
      "num_tokens": 85277763.0,
      "step": 38205
    },
    {
      "entropy": 5.718469476699829,
      "epoch": 4.291346099848374,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032911466009715254,
      "loss": 5.0483,
      "mean_token_accuracy": 0.2264597862958908,
      "num_tokens": 85288504.0,
      "step": 38210
    },
    {
      "entropy": 5.708159923553467,
      "epoch": 4.291907676756332,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032907568901182136,
      "loss": 4.916,
      "mean_token_accuracy": 0.23589191138744353,
      "num_tokens": 85299928.0,
      "step": 38215
    },
    {
      "entropy": 5.660965394973755,
      "epoch": 4.292469253664289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003290367162046993,
      "loss": 4.9254,
      "mean_token_accuracy": 0.23291152715682983,
      "num_tokens": 85311052.0,
      "step": 38220
    },
    {
      "entropy": 5.685665225982666,
      "epoch": 4.293030830572247,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032899774167702723,
      "loss": 4.987,
      "mean_token_accuracy": 0.22913981676101686,
      "num_tokens": 85322023.0,
      "step": 38225
    },
    {
      "entropy": 5.608619689941406,
      "epoch": 4.2935924074802045,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003289587654300461,
      "loss": 4.9498,
      "mean_token_accuracy": 0.23035528957843782,
      "num_tokens": 85334160.0,
      "step": 38230
    },
    {
      "entropy": 5.7665928363800045,
      "epoch": 4.294153984388162,
      "grad_norm": 1.234375,
      "learning_rate": 0.000328919787464997,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2180406481027603,
      "num_tokens": 85345378.0,
      "step": 38235
    },
    {
      "entropy": 5.708695554733277,
      "epoch": 4.29471556129612,
      "grad_norm": 1.15625,
      "learning_rate": 0.000328880807783121,
      "loss": 4.9285,
      "mean_token_accuracy": 0.23127275407314302,
      "num_tokens": 85357610.0,
      "step": 38240
    },
    {
      "entropy": 5.714637517929077,
      "epoch": 4.295277138204077,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032884182638565914,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22774459272623063,
      "num_tokens": 85368749.0,
      "step": 38245
    },
    {
      "entropy": 5.695669889450073,
      "epoch": 4.295838715112034,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032880284327385255,
      "loss": 4.962,
      "mean_token_accuracy": 0.2301877647638321,
      "num_tokens": 85379488.0,
      "step": 38250
    },
    {
      "entropy": 5.655576848983765,
      "epoch": 4.296400292019992,
      "grad_norm": 1.25,
      "learning_rate": 0.0003287638584489426,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23207154572010041,
      "num_tokens": 85390581.0,
      "step": 38255
    },
    {
      "entropy": 5.618980455398559,
      "epoch": 4.29696186892795,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003287248719121706,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22533250898122786,
      "num_tokens": 85402430.0,
      "step": 38260
    },
    {
      "entropy": 5.640824222564698,
      "epoch": 4.297523445835907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032868588366477785,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23615516126155853,
      "num_tokens": 85414932.0,
      "step": 38265
    },
    {
      "entropy": 5.672748994827271,
      "epoch": 4.298085022743865,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032864689370800564,
      "loss": 4.8673,
      "mean_token_accuracy": 0.2395637884736061,
      "num_tokens": 85426484.0,
      "step": 38270
    },
    {
      "entropy": 5.655053615570068,
      "epoch": 4.298646599651822,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003286079020430957,
      "loss": 4.8857,
      "mean_token_accuracy": 0.23220936506986617,
      "num_tokens": 85436864.0,
      "step": 38275
    },
    {
      "entropy": 5.588532304763794,
      "epoch": 4.29920817655978,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032856890867128926,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23511531054973603,
      "num_tokens": 85446303.0,
      "step": 38280
    },
    {
      "entropy": 5.638495588302613,
      "epoch": 4.2997697534677375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003285299135938279,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22983117550611495,
      "num_tokens": 85457218.0,
      "step": 38285
    },
    {
      "entropy": 5.6770172119140625,
      "epoch": 4.300331330375695,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032849091681195347,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22945036441087724,
      "num_tokens": 85469085.0,
      "step": 38290
    },
    {
      "entropy": 5.637843608856201,
      "epoch": 4.300892907283653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032845191832690747,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22831794768571853,
      "num_tokens": 85478966.0,
      "step": 38295
    },
    {
      "entropy": 5.631999063491821,
      "epoch": 4.30145448419161,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003284129181399317,
      "loss": 4.8764,
      "mean_token_accuracy": 0.24311271905899048,
      "num_tokens": 85489742.0,
      "step": 38300
    },
    {
      "entropy": 5.600444221496582,
      "epoch": 4.302016061099567,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003283739162522678,
      "loss": 4.8172,
      "mean_token_accuracy": 0.2464737832546234,
      "num_tokens": 85500766.0,
      "step": 38305
    },
    {
      "entropy": 5.672322368621826,
      "epoch": 4.302577638007525,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003283349126651578,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2270435318350792,
      "num_tokens": 85512206.0,
      "step": 38310
    },
    {
      "entropy": 5.636661148071289,
      "epoch": 4.303139214915483,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032829590737984345,
      "loss": 4.9197,
      "mean_token_accuracy": 0.23236653059720994,
      "num_tokens": 85523881.0,
      "step": 38315
    },
    {
      "entropy": 5.614172744750976,
      "epoch": 4.30370079182344,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032825690039756675,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23156623393297196,
      "num_tokens": 85535130.0,
      "step": 38320
    },
    {
      "entropy": 5.73844575881958,
      "epoch": 4.304262368731398,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032821789171956966,
      "loss": 4.9476,
      "mean_token_accuracy": 0.23044935762882232,
      "num_tokens": 85547261.0,
      "step": 38325
    },
    {
      "entropy": 5.711596059799194,
      "epoch": 4.304823945639355,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003281788813470943,
      "loss": 4.9461,
      "mean_token_accuracy": 0.23043714463710785,
      "num_tokens": 85558300.0,
      "step": 38330
    },
    {
      "entropy": 5.621511030197143,
      "epoch": 4.305385522547313,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003281398692813827,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23018787503242494,
      "num_tokens": 85569370.0,
      "step": 38335
    },
    {
      "entropy": 5.61142783164978,
      "epoch": 4.30594709945527,
      "grad_norm": 1.1875,
      "learning_rate": 0.000328100855523677,
      "loss": 4.8954,
      "mean_token_accuracy": 0.23888083845376967,
      "num_tokens": 85579527.0,
      "step": 38340
    },
    {
      "entropy": 5.698360872268677,
      "epoch": 4.306508676363228,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032806184007521955,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22724239230155946,
      "num_tokens": 85590791.0,
      "step": 38345
    },
    {
      "entropy": 5.626651906967163,
      "epoch": 4.307070253271186,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003280228229372525,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23527020514011382,
      "num_tokens": 85601210.0,
      "step": 38350
    },
    {
      "entropy": 5.693554353713989,
      "epoch": 4.307631830179143,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003279838041110182,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22819286435842515,
      "num_tokens": 85611811.0,
      "step": 38355
    },
    {
      "entropy": 5.652960109710693,
      "epoch": 4.3081934070871,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032794478359775904,
      "loss": 5.0464,
      "mean_token_accuracy": 0.21876865178346633,
      "num_tokens": 85622046.0,
      "step": 38360
    },
    {
      "entropy": 5.6862678050994875,
      "epoch": 4.308754983995058,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032790576139871743,
      "loss": 4.9257,
      "mean_token_accuracy": 0.23584501296281815,
      "num_tokens": 85633094.0,
      "step": 38365
    },
    {
      "entropy": 5.636904859542847,
      "epoch": 4.3093165609030155,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003278667375151359,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2249029442667961,
      "num_tokens": 85644224.0,
      "step": 38370
    },
    {
      "entropy": 5.673505353927612,
      "epoch": 4.309878137810974,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032782771194825685,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22247726321220399,
      "num_tokens": 85654950.0,
      "step": 38375
    },
    {
      "entropy": 5.706162548065185,
      "epoch": 4.310439714718931,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032778868469932304,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22703638225793837,
      "num_tokens": 85665250.0,
      "step": 38380
    },
    {
      "entropy": 5.729625797271728,
      "epoch": 4.311001291626888,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032774965576957705,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22406409680843353,
      "num_tokens": 85676856.0,
      "step": 38385
    },
    {
      "entropy": 5.827110719680786,
      "epoch": 4.311562868534846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003277106251602615,
      "loss": 5.0844,
      "mean_token_accuracy": 0.22489848881959915,
      "num_tokens": 85689733.0,
      "step": 38390
    },
    {
      "entropy": 5.670493030548096,
      "epoch": 4.312124445442803,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032767159287261914,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23890776932239532,
      "num_tokens": 85699678.0,
      "step": 38395
    },
    {
      "entropy": 5.681526708602905,
      "epoch": 4.312686022350761,
      "grad_norm": 1.265625,
      "learning_rate": 0.000327632558907893,
      "loss": 5.035,
      "mean_token_accuracy": 0.2219575360417366,
      "num_tokens": 85712130.0,
      "step": 38400
    },
    {
      "entropy": 5.674405241012574,
      "epoch": 4.313247599258719,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032759352326732573,
      "loss": 4.82,
      "mean_token_accuracy": 0.23809136748313903,
      "num_tokens": 85723025.0,
      "step": 38405
    },
    {
      "entropy": 5.599339962005615,
      "epoch": 4.313809176166676,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032755448595216027,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2377509817481041,
      "num_tokens": 85733551.0,
      "step": 38410
    },
    {
      "entropy": 5.682222461700439,
      "epoch": 4.314370753074634,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032751544696363945,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22655473947525023,
      "num_tokens": 85745633.0,
      "step": 38415
    },
    {
      "entropy": 5.674498844146728,
      "epoch": 4.314932329982591,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032747640630300664,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23279005140066147,
      "num_tokens": 85757522.0,
      "step": 38420
    },
    {
      "entropy": 5.7087654113769535,
      "epoch": 4.3154939068905485,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032743736397150455,
      "loss": 4.909,
      "mean_token_accuracy": 0.23955732882022857,
      "num_tokens": 85768666.0,
      "step": 38425
    },
    {
      "entropy": 5.609964895248413,
      "epoch": 4.316055483798507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003273983199703765,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23575494140386583,
      "num_tokens": 85780527.0,
      "step": 38430
    },
    {
      "entropy": 5.622522020339966,
      "epoch": 4.316617060706464,
      "grad_norm": 1.125,
      "learning_rate": 0.0003273592743008656,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22738635838031768,
      "num_tokens": 85792166.0,
      "step": 38435
    },
    {
      "entropy": 5.74838080406189,
      "epoch": 4.317178637614421,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032732022696421514,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2239519789814949,
      "num_tokens": 85803706.0,
      "step": 38440
    },
    {
      "entropy": 5.705405616760254,
      "epoch": 4.317740214522379,
      "grad_norm": 1.25,
      "learning_rate": 0.0003272811779616684,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2335553526878357,
      "num_tokens": 85815515.0,
      "step": 38445
    },
    {
      "entropy": 5.7039295673370365,
      "epoch": 4.318301791430336,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032724212729446864,
      "loss": 4.8759,
      "mean_token_accuracy": 0.2356896236538887,
      "num_tokens": 85826359.0,
      "step": 38450
    },
    {
      "entropy": 5.695631122589111,
      "epoch": 4.3188633683382935,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003272030749638593,
      "loss": 4.9638,
      "mean_token_accuracy": 0.23034477680921556,
      "num_tokens": 85837977.0,
      "step": 38455
    },
    {
      "entropy": 5.7192668437957765,
      "epoch": 4.319424945246252,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003271640209710837,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22712725549936294,
      "num_tokens": 85850356.0,
      "step": 38460
    },
    {
      "entropy": 5.694899272918701,
      "epoch": 4.319986522154209,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003271249653173855,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22679512202739716,
      "num_tokens": 85863101.0,
      "step": 38465
    },
    {
      "entropy": 5.705937337875366,
      "epoch": 4.320548099062167,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032708590800400826,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22397134304046631,
      "num_tokens": 85873374.0,
      "step": 38470
    },
    {
      "entropy": 5.755907154083252,
      "epoch": 4.321109675970124,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003270468490321955,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2263987496495247,
      "num_tokens": 85886167.0,
      "step": 38475
    },
    {
      "entropy": 5.675031852722168,
      "epoch": 4.321671252878081,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032700778840319086,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2290288284420967,
      "num_tokens": 85897040.0,
      "step": 38480
    },
    {
      "entropy": 5.637288904190063,
      "epoch": 4.3222328297860395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000326968726118238,
      "loss": 4.905,
      "mean_token_accuracy": 0.22935792654752732,
      "num_tokens": 85908386.0,
      "step": 38485
    },
    {
      "entropy": 5.611691331863403,
      "epoch": 4.322794406693997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032692966217858074,
      "loss": 4.8788,
      "mean_token_accuracy": 0.24960067570209504,
      "num_tokens": 85920313.0,
      "step": 38490
    },
    {
      "entropy": 5.690779781341552,
      "epoch": 4.323355983601954,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003268905965854629,
      "loss": 5.0006,
      "mean_token_accuracy": 0.22649172842502593,
      "num_tokens": 85931273.0,
      "step": 38495
    },
    {
      "entropy": 5.705136728286743,
      "epoch": 4.323917560509912,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032685152934012844,
      "loss": 4.9939,
      "mean_token_accuracy": 0.22559309601783753,
      "num_tokens": 85942952.0,
      "step": 38500
    },
    {
      "entropy": 5.650049448013306,
      "epoch": 4.324479137417869,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003268124604438211,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23242783695459365,
      "num_tokens": 85953331.0,
      "step": 38505
    },
    {
      "entropy": 5.653725051879883,
      "epoch": 4.3250407143258265,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032677338989778494,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2348245069384575,
      "num_tokens": 85965142.0,
      "step": 38510
    },
    {
      "entropy": 5.69839391708374,
      "epoch": 4.325602291233785,
      "grad_norm": 1.234375,
      "learning_rate": 0.000326734317703264,
      "loss": 4.9965,
      "mean_token_accuracy": 0.23394458144903182,
      "num_tokens": 85976258.0,
      "step": 38515
    },
    {
      "entropy": 5.687456941604614,
      "epoch": 4.326163868141742,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003266952438615023,
      "loss": 4.9423,
      "mean_token_accuracy": 0.23411486595869063,
      "num_tokens": 85987976.0,
      "step": 38520
    },
    {
      "entropy": 5.615362453460693,
      "epoch": 4.3267254450497,
      "grad_norm": 1.125,
      "learning_rate": 0.00032665616837374397,
      "loss": 4.8338,
      "mean_token_accuracy": 0.23027164041996,
      "num_tokens": 85999283.0,
      "step": 38525
    },
    {
      "entropy": 5.592931270599365,
      "epoch": 4.327287021957657,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003266170912412332,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23340798765420914,
      "num_tokens": 86010484.0,
      "step": 38530
    },
    {
      "entropy": 5.688470935821533,
      "epoch": 4.327848598865614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003265780124652143,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22281753718852998,
      "num_tokens": 86021582.0,
      "step": 38535
    },
    {
      "entropy": 5.6593883514404295,
      "epoch": 4.3284101757735725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032653893204693136,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22326709628105162,
      "num_tokens": 86033847.0,
      "step": 38540
    },
    {
      "entropy": 5.704416275024414,
      "epoch": 4.32897175268153,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003264998499876289,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23088097274303437,
      "num_tokens": 86044870.0,
      "step": 38545
    },
    {
      "entropy": 5.596376276016235,
      "epoch": 4.329533329589487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003264607662885512,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23833616971969604,
      "num_tokens": 86055807.0,
      "step": 38550
    },
    {
      "entropy": 5.644133186340332,
      "epoch": 4.330094906497445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032642168095094274,
      "loss": 5.021,
      "mean_token_accuracy": 0.22453206181526184,
      "num_tokens": 86066869.0,
      "step": 38555
    },
    {
      "entropy": 5.755071210861206,
      "epoch": 4.330656483405402,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003263825939760481,
      "loss": 5.0234,
      "mean_token_accuracy": 0.22124811708927156,
      "num_tokens": 86078870.0,
      "step": 38560
    },
    {
      "entropy": 5.638040828704834,
      "epoch": 4.33121806031336,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032634350536511167,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22607362866401673,
      "num_tokens": 86090940.0,
      "step": 38565
    },
    {
      "entropy": 5.625986909866333,
      "epoch": 4.3317796372213175,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00032630441511937813,
      "loss": 4.8439,
      "mean_token_accuracy": 0.23323971778154373,
      "num_tokens": 86101171.0,
      "step": 38570
    },
    {
      "entropy": 5.623922634124756,
      "epoch": 4.332341214129275,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032626532324009204,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22609673142433168,
      "num_tokens": 86111614.0,
      "step": 38575
    },
    {
      "entropy": 5.633909273147583,
      "epoch": 4.332902791037233,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003262262297284983,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23838466852903367,
      "num_tokens": 86124131.0,
      "step": 38580
    },
    {
      "entropy": 5.680058908462525,
      "epoch": 4.33346436794519,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032618713458584147,
      "loss": 4.9234,
      "mean_token_accuracy": 0.230645589530468,
      "num_tokens": 86135584.0,
      "step": 38585
    },
    {
      "entropy": 5.657023572921753,
      "epoch": 4.334025944853147,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003261480378133665,
      "loss": 4.835,
      "mean_token_accuracy": 0.23692988306283952,
      "num_tokens": 86146163.0,
      "step": 38590
    },
    {
      "entropy": 5.693690347671509,
      "epoch": 4.334587521761105,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003261089394123181,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22302511781454087,
      "num_tokens": 86157199.0,
      "step": 38595
    },
    {
      "entropy": 5.708629989624024,
      "epoch": 4.335149098669063,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003260698393839413,
      "loss": 4.9225,
      "mean_token_accuracy": 0.22754930406808854,
      "num_tokens": 86167523.0,
      "step": 38600
    },
    {
      "entropy": 5.705497884750367,
      "epoch": 4.335710675577021,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032603073772948096,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2282775267958641,
      "num_tokens": 86179341.0,
      "step": 38605
    },
    {
      "entropy": 5.692293548583985,
      "epoch": 4.336272252484978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032599163445018204,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2328698828816414,
      "num_tokens": 86190154.0,
      "step": 38610
    },
    {
      "entropy": 5.6109601020812985,
      "epoch": 4.336833829392935,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003259525295472899,
      "loss": 4.923,
      "mean_token_accuracy": 0.2291731283068657,
      "num_tokens": 86201598.0,
      "step": 38615
    },
    {
      "entropy": 5.665542316436768,
      "epoch": 4.337395406300893,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032591342302204935,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2275807797908783,
      "num_tokens": 86212033.0,
      "step": 38620
    },
    {
      "entropy": 5.627749347686768,
      "epoch": 4.3379569832088505,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003258743148757057,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2249969646334648,
      "num_tokens": 86222931.0,
      "step": 38625
    },
    {
      "entropy": 5.6451812267303465,
      "epoch": 4.338518560116808,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032583520510950405,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22569953501224518,
      "num_tokens": 86234022.0,
      "step": 38630
    },
    {
      "entropy": 5.603495359420776,
      "epoch": 4.339080137024766,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032579609372468984,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23269573897123336,
      "num_tokens": 86244267.0,
      "step": 38635
    },
    {
      "entropy": 5.643589305877685,
      "epoch": 4.339641713932723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032575698072250827,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2378285065293312,
      "num_tokens": 86254200.0,
      "step": 38640
    },
    {
      "entropy": 5.6671913146972654,
      "epoch": 4.34020329084068,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003257178661042047,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22245164811611176,
      "num_tokens": 86266523.0,
      "step": 38645
    },
    {
      "entropy": 5.718580198287964,
      "epoch": 4.340764867748638,
      "grad_norm": 1.25,
      "learning_rate": 0.00032567874987102463,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22857170253992082,
      "num_tokens": 86276949.0,
      "step": 38650
    },
    {
      "entropy": 5.689911031723023,
      "epoch": 4.341326444656596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003256396320242136,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22883136421442032,
      "num_tokens": 86287412.0,
      "step": 38655
    },
    {
      "entropy": 5.661675024032593,
      "epoch": 4.341888021564554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000325600512565017,
      "loss": 4.9075,
      "mean_token_accuracy": 0.23805847018957138,
      "num_tokens": 86298540.0,
      "step": 38660
    },
    {
      "entropy": 5.623710870742798,
      "epoch": 4.342449598472511,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032556139149468046,
      "loss": 4.8926,
      "mean_token_accuracy": 0.23759972900152207,
      "num_tokens": 86309351.0,
      "step": 38665
    },
    {
      "entropy": 5.766062021255493,
      "epoch": 4.343011175380468,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032552226881444956,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22800525277853012,
      "num_tokens": 86320873.0,
      "step": 38670
    },
    {
      "entropy": 5.700601100921631,
      "epoch": 4.343572752288426,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032548314452557006,
      "loss": 4.9959,
      "mean_token_accuracy": 0.2215322658419609,
      "num_tokens": 86332889.0,
      "step": 38675
    },
    {
      "entropy": 5.610351276397705,
      "epoch": 4.344134329196383,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032544401862928766,
      "loss": 4.8003,
      "mean_token_accuracy": 0.23685340881347655,
      "num_tokens": 86344155.0,
      "step": 38680
    },
    {
      "entropy": 5.636133813858033,
      "epoch": 4.344695906104341,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032540489112684804,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2343456819653511,
      "num_tokens": 86355718.0,
      "step": 38685
    },
    {
      "entropy": 5.74070143699646,
      "epoch": 4.345257483012299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032536576201949724,
      "loss": 5.0208,
      "mean_token_accuracy": 0.22273913025856018,
      "num_tokens": 86367174.0,
      "step": 38690
    },
    {
      "entropy": 5.702701044082642,
      "epoch": 4.345819059920256,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003253266313084811,
      "loss": 4.981,
      "mean_token_accuracy": 0.22775050848722458,
      "num_tokens": 86379086.0,
      "step": 38695
    },
    {
      "entropy": 5.735109472274781,
      "epoch": 4.346380636828213,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003252874989950454,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2309210181236267,
      "num_tokens": 86390113.0,
      "step": 38700
    },
    {
      "entropy": 5.7686464309692385,
      "epoch": 4.346942213736171,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032524836508043625,
      "loss": 5.044,
      "mean_token_accuracy": 0.2209057867527008,
      "num_tokens": 86400824.0,
      "step": 38705
    },
    {
      "entropy": 5.653359603881836,
      "epoch": 4.3475037906441285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032520922956589973,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2310551032423973,
      "num_tokens": 86412972.0,
      "step": 38710
    },
    {
      "entropy": 5.676255893707276,
      "epoch": 4.348065367552087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032517009245268175,
      "loss": 4.8925,
      "mean_token_accuracy": 0.23043349385261536,
      "num_tokens": 86425332.0,
      "step": 38715
    },
    {
      "entropy": 5.654563903808594,
      "epoch": 4.348626944460044,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032513095374202863,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23821083456277847,
      "num_tokens": 86435615.0,
      "step": 38720
    },
    {
      "entropy": 5.752492189407349,
      "epoch": 4.349188521368001,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032509181343518646,
      "loss": 5.0405,
      "mean_token_accuracy": 0.2320321097970009,
      "num_tokens": 86446875.0,
      "step": 38725
    },
    {
      "entropy": 5.688903093338013,
      "epoch": 4.349750098275959,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032505267153340163,
      "loss": 4.925,
      "mean_token_accuracy": 0.23303023278713225,
      "num_tokens": 86458284.0,
      "step": 38730
    },
    {
      "entropy": 5.758785200119019,
      "epoch": 4.350311675183916,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003250135280379202,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21574851721525193,
      "num_tokens": 86470977.0,
      "step": 38735
    },
    {
      "entropy": 5.703608131408691,
      "epoch": 4.350873252091874,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003249743829499888,
      "loss": 5.0081,
      "mean_token_accuracy": 0.22018079310655594,
      "num_tokens": 86482578.0,
      "step": 38740
    },
    {
      "entropy": 5.681200170516968,
      "epoch": 4.351434828999832,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003249352362708535,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23292198181152343,
      "num_tokens": 86493123.0,
      "step": 38745
    },
    {
      "entropy": 5.764856052398682,
      "epoch": 4.351996405907789,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000324896088001761,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2213633745908737,
      "num_tokens": 86503865.0,
      "step": 38750
    },
    {
      "entropy": 5.63877272605896,
      "epoch": 4.352557982815747,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003248569381439577,
      "loss": 4.9263,
      "mean_token_accuracy": 0.2303095743060112,
      "num_tokens": 86515006.0,
      "step": 38755
    },
    {
      "entropy": 5.658332014083863,
      "epoch": 4.353119559723704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003248177866986901,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2288641095161438,
      "num_tokens": 86526256.0,
      "step": 38760
    },
    {
      "entropy": 5.677366399765015,
      "epoch": 4.3536811366316615,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003247786336672049,
      "loss": 4.8603,
      "mean_token_accuracy": 0.24018530398607255,
      "num_tokens": 86537050.0,
      "step": 38765
    },
    {
      "entropy": 5.657522916793823,
      "epoch": 4.35424271353962,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032473947905074866,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23521381467580796,
      "num_tokens": 86547678.0,
      "step": 38770
    },
    {
      "entropy": 5.632687520980835,
      "epoch": 4.354804290447577,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032470032285056814,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23836649507284163,
      "num_tokens": 86558747.0,
      "step": 38775
    },
    {
      "entropy": 5.604736042022705,
      "epoch": 4.355365867355534,
      "grad_norm": 1.25,
      "learning_rate": 0.00032466116506791004,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22479851096868514,
      "num_tokens": 86569500.0,
      "step": 38780
    },
    {
      "entropy": 5.5928795337677,
      "epoch": 4.355927444263492,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032462200570402124,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22480981051921844,
      "num_tokens": 86581312.0,
      "step": 38785
    },
    {
      "entropy": 5.675066184997559,
      "epoch": 4.356489021171449,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003245828447601484,
      "loss": 4.9826,
      "mean_token_accuracy": 0.23216583281755448,
      "num_tokens": 86592526.0,
      "step": 38790
    },
    {
      "entropy": 5.651327466964721,
      "epoch": 4.357050598079407,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003245436822375387,
      "loss": 4.9746,
      "mean_token_accuracy": 0.23375888019800187,
      "num_tokens": 86604284.0,
      "step": 38795
    },
    {
      "entropy": 5.607264614105224,
      "epoch": 4.357612174987365,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003245045181374389,
      "loss": 4.8982,
      "mean_token_accuracy": 0.23223114162683486,
      "num_tokens": 86615567.0,
      "step": 38800
    },
    {
      "entropy": 5.599215173721314,
      "epoch": 4.358173751895322,
      "grad_norm": 1.21875,
      "learning_rate": 0.000324465352461096,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2402375504374504,
      "num_tokens": 86626188.0,
      "step": 38805
    },
    {
      "entropy": 5.756501197814941,
      "epoch": 4.35873532880328,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003244261852097572,
      "loss": 4.9935,
      "mean_token_accuracy": 0.22771529257297515,
      "num_tokens": 86638468.0,
      "step": 38810
    },
    {
      "entropy": 5.737773847579956,
      "epoch": 4.359296905711237,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032438701638466943,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2265259489417076,
      "num_tokens": 86649119.0,
      "step": 38815
    },
    {
      "entropy": 5.654578638076782,
      "epoch": 4.359858482619194,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032434784598707987,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2303483322262764,
      "num_tokens": 86660924.0,
      "step": 38820
    },
    {
      "entropy": 5.634189081192017,
      "epoch": 4.3604200595271525,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003243086740182358,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2350383296608925,
      "num_tokens": 86672655.0,
      "step": 38825
    },
    {
      "entropy": 5.653815126419067,
      "epoch": 4.36098163643511,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032426950047938436,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23878270238637925,
      "num_tokens": 86684146.0,
      "step": 38830
    },
    {
      "entropy": 5.779592752456665,
      "epoch": 4.361543213343067,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000324230325371773,
      "loss": 5.0938,
      "mean_token_accuracy": 0.21952737420797347,
      "num_tokens": 86695435.0,
      "step": 38835
    },
    {
      "entropy": 5.820628976821899,
      "epoch": 4.362104790251025,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003241911486966489,
      "loss": 5.1042,
      "mean_token_accuracy": 0.21843956857919694,
      "num_tokens": 86706387.0,
      "step": 38840
    },
    {
      "entropy": 5.664371538162231,
      "epoch": 4.362666367158982,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003241519704552596,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2315179005265236,
      "num_tokens": 86717661.0,
      "step": 38845
    },
    {
      "entropy": 5.678868532180786,
      "epoch": 4.36322794406694,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032411279064885253,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22744768410921096,
      "num_tokens": 86729045.0,
      "step": 38850
    },
    {
      "entropy": 5.702991962432861,
      "epoch": 4.363789520974898,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032407360927867506,
      "loss": 5.0682,
      "mean_token_accuracy": 0.22436111569404601,
      "num_tokens": 86740763.0,
      "step": 38855
    },
    {
      "entropy": 5.687169122695923,
      "epoch": 4.364351097882855,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003240344263459749,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22892104387283324,
      "num_tokens": 86752369.0,
      "step": 38860
    },
    {
      "entropy": 5.69080376625061,
      "epoch": 4.364912674790813,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032399524185199957,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22147663980722426,
      "num_tokens": 86765947.0,
      "step": 38865
    },
    {
      "entropy": 5.63343300819397,
      "epoch": 4.36547425169877,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032395605579799673,
      "loss": 4.8155,
      "mean_token_accuracy": 0.23598061352968216,
      "num_tokens": 86777202.0,
      "step": 38870
    },
    {
      "entropy": 5.699837636947632,
      "epoch": 4.366035828606727,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032391686818521414,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22664562612771988,
      "num_tokens": 86787838.0,
      "step": 38875
    },
    {
      "entropy": 5.751091194152832,
      "epoch": 4.3665974055146854,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032387767901489946,
      "loss": 5.0566,
      "mean_token_accuracy": 0.2245570734143257,
      "num_tokens": 86799875.0,
      "step": 38880
    },
    {
      "entropy": 5.704945182800293,
      "epoch": 4.367158982422643,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032383848828830054,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22716717123985292,
      "num_tokens": 86811018.0,
      "step": 38885
    },
    {
      "entropy": 5.647076272964478,
      "epoch": 4.3677205593306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032379929600666525,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22809570729732515,
      "num_tokens": 86821556.0,
      "step": 38890
    },
    {
      "entropy": 5.585342788696289,
      "epoch": 4.368282136238558,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032376010217124133,
      "loss": 4.7981,
      "mean_token_accuracy": 0.24172875434160232,
      "num_tokens": 86832041.0,
      "step": 38895
    },
    {
      "entropy": 5.6888045310974125,
      "epoch": 4.368843713146515,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032372090678327694,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23245298117399216,
      "num_tokens": 86844723.0,
      "step": 38900
    },
    {
      "entropy": 5.725958824157715,
      "epoch": 4.369405290054473,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032368170984401997,
      "loss": 4.96,
      "mean_token_accuracy": 0.2293907880783081,
      "num_tokens": 86856211.0,
      "step": 38905
    },
    {
      "entropy": 5.6863432884216305,
      "epoch": 4.3699668669624305,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032364251135471846,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2199055179953575,
      "num_tokens": 86867701.0,
      "step": 38910
    },
    {
      "entropy": 5.748887252807617,
      "epoch": 4.370528443870388,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003236033113166205,
      "loss": 5.0633,
      "mean_token_accuracy": 0.22405246049165725,
      "num_tokens": 86879838.0,
      "step": 38915
    },
    {
      "entropy": 5.705094909667968,
      "epoch": 4.371090020778346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032356410973097423,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2292623296380043,
      "num_tokens": 86891101.0,
      "step": 38920
    },
    {
      "entropy": 5.694620990753174,
      "epoch": 4.371651597686303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032352490659902795,
      "loss": 4.9224,
      "mean_token_accuracy": 0.23563063889741898,
      "num_tokens": 86902796.0,
      "step": 38925
    },
    {
      "entropy": 5.667128944396973,
      "epoch": 4.37221317459426,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003234857019220298,
      "loss": 4.981,
      "mean_token_accuracy": 0.224787075817585,
      "num_tokens": 86914381.0,
      "step": 38930
    },
    {
      "entropy": 5.751850986480713,
      "epoch": 4.372774751502218,
      "grad_norm": 1.171875,
      "learning_rate": 0.000323446495701228,
      "loss": 5.0912,
      "mean_token_accuracy": 0.22403351068496705,
      "num_tokens": 86926082.0,
      "step": 38935
    },
    {
      "entropy": 5.718506002426148,
      "epoch": 4.373336328410176,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003234072879378711,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2349979281425476,
      "num_tokens": 86936644.0,
      "step": 38940
    },
    {
      "entropy": 5.700245714187622,
      "epoch": 4.373897905318134,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003233680786332074,
      "loss": 4.9729,
      "mean_token_accuracy": 0.22959824055433273,
      "num_tokens": 86947586.0,
      "step": 38945
    },
    {
      "entropy": 5.710105133056641,
      "epoch": 4.374459482226091,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003233288677884853,
      "loss": 4.9362,
      "mean_token_accuracy": 0.23597662895917892,
      "num_tokens": 86958265.0,
      "step": 38950
    },
    {
      "entropy": 5.6991918087005615,
      "epoch": 4.375021059134048,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003232896554049533,
      "loss": 4.9246,
      "mean_token_accuracy": 0.23253436535596847,
      "num_tokens": 86969119.0,
      "step": 38955
    },
    {
      "entropy": 5.6428306102752686,
      "epoch": 4.375582636042006,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032325044148385986,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22309117317199706,
      "num_tokens": 86980232.0,
      "step": 38960
    },
    {
      "entropy": 5.635681629180908,
      "epoch": 4.3761442129499635,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032321122602645374,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2350428134202957,
      "num_tokens": 86993590.0,
      "step": 38965
    },
    {
      "entropy": 5.698475122451782,
      "epoch": 4.376705789857921,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003231720090339834,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2262442946434021,
      "num_tokens": 87004905.0,
      "step": 38970
    },
    {
      "entropy": 5.645428466796875,
      "epoch": 4.377267366765879,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032313279050769767,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23251426666975022,
      "num_tokens": 87016920.0,
      "step": 38975
    },
    {
      "entropy": 5.774557113647461,
      "epoch": 4.377828943673836,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003230935704488452,
      "loss": 5.0658,
      "mean_token_accuracy": 0.22259241640567778,
      "num_tokens": 87027740.0,
      "step": 38980
    },
    {
      "entropy": 5.743799829483033,
      "epoch": 4.378390520581794,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032305434885867483,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22580677270889282,
      "num_tokens": 87039742.0,
      "step": 38985
    },
    {
      "entropy": 5.626197099685669,
      "epoch": 4.378952097489751,
      "grad_norm": 1.25,
      "learning_rate": 0.00032301512573843535,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22212761789560317,
      "num_tokens": 87049750.0,
      "step": 38990
    },
    {
      "entropy": 5.686049699783325,
      "epoch": 4.3795136743977086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003229759010893756,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2399911478161812,
      "num_tokens": 87062890.0,
      "step": 38995
    },
    {
      "entropy": 5.707054615020752,
      "epoch": 4.380075251305667,
      "grad_norm": 1.125,
      "learning_rate": 0.0003229366749127446,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22990194112062454,
      "num_tokens": 87074450.0,
      "step": 39000
    },
    {
      "epoch": 4.380075251305667,
      "eval_entropy": 5.4586668978777935,
      "eval_loss": 5.060207843780518,
      "eval_mean_token_accuracy": 0.23139638186215986,
      "eval_num_tokens": 87074450.0,
      "eval_runtime": 18.4367,
      "eval_samples_per_second": 1682.139,
      "eval_steps_per_second": 210.288,
      "step": 39000
    },
    {
      "entropy": 5.814043664932251,
      "epoch": 4.380636828213624,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003228974472097913,
      "loss": 5.0516,
      "mean_token_accuracy": 0.2280524864792824,
      "num_tokens": 87085931.0,
      "step": 39005
    },
    {
      "entropy": 5.692433214187622,
      "epoch": 4.381198405121581,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032285821798176456,
      "loss": 5.0105,
      "mean_token_accuracy": 0.22737183421850204,
      "num_tokens": 87096655.0,
      "step": 39010
    },
    {
      "entropy": 5.719551467895508,
      "epoch": 4.381759982029539,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032281898722991376,
      "loss": 5.0293,
      "mean_token_accuracy": 0.22062167525291443,
      "num_tokens": 87108359.0,
      "step": 39015
    },
    {
      "entropy": 5.761613082885742,
      "epoch": 4.382321558937496,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032277975495548775,
      "loss": 5.0521,
      "mean_token_accuracy": 0.22338799089193345,
      "num_tokens": 87119028.0,
      "step": 39020
    },
    {
      "entropy": 5.631937265396118,
      "epoch": 4.382883135845454,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003227405211597359,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22984553575515748,
      "num_tokens": 87130574.0,
      "step": 39025
    },
    {
      "entropy": 5.691033792495728,
      "epoch": 4.383444712753412,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032270128584390735,
      "loss": 4.9109,
      "mean_token_accuracy": 0.231011700630188,
      "num_tokens": 87141052.0,
      "step": 39030
    },
    {
      "entropy": 5.707305860519409,
      "epoch": 4.384006289661369,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003226620490092513,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22511595636606216,
      "num_tokens": 87151323.0,
      "step": 39035
    },
    {
      "entropy": 5.664523124694824,
      "epoch": 4.384567866569327,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032262281065701716,
      "loss": 4.8762,
      "mean_token_accuracy": 0.22541665881872178,
      "num_tokens": 87162614.0,
      "step": 39040
    },
    {
      "entropy": 5.660956192016601,
      "epoch": 4.385129443477284,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003225835707884542,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2316904917359352,
      "num_tokens": 87172712.0,
      "step": 39045
    },
    {
      "entropy": 5.709676694869995,
      "epoch": 4.3856910203852415,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003225443294048119,
      "loss": 4.9347,
      "mean_token_accuracy": 0.23267614841461182,
      "num_tokens": 87183612.0,
      "step": 39050
    },
    {
      "entropy": 5.6930145740509035,
      "epoch": 4.3862525972932,
      "grad_norm": 1.375,
      "learning_rate": 0.00032250508650733976,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22667008489370347,
      "num_tokens": 87194707.0,
      "step": 39055
    },
    {
      "entropy": 5.605948877334595,
      "epoch": 4.386814174201157,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003224658420972871,
      "loss": 4.852,
      "mean_token_accuracy": 0.23601392805576324,
      "num_tokens": 87206451.0,
      "step": 39060
    },
    {
      "entropy": 5.6820313930511475,
      "epoch": 4.387375751109114,
      "grad_norm": 1.125,
      "learning_rate": 0.0003224265961759038,
      "loss": 5.043,
      "mean_token_accuracy": 0.22390164136886598,
      "num_tokens": 87217542.0,
      "step": 39065
    },
    {
      "entropy": 5.619008636474609,
      "epoch": 4.387937328017072,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003223873487444392,
      "loss": 4.869,
      "mean_token_accuracy": 0.2363324597477913,
      "num_tokens": 87228399.0,
      "step": 39070
    },
    {
      "entropy": 5.705646705627442,
      "epoch": 4.388498904925029,
      "grad_norm": 1.125,
      "learning_rate": 0.0003223480998041431,
      "loss": 4.9754,
      "mean_token_accuracy": 0.22823473066091537,
      "num_tokens": 87239875.0,
      "step": 39075
    },
    {
      "entropy": 5.694931936264038,
      "epoch": 4.389060481832987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032230884935626505,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2273623153567314,
      "num_tokens": 87252919.0,
      "step": 39080
    },
    {
      "entropy": 5.62820029258728,
      "epoch": 4.389622058740945,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000322269597402055,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22877780199050904,
      "num_tokens": 87264458.0,
      "step": 39085
    },
    {
      "entropy": 5.733487272262574,
      "epoch": 4.390183635648902,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003222303439427625,
      "loss": 4.9655,
      "mean_token_accuracy": 0.23201905339956283,
      "num_tokens": 87274935.0,
      "step": 39090
    },
    {
      "entropy": 5.625140810012818,
      "epoch": 4.39074521255686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032219108897963767,
      "loss": 4.895,
      "mean_token_accuracy": 0.2338988646864891,
      "num_tokens": 87285870.0,
      "step": 39095
    },
    {
      "entropy": 5.645923757553101,
      "epoch": 4.391306789464817,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003221518325139302,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23062293827533722,
      "num_tokens": 87297374.0,
      "step": 39100
    },
    {
      "entropy": 5.707010221481323,
      "epoch": 4.3918683663727744,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003221125745468901,
      "loss": 4.9186,
      "mean_token_accuracy": 0.23262429535388945,
      "num_tokens": 87307597.0,
      "step": 39105
    },
    {
      "entropy": 5.622785043716431,
      "epoch": 4.3924299432807326,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003220733150797673,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23282802999019622,
      "num_tokens": 87318459.0,
      "step": 39110
    },
    {
      "entropy": 5.739413452148438,
      "epoch": 4.39299152018869,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003220340541138119,
      "loss": 5.0903,
      "mean_token_accuracy": 0.22029670029878617,
      "num_tokens": 87330613.0,
      "step": 39115
    },
    {
      "entropy": 5.730040788650513,
      "epoch": 4.393553097096647,
      "grad_norm": 1.125,
      "learning_rate": 0.00032199479165027407,
      "loss": 5.0092,
      "mean_token_accuracy": 0.22499072402715684,
      "num_tokens": 87341988.0,
      "step": 39120
    },
    {
      "entropy": 5.685257387161255,
      "epoch": 4.394114674004605,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032195552769040376,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22772396802902223,
      "num_tokens": 87353532.0,
      "step": 39125
    },
    {
      "entropy": 5.697174549102783,
      "epoch": 4.394676250912562,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003219162622354513,
      "loss": 4.9856,
      "mean_token_accuracy": 0.23303370624780656,
      "num_tokens": 87364754.0,
      "step": 39130
    },
    {
      "entropy": 5.6241569995880125,
      "epoch": 4.39523782782052,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032187699528666674,
      "loss": 4.9268,
      "mean_token_accuracy": 0.22265757024288177,
      "num_tokens": 87375983.0,
      "step": 39135
    },
    {
      "entropy": 5.673055028915405,
      "epoch": 4.395799404728478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003218377268453006,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23984856009483338,
      "num_tokens": 87388169.0,
      "step": 39140
    },
    {
      "entropy": 5.776824951171875,
      "epoch": 4.396360981636435,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000321798456912603,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2319347456097603,
      "num_tokens": 87399341.0,
      "step": 39145
    },
    {
      "entropy": 5.676765871047974,
      "epoch": 4.396922558544393,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032175918548982437,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23045914024114608,
      "num_tokens": 87410941.0,
      "step": 39150
    },
    {
      "entropy": 5.641001796722412,
      "epoch": 4.39748413545235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032171991257821513,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23157504945993423,
      "num_tokens": 87421833.0,
      "step": 39155
    },
    {
      "entropy": 5.584724569320679,
      "epoch": 4.398045712360307,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003216806381790258,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23723750859498977,
      "num_tokens": 87431648.0,
      "step": 39160
    },
    {
      "entropy": 5.7024383544921875,
      "epoch": 4.3986072892682655,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003216413622935069,
      "loss": 5.017,
      "mean_token_accuracy": 0.2232268139719963,
      "num_tokens": 87444207.0,
      "step": 39165
    },
    {
      "entropy": 5.6532727718353275,
      "epoch": 4.399168866176223,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032160208492290886,
      "loss": 4.8589,
      "mean_token_accuracy": 0.23989565074443817,
      "num_tokens": 87455919.0,
      "step": 39170
    },
    {
      "entropy": 5.679591512680053,
      "epoch": 4.399730443084181,
      "grad_norm": 1.125,
      "learning_rate": 0.00032156280606848237,
      "loss": 4.895,
      "mean_token_accuracy": 0.23183702379465104,
      "num_tokens": 87466979.0,
      "step": 39175
    },
    {
      "entropy": 5.693000411987304,
      "epoch": 4.400292019992138,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003215235257314781,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2263645812869072,
      "num_tokens": 87478193.0,
      "step": 39180
    },
    {
      "entropy": 5.642584371566772,
      "epoch": 4.400853596900095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032148424391314665,
      "loss": 4.94,
      "mean_token_accuracy": 0.22985505759716035,
      "num_tokens": 87488961.0,
      "step": 39185
    },
    {
      "entropy": 5.739750003814697,
      "epoch": 4.401415173808053,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032144496061473895,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22632162123918534,
      "num_tokens": 87500508.0,
      "step": 39190
    },
    {
      "entropy": 5.692247676849365,
      "epoch": 4.401976750716011,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003214056758375056,
      "loss": 4.9066,
      "mean_token_accuracy": 0.22860590517520904,
      "num_tokens": 87510755.0,
      "step": 39195
    },
    {
      "entropy": 5.683482027053833,
      "epoch": 4.402538327623968,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003213663895826976,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2292216032743454,
      "num_tokens": 87522425.0,
      "step": 39200
    },
    {
      "entropy": 5.709164762496949,
      "epoch": 4.403099904531926,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003213271018515657,
      "loss": 4.9768,
      "mean_token_accuracy": 0.23536846786737442,
      "num_tokens": 87534209.0,
      "step": 39205
    },
    {
      "entropy": 5.606453227996826,
      "epoch": 4.403661481439883,
      "grad_norm": 1.15625,
      "learning_rate": 0.000321287812645361,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23038289844989776,
      "num_tokens": 87544647.0,
      "step": 39210
    },
    {
      "entropy": 5.6270647048950195,
      "epoch": 4.40422305834784,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032124852196533434,
      "loss": 4.8249,
      "mean_token_accuracy": 0.23826338946819306,
      "num_tokens": 87555924.0,
      "step": 39215
    },
    {
      "entropy": 5.626015043258667,
      "epoch": 4.404784635255798,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003212092298127368,
      "loss": 4.9126,
      "mean_token_accuracy": 0.23096441179513932,
      "num_tokens": 87566998.0,
      "step": 39220
    },
    {
      "entropy": 5.632910537719726,
      "epoch": 4.405346212163756,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032116993618881945,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22836151272058486,
      "num_tokens": 87579516.0,
      "step": 39225
    },
    {
      "entropy": 5.744158840179443,
      "epoch": 4.405907789071714,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032113064109483345,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22849342674016954,
      "num_tokens": 87591512.0,
      "step": 39230
    },
    {
      "entropy": 5.698963356018067,
      "epoch": 4.406469365979671,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032109134453202993,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22324783056974412,
      "num_tokens": 87602783.0,
      "step": 39235
    },
    {
      "entropy": 5.638091659545898,
      "epoch": 4.407030942887628,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032105204650166016,
      "loss": 4.906,
      "mean_token_accuracy": 0.23074611127376557,
      "num_tokens": 87613585.0,
      "step": 39240
    },
    {
      "entropy": 5.741014432907105,
      "epoch": 4.407592519795586,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032101274700497537,
      "loss": 4.9938,
      "mean_token_accuracy": 0.229786916077137,
      "num_tokens": 87624003.0,
      "step": 39245
    },
    {
      "entropy": 5.651638078689575,
      "epoch": 4.4081540967035435,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003209734460432269,
      "loss": 4.8966,
      "mean_token_accuracy": 0.2335831791162491,
      "num_tokens": 87634429.0,
      "step": 39250
    },
    {
      "entropy": 5.6240448474884035,
      "epoch": 4.408715673611501,
      "grad_norm": 1.15625,
      "learning_rate": 0.000320934143617666,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23306908309459687,
      "num_tokens": 87645887.0,
      "step": 39255
    },
    {
      "entropy": 5.7112181186676025,
      "epoch": 4.409277250519459,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003208948397295441,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22427288442850113,
      "num_tokens": 87657439.0,
      "step": 39260
    },
    {
      "entropy": 5.6224456310272215,
      "epoch": 4.409838827427416,
      "grad_norm": 1.25,
      "learning_rate": 0.00032085553438011285,
      "loss": 4.8927,
      "mean_token_accuracy": 0.22835176289081574,
      "num_tokens": 87668461.0,
      "step": 39265
    },
    {
      "entropy": 5.783779573440552,
      "epoch": 4.410400404335373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003208162275706235,
      "loss": 5.0049,
      "mean_token_accuracy": 0.223761123418808,
      "num_tokens": 87680595.0,
      "step": 39270
    },
    {
      "entropy": 5.649344968795776,
      "epoch": 4.410961981243331,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003207769193023277,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23343484997749328,
      "num_tokens": 87691563.0,
      "step": 39275
    },
    {
      "entropy": 5.728486585617065,
      "epoch": 4.411523558151289,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032073760957647705,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22785910218954086,
      "num_tokens": 87702052.0,
      "step": 39280
    },
    {
      "entropy": 5.630448579788208,
      "epoch": 4.412085135059247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003206982983943232,
      "loss": 4.937,
      "mean_token_accuracy": 0.22554800510406495,
      "num_tokens": 87713123.0,
      "step": 39285
    },
    {
      "entropy": 5.724777030944824,
      "epoch": 4.412646711967204,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003206589857571178,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21751996874809265,
      "num_tokens": 87723294.0,
      "step": 39290
    },
    {
      "entropy": 5.688076591491699,
      "epoch": 4.413208288875161,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00032061967166611255,
      "loss": 4.9256,
      "mean_token_accuracy": 0.23547120839357377,
      "num_tokens": 87733751.0,
      "step": 39295
    },
    {
      "entropy": 5.678538846969604,
      "epoch": 4.413769865783119,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00032058035612255934,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2392591193318367,
      "num_tokens": 87744584.0,
      "step": 39300
    },
    {
      "entropy": 5.689496326446533,
      "epoch": 4.4143314426910765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003205410391277098,
      "loss": 4.9373,
      "mean_token_accuracy": 0.235323004424572,
      "num_tokens": 87755420.0,
      "step": 39305
    },
    {
      "entropy": 5.718980598449707,
      "epoch": 4.414893019599034,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032050172068281594,
      "loss": 4.9604,
      "mean_token_accuracy": 0.23426315933465958,
      "num_tokens": 87767213.0,
      "step": 39310
    },
    {
      "entropy": 5.696855783462524,
      "epoch": 4.415454596506992,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003204624007891297,
      "loss": 4.9626,
      "mean_token_accuracy": 0.23009886592626572,
      "num_tokens": 87778058.0,
      "step": 39315
    },
    {
      "entropy": 5.693119144439697,
      "epoch": 4.416016173414949,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000320423079447903,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22794200628995895,
      "num_tokens": 87788279.0,
      "step": 39320
    },
    {
      "entropy": 5.667172956466675,
      "epoch": 4.416577750322907,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032038375666038776,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22907678931951522,
      "num_tokens": 87799643.0,
      "step": 39325
    },
    {
      "entropy": 5.647969007492065,
      "epoch": 4.417139327230864,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003203444324278361,
      "loss": 4.878,
      "mean_token_accuracy": 0.23537316024303437,
      "num_tokens": 87810493.0,
      "step": 39330
    },
    {
      "entropy": 5.66726508140564,
      "epoch": 4.4177009041388215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003203051067515002,
      "loss": 4.961,
      "mean_token_accuracy": 0.22590069025754927,
      "num_tokens": 87821876.0,
      "step": 39335
    },
    {
      "entropy": 5.670257663726806,
      "epoch": 4.41826248104678,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003202657796326321,
      "loss": 4.882,
      "mean_token_accuracy": 0.23841331899166107,
      "num_tokens": 87832058.0,
      "step": 39340
    },
    {
      "entropy": 5.696881246566773,
      "epoch": 4.418824057954737,
      "grad_norm": 1.15625,
      "learning_rate": 0.000320226451072484,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22838514149188996,
      "num_tokens": 87842549.0,
      "step": 39345
    },
    {
      "entropy": 5.608184957504273,
      "epoch": 4.419385634862694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032018712107230815,
      "loss": 4.9123,
      "mean_token_accuracy": 0.23131219744682313,
      "num_tokens": 87853507.0,
      "step": 39350
    },
    {
      "entropy": 5.7029376983642575,
      "epoch": 4.419947211770652,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003201477896333568,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22939281165599823,
      "num_tokens": 87864546.0,
      "step": 39355
    },
    {
      "entropy": 5.742338752746582,
      "epoch": 4.420508788678609,
      "grad_norm": 1.25,
      "learning_rate": 0.0003201084567568824,
      "loss": 4.897,
      "mean_token_accuracy": 0.23961483538150788,
      "num_tokens": 87876553.0,
      "step": 39360
    },
    {
      "entropy": 5.660269069671631,
      "epoch": 4.4210703655865675,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032006912244413727,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23082153648138046,
      "num_tokens": 87886257.0,
      "step": 39365
    },
    {
      "entropy": 5.642222785949707,
      "epoch": 4.421631942494525,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003200297866963737,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23662676960229873,
      "num_tokens": 87897704.0,
      "step": 39370
    },
    {
      "entropy": 5.634072780609131,
      "epoch": 4.422193519402482,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003199904495148443,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2377694994211197,
      "num_tokens": 87908892.0,
      "step": 39375
    },
    {
      "entropy": 5.6936821937561035,
      "epoch": 4.42275509631044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003199511109008016,
      "loss": 4.9858,
      "mean_token_accuracy": 0.23617703318595887,
      "num_tokens": 87921502.0,
      "step": 39380
    },
    {
      "entropy": 5.711408519744873,
      "epoch": 4.423316673218397,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031991177085549807,
      "loss": 4.9537,
      "mean_token_accuracy": 0.23413514196872712,
      "num_tokens": 87933250.0,
      "step": 39385
    },
    {
      "entropy": 5.64495849609375,
      "epoch": 4.4238782501263545,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031987242938018626,
      "loss": 4.8572,
      "mean_token_accuracy": 0.24054754078388213,
      "num_tokens": 87945715.0,
      "step": 39390
    },
    {
      "entropy": 5.611583471298218,
      "epoch": 4.424439827034313,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031983308647611896,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2446956068277359,
      "num_tokens": 87956973.0,
      "step": 39395
    },
    {
      "entropy": 5.7289392948150635,
      "epoch": 4.42500140394227,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003197937421445488,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2291952833533287,
      "num_tokens": 87968341.0,
      "step": 39400
    },
    {
      "entropy": 5.637761497497559,
      "epoch": 4.425562980850227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003197543963867285,
      "loss": 4.9578,
      "mean_token_accuracy": 0.23047932833433152,
      "num_tokens": 87978980.0,
      "step": 39405
    },
    {
      "entropy": 5.712554311752319,
      "epoch": 4.426124557758185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003197150492039108,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23704388737678528,
      "num_tokens": 87990612.0,
      "step": 39410
    },
    {
      "entropy": 5.717540550231933,
      "epoch": 4.426686134666142,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003196757005973487,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22903133034706116,
      "num_tokens": 88001052.0,
      "step": 39415
    },
    {
      "entropy": 5.653577041625977,
      "epoch": 4.4272477115741005,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031963635056829487,
      "loss": 4.9544,
      "mean_token_accuracy": 0.2288833811879158,
      "num_tokens": 88013102.0,
      "step": 39420
    },
    {
      "entropy": 5.704487752914429,
      "epoch": 4.427809288482058,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031959699911800237,
      "loss": 5.0192,
      "mean_token_accuracy": 0.23243589997291564,
      "num_tokens": 88023952.0,
      "step": 39425
    },
    {
      "entropy": 5.682995271682739,
      "epoch": 4.428370865390015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003195576462477241,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22937270551919936,
      "num_tokens": 88035148.0,
      "step": 39430
    },
    {
      "entropy": 5.695062780380249,
      "epoch": 4.428932442297973,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031951829195871315,
      "loss": 5.0003,
      "mean_token_accuracy": 0.22678448855876923,
      "num_tokens": 88046294.0,
      "step": 39435
    },
    {
      "entropy": 5.662312078475952,
      "epoch": 4.42949401920593,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031947893625222244,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22891543060541153,
      "num_tokens": 88058225.0,
      "step": 39440
    },
    {
      "entropy": 5.721632719039917,
      "epoch": 4.430055596113887,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031943957912950523,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22299874275922776,
      "num_tokens": 88069994.0,
      "step": 39445
    },
    {
      "entropy": 5.751102924346924,
      "epoch": 4.4306171730218455,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003194002205918145,
      "loss": 4.9464,
      "mean_token_accuracy": 0.23293985575437545,
      "num_tokens": 88081912.0,
      "step": 39450
    },
    {
      "entropy": 5.677507305145264,
      "epoch": 4.431178749929803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003193608606404036,
      "loss": 4.9279,
      "mean_token_accuracy": 0.23016296476125717,
      "num_tokens": 88092457.0,
      "step": 39455
    },
    {
      "entropy": 5.750240325927734,
      "epoch": 4.43174032683776,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003193214992765257,
      "loss": 5.0892,
      "mean_token_accuracy": 0.21522789001464843,
      "num_tokens": 88104407.0,
      "step": 39460
    },
    {
      "entropy": 5.650768280029297,
      "epoch": 4.432301903745718,
      "grad_norm": 1.25,
      "learning_rate": 0.00031928213650143407,
      "loss": 4.8838,
      "mean_token_accuracy": 0.23669250905513764,
      "num_tokens": 88114653.0,
      "step": 39465
    },
    {
      "entropy": 5.74715027809143,
      "epoch": 4.432863480653675,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000319242772316382,
      "loss": 4.9323,
      "mean_token_accuracy": 0.23921364843845366,
      "num_tokens": 88125477.0,
      "step": 39470
    },
    {
      "entropy": 5.641115331649781,
      "epoch": 4.433425057561633,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003192034067226228,
      "loss": 4.8429,
      "mean_token_accuracy": 0.23939171731472014,
      "num_tokens": 88136419.0,
      "step": 39475
    },
    {
      "entropy": 5.663536787033081,
      "epoch": 4.433986634469591,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00031916403972141017,
      "loss": 4.8643,
      "mean_token_accuracy": 0.23392075896263123,
      "num_tokens": 88146578.0,
      "step": 39480
    },
    {
      "entropy": 5.526888418197632,
      "epoch": 4.434548211377548,
      "grad_norm": 1.25,
      "learning_rate": 0.0003191246713139973,
      "loss": 4.7874,
      "mean_token_accuracy": 0.24074302166700362,
      "num_tokens": 88157351.0,
      "step": 39485
    },
    {
      "entropy": 5.705948734283448,
      "epoch": 4.435109788285506,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031908530150163777,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22990392297506332,
      "num_tokens": 88168460.0,
      "step": 39490
    },
    {
      "entropy": 5.7534764289855955,
      "epoch": 4.435671365193463,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031904593028558517,
      "loss": 4.964,
      "mean_token_accuracy": 0.2285271927714348,
      "num_tokens": 88178994.0,
      "step": 39495
    },
    {
      "entropy": 5.6805651664733885,
      "epoch": 4.43623294210142,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031900655766709305,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22908239662647248,
      "num_tokens": 88188916.0,
      "step": 39500
    },
    {
      "entropy": 5.6126447200775145,
      "epoch": 4.4367945190093785,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003189671836474151,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23188057243824006,
      "num_tokens": 88200833.0,
      "step": 39505
    },
    {
      "entropy": 5.70770583152771,
      "epoch": 4.437356095917336,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003189278082278049,
      "loss": 5.0366,
      "mean_token_accuracy": 0.22338384836912156,
      "num_tokens": 88212022.0,
      "step": 39510
    },
    {
      "entropy": 5.675034427642823,
      "epoch": 4.437917672825294,
      "grad_norm": 1.125,
      "learning_rate": 0.00031888843140951627,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23439015746116637,
      "num_tokens": 88222327.0,
      "step": 39515
    },
    {
      "entropy": 5.557827043533325,
      "epoch": 4.438479249733251,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003188490531938031,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2396608054637909,
      "num_tokens": 88233141.0,
      "step": 39520
    },
    {
      "entropy": 5.751549577713012,
      "epoch": 4.439040826641208,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031880967358191884,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21807009875774383,
      "num_tokens": 88245663.0,
      "step": 39525
    },
    {
      "entropy": 5.721306467056275,
      "epoch": 4.439602403549166,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031877029257511774,
      "loss": 4.9217,
      "mean_token_accuracy": 0.23083702027797698,
      "num_tokens": 88256291.0,
      "step": 39530
    },
    {
      "entropy": 5.732830667495728,
      "epoch": 4.440163980457124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031873091017465356,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2253860667347908,
      "num_tokens": 88268382.0,
      "step": 39535
    },
    {
      "entropy": 5.596661615371704,
      "epoch": 4.440725557365081,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003186915263817802,
      "loss": 4.8881,
      "mean_token_accuracy": 0.24007754921913146,
      "num_tokens": 88279638.0,
      "step": 39540
    },
    {
      "entropy": 5.678821802139282,
      "epoch": 4.441287134273039,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031865214119775173,
      "loss": 4.9776,
      "mean_token_accuracy": 0.2289220318198204,
      "num_tokens": 88290599.0,
      "step": 39545
    },
    {
      "entropy": 5.683985948562622,
      "epoch": 4.441848711180996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000318612754623822,
      "loss": 4.9283,
      "mean_token_accuracy": 0.23304898142814637,
      "num_tokens": 88301572.0,
      "step": 39550
    },
    {
      "entropy": 5.739064168930054,
      "epoch": 4.442410288088954,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031857336666124537,
      "loss": 4.9974,
      "mean_token_accuracy": 0.23123618811368943,
      "num_tokens": 88312978.0,
      "step": 39555
    },
    {
      "entropy": 5.736243867874146,
      "epoch": 4.442971864996911,
      "grad_norm": 1.25,
      "learning_rate": 0.0003185339773112758,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22340075224637984,
      "num_tokens": 88324023.0,
      "step": 39560
    },
    {
      "entropy": 5.659014177322388,
      "epoch": 4.443533441904869,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003184945865751675,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22532659471035005,
      "num_tokens": 88335366.0,
      "step": 39565
    },
    {
      "entropy": 5.6710405349731445,
      "epoch": 4.444095018812827,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031845519445417475,
      "loss": 4.9313,
      "mean_token_accuracy": 0.23062134385108948,
      "num_tokens": 88347786.0,
      "step": 39570
    },
    {
      "entropy": 5.747459745407104,
      "epoch": 4.444656595720784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003184158009495518,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22707706242799758,
      "num_tokens": 88358720.0,
      "step": 39575
    },
    {
      "entropy": 5.710166645050049,
      "epoch": 4.445218172628741,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003183764060625528,
      "loss": 4.976,
      "mean_token_accuracy": 0.22873493880033494,
      "num_tokens": 88370152.0,
      "step": 39580
    },
    {
      "entropy": 5.641630029678344,
      "epoch": 4.445779749536699,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031833700979443227,
      "loss": 4.9237,
      "mean_token_accuracy": 0.23059644997119905,
      "num_tokens": 88381385.0,
      "step": 39585
    },
    {
      "entropy": 5.692480659484863,
      "epoch": 4.4463413264446565,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031829761214644453,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22936591356992722,
      "num_tokens": 88391450.0,
      "step": 39590
    },
    {
      "entropy": 5.5989879131317135,
      "epoch": 4.446902903352614,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000318258213119844,
      "loss": 4.8561,
      "mean_token_accuracy": 0.24486086815595626,
      "num_tokens": 88402310.0,
      "step": 39595
    },
    {
      "entropy": 5.639358043670654,
      "epoch": 4.447464480260572,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003182188127158852,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21734818816184998,
      "num_tokens": 88412286.0,
      "step": 39600
    },
    {
      "entropy": 5.741382026672364,
      "epoch": 4.448026057168529,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003181794109358227,
      "loss": 4.97,
      "mean_token_accuracy": 0.23327964842319487,
      "num_tokens": 88423335.0,
      "step": 39605
    },
    {
      "entropy": 5.621005296707153,
      "epoch": 4.448587634076487,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003181400077809109,
      "loss": 4.888,
      "mean_token_accuracy": 0.23268462121486663,
      "num_tokens": 88434328.0,
      "step": 39610
    },
    {
      "entropy": 5.700366258621216,
      "epoch": 4.449149210984444,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031810060325240466,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2281843528151512,
      "num_tokens": 88445611.0,
      "step": 39615
    },
    {
      "entropy": 5.731183242797852,
      "epoch": 4.449710787892402,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003180611973515583,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2289839655160904,
      "num_tokens": 88456207.0,
      "step": 39620
    },
    {
      "entropy": 5.789615345001221,
      "epoch": 4.45027236480036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031802179007962687,
      "loss": 5.0808,
      "mean_token_accuracy": 0.22107764929533005,
      "num_tokens": 88467322.0,
      "step": 39625
    },
    {
      "entropy": 5.661843585968017,
      "epoch": 4.450833941708317,
      "grad_norm": 1.125,
      "learning_rate": 0.0003179823814378649,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2281666249036789,
      "num_tokens": 88479703.0,
      "step": 39630
    },
    {
      "entropy": 5.616714334487915,
      "epoch": 4.451395518616274,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031794297142752727,
      "loss": 4.8835,
      "mean_token_accuracy": 0.23328225910663605,
      "num_tokens": 88490284.0,
      "step": 39635
    },
    {
      "entropy": 5.673579263687134,
      "epoch": 4.451957095524232,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003179035600498687,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22924579530954362,
      "num_tokens": 88502232.0,
      "step": 39640
    },
    {
      "entropy": 5.6420286178588865,
      "epoch": 4.4525186724321895,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031786414730614423,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2363818973302841,
      "num_tokens": 88513788.0,
      "step": 39645
    },
    {
      "entropy": 5.6890379905700685,
      "epoch": 4.453080249340147,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003178247331976086,
      "loss": 4.9592,
      "mean_token_accuracy": 0.22972284108400345,
      "num_tokens": 88524942.0,
      "step": 39650
    },
    {
      "entropy": 5.605289268493652,
      "epoch": 4.453641826248105,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031778531772551696,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23565078228712083,
      "num_tokens": 88535692.0,
      "step": 39655
    },
    {
      "entropy": 5.639699220657349,
      "epoch": 4.454203403156062,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003177459008911241,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23143129348754882,
      "num_tokens": 88547007.0,
      "step": 39660
    },
    {
      "entropy": 5.760267210006714,
      "epoch": 4.45476498006402,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031770648269568524,
      "loss": 4.9932,
      "mean_token_accuracy": 0.23790469616651536,
      "num_tokens": 88558197.0,
      "step": 39665
    },
    {
      "entropy": 5.770211458206177,
      "epoch": 4.455326556971977,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003176670631404554,
      "loss": 5.0155,
      "mean_token_accuracy": 0.22612251192331315,
      "num_tokens": 88569488.0,
      "step": 39670
    },
    {
      "entropy": 5.70452766418457,
      "epoch": 4.4558881338799345,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031762764222668976,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22606841921806337,
      "num_tokens": 88582054.0,
      "step": 39675
    },
    {
      "entropy": 5.691221523284912,
      "epoch": 4.456449710787893,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031758821995564343,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23406733572483063,
      "num_tokens": 88592528.0,
      "step": 39680
    },
    {
      "entropy": 5.69893798828125,
      "epoch": 4.45701128769585,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003175487963285716,
      "loss": 4.9404,
      "mean_token_accuracy": 0.23594107627868652,
      "num_tokens": 88604136.0,
      "step": 39685
    },
    {
      "entropy": 5.7476794719696045,
      "epoch": 4.457572864603807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003175093713467297,
      "loss": 4.996,
      "mean_token_accuracy": 0.22968141883611679,
      "num_tokens": 88615629.0,
      "step": 39690
    },
    {
      "entropy": 5.596204471588135,
      "epoch": 4.458134441511765,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031746994501137283,
      "loss": 4.8408,
      "mean_token_accuracy": 0.24025336056947708,
      "num_tokens": 88626100.0,
      "step": 39695
    },
    {
      "entropy": 5.653398656845093,
      "epoch": 4.458696018419722,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003174305173237565,
      "loss": 4.9932,
      "mean_token_accuracy": 0.22567276358604432,
      "num_tokens": 88638948.0,
      "step": 39700
    },
    {
      "entropy": 5.723152685165405,
      "epoch": 4.4592575953276805,
      "grad_norm": 1.328125,
      "learning_rate": 0.000317391088285136,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22927793711423874,
      "num_tokens": 88650488.0,
      "step": 39705
    },
    {
      "entropy": 5.78495683670044,
      "epoch": 4.459819172235638,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003173516578967668,
      "loss": 5.0256,
      "mean_token_accuracy": 0.22308346927165984,
      "num_tokens": 88661584.0,
      "step": 39710
    },
    {
      "entropy": 5.6272297382354735,
      "epoch": 4.460380749143595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031731222615990443,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22617052346467972,
      "num_tokens": 88673051.0,
      "step": 39715
    },
    {
      "entropy": 5.681458950042725,
      "epoch": 4.460942326051553,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031727279307580436,
      "loss": 5.0399,
      "mean_token_accuracy": 0.22187387347221374,
      "num_tokens": 88684122.0,
      "step": 39720
    },
    {
      "entropy": 5.7564942836761475,
      "epoch": 4.46150390295951,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031723335864572216,
      "loss": 5.0415,
      "mean_token_accuracy": 0.22677181661128998,
      "num_tokens": 88694620.0,
      "step": 39725
    },
    {
      "entropy": 5.798727893829346,
      "epoch": 4.4620654798674675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031719392287091343,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2224842607975006,
      "num_tokens": 88706137.0,
      "step": 39730
    },
    {
      "entropy": 5.67785382270813,
      "epoch": 4.462627056775426,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003171544857526338,
      "loss": 4.8893,
      "mean_token_accuracy": 0.226445534825325,
      "num_tokens": 88717326.0,
      "step": 39735
    },
    {
      "entropy": 5.6583380699157715,
      "epoch": 4.463188633683383,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003171150472921391,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22934769093990326,
      "num_tokens": 88727125.0,
      "step": 39740
    },
    {
      "entropy": 5.658601379394531,
      "epoch": 4.463750210591341,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031707560749068487,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2249967485666275,
      "num_tokens": 88737098.0,
      "step": 39745
    },
    {
      "entropy": 5.696771478652954,
      "epoch": 4.464311787499298,
      "grad_norm": 1.21875,
      "learning_rate": 0.000317036166349527,
      "loss": 4.939,
      "mean_token_accuracy": 0.23366841226816176,
      "num_tokens": 88748454.0,
      "step": 39750
    },
    {
      "entropy": 5.694214677810669,
      "epoch": 4.464873364407255,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003169967238699213,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23092080354690553,
      "num_tokens": 88760165.0,
      "step": 39755
    },
    {
      "entropy": 5.674565839767456,
      "epoch": 4.4654349413152135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031695728005312363,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22837663292884827,
      "num_tokens": 88772124.0,
      "step": 39760
    },
    {
      "entropy": 5.6015284061431885,
      "epoch": 4.465996518223171,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003169178349003898,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2301843374967575,
      "num_tokens": 88782429.0,
      "step": 39765
    },
    {
      "entropy": 5.673550224304199,
      "epoch": 4.466558095131128,
      "grad_norm": 1.171875,
      "learning_rate": 0.000316878388412976,
      "loss": 5.0128,
      "mean_token_accuracy": 0.22664800584316253,
      "num_tokens": 88794060.0,
      "step": 39770
    },
    {
      "entropy": 5.7512855052948,
      "epoch": 4.467119672039086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000316838940592138,
      "loss": 4.9824,
      "mean_token_accuracy": 0.23308525383472442,
      "num_tokens": 88806174.0,
      "step": 39775
    },
    {
      "entropy": 5.6103777408599855,
      "epoch": 4.467681248947043,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031679949143913183,
      "loss": 4.8304,
      "mean_token_accuracy": 0.2347559079527855,
      "num_tokens": 88817587.0,
      "step": 39780
    },
    {
      "entropy": 5.723797225952149,
      "epoch": 4.468242825855,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003167600409552137,
      "loss": 4.9058,
      "mean_token_accuracy": 0.23525333851575853,
      "num_tokens": 88829435.0,
      "step": 39785
    },
    {
      "entropy": 5.675857305526733,
      "epoch": 4.4688044027629585,
      "grad_norm": 1.25,
      "learning_rate": 0.0003167205891416397,
      "loss": 5.0125,
      "mean_token_accuracy": 0.22312256544828415,
      "num_tokens": 88840244.0,
      "step": 39790
    },
    {
      "entropy": 5.780148649215699,
      "epoch": 4.469365979670916,
      "grad_norm": 1.21875,
      "learning_rate": 0.000316681135999666,
      "loss": 5.0242,
      "mean_token_accuracy": 0.23428492844104767,
      "num_tokens": 88852085.0,
      "step": 39795
    },
    {
      "entropy": 5.749989652633667,
      "epoch": 4.469927556578874,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003166416815305486,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2243681237101555,
      "num_tokens": 88864605.0,
      "step": 39800
    },
    {
      "entropy": 5.749749994277954,
      "epoch": 4.470489133486831,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031660222573554406,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23337316811084746,
      "num_tokens": 88876209.0,
      "step": 39805
    },
    {
      "entropy": 5.636611986160278,
      "epoch": 4.471050710394788,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003165627686159085,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23365958034992218,
      "num_tokens": 88886346.0,
      "step": 39810
    },
    {
      "entropy": 5.61612639427185,
      "epoch": 4.471612287302746,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003165233101728982,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22517409771680832,
      "num_tokens": 88898190.0,
      "step": 39815
    },
    {
      "entropy": 5.7148905277252195,
      "epoch": 4.472173864210704,
      "grad_norm": 1.125,
      "learning_rate": 0.00031648385040776965,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21230324655771254,
      "num_tokens": 88909957.0,
      "step": 39820
    },
    {
      "entropy": 5.8655736446380615,
      "epoch": 4.472735441118661,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003164443893217793,
      "loss": 5.1079,
      "mean_token_accuracy": 0.22051628082990646,
      "num_tokens": 88922993.0,
      "step": 39825
    },
    {
      "entropy": 5.605487537384033,
      "epoch": 4.473297018026619,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003164049269161834,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23708995580673217,
      "num_tokens": 88934506.0,
      "step": 39830
    },
    {
      "entropy": 5.677433204650879,
      "epoch": 4.473858594934576,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031636546319223845,
      "loss": 4.9398,
      "mean_token_accuracy": 0.23182932138442994,
      "num_tokens": 88946963.0,
      "step": 39835
    },
    {
      "entropy": 5.7310638427734375,
      "epoch": 4.474420171842533,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003163259981512013,
      "loss": 5.0045,
      "mean_token_accuracy": 0.23013526052236558,
      "num_tokens": 88956995.0,
      "step": 39840
    },
    {
      "entropy": 5.762968873977661,
      "epoch": 4.4749817487504915,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003162865317943283,
      "loss": 4.9948,
      "mean_token_accuracy": 0.23017260134220124,
      "num_tokens": 88967507.0,
      "step": 39845
    },
    {
      "entropy": 5.782586765289307,
      "epoch": 4.475543325658449,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031624706412287614,
      "loss": 5.0021,
      "mean_token_accuracy": 0.22877947837114335,
      "num_tokens": 88979943.0,
      "step": 39850
    },
    {
      "entropy": 5.747351741790771,
      "epoch": 4.476104902566407,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003162075951381014,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2235921561717987,
      "num_tokens": 88990840.0,
      "step": 39855
    },
    {
      "entropy": 5.704897356033325,
      "epoch": 4.476666479474364,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031616812484126084,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2242915987968445,
      "num_tokens": 89002245.0,
      "step": 39860
    },
    {
      "entropy": 5.78555998802185,
      "epoch": 4.477228056382321,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003161286532336113,
      "loss": 4.9966,
      "mean_token_accuracy": 0.23261698186397553,
      "num_tokens": 89013520.0,
      "step": 39865
    },
    {
      "entropy": 5.728280973434448,
      "epoch": 4.477789633290279,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031608918031640946,
      "loss": 4.92,
      "mean_token_accuracy": 0.22938950955867768,
      "num_tokens": 89025377.0,
      "step": 39870
    },
    {
      "entropy": 5.660519695281982,
      "epoch": 4.478351210198237,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003160497060909121,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2264182224869728,
      "num_tokens": 89035299.0,
      "step": 39875
    },
    {
      "entropy": 5.708780717849732,
      "epoch": 4.478912787106194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031601023055837624,
      "loss": 4.984,
      "mean_token_accuracy": 0.22799801379442214,
      "num_tokens": 89046255.0,
      "step": 39880
    },
    {
      "entropy": 5.7344951152801515,
      "epoch": 4.479474364014152,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00031597075372005875,
      "loss": 5.0419,
      "mean_token_accuracy": 0.22381750941276551,
      "num_tokens": 89057873.0,
      "step": 39885
    },
    {
      "entropy": 5.674070024490357,
      "epoch": 4.480035940922109,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031593127557721657,
      "loss": 4.8522,
      "mean_token_accuracy": 0.2328539624810219,
      "num_tokens": 89068830.0,
      "step": 39890
    },
    {
      "entropy": 5.64543046951294,
      "epoch": 4.480597517830067,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003158917961311067,
      "loss": 4.9506,
      "mean_token_accuracy": 0.23480495661497117,
      "num_tokens": 89080328.0,
      "step": 39895
    },
    {
      "entropy": 5.7261772632598875,
      "epoch": 4.481159094738024,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003158523153829862,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22734831124544144,
      "num_tokens": 89090118.0,
      "step": 39900
    },
    {
      "entropy": 5.743396377563476,
      "epoch": 4.481720671645982,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003158128333341121,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2321227088570595,
      "num_tokens": 89103338.0,
      "step": 39905
    },
    {
      "entropy": 5.70250940322876,
      "epoch": 4.48228224855394,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031577334998574155,
      "loss": 4.914,
      "mean_token_accuracy": 0.23187376111745833,
      "num_tokens": 89113721.0,
      "step": 39910
    },
    {
      "entropy": 5.791436052322387,
      "epoch": 4.482843825461897,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003157338653391317,
      "loss": 5.0207,
      "mean_token_accuracy": 0.22688086777925492,
      "num_tokens": 89125425.0,
      "step": 39915
    },
    {
      "entropy": 5.71312198638916,
      "epoch": 4.483405402369854,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003156943793955398,
      "loss": 4.9883,
      "mean_token_accuracy": 0.22831060886383056,
      "num_tokens": 89135407.0,
      "step": 39920
    },
    {
      "entropy": 5.724767398834229,
      "epoch": 4.483966979277812,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031565489215622304,
      "loss": 4.9931,
      "mean_token_accuracy": 0.226283098757267,
      "num_tokens": 89147563.0,
      "step": 39925
    },
    {
      "entropy": 5.753930521011353,
      "epoch": 4.4845285561857695,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003156154036224388,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2305379405617714,
      "num_tokens": 89158336.0,
      "step": 39930
    },
    {
      "entropy": 5.649166250228882,
      "epoch": 4.485090133093728,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003155759137954444,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23591730296611785,
      "num_tokens": 89168894.0,
      "step": 39935
    },
    {
      "entropy": 5.757972526550293,
      "epoch": 4.485651710001685,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031553642267649715,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2345918521285057,
      "num_tokens": 89180701.0,
      "step": 39940
    },
    {
      "entropy": 5.715134000778198,
      "epoch": 4.486213286909642,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003154969302668544,
      "loss": 4.9323,
      "mean_token_accuracy": 0.23200661689043045,
      "num_tokens": 89191192.0,
      "step": 39945
    },
    {
      "entropy": 5.681817102432251,
      "epoch": 4.4867748638176,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003154574365677737,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23420778959989547,
      "num_tokens": 89202377.0,
      "step": 39950
    },
    {
      "entropy": 5.60348596572876,
      "epoch": 4.487336440725557,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003154179415805126,
      "loss": 4.8523,
      "mean_token_accuracy": 0.24177670031785964,
      "num_tokens": 89212869.0,
      "step": 39955
    },
    {
      "entropy": 5.688566589355469,
      "epoch": 4.487898017633515,
      "grad_norm": 1.125,
      "learning_rate": 0.0003153784453063285,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22753900289535522,
      "num_tokens": 89224558.0,
      "step": 39960
    },
    {
      "entropy": 5.621765279769898,
      "epoch": 4.488459594541473,
      "grad_norm": 1.125,
      "learning_rate": 0.0003153389477464791,
      "loss": 4.87,
      "mean_token_accuracy": 0.2363751545548439,
      "num_tokens": 89235645.0,
      "step": 39965
    },
    {
      "entropy": 5.625060081481934,
      "epoch": 4.48902117144943,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003152994489022219,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23019082695245743,
      "num_tokens": 89247063.0,
      "step": 39970
    },
    {
      "entropy": 5.674606466293335,
      "epoch": 4.489582748357387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031525994877481465,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22563959509134293,
      "num_tokens": 89258297.0,
      "step": 39975
    },
    {
      "entropy": 5.664289951324463,
      "epoch": 4.490144325265345,
      "grad_norm": 1.203125,
      "learning_rate": 0.000315220447365515,
      "loss": 4.934,
      "mean_token_accuracy": 0.22991891503334044,
      "num_tokens": 89269245.0,
      "step": 39980
    },
    {
      "entropy": 5.762246084213257,
      "epoch": 4.4907059021733025,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031518094467558074,
      "loss": 5.0014,
      "mean_token_accuracy": 0.22427121549844742,
      "num_tokens": 89281959.0,
      "step": 39985
    },
    {
      "entropy": 5.699004411697388,
      "epoch": 4.491267479081261,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031514144070626956,
      "loss": 4.8769,
      "mean_token_accuracy": 0.24343404769897461,
      "num_tokens": 89292656.0,
      "step": 39990
    },
    {
      "entropy": 5.691781091690063,
      "epoch": 4.491829055989218,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031510193545883944,
      "loss": 4.9823,
      "mean_token_accuracy": 0.23053805381059647,
      "num_tokens": 89304198.0,
      "step": 39995
    },
    {
      "entropy": 5.68519344329834,
      "epoch": 4.492390632897175,
      "grad_norm": 1.296875,
      "learning_rate": 0.000315062428934548,
      "loss": 4.9601,
      "mean_token_accuracy": 0.23576875478029252,
      "num_tokens": 89315443.0,
      "step": 40000
    },
    {
      "entropy": 5.637674283981323,
      "epoch": 4.492952209805133,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031502292113465334,
      "loss": 4.9763,
      "mean_token_accuracy": 0.23007998168468474,
      "num_tokens": 89326597.0,
      "step": 40005
    },
    {
      "entropy": 5.7258213520050045,
      "epoch": 4.49351378671309,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031498341206041337,
      "loss": 4.9632,
      "mean_token_accuracy": 0.23251817971467972,
      "num_tokens": 89338385.0,
      "step": 40010
    },
    {
      "entropy": 5.665445280075073,
      "epoch": 4.4940753636210475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000314943901713086,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23952847272157668,
      "num_tokens": 89350950.0,
      "step": 40015
    },
    {
      "entropy": 5.664525413513184,
      "epoch": 4.494636940529006,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003149043900939294,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22501963526010513,
      "num_tokens": 89361154.0,
      "step": 40020
    },
    {
      "entropy": 5.720822525024414,
      "epoch": 4.495198517436963,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031486487720420146,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22730011790990828,
      "num_tokens": 89373188.0,
      "step": 40025
    },
    {
      "entropy": 5.7050577163696286,
      "epoch": 4.49576009434492,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031482536304516047,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23777014166116714,
      "num_tokens": 89383651.0,
      "step": 40030
    },
    {
      "entropy": 5.626878786087036,
      "epoch": 4.496321671252878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003147858476180644,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23237667232751846,
      "num_tokens": 89395385.0,
      "step": 40035
    },
    {
      "entropy": 5.6380311965942385,
      "epoch": 4.496883248160835,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031474633092417155,
      "loss": 4.9293,
      "mean_token_accuracy": 0.229825359582901,
      "num_tokens": 89406071.0,
      "step": 40040
    },
    {
      "entropy": 5.737736940383911,
      "epoch": 4.4974448250687935,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031470681296474006,
      "loss": 5.0514,
      "mean_token_accuracy": 0.22572692781686782,
      "num_tokens": 89417044.0,
      "step": 40045
    },
    {
      "entropy": 5.671350574493408,
      "epoch": 4.498006401976751,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003146672937410283,
      "loss": 4.9076,
      "mean_token_accuracy": 0.22628950774669648,
      "num_tokens": 89428253.0,
      "step": 40050
    },
    {
      "entropy": 5.693119668960572,
      "epoch": 4.498567978884708,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031462777325429447,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22749807983636855,
      "num_tokens": 89438760.0,
      "step": 40055
    },
    {
      "entropy": 5.639578485488892,
      "epoch": 4.499129555792666,
      "grad_norm": 1.171875,
      "learning_rate": 0.000314588251505797,
      "loss": 4.9049,
      "mean_token_accuracy": 0.24564722329378127,
      "num_tokens": 89450335.0,
      "step": 40060
    },
    {
      "entropy": 5.746089744567871,
      "epoch": 4.499691132700623,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00031454872849679425,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22569023370742797,
      "num_tokens": 89461077.0,
      "step": 40065
    },
    {
      "entropy": 5.741124105453491,
      "epoch": 4.500252709608581,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003145092042285446,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2269244447350502,
      "num_tokens": 89472281.0,
      "step": 40070
    },
    {
      "entropy": 5.678004407882691,
      "epoch": 4.500814286516539,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003144696787023066,
      "loss": 4.9192,
      "mean_token_accuracy": 0.2303637981414795,
      "num_tokens": 89483030.0,
      "step": 40075
    },
    {
      "entropy": 5.608297204971313,
      "epoch": 4.501375863424496,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031443015191933864,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23218373209238052,
      "num_tokens": 89493355.0,
      "step": 40080
    },
    {
      "entropy": 5.695747947692871,
      "epoch": 4.501937440332454,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031439062388089953,
      "loss": 4.9819,
      "mean_token_accuracy": 0.23383208364248276,
      "num_tokens": 89504253.0,
      "step": 40085
    },
    {
      "entropy": 5.636801671981812,
      "epoch": 4.502499017240411,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003143510945882475,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22836968451738357,
      "num_tokens": 89514720.0,
      "step": 40090
    },
    {
      "entropy": 5.695025825500489,
      "epoch": 4.503060594148368,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031431156404264143,
      "loss": 4.942,
      "mean_token_accuracy": 0.23538573384284972,
      "num_tokens": 89526583.0,
      "step": 40095
    },
    {
      "entropy": 5.687039613723755,
      "epoch": 4.5036221710563265,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031427203224533996,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23321066945791244,
      "num_tokens": 89537791.0,
      "step": 40100
    },
    {
      "entropy": 5.748231601715088,
      "epoch": 4.504183747964284,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031423249919760167,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2258245453238487,
      "num_tokens": 89549480.0,
      "step": 40105
    },
    {
      "entropy": 5.681352281570435,
      "epoch": 4.504745324872241,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003141929649006854,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22453170418739318,
      "num_tokens": 89559531.0,
      "step": 40110
    },
    {
      "entropy": 5.6705962181091305,
      "epoch": 4.505306901780199,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031415342935585,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22486380487680435,
      "num_tokens": 89570238.0,
      "step": 40115
    },
    {
      "entropy": 5.6689124584198,
      "epoch": 4.505868478688156,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003141138925643542,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2324486956000328,
      "num_tokens": 89581300.0,
      "step": 40120
    },
    {
      "entropy": 5.7589866638183596,
      "epoch": 4.506430055596114,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003140743545274569,
      "loss": 4.9943,
      "mean_token_accuracy": 0.22504935264587403,
      "num_tokens": 89592461.0,
      "step": 40125
    },
    {
      "entropy": 5.7851951122283936,
      "epoch": 4.5069916325040715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031403481524641696,
      "loss": 5.0775,
      "mean_token_accuracy": 0.21440664827823638,
      "num_tokens": 89603380.0,
      "step": 40130
    },
    {
      "entropy": 5.693485069274902,
      "epoch": 4.507553209412029,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031399527472249336,
      "loss": 4.97,
      "mean_token_accuracy": 0.22592180222272873,
      "num_tokens": 89614268.0,
      "step": 40135
    },
    {
      "entropy": 5.735186338424683,
      "epoch": 4.508114786319987,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031395573295694507,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22519967705011368,
      "num_tokens": 89625480.0,
      "step": 40140
    },
    {
      "entropy": 5.7214026927948,
      "epoch": 4.508676363227944,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031391618995103116,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22971094846725465,
      "num_tokens": 89637437.0,
      "step": 40145
    },
    {
      "entropy": 5.694496297836304,
      "epoch": 4.509237940135901,
      "grad_norm": 1.125,
      "learning_rate": 0.0003138766457060107,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2369756743311882,
      "num_tokens": 89649196.0,
      "step": 40150
    },
    {
      "entropy": 5.576566076278686,
      "epoch": 4.509799517043859,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003138371002231427,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23520082533359526,
      "num_tokens": 89659536.0,
      "step": 40155
    },
    {
      "entropy": 5.701779365539551,
      "epoch": 4.510361093951817,
      "grad_norm": 1.25,
      "learning_rate": 0.0003137975535036864,
      "loss": 4.9898,
      "mean_token_accuracy": 0.23028669655323028,
      "num_tokens": 89671021.0,
      "step": 40160
    },
    {
      "entropy": 5.6637428283691404,
      "epoch": 4.510922670859774,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031375800554890094,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2373283252120018,
      "num_tokens": 89681689.0,
      "step": 40165
    },
    {
      "entropy": 5.6877930641174315,
      "epoch": 4.511484247767732,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031371845636004556,
      "loss": 4.9358,
      "mean_token_accuracy": 0.2271155297756195,
      "num_tokens": 89692353.0,
      "step": 40170
    },
    {
      "entropy": 5.658427000045776,
      "epoch": 4.512045824675689,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003136789059383795,
      "loss": 4.9667,
      "mean_token_accuracy": 0.23157906234264375,
      "num_tokens": 89702852.0,
      "step": 40175
    },
    {
      "entropy": 5.723934602737427,
      "epoch": 4.512607401583647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003136393542851622,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22558408826589585,
      "num_tokens": 89716125.0,
      "step": 40180
    },
    {
      "entropy": 5.687757873535157,
      "epoch": 4.5131689784916045,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031359980140165274,
      "loss": 4.9112,
      "mean_token_accuracy": 0.2268664076924324,
      "num_tokens": 89726611.0,
      "step": 40185
    },
    {
      "entropy": 5.67101263999939,
      "epoch": 4.513730555399562,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003135602472891106,
      "loss": 5.0252,
      "mean_token_accuracy": 0.22750550657510757,
      "num_tokens": 89738563.0,
      "step": 40190
    },
    {
      "entropy": 5.65364818572998,
      "epoch": 4.51429213230752,
      "grad_norm": 1.25,
      "learning_rate": 0.00031352069194879527,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2313015416264534,
      "num_tokens": 89748842.0,
      "step": 40195
    },
    {
      "entropy": 5.651628112792968,
      "epoch": 4.514853709215477,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003134811353819662,
      "loss": 4.9345,
      "mean_token_accuracy": 0.23225504755973816,
      "num_tokens": 89760747.0,
      "step": 40200
    },
    {
      "entropy": 5.667192173004151,
      "epoch": 4.515415286123434,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003134415775898828,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2293892651796341,
      "num_tokens": 89773059.0,
      "step": 40205
    },
    {
      "entropy": 5.714333868026733,
      "epoch": 4.515976863031392,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003134020185738047,
      "loss": 4.906,
      "mean_token_accuracy": 0.23577733635902404,
      "num_tokens": 89783347.0,
      "step": 40210
    },
    {
      "entropy": 5.683548831939698,
      "epoch": 4.51653843993935,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031336245833499146,
      "loss": 4.9521,
      "mean_token_accuracy": 0.23541935831308364,
      "num_tokens": 89793094.0,
      "step": 40215
    },
    {
      "entropy": 5.6897940158844,
      "epoch": 4.517100016847307,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003133228968747026,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22767571061849595,
      "num_tokens": 89804591.0,
      "step": 40220
    },
    {
      "entropy": 5.728367328643799,
      "epoch": 4.517661593755265,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003132833341941979,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2306938111782074,
      "num_tokens": 89815749.0,
      "step": 40225
    },
    {
      "entropy": 5.742270040512085,
      "epoch": 4.518223170663222,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000313243770294737,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22774627506732942,
      "num_tokens": 89826940.0,
      "step": 40230
    },
    {
      "entropy": 5.767226934432983,
      "epoch": 4.51878474757118,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031320420517757955,
      "loss": 5.0262,
      "mean_token_accuracy": 0.22713571637868882,
      "num_tokens": 89838955.0,
      "step": 40235
    },
    {
      "entropy": 5.676240158081055,
      "epoch": 4.519346324479137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003131646388439854,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22472320050001143,
      "num_tokens": 89850177.0,
      "step": 40240
    },
    {
      "entropy": 5.656711387634277,
      "epoch": 4.519907901387095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003131250712952144,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22707660794258117,
      "num_tokens": 89861185.0,
      "step": 40245
    },
    {
      "entropy": 5.7373090267181395,
      "epoch": 4.520469478295053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031308550253252634,
      "loss": 4.9821,
      "mean_token_accuracy": 0.23034353107213973,
      "num_tokens": 89873000.0,
      "step": 40250
    },
    {
      "entropy": 5.696518182754517,
      "epoch": 4.52103105520301,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031304593255718113,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22733585238456727,
      "num_tokens": 89884223.0,
      "step": 40255
    },
    {
      "entropy": 5.767540645599365,
      "epoch": 4.521592632110968,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031300636137043857,
      "loss": 5.042,
      "mean_token_accuracy": 0.22526390552520753,
      "num_tokens": 89895655.0,
      "step": 40260
    },
    {
      "entropy": 5.6758486270904545,
      "epoch": 4.522154209018925,
      "grad_norm": 1.25,
      "learning_rate": 0.00031296678897355877,
      "loss": 4.866,
      "mean_token_accuracy": 0.22567643076181412,
      "num_tokens": 89906475.0,
      "step": 40265
    },
    {
      "entropy": 5.642109775543213,
      "epoch": 4.5227157859268825,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003129272153678017,
      "loss": 4.956,
      "mean_token_accuracy": 0.23231690675020217,
      "num_tokens": 89917523.0,
      "step": 40270
    },
    {
      "entropy": 5.5875648021697994,
      "epoch": 4.523277362834841,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003128876405544274,
      "loss": 4.7965,
      "mean_token_accuracy": 0.2363772436976433,
      "num_tokens": 89928106.0,
      "step": 40275
    },
    {
      "entropy": 5.640685510635376,
      "epoch": 4.523838939742798,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031284806453469596,
      "loss": 4.8708,
      "mean_token_accuracy": 0.2370930090546608,
      "num_tokens": 89938638.0,
      "step": 40280
    },
    {
      "entropy": 5.695898056030273,
      "epoch": 4.524400516650755,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031280848730986735,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23763324916362763,
      "num_tokens": 89949760.0,
      "step": 40285
    },
    {
      "entropy": 5.774521207809448,
      "epoch": 4.524962093558713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003127689088812019,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2237669199705124,
      "num_tokens": 89960578.0,
      "step": 40290
    },
    {
      "entropy": 5.643291330337524,
      "epoch": 4.52552367046667,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003127293292499598,
      "loss": 4.9178,
      "mean_token_accuracy": 0.23399037420749663,
      "num_tokens": 89971370.0,
      "step": 40295
    },
    {
      "entropy": 5.666683101654053,
      "epoch": 4.526085247374628,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003126897484174012,
      "loss": 4.8516,
      "mean_token_accuracy": 0.2324277564883232,
      "num_tokens": 89982068.0,
      "step": 40300
    },
    {
      "entropy": 5.674026679992676,
      "epoch": 4.526646824282586,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003126501663847863,
      "loss": 4.8461,
      "mean_token_accuracy": 0.23670774698257446,
      "num_tokens": 89993108.0,
      "step": 40305
    },
    {
      "entropy": 5.676106882095337,
      "epoch": 4.527208401190543,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003126105831533756,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22285795956850052,
      "num_tokens": 90004604.0,
      "step": 40310
    },
    {
      "entropy": 5.740215730667114,
      "epoch": 4.527769978098501,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003125709987244293,
      "loss": 5.0186,
      "mean_token_accuracy": 0.2237156093120575,
      "num_tokens": 90015241.0,
      "step": 40315
    },
    {
      "entropy": 5.583602237701416,
      "epoch": 4.528331555006458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031253141309920784,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23567896336317062,
      "num_tokens": 90026511.0,
      "step": 40320
    },
    {
      "entropy": 5.586885452270508,
      "epoch": 4.5288931319144154,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031249182627897165,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23516704887151718,
      "num_tokens": 90038994.0,
      "step": 40325
    },
    {
      "entropy": 5.697541904449463,
      "epoch": 4.5294547088223736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031245223826498116,
      "loss": 4.963,
      "mean_token_accuracy": 0.21980032324790955,
      "num_tokens": 90049183.0,
      "step": 40330
    },
    {
      "entropy": 5.677015113830566,
      "epoch": 4.530016285730331,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031241264905849687,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2329923614859581,
      "num_tokens": 90060642.0,
      "step": 40335
    },
    {
      "entropy": 5.688725280761719,
      "epoch": 4.530577862638288,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031237305866077926,
      "loss": 4.9908,
      "mean_token_accuracy": 0.22536271661520005,
      "num_tokens": 90069709.0,
      "step": 40340
    },
    {
      "entropy": 5.707696056365966,
      "epoch": 4.531139439546246,
      "grad_norm": 1.25,
      "learning_rate": 0.00031233346707308906,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22809722870588303,
      "num_tokens": 90080219.0,
      "step": 40345
    },
    {
      "entropy": 5.712948036193848,
      "epoch": 4.531701016454203,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031229387429668675,
      "loss": 5.048,
      "mean_token_accuracy": 0.22082188874483108,
      "num_tokens": 90091769.0,
      "step": 40350
    },
    {
      "entropy": 5.731015586853028,
      "epoch": 4.5322625933621605,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031225428033283305,
      "loss": 4.9285,
      "mean_token_accuracy": 0.23337124586105346,
      "num_tokens": 90102273.0,
      "step": 40355
    },
    {
      "entropy": 5.684999704360962,
      "epoch": 4.532824170270119,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003122146851827886,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22478949874639512,
      "num_tokens": 90112311.0,
      "step": 40360
    },
    {
      "entropy": 5.597624588012695,
      "epoch": 4.533385747178076,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031217508884781413,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2343440607190132,
      "num_tokens": 90123361.0,
      "step": 40365
    },
    {
      "entropy": 5.631950807571411,
      "epoch": 4.533947324086034,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003121354913291704,
      "loss": 4.8333,
      "mean_token_accuracy": 0.23634476810693741,
      "num_tokens": 90134811.0,
      "step": 40370
    },
    {
      "entropy": 5.700833034515381,
      "epoch": 4.534508900993991,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003120958926281182,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22746268510818482,
      "num_tokens": 90145218.0,
      "step": 40375
    },
    {
      "entropy": 5.743531084060669,
      "epoch": 4.535070477901948,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031205629274591844,
      "loss": 5.0182,
      "mean_token_accuracy": 0.22988623082637788,
      "num_tokens": 90157217.0,
      "step": 40380
    },
    {
      "entropy": 5.720278644561768,
      "epoch": 4.5356320548099065,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031201669168383193,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22991358786821364,
      "num_tokens": 90167752.0,
      "step": 40385
    },
    {
      "entropy": 5.670735454559326,
      "epoch": 4.536193631717864,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031197708944311964,
      "loss": 4.917,
      "mean_token_accuracy": 0.23233776837587355,
      "num_tokens": 90178309.0,
      "step": 40390
    },
    {
      "entropy": 5.601439189910889,
      "epoch": 4.536755208625821,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003119374860250425,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2429985761642456,
      "num_tokens": 90189158.0,
      "step": 40395
    },
    {
      "entropy": 5.695719051361084,
      "epoch": 4.537316785533779,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003118978814308614,
      "loss": 4.9316,
      "mean_token_accuracy": 0.23010877519845963,
      "num_tokens": 90199611.0,
      "step": 40400
    },
    {
      "entropy": 5.7399147033691404,
      "epoch": 4.537878362441736,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003118582756618375,
      "loss": 4.9594,
      "mean_token_accuracy": 0.22912045419216157,
      "num_tokens": 90210852.0,
      "step": 40405
    },
    {
      "entropy": 5.734960222244263,
      "epoch": 4.5384399393496935,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003118186687192318,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22493079751729966,
      "num_tokens": 90221193.0,
      "step": 40410
    },
    {
      "entropy": 5.677961158752441,
      "epoch": 4.539001516257652,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031177906060430536,
      "loss": 4.9528,
      "mean_token_accuracy": 0.22471413761377335,
      "num_tokens": 90232811.0,
      "step": 40415
    },
    {
      "entropy": 5.679717397689819,
      "epoch": 4.539563093165609,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031173945131831936,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22865983992815017,
      "num_tokens": 90243566.0,
      "step": 40420
    },
    {
      "entropy": 5.720267295837402,
      "epoch": 4.540124670073567,
      "grad_norm": 1.28125,
      "learning_rate": 0.000311699840862535,
      "loss": 4.977,
      "mean_token_accuracy": 0.23768015205860138,
      "num_tokens": 90255292.0,
      "step": 40425
    },
    {
      "entropy": 5.727870130538941,
      "epoch": 4.540686246981524,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003116602292382135,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2298446476459503,
      "num_tokens": 90266691.0,
      "step": 40430
    },
    {
      "entropy": 5.694303512573242,
      "epoch": 4.541247823889481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031162061644661606,
      "loss": 4.9584,
      "mean_token_accuracy": 0.23330926448106765,
      "num_tokens": 90278240.0,
      "step": 40435
    },
    {
      "entropy": 5.7512119770050045,
      "epoch": 4.5418094007974394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000311581002489004,
      "loss": 4.9806,
      "mean_token_accuracy": 0.23262449055910112,
      "num_tokens": 90288203.0,
      "step": 40440
    },
    {
      "entropy": 5.64569787979126,
      "epoch": 4.542370977705397,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003115413873666387,
      "loss": 4.9773,
      "mean_token_accuracy": 0.23033352047204972,
      "num_tokens": 90299326.0,
      "step": 40445
    },
    {
      "entropy": 5.6664165496826175,
      "epoch": 4.542932554613355,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003115017710807813,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23107277750968933,
      "num_tokens": 90309504.0,
      "step": 40450
    },
    {
      "entropy": 5.6657623767852785,
      "epoch": 4.543494131521312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003114621536326934,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2358606666326523,
      "num_tokens": 90320919.0,
      "step": 40455
    },
    {
      "entropy": 5.661131858825684,
      "epoch": 4.544055708429269,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003114225350236365,
      "loss": 4.8306,
      "mean_token_accuracy": 0.24239781647920608,
      "num_tokens": 90331864.0,
      "step": 40460
    },
    {
      "entropy": 5.6766510009765625,
      "epoch": 4.544617285337227,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003113829152548719,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2303425446152687,
      "num_tokens": 90342563.0,
      "step": 40465
    },
    {
      "entropy": 5.632361030578613,
      "epoch": 4.5451788622451845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003113432943276611,
      "loss": 4.8485,
      "mean_token_accuracy": 0.22844285070896148,
      "num_tokens": 90353464.0,
      "step": 40470
    },
    {
      "entropy": 5.651726341247558,
      "epoch": 4.545740439153142,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031130367224326585,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2307778164744377,
      "num_tokens": 90363255.0,
      "step": 40475
    },
    {
      "entropy": 5.731158494949341,
      "epoch": 4.5463020160611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003112640490029475,
      "loss": 5.0062,
      "mean_token_accuracy": 0.22679139971733092,
      "num_tokens": 90374478.0,
      "step": 40480
    },
    {
      "entropy": 5.734487104415893,
      "epoch": 4.546863592969057,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031122442460796774,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2343481957912445,
      "num_tokens": 90386916.0,
      "step": 40485
    },
    {
      "entropy": 5.63499755859375,
      "epoch": 4.547425169877014,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003111847990595883,
      "loss": 4.8246,
      "mean_token_accuracy": 0.23388212025165558,
      "num_tokens": 90397407.0,
      "step": 40490
    },
    {
      "entropy": 5.679430150985718,
      "epoch": 4.547986746784972,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003111451723590708,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22504303604364395,
      "num_tokens": 90408762.0,
      "step": 40495
    },
    {
      "entropy": 5.663250064849853,
      "epoch": 4.54854832369293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000311105544507677,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22064884454011918,
      "num_tokens": 90418942.0,
      "step": 40500
    },
    {
      "entropy": 5.807122087478637,
      "epoch": 4.549109900600888,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031106591550666865,
      "loss": 5.087,
      "mean_token_accuracy": 0.2184985861182213,
      "num_tokens": 90430052.0,
      "step": 40505
    },
    {
      "entropy": 5.756696033477783,
      "epoch": 4.549671477508845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003110262853573076,
      "loss": 5.0201,
      "mean_token_accuracy": 0.23280788362026214,
      "num_tokens": 90442631.0,
      "step": 40510
    },
    {
      "entropy": 5.7227130889892575,
      "epoch": 4.550233054416802,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003109866540608556,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22898348420858383,
      "num_tokens": 90454429.0,
      "step": 40515
    },
    {
      "entropy": 5.696552896499634,
      "epoch": 4.55079463132476,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031094702161857455,
      "loss": 5.0281,
      "mean_token_accuracy": 0.22239596545696258,
      "num_tokens": 90466367.0,
      "step": 40520
    },
    {
      "entropy": 5.7035223007202145,
      "epoch": 4.5513562082327175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031090738803172646,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23541707098484038,
      "num_tokens": 90478480.0,
      "step": 40525
    },
    {
      "entropy": 5.678327417373657,
      "epoch": 4.551917785140675,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031086775330157324,
      "loss": 4.925,
      "mean_token_accuracy": 0.22299264669418334,
      "num_tokens": 90490501.0,
      "step": 40530
    },
    {
      "entropy": 5.696375226974487,
      "epoch": 4.552479362048633,
      "grad_norm": 1.375,
      "learning_rate": 0.0003108281174293768,
      "loss": 4.925,
      "mean_token_accuracy": 0.22598331421613693,
      "num_tokens": 90500647.0,
      "step": 40535
    },
    {
      "entropy": 5.656865167617798,
      "epoch": 4.55304093895659,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031078848041639926,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23632307946681977,
      "num_tokens": 90510963.0,
      "step": 40540
    },
    {
      "entropy": 5.7558643341064455,
      "epoch": 4.553602515864547,
      "grad_norm": 1.125,
      "learning_rate": 0.00031074884226390246,
      "loss": 4.9953,
      "mean_token_accuracy": 0.23221785724163055,
      "num_tokens": 90523475.0,
      "step": 40545
    },
    {
      "entropy": 5.686422300338745,
      "epoch": 4.554164092772505,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003107092029731488,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2302297279238701,
      "num_tokens": 90535348.0,
      "step": 40550
    },
    {
      "entropy": 5.614973688125611,
      "epoch": 4.5547256696804626,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031066956254540025,
      "loss": 4.8918,
      "mean_token_accuracy": 0.23107147067785264,
      "num_tokens": 90545841.0,
      "step": 40555
    },
    {
      "entropy": 5.6776806831359865,
      "epoch": 4.555287246588421,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031062992098191897,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22893864512443543,
      "num_tokens": 90557221.0,
      "step": 40560
    },
    {
      "entropy": 5.621248769760132,
      "epoch": 4.555848823496378,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003105902782839672,
      "loss": 4.8102,
      "mean_token_accuracy": 0.23769522756338118,
      "num_tokens": 90568080.0,
      "step": 40565
    },
    {
      "entropy": 5.702026796340943,
      "epoch": 4.556410400404335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003105506344528071,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22825413048267365,
      "num_tokens": 90579075.0,
      "step": 40570
    },
    {
      "entropy": 5.712372875213623,
      "epoch": 4.556971977312293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000310510989489701,
      "loss": 5.0075,
      "mean_token_accuracy": 0.22970953285694123,
      "num_tokens": 90589937.0,
      "step": 40575
    },
    {
      "entropy": 5.67779221534729,
      "epoch": 4.55753355422025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031047134339591136,
      "loss": 4.9085,
      "mean_token_accuracy": 0.236307530105114,
      "num_tokens": 90600262.0,
      "step": 40580
    },
    {
      "entropy": 5.76380410194397,
      "epoch": 4.558095131128208,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031043169617270033,
      "loss": 5.0288,
      "mean_token_accuracy": 0.22822521924972533,
      "num_tokens": 90612161.0,
      "step": 40585
    },
    {
      "entropy": 5.75427622795105,
      "epoch": 4.558656708036166,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031039204782133033,
      "loss": 4.9782,
      "mean_token_accuracy": 0.22518151253461838,
      "num_tokens": 90624304.0,
      "step": 40590
    },
    {
      "entropy": 5.696613407135009,
      "epoch": 4.559218284944123,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003103523983430638,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2248113974928856,
      "num_tokens": 90636196.0,
      "step": 40595
    },
    {
      "entropy": 5.665134429931641,
      "epoch": 4.55977986185208,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031031274773916326,
      "loss": 4.905,
      "mean_token_accuracy": 0.23307977467775345,
      "num_tokens": 90649024.0,
      "step": 40600
    },
    {
      "entropy": 5.648466396331787,
      "epoch": 4.560341438760038,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031027309601089104,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22545091211795806,
      "num_tokens": 90661110.0,
      "step": 40605
    },
    {
      "entropy": 5.643390274047851,
      "epoch": 4.5609030156679955,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003102334431595098,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2308002233505249,
      "num_tokens": 90671691.0,
      "step": 40610
    },
    {
      "entropy": 5.715362596511841,
      "epoch": 4.561464592575954,
      "grad_norm": 1.296875,
      "learning_rate": 0.00031019378918628203,
      "loss": 4.9046,
      "mean_token_accuracy": 0.23666052967309953,
      "num_tokens": 90683147.0,
      "step": 40615
    },
    {
      "entropy": 5.69775071144104,
      "epoch": 4.562026169483911,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031015413409247046,
      "loss": 4.9572,
      "mean_token_accuracy": 0.23181834816932678,
      "num_tokens": 90694614.0,
      "step": 40620
    },
    {
      "entropy": 5.662872076034546,
      "epoch": 4.562587746391868,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031011447787933756,
      "loss": 5.0054,
      "mean_token_accuracy": 0.224258092045784,
      "num_tokens": 90705326.0,
      "step": 40625
    },
    {
      "entropy": 5.684609460830688,
      "epoch": 4.563149323299826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003100748205481461,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23093536943197251,
      "num_tokens": 90716140.0,
      "step": 40630
    },
    {
      "entropy": 5.598288345336914,
      "epoch": 4.563710900207783,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003100351621001587,
      "loss": 4.7533,
      "mean_token_accuracy": 0.24461901932954788,
      "num_tokens": 90727181.0,
      "step": 40635
    },
    {
      "entropy": 5.622074794769287,
      "epoch": 4.5642724771157415,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003099955025366382,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2332538515329361,
      "num_tokens": 90737931.0,
      "step": 40640
    },
    {
      "entropy": 5.624098014831543,
      "epoch": 4.564834054023699,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030995584185884733,
      "loss": 4.9458,
      "mean_token_accuracy": 0.2324726715683937,
      "num_tokens": 90749011.0,
      "step": 40645
    },
    {
      "entropy": 5.6673613548278805,
      "epoch": 4.565395630931656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003099161800680488,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23224325329065323,
      "num_tokens": 90760413.0,
      "step": 40650
    },
    {
      "entropy": 5.6392090797424315,
      "epoch": 4.565957207839614,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003098765171655056,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2261739417910576,
      "num_tokens": 90770973.0,
      "step": 40655
    },
    {
      "entropy": 5.695622348785401,
      "epoch": 4.566518784747571,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003098368531524806,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2346402257680893,
      "num_tokens": 90782204.0,
      "step": 40660
    },
    {
      "entropy": 5.70778260231018,
      "epoch": 4.567080361655528,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003097971880302367,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2330818310379982,
      "num_tokens": 90791906.0,
      "step": 40665
    },
    {
      "entropy": 5.648031759262085,
      "epoch": 4.5676419385634865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030975752180003674,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23130397796630858,
      "num_tokens": 90802153.0,
      "step": 40670
    },
    {
      "entropy": 5.636433267593384,
      "epoch": 4.568203515471444,
      "grad_norm": 1.28125,
      "learning_rate": 0.000309717854463144,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2309788167476654,
      "num_tokens": 90813129.0,
      "step": 40675
    },
    {
      "entropy": 5.738002490997315,
      "epoch": 4.568765092379401,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003096781860208211,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22688987106084824,
      "num_tokens": 90823445.0,
      "step": 40680
    },
    {
      "entropy": 5.685943126678467,
      "epoch": 4.569326669287359,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003096385164743314,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2294995069503784,
      "num_tokens": 90834823.0,
      "step": 40685
    },
    {
      "entropy": 5.590376329421997,
      "epoch": 4.569888246195316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030959884582493793,
      "loss": 4.7882,
      "mean_token_accuracy": 0.23679041713476182,
      "num_tokens": 90845503.0,
      "step": 40690
    },
    {
      "entropy": 5.745629453659058,
      "epoch": 4.570449823103274,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003095591740739038,
      "loss": 5.007,
      "mean_token_accuracy": 0.22438061088323594,
      "num_tokens": 90857562.0,
      "step": 40695
    },
    {
      "entropy": 5.720430707931518,
      "epoch": 4.571011400011232,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003095195012224921,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22659128159284592,
      "num_tokens": 90868207.0,
      "step": 40700
    },
    {
      "entropy": 5.623395919799805,
      "epoch": 4.571572976919189,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003094798272719661,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22628589421510698,
      "num_tokens": 90879710.0,
      "step": 40705
    },
    {
      "entropy": 5.728093194961548,
      "epoch": 4.572134553827147,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030944015222358906,
      "loss": 4.93,
      "mean_token_accuracy": 0.22935021966695784,
      "num_tokens": 90890702.0,
      "step": 40710
    },
    {
      "entropy": 5.674951028823853,
      "epoch": 4.572696130735104,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030940047607862415,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23470368832349778,
      "num_tokens": 90902171.0,
      "step": 40715
    },
    {
      "entropy": 5.689206314086914,
      "epoch": 4.573257707643061,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003093607988383348,
      "loss": 4.908,
      "mean_token_accuracy": 0.22975191622972488,
      "num_tokens": 90913735.0,
      "step": 40720
    },
    {
      "entropy": 5.686890983581543,
      "epoch": 4.5738192845510195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003093211205039842,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22722995579242705,
      "num_tokens": 90924137.0,
      "step": 40725
    },
    {
      "entropy": 5.65024356842041,
      "epoch": 4.574380861458977,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003092814410768359,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23474815040826796,
      "num_tokens": 90935214.0,
      "step": 40730
    },
    {
      "entropy": 5.6441216468811035,
      "epoch": 4.574942438366934,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030924176055815327,
      "loss": 4.8771,
      "mean_token_accuracy": 0.23437264710664749,
      "num_tokens": 90946797.0,
      "step": 40735
    },
    {
      "entropy": 5.622460556030274,
      "epoch": 4.575504015274892,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030920207894919957,
      "loss": 4.8932,
      "mean_token_accuracy": 0.23208916634321214,
      "num_tokens": 90958055.0,
      "step": 40740
    },
    {
      "entropy": 5.635843563079834,
      "epoch": 4.576065592182849,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003091623962512385,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23361043483018876,
      "num_tokens": 90969516.0,
      "step": 40745
    },
    {
      "entropy": 5.576106214523316,
      "epoch": 4.576627169090807,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030912271246553346,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23149093240499496,
      "num_tokens": 90980536.0,
      "step": 40750
    },
    {
      "entropy": 5.679629278182984,
      "epoch": 4.577188745998765,
      "grad_norm": 1.328125,
      "learning_rate": 0.000309083027593348,
      "loss": 4.9714,
      "mean_token_accuracy": 0.23373462557792662,
      "num_tokens": 90992150.0,
      "step": 40755
    },
    {
      "entropy": 5.698806285858154,
      "epoch": 4.577750322906722,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030904334163594577,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2275987833738327,
      "num_tokens": 91003323.0,
      "step": 40760
    },
    {
      "entropy": 5.629584932327271,
      "epoch": 4.57831189981468,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003090036545945904,
      "loss": 4.8309,
      "mean_token_accuracy": 0.24124495238065718,
      "num_tokens": 91014238.0,
      "step": 40765
    },
    {
      "entropy": 5.672249221801758,
      "epoch": 4.578873476722637,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030896396647054535,
      "loss": 4.9387,
      "mean_token_accuracy": 0.22854862362146378,
      "num_tokens": 91025375.0,
      "step": 40770
    },
    {
      "entropy": 5.633003568649292,
      "epoch": 4.579435053630594,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003089242772650744,
      "loss": 4.971,
      "mean_token_accuracy": 0.22803101390600206,
      "num_tokens": 91036498.0,
      "step": 40775
    },
    {
      "entropy": 5.641131448745727,
      "epoch": 4.579996630538552,
      "grad_norm": 1.25,
      "learning_rate": 0.00030888458697944144,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2336245983839035,
      "num_tokens": 91046909.0,
      "step": 40780
    },
    {
      "entropy": 5.754197120666504,
      "epoch": 4.58055820744651,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030884489561491005,
      "loss": 5.0326,
      "mean_token_accuracy": 0.22465491890907288,
      "num_tokens": 91056052.0,
      "step": 40785
    },
    {
      "entropy": 5.849635982513428,
      "epoch": 4.581119784354467,
      "grad_norm": 1.15625,
      "learning_rate": 0.000308805203172744,
      "loss": 5.1145,
      "mean_token_accuracy": 0.21689886450767518,
      "num_tokens": 91067614.0,
      "step": 40790
    },
    {
      "entropy": 5.610347938537598,
      "epoch": 4.581681361262425,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030876550965420725,
      "loss": 4.8562,
      "mean_token_accuracy": 0.23566687107086182,
      "num_tokens": 91078539.0,
      "step": 40795
    },
    {
      "entropy": 5.689392042160034,
      "epoch": 4.582242938170382,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003087258150605635,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22748344540596008,
      "num_tokens": 91088909.0,
      "step": 40800
    },
    {
      "entropy": 5.629751825332642,
      "epoch": 4.58280451507834,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003086861193930767,
      "loss": 4.9427,
      "mean_token_accuracy": 0.23177082240581512,
      "num_tokens": 91101073.0,
      "step": 40805
    },
    {
      "entropy": 5.68203911781311,
      "epoch": 4.5833660919862975,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003086464226530109,
      "loss": 4.9091,
      "mean_token_accuracy": 0.23157355040311814,
      "num_tokens": 91113518.0,
      "step": 40810
    },
    {
      "entropy": 5.7389098644256595,
      "epoch": 4.583927668894255,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030860672484162987,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23473233133554458,
      "num_tokens": 91124287.0,
      "step": 40815
    },
    {
      "entropy": 5.7634271621704105,
      "epoch": 4.584489245802213,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003085670259601977,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22933127284049987,
      "num_tokens": 91134828.0,
      "step": 40820
    },
    {
      "entropy": 5.684616184234619,
      "epoch": 4.58505082271017,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030852732600997846,
      "loss": 4.969,
      "mean_token_accuracy": 0.22563833147287368,
      "num_tokens": 91146647.0,
      "step": 40825
    },
    {
      "entropy": 5.643408536911011,
      "epoch": 4.585612399618128,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030848762499223616,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22762055248022078,
      "num_tokens": 91156769.0,
      "step": 40830
    },
    {
      "entropy": 5.694388818740845,
      "epoch": 4.586173976526085,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003084479229082348,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2366035908460617,
      "num_tokens": 91168876.0,
      "step": 40835
    },
    {
      "entropy": 5.766136503219604,
      "epoch": 4.586735553434043,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003084082197592387,
      "loss": 5.0595,
      "mean_token_accuracy": 0.22644151300191878,
      "num_tokens": 91178587.0,
      "step": 40840
    },
    {
      "entropy": 5.642407035827636,
      "epoch": 4.587297130342001,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030836851554651184,
      "loss": 4.8377,
      "mean_token_accuracy": 0.2414844274520874,
      "num_tokens": 91190210.0,
      "step": 40845
    },
    {
      "entropy": 5.682376956939697,
      "epoch": 4.587858707249958,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030832881027131854,
      "loss": 4.9566,
      "mean_token_accuracy": 0.23313515782356262,
      "num_tokens": 91200921.0,
      "step": 40850
    },
    {
      "entropy": 5.669706630706787,
      "epoch": 4.588420284157915,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030828910393492306,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2351273387670517,
      "num_tokens": 91211553.0,
      "step": 40855
    },
    {
      "entropy": 5.707495212554932,
      "epoch": 4.588981861065873,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030824939653858953,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22760127931833268,
      "num_tokens": 91222224.0,
      "step": 40860
    },
    {
      "entropy": 5.722015857696533,
      "epoch": 4.5895434379738305,
      "grad_norm": 1.125,
      "learning_rate": 0.00030820968808358223,
      "loss": 4.8988,
      "mean_token_accuracy": 0.23260479271411896,
      "num_tokens": 91233645.0,
      "step": 40865
    },
    {
      "entropy": 5.732394170761109,
      "epoch": 4.590105014881788,
      "grad_norm": 1.25,
      "learning_rate": 0.00030816997857116573,
      "loss": 4.9329,
      "mean_token_accuracy": 0.23077697157859803,
      "num_tokens": 91245194.0,
      "step": 40870
    },
    {
      "entropy": 5.646994018554688,
      "epoch": 4.590666591789746,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003081302680026042,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22861131131649018,
      "num_tokens": 91255124.0,
      "step": 40875
    },
    {
      "entropy": 5.716728973388672,
      "epoch": 4.591228168697703,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000308090556379162,
      "loss": 5.0426,
      "mean_token_accuracy": 0.22344864010810853,
      "num_tokens": 91267129.0,
      "step": 40880
    },
    {
      "entropy": 5.676351690292359,
      "epoch": 4.591789745605661,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030805084370210377,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2270968809723854,
      "num_tokens": 91277614.0,
      "step": 40885
    },
    {
      "entropy": 5.708772659301758,
      "epoch": 4.592351322513618,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030801112997269386,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22736005932092668,
      "num_tokens": 91289856.0,
      "step": 40890
    },
    {
      "entropy": 5.630273866653442,
      "epoch": 4.5929128994215755,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003079714151921967,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22975433766841888,
      "num_tokens": 91300593.0,
      "step": 40895
    },
    {
      "entropy": 5.680385303497315,
      "epoch": 4.593474476329534,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030793169936187695,
      "loss": 4.8589,
      "mean_token_accuracy": 0.24352422207593918,
      "num_tokens": 91312934.0,
      "step": 40900
    },
    {
      "entropy": 5.619242238998413,
      "epoch": 4.594036053237491,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030789198248299906,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23242080509662627,
      "num_tokens": 91323856.0,
      "step": 40905
    },
    {
      "entropy": 5.70281753540039,
      "epoch": 4.594597630145448,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030785226455682773,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22632498294115067,
      "num_tokens": 91334545.0,
      "step": 40910
    },
    {
      "entropy": 5.768341588973999,
      "epoch": 4.595159207053406,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003078125455846276,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2296860158443451,
      "num_tokens": 91346006.0,
      "step": 40915
    },
    {
      "entropy": 5.719723081588745,
      "epoch": 4.595720783961363,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030777282556766316,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21593126952648162,
      "num_tokens": 91356710.0,
      "step": 40920
    },
    {
      "entropy": 5.6724937915802,
      "epoch": 4.596282360869321,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030773310450719937,
      "loss": 4.888,
      "mean_token_accuracy": 0.22795527577400207,
      "num_tokens": 91367747.0,
      "step": 40925
    },
    {
      "entropy": 5.728208351135254,
      "epoch": 4.596843937777279,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003076933824045008,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22827720642089844,
      "num_tokens": 91379079.0,
      "step": 40930
    },
    {
      "entropy": 5.67640790939331,
      "epoch": 4.597405514685236,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030765365926083214,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22471380233764648,
      "num_tokens": 91390264.0,
      "step": 40935
    },
    {
      "entropy": 5.6565135478973385,
      "epoch": 4.597967091593194,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003076139350774584,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23734679520130159,
      "num_tokens": 91401322.0,
      "step": 40940
    },
    {
      "entropy": 5.651098823547363,
      "epoch": 4.598528668501151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003075742098556444,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2294853612780571,
      "num_tokens": 91412277.0,
      "step": 40945
    },
    {
      "entropy": 5.594609546661377,
      "epoch": 4.5990902454091085,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003075344835966548,
      "loss": 4.8741,
      "mean_token_accuracy": 0.22883608490228652,
      "num_tokens": 91422941.0,
      "step": 40950
    },
    {
      "entropy": 5.660964441299439,
      "epoch": 4.599651822317067,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003074947563017547,
      "loss": 4.9038,
      "mean_token_accuracy": 0.232273131608963,
      "num_tokens": 91434194.0,
      "step": 40955
    },
    {
      "entropy": 5.664493274688721,
      "epoch": 4.600213399225024,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003074550279722089,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2293860763311386,
      "num_tokens": 91445330.0,
      "step": 40960
    },
    {
      "entropy": 5.71310076713562,
      "epoch": 4.600774976132981,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003074152986092825,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22969044446945192,
      "num_tokens": 91456762.0,
      "step": 40965
    },
    {
      "entropy": 5.643471670150757,
      "epoch": 4.601336553040939,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003073755682142404,
      "loss": 4.8263,
      "mean_token_accuracy": 0.23225726038217545,
      "num_tokens": 91467272.0,
      "step": 40970
    },
    {
      "entropy": 5.712204647064209,
      "epoch": 4.601898129948896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030733583678834765,
      "loss": 5.0426,
      "mean_token_accuracy": 0.22455989569425583,
      "num_tokens": 91478188.0,
      "step": 40975
    },
    {
      "entropy": 5.659331369400024,
      "epoch": 4.602459706856854,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030729610433286926,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2312658354640007,
      "num_tokens": 91489759.0,
      "step": 40980
    },
    {
      "entropy": 5.7262811183929445,
      "epoch": 4.603021283764812,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030725637084907037,
      "loss": 4.945,
      "mean_token_accuracy": 0.23182709962129594,
      "num_tokens": 91500705.0,
      "step": 40985
    },
    {
      "entropy": 5.654567289352417,
      "epoch": 4.603582860672769,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003072166363382162,
      "loss": 4.9552,
      "mean_token_accuracy": 0.23011601865291595,
      "num_tokens": 91511832.0,
      "step": 40990
    },
    {
      "entropy": 5.704951429367066,
      "epoch": 4.604144437580727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003071769008015717,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22659798562526703,
      "num_tokens": 91522454.0,
      "step": 40995
    },
    {
      "entropy": 5.72240104675293,
      "epoch": 4.604706014488684,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003071371642404024,
      "loss": 5.0253,
      "mean_token_accuracy": 0.22746056020259858,
      "num_tokens": 91533987.0,
      "step": 41000
    },
    {
      "entropy": 5.681027460098266,
      "epoch": 4.605267591396641,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003070974266559732,
      "loss": 4.9337,
      "mean_token_accuracy": 0.23229524940252305,
      "num_tokens": 91545000.0,
      "step": 41005
    },
    {
      "entropy": 5.803507709503174,
      "epoch": 4.6058291683045995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003070576880495495,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2276366174221039,
      "num_tokens": 91556481.0,
      "step": 41010
    },
    {
      "entropy": 5.661529779434204,
      "epoch": 4.606390745212557,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003070179484223966,
      "loss": 4.8569,
      "mean_token_accuracy": 0.23887065947055816,
      "num_tokens": 91568668.0,
      "step": 41015
    },
    {
      "entropy": 5.6335062980651855,
      "epoch": 4.606952322120515,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003069782077757797,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2366312637925148,
      "num_tokens": 91579884.0,
      "step": 41020
    },
    {
      "entropy": 5.7362494468688965,
      "epoch": 4.607513899028472,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030693846611096437,
      "loss": 5.0472,
      "mean_token_accuracy": 0.23121195882558823,
      "num_tokens": 91590488.0,
      "step": 41025
    },
    {
      "entropy": 5.71304407119751,
      "epoch": 4.608075475936429,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030689872342921595,
      "loss": 4.92,
      "mean_token_accuracy": 0.229910209774971,
      "num_tokens": 91602163.0,
      "step": 41030
    },
    {
      "entropy": 5.669404697418213,
      "epoch": 4.608637052844387,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030685897973179967,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23722280710935592,
      "num_tokens": 91614251.0,
      "step": 41035
    },
    {
      "entropy": 5.696557998657227,
      "epoch": 4.609198629752345,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030681923501998124,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23362279683351517,
      "num_tokens": 91625102.0,
      "step": 41040
    },
    {
      "entropy": 5.740299654006958,
      "epoch": 4.609760206660302,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030677948929502596,
      "loss": 4.9635,
      "mean_token_accuracy": 0.23302365988492965,
      "num_tokens": 91636613.0,
      "step": 41045
    },
    {
      "entropy": 5.657492017745971,
      "epoch": 4.61032178356826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030673974255819944,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2303254172205925,
      "num_tokens": 91647648.0,
      "step": 41050
    },
    {
      "entropy": 5.829934597015381,
      "epoch": 4.610883360476217,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003066999948107672,
      "loss": 5.0469,
      "mean_token_accuracy": 0.22086264938116074,
      "num_tokens": 91659160.0,
      "step": 41055
    },
    {
      "entropy": 5.684950590133667,
      "epoch": 4.611444937384174,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003066602460539948,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2400971382856369,
      "num_tokens": 91670973.0,
      "step": 41060
    },
    {
      "entropy": 5.706352663040161,
      "epoch": 4.6120065142921325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030662049628914787,
      "loss": 5.039,
      "mean_token_accuracy": 0.22818835228681564,
      "num_tokens": 91682710.0,
      "step": 41065
    },
    {
      "entropy": 5.7379742622375485,
      "epoch": 4.61256809120009,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003065807455174921,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22662048488855363,
      "num_tokens": 91694008.0,
      "step": 41070
    },
    {
      "entropy": 5.652651166915893,
      "epoch": 4.613129668108048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003065409937402932,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22764723747968674,
      "num_tokens": 91705289.0,
      "step": 41075
    },
    {
      "entropy": 5.6702048778533936,
      "epoch": 4.613691245016005,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030650124095881674,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22938291728496552,
      "num_tokens": 91717078.0,
      "step": 41080
    },
    {
      "entropy": 5.633471822738647,
      "epoch": 4.614252821923962,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030646148717432855,
      "loss": 4.9328,
      "mean_token_accuracy": 0.23012931644916534,
      "num_tokens": 91728971.0,
      "step": 41085
    },
    {
      "entropy": 5.6607218265533445,
      "epoch": 4.61481439883192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030642173238809453,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23652640879154205,
      "num_tokens": 91739496.0,
      "step": 41090
    },
    {
      "entropy": 5.72120885848999,
      "epoch": 4.615375975739878,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003063819766013802,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22180007845163346,
      "num_tokens": 91749732.0,
      "step": 41095
    },
    {
      "entropy": 5.660435390472412,
      "epoch": 4.615937552647835,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003063422198154516,
      "loss": 4.8733,
      "mean_token_accuracy": 0.22302502393722534,
      "num_tokens": 91758772.0,
      "step": 41100
    },
    {
      "entropy": 5.683911085128784,
      "epoch": 4.616499129555793,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003063024620315746,
      "loss": 4.8512,
      "mean_token_accuracy": 0.23616029769182206,
      "num_tokens": 91770012.0,
      "step": 41105
    },
    {
      "entropy": 5.6839924335479735,
      "epoch": 4.61706070646375,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003062627032510151,
      "loss": 4.981,
      "mean_token_accuracy": 0.22230131179094315,
      "num_tokens": 91781926.0,
      "step": 41110
    },
    {
      "entropy": 5.709225273132324,
      "epoch": 4.617622283371707,
      "grad_norm": 1.15625,
      "learning_rate": 0.000306222943475039,
      "loss": 5.0324,
      "mean_token_accuracy": 0.22950912564992904,
      "num_tokens": 91794598.0,
      "step": 41115
    },
    {
      "entropy": 5.70530891418457,
      "epoch": 4.618183860279665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003061831827049123,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22466447204351425,
      "num_tokens": 91805898.0,
      "step": 41120
    },
    {
      "entropy": 5.722815418243409,
      "epoch": 4.618745437187623,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00030614342094190095,
      "loss": 4.9373,
      "mean_token_accuracy": 0.23041225522756575,
      "num_tokens": 91816369.0,
      "step": 41125
    },
    {
      "entropy": 5.549451446533203,
      "epoch": 4.619307014095581,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030610365818727097,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22966053783893586,
      "num_tokens": 91827223.0,
      "step": 41130
    },
    {
      "entropy": 5.672425031661987,
      "epoch": 4.619868591003538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003060638944422884,
      "loss": 4.9775,
      "mean_token_accuracy": 0.225578935444355,
      "num_tokens": 91838898.0,
      "step": 41135
    },
    {
      "entropy": 5.720216608047485,
      "epoch": 4.620430167911495,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030602412970821944,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2247800886631012,
      "num_tokens": 91850660.0,
      "step": 41140
    },
    {
      "entropy": 5.693855237960816,
      "epoch": 4.620991744819453,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003059843639863301,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22772235423326492,
      "num_tokens": 91860594.0,
      "step": 41145
    },
    {
      "entropy": 5.6737017154693605,
      "epoch": 4.6215533217274105,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030594459727788665,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22557242810726166,
      "num_tokens": 91872209.0,
      "step": 41150
    },
    {
      "entropy": 5.721431303024292,
      "epoch": 4.622114898635368,
      "grad_norm": 1.125,
      "learning_rate": 0.00030590482958415524,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22765921503305436,
      "num_tokens": 91885241.0,
      "step": 41155
    },
    {
      "entropy": 5.800809049606324,
      "epoch": 4.622676475543326,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030586506090640205,
      "loss": 5.0014,
      "mean_token_accuracy": 0.23890303671360016,
      "num_tokens": 91896999.0,
      "step": 41160
    },
    {
      "entropy": 5.643677282333374,
      "epoch": 4.623238052451283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003058252912458933,
      "loss": 4.8843,
      "mean_token_accuracy": 0.22792392671108247,
      "num_tokens": 91908554.0,
      "step": 41165
    },
    {
      "entropy": 5.63778657913208,
      "epoch": 4.62379962935924,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003057855206038954,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22733526229858397,
      "num_tokens": 91920013.0,
      "step": 41170
    },
    {
      "entropy": 5.720777463912964,
      "epoch": 4.624361206267198,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003057457489816745,
      "loss": 4.967,
      "mean_token_accuracy": 0.2280610144138336,
      "num_tokens": 91930951.0,
      "step": 41175
    },
    {
      "entropy": 5.6570258140563965,
      "epoch": 4.624922783175156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003057059763804971,
      "loss": 4.9317,
      "mean_token_accuracy": 0.23462909162044526,
      "num_tokens": 91941718.0,
      "step": 41180
    },
    {
      "entropy": 5.594879245758056,
      "epoch": 4.625484360083114,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003056662028016295,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23470367342233658,
      "num_tokens": 91952756.0,
      "step": 41185
    },
    {
      "entropy": 5.67078332901001,
      "epoch": 4.626045936991071,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030562642824633805,
      "loss": 4.9194,
      "mean_token_accuracy": 0.23233454525470734,
      "num_tokens": 91964200.0,
      "step": 41190
    },
    {
      "entropy": 5.672557163238525,
      "epoch": 4.626607513899028,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003055866527158893,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22412475645542146,
      "num_tokens": 91975032.0,
      "step": 41195
    },
    {
      "entropy": 5.661380910873413,
      "epoch": 4.627169090806986,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003055468762115497,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23042073249816894,
      "num_tokens": 91986566.0,
      "step": 41200
    },
    {
      "entropy": 5.65772557258606,
      "epoch": 4.6277306677149435,
      "grad_norm": 1.21875,
      "learning_rate": 0.00030550709873458567,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23843657672405244,
      "num_tokens": 91996944.0,
      "step": 41205
    },
    {
      "entropy": 5.68788480758667,
      "epoch": 4.628292244622902,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030546732028626373,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23927495628595352,
      "num_tokens": 92007692.0,
      "step": 41210
    },
    {
      "entropy": 5.669079256057739,
      "epoch": 4.628853821530859,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030542754086785055,
      "loss": 4.9,
      "mean_token_accuracy": 0.23612950146198272,
      "num_tokens": 92017514.0,
      "step": 41215
    },
    {
      "entropy": 5.66175332069397,
      "epoch": 4.629415398438816,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003053877604806127,
      "loss": 4.8708,
      "mean_token_accuracy": 0.23018234074115754,
      "num_tokens": 92028311.0,
      "step": 41220
    },
    {
      "entropy": 5.5964418888092045,
      "epoch": 4.629976975346774,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030534797912581664,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2396205633878708,
      "num_tokens": 92038795.0,
      "step": 41225
    },
    {
      "entropy": 5.579810714721679,
      "epoch": 4.630538552254731,
      "grad_norm": 1.25,
      "learning_rate": 0.00030530819680472926,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23078676909208298,
      "num_tokens": 92049578.0,
      "step": 41230
    },
    {
      "entropy": 5.802322769165039,
      "epoch": 4.6311001291626885,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030526841351861717,
      "loss": 5.083,
      "mean_token_accuracy": 0.22324947118759156,
      "num_tokens": 92062868.0,
      "step": 41235
    },
    {
      "entropy": 5.789940071105957,
      "epoch": 4.631661706070647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030522862926874697,
      "loss": 4.9492,
      "mean_token_accuracy": 0.23678210526704788,
      "num_tokens": 92073819.0,
      "step": 41240
    },
    {
      "entropy": 5.719776248931884,
      "epoch": 4.632223282978604,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003051888440563855,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21943332105875016,
      "num_tokens": 92085212.0,
      "step": 41245
    },
    {
      "entropy": 5.698294734954834,
      "epoch": 4.632784859886561,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003051490578827996,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2328324243426323,
      "num_tokens": 92096807.0,
      "step": 41250
    },
    {
      "entropy": 5.684759569168091,
      "epoch": 4.633346436794519,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030510927074925594,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2358625203371048,
      "num_tokens": 92108619.0,
      "step": 41255
    },
    {
      "entropy": 5.691089153289795,
      "epoch": 4.633908013702476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030506948265702144,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22871422916650772,
      "num_tokens": 92120815.0,
      "step": 41260
    },
    {
      "entropy": 5.6474565982818605,
      "epoch": 4.6344695906104345,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000305029693607363,
      "loss": 4.9142,
      "mean_token_accuracy": 0.22873939275741578,
      "num_tokens": 92131957.0,
      "step": 41265
    },
    {
      "entropy": 5.657431888580322,
      "epoch": 4.635031167518392,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030498990360154746,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23606175184249878,
      "num_tokens": 92143936.0,
      "step": 41270
    },
    {
      "entropy": 5.704864263534546,
      "epoch": 4.635592744426349,
      "grad_norm": 1.21875,
      "learning_rate": 0.00030495011264084165,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23256365954875946,
      "num_tokens": 92155320.0,
      "step": 41275
    },
    {
      "entropy": 5.732161188125611,
      "epoch": 4.636154321334307,
      "grad_norm": 1.21875,
      "learning_rate": 0.00030491032072651266,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2358216166496277,
      "num_tokens": 92166831.0,
      "step": 41280
    },
    {
      "entropy": 5.719066429138183,
      "epoch": 4.636715898242264,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030487052785982757,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22616610676050186,
      "num_tokens": 92177927.0,
      "step": 41285
    },
    {
      "entropy": 5.70571117401123,
      "epoch": 4.6372774751502215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003048307340420532,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2239794358611107,
      "num_tokens": 92188755.0,
      "step": 41290
    },
    {
      "entropy": 5.696298980712891,
      "epoch": 4.63783905205818,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003047909392744566,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23000190407037735,
      "num_tokens": 92199156.0,
      "step": 41295
    },
    {
      "entropy": 5.662765026092529,
      "epoch": 4.638400628966137,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030475114355830513,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2364081025123596,
      "num_tokens": 92211080.0,
      "step": 41300
    },
    {
      "entropy": 5.7060546875,
      "epoch": 4.638962205874094,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030471134689486563,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22784582078456878,
      "num_tokens": 92220734.0,
      "step": 41305
    },
    {
      "entropy": 5.731780624389648,
      "epoch": 4.639523782782052,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003046715492854053,
      "loss": 5.0473,
      "mean_token_accuracy": 0.22257788181304933,
      "num_tokens": 92231166.0,
      "step": 41310
    },
    {
      "entropy": 5.738078451156616,
      "epoch": 4.640085359690009,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030463175073119133,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22470748275518418,
      "num_tokens": 92241920.0,
      "step": 41315
    },
    {
      "entropy": 5.652801084518432,
      "epoch": 4.6406469365979675,
      "grad_norm": 1.125,
      "learning_rate": 0.00030459195123349095,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2302542433142662,
      "num_tokens": 92253745.0,
      "step": 41320
    },
    {
      "entropy": 5.667269849777222,
      "epoch": 4.641208513505925,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003045521507935714,
      "loss": 4.9476,
      "mean_token_accuracy": 0.23088291436433792,
      "num_tokens": 92263672.0,
      "step": 41325
    },
    {
      "entropy": 5.726401615142822,
      "epoch": 4.641770090413882,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003045123494126999,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2274210348725319,
      "num_tokens": 92275124.0,
      "step": 41330
    },
    {
      "entropy": 5.776758146286011,
      "epoch": 4.64233166732184,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003044725470921437,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22492340207099915,
      "num_tokens": 92287077.0,
      "step": 41335
    },
    {
      "entropy": 5.733022022247314,
      "epoch": 4.642893244229797,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003044327438331702,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22980345487594606,
      "num_tokens": 92298649.0,
      "step": 41340
    },
    {
      "entropy": 5.642969846725464,
      "epoch": 4.643454821137754,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003043929396370467,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22886722385883332,
      "num_tokens": 92309589.0,
      "step": 41345
    },
    {
      "entropy": 5.587350511550904,
      "epoch": 4.6440163980457125,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030435313450504056,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23025670349597932,
      "num_tokens": 92319927.0,
      "step": 41350
    },
    {
      "entropy": 5.700066137313843,
      "epoch": 4.64457797495367,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003043133284384193,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22788115739822387,
      "num_tokens": 92333156.0,
      "step": 41355
    },
    {
      "entropy": 5.626075506210327,
      "epoch": 4.645139551861627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003042735214384503,
      "loss": 4.7983,
      "mean_token_accuracy": 0.23568016141653061,
      "num_tokens": 92345228.0,
      "step": 41360
    },
    {
      "entropy": 5.72742977142334,
      "epoch": 4.645701128769585,
      "grad_norm": 1.296875,
      "learning_rate": 0.000304233713506401,
      "loss": 4.986,
      "mean_token_accuracy": 0.23191258758306504,
      "num_tokens": 92355922.0,
      "step": 41365
    },
    {
      "entropy": 5.684233903884888,
      "epoch": 4.646262705677542,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003041939046435389,
      "loss": 4.9449,
      "mean_token_accuracy": 0.23434391468763352,
      "num_tokens": 92367255.0,
      "step": 41370
    },
    {
      "entropy": 5.603784513473511,
      "epoch": 4.6468242825855,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030415409485113156,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23470613211393357,
      "num_tokens": 92378577.0,
      "step": 41375
    },
    {
      "entropy": 5.685605001449585,
      "epoch": 4.647385859493458,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00030411428413044657,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22903090119361877,
      "num_tokens": 92389822.0,
      "step": 41380
    },
    {
      "entropy": 5.777561140060425,
      "epoch": 4.647947436401415,
      "grad_norm": 1.25,
      "learning_rate": 0.00030407447248275137,
      "loss": 5.0437,
      "mean_token_accuracy": 0.22228287607431413,
      "num_tokens": 92400682.0,
      "step": 41385
    },
    {
      "entropy": 5.751329803466797,
      "epoch": 4.648509013309373,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003040346599093138,
      "loss": 4.9244,
      "mean_token_accuracy": 0.23972146809101105,
      "num_tokens": 92412194.0,
      "step": 41390
    },
    {
      "entropy": 5.703125619888306,
      "epoch": 4.64907059021733,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003039948464114014,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23594462275505065,
      "num_tokens": 92423311.0,
      "step": 41395
    },
    {
      "entropy": 5.6720606803894045,
      "epoch": 4.649632167125288,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030395503199028186,
      "loss": 5.0235,
      "mean_token_accuracy": 0.22659768164157867,
      "num_tokens": 92434008.0,
      "step": 41400
    },
    {
      "entropy": 5.658440160751343,
      "epoch": 4.6501937440332455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003039152166472228,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23747073858976364,
      "num_tokens": 92445887.0,
      "step": 41405
    },
    {
      "entropy": 5.777170658111572,
      "epoch": 4.650755320941203,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003038754003834921,
      "loss": 5.0051,
      "mean_token_accuracy": 0.23649463355541228,
      "num_tokens": 92457856.0,
      "step": 41410
    },
    {
      "entropy": 5.778076839447022,
      "epoch": 4.651316897849161,
      "grad_norm": 1.125,
      "learning_rate": 0.00030383558320035744,
      "loss": 4.9773,
      "mean_token_accuracy": 0.23835514932870866,
      "num_tokens": 92469976.0,
      "step": 41415
    },
    {
      "entropy": 5.645499849319458,
      "epoch": 4.651878474757118,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030379576509908656,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22969358563423156,
      "num_tokens": 92481035.0,
      "step": 41420
    },
    {
      "entropy": 5.651830959320068,
      "epoch": 4.652440051665075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030375594608094737,
      "loss": 4.961,
      "mean_token_accuracy": 0.2270359978079796,
      "num_tokens": 92492145.0,
      "step": 41425
    },
    {
      "entropy": 5.676679372787476,
      "epoch": 4.653001628573033,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003037161261472078,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2279564067721367,
      "num_tokens": 92504541.0,
      "step": 41430
    },
    {
      "entropy": 5.735292959213257,
      "epoch": 4.653563205480991,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030367630529913554,
      "loss": 5.0095,
      "mean_token_accuracy": 0.22780987173318862,
      "num_tokens": 92516911.0,
      "step": 41435
    },
    {
      "entropy": 5.746888542175293,
      "epoch": 4.654124782388948,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030363648353799857,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22717862725257873,
      "num_tokens": 92526619.0,
      "step": 41440
    },
    {
      "entropy": 5.771391534805298,
      "epoch": 4.654686359296906,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030359666086506493,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2276647299528122,
      "num_tokens": 92536579.0,
      "step": 41445
    },
    {
      "entropy": 5.6733907699584964,
      "epoch": 4.655247936204863,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030355683728160247,
      "loss": 4.8659,
      "mean_token_accuracy": 0.23907644152641297,
      "num_tokens": 92548147.0,
      "step": 41450
    },
    {
      "entropy": 5.6803281784057615,
      "epoch": 4.655809513112821,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003035170127888793,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2323070541024208,
      "num_tokens": 92559335.0,
      "step": 41455
    },
    {
      "entropy": 5.680207395553589,
      "epoch": 4.656371090020778,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030347718738816333,
      "loss": 4.864,
      "mean_token_accuracy": 0.2280697613954544,
      "num_tokens": 92571527.0,
      "step": 41460
    },
    {
      "entropy": 5.716269683837891,
      "epoch": 4.656932666928736,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030343736108072266,
      "loss": 4.9609,
      "mean_token_accuracy": 0.23386631309986114,
      "num_tokens": 92581996.0,
      "step": 41465
    },
    {
      "entropy": 5.65674467086792,
      "epoch": 4.657494243836694,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003033975338678254,
      "loss": 4.8901,
      "mean_token_accuracy": 0.23762264400720595,
      "num_tokens": 92593134.0,
      "step": 41470
    },
    {
      "entropy": 5.656566190719604,
      "epoch": 4.658055820744651,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030335770575073967,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2317889466881752,
      "num_tokens": 92605196.0,
      "step": 41475
    },
    {
      "entropy": 5.6537432193756105,
      "epoch": 4.658617397652608,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003033178767307336,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22594739496707916,
      "num_tokens": 92615151.0,
      "step": 41480
    },
    {
      "entropy": 5.707935953140259,
      "epoch": 4.659178974560566,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003032780468090753,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2302631437778473,
      "num_tokens": 92625723.0,
      "step": 41485
    },
    {
      "entropy": 5.720123434066773,
      "epoch": 4.6597405514685235,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003032382159870331,
      "loss": 5.0137,
      "mean_token_accuracy": 0.2334391936659813,
      "num_tokens": 92636699.0,
      "step": 41490
    },
    {
      "entropy": 5.6919190883636475,
      "epoch": 4.660302128376481,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030319838426587505,
      "loss": 4.9731,
      "mean_token_accuracy": 0.22393725365400313,
      "num_tokens": 92646977.0,
      "step": 41495
    },
    {
      "entropy": 5.637393426895142,
      "epoch": 4.660863705284439,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030315855164686956,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23785743713378907,
      "num_tokens": 92658021.0,
      "step": 41500
    },
    {
      "entropy": 5.627028274536133,
      "epoch": 4.661425282192396,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003031187181312848,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23941361755132676,
      "num_tokens": 92668573.0,
      "step": 41505
    },
    {
      "entropy": 5.702692174911499,
      "epoch": 4.661986859100354,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030307888372038927,
      "loss": 4.9781,
      "mean_token_accuracy": 0.2349554106593132,
      "num_tokens": 92680286.0,
      "step": 41510
    },
    {
      "entropy": 5.679331731796265,
      "epoch": 4.662548436008311,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030303904841545113,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22534120082855225,
      "num_tokens": 92692309.0,
      "step": 41515
    },
    {
      "entropy": 5.6138505935668945,
      "epoch": 4.663110012916269,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003029992122177388,
      "loss": 4.8924,
      "mean_token_accuracy": 0.2373897299170494,
      "num_tokens": 92703194.0,
      "step": 41520
    },
    {
      "entropy": 5.807650184631347,
      "epoch": 4.663671589824227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030295937512852075,
      "loss": 5.0619,
      "mean_token_accuracy": 0.22571446895599365,
      "num_tokens": 92714218.0,
      "step": 41525
    },
    {
      "entropy": 5.7299586772918705,
      "epoch": 4.664233166732184,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003029195371490652,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23644038736820222,
      "num_tokens": 92726422.0,
      "step": 41530
    },
    {
      "entropy": 5.682104539871216,
      "epoch": 4.664794743640142,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030287969828064087,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22656573057174684,
      "num_tokens": 92736835.0,
      "step": 41535
    },
    {
      "entropy": 5.690357446670532,
      "epoch": 4.665356320548099,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030283985852451624,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23469751924276352,
      "num_tokens": 92746665.0,
      "step": 41540
    },
    {
      "entropy": 5.707977342605591,
      "epoch": 4.6659178974560565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003028000178819596,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23119032979011536,
      "num_tokens": 92758059.0,
      "step": 41545
    },
    {
      "entropy": 5.6190708637237545,
      "epoch": 4.666479474364014,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030276017635423965,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23498113602399825,
      "num_tokens": 92768982.0,
      "step": 41550
    },
    {
      "entropy": 5.620395135879517,
      "epoch": 4.667041051271972,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030272033394262483,
      "loss": 4.8218,
      "mean_token_accuracy": 0.236016845703125,
      "num_tokens": 92779117.0,
      "step": 41555
    },
    {
      "entropy": 5.6564394474029545,
      "epoch": 4.667602628179929,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030268049064838395,
      "loss": 4.941,
      "mean_token_accuracy": 0.2278914511203766,
      "num_tokens": 92790595.0,
      "step": 41560
    },
    {
      "entropy": 5.720668172836303,
      "epoch": 4.668164205087887,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030264064647278537,
      "loss": 4.9746,
      "mean_token_accuracy": 0.23578492999076844,
      "num_tokens": 92801661.0,
      "step": 41565
    },
    {
      "entropy": 5.59015793800354,
      "epoch": 4.668725781995844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000302600801417098,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23633761405944825,
      "num_tokens": 92813223.0,
      "step": 41570
    },
    {
      "entropy": 5.635608243942261,
      "epoch": 4.6692873589038015,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030256095548259035,
      "loss": 4.9459,
      "mean_token_accuracy": 0.23074203133583068,
      "num_tokens": 92823839.0,
      "step": 41575
    },
    {
      "entropy": 5.57327184677124,
      "epoch": 4.66984893581176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003025211086705312,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23466433435678483,
      "num_tokens": 92834739.0,
      "step": 41580
    },
    {
      "entropy": 5.708894824981689,
      "epoch": 4.670410512719717,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003024812609821893,
      "loss": 4.9518,
      "mean_token_accuracy": 0.23089035600423813,
      "num_tokens": 92846369.0,
      "step": 41585
    },
    {
      "entropy": 5.632703590393066,
      "epoch": 4.670972089627675,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003024414124188333,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23503064811229707,
      "num_tokens": 92857963.0,
      "step": 41590
    },
    {
      "entropy": 5.6823889255523685,
      "epoch": 4.671533666535632,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003024015629817322,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22984385937452317,
      "num_tokens": 92868922.0,
      "step": 41595
    },
    {
      "entropy": 5.664804887771607,
      "epoch": 4.672095243443589,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003023617126721547,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23024675250053406,
      "num_tokens": 92881107.0,
      "step": 41600
    },
    {
      "entropy": 5.705058813095093,
      "epoch": 4.6726568203515475,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003023218614913696,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2312229558825493,
      "num_tokens": 92891538.0,
      "step": 41605
    },
    {
      "entropy": 5.695948553085327,
      "epoch": 4.673218397259505,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003022820094406458,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22123107314109802,
      "num_tokens": 92903015.0,
      "step": 41610
    },
    {
      "entropy": 5.662689065933227,
      "epoch": 4.673779974167462,
      "grad_norm": 1.125,
      "learning_rate": 0.0003022421565212524,
      "loss": 4.8648,
      "mean_token_accuracy": 0.23268385976552963,
      "num_tokens": 92914237.0,
      "step": 41615
    },
    {
      "entropy": 5.6773216247558596,
      "epoch": 4.67434155107542,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030220230273445806,
      "loss": 4.8752,
      "mean_token_accuracy": 0.2329510197043419,
      "num_tokens": 92925351.0,
      "step": 41620
    },
    {
      "entropy": 5.667792844772339,
      "epoch": 4.674903127983377,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030216244808153186,
      "loss": 4.9307,
      "mean_token_accuracy": 0.23570711612701417,
      "num_tokens": 92935472.0,
      "step": 41625
    },
    {
      "entropy": 5.717830228805542,
      "epoch": 4.6754647048913345,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003021225925637428,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2201578676700592,
      "num_tokens": 92948328.0,
      "step": 41630
    },
    {
      "entropy": 5.6510618209838865,
      "epoch": 4.676026281799293,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003020827361823598,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2356917977333069,
      "num_tokens": 92958458.0,
      "step": 41635
    },
    {
      "entropy": 5.700226068496704,
      "epoch": 4.67658785870725,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030204287893865204,
      "loss": 4.9493,
      "mean_token_accuracy": 0.22937211245298386,
      "num_tokens": 92969870.0,
      "step": 41640
    },
    {
      "entropy": 5.642238330841065,
      "epoch": 4.677149435615208,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030200302083388856,
      "loss": 4.9226,
      "mean_token_accuracy": 0.22960188388824462,
      "num_tokens": 92981597.0,
      "step": 41645
    },
    {
      "entropy": 5.7396204471588135,
      "epoch": 4.677711012523165,
      "grad_norm": 1.3125,
      "learning_rate": 0.00030196316186933834,
      "loss": 4.9785,
      "mean_token_accuracy": 0.23513265550136567,
      "num_tokens": 92992525.0,
      "step": 41650
    },
    {
      "entropy": 5.696923303604126,
      "epoch": 4.678272589431122,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030192330204627054,
      "loss": 4.9288,
      "mean_token_accuracy": 0.23420287668704987,
      "num_tokens": 93003652.0,
      "step": 41655
    },
    {
      "entropy": 5.7869397640228275,
      "epoch": 4.6788341663390804,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003018834413659545,
      "loss": 5.0108,
      "mean_token_accuracy": 0.2303907096385956,
      "num_tokens": 93015794.0,
      "step": 41660
    },
    {
      "entropy": 5.640716171264648,
      "epoch": 4.679395743247038,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003018435798296592,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23594279140233992,
      "num_tokens": 93026454.0,
      "step": 41665
    },
    {
      "entropy": 5.62626576423645,
      "epoch": 4.679957320154995,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003018037174386539,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23096589148044586,
      "num_tokens": 93039263.0,
      "step": 41670
    },
    {
      "entropy": 5.717565488815308,
      "epoch": 4.680518897062953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003017638541942078,
      "loss": 4.9526,
      "mean_token_accuracy": 0.23243594616651536,
      "num_tokens": 93050437.0,
      "step": 41675
    },
    {
      "entropy": 5.734341049194336,
      "epoch": 4.68108047397091,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030172399009759034,
      "loss": 4.972,
      "mean_token_accuracy": 0.22853637486696243,
      "num_tokens": 93060602.0,
      "step": 41680
    },
    {
      "entropy": 5.725896835327148,
      "epoch": 4.681642050878867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003016841251500705,
      "loss": 4.9844,
      "mean_token_accuracy": 0.22837699502706527,
      "num_tokens": 93073062.0,
      "step": 41685
    },
    {
      "entropy": 5.696927881240844,
      "epoch": 4.6822036277868255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030164425935291794,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2303851619362831,
      "num_tokens": 93083150.0,
      "step": 41690
    },
    {
      "entropy": 5.720427513122559,
      "epoch": 4.682765204694783,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003016043927074018,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22400375902652742,
      "num_tokens": 93094506.0,
      "step": 41695
    },
    {
      "entropy": 5.597066116333008,
      "epoch": 4.683326781602741,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003015645252147914,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23420009911060333,
      "num_tokens": 93105148.0,
      "step": 41700
    },
    {
      "entropy": 5.636464786529541,
      "epoch": 4.683888358510698,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003015246568763562,
      "loss": 4.9591,
      "mean_token_accuracy": 0.22614929974079132,
      "num_tokens": 93116470.0,
      "step": 41705
    },
    {
      "entropy": 5.620119333267212,
      "epoch": 4.684449935418655,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030148478769336564,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22818871438503266,
      "num_tokens": 93126662.0,
      "step": 41710
    },
    {
      "entropy": 5.713389921188354,
      "epoch": 4.685011512326613,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003014449176670892,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22816567867994308,
      "num_tokens": 93137690.0,
      "step": 41715
    },
    {
      "entropy": 5.629846572875977,
      "epoch": 4.685573089234571,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030140504679879637,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2327523946762085,
      "num_tokens": 93149760.0,
      "step": 41720
    },
    {
      "entropy": 5.673941135406494,
      "epoch": 4.686134666142529,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003013651750897566,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2407824844121933,
      "num_tokens": 93162046.0,
      "step": 41725
    },
    {
      "entropy": 5.714065313339233,
      "epoch": 4.686696243050486,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003013253025412393,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23684634417295455,
      "num_tokens": 93172912.0,
      "step": 41730
    },
    {
      "entropy": 5.7254064083099365,
      "epoch": 4.687257819958443,
      "grad_norm": 1.25,
      "learning_rate": 0.00030128542915451427,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22296549379825592,
      "num_tokens": 93184595.0,
      "step": 41735
    },
    {
      "entropy": 5.667767477035523,
      "epoch": 4.6878193968664,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030124555493085095,
      "loss": 4.8783,
      "mean_token_accuracy": 0.2333732396364212,
      "num_tokens": 93194753.0,
      "step": 41740
    },
    {
      "entropy": 5.699838066101075,
      "epoch": 4.6883809737743585,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003012056798715189,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2338396966457367,
      "num_tokens": 93204259.0,
      "step": 41745
    },
    {
      "entropy": 5.660109519958496,
      "epoch": 4.688942550682316,
      "grad_norm": 1.234375,
      "learning_rate": 0.000301165803977788,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2423437714576721,
      "num_tokens": 93214984.0,
      "step": 41750
    },
    {
      "entropy": 5.651834201812744,
      "epoch": 4.689504127590274,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003011259272509277,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2270142465829849,
      "num_tokens": 93227004.0,
      "step": 41755
    },
    {
      "entropy": 5.789463186264038,
      "epoch": 4.690065704498231,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003010860496922077,
      "loss": 5.0375,
      "mean_token_accuracy": 0.22164745032787322,
      "num_tokens": 93238485.0,
      "step": 41760
    },
    {
      "entropy": 5.71214451789856,
      "epoch": 4.690627281406188,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030104617130289783,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22819675356149674,
      "num_tokens": 93249318.0,
      "step": 41765
    },
    {
      "entropy": 5.7278073787689205,
      "epoch": 4.691188858314146,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003010062920842677,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2370499312877655,
      "num_tokens": 93262040.0,
      "step": 41770
    },
    {
      "entropy": 5.686694145202637,
      "epoch": 4.6917504352221036,
      "grad_norm": 1.21875,
      "learning_rate": 0.00030096641203758716,
      "loss": 5.0136,
      "mean_token_accuracy": 0.22583015412092208,
      "num_tokens": 93272766.0,
      "step": 41775
    },
    {
      "entropy": 5.796385383605957,
      "epoch": 4.692312012130062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030092653116412605,
      "loss": 4.981,
      "mean_token_accuracy": 0.23367751240730286,
      "num_tokens": 93283179.0,
      "step": 41780
    },
    {
      "entropy": 5.703200483322144,
      "epoch": 4.692873589038019,
      "grad_norm": 1.203125,
      "learning_rate": 0.000300886649465154,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22979574054479598,
      "num_tokens": 93293484.0,
      "step": 41785
    },
    {
      "entropy": 5.657589626312256,
      "epoch": 4.693435165945976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003008467669419411,
      "loss": 4.9049,
      "mean_token_accuracy": 0.227199923992157,
      "num_tokens": 93305611.0,
      "step": 41790
    },
    {
      "entropy": 5.729392957687378,
      "epoch": 4.693996742853934,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003008068835957571,
      "loss": 4.8909,
      "mean_token_accuracy": 0.23201024830341338,
      "num_tokens": 93316495.0,
      "step": 41795
    },
    {
      "entropy": 5.674944829940796,
      "epoch": 4.694558319761891,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003007669994278719,
      "loss": 4.8582,
      "mean_token_accuracy": 0.24040453881025314,
      "num_tokens": 93328484.0,
      "step": 41800
    },
    {
      "entropy": 5.713298797607422,
      "epoch": 4.695119896669849,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003007271144395554,
      "loss": 4.9826,
      "mean_token_accuracy": 0.22384965866804124,
      "num_tokens": 93339826.0,
      "step": 41805
    },
    {
      "entropy": 5.673054885864258,
      "epoch": 4.695681473577807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003006872286320777,
      "loss": 4.939,
      "mean_token_accuracy": 0.23839107155799866,
      "num_tokens": 93351923.0,
      "step": 41810
    },
    {
      "entropy": 5.649516534805298,
      "epoch": 4.696243050485764,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003006473420067086,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23652019947767258,
      "num_tokens": 93363813.0,
      "step": 41815
    },
    {
      "entropy": 5.70139627456665,
      "epoch": 4.696804627393721,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003006074545647182,
      "loss": 5.0162,
      "mean_token_accuracy": 0.2362995520234108,
      "num_tokens": 93374905.0,
      "step": 41820
    },
    {
      "entropy": 5.73781533241272,
      "epoch": 4.697366204301679,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030056756630737655,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2296364948153496,
      "num_tokens": 93386473.0,
      "step": 41825
    },
    {
      "entropy": 5.710792779922485,
      "epoch": 4.6979277812096365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00030052767723595365,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2289835050702095,
      "num_tokens": 93398010.0,
      "step": 41830
    },
    {
      "entropy": 5.736563396453858,
      "epoch": 4.698489358117595,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003004877873517196,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2238424926996231,
      "num_tokens": 93410692.0,
      "step": 41835
    },
    {
      "entropy": 5.682119655609131,
      "epoch": 4.699050935025552,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00030044789665594444,
      "loss": 4.8982,
      "mean_token_accuracy": 0.2261192351579666,
      "num_tokens": 93421830.0,
      "step": 41840
    },
    {
      "entropy": 5.7883391857147215,
      "epoch": 4.699612511933509,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003004080051498985,
      "loss": 5.0791,
      "mean_token_accuracy": 0.22008128315210343,
      "num_tokens": 93434136.0,
      "step": 41845
    },
    {
      "entropy": 5.639190196990967,
      "epoch": 4.700174088841467,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030036811283485184,
      "loss": 4.8376,
      "mean_token_accuracy": 0.24380816370248795,
      "num_tokens": 93444559.0,
      "step": 41850
    },
    {
      "entropy": 5.695085763931274,
      "epoch": 4.700735665749424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030032821971207453,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22869841903448104,
      "num_tokens": 93456258.0,
      "step": 41855
    },
    {
      "entropy": 5.726916790008545,
      "epoch": 4.701297242657382,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003002883257828369,
      "loss": 4.9217,
      "mean_token_accuracy": 0.23085129261016846,
      "num_tokens": 93467402.0,
      "step": 41860
    },
    {
      "entropy": 5.69914665222168,
      "epoch": 4.70185881956534,
      "grad_norm": 1.125,
      "learning_rate": 0.00030024843104840924,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22639172673225402,
      "num_tokens": 93477515.0,
      "step": 41865
    },
    {
      "entropy": 5.730254745483398,
      "epoch": 4.702420396473297,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030020853551006164,
      "loss": 4.9971,
      "mean_token_accuracy": 0.2264176458120346,
      "num_tokens": 93488173.0,
      "step": 41870
    },
    {
      "entropy": 5.669711542129517,
      "epoch": 4.702981973381254,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003001686391690645,
      "loss": 4.8186,
      "mean_token_accuracy": 0.2390141487121582,
      "num_tokens": 93498236.0,
      "step": 41875
    },
    {
      "entropy": 5.663440370559693,
      "epoch": 4.703543550289212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030012874202668823,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23411034792661667,
      "num_tokens": 93510162.0,
      "step": 41880
    },
    {
      "entropy": 5.720016288757324,
      "epoch": 4.7041051271971694,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030008884408420293,
      "loss": 4.9715,
      "mean_token_accuracy": 0.23108863234519958,
      "num_tokens": 93521344.0,
      "step": 41885
    },
    {
      "entropy": 5.7269322872161865,
      "epoch": 4.7046667041051276,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003000489453428792,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2268797531723976,
      "num_tokens": 93533116.0,
      "step": 41890
    },
    {
      "entropy": 5.728123903274536,
      "epoch": 4.705228281013085,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003000090458039873,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2331618383526802,
      "num_tokens": 93544612.0,
      "step": 41895
    },
    {
      "entropy": 5.735495233535767,
      "epoch": 4.705789857921042,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029996914546879773,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22895953059196472,
      "num_tokens": 93556033.0,
      "step": 41900
    },
    {
      "entropy": 5.764012098312378,
      "epoch": 4.706351434829,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002999292443385809,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2205958306789398,
      "num_tokens": 93567250.0,
      "step": 41905
    },
    {
      "entropy": 5.738847398757935,
      "epoch": 4.706913011736957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002998893424146072,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22203452736139298,
      "num_tokens": 93578521.0,
      "step": 41910
    },
    {
      "entropy": 5.734733629226684,
      "epoch": 4.707474588644915,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002998494396981472,
      "loss": 4.9784,
      "mean_token_accuracy": 0.23406610637903214,
      "num_tokens": 93589809.0,
      "step": 41915
    },
    {
      "entropy": 5.646167421340943,
      "epoch": 4.708036165552873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002998095361904714,
      "loss": 4.8416,
      "mean_token_accuracy": 0.24219832569360733,
      "num_tokens": 93599688.0,
      "step": 41920
    },
    {
      "entropy": 5.60947618484497,
      "epoch": 4.70859774246083,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002997696318928503,
      "loss": 4.9315,
      "mean_token_accuracy": 0.23316064476966858,
      "num_tokens": 93610575.0,
      "step": 41925
    },
    {
      "entropy": 5.62171516418457,
      "epoch": 4.709159319368787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002997297268065546,
      "loss": 4.8808,
      "mean_token_accuracy": 0.24039683640003204,
      "num_tokens": 93622329.0,
      "step": 41930
    },
    {
      "entropy": 5.647642993927002,
      "epoch": 4.709720896276745,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029968982093285475,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22997348457574845,
      "num_tokens": 93632686.0,
      "step": 41935
    },
    {
      "entropy": 5.767669343948365,
      "epoch": 4.710282473184702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029964991427302147,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23027051538228988,
      "num_tokens": 93643086.0,
      "step": 41940
    },
    {
      "entropy": 5.776145076751709,
      "epoch": 4.7108440500926605,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002996100068283253,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2241852954030037,
      "num_tokens": 93654457.0,
      "step": 41945
    },
    {
      "entropy": 5.7283220291137695,
      "epoch": 4.711405627000618,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029957009860003697,
      "loss": 4.9888,
      "mean_token_accuracy": 0.22208400070667267,
      "num_tokens": 93665806.0,
      "step": 41950
    },
    {
      "entropy": 5.671831750869751,
      "epoch": 4.711967203908575,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029953018958942717,
      "loss": 4.9587,
      "mean_token_accuracy": 0.223895063996315,
      "num_tokens": 93676685.0,
      "step": 41955
    },
    {
      "entropy": 5.7653522968292235,
      "epoch": 4.712528780816533,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002994902797977667,
      "loss": 5.0477,
      "mean_token_accuracy": 0.22288326621055604,
      "num_tokens": 93688111.0,
      "step": 41960
    },
    {
      "entropy": 5.703473854064941,
      "epoch": 4.71309035772449,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029945036922632615,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22930506020784377,
      "num_tokens": 93698919.0,
      "step": 41965
    },
    {
      "entropy": 5.716944742202759,
      "epoch": 4.713651934632448,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029941045787637636,
      "loss": 4.9594,
      "mean_token_accuracy": 0.23475779294967652,
      "num_tokens": 93711367.0,
      "step": 41970
    },
    {
      "entropy": 5.680512619018555,
      "epoch": 4.714213511540406,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002993705457491882,
      "loss": 4.9132,
      "mean_token_accuracy": 0.23264397233724593,
      "num_tokens": 93722649.0,
      "step": 41975
    },
    {
      "entropy": 5.716482925415039,
      "epoch": 4.714775088448363,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029933063284603235,
      "loss": 5.0086,
      "mean_token_accuracy": 0.2297607406973839,
      "num_tokens": 93733873.0,
      "step": 41980
    },
    {
      "entropy": 5.657264518737793,
      "epoch": 4.715336665356321,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002992907191681797,
      "loss": 4.7821,
      "mean_token_accuracy": 0.2448120415210724,
      "num_tokens": 93744954.0,
      "step": 41985
    },
    {
      "entropy": 5.7066998958587645,
      "epoch": 4.715898242264278,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029925080471690117,
      "loss": 4.9221,
      "mean_token_accuracy": 0.2283354014158249,
      "num_tokens": 93756607.0,
      "step": 41990
    },
    {
      "entropy": 5.630630588531494,
      "epoch": 4.716459819172235,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029921088949346765,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23203436881303788,
      "num_tokens": 93767496.0,
      "step": 41995
    },
    {
      "entropy": 5.620767211914062,
      "epoch": 4.717021396080193,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002991709734991499,
      "loss": 4.8598,
      "mean_token_accuracy": 0.225998218357563,
      "num_tokens": 93778432.0,
      "step": 42000
    },
    {
      "epoch": 4.717021396080193,
      "eval_entropy": 5.418499280842488,
      "eval_loss": 5.055675029754639,
      "eval_mean_token_accuracy": 0.2321538851529875,
      "eval_num_tokens": 93778432.0,
      "eval_runtime": 18.4181,
      "eval_samples_per_second": 1683.836,
      "eval_steps_per_second": 210.5,
      "step": 42000
    },
    {
      "entropy": 5.656275272369385,
      "epoch": 4.717582972988151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029913105673521907,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23047299087047576,
      "num_tokens": 93789652.0,
      "step": 42005
    },
    {
      "entropy": 5.699820852279663,
      "epoch": 4.718144549896108,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000299091139202946,
      "loss": 4.891,
      "mean_token_accuracy": 0.23276606053113938,
      "num_tokens": 93800386.0,
      "step": 42010
    },
    {
      "entropy": 5.734050321578979,
      "epoch": 4.718706126804066,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029905122090360167,
      "loss": 4.9421,
      "mean_token_accuracy": 0.23269993513822557,
      "num_tokens": 93810184.0,
      "step": 42015
    },
    {
      "entropy": 5.67419023513794,
      "epoch": 4.719267703712023,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002990113018384572,
      "loss": 4.8721,
      "mean_token_accuracy": 0.23384375125169754,
      "num_tokens": 93820139.0,
      "step": 42020
    },
    {
      "entropy": 5.683533763885498,
      "epoch": 4.719829280619981,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002989713820087836,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23728434294462203,
      "num_tokens": 93831323.0,
      "step": 42025
    },
    {
      "entropy": 5.728961610794068,
      "epoch": 4.7203908575279385,
      "grad_norm": 1.25,
      "learning_rate": 0.00029893146141585187,
      "loss": 4.939,
      "mean_token_accuracy": 0.23108334243297576,
      "num_tokens": 93841160.0,
      "step": 42030
    },
    {
      "entropy": 5.7554737567901615,
      "epoch": 4.720952434435896,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029889154006093306,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2189517453312874,
      "num_tokens": 93851530.0,
      "step": 42035
    },
    {
      "entropy": 5.742127180099487,
      "epoch": 4.721514011343854,
      "grad_norm": 1.25,
      "learning_rate": 0.0002988516179452984,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2381913959980011,
      "num_tokens": 93862466.0,
      "step": 42040
    },
    {
      "entropy": 5.747883319854736,
      "epoch": 4.722075588251811,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029881169507021895,
      "loss": 4.9154,
      "mean_token_accuracy": 0.24238671064376832,
      "num_tokens": 93874399.0,
      "step": 42045
    },
    {
      "entropy": 5.703614664077759,
      "epoch": 4.722637165159768,
      "grad_norm": 1.140625,
      "learning_rate": 0.000298771771436966,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23531084060668944,
      "num_tokens": 93885054.0,
      "step": 42050
    },
    {
      "entropy": 5.676351404190063,
      "epoch": 4.723198742067726,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002987318470468105,
      "loss": 4.8762,
      "mean_token_accuracy": 0.24497703909873964,
      "num_tokens": 93894813.0,
      "step": 42055
    },
    {
      "entropy": 5.587923288345337,
      "epoch": 4.723760318975684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029869192190102385,
      "loss": 4.8981,
      "mean_token_accuracy": 0.24063005149364472,
      "num_tokens": 93905625.0,
      "step": 42060
    },
    {
      "entropy": 5.629213953018189,
      "epoch": 4.724321895883641,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029865199600087724,
      "loss": 4.9164,
      "mean_token_accuracy": 0.23485940396785737,
      "num_tokens": 93917539.0,
      "step": 42065
    },
    {
      "entropy": 5.7007452011108395,
      "epoch": 4.724883472791599,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002986120693476418,
      "loss": 4.9002,
      "mean_token_accuracy": 0.23062665909528732,
      "num_tokens": 93927307.0,
      "step": 42070
    },
    {
      "entropy": 5.707657384872436,
      "epoch": 4.725445049699556,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029857214194258896,
      "loss": 4.9128,
      "mean_token_accuracy": 0.23804556280374528,
      "num_tokens": 93938252.0,
      "step": 42075
    },
    {
      "entropy": 5.710659790039062,
      "epoch": 4.726006626607514,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029853221378699,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23162503242492677,
      "num_tokens": 93947762.0,
      "step": 42080
    },
    {
      "entropy": 5.712592744827271,
      "epoch": 4.7265682035154715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002984922848821162,
      "loss": 5.0192,
      "mean_token_accuracy": 0.22740862369537354,
      "num_tokens": 93958410.0,
      "step": 42085
    },
    {
      "entropy": 5.693926572799683,
      "epoch": 4.727129780423429,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029845235522923903,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23458654284477234,
      "num_tokens": 93969610.0,
      "step": 42090
    },
    {
      "entropy": 5.7103088855743405,
      "epoch": 4.727691357331387,
      "grad_norm": 1.46875,
      "learning_rate": 0.00029841242482962966,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2294575035572052,
      "num_tokens": 93978404.0,
      "step": 42095
    },
    {
      "entropy": 5.620197057723999,
      "epoch": 4.728252934239344,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029837249368455963,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22772636413574218,
      "num_tokens": 93989578.0,
      "step": 42100
    },
    {
      "entropy": 5.738563108444214,
      "epoch": 4.728814511147302,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002983325617953004,
      "loss": 5.0097,
      "mean_token_accuracy": 0.22475008964538573,
      "num_tokens": 94000010.0,
      "step": 42105
    },
    {
      "entropy": 5.705191612243652,
      "epoch": 4.729376088055259,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029829262916312326,
      "loss": 4.8808,
      "mean_token_accuracy": 0.23645493239164353,
      "num_tokens": 94009386.0,
      "step": 42110
    },
    {
      "entropy": 5.627643060684204,
      "epoch": 4.7299376649632165,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029825269578929986,
      "loss": 4.8319,
      "mean_token_accuracy": 0.236871075630188,
      "num_tokens": 94021594.0,
      "step": 42115
    },
    {
      "entropy": 5.713966417312622,
      "epoch": 4.730499241871174,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029821276167510153,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22727310806512832,
      "num_tokens": 94033518.0,
      "step": 42120
    },
    {
      "entropy": 5.653714227676391,
      "epoch": 4.731060818779132,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002981728268217999,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2246365636587143,
      "num_tokens": 94044174.0,
      "step": 42125
    },
    {
      "entropy": 5.685425472259522,
      "epoch": 4.731622395687089,
      "grad_norm": 1.25,
      "learning_rate": 0.0002981328912306665,
      "loss": 4.8937,
      "mean_token_accuracy": 0.23306729942560195,
      "num_tokens": 94056000.0,
      "step": 42130
    },
    {
      "entropy": 5.766964244842529,
      "epoch": 4.732183972595047,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029809295490297285,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2298567622900009,
      "num_tokens": 94067274.0,
      "step": 42135
    },
    {
      "entropy": 5.681794786453247,
      "epoch": 4.732745549503004,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002980530178399906,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22744949907064438,
      "num_tokens": 94079113.0,
      "step": 42140
    },
    {
      "entropy": 5.649424028396607,
      "epoch": 4.733307126410962,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002980130800429912,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22832608520984649,
      "num_tokens": 94090438.0,
      "step": 42145
    },
    {
      "entropy": 5.665004062652588,
      "epoch": 4.73386870331892,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029797314151324653,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22809159457683564,
      "num_tokens": 94101153.0,
      "step": 42150
    },
    {
      "entropy": 5.797663640975952,
      "epoch": 4.734430280226877,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029793320225202807,
      "loss": 5.0686,
      "mean_token_accuracy": 0.22796513885259628,
      "num_tokens": 94112366.0,
      "step": 42155
    },
    {
      "entropy": 5.669768285751343,
      "epoch": 4.734991857134835,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002978932622606075,
      "loss": 4.9379,
      "mean_token_accuracy": 0.23147801011800767,
      "num_tokens": 94122919.0,
      "step": 42160
    },
    {
      "entropy": 5.7039210319519045,
      "epoch": 4.735553434042792,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029785332154025676,
      "loss": 5.0248,
      "mean_token_accuracy": 0.2298448473215103,
      "num_tokens": 94135158.0,
      "step": 42165
    },
    {
      "entropy": 5.722611379623413,
      "epoch": 4.7361150109507495,
      "grad_norm": 1.25,
      "learning_rate": 0.0002978133800922473,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2211513713002205,
      "num_tokens": 94145773.0,
      "step": 42170
    },
    {
      "entropy": 5.634186124801635,
      "epoch": 4.736676587858708,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002977734379178509,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2330506071448326,
      "num_tokens": 94157260.0,
      "step": 42175
    },
    {
      "entropy": 5.679907369613647,
      "epoch": 4.737238164766665,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002977334950183394,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2316728040575981,
      "num_tokens": 94166513.0,
      "step": 42180
    },
    {
      "entropy": 5.78766565322876,
      "epoch": 4.737799741674622,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029769355139498465,
      "loss": 5.0467,
      "mean_token_accuracy": 0.2113572433590889,
      "num_tokens": 94178565.0,
      "step": 42185
    },
    {
      "entropy": 5.661460590362549,
      "epoch": 4.73836131858258,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002976536070490585,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23372258394956588,
      "num_tokens": 94189319.0,
      "step": 42190
    },
    {
      "entropy": 5.72606406211853,
      "epoch": 4.738922895490537,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029761366198183264,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22948258072137834,
      "num_tokens": 94203030.0,
      "step": 42195
    },
    {
      "entropy": 5.718115949630738,
      "epoch": 4.739484472398495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029757371619457904,
      "loss": 4.982,
      "mean_token_accuracy": 0.22807787358760834,
      "num_tokens": 94214901.0,
      "step": 42200
    },
    {
      "entropy": 5.694693565368652,
      "epoch": 4.740046049306453,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029753376968856955,
      "loss": 4.896,
      "mean_token_accuracy": 0.2341851457953453,
      "num_tokens": 94226833.0,
      "step": 42205
    },
    {
      "entropy": 5.7452984809875485,
      "epoch": 4.74060762621441,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029749382246507604,
      "loss": 4.9446,
      "mean_token_accuracy": 0.23707537204027176,
      "num_tokens": 94239464.0,
      "step": 42210
    },
    {
      "entropy": 5.678625822067261,
      "epoch": 4.741169203122368,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029745387452537056,
      "loss": 4.8661,
      "mean_token_accuracy": 0.23965879827737807,
      "num_tokens": 94250692.0,
      "step": 42215
    },
    {
      "entropy": 5.632197666168213,
      "epoch": 4.741730780030325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029741392587072504,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22703367173671724,
      "num_tokens": 94261130.0,
      "step": 42220
    },
    {
      "entropy": 5.675814771652222,
      "epoch": 4.742292356938282,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029737397650241135,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22891814708709718,
      "num_tokens": 94271581.0,
      "step": 42225
    },
    {
      "entropy": 5.710898780822754,
      "epoch": 4.7428539338462405,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002973340264217016,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22846475541591643,
      "num_tokens": 94283353.0,
      "step": 42230
    },
    {
      "entropy": 5.722050285339355,
      "epoch": 4.743415510754198,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029729407562986774,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2270229399204254,
      "num_tokens": 94294669.0,
      "step": 42235
    },
    {
      "entropy": 5.714051866531372,
      "epoch": 4.743977087662155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029725412412818196,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23473307639360427,
      "num_tokens": 94306127.0,
      "step": 42240
    },
    {
      "entropy": 5.643300580978393,
      "epoch": 4.744538664570113,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00029721417191791617,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23501307964324952,
      "num_tokens": 94316621.0,
      "step": 42245
    },
    {
      "entropy": 5.681365966796875,
      "epoch": 4.74510024147807,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029717421900034253,
      "loss": 4.9397,
      "mean_token_accuracy": 0.23235629349946976,
      "num_tokens": 94327715.0,
      "step": 42250
    },
    {
      "entropy": 5.695126676559449,
      "epoch": 4.7456618183860275,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002971342653767332,
      "loss": 4.8906,
      "mean_token_accuracy": 0.23246212750673295,
      "num_tokens": 94339452.0,
      "step": 42255
    },
    {
      "entropy": 5.6894923686981205,
      "epoch": 4.746223395293986,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029709431104836027,
      "loss": 4.9148,
      "mean_token_accuracy": 0.23525465577840804,
      "num_tokens": 94350309.0,
      "step": 42260
    },
    {
      "entropy": 5.717938375473023,
      "epoch": 4.746784972201943,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029705435601649584,
      "loss": 5.0305,
      "mean_token_accuracy": 0.2196480616927147,
      "num_tokens": 94361665.0,
      "step": 42265
    },
    {
      "entropy": 5.827962875366211,
      "epoch": 4.747346549109901,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002970144002824122,
      "loss": 5.0314,
      "mean_token_accuracy": 0.22364278882741928,
      "num_tokens": 94374141.0,
      "step": 42270
    },
    {
      "entropy": 5.682801342010498,
      "epoch": 4.747908126017858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002969744438473815,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22189324498176574,
      "num_tokens": 94384379.0,
      "step": 42275
    },
    {
      "entropy": 5.661912250518799,
      "epoch": 4.748469702925815,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002969344867126761,
      "loss": 4.8496,
      "mean_token_accuracy": 0.23822988718748092,
      "num_tokens": 94394747.0,
      "step": 42280
    },
    {
      "entropy": 5.665245056152344,
      "epoch": 4.7490312798337735,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029689452887956804,
      "loss": 4.9912,
      "mean_token_accuracy": 0.223173850774765,
      "num_tokens": 94407530.0,
      "step": 42285
    },
    {
      "entropy": 5.72736349105835,
      "epoch": 4.749592856741731,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029685457034932966,
      "loss": 4.9894,
      "mean_token_accuracy": 0.22638170570135116,
      "num_tokens": 94420031.0,
      "step": 42290
    },
    {
      "entropy": 5.715549182891846,
      "epoch": 4.750154433649689,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002968146111232333,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22974250763654708,
      "num_tokens": 94429634.0,
      "step": 42295
    },
    {
      "entropy": 5.710513591766357,
      "epoch": 4.750716010557646,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029677465120255133,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2260780319571495,
      "num_tokens": 94440310.0,
      "step": 42300
    },
    {
      "entropy": 5.702475118637085,
      "epoch": 4.751277587465603,
      "grad_norm": 1.25,
      "learning_rate": 0.0002967346905885559,
      "loss": 5.0057,
      "mean_token_accuracy": 0.225387866795063,
      "num_tokens": 94451050.0,
      "step": 42305
    },
    {
      "entropy": 5.7217284679412845,
      "epoch": 4.7518391643735605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029669472928251954,
      "loss": 5.0076,
      "mean_token_accuracy": 0.23606036603450775,
      "num_tokens": 94462210.0,
      "step": 42310
    },
    {
      "entropy": 5.602719116210937,
      "epoch": 4.752400741281519,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029665476728571466,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23228590041399003,
      "num_tokens": 94472716.0,
      "step": 42315
    },
    {
      "entropy": 5.645095872879028,
      "epoch": 4.752962318189476,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002966148045994135,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23517972379922866,
      "num_tokens": 94482446.0,
      "step": 42320
    },
    {
      "entropy": 5.668193006515503,
      "epoch": 4.753523895097434,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002965748412248886,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2347920298576355,
      "num_tokens": 94492476.0,
      "step": 42325
    },
    {
      "entropy": 5.62559871673584,
      "epoch": 4.754085472005391,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002965348771634124,
      "loss": 4.8538,
      "mean_token_accuracy": 0.2342485949397087,
      "num_tokens": 94502607.0,
      "step": 42330
    },
    {
      "entropy": 5.6307337284088135,
      "epoch": 4.754647048913348,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029649491241625736,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23529257625341415,
      "num_tokens": 94512528.0,
      "step": 42335
    },
    {
      "entropy": 5.868231105804443,
      "epoch": 4.755208625821306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029645494698469604,
      "loss": 5.1208,
      "mean_token_accuracy": 0.22230370342731476,
      "num_tokens": 94524890.0,
      "step": 42340
    },
    {
      "entropy": 5.647023391723633,
      "epoch": 4.755770202729264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029641498087000083,
      "loss": 4.8209,
      "mean_token_accuracy": 0.24302908927202224,
      "num_tokens": 94536483.0,
      "step": 42345
    },
    {
      "entropy": 5.75216383934021,
      "epoch": 4.756331779637222,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002963750140734443,
      "loss": 5.0795,
      "mean_token_accuracy": 0.22079874575138092,
      "num_tokens": 94547235.0,
      "step": 42350
    },
    {
      "entropy": 5.744336795806885,
      "epoch": 4.756893356545179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002963350465962991,
      "loss": 4.9588,
      "mean_token_accuracy": 0.23111457079648973,
      "num_tokens": 94558511.0,
      "step": 42355
    },
    {
      "entropy": 5.659914493560791,
      "epoch": 4.757454933453136,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002962950784398376,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22753272205591202,
      "num_tokens": 94569949.0,
      "step": 42360
    },
    {
      "entropy": 5.6622357845306395,
      "epoch": 4.758016510361094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002962551096053327,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2302081987261772,
      "num_tokens": 94581353.0,
      "step": 42365
    },
    {
      "entropy": 5.633219861984253,
      "epoch": 4.7585780872690515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029621514009405675,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23576153069734573,
      "num_tokens": 94591653.0,
      "step": 42370
    },
    {
      "entropy": 5.678385496139526,
      "epoch": 4.759139664177009,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002961751699072826,
      "loss": 4.9606,
      "mean_token_accuracy": 0.22544604390859604,
      "num_tokens": 94603615.0,
      "step": 42375
    },
    {
      "entropy": 5.719779205322266,
      "epoch": 4.759701241084967,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002961351990462827,
      "loss": 5.0194,
      "mean_token_accuracy": 0.227880297601223,
      "num_tokens": 94614813.0,
      "step": 42380
    },
    {
      "entropy": 5.69309983253479,
      "epoch": 4.760262817992924,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029609522751232996,
      "loss": 4.826,
      "mean_token_accuracy": 0.23795610368251802,
      "num_tokens": 94624483.0,
      "step": 42385
    },
    {
      "entropy": 5.744231748580932,
      "epoch": 4.760824394900881,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029605525530669705,
      "loss": 4.9591,
      "mean_token_accuracy": 0.22906437516212463,
      "num_tokens": 94634332.0,
      "step": 42390
    },
    {
      "entropy": 5.659697151184082,
      "epoch": 4.761385971808839,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029601528243065656,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23505012392997743,
      "num_tokens": 94645019.0,
      "step": 42395
    },
    {
      "entropy": 5.7049524784088135,
      "epoch": 4.761947548716797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029597530888548135,
      "loss": 5.0174,
      "mean_token_accuracy": 0.22980674356222153,
      "num_tokens": 94656393.0,
      "step": 42400
    },
    {
      "entropy": 5.662168025970459,
      "epoch": 4.762509125624755,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00029593533467244415,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23509866297245025,
      "num_tokens": 94666730.0,
      "step": 42405
    },
    {
      "entropy": 5.646588563919067,
      "epoch": 4.763070702532712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002958953597928178,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22774817794561386,
      "num_tokens": 94677937.0,
      "step": 42410
    },
    {
      "entropy": 5.705776834487915,
      "epoch": 4.763632279440669,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002958553842478751,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2277164176106453,
      "num_tokens": 94688141.0,
      "step": 42415
    },
    {
      "entropy": 5.6888049125671385,
      "epoch": 4.764193856348627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029581540803888886,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2368408665060997,
      "num_tokens": 94700055.0,
      "step": 42420
    },
    {
      "entropy": 5.722445869445801,
      "epoch": 4.7647554332565845,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029577543116713197,
      "loss": 5.01,
      "mean_token_accuracy": 0.22507500797510147,
      "num_tokens": 94711858.0,
      "step": 42425
    },
    {
      "entropy": 5.700692987442016,
      "epoch": 4.765317010164542,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029573545363387725,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22772753238677979,
      "num_tokens": 94722691.0,
      "step": 42430
    },
    {
      "entropy": 5.7085095882415775,
      "epoch": 4.7658785870725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002956954754403977,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22767432332038878,
      "num_tokens": 94733570.0,
      "step": 42435
    },
    {
      "entropy": 5.681515789031982,
      "epoch": 4.766440163980457,
      "grad_norm": 1.25,
      "learning_rate": 0.0002956554965879662,
      "loss": 4.9069,
      "mean_token_accuracy": 0.23826039284467698,
      "num_tokens": 94744719.0,
      "step": 42440
    },
    {
      "entropy": 5.7565898418426515,
      "epoch": 4.767001740888414,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002956155170778556,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22150700241327287,
      "num_tokens": 94755272.0,
      "step": 42445
    },
    {
      "entropy": 5.790570449829102,
      "epoch": 4.767563317796372,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002955755369113389,
      "loss": 5.0591,
      "mean_token_accuracy": 0.22622178196907045,
      "num_tokens": 94767178.0,
      "step": 42450
    },
    {
      "entropy": 5.686270618438721,
      "epoch": 4.7681248947043295,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029553555608968925,
      "loss": 4.93,
      "mean_token_accuracy": 0.22574947029352188,
      "num_tokens": 94777088.0,
      "step": 42455
    },
    {
      "entropy": 5.768286609649659,
      "epoch": 4.768686471612288,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002954955746141795,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22475292831659316,
      "num_tokens": 94787995.0,
      "step": 42460
    },
    {
      "entropy": 5.727022266387939,
      "epoch": 4.769248048520245,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002954555924860827,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22997929900884628,
      "num_tokens": 94797810.0,
      "step": 42465
    },
    {
      "entropy": 5.744407844543457,
      "epoch": 4.769809625428202,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002954156097066718,
      "loss": 4.9339,
      "mean_token_accuracy": 0.23449183702468873,
      "num_tokens": 94809725.0,
      "step": 42470
    },
    {
      "entropy": 5.743777847290039,
      "epoch": 4.77037120233616,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002953756262772201,
      "loss": 4.9416,
      "mean_token_accuracy": 0.23356932997703553,
      "num_tokens": 94819740.0,
      "step": 42475
    },
    {
      "entropy": 5.7680213928222654,
      "epoch": 4.770932779244117,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002953356421990005,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2222810670733452,
      "num_tokens": 94830419.0,
      "step": 42480
    },
    {
      "entropy": 5.729334831237793,
      "epoch": 4.7714943561520755,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002952956574732862,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2223428890109062,
      "num_tokens": 94842796.0,
      "step": 42485
    },
    {
      "entropy": 5.73942232131958,
      "epoch": 4.772055933060033,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002952556721013503,
      "loss": 5.0243,
      "mean_token_accuracy": 0.23164262622594833,
      "num_tokens": 94854065.0,
      "step": 42490
    },
    {
      "entropy": 5.756405973434449,
      "epoch": 4.77261750996799,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002952156860844659,
      "loss": 4.8882,
      "mean_token_accuracy": 0.24084192812442778,
      "num_tokens": 94866264.0,
      "step": 42495
    },
    {
      "entropy": 5.730883407592773,
      "epoch": 4.773179086875947,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002951756994239063,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22592624723911287,
      "num_tokens": 94878549.0,
      "step": 42500
    },
    {
      "entropy": 5.734110116958618,
      "epoch": 4.773740663783905,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029513571212094443,
      "loss": 4.946,
      "mean_token_accuracy": 0.23279216289520263,
      "num_tokens": 94890527.0,
      "step": 42505
    },
    {
      "entropy": 5.6747253894805905,
      "epoch": 4.7743022406918625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002950957241768539,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22734506875276567,
      "num_tokens": 94901424.0,
      "step": 42510
    },
    {
      "entropy": 5.668261337280273,
      "epoch": 4.774863817599821,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029505573559290756,
      "loss": 4.882,
      "mean_token_accuracy": 0.23581726849079132,
      "num_tokens": 94912153.0,
      "step": 42515
    },
    {
      "entropy": 5.711395502090454,
      "epoch": 4.775425394507778,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029501574637037884,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23004238605499266,
      "num_tokens": 94923912.0,
      "step": 42520
    },
    {
      "entropy": 5.647719192504883,
      "epoch": 4.775986971415735,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000294975756510541,
      "loss": 4.8299,
      "mean_token_accuracy": 0.23794691264629364,
      "num_tokens": 94934768.0,
      "step": 42525
    },
    {
      "entropy": 5.6988921642303465,
      "epoch": 4.776548548323693,
      "grad_norm": 1.25,
      "learning_rate": 0.0002949357660146673,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21919696033000946,
      "num_tokens": 94946617.0,
      "step": 42530
    },
    {
      "entropy": 5.792539310455322,
      "epoch": 4.77711012523165,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002948957748840311,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22329760044813157,
      "num_tokens": 94958863.0,
      "step": 42535
    },
    {
      "entropy": 5.7499223232269285,
      "epoch": 4.7776717021396085,
      "grad_norm": 1.25,
      "learning_rate": 0.00029485578311990574,
      "loss": 4.9473,
      "mean_token_accuracy": 0.23350555449724197,
      "num_tokens": 94969930.0,
      "step": 42540
    },
    {
      "entropy": 5.722478771209717,
      "epoch": 4.778233279047566,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029481579072356456,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22631313800811767,
      "num_tokens": 94980858.0,
      "step": 42545
    },
    {
      "entropy": 5.8097734451293945,
      "epoch": 4.778794855955523,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002947757976962809,
      "loss": 5.0574,
      "mean_token_accuracy": 0.22448773831129074,
      "num_tokens": 94993075.0,
      "step": 42550
    },
    {
      "entropy": 5.653691720962525,
      "epoch": 4.779356432863481,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029473580403932814,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22867883294820784,
      "num_tokens": 95003849.0,
      "step": 42555
    },
    {
      "entropy": 5.7236145496368405,
      "epoch": 4.779918009771438,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002946958097539798,
      "loss": 4.9556,
      "mean_token_accuracy": 0.23006251007318496,
      "num_tokens": 95015766.0,
      "step": 42560
    },
    {
      "entropy": 5.682914066314697,
      "epoch": 4.780479586679395,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002946558148415092,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2299383357167244,
      "num_tokens": 95026690.0,
      "step": 42565
    },
    {
      "entropy": 5.662400007247925,
      "epoch": 4.7810411635873535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002946158193031898,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2387446016073227,
      "num_tokens": 95038591.0,
      "step": 42570
    },
    {
      "entropy": 5.7839727878570555,
      "epoch": 4.781602740495311,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029457582314029507,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2230371803045273,
      "num_tokens": 95050268.0,
      "step": 42575
    },
    {
      "entropy": 5.760484600067139,
      "epoch": 4.782164317403268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029453582635409866,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22926872074604035,
      "num_tokens": 95061448.0,
      "step": 42580
    },
    {
      "entropy": 5.734341812133789,
      "epoch": 4.782725894311226,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002944958289458739,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22769576013088227,
      "num_tokens": 95072331.0,
      "step": 42585
    },
    {
      "entropy": 5.721552133560181,
      "epoch": 4.783287471219183,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029445583091689437,
      "loss": 4.9485,
      "mean_token_accuracy": 0.22875531911849975,
      "num_tokens": 95082536.0,
      "step": 42590
    },
    {
      "entropy": 5.666303205490112,
      "epoch": 4.783849048127141,
      "grad_norm": 1.25,
      "learning_rate": 0.0002944158322684336,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23819584995508195,
      "num_tokens": 95093877.0,
      "step": 42595
    },
    {
      "entropy": 5.74659948348999,
      "epoch": 4.784410625035099,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029437583300176526,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2336931750178337,
      "num_tokens": 95104672.0,
      "step": 42600
    },
    {
      "entropy": 5.698288679122925,
      "epoch": 4.784972201943056,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029433583311816284,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22472794651985167,
      "num_tokens": 95115206.0,
      "step": 42605
    },
    {
      "entropy": 5.67097430229187,
      "epoch": 4.785533778851014,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002942958326189,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2272953823208809,
      "num_tokens": 95125142.0,
      "step": 42610
    },
    {
      "entropy": 5.704036808013916,
      "epoch": 4.786095355758971,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029425583150525044,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2297898843884468,
      "num_tokens": 95136108.0,
      "step": 42615
    },
    {
      "entropy": 5.702893400192261,
      "epoch": 4.786656932666928,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029421582977848763,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2282174199819565,
      "num_tokens": 95146452.0,
      "step": 42620
    },
    {
      "entropy": 5.589194583892822,
      "epoch": 4.7872185095748865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029417582743988547,
      "loss": 4.7276,
      "mean_token_accuracy": 0.23673201352357864,
      "num_tokens": 95156665.0,
      "step": 42625
    },
    {
      "entropy": 5.690122079849243,
      "epoch": 4.787780086482844,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029413582449071745,
      "loss": 4.9077,
      "mean_token_accuracy": 0.23169521391391754,
      "num_tokens": 95167092.0,
      "step": 42630
    },
    {
      "entropy": 5.6805966854095455,
      "epoch": 4.788341663390801,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002940958209322574,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23892539292573928,
      "num_tokens": 95178158.0,
      "step": 42635
    },
    {
      "entropy": 5.6403160572052,
      "epoch": 4.788903240298759,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029405581676577897,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22986772507429123,
      "num_tokens": 95189432.0,
      "step": 42640
    },
    {
      "entropy": 5.6569112777709964,
      "epoch": 4.789464817206716,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000294015811992556,
      "loss": 4.9177,
      "mean_token_accuracy": 0.23246700763702394,
      "num_tokens": 95200682.0,
      "step": 42645
    },
    {
      "entropy": 5.647894620895386,
      "epoch": 4.790026394114674,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002939758066138621,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23622154146432878,
      "num_tokens": 95212643.0,
      "step": 42650
    },
    {
      "entropy": 5.699855041503906,
      "epoch": 4.790587971022632,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002939358006309712,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23183035403490065,
      "num_tokens": 95223098.0,
      "step": 42655
    },
    {
      "entropy": 5.666549015045166,
      "epoch": 4.791149547930589,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002938957940451571,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23290950655937195,
      "num_tokens": 95232946.0,
      "step": 42660
    },
    {
      "entropy": 5.729850101470947,
      "epoch": 4.791711124838547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029385578685769346,
      "loss": 4.9648,
      "mean_token_accuracy": 0.23082300275564194,
      "num_tokens": 95243966.0,
      "step": 42665
    },
    {
      "entropy": 5.691252946853638,
      "epoch": 4.792272701746504,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002938157790698544,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22984337508678437,
      "num_tokens": 95255251.0,
      "step": 42670
    },
    {
      "entropy": 5.632069158554077,
      "epoch": 4.792834278654462,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002937757706829136,
      "loss": 4.854,
      "mean_token_accuracy": 0.23705863654613496,
      "num_tokens": 95267517.0,
      "step": 42675
    },
    {
      "entropy": 5.697246885299682,
      "epoch": 4.793395855562419,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000293735761698145,
      "loss": 4.9743,
      "mean_token_accuracy": 0.23115744739770888,
      "num_tokens": 95277508.0,
      "step": 42680
    },
    {
      "entropy": 5.619847679138184,
      "epoch": 4.793957432470377,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002936957521168224,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23606630116701127,
      "num_tokens": 95289757.0,
      "step": 42685
    },
    {
      "entropy": 5.775667285919189,
      "epoch": 4.794519009378335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002936557419402198,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22791101634502411,
      "num_tokens": 95301541.0,
      "step": 42690
    },
    {
      "entropy": 5.678779125213623,
      "epoch": 4.795080586286292,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002936157311696112,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22889723330736161,
      "num_tokens": 95312840.0,
      "step": 42695
    },
    {
      "entropy": 5.653763628005981,
      "epoch": 4.795642163194249,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002935757198062705,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2405130922794342,
      "num_tokens": 95322909.0,
      "step": 42700
    },
    {
      "entropy": 5.550262355804444,
      "epoch": 4.796203740102207,
      "grad_norm": 1.125,
      "learning_rate": 0.0002935357078514717,
      "loss": 4.7781,
      "mean_token_accuracy": 0.2378532826900482,
      "num_tokens": 95333437.0,
      "step": 42705
    },
    {
      "entropy": 5.635079574584961,
      "epoch": 4.7967653170101645,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029349569530648877,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2337607264518738,
      "num_tokens": 95344477.0,
      "step": 42710
    },
    {
      "entropy": 5.744970464706421,
      "epoch": 4.797326893918122,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002934556821725957,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22508141547441482,
      "num_tokens": 95354354.0,
      "step": 42715
    },
    {
      "entropy": 5.6969352722167965,
      "epoch": 4.79788847082608,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00029341566845106647,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22928321212530137,
      "num_tokens": 95365026.0,
      "step": 42720
    },
    {
      "entropy": 5.765226888656616,
      "epoch": 4.798450047734037,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002933756541431753,
      "loss": 4.9986,
      "mean_token_accuracy": 0.23132768273353577,
      "num_tokens": 95376625.0,
      "step": 42725
    },
    {
      "entropy": 5.680867528915405,
      "epoch": 4.799011624641995,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029333563925019616,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22569175511598588,
      "num_tokens": 95387066.0,
      "step": 42730
    },
    {
      "entropy": 5.6330585956573485,
      "epoch": 4.799573201549952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029329562377340327,
      "loss": 4.908,
      "mean_token_accuracy": 0.23281950056552886,
      "num_tokens": 95398623.0,
      "step": 42735
    },
    {
      "entropy": 5.752023696899414,
      "epoch": 4.80013477845791,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002932556077140704,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22758084088563918,
      "num_tokens": 95412514.0,
      "step": 42740
    },
    {
      "entropy": 5.771838617324829,
      "epoch": 4.800696355365868,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029321559107347206,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22335057854652404,
      "num_tokens": 95422885.0,
      "step": 42745
    },
    {
      "entropy": 5.738114595413208,
      "epoch": 4.801257932273825,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029317557385288216,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2297970935702324,
      "num_tokens": 95433659.0,
      "step": 42750
    },
    {
      "entropy": 5.74195556640625,
      "epoch": 4.801819509181782,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000293135556053575,
      "loss": 4.9735,
      "mean_token_accuracy": 0.225498566031456,
      "num_tokens": 95444161.0,
      "step": 42755
    },
    {
      "entropy": 5.689209985733032,
      "epoch": 4.80238108608974,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029309553767682467,
      "loss": 4.9034,
      "mean_token_accuracy": 0.23140663355588914,
      "num_tokens": 95454840.0,
      "step": 42760
    },
    {
      "entropy": 5.75335373878479,
      "epoch": 4.8029426629976975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002930555187239054,
      "loss": 4.952,
      "mean_token_accuracy": 0.23195367455482482,
      "num_tokens": 95466766.0,
      "step": 42765
    },
    {
      "entropy": 5.677689743041992,
      "epoch": 4.803504239905655,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029301549919609146,
      "loss": 4.948,
      "mean_token_accuracy": 0.23495631366968156,
      "num_tokens": 95477767.0,
      "step": 42770
    },
    {
      "entropy": 5.648311567306519,
      "epoch": 4.804065816813613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029297547909465697,
      "loss": 4.9014,
      "mean_token_accuracy": 0.22633554339408873,
      "num_tokens": 95489428.0,
      "step": 42775
    },
    {
      "entropy": 5.755836153030396,
      "epoch": 4.80462739372157,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002929354584208763,
      "loss": 5.0422,
      "mean_token_accuracy": 0.22724674344062806,
      "num_tokens": 95500844.0,
      "step": 42780
    },
    {
      "entropy": 5.698389196395874,
      "epoch": 4.805188970629528,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002928954371760237,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2408514678478241,
      "num_tokens": 95511829.0,
      "step": 42785
    },
    {
      "entropy": 5.758673906326294,
      "epoch": 4.805750547537485,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002928554153613734,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2278372198343277,
      "num_tokens": 95523366.0,
      "step": 42790
    },
    {
      "entropy": 5.612120246887207,
      "epoch": 4.8063121244454425,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002928153929781997,
      "loss": 4.81,
      "mean_token_accuracy": 0.2440278246998787,
      "num_tokens": 95533495.0,
      "step": 42795
    },
    {
      "entropy": 5.658991956710816,
      "epoch": 4.806873701353401,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029277537002777703,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2315381720662117,
      "num_tokens": 95545375.0,
      "step": 42800
    },
    {
      "entropy": 5.5967018604278564,
      "epoch": 4.807435278261358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029273534651137967,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2361537277698517,
      "num_tokens": 95557212.0,
      "step": 42805
    },
    {
      "entropy": 5.811467027664184,
      "epoch": 4.807996855169315,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029269532243028195,
      "loss": 5.087,
      "mean_token_accuracy": 0.22432696670293809,
      "num_tokens": 95568608.0,
      "step": 42810
    },
    {
      "entropy": 5.600391817092896,
      "epoch": 4.808558432077273,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002926552977857584,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23195265829563141,
      "num_tokens": 95579069.0,
      "step": 42815
    },
    {
      "entropy": 5.7210523128509525,
      "epoch": 4.80912000898523,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029261527257908325,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2306963622570038,
      "num_tokens": 95589244.0,
      "step": 42820
    },
    {
      "entropy": 5.675212287902832,
      "epoch": 4.809681585893188,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029257524681153106,
      "loss": 4.9587,
      "mean_token_accuracy": 0.23614521324634552,
      "num_tokens": 95600277.0,
      "step": 42825
    },
    {
      "entropy": 5.7552947998046875,
      "epoch": 4.810243162801146,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002925352204843761,
      "loss": 4.9561,
      "mean_token_accuracy": 0.23076430857181549,
      "num_tokens": 95609913.0,
      "step": 42830
    },
    {
      "entropy": 5.845233392715454,
      "epoch": 4.810804739709103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000292495193598893,
      "loss": 5.0423,
      "mean_token_accuracy": 0.22694491744041442,
      "num_tokens": 95623142.0,
      "step": 42835
    },
    {
      "entropy": 5.679775142669678,
      "epoch": 4.811366316617061,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029245516615635615,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22150918394327163,
      "num_tokens": 95633711.0,
      "step": 42840
    },
    {
      "entropy": 5.68955774307251,
      "epoch": 4.811927893525018,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029241513815804,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23391638398170472,
      "num_tokens": 95646956.0,
      "step": 42845
    },
    {
      "entropy": 5.78933162689209,
      "epoch": 4.8124894704329755,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029237510960521904,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22384982109069823,
      "num_tokens": 95659260.0,
      "step": 42850
    },
    {
      "entropy": 5.792727947235107,
      "epoch": 4.813051047340934,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000292335080499168,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2230590045452118,
      "num_tokens": 95671122.0,
      "step": 42855
    },
    {
      "entropy": 5.668305826187134,
      "epoch": 4.813612624248891,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029229505084116123,
      "loss": 4.919,
      "mean_token_accuracy": 0.22652896791696547,
      "num_tokens": 95681513.0,
      "step": 42860
    },
    {
      "entropy": 5.656750822067261,
      "epoch": 4.814174201156849,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002922550206324732,
      "loss": 4.8899,
      "mean_token_accuracy": 0.23970165252685546,
      "num_tokens": 95691670.0,
      "step": 42865
    },
    {
      "entropy": 5.673542594909668,
      "epoch": 4.814735778064806,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002922149898743787,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2280026227235794,
      "num_tokens": 95702359.0,
      "step": 42870
    },
    {
      "entropy": 5.73592734336853,
      "epoch": 4.815297354972763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002921749585681523,
      "loss": 4.9653,
      "mean_token_accuracy": 0.24099575281143187,
      "num_tokens": 95714069.0,
      "step": 42875
    },
    {
      "entropy": 5.71402325630188,
      "epoch": 4.8158589318807214,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029213492671506855,
      "loss": 4.921,
      "mean_token_accuracy": 0.22349916100502015,
      "num_tokens": 95724420.0,
      "step": 42880
    },
    {
      "entropy": 5.6939538478851315,
      "epoch": 4.816420508788679,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029209489431640203,
      "loss": 4.9128,
      "mean_token_accuracy": 0.23566330671310426,
      "num_tokens": 95735229.0,
      "step": 42885
    },
    {
      "entropy": 5.612822389602661,
      "epoch": 4.816982085696636,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002920548613734275,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23854623585939408,
      "num_tokens": 95745246.0,
      "step": 42890
    },
    {
      "entropy": 5.650992488861084,
      "epoch": 4.817543662604594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029201482788741953,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22532814145088195,
      "num_tokens": 95755752.0,
      "step": 42895
    },
    {
      "entropy": 5.71661262512207,
      "epoch": 4.818105239512551,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002919747938596528,
      "loss": 5.0139,
      "mean_token_accuracy": 0.22245561629533767,
      "num_tokens": 95766476.0,
      "step": 42900
    },
    {
      "entropy": 5.784330940246582,
      "epoch": 4.818666816420508,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002919347592914021,
      "loss": 5.008,
      "mean_token_accuracy": 0.22095311880111695,
      "num_tokens": 95776862.0,
      "step": 42905
    },
    {
      "entropy": 5.699935054779052,
      "epoch": 4.8192283933284665,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029189472418394215,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2317724972963333,
      "num_tokens": 95788216.0,
      "step": 42910
    },
    {
      "entropy": 5.6969328880310055,
      "epoch": 4.819789970236424,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029185468853854755,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2349299520254135,
      "num_tokens": 95799050.0,
      "step": 42915
    },
    {
      "entropy": 5.7361208438873295,
      "epoch": 4.820351547144382,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002918146523564932,
      "loss": 4.946,
      "mean_token_accuracy": 0.22825605124235154,
      "num_tokens": 95810763.0,
      "step": 42920
    },
    {
      "entropy": 5.756783866882325,
      "epoch": 4.820913124052339,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029177461563905375,
      "loss": 4.9385,
      "mean_token_accuracy": 0.23308091312646867,
      "num_tokens": 95821730.0,
      "step": 42925
    },
    {
      "entropy": 5.76891508102417,
      "epoch": 4.821474700960296,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029173457838750405,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22424596101045607,
      "num_tokens": 95834037.0,
      "step": 42930
    },
    {
      "entropy": 5.640110445022583,
      "epoch": 4.822036277868254,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029169454060311886,
      "loss": 4.894,
      "mean_token_accuracy": 0.23008856922388077,
      "num_tokens": 95845074.0,
      "step": 42935
    },
    {
      "entropy": 5.761757755279541,
      "epoch": 4.822597854776212,
      "grad_norm": 1.265625,
      "learning_rate": 0.000291654502287173,
      "loss": 5.028,
      "mean_token_accuracy": 0.23067302256822586,
      "num_tokens": 95855439.0,
      "step": 42940
    },
    {
      "entropy": 5.830123710632324,
      "epoch": 4.823159431684169,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002916144634409414,
      "loss": 5.0172,
      "mean_token_accuracy": 0.23161375671625137,
      "num_tokens": 95866511.0,
      "step": 42945
    },
    {
      "entropy": 5.763914489746094,
      "epoch": 4.823721008592127,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029157442406569885,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23085851222276688,
      "num_tokens": 95877795.0,
      "step": 42950
    },
    {
      "entropy": 5.698693895339966,
      "epoch": 4.824282585500084,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002915343841627201,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22876168936491012,
      "num_tokens": 95889039.0,
      "step": 42955
    },
    {
      "entropy": 5.68127851486206,
      "epoch": 4.824844162408041,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002914943437332802,
      "loss": 4.8695,
      "mean_token_accuracy": 0.235785773396492,
      "num_tokens": 95900101.0,
      "step": 42960
    },
    {
      "entropy": 5.668141031265259,
      "epoch": 4.8254057393159995,
      "grad_norm": 1.25,
      "learning_rate": 0.00029145430277865406,
      "loss": 4.8661,
      "mean_token_accuracy": 0.23275691270828247,
      "num_tokens": 95910023.0,
      "step": 42965
    },
    {
      "entropy": 5.641350936889649,
      "epoch": 4.825967316223957,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002914142613001165,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23860234916210174,
      "num_tokens": 95920610.0,
      "step": 42970
    },
    {
      "entropy": 5.625864744186401,
      "epoch": 4.826528893131915,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029137421929894235,
      "loss": 4.8731,
      "mean_token_accuracy": 0.2305596128106117,
      "num_tokens": 95931521.0,
      "step": 42975
    },
    {
      "entropy": 5.827869892120361,
      "epoch": 4.827090470039872,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002913341767764069,
      "loss": 5.008,
      "mean_token_accuracy": 0.23592042922973633,
      "num_tokens": 95943179.0,
      "step": 42980
    },
    {
      "entropy": 5.766996669769287,
      "epoch": 4.827652046947829,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029129413373378484,
      "loss": 4.9435,
      "mean_token_accuracy": 0.23591773808002472,
      "num_tokens": 95955282.0,
      "step": 42985
    },
    {
      "entropy": 5.6780304431915285,
      "epoch": 4.828213623855787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029125409017235126,
      "loss": 4.8505,
      "mean_token_accuracy": 0.2339387208223343,
      "num_tokens": 95966296.0,
      "step": 42990
    },
    {
      "entropy": 5.738723421096802,
      "epoch": 4.828775200763745,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002912140460933812,
      "loss": 4.9533,
      "mean_token_accuracy": 0.23078607767820358,
      "num_tokens": 95977039.0,
      "step": 42995
    },
    {
      "entropy": 5.765273904800415,
      "epoch": 4.829336777671702,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029117400149814953,
      "loss": 5.075,
      "mean_token_accuracy": 0.22283040285110473,
      "num_tokens": 95988041.0,
      "step": 43000
    },
    {
      "entropy": 5.662926959991455,
      "epoch": 4.82989835457966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002911339563879315,
      "loss": 4.8496,
      "mean_token_accuracy": 0.24416604042053222,
      "num_tokens": 95998950.0,
      "step": 43005
    },
    {
      "entropy": 5.669920635223389,
      "epoch": 4.830459931487617,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002910939107640019,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2279868721961975,
      "num_tokens": 96010380.0,
      "step": 43010
    },
    {
      "entropy": 5.682880020141601,
      "epoch": 4.831021508395574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000291053864627636,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2290787786245346,
      "num_tokens": 96021860.0,
      "step": 43015
    },
    {
      "entropy": 5.746015977859497,
      "epoch": 4.831583085303532,
      "grad_norm": 1.3125,
      "learning_rate": 0.00029101381798010885,
      "loss": 4.9251,
      "mean_token_accuracy": 0.23504430800676346,
      "num_tokens": 96034127.0,
      "step": 43020
    },
    {
      "entropy": 5.702151489257813,
      "epoch": 4.83214466221149,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002909737708226954,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2320763111114502,
      "num_tokens": 96045029.0,
      "step": 43025
    },
    {
      "entropy": 5.7114208221435545,
      "epoch": 4.832706239119448,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029093372315667105,
      "loss": 4.8976,
      "mean_token_accuracy": 0.2356528654694557,
      "num_tokens": 96055409.0,
      "step": 43030
    },
    {
      "entropy": 5.713522291183471,
      "epoch": 4.833267816027405,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029089367498331075,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2313860610127449,
      "num_tokens": 96064928.0,
      "step": 43035
    },
    {
      "entropy": 5.728854846954346,
      "epoch": 4.833829392935362,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029085362630388965,
      "loss": 5.0149,
      "mean_token_accuracy": 0.23561418652534485,
      "num_tokens": 96077111.0,
      "step": 43040
    },
    {
      "entropy": 5.680845928192139,
      "epoch": 4.83439096984332,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002908135771196829,
      "loss": 4.8466,
      "mean_token_accuracy": 0.24517299681901933,
      "num_tokens": 96088850.0,
      "step": 43045
    },
    {
      "entropy": 5.726765394210815,
      "epoch": 4.8349525467512775,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002907735274319657,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23365541994571687,
      "num_tokens": 96099273.0,
      "step": 43050
    },
    {
      "entropy": 5.726725721359253,
      "epoch": 4.835514123659236,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002907334772420134,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22015268057584764,
      "num_tokens": 96111332.0,
      "step": 43055
    },
    {
      "entropy": 5.735763263702393,
      "epoch": 4.836075700567193,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000290693426551101,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22918733805418015,
      "num_tokens": 96122225.0,
      "step": 43060
    },
    {
      "entropy": 5.704058265686035,
      "epoch": 4.83663727747515,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029065337536050384,
      "loss": 4.9131,
      "mean_token_accuracy": 0.24236684739589692,
      "num_tokens": 96133465.0,
      "step": 43065
    },
    {
      "entropy": 5.650646495819092,
      "epoch": 4.837198854383108,
      "grad_norm": 1.25,
      "learning_rate": 0.00029061332367149705,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2407153144478798,
      "num_tokens": 96145036.0,
      "step": 43070
    },
    {
      "entropy": 5.573168230056763,
      "epoch": 4.837760431291065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029057327148535614,
      "loss": 4.7959,
      "mean_token_accuracy": 0.24309304654598235,
      "num_tokens": 96154986.0,
      "step": 43075
    },
    {
      "entropy": 5.722524356842041,
      "epoch": 4.838322008199023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000290533218803356,
      "loss": 5.0162,
      "mean_token_accuracy": 0.226569364964962,
      "num_tokens": 96166271.0,
      "step": 43080
    },
    {
      "entropy": 5.720946264266968,
      "epoch": 4.838883585106981,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029049316562677235,
      "loss": 5.0025,
      "mean_token_accuracy": 0.23145142942667007,
      "num_tokens": 96177203.0,
      "step": 43085
    },
    {
      "entropy": 5.730516338348389,
      "epoch": 4.839445162014938,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029045311195688016,
      "loss": 4.9709,
      "mean_token_accuracy": 0.22762647569179534,
      "num_tokens": 96188749.0,
      "step": 43090
    },
    {
      "entropy": 5.7314661026000975,
      "epoch": 4.840006738922895,
      "grad_norm": 1.296875,
      "learning_rate": 0.000290413057794955,
      "loss": 4.9599,
      "mean_token_accuracy": 0.23478259593248368,
      "num_tokens": 96198701.0,
      "step": 43095
    },
    {
      "entropy": 5.857251501083374,
      "epoch": 4.840568315830853,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029037300314227205,
      "loss": 5.0941,
      "mean_token_accuracy": 0.22676442116498946,
      "num_tokens": 96211016.0,
      "step": 43100
    },
    {
      "entropy": 5.802139091491699,
      "epoch": 4.8411298927388104,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029033294800010666,
      "loss": 4.9783,
      "mean_token_accuracy": 0.22572473883628846,
      "num_tokens": 96222449.0,
      "step": 43105
    },
    {
      "entropy": 5.656638050079346,
      "epoch": 4.8416914696467686,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029029289236973436,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22926890403032302,
      "num_tokens": 96234434.0,
      "step": 43110
    },
    {
      "entropy": 5.667249155044556,
      "epoch": 4.842253046554726,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029025283625243043,
      "loss": 4.9197,
      "mean_token_accuracy": 0.23017847239971162,
      "num_tokens": 96245612.0,
      "step": 43115
    },
    {
      "entropy": 5.722255039215088,
      "epoch": 4.842814623462683,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029021277964947023,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2346621721982956,
      "num_tokens": 96256790.0,
      "step": 43120
    },
    {
      "entropy": 5.782699918746948,
      "epoch": 4.843376200370641,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029017272256212927,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2274271219968796,
      "num_tokens": 96267562.0,
      "step": 43125
    },
    {
      "entropy": 5.711585569381714,
      "epoch": 4.843937777278598,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002901326649916829,
      "loss": 4.891,
      "mean_token_accuracy": 0.23246378004550933,
      "num_tokens": 96278469.0,
      "step": 43130
    },
    {
      "entropy": 5.730046558380127,
      "epoch": 4.8444993541865555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029009260693940665,
      "loss": 5.0572,
      "mean_token_accuracy": 0.22502431422472,
      "num_tokens": 96289460.0,
      "step": 43135
    },
    {
      "entropy": 5.7404906272888185,
      "epoch": 4.845060931094514,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002900525484065759,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2256675347685814,
      "num_tokens": 96300963.0,
      "step": 43140
    },
    {
      "entropy": 5.692561149597168,
      "epoch": 4.845622508002471,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029001248939446617,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23349389880895616,
      "num_tokens": 96311521.0,
      "step": 43145
    },
    {
      "entropy": 5.69579119682312,
      "epoch": 4.846184084910428,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000289972429904353,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22492395639419555,
      "num_tokens": 96321959.0,
      "step": 43150
    },
    {
      "entropy": 5.832092142105102,
      "epoch": 4.846745661818386,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028993236993751174,
      "loss": 5.1377,
      "mean_token_accuracy": 0.22423526346683503,
      "num_tokens": 96333916.0,
      "step": 43155
    },
    {
      "entropy": 5.766913795471192,
      "epoch": 4.847307238726343,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028989230949521817,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23264532685279846,
      "num_tokens": 96345081.0,
      "step": 43160
    },
    {
      "entropy": 5.659201383590698,
      "epoch": 4.8478688156343015,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028985224857874765,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22738554775714875,
      "num_tokens": 96356394.0,
      "step": 43165
    },
    {
      "entropy": 5.581744003295898,
      "epoch": 4.848430392542259,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028981218718937573,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23303744494915007,
      "num_tokens": 96366952.0,
      "step": 43170
    },
    {
      "entropy": 5.693431568145752,
      "epoch": 4.848991969450216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028977212532837804,
      "loss": 4.994,
      "mean_token_accuracy": 0.21920443922281266,
      "num_tokens": 96378165.0,
      "step": 43175
    },
    {
      "entropy": 5.828152799606324,
      "epoch": 4.849553546358174,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002897320629970302,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2409444198012352,
      "num_tokens": 96390606.0,
      "step": 43180
    },
    {
      "entropy": 5.683108901977539,
      "epoch": 4.850115123266131,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028969200019660774,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2376352772116661,
      "num_tokens": 96401089.0,
      "step": 43185
    },
    {
      "entropy": 5.695379590988159,
      "epoch": 4.8506767001740885,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002896519369283862,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2292659342288971,
      "num_tokens": 96411908.0,
      "step": 43190
    },
    {
      "entropy": 5.648802757263184,
      "epoch": 4.851238277082047,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002896118731936414,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2377168357372284,
      "num_tokens": 96422859.0,
      "step": 43195
    },
    {
      "entropy": 5.699770259857178,
      "epoch": 4.851799853990004,
      "grad_norm": 1.125,
      "learning_rate": 0.0002895718089936489,
      "loss": 4.937,
      "mean_token_accuracy": 0.23078078031539917,
      "num_tokens": 96435122.0,
      "step": 43200
    },
    {
      "entropy": 5.724276351928711,
      "epoch": 4.852361430897961,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002895317443296843,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2235727369785309,
      "num_tokens": 96446916.0,
      "step": 43205
    },
    {
      "entropy": 5.70645489692688,
      "epoch": 4.852923007805919,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028949167920302337,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2272409528493881,
      "num_tokens": 96458241.0,
      "step": 43210
    },
    {
      "entropy": 5.739713191986084,
      "epoch": 4.853484584713876,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028945161361494177,
      "loss": 4.941,
      "mean_token_accuracy": 0.22695276886224747,
      "num_tokens": 96468581.0,
      "step": 43215
    },
    {
      "entropy": 5.696501874923706,
      "epoch": 4.854046161621834,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002894115475667151,
      "loss": 4.8986,
      "mean_token_accuracy": 0.23591605871915816,
      "num_tokens": 96479119.0,
      "step": 43220
    },
    {
      "entropy": 5.750448179244995,
      "epoch": 4.854607738529792,
      "grad_norm": 1.40625,
      "learning_rate": 0.00028937148105961924,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22238780111074447,
      "num_tokens": 96490306.0,
      "step": 43225
    },
    {
      "entropy": 5.620554828643799,
      "epoch": 4.855169315437749,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002893314140949298,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23484330773353576,
      "num_tokens": 96501222.0,
      "step": 43230
    },
    {
      "entropy": 5.698366403579712,
      "epoch": 4.855730892345707,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028929134667392265,
      "loss": 5.0156,
      "mean_token_accuracy": 0.23005005866289138,
      "num_tokens": 96513283.0,
      "step": 43235
    },
    {
      "entropy": 5.704756593704223,
      "epoch": 4.856292469253664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028925127879787343,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22902049422264098,
      "num_tokens": 96523546.0,
      "step": 43240
    },
    {
      "entropy": 5.712658452987671,
      "epoch": 4.856854046161622,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002892112104680579,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2312276691198349,
      "num_tokens": 96534906.0,
      "step": 43245
    },
    {
      "entropy": 5.65229754447937,
      "epoch": 4.8574156230695795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028917114168575196,
      "loss": 4.8846,
      "mean_token_accuracy": 0.233695325255394,
      "num_tokens": 96547029.0,
      "step": 43250
    },
    {
      "entropy": 5.6991126537323,
      "epoch": 4.857977199977537,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002891310724522314,
      "loss": 4.8471,
      "mean_token_accuracy": 0.2437685325741768,
      "num_tokens": 96558632.0,
      "step": 43255
    },
    {
      "entropy": 5.732682228088379,
      "epoch": 4.858538776885495,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028909100276877203,
      "loss": 4.9661,
      "mean_token_accuracy": 0.23806638717651368,
      "num_tokens": 96568644.0,
      "step": 43260
    },
    {
      "entropy": 5.657897138595581,
      "epoch": 4.859100353793452,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028905093263664966,
      "loss": 4.8788,
      "mean_token_accuracy": 0.23384026139974595,
      "num_tokens": 96578844.0,
      "step": 43265
    },
    {
      "entropy": 5.7050042152404785,
      "epoch": 4.859661930701409,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028901086205714016,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22780286967754365,
      "num_tokens": 96590715.0,
      "step": 43270
    },
    {
      "entropy": 5.704536819458008,
      "epoch": 4.860223507609367,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028897079103151936,
      "loss": 4.9257,
      "mean_token_accuracy": 0.23180724531412125,
      "num_tokens": 96601001.0,
      "step": 43275
    },
    {
      "entropy": 5.6353028297424315,
      "epoch": 4.860785084517325,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028893071956106315,
      "loss": 4.869,
      "mean_token_accuracy": 0.23339757174253464,
      "num_tokens": 96612591.0,
      "step": 43280
    },
    {
      "entropy": 5.612305307388306,
      "epoch": 4.861346661425282,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028889064764704735,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22915439754724504,
      "num_tokens": 96624353.0,
      "step": 43285
    },
    {
      "entropy": 5.711311006546021,
      "epoch": 4.86190823833324,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028885057529074804,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2331239327788353,
      "num_tokens": 96634293.0,
      "step": 43290
    },
    {
      "entropy": 5.80243501663208,
      "epoch": 4.862469815241197,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028881050249344103,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22478341013193132,
      "num_tokens": 96645308.0,
      "step": 43295
    },
    {
      "entropy": 5.748935651779175,
      "epoch": 4.863031392149155,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002887704292564022,
      "loss": 4.9733,
      "mean_token_accuracy": 0.23021266013383865,
      "num_tokens": 96655198.0,
      "step": 43300
    },
    {
      "entropy": 5.6689857006072994,
      "epoch": 4.8635929690571125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028873035558090766,
      "loss": 4.9369,
      "mean_token_accuracy": 0.23003607094287873,
      "num_tokens": 96665753.0,
      "step": 43305
    },
    {
      "entropy": 5.688166332244873,
      "epoch": 4.86415454596507,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002886902814682332,
      "loss": 4.937,
      "mean_token_accuracy": 0.2326132297515869,
      "num_tokens": 96676038.0,
      "step": 43310
    },
    {
      "entropy": 5.771292591094971,
      "epoch": 4.864716122873028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028865020691965485,
      "loss": 4.8918,
      "mean_token_accuracy": 0.23147742152214051,
      "num_tokens": 96687893.0,
      "step": 43315
    },
    {
      "entropy": 5.79625391960144,
      "epoch": 4.865277699780985,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028861013193644865,
      "loss": 5.0305,
      "mean_token_accuracy": 0.22693181931972503,
      "num_tokens": 96698937.0,
      "step": 43320
    },
    {
      "entropy": 5.66983733177185,
      "epoch": 4.865839276688942,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002885700565198906,
      "loss": 4.9017,
      "mean_token_accuracy": 0.23178031295537949,
      "num_tokens": 96710117.0,
      "step": 43325
    },
    {
      "entropy": 5.656578493118286,
      "epoch": 4.8664008535969,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028852998067125675,
      "loss": 4.8733,
      "mean_token_accuracy": 0.23297743052244185,
      "num_tokens": 96719831.0,
      "step": 43330
    },
    {
      "entropy": 5.7079226016998295,
      "epoch": 4.8669624305048576,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002884899043918229,
      "loss": 5.0364,
      "mean_token_accuracy": 0.22651194632053376,
      "num_tokens": 96732127.0,
      "step": 43335
    },
    {
      "entropy": 5.75765290260315,
      "epoch": 4.867524007412815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002884498276828653,
      "loss": 5.0473,
      "mean_token_accuracy": 0.22277527302503586,
      "num_tokens": 96744406.0,
      "step": 43340
    },
    {
      "entropy": 5.741248273849488,
      "epoch": 4.868085584320773,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028840975054566007,
      "loss": 4.9359,
      "mean_token_accuracy": 0.24196452796459197,
      "num_tokens": 96754810.0,
      "step": 43345
    },
    {
      "entropy": 5.660648918151855,
      "epoch": 4.86864716122873,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028836967298148317,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23444783687591553,
      "num_tokens": 96765520.0,
      "step": 43350
    },
    {
      "entropy": 5.758445358276367,
      "epoch": 4.869208738136688,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028832959499161065,
      "loss": 5.069,
      "mean_token_accuracy": 0.2260325476527214,
      "num_tokens": 96777262.0,
      "step": 43355
    },
    {
      "entropy": 5.706459808349609,
      "epoch": 4.869770315044645,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002882895165773187,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2294507533311844,
      "num_tokens": 96788452.0,
      "step": 43360
    },
    {
      "entropy": 5.681039428710937,
      "epoch": 4.870331891952603,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002882494377398833,
      "loss": 4.9107,
      "mean_token_accuracy": 0.23629775494337082,
      "num_tokens": 96799402.0,
      "step": 43365
    },
    {
      "entropy": 5.763542604446411,
      "epoch": 4.870893468860561,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028820935848058066,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2306209459900856,
      "num_tokens": 96810467.0,
      "step": 43370
    },
    {
      "entropy": 5.689184522628784,
      "epoch": 4.871455045768518,
      "grad_norm": 1.25,
      "learning_rate": 0.000288169278800687,
      "loss": 4.8694,
      "mean_token_accuracy": 0.23528995513916015,
      "num_tokens": 96821480.0,
      "step": 43375
    },
    {
      "entropy": 5.640556573867798,
      "epoch": 4.872016622676475,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028812919870147837,
      "loss": 4.8665,
      "mean_token_accuracy": 0.24443240761756896,
      "num_tokens": 96832473.0,
      "step": 43380
    },
    {
      "entropy": 5.707164239883423,
      "epoch": 4.872578199584433,
      "grad_norm": 1.125,
      "learning_rate": 0.00028808911818423096,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2286904886364937,
      "num_tokens": 96844019.0,
      "step": 43385
    },
    {
      "entropy": 5.649542570114136,
      "epoch": 4.8731397764923905,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028804903725022094,
      "loss": 4.918,
      "mean_token_accuracy": 0.2297379344701767,
      "num_tokens": 96853902.0,
      "step": 43390
    },
    {
      "entropy": 5.735810089111328,
      "epoch": 4.873701353400348,
      "grad_norm": 1.25,
      "learning_rate": 0.0002880089559007245,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23172398507595063,
      "num_tokens": 96865066.0,
      "step": 43395
    },
    {
      "entropy": 5.722618627548218,
      "epoch": 4.874262930308306,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002879688741370179,
      "loss": 5.0338,
      "mean_token_accuracy": 0.22790073454380036,
      "num_tokens": 96876329.0,
      "step": 43400
    },
    {
      "entropy": 5.688831996917725,
      "epoch": 4.874824507216263,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002879287919603773,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2320888504385948,
      "num_tokens": 96887108.0,
      "step": 43405
    },
    {
      "entropy": 5.716056489944458,
      "epoch": 4.875386084124221,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028788870937207896,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22557371109724045,
      "num_tokens": 96898956.0,
      "step": 43410
    },
    {
      "entropy": 5.689077854156494,
      "epoch": 4.875947661032178,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002878486263733991,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22871710509061813,
      "num_tokens": 96910889.0,
      "step": 43415
    },
    {
      "entropy": 5.682149648666382,
      "epoch": 4.876509237940136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000287808542965614,
      "loss": 4.8611,
      "mean_token_accuracy": 0.2338844731450081,
      "num_tokens": 96921682.0,
      "step": 43420
    },
    {
      "entropy": 5.722808074951172,
      "epoch": 4.877070814848094,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028776845914999996,
      "loss": 4.9096,
      "mean_token_accuracy": 0.23592510670423508,
      "num_tokens": 96933502.0,
      "step": 43425
    },
    {
      "entropy": 5.6127601146698,
      "epoch": 4.877632391756051,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002877283749278332,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23586352914571762,
      "num_tokens": 96945621.0,
      "step": 43430
    },
    {
      "entropy": 5.779911231994629,
      "epoch": 4.878193968664009,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028768829030039,
      "loss": 5.058,
      "mean_token_accuracy": 0.22584817558526993,
      "num_tokens": 96957164.0,
      "step": 43435
    },
    {
      "entropy": 5.778891611099243,
      "epoch": 4.878755545571966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028764820526894675,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22863264083862306,
      "num_tokens": 96968646.0,
      "step": 43440
    },
    {
      "entropy": 5.767228460311889,
      "epoch": 4.879317122479923,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028760811983477976,
      "loss": 4.9029,
      "mean_token_accuracy": 0.23893748521804808,
      "num_tokens": 96979855.0,
      "step": 43445
    },
    {
      "entropy": 5.736949491500854,
      "epoch": 4.8798786993878815,
      "grad_norm": 1.25,
      "learning_rate": 0.00028756803399916534,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2385108456015587,
      "num_tokens": 96989603.0,
      "step": 43450
    },
    {
      "entropy": 5.702169370651245,
      "epoch": 4.880440276295839,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002875279477633798,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23474832326173783,
      "num_tokens": 96999511.0,
      "step": 43455
    },
    {
      "entropy": 5.716589736938476,
      "epoch": 4.881001853203796,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002874878611286995,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2307572454214096,
      "num_tokens": 97011210.0,
      "step": 43460
    },
    {
      "entropy": 5.692218351364136,
      "epoch": 4.881563430111754,
      "grad_norm": 1.4375,
      "learning_rate": 0.000287447774096401,
      "loss": 4.9545,
      "mean_token_accuracy": 0.23913467526435853,
      "num_tokens": 97022351.0,
      "step": 43465
    },
    {
      "entropy": 5.733716917037964,
      "epoch": 4.882125007019711,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028740768666776046,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2317238450050354,
      "num_tokens": 97033269.0,
      "step": 43470
    },
    {
      "entropy": 5.746064281463623,
      "epoch": 4.8826865839276685,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028736759884405433,
      "loss": 4.9774,
      "mean_token_accuracy": 0.23127317130565644,
      "num_tokens": 97045278.0,
      "step": 43475
    },
    {
      "entropy": 5.667891216278076,
      "epoch": 4.883248160835627,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002873275106265591,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2431857630610466,
      "num_tokens": 97055617.0,
      "step": 43480
    },
    {
      "entropy": 5.742639684677124,
      "epoch": 4.883809737743584,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028728742201655116,
      "loss": 5.0528,
      "mean_token_accuracy": 0.22772389352321626,
      "num_tokens": 97068815.0,
      "step": 43485
    },
    {
      "entropy": 5.704844045639038,
      "epoch": 4.884371314651542,
      "grad_norm": 1.28125,
      "learning_rate": 0.000287247333015307,
      "loss": 5.0425,
      "mean_token_accuracy": 0.22119652777910231,
      "num_tokens": 97078812.0,
      "step": 43490
    },
    {
      "entropy": 5.718068075180054,
      "epoch": 4.884932891559499,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002872072436241029,
      "loss": 4.85,
      "mean_token_accuracy": 0.25236002206802366,
      "num_tokens": 97090423.0,
      "step": 43495
    },
    {
      "entropy": 5.678114891052246,
      "epoch": 4.885494468467456,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002871671538442155,
      "loss": 4.7912,
      "mean_token_accuracy": 0.24117551743984222,
      "num_tokens": 97100756.0,
      "step": 43500
    },
    {
      "entropy": 5.6155198097229,
      "epoch": 4.8860560453754145,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002871270636769212,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23707357794046402,
      "num_tokens": 97111016.0,
      "step": 43505
    },
    {
      "entropy": 5.716490697860718,
      "epoch": 4.886617622283372,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002870869731234965,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2248105749487877,
      "num_tokens": 97122148.0,
      "step": 43510
    },
    {
      "entropy": 5.69991626739502,
      "epoch": 4.887179199191329,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002870468821852178,
      "loss": 4.997,
      "mean_token_accuracy": 0.22681756019592286,
      "num_tokens": 97133301.0,
      "step": 43515
    },
    {
      "entropy": 5.6912980556488035,
      "epoch": 4.887740776099287,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028700679086336175,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2281586468219757,
      "num_tokens": 97144637.0,
      "step": 43520
    },
    {
      "entropy": 5.670819711685181,
      "epoch": 4.888302353007244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002869666991592048,
      "loss": 4.7741,
      "mean_token_accuracy": 0.24191081374883652,
      "num_tokens": 97156047.0,
      "step": 43525
    },
    {
      "entropy": 5.677797794342041,
      "epoch": 4.8888639299152015,
      "grad_norm": 1.25,
      "learning_rate": 0.0002869266070740235,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2278374120593071,
      "num_tokens": 97167767.0,
      "step": 43530
    },
    {
      "entropy": 5.591698122024536,
      "epoch": 4.88942550682316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002868865146090944,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22904103696346284,
      "num_tokens": 97179326.0,
      "step": 43535
    },
    {
      "entropy": 5.701693344116211,
      "epoch": 4.889987083731117,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028684642176569414,
      "loss": 4.9722,
      "mean_token_accuracy": 0.23468440771102905,
      "num_tokens": 97190437.0,
      "step": 43540
    },
    {
      "entropy": 5.648698472976685,
      "epoch": 4.890548660639075,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002868063285450992,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23827028572559356,
      "num_tokens": 97200913.0,
      "step": 43545
    },
    {
      "entropy": 5.629736804962159,
      "epoch": 4.891110237547032,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002867662349485861,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23432961106300354,
      "num_tokens": 97210935.0,
      "step": 43550
    },
    {
      "entropy": 5.66602520942688,
      "epoch": 4.891671814454989,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028672614097743156,
      "loss": 4.8618,
      "mean_token_accuracy": 0.23286574184894562,
      "num_tokens": 97221568.0,
      "step": 43555
    },
    {
      "entropy": 5.675609540939331,
      "epoch": 4.892233391362947,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028668604663291213,
      "loss": 4.9105,
      "mean_token_accuracy": 0.23182505071163179,
      "num_tokens": 97232149.0,
      "step": 43560
    },
    {
      "entropy": 5.728280735015869,
      "epoch": 4.892794968270905,
      "grad_norm": 1.125,
      "learning_rate": 0.00028664595191630455,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22998666614294053,
      "num_tokens": 97242800.0,
      "step": 43565
    },
    {
      "entropy": 5.763154554367065,
      "epoch": 4.893356545178862,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028660585682888524,
      "loss": 4.9592,
      "mean_token_accuracy": 0.23328208178281784,
      "num_tokens": 97253681.0,
      "step": 43570
    },
    {
      "entropy": 5.772428798675537,
      "epoch": 4.89391812208682,
      "grad_norm": 1.25,
      "learning_rate": 0.00028656576137193096,
      "loss": 4.9614,
      "mean_token_accuracy": 0.23062064349651337,
      "num_tokens": 97264949.0,
      "step": 43575
    },
    {
      "entropy": 5.731955623626709,
      "epoch": 4.894479698994777,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002865256655467184,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2378476470708847,
      "num_tokens": 97276670.0,
      "step": 43580
    },
    {
      "entropy": 5.625547266006469,
      "epoch": 4.895041275902734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002864855693545241,
      "loss": 4.7979,
      "mean_token_accuracy": 0.24104085713624954,
      "num_tokens": 97287134.0,
      "step": 43585
    },
    {
      "entropy": 5.6778709411621096,
      "epoch": 4.8956028528106925,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028644547279662485,
      "loss": 4.9569,
      "mean_token_accuracy": 0.23320836573839188,
      "num_tokens": 97298250.0,
      "step": 43590
    },
    {
      "entropy": 5.628283071517944,
      "epoch": 4.89616442971865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028640537587429734,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23024081587791442,
      "num_tokens": 97309144.0,
      "step": 43595
    },
    {
      "entropy": 5.725813722610473,
      "epoch": 4.896726006626608,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002863652785888182,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22768300771713257,
      "num_tokens": 97320586.0,
      "step": 43600
    },
    {
      "entropy": 5.603799819946289,
      "epoch": 4.897287583534565,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002863251809414642,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23937083929777145,
      "num_tokens": 97330924.0,
      "step": 43605
    },
    {
      "entropy": 5.744052076339722,
      "epoch": 4.897849160442522,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000286285082933512,
      "loss": 4.9967,
      "mean_token_accuracy": 0.23460466116666795,
      "num_tokens": 97341926.0,
      "step": 43610
    },
    {
      "entropy": 5.737836122512817,
      "epoch": 4.89841073735048,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002862449845662385,
      "loss": 5.0005,
      "mean_token_accuracy": 0.2281937062740326,
      "num_tokens": 97353809.0,
      "step": 43615
    },
    {
      "entropy": 5.729521656036377,
      "epoch": 4.898972314258438,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002862048858409201,
      "loss": 4.974,
      "mean_token_accuracy": 0.232970130443573,
      "num_tokens": 97365317.0,
      "step": 43620
    },
    {
      "entropy": 5.746791744232178,
      "epoch": 4.899533891166396,
      "grad_norm": 1.40625,
      "learning_rate": 0.000286164786758834,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22733145207166672,
      "num_tokens": 97375581.0,
      "step": 43625
    },
    {
      "entropy": 5.854024362564087,
      "epoch": 4.900095468074353,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028612468732125667,
      "loss": 5.0741,
      "mean_token_accuracy": 0.2309720143675804,
      "num_tokens": 97387371.0,
      "step": 43630
    },
    {
      "entropy": 5.667828798294067,
      "epoch": 4.90065704498231,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002860845875294651,
      "loss": 4.9048,
      "mean_token_accuracy": 0.2356783702969551,
      "num_tokens": 97398929.0,
      "step": 43635
    },
    {
      "entropy": 5.802333688735962,
      "epoch": 4.901218621890268,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002860444873847358,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2190933421254158,
      "num_tokens": 97410469.0,
      "step": 43640
    },
    {
      "entropy": 5.679329347610474,
      "epoch": 4.9017801987982255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028600438688834587,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2360340252518654,
      "num_tokens": 97422253.0,
      "step": 43645
    },
    {
      "entropy": 5.7316149234771725,
      "epoch": 4.902341775706183,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000285964286041572,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2192739099264145,
      "num_tokens": 97433384.0,
      "step": 43650
    },
    {
      "entropy": 5.722508430480957,
      "epoch": 4.902903352614141,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028592418484569095,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2278374880552292,
      "num_tokens": 97444755.0,
      "step": 43655
    },
    {
      "entropy": 5.608914232254028,
      "epoch": 4.903464929522098,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002858840833019795,
      "loss": 4.7784,
      "mean_token_accuracy": 0.24033973813056947,
      "num_tokens": 97456156.0,
      "step": 43660
    },
    {
      "entropy": 5.664604187011719,
      "epoch": 4.904026506430055,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002858439814117148,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23685464411973953,
      "num_tokens": 97467494.0,
      "step": 43665
    },
    {
      "entropy": 5.655533742904663,
      "epoch": 4.904588083338013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028580387917617343,
      "loss": 4.8594,
      "mean_token_accuracy": 0.235744109749794,
      "num_tokens": 97478317.0,
      "step": 43670
    },
    {
      "entropy": 5.690920495986939,
      "epoch": 4.9051496602459705,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002857637765966324,
      "loss": 4.9292,
      "mean_token_accuracy": 0.23447647839784622,
      "num_tokens": 97488530.0,
      "step": 43675
    },
    {
      "entropy": 5.681792402267456,
      "epoch": 4.905711237153929,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002857236736743685,
      "loss": 4.8607,
      "mean_token_accuracy": 0.2393616482615471,
      "num_tokens": 97501148.0,
      "step": 43680
    },
    {
      "entropy": 5.6907459735870365,
      "epoch": 4.906272814061886,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002856835704106587,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23406973332166672,
      "num_tokens": 97512136.0,
      "step": 43685
    },
    {
      "entropy": 5.706400918960571,
      "epoch": 4.906834390969843,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028564346680677995,
      "loss": 4.9208,
      "mean_token_accuracy": 0.23416323810815812,
      "num_tokens": 97523787.0,
      "step": 43690
    },
    {
      "entropy": 5.770781564712524,
      "epoch": 4.907395967877801,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002856033628640089,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22985766381025313,
      "num_tokens": 97534851.0,
      "step": 43695
    },
    {
      "entropy": 5.693415307998658,
      "epoch": 4.907957544785758,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002855632585836228,
      "loss": 4.871,
      "mean_token_accuracy": 0.2313727006316185,
      "num_tokens": 97545160.0,
      "step": 43700
    },
    {
      "entropy": 5.720825481414795,
      "epoch": 4.908519121693716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028552315396689846,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2287509486079216,
      "num_tokens": 97556321.0,
      "step": 43705
    },
    {
      "entropy": 5.712170505523682,
      "epoch": 4.909080698601674,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002854830490151128,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22946774512529372,
      "num_tokens": 97567406.0,
      "step": 43710
    },
    {
      "entropy": 5.703437232971192,
      "epoch": 4.909642275509631,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028544294372954276,
      "loss": 4.876,
      "mean_token_accuracy": 0.23377344906330108,
      "num_tokens": 97577998.0,
      "step": 43715
    },
    {
      "entropy": 5.664411973953247,
      "epoch": 4.910203852417588,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002854028381114654,
      "loss": 4.9104,
      "mean_token_accuracy": 0.22963576912879943,
      "num_tokens": 97588262.0,
      "step": 43720
    },
    {
      "entropy": 5.655009031295776,
      "epoch": 4.910765429325546,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028536273216215756,
      "loss": 4.8323,
      "mean_token_accuracy": 0.23671152740716933,
      "num_tokens": 97600011.0,
      "step": 43725
    },
    {
      "entropy": 5.8152384757995605,
      "epoch": 4.9113270062335035,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028532262588289636,
      "loss": 5.0881,
      "mean_token_accuracy": 0.23243529498577117,
      "num_tokens": 97612037.0,
      "step": 43730
    },
    {
      "entropy": 5.69534215927124,
      "epoch": 4.911888583141462,
      "grad_norm": 1.25,
      "learning_rate": 0.00028528251927495874,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2387850195169449,
      "num_tokens": 97622643.0,
      "step": 43735
    },
    {
      "entropy": 5.726867580413819,
      "epoch": 4.912450160049419,
      "grad_norm": 1.25,
      "learning_rate": 0.00028524241233962176,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22854717671871186,
      "num_tokens": 97634720.0,
      "step": 43740
    },
    {
      "entropy": 5.8056793212890625,
      "epoch": 4.913011736957376,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028520230507816236,
      "loss": 5.0934,
      "mean_token_accuracy": 0.22050099968910217,
      "num_tokens": 97646314.0,
      "step": 43745
    },
    {
      "entropy": 5.748988151550293,
      "epoch": 4.913573313865334,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028516219749185755,
      "loss": 4.9354,
      "mean_token_accuracy": 0.23261468708515168,
      "num_tokens": 97657478.0,
      "step": 43750
    },
    {
      "entropy": 5.7578216075897215,
      "epoch": 4.914134890773291,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002851220895819845,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2324781075119972,
      "num_tokens": 97668084.0,
      "step": 43755
    },
    {
      "entropy": 5.7205033779144285,
      "epoch": 4.914696467681249,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028508198134982016,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22723958939313887,
      "num_tokens": 97678769.0,
      "step": 43760
    },
    {
      "entropy": 5.696396112442017,
      "epoch": 4.915258044589207,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028504187279664164,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22753168195486068,
      "num_tokens": 97689587.0,
      "step": 43765
    },
    {
      "entropy": 5.703232479095459,
      "epoch": 4.915819621497164,
      "grad_norm": 1.15625,
      "learning_rate": 0.000285001763923726,
      "loss": 4.9681,
      "mean_token_accuracy": 0.23288268893957137,
      "num_tokens": 97700334.0,
      "step": 43770
    },
    {
      "entropy": 5.658030080795288,
      "epoch": 4.916381198405121,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002849616547323503,
      "loss": 4.8635,
      "mean_token_accuracy": 0.24084721505641937,
      "num_tokens": 97713038.0,
      "step": 43775
    },
    {
      "entropy": 5.710124874114991,
      "epoch": 4.916942775313079,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028492154522379167,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22996796667575836,
      "num_tokens": 97724651.0,
      "step": 43780
    },
    {
      "entropy": 5.728899145126343,
      "epoch": 4.917504352221036,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002848814353993273,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22458851039409639,
      "num_tokens": 97735862.0,
      "step": 43785
    },
    {
      "entropy": 5.727177143096924,
      "epoch": 4.9180659291289945,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028484132526023404,
      "loss": 4.933,
      "mean_token_accuracy": 0.2312137857079506,
      "num_tokens": 97747356.0,
      "step": 43790
    },
    {
      "entropy": 5.685667705535889,
      "epoch": 4.918627506036952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002848012148077892,
      "loss": 4.9476,
      "mean_token_accuracy": 0.22693461924791336,
      "num_tokens": 97758670.0,
      "step": 43795
    },
    {
      "entropy": 5.742665529251099,
      "epoch": 4.919189082944909,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028476110404326984,
      "loss": 4.9966,
      "mean_token_accuracy": 0.2343623921275139,
      "num_tokens": 97770291.0,
      "step": 43800
    },
    {
      "entropy": 5.692098999023438,
      "epoch": 4.919750659852867,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002847209929679532,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22927282154560089,
      "num_tokens": 97781567.0,
      "step": 43805
    },
    {
      "entropy": 5.6511955738067625,
      "epoch": 4.920312236760824,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002846808815831164,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23303983807563783,
      "num_tokens": 97792281.0,
      "step": 43810
    },
    {
      "entropy": 5.573779964447022,
      "epoch": 4.920873813668782,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002846407698900365,
      "loss": 4.8012,
      "mean_token_accuracy": 0.23677643835544587,
      "num_tokens": 97803606.0,
      "step": 43815
    },
    {
      "entropy": 5.658079195022583,
      "epoch": 4.92143539057674,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028460065788999073,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23346869349479676,
      "num_tokens": 97816024.0,
      "step": 43820
    },
    {
      "entropy": 5.727068042755127,
      "epoch": 4.921996967484697,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002845605455842564,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23962558060884476,
      "num_tokens": 97827121.0,
      "step": 43825
    },
    {
      "entropy": 5.74198727607727,
      "epoch": 4.922558544392655,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002845204329741106,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23154103606939316,
      "num_tokens": 97838758.0,
      "step": 43830
    },
    {
      "entropy": 5.564477014541626,
      "epoch": 4.923120121300612,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002844803200608304,
      "loss": 4.7855,
      "mean_token_accuracy": 0.246735817193985,
      "num_tokens": 97850208.0,
      "step": 43835
    },
    {
      "entropy": 5.564982891082764,
      "epoch": 4.923681698208569,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002844402068456931,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23126907497644425,
      "num_tokens": 97860210.0,
      "step": 43840
    },
    {
      "entropy": 5.660295724868774,
      "epoch": 4.9242432751165275,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002844000933299761,
      "loss": 4.908,
      "mean_token_accuracy": 0.22911003977060318,
      "num_tokens": 97871348.0,
      "step": 43845
    },
    {
      "entropy": 5.725089693069458,
      "epoch": 4.924804852024485,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028435997951495643,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2348041608929634,
      "num_tokens": 97882767.0,
      "step": 43850
    },
    {
      "entropy": 5.754593229293823,
      "epoch": 4.925366428932442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002843198654019114,
      "loss": 4.9843,
      "mean_token_accuracy": 0.23256502747535707,
      "num_tokens": 97895690.0,
      "step": 43855
    },
    {
      "entropy": 5.704097700119019,
      "epoch": 4.9259280058404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002842797509921182,
      "loss": 4.9416,
      "mean_token_accuracy": 0.23670366257429123,
      "num_tokens": 97906951.0,
      "step": 43860
    },
    {
      "entropy": 5.722225761413574,
      "epoch": 4.926489582748357,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028423963628685414,
      "loss": 4.976,
      "mean_token_accuracy": 0.22699511796236038,
      "num_tokens": 97918538.0,
      "step": 43865
    },
    {
      "entropy": 5.679229402542115,
      "epoch": 4.927051159656315,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002841995212873965,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23725519329309464,
      "num_tokens": 97930175.0,
      "step": 43870
    },
    {
      "entropy": 5.683222055435181,
      "epoch": 4.927612736564273,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002841594059950225,
      "loss": 4.9782,
      "mean_token_accuracy": 0.23092785775661467,
      "num_tokens": 97940870.0,
      "step": 43875
    },
    {
      "entropy": 5.659820032119751,
      "epoch": 4.92817431347223,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028411929041100947,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23096588701009751,
      "num_tokens": 97951492.0,
      "step": 43880
    },
    {
      "entropy": 5.6477690696716305,
      "epoch": 4.928735890380188,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002840791745366347,
      "loss": 4.89,
      "mean_token_accuracy": 0.23736446499824523,
      "num_tokens": 97962243.0,
      "step": 43885
    },
    {
      "entropy": 5.664938640594483,
      "epoch": 4.929297467288145,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028403905837317547,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2373310536146164,
      "num_tokens": 97973150.0,
      "step": 43890
    },
    {
      "entropy": 5.686705303192139,
      "epoch": 4.929859044196102,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028399894192190916,
      "loss": 4.941,
      "mean_token_accuracy": 0.22987054288387299,
      "num_tokens": 97985314.0,
      "step": 43895
    },
    {
      "entropy": 5.706192207336426,
      "epoch": 4.93042062110406,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028395882518411305,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23430256098508834,
      "num_tokens": 97997065.0,
      "step": 43900
    },
    {
      "entropy": 5.675081014633179,
      "epoch": 4.930982198012018,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002839187081610645,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23158821016550063,
      "num_tokens": 98007576.0,
      "step": 43905
    },
    {
      "entropy": 5.655984830856323,
      "epoch": 4.931543774919975,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002838785908540407,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23199134916067124,
      "num_tokens": 98017870.0,
      "step": 43910
    },
    {
      "entropy": 5.699787282943726,
      "epoch": 4.932105351827933,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028383847326431924,
      "loss": 5.0039,
      "mean_token_accuracy": 0.22498496174812316,
      "num_tokens": 98029076.0,
      "step": 43915
    },
    {
      "entropy": 5.698687267303467,
      "epoch": 4.93266692873589,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028379835539317727,
      "loss": 4.871,
      "mean_token_accuracy": 0.236060494184494,
      "num_tokens": 98042099.0,
      "step": 43920
    },
    {
      "entropy": 5.677043485641479,
      "epoch": 4.933228505643848,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002837582372418923,
      "loss": 4.9141,
      "mean_token_accuracy": 0.23485240936279297,
      "num_tokens": 98053335.0,
      "step": 43925
    },
    {
      "entropy": 5.733878231048584,
      "epoch": 4.9337900825518055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002837181188117417,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22905307859182358,
      "num_tokens": 98063723.0,
      "step": 43930
    },
    {
      "entropy": 5.656060743331909,
      "epoch": 4.934351659459763,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002836780001040028,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23424506932497025,
      "num_tokens": 98074756.0,
      "step": 43935
    },
    {
      "entropy": 5.658980464935302,
      "epoch": 4.934913236367721,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000283637881119953,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23535499572753907,
      "num_tokens": 98086584.0,
      "step": 43940
    },
    {
      "entropy": 5.7354501247406,
      "epoch": 4.935474813275678,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028359776186086965,
      "loss": 5.1017,
      "mean_token_accuracy": 0.22362194955348969,
      "num_tokens": 98099996.0,
      "step": 43945
    },
    {
      "entropy": 5.685378837585449,
      "epoch": 4.936036390183635,
      "grad_norm": 1.25,
      "learning_rate": 0.0002835576423280303,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23208689242601394,
      "num_tokens": 98111740.0,
      "step": 43950
    },
    {
      "entropy": 5.760073184967041,
      "epoch": 4.936597967091593,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002835175225227123,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2302089586853981,
      "num_tokens": 98123946.0,
      "step": 43955
    },
    {
      "entropy": 5.731243371963501,
      "epoch": 4.937159543999551,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002834774024461931,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22930005937814713,
      "num_tokens": 98134980.0,
      "step": 43960
    },
    {
      "entropy": 5.701988077163696,
      "epoch": 4.937721120907508,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028343728209975,
      "loss": 4.9722,
      "mean_token_accuracy": 0.23250980228185653,
      "num_tokens": 98145639.0,
      "step": 43965
    },
    {
      "entropy": 5.697753286361694,
      "epoch": 4.938282697815466,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028339716148466067,
      "loss": 4.9724,
      "mean_token_accuracy": 0.23516679257154466,
      "num_tokens": 98156078.0,
      "step": 43970
    },
    {
      "entropy": 5.7515496730804445,
      "epoch": 4.938844274723423,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028335704060220246,
      "loss": 5.0406,
      "mean_token_accuracy": 0.22832592725753784,
      "num_tokens": 98168486.0,
      "step": 43975
    },
    {
      "entropy": 5.800979280471802,
      "epoch": 4.939405851631381,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028331691945365274,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22815954834222793,
      "num_tokens": 98178196.0,
      "step": 43980
    },
    {
      "entropy": 5.68770055770874,
      "epoch": 4.9399674285393385,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002832767980402892,
      "loss": 4.9337,
      "mean_token_accuracy": 0.2327951341867447,
      "num_tokens": 98188912.0,
      "step": 43985
    },
    {
      "entropy": 5.704099178314209,
      "epoch": 4.940529005447296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002832366763633892,
      "loss": 5.0312,
      "mean_token_accuracy": 0.22363149225711823,
      "num_tokens": 98200207.0,
      "step": 43990
    },
    {
      "entropy": 5.6212828159332275,
      "epoch": 4.941090582355254,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028319655442423015,
      "loss": 4.8317,
      "mean_token_accuracy": 0.24162093549966812,
      "num_tokens": 98210846.0,
      "step": 43995
    },
    {
      "entropy": 5.79054069519043,
      "epoch": 4.941652159263211,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028315643222408965,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2199118047952652,
      "num_tokens": 98222303.0,
      "step": 44000
    },
    {
      "entropy": 5.646766710281372,
      "epoch": 4.942213736171169,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028311630976424526,
      "loss": 4.884,
      "mean_token_accuracy": 0.23308131247758865,
      "num_tokens": 98233374.0,
      "step": 44005
    },
    {
      "entropy": 5.773446798324585,
      "epoch": 4.942775313079126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002830761870459743,
      "loss": 5.0615,
      "mean_token_accuracy": 0.23040904104709625,
      "num_tokens": 98244840.0,
      "step": 44010
    },
    {
      "entropy": 5.651413345336914,
      "epoch": 4.9433368899870835,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002830360640705545,
      "loss": 4.9024,
      "mean_token_accuracy": 0.23284354358911513,
      "num_tokens": 98254829.0,
      "step": 44015
    },
    {
      "entropy": 5.701436328887939,
      "epoch": 4.943898466895042,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028299594083926324,
      "loss": 4.966,
      "mean_token_accuracy": 0.23291689455509185,
      "num_tokens": 98266296.0,
      "step": 44020
    },
    {
      "entropy": 5.640690755844116,
      "epoch": 4.944460043802999,
      "grad_norm": 1.375,
      "learning_rate": 0.00028295581735337817,
      "loss": 4.9053,
      "mean_token_accuracy": 0.24378354847431183,
      "num_tokens": 98277723.0,
      "step": 44025
    },
    {
      "entropy": 5.729151678085327,
      "epoch": 4.945021620710956,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028291569361417674,
      "loss": 4.9538,
      "mean_token_accuracy": 0.23862432837486267,
      "num_tokens": 98288799.0,
      "step": 44030
    },
    {
      "entropy": 5.695339632034302,
      "epoch": 4.945583197618914,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002828755696229366,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22553272545337677,
      "num_tokens": 98299731.0,
      "step": 44035
    },
    {
      "entropy": 5.713605546951294,
      "epoch": 4.946144774526871,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002828354453809352,
      "loss": 4.87,
      "mean_token_accuracy": 0.23350426703691482,
      "num_tokens": 98310607.0,
      "step": 44040
    },
    {
      "entropy": 5.65521388053894,
      "epoch": 4.946706351434829,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002827953208894503,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23481371104717255,
      "num_tokens": 98320149.0,
      "step": 44045
    },
    {
      "entropy": 5.666835355758667,
      "epoch": 4.947267928342787,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002827551961497593,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22932618260383605,
      "num_tokens": 98330816.0,
      "step": 44050
    },
    {
      "entropy": 5.622478294372558,
      "epoch": 4.947829505250744,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002827150711631398,
      "loss": 4.8029,
      "mean_token_accuracy": 0.23860319703817368,
      "num_tokens": 98342041.0,
      "step": 44055
    },
    {
      "entropy": 5.702459001541138,
      "epoch": 4.948391082158702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028267494593086945,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22741432040929793,
      "num_tokens": 98353891.0,
      "step": 44060
    },
    {
      "entropy": 5.747168493270874,
      "epoch": 4.948952659066659,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028263482045422597,
      "loss": 5.012,
      "mean_token_accuracy": 0.23083198070526123,
      "num_tokens": 98364931.0,
      "step": 44065
    },
    {
      "entropy": 5.783327960968018,
      "epoch": 4.9495142359746165,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002825946947344867,
      "loss": 4.9621,
      "mean_token_accuracy": 0.23569077253341675,
      "num_tokens": 98376787.0,
      "step": 44070
    },
    {
      "entropy": 5.752938938140869,
      "epoch": 4.950075812882575,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002825545687729295,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22574138790369033,
      "num_tokens": 98387750.0,
      "step": 44075
    },
    {
      "entropy": 5.653963613510132,
      "epoch": 4.950637389790532,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028251444257083186,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2380245089530945,
      "num_tokens": 98399309.0,
      "step": 44080
    },
    {
      "entropy": 5.63860969543457,
      "epoch": 4.951198966698489,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028247431612947146,
      "loss": 4.8487,
      "mean_token_accuracy": 0.22933264076709747,
      "num_tokens": 98410850.0,
      "step": 44085
    },
    {
      "entropy": 5.7653247833251955,
      "epoch": 4.951760543606447,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002824341894501259,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22705175429582597,
      "num_tokens": 98422166.0,
      "step": 44090
    },
    {
      "entropy": 5.717971706390381,
      "epoch": 4.952322120514404,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028239406253407286,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22902271896600723,
      "num_tokens": 98432885.0,
      "step": 44095
    },
    {
      "entropy": 5.725024700164795,
      "epoch": 4.952883697422362,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028235393538259007,
      "loss": 4.9547,
      "mean_token_accuracy": 0.23068824261426926,
      "num_tokens": 98444513.0,
      "step": 44100
    },
    {
      "entropy": 5.718712377548218,
      "epoch": 4.95344527433032,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002823138079969551,
      "loss": 4.9432,
      "mean_token_accuracy": 0.2304680675268173,
      "num_tokens": 98456009.0,
      "step": 44105
    },
    {
      "entropy": 5.806165409088135,
      "epoch": 4.954006851238277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028227368037844557,
      "loss": 5.0248,
      "mean_token_accuracy": 0.22819639891386032,
      "num_tokens": 98466282.0,
      "step": 44110
    },
    {
      "entropy": 5.733563041687011,
      "epoch": 4.954568428146235,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028223355252833933,
      "loss": 4.9493,
      "mean_token_accuracy": 0.23085199743509294,
      "num_tokens": 98478334.0,
      "step": 44115
    },
    {
      "entropy": 5.725430870056153,
      "epoch": 4.955130005054192,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028219342444791396,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22913835644721986,
      "num_tokens": 98488446.0,
      "step": 44120
    },
    {
      "entropy": 5.657358026504516,
      "epoch": 4.955691581962149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028215329613844715,
      "loss": 4.7962,
      "mean_token_accuracy": 0.24076959043741225,
      "num_tokens": 98499698.0,
      "step": 44125
    },
    {
      "entropy": 5.721067142486572,
      "epoch": 4.9562531588701075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028211316760121656,
      "loss": 4.9666,
      "mean_token_accuracy": 0.23681317418813705,
      "num_tokens": 98509980.0,
      "step": 44130
    },
    {
      "entropy": 5.805875301361084,
      "epoch": 4.956814735778065,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002820730388375001,
      "loss": 5.0007,
      "mean_token_accuracy": 0.23556143492460252,
      "num_tokens": 98521406.0,
      "step": 44135
    },
    {
      "entropy": 5.810909605026245,
      "epoch": 4.957376312686022,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002820329098485752,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2304034173488617,
      "num_tokens": 98532845.0,
      "step": 44140
    },
    {
      "entropy": 5.649046993255615,
      "epoch": 4.95793788959398,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028199278063571976,
      "loss": 4.8243,
      "mean_token_accuracy": 0.24094609469175338,
      "num_tokens": 98543639.0,
      "step": 44145
    },
    {
      "entropy": 5.74506106376648,
      "epoch": 4.958499466501937,
      "grad_norm": 1.125,
      "learning_rate": 0.00028195265120021157,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2272594153881073,
      "num_tokens": 98554604.0,
      "step": 44150
    },
    {
      "entropy": 5.705216264724731,
      "epoch": 4.9590610434098945,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028191252154332816,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22363436669111253,
      "num_tokens": 98566388.0,
      "step": 44155
    },
    {
      "entropy": 5.714275693893432,
      "epoch": 4.959622620317853,
      "grad_norm": 1.328125,
      "learning_rate": 0.00028187239166634737,
      "loss": 4.9989,
      "mean_token_accuracy": 0.23040703535079957,
      "num_tokens": 98577215.0,
      "step": 44160
    },
    {
      "entropy": 5.7931702613830565,
      "epoch": 4.96018419722581,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028183226157054705,
      "loss": 5.1205,
      "mean_token_accuracy": 0.22671267688274382,
      "num_tokens": 98588640.0,
      "step": 44165
    },
    {
      "entropy": 5.745516395568847,
      "epoch": 4.960745774133768,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002817921312572048,
      "loss": 4.9059,
      "mean_token_accuracy": 0.2373422220349312,
      "num_tokens": 98599710.0,
      "step": 44170
    },
    {
      "entropy": 5.704620361328125,
      "epoch": 4.961307351041725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002817520007275985,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2292394682765007,
      "num_tokens": 98611510.0,
      "step": 44175
    },
    {
      "entropy": 5.685167551040649,
      "epoch": 4.961868927949682,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002817118699830058,
      "loss": 4.8982,
      "mean_token_accuracy": 0.23270812779664993,
      "num_tokens": 98622982.0,
      "step": 44180
    },
    {
      "entropy": 5.819343423843383,
      "epoch": 4.9624305048576405,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002816717390247046,
      "loss": 5.0595,
      "mean_token_accuracy": 0.23180107027292252,
      "num_tokens": 98634216.0,
      "step": 44185
    },
    {
      "entropy": 5.785608291625977,
      "epoch": 4.962992081765598,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002816316078539727,
      "loss": 5.0463,
      "mean_token_accuracy": 0.23045540750026702,
      "num_tokens": 98646469.0,
      "step": 44190
    },
    {
      "entropy": 5.784533596038818,
      "epoch": 4.963553658673556,
      "grad_norm": 1.0,
      "learning_rate": 0.00028159147647208784,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2204393208026886,
      "num_tokens": 98659249.0,
      "step": 44195
    },
    {
      "entropy": 5.694730758666992,
      "epoch": 4.964115235581513,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002815513448803277,
      "loss": 4.85,
      "mean_token_accuracy": 0.23787922710180281,
      "num_tokens": 98671322.0,
      "step": 44200
    },
    {
      "entropy": 5.729255294799804,
      "epoch": 4.96467681248947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002815112130799703,
      "loss": 4.9777,
      "mean_token_accuracy": 0.23442451506853104,
      "num_tokens": 98684236.0,
      "step": 44205
    },
    {
      "entropy": 5.635502815246582,
      "epoch": 4.965238389397428,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028147108107229324,
      "loss": 4.856,
      "mean_token_accuracy": 0.23120892643928528,
      "num_tokens": 98694581.0,
      "step": 44210
    },
    {
      "entropy": 5.717007827758789,
      "epoch": 4.965799966305386,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028143094885857455,
      "loss": 4.967,
      "mean_token_accuracy": 0.22831621915102004,
      "num_tokens": 98705002.0,
      "step": 44215
    },
    {
      "entropy": 5.700779390335083,
      "epoch": 4.966361543213343,
      "grad_norm": 1.28125,
      "learning_rate": 0.000281390816440092,
      "loss": 4.9298,
      "mean_token_accuracy": 0.23178231865167617,
      "num_tokens": 98715131.0,
      "step": 44220
    },
    {
      "entropy": 5.623217582702637,
      "epoch": 4.966923120121301,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002813506838181233,
      "loss": 4.8213,
      "mean_token_accuracy": 0.24009584188461303,
      "num_tokens": 98724992.0,
      "step": 44225
    },
    {
      "entropy": 5.692883014678955,
      "epoch": 4.967484697029258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002813105509939464,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22823852598667144,
      "num_tokens": 98736651.0,
      "step": 44230
    },
    {
      "entropy": 5.748697757720947,
      "epoch": 4.968046273937215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000281270417968839,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2354969561100006,
      "num_tokens": 98747365.0,
      "step": 44235
    },
    {
      "entropy": 5.720853853225708,
      "epoch": 4.968607850845173,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002812302847440791,
      "loss": 5.0182,
      "mean_token_accuracy": 0.22603610754013062,
      "num_tokens": 98759610.0,
      "step": 44240
    },
    {
      "entropy": 5.7220635414123535,
      "epoch": 4.969169427753131,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028119015132094455,
      "loss": 4.9534,
      "mean_token_accuracy": 0.23707758635282516,
      "num_tokens": 98770166.0,
      "step": 44245
    },
    {
      "entropy": 5.723124027252197,
      "epoch": 4.969731004661089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002811500177007132,
      "loss": 4.9561,
      "mean_token_accuracy": 0.23271597921848297,
      "num_tokens": 98782416.0,
      "step": 44250
    },
    {
      "entropy": 5.744660234451294,
      "epoch": 4.970292581569046,
      "grad_norm": 1.125,
      "learning_rate": 0.00028110988388466287,
      "loss": 4.9834,
      "mean_token_accuracy": 0.22783474624156952,
      "num_tokens": 98793341.0,
      "step": 44255
    },
    {
      "entropy": 5.611420011520385,
      "epoch": 4.970854158477003,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002810697498740715,
      "loss": 4.8691,
      "mean_token_accuracy": 0.23662999123334885,
      "num_tokens": 98803642.0,
      "step": 44260
    },
    {
      "entropy": 5.665055418014527,
      "epoch": 4.971415735384961,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002810296156702169,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2347092092037201,
      "num_tokens": 98814650.0,
      "step": 44265
    },
    {
      "entropy": 5.733565044403076,
      "epoch": 4.9719773122929185,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028098948127437695,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2296796604990959,
      "num_tokens": 98825928.0,
      "step": 44270
    },
    {
      "entropy": 5.741301393508911,
      "epoch": 4.972538889200876,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002809493466878297,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23006316572427749,
      "num_tokens": 98837447.0,
      "step": 44275
    },
    {
      "entropy": 5.680266666412353,
      "epoch": 4.973100466108834,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028090921191185295,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2299267277121544,
      "num_tokens": 98847720.0,
      "step": 44280
    },
    {
      "entropy": 5.620090961456299,
      "epoch": 4.973662043016791,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002808690769477245,
      "loss": 4.8591,
      "mean_token_accuracy": 0.24003641158342362,
      "num_tokens": 98859315.0,
      "step": 44285
    },
    {
      "entropy": 5.637885522842407,
      "epoch": 4.974223619924748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002808289417967225,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22862363308668138,
      "num_tokens": 98869921.0,
      "step": 44290
    },
    {
      "entropy": 5.676765251159668,
      "epoch": 4.974785196832706,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028078880646012467,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23998788446187974,
      "num_tokens": 98881473.0,
      "step": 44295
    },
    {
      "entropy": 5.689709711074829,
      "epoch": 4.975346773740664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028074867093920897,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23362044841051102,
      "num_tokens": 98892601.0,
      "step": 44300
    },
    {
      "entropy": 5.7459310531616214,
      "epoch": 4.975908350648622,
      "grad_norm": 1.125,
      "learning_rate": 0.00028070853523525337,
      "loss": 4.9858,
      "mean_token_accuracy": 0.2299649730324745,
      "num_tokens": 98903823.0,
      "step": 44305
    },
    {
      "entropy": 5.672928380966186,
      "epoch": 4.976469927556579,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028066839934953573,
      "loss": 4.9624,
      "mean_token_accuracy": 0.22647807598114014,
      "num_tokens": 98914266.0,
      "step": 44310
    },
    {
      "entropy": 5.687666702270508,
      "epoch": 4.977031504464536,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028062826328333415,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22797092646360398,
      "num_tokens": 98924693.0,
      "step": 44315
    },
    {
      "entropy": 5.6677971363067625,
      "epoch": 4.977593081372494,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002805881270379265,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22816374003887177,
      "num_tokens": 98936081.0,
      "step": 44320
    },
    {
      "entropy": 5.694923686981201,
      "epoch": 4.9781546582804515,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002805479906145906,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23110885322093963,
      "num_tokens": 98946495.0,
      "step": 44325
    },
    {
      "entropy": 5.676246595382691,
      "epoch": 4.978716235188409,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002805078540146046,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23182678818702698,
      "num_tokens": 98956873.0,
      "step": 44330
    },
    {
      "entropy": 5.640589904785156,
      "epoch": 4.979277812096367,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002804677172392464,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23425389975309371,
      "num_tokens": 98968349.0,
      "step": 44335
    },
    {
      "entropy": 5.663880157470703,
      "epoch": 4.979839389004324,
      "grad_norm": 1.203125,
      "learning_rate": 0.000280427580289794,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23623311817646026,
      "num_tokens": 98978558.0,
      "step": 44340
    },
    {
      "entropy": 5.767185068130493,
      "epoch": 4.980400965912281,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028038744316752524,
      "loss": 5.0555,
      "mean_token_accuracy": 0.23061203807592393,
      "num_tokens": 98990435.0,
      "step": 44345
    },
    {
      "entropy": 5.75434718132019,
      "epoch": 4.980962542820239,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028034730587371824,
      "loss": 4.9254,
      "mean_token_accuracy": 0.231184720993042,
      "num_tokens": 99001516.0,
      "step": 44350
    },
    {
      "entropy": 5.721002578735352,
      "epoch": 4.9815241197281965,
      "grad_norm": 1.25,
      "learning_rate": 0.00028030716840965096,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2300449639558792,
      "num_tokens": 99012810.0,
      "step": 44355
    },
    {
      "entropy": 5.767683601379394,
      "epoch": 4.982085696636155,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028026703077660143,
      "loss": 5.031,
      "mean_token_accuracy": 0.21979927867650986,
      "num_tokens": 99023342.0,
      "step": 44360
    },
    {
      "entropy": 5.648006772994995,
      "epoch": 4.982647273544112,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002802268929758475,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23902892023324968,
      "num_tokens": 99034047.0,
      "step": 44365
    },
    {
      "entropy": 5.725071048736572,
      "epoch": 4.983208850452069,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002801867550086674,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22991015762090683,
      "num_tokens": 99045333.0,
      "step": 44370
    },
    {
      "entropy": 5.738540887832642,
      "epoch": 4.983770427360027,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002801466168763389,
      "loss": 4.9929,
      "mean_token_accuracy": 0.23321648240089415,
      "num_tokens": 99057582.0,
      "step": 44375
    },
    {
      "entropy": 5.730814695358276,
      "epoch": 4.984332004267984,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002801064785801401,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22815045863389968,
      "num_tokens": 99067940.0,
      "step": 44380
    },
    {
      "entropy": 5.729376173019409,
      "epoch": 4.9848935811759425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028006634012134915,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22505124062299728,
      "num_tokens": 99079408.0,
      "step": 44385
    },
    {
      "entropy": 5.668135070800782,
      "epoch": 4.9854551580839,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028002620150124386,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23581558912992479,
      "num_tokens": 99090320.0,
      "step": 44390
    },
    {
      "entropy": 5.723673152923584,
      "epoch": 4.986016734991857,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027998606272110237,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22431714236736297,
      "num_tokens": 99101503.0,
      "step": 44395
    },
    {
      "entropy": 5.783992719650269,
      "epoch": 4.986578311899815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002799459237822028,
      "loss": 5.0085,
      "mean_token_accuracy": 0.22508209496736525,
      "num_tokens": 99112768.0,
      "step": 44400
    },
    {
      "entropy": 5.7114263534545895,
      "epoch": 4.987139888807772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027990578468582304,
      "loss": 4.9327,
      "mean_token_accuracy": 0.23365508317947387,
      "num_tokens": 99123473.0,
      "step": 44405
    },
    {
      "entropy": 5.761605930328369,
      "epoch": 4.9877014657157295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002798656454332412,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2320594757795334,
      "num_tokens": 99136005.0,
      "step": 44410
    },
    {
      "entropy": 5.784644842147827,
      "epoch": 4.988263042623688,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027982550602573524,
      "loss": 5.0092,
      "mean_token_accuracy": 0.2349707677960396,
      "num_tokens": 99147494.0,
      "step": 44415
    },
    {
      "entropy": 5.790716743469238,
      "epoch": 4.988824619531645,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002797853664645834,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22385508120059966,
      "num_tokens": 99159255.0,
      "step": 44420
    },
    {
      "entropy": 5.737120866775513,
      "epoch": 4.989386196439602,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002797452267510637,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2346154510974884,
      "num_tokens": 99169406.0,
      "step": 44425
    },
    {
      "entropy": 5.669383382797241,
      "epoch": 4.98994777334756,
      "grad_norm": 1.21875,
      "learning_rate": 0.000279705086886454,
      "loss": 4.926,
      "mean_token_accuracy": 0.23649099022150039,
      "num_tokens": 99180392.0,
      "step": 44430
    },
    {
      "entropy": 5.667870664596558,
      "epoch": 4.990509350255517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002796649468720326,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23260220140218735,
      "num_tokens": 99190387.0,
      "step": 44435
    },
    {
      "entropy": 5.741183090209961,
      "epoch": 4.9910709271634754,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027962480670907743,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2300523743033409,
      "num_tokens": 99202070.0,
      "step": 44440
    },
    {
      "entropy": 5.772120571136474,
      "epoch": 4.991632504071433,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027958466639886663,
      "loss": 4.9933,
      "mean_token_accuracy": 0.22991181910037994,
      "num_tokens": 99213323.0,
      "step": 44445
    },
    {
      "entropy": 5.735208129882812,
      "epoch": 4.99219408097939,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027954452594267825,
      "loss": 4.9413,
      "mean_token_accuracy": 0.2305689349770546,
      "num_tokens": 99224178.0,
      "step": 44450
    },
    {
      "entropy": 5.806602907180786,
      "epoch": 4.992755657887348,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027950438534179046,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2314574673771858,
      "num_tokens": 99234538.0,
      "step": 44455
    },
    {
      "entropy": 5.749233150482178,
      "epoch": 4.993317234795305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027946424459748126,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22919864803552628,
      "num_tokens": 99245737.0,
      "step": 44460
    },
    {
      "entropy": 5.714778470993042,
      "epoch": 4.993878811703262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027942410371102877,
      "loss": 4.9396,
      "mean_token_accuracy": 0.23036699891090393,
      "num_tokens": 99257047.0,
      "step": 44465
    },
    {
      "entropy": 5.795732688903809,
      "epoch": 4.9944403886112205,
      "grad_norm": 1.25,
      "learning_rate": 0.0002793839626837111,
      "loss": 5.0451,
      "mean_token_accuracy": 0.2258838027715683,
      "num_tokens": 99267664.0,
      "step": 44470
    },
    {
      "entropy": 5.6479674816131595,
      "epoch": 4.995001965519178,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027934382151680644,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22574071288108827,
      "num_tokens": 99279270.0,
      "step": 44475
    },
    {
      "entropy": 5.698285293579102,
      "epoch": 4.995563542427135,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027930368021159273,
      "loss": 4.8882,
      "mean_token_accuracy": 0.23377573490142822,
      "num_tokens": 99289724.0,
      "step": 44480
    },
    {
      "entropy": 5.806068754196167,
      "epoch": 4.996125119335093,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002792635387693481,
      "loss": 4.9814,
      "mean_token_accuracy": 0.23163032829761504,
      "num_tokens": 99300326.0,
      "step": 44485
    },
    {
      "entropy": 5.714386367797852,
      "epoch": 4.99668669624305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027922339719135093,
      "loss": 4.8908,
      "mean_token_accuracy": 0.23651191741228103,
      "num_tokens": 99312073.0,
      "step": 44490
    },
    {
      "entropy": 5.763074779510498,
      "epoch": 4.997248273151008,
      "grad_norm": 1.203125,
      "learning_rate": 0.000279183255478879,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2278701975941658,
      "num_tokens": 99323938.0,
      "step": 44495
    },
    {
      "entropy": 5.684124183654785,
      "epoch": 4.997809850058966,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027914311363321073,
      "loss": 4.9002,
      "mean_token_accuracy": 0.2302742198109627,
      "num_tokens": 99334773.0,
      "step": 44500
    },
    {
      "entropy": 5.685778713226318,
      "epoch": 4.998371426966923,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027910297165562406,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23846668899059295,
      "num_tokens": 99345688.0,
      "step": 44505
    },
    {
      "entropy": 5.757121467590332,
      "epoch": 4.998933003874881,
      "grad_norm": 1.25,
      "learning_rate": 0.00027906282954739714,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2330136090517044,
      "num_tokens": 99356307.0,
      "step": 44510
    },
    {
      "entropy": 5.665755653381348,
      "epoch": 4.999494580782838,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027902268730980816,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23955320864915847,
      "num_tokens": 99367348.0,
      "step": 44515
    },
    {
      "entropy": 5.727226310306126,
      "epoch": 5.0,
      "grad_norm": 6.15625,
      "learning_rate": 0.0002789825449441353,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2383095638619529,
      "num_tokens": 99377525.0,
      "step": 44520
    },
    {
      "entropy": 5.716814279556274,
      "epoch": 5.000561576907957,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002789424024516566,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23673781901597976,
      "num_tokens": 99388342.0,
      "step": 44525
    },
    {
      "entropy": 5.758033227920532,
      "epoch": 5.001123153815915,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027890225983365034,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2342769280076027,
      "num_tokens": 99400387.0,
      "step": 44530
    },
    {
      "entropy": 5.61524133682251,
      "epoch": 5.0016847307238725,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002788621170913946,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23646885007619858,
      "num_tokens": 99411232.0,
      "step": 44535
    },
    {
      "entropy": 5.631529760360718,
      "epoch": 5.00224630763183,
      "grad_norm": 1.25,
      "learning_rate": 0.00027882197422616746,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23288278877735138,
      "num_tokens": 99422839.0,
      "step": 44540
    },
    {
      "entropy": 5.746863174438476,
      "epoch": 5.002807884539788,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002787818312392473,
      "loss": 4.88,
      "mean_token_accuracy": 0.23630152493715287,
      "num_tokens": 99434835.0,
      "step": 44545
    },
    {
      "entropy": 5.699570083618164,
      "epoch": 5.003369461447745,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027874168813191207,
      "loss": 4.8808,
      "mean_token_accuracy": 0.23727004379034042,
      "num_tokens": 99445705.0,
      "step": 44550
    },
    {
      "entropy": 5.674230480194092,
      "epoch": 5.003931038355703,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002787015449054401,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2373849019408226,
      "num_tokens": 99457127.0,
      "step": 44555
    },
    {
      "entropy": 5.632333278656006,
      "epoch": 5.00449261526366,
      "grad_norm": 1.25,
      "learning_rate": 0.0002786614015611095,
      "loss": 4.845,
      "mean_token_accuracy": 0.23725647628307342,
      "num_tokens": 99468653.0,
      "step": 44560
    },
    {
      "entropy": 5.809616899490356,
      "epoch": 5.005054192171618,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027862125810019844,
      "loss": 5.0746,
      "mean_token_accuracy": 0.22167638391256334,
      "num_tokens": 99481238.0,
      "step": 44565
    },
    {
      "entropy": 5.720807361602783,
      "epoch": 5.005615769079576,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027858111452398515,
      "loss": 4.8612,
      "mean_token_accuracy": 0.24150482714176177,
      "num_tokens": 99493025.0,
      "step": 44570
    },
    {
      "entropy": 5.678908014297486,
      "epoch": 5.006177345987533,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027854097083374773,
      "loss": 4.9307,
      "mean_token_accuracy": 0.23772085458040237,
      "num_tokens": 99504675.0,
      "step": 44575
    },
    {
      "entropy": 5.745387601852417,
      "epoch": 5.00673892289549,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002785008270307645,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23822046667337418,
      "num_tokens": 99516005.0,
      "step": 44580
    },
    {
      "entropy": 5.800759649276733,
      "epoch": 5.007300499803448,
      "grad_norm": 1.25,
      "learning_rate": 0.0002784606831163136,
      "loss": 5.0413,
      "mean_token_accuracy": 0.22228946089744567,
      "num_tokens": 99526917.0,
      "step": 44585
    },
    {
      "entropy": 5.640524911880493,
      "epoch": 5.0078620767114055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027842053909167314,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2334575578570366,
      "num_tokens": 99538799.0,
      "step": 44590
    },
    {
      "entropy": 5.69981164932251,
      "epoch": 5.008423653619364,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027838039495812133,
      "loss": 4.8938,
      "mean_token_accuracy": 0.23995261937379836,
      "num_tokens": 99550431.0,
      "step": 44595
    },
    {
      "entropy": 5.66878490447998,
      "epoch": 5.008985230527321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027834025071693655,
      "loss": 4.8341,
      "mean_token_accuracy": 0.2354969173669815,
      "num_tokens": 99561562.0,
      "step": 44600
    },
    {
      "entropy": 5.6604729175567625,
      "epoch": 5.009546807435278,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027830010636939684,
      "loss": 4.827,
      "mean_token_accuracy": 0.2386807456612587,
      "num_tokens": 99572101.0,
      "step": 44605
    },
    {
      "entropy": 5.692048215866089,
      "epoch": 5.010108384343236,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027825996191678043,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23031797409057617,
      "num_tokens": 99583186.0,
      "step": 44610
    },
    {
      "entropy": 5.677436971664429,
      "epoch": 5.010669961251193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027821981736036555,
      "loss": 4.8762,
      "mean_token_accuracy": 0.23972527235746383,
      "num_tokens": 99593716.0,
      "step": 44615
    },
    {
      "entropy": 5.773723220825195,
      "epoch": 5.011231538159151,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002781796727014305,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2273658737540245,
      "num_tokens": 99605798.0,
      "step": 44620
    },
    {
      "entropy": 5.694684457778931,
      "epoch": 5.011793115067109,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002781395279412535,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23223683834075928,
      "num_tokens": 99617938.0,
      "step": 44625
    },
    {
      "entropy": 5.708831644058227,
      "epoch": 5.012354691975066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027809938308111254,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23769476860761643,
      "num_tokens": 99628761.0,
      "step": 44630
    },
    {
      "entropy": 5.699739027023315,
      "epoch": 5.012916268883023,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027805923812228615,
      "loss": 4.928,
      "mean_token_accuracy": 0.23276339322328568,
      "num_tokens": 99640072.0,
      "step": 44635
    },
    {
      "entropy": 5.690370321273804,
      "epoch": 5.013477845790981,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027801909306605243,
      "loss": 4.959,
      "mean_token_accuracy": 0.2302428036928177,
      "num_tokens": 99651280.0,
      "step": 44640
    },
    {
      "entropy": 5.695747518539429,
      "epoch": 5.014039422698938,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002779789479136896,
      "loss": 4.8876,
      "mean_token_accuracy": 0.2331558182835579,
      "num_tokens": 99661574.0,
      "step": 44645
    },
    {
      "entropy": 5.692471837997436,
      "epoch": 5.0146009996068965,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002779388026664759,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23961824923753738,
      "num_tokens": 99671092.0,
      "step": 44650
    },
    {
      "entropy": 5.837254476547241,
      "epoch": 5.015162576514854,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002778986573256896,
      "loss": 5.033,
      "mean_token_accuracy": 0.23124216347932816,
      "num_tokens": 99682104.0,
      "step": 44655
    },
    {
      "entropy": 5.7790367603302,
      "epoch": 5.015724153422811,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002778585118926089,
      "loss": 4.9668,
      "mean_token_accuracy": 0.23275788277387618,
      "num_tokens": 99692152.0,
      "step": 44660
    },
    {
      "entropy": 5.684384346008301,
      "epoch": 5.016285730330769,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027781836636851207,
      "loss": 4.8648,
      "mean_token_accuracy": 0.23751218169927596,
      "num_tokens": 99703807.0,
      "step": 44665
    },
    {
      "entropy": 5.810228252410889,
      "epoch": 5.016847307238726,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027777822075467735,
      "loss": 4.957,
      "mean_token_accuracy": 0.23504173308610915,
      "num_tokens": 99714039.0,
      "step": 44670
    },
    {
      "entropy": 5.646932935714721,
      "epoch": 5.0174088841466835,
      "grad_norm": 1.21875,
      "learning_rate": 0.000277738075052383,
      "loss": 4.8754,
      "mean_token_accuracy": 0.23345767855644226,
      "num_tokens": 99725215.0,
      "step": 44675
    },
    {
      "entropy": 5.669582366943359,
      "epoch": 5.017970461054642,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002776979292629074,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23436338901519777,
      "num_tokens": 99736281.0,
      "step": 44680
    },
    {
      "entropy": 5.737140655517578,
      "epoch": 5.018532037962599,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002776577833875285,
      "loss": 4.9028,
      "mean_token_accuracy": 0.2379188969731331,
      "num_tokens": 99747544.0,
      "step": 44685
    },
    {
      "entropy": 5.676158237457275,
      "epoch": 5.019093614870557,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027761763742752485,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23765848129987716,
      "num_tokens": 99758806.0,
      "step": 44690
    },
    {
      "entropy": 5.701710939407349,
      "epoch": 5.019655191778514,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002775774913841746,
      "loss": 4.8656,
      "mean_token_accuracy": 0.23003000169992446,
      "num_tokens": 99769477.0,
      "step": 44695
    },
    {
      "entropy": 5.667230272293091,
      "epoch": 5.020216768686471,
      "grad_norm": 1.171875,
      "learning_rate": 0.000277537345258756,
      "loss": 4.8849,
      "mean_token_accuracy": 0.23019291162490846,
      "num_tokens": 99780283.0,
      "step": 44700
    },
    {
      "entropy": 5.794473266601562,
      "epoch": 5.0207783455944295,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027749719905254737,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2366488069295883,
      "num_tokens": 99790119.0,
      "step": 44705
    },
    {
      "entropy": 5.749775695800781,
      "epoch": 5.021339922502387,
      "grad_norm": 1.21875,
      "learning_rate": 0.000277457052766827,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22922950387001037,
      "num_tokens": 99801619.0,
      "step": 44710
    },
    {
      "entropy": 5.690867519378662,
      "epoch": 5.021901499410344,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027741690640287306,
      "loss": 4.8651,
      "mean_token_accuracy": 0.2444425567984581,
      "num_tokens": 99812972.0,
      "step": 44715
    },
    {
      "entropy": 5.665153312683105,
      "epoch": 5.022463076318302,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002773767599619639,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2357382595539093,
      "num_tokens": 99823667.0,
      "step": 44720
    },
    {
      "entropy": 5.6408487319946286,
      "epoch": 5.023024653226259,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027733661344537776,
      "loss": 4.7777,
      "mean_token_accuracy": 0.24484476894140245,
      "num_tokens": 99834654.0,
      "step": 44725
    },
    {
      "entropy": 5.7478149890899655,
      "epoch": 5.0235862301342165,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000277296466854393,
      "loss": 4.9666,
      "mean_token_accuracy": 0.24019135385751725,
      "num_tokens": 99846393.0,
      "step": 44730
    },
    {
      "entropy": 5.686874628067017,
      "epoch": 5.024147807042175,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002772563201902877,
      "loss": 4.8187,
      "mean_token_accuracy": 0.24423286616802214,
      "num_tokens": 99857206.0,
      "step": 44735
    },
    {
      "entropy": 5.815841436386108,
      "epoch": 5.024709383950132,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027721617345434035,
      "loss": 5.0333,
      "mean_token_accuracy": 0.2379532814025879,
      "num_tokens": 99871497.0,
      "step": 44740
    },
    {
      "entropy": 5.6725541114807125,
      "epoch": 5.02527096085809,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027717602664782925,
      "loss": 4.954,
      "mean_token_accuracy": 0.22501264214515687,
      "num_tokens": 99882337.0,
      "step": 44745
    },
    {
      "entropy": 5.649121570587158,
      "epoch": 5.025832537766047,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002771358797720326,
      "loss": 4.8262,
      "mean_token_accuracy": 0.24108425080776213,
      "num_tokens": 99892863.0,
      "step": 44750
    },
    {
      "entropy": 5.667015886306762,
      "epoch": 5.026394114674004,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027709573282822864,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23602715283632278,
      "num_tokens": 99903566.0,
      "step": 44755
    },
    {
      "entropy": 5.737066555023193,
      "epoch": 5.026955691581962,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002770555858176958,
      "loss": 4.8655,
      "mean_token_accuracy": 0.2416760429739952,
      "num_tokens": 99913466.0,
      "step": 44760
    },
    {
      "entropy": 5.713699913024902,
      "epoch": 5.02751726848992,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002770154387417123,
      "loss": 4.9679,
      "mean_token_accuracy": 0.23322279900312423,
      "num_tokens": 99923579.0,
      "step": 44765
    },
    {
      "entropy": 5.76369571685791,
      "epoch": 5.028078845397877,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002769752916015565,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22635047137737274,
      "num_tokens": 99935471.0,
      "step": 44770
    },
    {
      "entropy": 5.8000670909881595,
      "epoch": 5.028640422305835,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027693514439850646,
      "loss": 4.9622,
      "mean_token_accuracy": 0.23693124949932098,
      "num_tokens": 99947082.0,
      "step": 44775
    },
    {
      "entropy": 5.756147146224976,
      "epoch": 5.029201999213792,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027689499713384085,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23181576430797576,
      "num_tokens": 99958186.0,
      "step": 44780
    },
    {
      "entropy": 5.747011995315551,
      "epoch": 5.02976357612175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002768548498088378,
      "loss": 4.9673,
      "mean_token_accuracy": 0.23197509944438935,
      "num_tokens": 99970928.0,
      "step": 44785
    },
    {
      "entropy": 5.626775503158569,
      "epoch": 5.0303251530297075,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002768147024247756,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23794748783111572,
      "num_tokens": 99982549.0,
      "step": 44790
    },
    {
      "entropy": 5.638612174987793,
      "epoch": 5.030886729937665,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002767745549829325,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23876197785139083,
      "num_tokens": 99993835.0,
      "step": 44795
    },
    {
      "entropy": 5.7610198020935055,
      "epoch": 5.031448306845623,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027673440748458687,
      "loss": 4.9136,
      "mean_token_accuracy": 0.23268130123615266,
      "num_tokens": 100004336.0,
      "step": 44800
    },
    {
      "entropy": 5.679032278060913,
      "epoch": 5.03200988375358,
      "grad_norm": 1.25,
      "learning_rate": 0.00027669425993101716,
      "loss": 4.9114,
      "mean_token_accuracy": 0.236665478348732,
      "num_tokens": 100015334.0,
      "step": 44805
    },
    {
      "entropy": 5.718606805801391,
      "epoch": 5.032571460661537,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002766541123235014,
      "loss": 4.9561,
      "mean_token_accuracy": 0.2249349519610405,
      "num_tokens": 100026665.0,
      "step": 44810
    },
    {
      "entropy": 5.659159183502197,
      "epoch": 5.033133037569495,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002766139646633182,
      "loss": 4.8518,
      "mean_token_accuracy": 0.2396455243229866,
      "num_tokens": 100037263.0,
      "step": 44815
    },
    {
      "entropy": 5.667156314849853,
      "epoch": 5.033694614477453,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027657381695174563,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22935565412044526,
      "num_tokens": 100048482.0,
      "step": 44820
    },
    {
      "entropy": 5.646645498275757,
      "epoch": 5.03425619138541,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027653366919006217,
      "loss": 4.9004,
      "mean_token_accuracy": 0.23064118176698684,
      "num_tokens": 100059351.0,
      "step": 44825
    },
    {
      "entropy": 5.712008810043335,
      "epoch": 5.034817768293368,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002764935213795461,
      "loss": 4.7731,
      "mean_token_accuracy": 0.24459931254386902,
      "num_tokens": 100069924.0,
      "step": 44830
    },
    {
      "entropy": 5.666696977615357,
      "epoch": 5.035379345201325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002764533735214757,
      "loss": 4.8191,
      "mean_token_accuracy": 0.24493090212345123,
      "num_tokens": 100080733.0,
      "step": 44835
    },
    {
      "entropy": 5.6637434482574465,
      "epoch": 5.035940922109283,
      "grad_norm": 1.3125,
      "learning_rate": 0.00027641322561712935,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23304588198661805,
      "num_tokens": 100091834.0,
      "step": 44840
    },
    {
      "entropy": 5.830565118789673,
      "epoch": 5.0365024990172405,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027637307766778524,
      "loss": 5.0903,
      "mean_token_accuracy": 0.22275288850069047,
      "num_tokens": 100103872.0,
      "step": 44845
    },
    {
      "entropy": 5.799667263031006,
      "epoch": 5.037064075925198,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000276332929674722,
      "loss": 4.9631,
      "mean_token_accuracy": 0.23196862190961837,
      "num_tokens": 100114391.0,
      "step": 44850
    },
    {
      "entropy": 5.730702066421509,
      "epoch": 5.037625652833156,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027629278163921763,
      "loss": 4.8818,
      "mean_token_accuracy": 0.24378167390823363,
      "num_tokens": 100125000.0,
      "step": 44855
    },
    {
      "entropy": 5.745064115524292,
      "epoch": 5.038187229741113,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027625263356255067,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2374044254422188,
      "num_tokens": 100135394.0,
      "step": 44860
    },
    {
      "entropy": 5.6431139469146725,
      "epoch": 5.03874880664907,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002762124854459992,
      "loss": 4.8135,
      "mean_token_accuracy": 0.24476829618215562,
      "num_tokens": 100146610.0,
      "step": 44865
    },
    {
      "entropy": 5.6953528881072994,
      "epoch": 5.039310383557028,
      "grad_norm": 1.359375,
      "learning_rate": 0.00027617233729084194,
      "loss": 4.9263,
      "mean_token_accuracy": 0.23628598153591157,
      "num_tokens": 100156936.0,
      "step": 44870
    },
    {
      "entropy": 5.694996309280396,
      "epoch": 5.0398719604649855,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027613218909835693,
      "loss": 4.8812,
      "mean_token_accuracy": 0.23806748986244203,
      "num_tokens": 100168406.0,
      "step": 44875
    },
    {
      "entropy": 5.633799314498901,
      "epoch": 5.040433537372944,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002760920408698225,
      "loss": 4.8585,
      "mean_token_accuracy": 0.2288263663649559,
      "num_tokens": 100179317.0,
      "step": 44880
    },
    {
      "entropy": 5.634406900405883,
      "epoch": 5.040995114280901,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027605189260651716,
      "loss": 4.7773,
      "mean_token_accuracy": 0.24233395904302596,
      "num_tokens": 100190916.0,
      "step": 44885
    },
    {
      "entropy": 5.6495870590209964,
      "epoch": 5.041556691188858,
      "grad_norm": 1.25,
      "learning_rate": 0.0002760117443097191,
      "loss": 4.7803,
      "mean_token_accuracy": 0.2486661657691002,
      "num_tokens": 100201437.0,
      "step": 44890
    },
    {
      "entropy": 5.662458848953247,
      "epoch": 5.042118268096816,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027597159598070675,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2391245484352112,
      "num_tokens": 100212405.0,
      "step": 44895
    },
    {
      "entropy": 5.633211755752564,
      "epoch": 5.042679845004773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027593144762075837,
      "loss": 4.8034,
      "mean_token_accuracy": 0.24386757016181945,
      "num_tokens": 100223428.0,
      "step": 44900
    },
    {
      "entropy": 5.59162220954895,
      "epoch": 5.043241421912731,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027589129923115234,
      "loss": 4.7938,
      "mean_token_accuracy": 0.24118905365467072,
      "num_tokens": 100234128.0,
      "step": 44905
    },
    {
      "entropy": 5.696239376068116,
      "epoch": 5.043802998820689,
      "grad_norm": 1.296875,
      "learning_rate": 0.000275851150813167,
      "loss": 4.9186,
      "mean_token_accuracy": 0.23246662318706512,
      "num_tokens": 100244920.0,
      "step": 44910
    },
    {
      "entropy": 5.6541790008544925,
      "epoch": 5.044364575728646,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002758110023680807,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2388801172375679,
      "num_tokens": 100256518.0,
      "step": 44915
    },
    {
      "entropy": 5.749501323699951,
      "epoch": 5.044926152636603,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002757708538971717,
      "loss": 4.9336,
      "mean_token_accuracy": 0.23850459456443787,
      "num_tokens": 100268203.0,
      "step": 44920
    },
    {
      "entropy": 5.711317157745361,
      "epoch": 5.045487729544561,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002757307054017185,
      "loss": 4.8226,
      "mean_token_accuracy": 0.2410469889640808,
      "num_tokens": 100279440.0,
      "step": 44925
    },
    {
      "entropy": 5.724351596832276,
      "epoch": 5.0460493064525185,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002756905568829994,
      "loss": 4.9472,
      "mean_token_accuracy": 0.23498676866292953,
      "num_tokens": 100290616.0,
      "step": 44930
    },
    {
      "entropy": 5.729852199554443,
      "epoch": 5.046610883360477,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027565040834229266,
      "loss": 4.96,
      "mean_token_accuracy": 0.23032576739788055,
      "num_tokens": 100301602.0,
      "step": 44935
    },
    {
      "entropy": 5.718662738800049,
      "epoch": 5.047172460268434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027561025978087663,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23886573165655137,
      "num_tokens": 100313311.0,
      "step": 44940
    },
    {
      "entropy": 5.736440801620484,
      "epoch": 5.047734037176391,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027557011120002975,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2323760837316513,
      "num_tokens": 100325394.0,
      "step": 44945
    },
    {
      "entropy": 5.70048508644104,
      "epoch": 5.048295614084349,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002755299626010302,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23901067972183226,
      "num_tokens": 100336643.0,
      "step": 44950
    },
    {
      "entropy": 5.712715578079224,
      "epoch": 5.048857190992306,
      "grad_norm": 1.25,
      "learning_rate": 0.0002754898139851565,
      "loss": 4.9324,
      "mean_token_accuracy": 0.235595640540123,
      "num_tokens": 100348409.0,
      "step": 44955
    },
    {
      "entropy": 5.706442642211914,
      "epoch": 5.049418767900264,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027544966535368695,
      "loss": 4.9507,
      "mean_token_accuracy": 0.23555206507444382,
      "num_tokens": 100360115.0,
      "step": 44960
    },
    {
      "entropy": 5.679500675201416,
      "epoch": 5.049980344808222,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002754095167078999,
      "loss": 4.8681,
      "mean_token_accuracy": 0.2364407017827034,
      "num_tokens": 100369618.0,
      "step": 44965
    },
    {
      "entropy": 5.681957864761353,
      "epoch": 5.050541921716179,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002753693680490736,
      "loss": 4.9033,
      "mean_token_accuracy": 0.23480305671691895,
      "num_tokens": 100379470.0,
      "step": 44970
    },
    {
      "entropy": 5.6388462543487545,
      "epoch": 5.051103498624137,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002753292193784865,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23190362751483917,
      "num_tokens": 100390490.0,
      "step": 44975
    },
    {
      "entropy": 5.672218894958496,
      "epoch": 5.051665075532094,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027528907069741696,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23604730069637297,
      "num_tokens": 100402817.0,
      "step": 44980
    },
    {
      "entropy": 5.80064902305603,
      "epoch": 5.052226652440051,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002752489220071433,
      "loss": 4.9776,
      "mean_token_accuracy": 0.22542739808559417,
      "num_tokens": 100413411.0,
      "step": 44985
    },
    {
      "entropy": 5.622895383834839,
      "epoch": 5.0527882293480095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002752087733089438,
      "loss": 4.881,
      "mean_token_accuracy": 0.2347123220562935,
      "num_tokens": 100425118.0,
      "step": 44990
    },
    {
      "entropy": 5.796891117095948,
      "epoch": 5.053349806255967,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002751686246040969,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22688751816749572,
      "num_tokens": 100436544.0,
      "step": 44995
    },
    {
      "entropy": 5.737780284881592,
      "epoch": 5.053911383163924,
      "grad_norm": 1.25,
      "learning_rate": 0.000275128475893881,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23849175721406937,
      "num_tokens": 100448315.0,
      "step": 45000
    },
    {
      "epoch": 5.053911383163924,
      "eval_entropy": 5.418934946180712,
      "eval_loss": 5.051808834075928,
      "eval_mean_token_accuracy": 0.2336572439654574,
      "eval_num_tokens": 100448315.0,
      "eval_runtime": 18.3592,
      "eval_samples_per_second": 1689.233,
      "eval_steps_per_second": 211.175,
      "step": 45000
    },
    {
      "entropy": 5.637003469467163,
      "epoch": 5.054472960071882,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027508832717957433,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23933654874563218,
      "num_tokens": 100459548.0,
      "step": 45005
    },
    {
      "entropy": 5.786850690841675,
      "epoch": 5.055034536979839,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002750481784624553,
      "loss": 4.9046,
      "mean_token_accuracy": 0.23480217009782792,
      "num_tokens": 100468687.0,
      "step": 45010
    },
    {
      "entropy": 5.685141897201538,
      "epoch": 5.0555961138877965,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027500802974380224,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2329969063401222,
      "num_tokens": 100479607.0,
      "step": 45015
    },
    {
      "entropy": 5.709640407562256,
      "epoch": 5.056157690795755,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002749678810248935,
      "loss": 4.9676,
      "mean_token_accuracy": 0.23641930073499678,
      "num_tokens": 100492245.0,
      "step": 45020
    },
    {
      "entropy": 5.6977269649505615,
      "epoch": 5.056719267703712,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002749277323070075,
      "loss": 4.827,
      "mean_token_accuracy": 0.24091786444187163,
      "num_tokens": 100503973.0,
      "step": 45025
    },
    {
      "entropy": 5.7027050971984865,
      "epoch": 5.05728084461167,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027488758359142245,
      "loss": 4.8702,
      "mean_token_accuracy": 0.24007397294044494,
      "num_tokens": 100516003.0,
      "step": 45030
    },
    {
      "entropy": 5.72406325340271,
      "epoch": 5.057842421519627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002748474348794167,
      "loss": 4.9778,
      "mean_token_accuracy": 0.23266079276800156,
      "num_tokens": 100528130.0,
      "step": 45035
    },
    {
      "entropy": 5.660383796691894,
      "epoch": 5.058403998427584,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027480728617226884,
      "loss": 4.8782,
      "mean_token_accuracy": 0.24113683849573136,
      "num_tokens": 100539532.0,
      "step": 45040
    },
    {
      "entropy": 5.708705282211303,
      "epoch": 5.0589655753355425,
      "grad_norm": 1.234375,
      "learning_rate": 0.000274767137471257,
      "loss": 4.9823,
      "mean_token_accuracy": 0.22914164066314696,
      "num_tokens": 100550520.0,
      "step": 45045
    },
    {
      "entropy": 5.676995420455933,
      "epoch": 5.0595271522435,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027472698877765967,
      "loss": 4.7796,
      "mean_token_accuracy": 0.24327741414308549,
      "num_tokens": 100561596.0,
      "step": 45050
    },
    {
      "entropy": 5.693200778961182,
      "epoch": 5.060088729151457,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002746868400927551,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23756448477506636,
      "num_tokens": 100571238.0,
      "step": 45055
    },
    {
      "entropy": 5.726655149459839,
      "epoch": 5.060650306059415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002746466914178216,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22751846313476562,
      "num_tokens": 100582687.0,
      "step": 45060
    },
    {
      "entropy": 5.700425529479981,
      "epoch": 5.061211882967372,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002746065427541376,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23009353280067443,
      "num_tokens": 100594270.0,
      "step": 45065
    },
    {
      "entropy": 5.6968724727630615,
      "epoch": 5.06177345987533,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027456639410298144,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2298058032989502,
      "num_tokens": 100605449.0,
      "step": 45070
    },
    {
      "entropy": 5.62893853187561,
      "epoch": 5.062335036783288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002745262454656314,
      "loss": 4.7922,
      "mean_token_accuracy": 0.24465217888355256,
      "num_tokens": 100616985.0,
      "step": 45075
    },
    {
      "entropy": 5.670635461807251,
      "epoch": 5.062896613691245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000274486096843366,
      "loss": 4.87,
      "mean_token_accuracy": 0.23842044919729233,
      "num_tokens": 100628088.0,
      "step": 45080
    },
    {
      "entropy": 5.792975044250488,
      "epoch": 5.063458190599203,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027444594823746344,
      "loss": 4.9987,
      "mean_token_accuracy": 0.2349924311041832,
      "num_tokens": 100639805.0,
      "step": 45085
    },
    {
      "entropy": 5.842933940887451,
      "epoch": 5.06401976750716,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002744057996492021,
      "loss": 5.0798,
      "mean_token_accuracy": 0.2199348494410515,
      "num_tokens": 100650437.0,
      "step": 45090
    },
    {
      "entropy": 5.664734792709351,
      "epoch": 5.064581344415117,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002743656510798603,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23660925030708313,
      "num_tokens": 100661466.0,
      "step": 45095
    },
    {
      "entropy": 5.735499620437622,
      "epoch": 5.065142921323075,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002743255025307165,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23966730684041976,
      "num_tokens": 100671706.0,
      "step": 45100
    },
    {
      "entropy": 5.683469676971436,
      "epoch": 5.065704498231033,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027428535400304886,
      "loss": 4.868,
      "mean_token_accuracy": 0.23457978218793868,
      "num_tokens": 100682592.0,
      "step": 45105
    },
    {
      "entropy": 5.70570592880249,
      "epoch": 5.06626607513899,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002742452054981358,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23575789332389832,
      "num_tokens": 100693893.0,
      "step": 45110
    },
    {
      "entropy": 5.611766386032104,
      "epoch": 5.066827652046948,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002742050570172558,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23776651918888092,
      "num_tokens": 100705455.0,
      "step": 45115
    },
    {
      "entropy": 5.693363189697266,
      "epoch": 5.067389228954905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000274164908561687,
      "loss": 4.8248,
      "mean_token_accuracy": 0.2341512680053711,
      "num_tokens": 100714945.0,
      "step": 45120
    },
    {
      "entropy": 5.73571548461914,
      "epoch": 5.067950805862863,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027412476013270786,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22703395783901215,
      "num_tokens": 100726299.0,
      "step": 45125
    },
    {
      "entropy": 5.769162702560425,
      "epoch": 5.0685123827708205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027408461173159676,
      "loss": 5.0388,
      "mean_token_accuracy": 0.22595582604408265,
      "num_tokens": 100739415.0,
      "step": 45130
    },
    {
      "entropy": 5.784015417098999,
      "epoch": 5.069073959678778,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027404446335963195,
      "loss": 5.0328,
      "mean_token_accuracy": 0.22637763768434524,
      "num_tokens": 100751018.0,
      "step": 45135
    },
    {
      "entropy": 5.656726741790772,
      "epoch": 5.069635536586736,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002740043150180918,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23723266869783402,
      "num_tokens": 100761425.0,
      "step": 45140
    },
    {
      "entropy": 5.732597351074219,
      "epoch": 5.070197113494693,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002739641667082546,
      "loss": 4.924,
      "mean_token_accuracy": 0.23301879167556763,
      "num_tokens": 100772906.0,
      "step": 45145
    },
    {
      "entropy": 5.721082973480224,
      "epoch": 5.07075869040265,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027392401843139876,
      "loss": 4.96,
      "mean_token_accuracy": 0.23427198678255082,
      "num_tokens": 100784106.0,
      "step": 45150
    },
    {
      "entropy": 5.726725387573242,
      "epoch": 5.071320267310608,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002738838701888026,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2245401069521904,
      "num_tokens": 100795066.0,
      "step": 45155
    },
    {
      "entropy": 5.664318418502807,
      "epoch": 5.071881844218566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002738437219817444,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2374735713005066,
      "num_tokens": 100806936.0,
      "step": 45160
    },
    {
      "entropy": 5.572559976577759,
      "epoch": 5.072443421126524,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027380357381150263,
      "loss": 4.7725,
      "mean_token_accuracy": 0.23574666380882264,
      "num_tokens": 100817483.0,
      "step": 45165
    },
    {
      "entropy": 5.776484727859497,
      "epoch": 5.073004998034481,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002737634256793555,
      "loss": 5.0132,
      "mean_token_accuracy": 0.22833300083875657,
      "num_tokens": 100828482.0,
      "step": 45170
    },
    {
      "entropy": 5.708509922027588,
      "epoch": 5.073566574942438,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027372327758658135,
      "loss": 4.942,
      "mean_token_accuracy": 0.23790284544229506,
      "num_tokens": 100839745.0,
      "step": 45175
    },
    {
      "entropy": 5.663076972961425,
      "epoch": 5.074128151850396,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027368312953445845,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2321598306298256,
      "num_tokens": 100850441.0,
      "step": 45180
    },
    {
      "entropy": 5.614814949035645,
      "epoch": 5.0746897287583534,
      "grad_norm": 1.25,
      "learning_rate": 0.0002736429815242653,
      "loss": 4.8602,
      "mean_token_accuracy": 0.23462419360876083,
      "num_tokens": 100861428.0,
      "step": 45185
    },
    {
      "entropy": 5.6725634098052975,
      "epoch": 5.075251305666311,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002736028335572802,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2369949236512184,
      "num_tokens": 100872874.0,
      "step": 45190
    },
    {
      "entropy": 5.7517955780029295,
      "epoch": 5.075812882574269,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027356268563478135,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21903525590896605,
      "num_tokens": 100884150.0,
      "step": 45195
    },
    {
      "entropy": 5.710903549194336,
      "epoch": 5.076374459482226,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002735225377580472,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23735448122024536,
      "num_tokens": 100896440.0,
      "step": 45200
    },
    {
      "entropy": 5.715531206130981,
      "epoch": 5.076936036390183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000273482389928356,
      "loss": 4.8887,
      "mean_token_accuracy": 0.24396452754735948,
      "num_tokens": 100906854.0,
      "step": 45205
    },
    {
      "entropy": 5.676136875152588,
      "epoch": 5.077497613298141,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002734422421469861,
      "loss": 4.8829,
      "mean_token_accuracy": 0.24282888174057007,
      "num_tokens": 100918644.0,
      "step": 45210
    },
    {
      "entropy": 5.70551118850708,
      "epoch": 5.0780591902060985,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002734020944152157,
      "loss": 4.8835,
      "mean_token_accuracy": 0.2341565579175949,
      "num_tokens": 100929313.0,
      "step": 45215
    },
    {
      "entropy": 5.748371553421021,
      "epoch": 5.078620767114057,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002733619467343234,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23360079526901245,
      "num_tokens": 100940406.0,
      "step": 45220
    },
    {
      "entropy": 5.6603920459747314,
      "epoch": 5.079182344022014,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002733217991055873,
      "loss": 4.8438,
      "mean_token_accuracy": 0.24195905923843383,
      "num_tokens": 100951086.0,
      "step": 45225
    },
    {
      "entropy": 5.6998138427734375,
      "epoch": 5.079743920929971,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002732816515302858,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2341106116771698,
      "num_tokens": 100961417.0,
      "step": 45230
    },
    {
      "entropy": 5.69844274520874,
      "epoch": 5.080305497837929,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002732415040096971,
      "loss": 4.9459,
      "mean_token_accuracy": 0.23727672398090363,
      "num_tokens": 100973667.0,
      "step": 45235
    },
    {
      "entropy": 5.749660444259644,
      "epoch": 5.080867074745886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027320135654509964,
      "loss": 4.9523,
      "mean_token_accuracy": 0.23205482065677643,
      "num_tokens": 100984891.0,
      "step": 45240
    },
    {
      "entropy": 5.703946495056153,
      "epoch": 5.081428651653844,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002731612091377717,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23267505019903184,
      "num_tokens": 100997273.0,
      "step": 45245
    },
    {
      "entropy": 5.5966408252716064,
      "epoch": 5.081990228561802,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027312106178899154,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2518613785505295,
      "num_tokens": 101007341.0,
      "step": 45250
    },
    {
      "entropy": 5.753765058517456,
      "epoch": 5.082551805469759,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002730809145000374,
      "loss": 4.9585,
      "mean_token_accuracy": 0.23450706601142884,
      "num_tokens": 101017471.0,
      "step": 45255
    },
    {
      "entropy": 5.675786542892456,
      "epoch": 5.083113382377717,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002730407672721878,
      "loss": 4.8317,
      "mean_token_accuracy": 0.24132624864578248,
      "num_tokens": 101028726.0,
      "step": 45260
    },
    {
      "entropy": 5.860834836959839,
      "epoch": 5.083674959285674,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000273000620106721,
      "loss": 5.0489,
      "mean_token_accuracy": 0.22146134823560715,
      "num_tokens": 101040824.0,
      "step": 45265
    },
    {
      "entropy": 5.725242614746094,
      "epoch": 5.0842365361936315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002729604730049151,
      "loss": 4.939,
      "mean_token_accuracy": 0.23325855880975724,
      "num_tokens": 101051849.0,
      "step": 45270
    },
    {
      "entropy": 5.670284748077393,
      "epoch": 5.08479811310159,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002729203259680486,
      "loss": 4.9262,
      "mean_token_accuracy": 0.23263280242681503,
      "num_tokens": 101062667.0,
      "step": 45275
    },
    {
      "entropy": 5.68509430885315,
      "epoch": 5.085359690009547,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002728801789973998,
      "loss": 4.8807,
      "mean_token_accuracy": 0.23585725724697112,
      "num_tokens": 101073679.0,
      "step": 45280
    },
    {
      "entropy": 5.821037244796753,
      "epoch": 5.085921266917504,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002728400320942468,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22826295346021652,
      "num_tokens": 101085394.0,
      "step": 45285
    },
    {
      "entropy": 5.8216582298278805,
      "epoch": 5.086482843825462,
      "grad_norm": 1.234375,
      "learning_rate": 0.000272799885259868,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22722966223955154,
      "num_tokens": 101096748.0,
      "step": 45290
    },
    {
      "entropy": 5.633794116973877,
      "epoch": 5.087044420733419,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002727597384955418,
      "loss": 4.732,
      "mean_token_accuracy": 0.2501805365085602,
      "num_tokens": 101107704.0,
      "step": 45295
    },
    {
      "entropy": 5.676737689971924,
      "epoch": 5.0876059976413766,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027271959180254636,
      "loss": 4.8739,
      "mean_token_accuracy": 0.23864730149507524,
      "num_tokens": 101119890.0,
      "step": 45300
    },
    {
      "entropy": 5.679659509658814,
      "epoch": 5.088167574549335,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027267944518215997,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2323233738541603,
      "num_tokens": 101130172.0,
      "step": 45305
    },
    {
      "entropy": 5.682968282699585,
      "epoch": 5.088729151457292,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027263929863566104,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2448131650686264,
      "num_tokens": 101141374.0,
      "step": 45310
    },
    {
      "entropy": 5.726422452926636,
      "epoch": 5.08929072836525,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002725991521643277,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2265941545367241,
      "num_tokens": 101155192.0,
      "step": 45315
    },
    {
      "entropy": 5.765216112136841,
      "epoch": 5.089852305273207,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002725590057694383,
      "loss": 4.9448,
      "mean_token_accuracy": 0.23140295892953872,
      "num_tokens": 101168438.0,
      "step": 45320
    },
    {
      "entropy": 5.756173515319825,
      "epoch": 5.090413882181164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027251885945227093,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23213860690593718,
      "num_tokens": 101180609.0,
      "step": 45325
    },
    {
      "entropy": 5.740698671340942,
      "epoch": 5.0909754590891225,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027247871321410425,
      "loss": 4.871,
      "mean_token_accuracy": 0.24205769896507262,
      "num_tokens": 101191211.0,
      "step": 45330
    },
    {
      "entropy": 5.664555597305298,
      "epoch": 5.09153703599708,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027243856705621623,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23139839768409728,
      "num_tokens": 101202706.0,
      "step": 45335
    },
    {
      "entropy": 5.668880558013916,
      "epoch": 5.092098612905037,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027239842097988526,
      "loss": 4.9105,
      "mean_token_accuracy": 0.23272136151790618,
      "num_tokens": 101213890.0,
      "step": 45340
    },
    {
      "entropy": 5.794045305252075,
      "epoch": 5.092660189812995,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002723582749863896,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2269184872508049,
      "num_tokens": 101225946.0,
      "step": 45345
    },
    {
      "entropy": 5.789025020599365,
      "epoch": 5.093221766720952,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027231812907700746,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22952473908662796,
      "num_tokens": 101236269.0,
      "step": 45350
    },
    {
      "entropy": 5.737592363357544,
      "epoch": 5.09378334362891,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002722779832530171,
      "loss": 4.9226,
      "mean_token_accuracy": 0.23984200209379197,
      "num_tokens": 101247641.0,
      "step": 45355
    },
    {
      "entropy": 5.675487852096557,
      "epoch": 5.094344920536868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027223783751569687,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23153618723154068,
      "num_tokens": 101259271.0,
      "step": 45360
    },
    {
      "entropy": 5.608890533447266,
      "epoch": 5.094906497444825,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027219769186632493,
      "loss": 4.7944,
      "mean_token_accuracy": 0.24326706528663636,
      "num_tokens": 101269709.0,
      "step": 45365
    },
    {
      "entropy": 5.725886249542237,
      "epoch": 5.095468074352783,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002721575463061796,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22448640167713166,
      "num_tokens": 101280559.0,
      "step": 45370
    },
    {
      "entropy": 5.769709539413452,
      "epoch": 5.09602965126074,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027211740083653903,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23534354865550994,
      "num_tokens": 101292356.0,
      "step": 45375
    },
    {
      "entropy": 5.7255411624908445,
      "epoch": 5.096591228168697,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002720772554586816,
      "loss": 4.8857,
      "mean_token_accuracy": 0.23810740560293198,
      "num_tokens": 101303881.0,
      "step": 45380
    },
    {
      "entropy": 5.6644384384155275,
      "epoch": 5.0971528050766555,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002720371101738855,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2417706146836281,
      "num_tokens": 101315310.0,
      "step": 45385
    },
    {
      "entropy": 5.60558762550354,
      "epoch": 5.097714381984613,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027199696498342893,
      "loss": 4.8524,
      "mean_token_accuracy": 0.2379390224814415,
      "num_tokens": 101327085.0,
      "step": 45390
    },
    {
      "entropy": 5.654493570327759,
      "epoch": 5.09827595889257,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002719568198885901,
      "loss": 4.8701,
      "mean_token_accuracy": 0.23879973143339156,
      "num_tokens": 101340473.0,
      "step": 45395
    },
    {
      "entropy": 5.7238305568695065,
      "epoch": 5.098837535800528,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002719166748906473,
      "loss": 4.8842,
      "mean_token_accuracy": 0.24334581643342973,
      "num_tokens": 101353434.0,
      "step": 45400
    },
    {
      "entropy": 5.653680515289307,
      "epoch": 5.099399112708485,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027187652999087887,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2317749708890915,
      "num_tokens": 101363782.0,
      "step": 45405
    },
    {
      "entropy": 5.705436563491821,
      "epoch": 5.099960689616443,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027183638519056284,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22947713881731033,
      "num_tokens": 101374985.0,
      "step": 45410
    },
    {
      "entropy": 5.682896900177002,
      "epoch": 5.1005222665244005,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002717962404909776,
      "loss": 4.8391,
      "mean_token_accuracy": 0.24047937840223313,
      "num_tokens": 101386916.0,
      "step": 45415
    },
    {
      "entropy": 5.649651432037354,
      "epoch": 5.101083843432358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027175609589340135,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23437010645866393,
      "num_tokens": 101397656.0,
      "step": 45420
    },
    {
      "entropy": 5.637384653091431,
      "epoch": 5.101645420340316,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002717159513991122,
      "loss": 4.8431,
      "mean_token_accuracy": 0.2426463097333908,
      "num_tokens": 101408725.0,
      "step": 45425
    },
    {
      "entropy": 5.581564903259277,
      "epoch": 5.102206997248273,
      "grad_norm": 1.25,
      "learning_rate": 0.00027167580700938844,
      "loss": 4.7912,
      "mean_token_accuracy": 0.2449488580226898,
      "num_tokens": 101419580.0,
      "step": 45430
    },
    {
      "entropy": 5.721873664855957,
      "epoch": 5.10276857415623,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027163566272550826,
      "loss": 4.9067,
      "mean_token_accuracy": 0.23023221045732498,
      "num_tokens": 101430735.0,
      "step": 45435
    },
    {
      "entropy": 5.695324087142945,
      "epoch": 5.103330151064188,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027159551854874986,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23703051507472991,
      "num_tokens": 101440544.0,
      "step": 45440
    },
    {
      "entropy": 5.732795476913452,
      "epoch": 5.103891727972146,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002715553744803915,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22934295684099198,
      "num_tokens": 101450793.0,
      "step": 45445
    },
    {
      "entropy": 5.698628330230713,
      "epoch": 5.104453304880104,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002715152305217114,
      "loss": 5.0321,
      "mean_token_accuracy": 0.2270260974764824,
      "num_tokens": 101463593.0,
      "step": 45450
    },
    {
      "entropy": 5.67398772239685,
      "epoch": 5.105014881788061,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002714750866739877,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2304928183555603,
      "num_tokens": 101474334.0,
      "step": 45455
    },
    {
      "entropy": 5.744370698928833,
      "epoch": 5.105576458696018,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002714349429384986,
      "loss": 4.9475,
      "mean_token_accuracy": 0.23231513798236847,
      "num_tokens": 101485356.0,
      "step": 45460
    },
    {
      "entropy": 5.683604288101196,
      "epoch": 5.106138035603976,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027139479931652226,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2354555234313011,
      "num_tokens": 101495939.0,
      "step": 45465
    },
    {
      "entropy": 5.684259700775146,
      "epoch": 5.1066996125119335,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002713546558093368,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23192273676395417,
      "num_tokens": 101505563.0,
      "step": 45470
    },
    {
      "entropy": 5.589453220367432,
      "epoch": 5.107261189419891,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002713145124182207,
      "loss": 4.7818,
      "mean_token_accuracy": 0.24834335893392562,
      "num_tokens": 101516829.0,
      "step": 45475
    },
    {
      "entropy": 5.706179475784301,
      "epoch": 5.107822766327849,
      "grad_norm": 1.25,
      "learning_rate": 0.00027127436914445194,
      "loss": 4.9227,
      "mean_token_accuracy": 0.23024535477161406,
      "num_tokens": 101527427.0,
      "step": 45480
    },
    {
      "entropy": 5.708833932876587,
      "epoch": 5.108384343235806,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002712342259893086,
      "loss": 4.8719,
      "mean_token_accuracy": 0.24223242104053497,
      "num_tokens": 101537991.0,
      "step": 45485
    },
    {
      "entropy": 5.692278814315796,
      "epoch": 5.108945920143763,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027119408295406906,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22476522177457808,
      "num_tokens": 101549485.0,
      "step": 45490
    },
    {
      "entropy": 5.718194103240966,
      "epoch": 5.109507497051721,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002711539400400114,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2295119881629944,
      "num_tokens": 101560719.0,
      "step": 45495
    },
    {
      "entropy": 5.730129718780518,
      "epoch": 5.110069073959679,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027111379724841373,
      "loss": 5.005,
      "mean_token_accuracy": 0.22577994614839553,
      "num_tokens": 101572118.0,
      "step": 45500
    },
    {
      "entropy": 5.794856214523316,
      "epoch": 5.110630650867637,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027107365458055424,
      "loss": 4.9476,
      "mean_token_accuracy": 0.23801043182611464,
      "num_tokens": 101582050.0,
      "step": 45505
    },
    {
      "entropy": 5.669669818878174,
      "epoch": 5.111192227775594,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002710335120377112,
      "loss": 4.7596,
      "mean_token_accuracy": 0.2416050374507904,
      "num_tokens": 101592529.0,
      "step": 45510
    },
    {
      "entropy": 5.71156005859375,
      "epoch": 5.111753804683551,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027099336962116253,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23147226125001907,
      "num_tokens": 101603106.0,
      "step": 45515
    },
    {
      "entropy": 5.7719056606292725,
      "epoch": 5.112315381591509,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027095322733218665,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22916573882102967,
      "num_tokens": 101614171.0,
      "step": 45520
    },
    {
      "entropy": 5.776117467880249,
      "epoch": 5.112876958499466,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002709130851720616,
      "loss": 4.9887,
      "mean_token_accuracy": 0.23071971535682678,
      "num_tokens": 101626311.0,
      "step": 45525
    },
    {
      "entropy": 5.604635381698609,
      "epoch": 5.113438535407424,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027087294314206544,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23938422352075578,
      "num_tokens": 101637923.0,
      "step": 45530
    },
    {
      "entropy": 5.759845924377442,
      "epoch": 5.114000112315382,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002708328012434764,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22049022167921067,
      "num_tokens": 101647943.0,
      "step": 45535
    },
    {
      "entropy": 5.728460645675659,
      "epoch": 5.114561689223339,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002707926594775725,
      "loss": 4.9253,
      "mean_token_accuracy": 0.2323511943221092,
      "num_tokens": 101660095.0,
      "step": 45540
    },
    {
      "entropy": 5.65558648109436,
      "epoch": 5.115123266131297,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027075251784563196,
      "loss": 4.8495,
      "mean_token_accuracy": 0.24050552695989608,
      "num_tokens": 101670092.0,
      "step": 45545
    },
    {
      "entropy": 5.6856849670410154,
      "epoch": 5.115684843039254,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000270712376348933,
      "loss": 4.9348,
      "mean_token_accuracy": 0.23272418826818467,
      "num_tokens": 101681692.0,
      "step": 45550
    },
    {
      "entropy": 5.70384407043457,
      "epoch": 5.1162464199472115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002706722349887536,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23301399499177933,
      "num_tokens": 101692569.0,
      "step": 45555
    },
    {
      "entropy": 5.811294794082642,
      "epoch": 5.11680799685517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002706320937663719,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2328350767493248,
      "num_tokens": 101704185.0,
      "step": 45560
    },
    {
      "entropy": 5.747335767745971,
      "epoch": 5.117369573763127,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002705919526830661,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2318730518221855,
      "num_tokens": 101715500.0,
      "step": 45565
    },
    {
      "entropy": 5.745405864715576,
      "epoch": 5.117931150671084,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027055181174011415,
      "loss": 4.9822,
      "mean_token_accuracy": 0.2370037242770195,
      "num_tokens": 101727562.0,
      "step": 45570
    },
    {
      "entropy": 5.721790075302124,
      "epoch": 5.118492727579042,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00027051167093879424,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23963865339756013,
      "num_tokens": 101739269.0,
      "step": 45575
    },
    {
      "entropy": 5.716675472259522,
      "epoch": 5.119054304486999,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002704715302803846,
      "loss": 4.9608,
      "mean_token_accuracy": 0.23363710343837737,
      "num_tokens": 101750189.0,
      "step": 45580
    },
    {
      "entropy": 5.731708335876465,
      "epoch": 5.119615881394957,
      "grad_norm": 1.25,
      "learning_rate": 0.00027043138976616304,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22929249703884125,
      "num_tokens": 101760824.0,
      "step": 45585
    },
    {
      "entropy": 5.646698808670044,
      "epoch": 5.120177458302915,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027039124939740787,
      "loss": 4.7852,
      "mean_token_accuracy": 0.24643949568271636,
      "num_tokens": 101772647.0,
      "step": 45590
    },
    {
      "entropy": 5.718174600601197,
      "epoch": 5.120739035210872,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027035110917539713,
      "loss": 4.9754,
      "mean_token_accuracy": 0.23367606550455094,
      "num_tokens": 101783020.0,
      "step": 45595
    },
    {
      "entropy": 5.737560272216797,
      "epoch": 5.12130061211883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027031096910140877,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2396194666624069,
      "num_tokens": 101793372.0,
      "step": 45600
    },
    {
      "entropy": 5.684696960449219,
      "epoch": 5.121862189026787,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027027082917672107,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23709076046943664,
      "num_tokens": 101805077.0,
      "step": 45605
    },
    {
      "entropy": 5.6864604473114015,
      "epoch": 5.1224237659347445,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027023068940261203,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23662899434566498,
      "num_tokens": 101816334.0,
      "step": 45610
    },
    {
      "entropy": 5.7692468643188475,
      "epoch": 5.122985342842703,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002701905497803595,
      "loss": 5.052,
      "mean_token_accuracy": 0.22658927142620086,
      "num_tokens": 101827812.0,
      "step": 45615
    },
    {
      "entropy": 5.663055801391602,
      "epoch": 5.12354691975066,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002701504103112419,
      "loss": 4.7779,
      "mean_token_accuracy": 0.2449037104845047,
      "num_tokens": 101839400.0,
      "step": 45620
    },
    {
      "entropy": 5.7624317646026615,
      "epoch": 5.124108496658617,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027011027099653706,
      "loss": 4.9788,
      "mean_token_accuracy": 0.23947205990552903,
      "num_tokens": 101851948.0,
      "step": 45625
    },
    {
      "entropy": 5.684853744506836,
      "epoch": 5.124670073566575,
      "grad_norm": 1.25,
      "learning_rate": 0.0002700701318375232,
      "loss": 4.9681,
      "mean_token_accuracy": 0.23092929422855377,
      "num_tokens": 101862985.0,
      "step": 45630
    },
    {
      "entropy": 5.6761070728302006,
      "epoch": 5.125231650474532,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002700299928354781,
      "loss": 4.8694,
      "mean_token_accuracy": 0.23522304147481918,
      "num_tokens": 101873247.0,
      "step": 45635
    },
    {
      "entropy": 5.738183641433716,
      "epoch": 5.12579322738249,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026998985399168005,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2333667457103729,
      "num_tokens": 101883686.0,
      "step": 45640
    },
    {
      "entropy": 5.775682687759399,
      "epoch": 5.126354804290448,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026994971530740696,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2319033369421959,
      "num_tokens": 101894905.0,
      "step": 45645
    },
    {
      "entropy": 5.6823021411895756,
      "epoch": 5.126916381198405,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026990957678393684,
      "loss": 4.9116,
      "mean_token_accuracy": 0.23382880091667174,
      "num_tokens": 101906671.0,
      "step": 45650
    },
    {
      "entropy": 5.600595569610595,
      "epoch": 5.127477958106363,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026986943842254783,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23507247418165206,
      "num_tokens": 101917440.0,
      "step": 45655
    },
    {
      "entropy": 5.710738897323608,
      "epoch": 5.12803953501432,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002698293002245179,
      "loss": 4.931,
      "mean_token_accuracy": 0.23318861722946166,
      "num_tokens": 101929177.0,
      "step": 45660
    },
    {
      "entropy": 5.769397163391114,
      "epoch": 5.128601111922277,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026978916219112505,
      "loss": 4.9605,
      "mean_token_accuracy": 0.23295588791370392,
      "num_tokens": 101940034.0,
      "step": 45665
    },
    {
      "entropy": 5.815218305587768,
      "epoch": 5.1291626888302355,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026974902432364727,
      "loss": 4.995,
      "mean_token_accuracy": 0.2264265835285187,
      "num_tokens": 101952108.0,
      "step": 45670
    },
    {
      "entropy": 5.759220695495605,
      "epoch": 5.129724265738193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026970888662336263,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2320260301232338,
      "num_tokens": 101963617.0,
      "step": 45675
    },
    {
      "entropy": 5.729390668869018,
      "epoch": 5.130285842646151,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002696687490915491,
      "loss": 4.9083,
      "mean_token_accuracy": 0.24325990527868271,
      "num_tokens": 101975565.0,
      "step": 45680
    },
    {
      "entropy": 5.655146169662475,
      "epoch": 5.130847419554108,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026962861172948456,
      "loss": 4.9224,
      "mean_token_accuracy": 0.23775609135627745,
      "num_tokens": 101988048.0,
      "step": 45685
    },
    {
      "entropy": 5.6547283172607425,
      "epoch": 5.131408996462065,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026958847453844713,
      "loss": 4.8664,
      "mean_token_accuracy": 0.2340142160654068,
      "num_tokens": 101999045.0,
      "step": 45690
    },
    {
      "entropy": 5.750152635574341,
      "epoch": 5.131970573370023,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026954833751971486,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2264761060476303,
      "num_tokens": 102009679.0,
      "step": 45695
    },
    {
      "entropy": 5.682682847976684,
      "epoch": 5.132532150277981,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002695082006745656,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22670343965291978,
      "num_tokens": 102021378.0,
      "step": 45700
    },
    {
      "entropy": 5.722785758972168,
      "epoch": 5.133093727185938,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002694680640042774,
      "loss": 4.9232,
      "mean_token_accuracy": 0.2380713328719139,
      "num_tokens": 102031037.0,
      "step": 45705
    },
    {
      "entropy": 5.70200743675232,
      "epoch": 5.133655304093896,
      "grad_norm": 1.25,
      "learning_rate": 0.00026942792751012814,
      "loss": 4.8643,
      "mean_token_accuracy": 0.2394473910331726,
      "num_tokens": 102041767.0,
      "step": 45710
    },
    {
      "entropy": 5.6961095333099365,
      "epoch": 5.134216881001853,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026938779119339585,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23605036735534668,
      "num_tokens": 102052792.0,
      "step": 45715
    },
    {
      "entropy": 5.7038328647613525,
      "epoch": 5.13477845790981,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002693476550553584,
      "loss": 4.891,
      "mean_token_accuracy": 0.23962925672531127,
      "num_tokens": 102064238.0,
      "step": 45720
    },
    {
      "entropy": 5.720402956008911,
      "epoch": 5.1353400348177685,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026930751909729384,
      "loss": 4.9201,
      "mean_token_accuracy": 0.23530001193284988,
      "num_tokens": 102075557.0,
      "step": 45725
    },
    {
      "entropy": 5.663139963150025,
      "epoch": 5.135901611725726,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026926738332048013,
      "loss": 4.8164,
      "mean_token_accuracy": 0.24508301317691802,
      "num_tokens": 102086691.0,
      "step": 45730
    },
    {
      "entropy": 5.699800491333008,
      "epoch": 5.136463188633684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026922724772619516,
      "loss": 4.8643,
      "mean_token_accuracy": 0.23743362128734588,
      "num_tokens": 102097215.0,
      "step": 45735
    },
    {
      "entropy": 5.716428422927857,
      "epoch": 5.137024765541641,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002691871123157168,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2313874989748001,
      "num_tokens": 102108408.0,
      "step": 45740
    },
    {
      "entropy": 5.735626125335694,
      "epoch": 5.137586342449598,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002691469770903231,
      "loss": 4.9572,
      "mean_token_accuracy": 0.2394236370921135,
      "num_tokens": 102120594.0,
      "step": 45745
    },
    {
      "entropy": 5.712230443954468,
      "epoch": 5.138147919357556,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002691068420512919,
      "loss": 4.8861,
      "mean_token_accuracy": 0.23467095643281938,
      "num_tokens": 102132306.0,
      "step": 45750
    },
    {
      "entropy": 5.756574392318726,
      "epoch": 5.1387094962655135,
      "grad_norm": 1.265625,
      "learning_rate": 0.000269066707199901,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2354958564043045,
      "num_tokens": 102143271.0,
      "step": 45755
    },
    {
      "entropy": 5.7208836555480955,
      "epoch": 5.139271073173471,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026902657253742855,
      "loss": 4.9505,
      "mean_token_accuracy": 0.23689593523740768,
      "num_tokens": 102154657.0,
      "step": 45760
    },
    {
      "entropy": 5.689207744598389,
      "epoch": 5.139832650081429,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002689864380651524,
      "loss": 4.8219,
      "mean_token_accuracy": 0.23735203444957734,
      "num_tokens": 102167135.0,
      "step": 45765
    },
    {
      "entropy": 5.604485702514649,
      "epoch": 5.140394226989386,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002689463037843504,
      "loss": 4.8174,
      "mean_token_accuracy": 0.2404410019516945,
      "num_tokens": 102178485.0,
      "step": 45770
    },
    {
      "entropy": 5.685472726821899,
      "epoch": 5.140955803897343,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002689061696963004,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2314574122428894,
      "num_tokens": 102189378.0,
      "step": 45775
    },
    {
      "entropy": 5.686957597732544,
      "epoch": 5.141517380805301,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002688660358022803,
      "loss": 4.9387,
      "mean_token_accuracy": 0.23222472965717317,
      "num_tokens": 102199864.0,
      "step": 45780
    },
    {
      "entropy": 5.6347620487213135,
      "epoch": 5.142078957713259,
      "grad_norm": 1.328125,
      "learning_rate": 0.000268825902103568,
      "loss": 4.818,
      "mean_token_accuracy": 0.23487171828746795,
      "num_tokens": 102210699.0,
      "step": 45785
    },
    {
      "entropy": 5.72046046257019,
      "epoch": 5.142640534621217,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002687857686014414,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2351283922791481,
      "num_tokens": 102221724.0,
      "step": 45790
    },
    {
      "entropy": 5.642383670806884,
      "epoch": 5.143202111529174,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002687456352971783,
      "loss": 4.7195,
      "mean_token_accuracy": 0.2519736662507057,
      "num_tokens": 102232023.0,
      "step": 45795
    },
    {
      "entropy": 5.642626428604126,
      "epoch": 5.143763688437131,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002687055021920566,
      "loss": 4.857,
      "mean_token_accuracy": 0.2405129551887512,
      "num_tokens": 102242526.0,
      "step": 45800
    },
    {
      "entropy": 5.591801023483276,
      "epoch": 5.144325265345089,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002686653692873542,
      "loss": 4.9091,
      "mean_token_accuracy": 0.23347246944904326,
      "num_tokens": 102253271.0,
      "step": 45805
    },
    {
      "entropy": 5.7091240882873535,
      "epoch": 5.1448868422530465,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026862523658434883,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2328251674771309,
      "num_tokens": 102263089.0,
      "step": 45810
    },
    {
      "entropy": 5.669428205490112,
      "epoch": 5.145448419161004,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002685851040843185,
      "loss": 4.8896,
      "mean_token_accuracy": 0.23980252742767333,
      "num_tokens": 102273312.0,
      "step": 45815
    },
    {
      "entropy": 5.740741777420044,
      "epoch": 5.146009996068962,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026854497178854087,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2286195531487465,
      "num_tokens": 102283940.0,
      "step": 45820
    },
    {
      "entropy": 5.757841730117798,
      "epoch": 5.146571572976919,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002685048396982939,
      "loss": 4.9219,
      "mean_token_accuracy": 0.23151046633720399,
      "num_tokens": 102295666.0,
      "step": 45825
    },
    {
      "entropy": 5.675566625595093,
      "epoch": 5.147133149884877,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026846470781485517,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23268258273601533,
      "num_tokens": 102306463.0,
      "step": 45830
    },
    {
      "entropy": 5.646671962738037,
      "epoch": 5.147694726792834,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002684245761395028,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2406246244907379,
      "num_tokens": 102317443.0,
      "step": 45835
    },
    {
      "entropy": 5.693761110305786,
      "epoch": 5.148256303700792,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026838444467351447,
      "loss": 4.9212,
      "mean_token_accuracy": 0.23371184915304183,
      "num_tokens": 102327924.0,
      "step": 45840
    },
    {
      "entropy": 5.82807354927063,
      "epoch": 5.14881788060875,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026834431341816805,
      "loss": 5.0294,
      "mean_token_accuracy": 0.22676809132099152,
      "num_tokens": 102339364.0,
      "step": 45845
    },
    {
      "entropy": 5.7200172424316404,
      "epoch": 5.149379457516707,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026830418237474126,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2325504258275032,
      "num_tokens": 102350947.0,
      "step": 45850
    },
    {
      "entropy": 5.755914306640625,
      "epoch": 5.149941034424664,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002682640515445118,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2291715055704117,
      "num_tokens": 102362412.0,
      "step": 45855
    },
    {
      "entropy": 5.690887451171875,
      "epoch": 5.150502611332622,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002682239209287577,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23583494126796722,
      "num_tokens": 102373630.0,
      "step": 45860
    },
    {
      "entropy": 5.66808500289917,
      "epoch": 5.151064188240579,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026818379052875646,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22947173416614533,
      "num_tokens": 102385081.0,
      "step": 45865
    },
    {
      "entropy": 5.733764266967773,
      "epoch": 5.1516257651485375,
      "grad_norm": 1.15625,
      "learning_rate": 0.000268143660345786,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23574648201465606,
      "num_tokens": 102396519.0,
      "step": 45870
    },
    {
      "entropy": 5.753663158416748,
      "epoch": 5.152187342056495,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002681035303811241,
      "loss": 4.959,
      "mean_token_accuracy": 0.23284192085266114,
      "num_tokens": 102406326.0,
      "step": 45875
    },
    {
      "entropy": 5.793155717849731,
      "epoch": 5.152748918964452,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026806340063604845,
      "loss": 4.9877,
      "mean_token_accuracy": 0.22742173671722413,
      "num_tokens": 102416965.0,
      "step": 45880
    },
    {
      "entropy": 5.694823694229126,
      "epoch": 5.15331049587241,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026802327111183686,
      "loss": 4.852,
      "mean_token_accuracy": 0.23374133259058,
      "num_tokens": 102428414.0,
      "step": 45885
    },
    {
      "entropy": 5.689474725723267,
      "epoch": 5.153872072780367,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026798314180976696,
      "loss": 4.8462,
      "mean_token_accuracy": 0.24234450459480286,
      "num_tokens": 102438665.0,
      "step": 45890
    },
    {
      "entropy": 5.730973911285401,
      "epoch": 5.1544336496883245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026794301273111654,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2274249643087387,
      "num_tokens": 102449828.0,
      "step": 45895
    },
    {
      "entropy": 5.722751188278198,
      "epoch": 5.154995226596283,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002679028838771633,
      "loss": 4.9493,
      "mean_token_accuracy": 0.23470087945461274,
      "num_tokens": 102461007.0,
      "step": 45900
    },
    {
      "entropy": 5.658668661117554,
      "epoch": 5.15555680350424,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026786275524918503,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2421344056725502,
      "num_tokens": 102472372.0,
      "step": 45905
    },
    {
      "entropy": 5.66387186050415,
      "epoch": 5.156118380412197,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002678226268484594,
      "loss": 4.8083,
      "mean_token_accuracy": 0.23626537919044494,
      "num_tokens": 102482361.0,
      "step": 45910
    },
    {
      "entropy": 5.571767091751099,
      "epoch": 5.156679957320155,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002677824986762641,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23664769381284714,
      "num_tokens": 102493058.0,
      "step": 45915
    },
    {
      "entropy": 5.771416664123535,
      "epoch": 5.157241534228112,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002677423707338769,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2348492905497551,
      "num_tokens": 102505242.0,
      "step": 45920
    },
    {
      "entropy": 5.738821172714234,
      "epoch": 5.1578031111360705,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002677022430225754,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2421805739402771,
      "num_tokens": 102517580.0,
      "step": 45925
    },
    {
      "entropy": 5.761465835571289,
      "epoch": 5.158364688044028,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026766211554363727,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2370716467499733,
      "num_tokens": 102527336.0,
      "step": 45930
    },
    {
      "entropy": 5.707497835159302,
      "epoch": 5.158926264951985,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002676219882983403,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2259282350540161,
      "num_tokens": 102538511.0,
      "step": 45935
    },
    {
      "entropy": 5.746284151077271,
      "epoch": 5.159487841859943,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026758186128796204,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2371939867734909,
      "num_tokens": 102549981.0,
      "step": 45940
    },
    {
      "entropy": 5.814321517944336,
      "epoch": 5.1600494187679,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026754173451378014,
      "loss": 4.9969,
      "mean_token_accuracy": 0.23362721204757692,
      "num_tokens": 102560504.0,
      "step": 45945
    },
    {
      "entropy": 5.752322340011597,
      "epoch": 5.1606109956758575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026750160797707245,
      "loss": 4.9473,
      "mean_token_accuracy": 0.23378413766622544,
      "num_tokens": 102571136.0,
      "step": 45950
    },
    {
      "entropy": 5.709217977523804,
      "epoch": 5.161172572583816,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002674614816791164,
      "loss": 4.8646,
      "mean_token_accuracy": 0.24356432110071183,
      "num_tokens": 102582879.0,
      "step": 45955
    },
    {
      "entropy": 5.75213589668274,
      "epoch": 5.161734149491773,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026742135562118967,
      "loss": 4.9357,
      "mean_token_accuracy": 0.23615251183509828,
      "num_tokens": 102593171.0,
      "step": 45960
    },
    {
      "entropy": 5.786683416366577,
      "epoch": 5.16229572639973,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002673812298045699,
      "loss": 5.0144,
      "mean_token_accuracy": 0.22862644940614701,
      "num_tokens": 102603816.0,
      "step": 45965
    },
    {
      "entropy": 5.717014169692993,
      "epoch": 5.162857303307688,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026734110423053483,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23882527351379396,
      "num_tokens": 102614360.0,
      "step": 45970
    },
    {
      "entropy": 5.77713794708252,
      "epoch": 5.163418880215645,
      "grad_norm": 1.25,
      "learning_rate": 0.00026730097890036175,
      "loss": 4.9659,
      "mean_token_accuracy": 0.22840667366981507,
      "num_tokens": 102625320.0,
      "step": 45975
    },
    {
      "entropy": 5.702233076095581,
      "epoch": 5.163980457123603,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026726085381532867,
      "loss": 4.909,
      "mean_token_accuracy": 0.23508263677358626,
      "num_tokens": 102635728.0,
      "step": 45980
    },
    {
      "entropy": 5.637184000015258,
      "epoch": 5.164542034031561,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000267220728976713,
      "loss": 4.798,
      "mean_token_accuracy": 0.2372418999671936,
      "num_tokens": 102645880.0,
      "step": 45985
    },
    {
      "entropy": 5.711064100265503,
      "epoch": 5.165103610939518,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002671806043857923,
      "loss": 4.9423,
      "mean_token_accuracy": 0.23317295610904692,
      "num_tokens": 102657854.0,
      "step": 45990
    },
    {
      "entropy": 5.76537675857544,
      "epoch": 5.165665187847476,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002671404800438442,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2365706220269203,
      "num_tokens": 102669637.0,
      "step": 45995
    },
    {
      "entropy": 5.719860315322876,
      "epoch": 5.166226764755433,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002671003559521463,
      "loss": 4.8469,
      "mean_token_accuracy": 0.24016601741313934,
      "num_tokens": 102680520.0,
      "step": 46000
    },
    {
      "entropy": 5.662162446975708,
      "epoch": 5.16678834166339,
      "grad_norm": 1.25,
      "learning_rate": 0.00026706023211197613,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22819326668977738,
      "num_tokens": 102691633.0,
      "step": 46005
    },
    {
      "entropy": 5.715544509887695,
      "epoch": 5.1673499185713485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002670201085246111,
      "loss": 4.95,
      "mean_token_accuracy": 0.2289068803191185,
      "num_tokens": 102703171.0,
      "step": 46010
    },
    {
      "entropy": 5.756628179550171,
      "epoch": 5.167911495479306,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026697998519132905,
      "loss": 4.9459,
      "mean_token_accuracy": 0.23570191413164138,
      "num_tokens": 102714636.0,
      "step": 46015
    },
    {
      "entropy": 5.722197771072388,
      "epoch": 5.168473072387264,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026693986211340736,
      "loss": 5.0616,
      "mean_token_accuracy": 0.2318487897515297,
      "num_tokens": 102725554.0,
      "step": 46020
    },
    {
      "entropy": 5.731254768371582,
      "epoch": 5.169034649295221,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026689973929212363,
      "loss": 4.9728,
      "mean_token_accuracy": 0.22987303435802459,
      "num_tokens": 102736705.0,
      "step": 46025
    },
    {
      "entropy": 5.798680114746094,
      "epoch": 5.169596226203178,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026685961672875524,
      "loss": 5.0217,
      "mean_token_accuracy": 0.22451857924461366,
      "num_tokens": 102749187.0,
      "step": 46030
    },
    {
      "entropy": 5.719456148147583,
      "epoch": 5.170157803111136,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002668194944245799,
      "loss": 4.8988,
      "mean_token_accuracy": 0.23000696897506714,
      "num_tokens": 102760095.0,
      "step": 46035
    },
    {
      "entropy": 5.711248397827148,
      "epoch": 5.170719380019094,
      "grad_norm": 1.296875,
      "learning_rate": 0.000266779372380875,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2286094009876251,
      "num_tokens": 102771086.0,
      "step": 46040
    },
    {
      "entropy": 5.74081597328186,
      "epoch": 5.171280956927051,
      "grad_norm": 1.28125,
      "learning_rate": 0.000266739250598918,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22614069581031798,
      "num_tokens": 102782024.0,
      "step": 46045
    },
    {
      "entropy": 5.810883665084839,
      "epoch": 5.171842533835009,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026669912907998657,
      "loss": 4.9792,
      "mean_token_accuracy": 0.23011634796857833,
      "num_tokens": 102793654.0,
      "step": 46050
    },
    {
      "entropy": 5.843155193328857,
      "epoch": 5.172404110742966,
      "grad_norm": 1.296875,
      "learning_rate": 0.000266659007825358,
      "loss": 5.0758,
      "mean_token_accuracy": 0.22579529285430908,
      "num_tokens": 102804908.0,
      "step": 46055
    },
    {
      "entropy": 5.62238130569458,
      "epoch": 5.172965687650924,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026661888683631,
      "loss": 4.8224,
      "mean_token_accuracy": 0.24064296185970308,
      "num_tokens": 102815399.0,
      "step": 46060
    },
    {
      "entropy": 5.647350740432739,
      "epoch": 5.1735272645588815,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026657876611411974,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23675123155117034,
      "num_tokens": 102825734.0,
      "step": 46065
    },
    {
      "entropy": 5.656853771209716,
      "epoch": 5.174088841466839,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026653864566006494,
      "loss": 4.7527,
      "mean_token_accuracy": 0.24145993143320083,
      "num_tokens": 102836251.0,
      "step": 46070
    },
    {
      "entropy": 5.734486103057861,
      "epoch": 5.174650418374797,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002664985254754229,
      "loss": 5.0174,
      "mean_token_accuracy": 0.22455836534500123,
      "num_tokens": 102847626.0,
      "step": 46075
    },
    {
      "entropy": 5.703305339813232,
      "epoch": 5.175211995282754,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026645840556147103,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2443126007914543,
      "num_tokens": 102859424.0,
      "step": 46080
    },
    {
      "entropy": 5.713838720321656,
      "epoch": 5.175773572190711,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002664182859194869,
      "loss": 4.88,
      "mean_token_accuracy": 0.23837772458791734,
      "num_tokens": 102870466.0,
      "step": 46085
    },
    {
      "entropy": 5.6931405544281,
      "epoch": 5.176335149098669,
      "grad_norm": 1.25,
      "learning_rate": 0.0002663781665507479,
      "loss": 4.8928,
      "mean_token_accuracy": 0.2313981145620346,
      "num_tokens": 102882181.0,
      "step": 46090
    },
    {
      "entropy": 5.689406204223633,
      "epoch": 5.1768967260066265,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026633804745653146,
      "loss": 4.9428,
      "mean_token_accuracy": 0.2338812232017517,
      "num_tokens": 102892562.0,
      "step": 46095
    },
    {
      "entropy": 5.643070793151855,
      "epoch": 5.177458302914584,
      "grad_norm": 1.25,
      "learning_rate": 0.00026629792863811484,
      "loss": 4.8234,
      "mean_token_accuracy": 0.2463391825556755,
      "num_tokens": 102904380.0,
      "step": 46100
    },
    {
      "entropy": 5.628071641921997,
      "epoch": 5.178019879822542,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002662578100967756,
      "loss": 4.873,
      "mean_token_accuracy": 0.22503396570682527,
      "num_tokens": 102916018.0,
      "step": 46105
    },
    {
      "entropy": 5.701686573028565,
      "epoch": 5.178581456730499,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002662176918337911,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2240095168352127,
      "num_tokens": 102927885.0,
      "step": 46110
    },
    {
      "entropy": 5.67809739112854,
      "epoch": 5.179143033638457,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002661775738504387,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23877330720424653,
      "num_tokens": 102938343.0,
      "step": 46115
    },
    {
      "entropy": 5.6662344455719,
      "epoch": 5.179704610546414,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026613745614799574,
      "loss": 4.904,
      "mean_token_accuracy": 0.2326272875070572,
      "num_tokens": 102949341.0,
      "step": 46120
    },
    {
      "entropy": 5.773061227798462,
      "epoch": 5.180266187454372,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002660973387277396,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2387877359986305,
      "num_tokens": 102959722.0,
      "step": 46125
    },
    {
      "entropy": 5.789424800872803,
      "epoch": 5.18082776436233,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002660572215909477,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2374662458896637,
      "num_tokens": 102971423.0,
      "step": 46130
    },
    {
      "entropy": 5.71287727355957,
      "epoch": 5.181389341270287,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002660171047388973,
      "loss": 4.8949,
      "mean_token_accuracy": 0.24188187569379807,
      "num_tokens": 102983218.0,
      "step": 46135
    },
    {
      "entropy": 5.724604511260987,
      "epoch": 5.181950918178244,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026597698817286576,
      "loss": 4.9332,
      "mean_token_accuracy": 0.23642298430204392,
      "num_tokens": 102995287.0,
      "step": 46140
    },
    {
      "entropy": 5.691929149627685,
      "epoch": 5.182512495086202,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026593687189413044,
      "loss": 4.9155,
      "mean_token_accuracy": 0.23756024688482286,
      "num_tokens": 103006578.0,
      "step": 46145
    },
    {
      "entropy": 5.763690328598022,
      "epoch": 5.1830740719941595,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026589675590396865,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22530430406332017,
      "num_tokens": 103017959.0,
      "step": 46150
    },
    {
      "entropy": 5.715850687026977,
      "epoch": 5.183635648902117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002658566402036576,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22872209399938584,
      "num_tokens": 103029931.0,
      "step": 46155
    },
    {
      "entropy": 5.704902648925781,
      "epoch": 5.184197225810075,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002658165247944747,
      "loss": 4.8456,
      "mean_token_accuracy": 0.24254077970981597,
      "num_tokens": 103041603.0,
      "step": 46160
    },
    {
      "entropy": 5.671788311004638,
      "epoch": 5.184758802718032,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026577640967769727,
      "loss": 4.8586,
      "mean_token_accuracy": 0.23471288681030272,
      "num_tokens": 103053353.0,
      "step": 46165
    },
    {
      "entropy": 5.642350673675537,
      "epoch": 5.18532037962599,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026573629485460257,
      "loss": 4.7871,
      "mean_token_accuracy": 0.24430087208747864,
      "num_tokens": 103064295.0,
      "step": 46170
    },
    {
      "entropy": 5.734515047073364,
      "epoch": 5.185881956533947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026569618032646773,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2418116569519043,
      "num_tokens": 103075582.0,
      "step": 46175
    },
    {
      "entropy": 5.694663572311401,
      "epoch": 5.186443533441905,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026565606609457016,
      "loss": 4.9129,
      "mean_token_accuracy": 0.23558801859617234,
      "num_tokens": 103087969.0,
      "step": 46180
    },
    {
      "entropy": 5.71925835609436,
      "epoch": 5.187005110349863,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002656159521601871,
      "loss": 4.9294,
      "mean_token_accuracy": 0.23250861316919327,
      "num_tokens": 103099505.0,
      "step": 46185
    },
    {
      "entropy": 5.790623664855957,
      "epoch": 5.18756668725782,
      "grad_norm": 1.25,
      "learning_rate": 0.0002655758385245956,
      "loss": 4.98,
      "mean_token_accuracy": 0.22739075422286986,
      "num_tokens": 103110706.0,
      "step": 46190
    },
    {
      "entropy": 5.70259108543396,
      "epoch": 5.188128264165777,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002655357251890732,
      "loss": 4.8258,
      "mean_token_accuracy": 0.2433549791574478,
      "num_tokens": 103121129.0,
      "step": 46195
    },
    {
      "entropy": 5.66606650352478,
      "epoch": 5.188689841073735,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026549561215489695,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2310041755437851,
      "num_tokens": 103130706.0,
      "step": 46200
    },
    {
      "entropy": 5.700936937332154,
      "epoch": 5.189251417981692,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002654554994233442,
      "loss": 4.962,
      "mean_token_accuracy": 0.22976129800081252,
      "num_tokens": 103140441.0,
      "step": 46205
    },
    {
      "entropy": 5.628775930404663,
      "epoch": 5.1898129948896505,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002654153869956919,
      "loss": 4.8355,
      "mean_token_accuracy": 0.2384761765599251,
      "num_tokens": 103151245.0,
      "step": 46210
    },
    {
      "entropy": 5.706151294708252,
      "epoch": 5.190374571797608,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002653752748732175,
      "loss": 4.9031,
      "mean_token_accuracy": 0.2359470784664154,
      "num_tokens": 103161985.0,
      "step": 46215
    },
    {
      "entropy": 5.71307258605957,
      "epoch": 5.190936148705565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000265335163057198,
      "loss": 4.926,
      "mean_token_accuracy": 0.23652025759220124,
      "num_tokens": 103174471.0,
      "step": 46220
    },
    {
      "entropy": 5.6712359428405765,
      "epoch": 5.191497725613523,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002652950515489107,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23818518817424775,
      "num_tokens": 103184853.0,
      "step": 46225
    },
    {
      "entropy": 5.7432074546813965,
      "epoch": 5.19205930252148,
      "grad_norm": 1.25,
      "learning_rate": 0.0002652549403496327,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2272544413805008,
      "num_tokens": 103195909.0,
      "step": 46230
    },
    {
      "entropy": 5.788003444671631,
      "epoch": 5.1926208794294375,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026521482946064123,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22812086194753647,
      "num_tokens": 103206557.0,
      "step": 46235
    },
    {
      "entropy": 5.687548017501831,
      "epoch": 5.193182456337396,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002651747188832134,
      "loss": 4.8402,
      "mean_token_accuracy": 0.239799103140831,
      "num_tokens": 103216887.0,
      "step": 46240
    },
    {
      "entropy": 5.688014698028565,
      "epoch": 5.193744033245353,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002651346086186262,
      "loss": 4.9405,
      "mean_token_accuracy": 0.23291046023368836,
      "num_tokens": 103228879.0,
      "step": 46245
    },
    {
      "entropy": 5.675361251831054,
      "epoch": 5.194305610153311,
      "grad_norm": 1.140625,
      "learning_rate": 0.000265094498668157,
      "loss": 4.8625,
      "mean_token_accuracy": 0.23333002179861068,
      "num_tokens": 103240564.0,
      "step": 46250
    },
    {
      "entropy": 5.683162832260132,
      "epoch": 5.194867187061268,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026505438903308285,
      "loss": 4.8896,
      "mean_token_accuracy": 0.2358222246170044,
      "num_tokens": 103251297.0,
      "step": 46255
    },
    {
      "entropy": 5.730426263809204,
      "epoch": 5.195428763969225,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026501427971468056,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2460133343935013,
      "num_tokens": 103262041.0,
      "step": 46260
    },
    {
      "entropy": 5.6866168022155765,
      "epoch": 5.1959903408771835,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002649741707142277,
      "loss": 4.8674,
      "mean_token_accuracy": 0.24237967878580094,
      "num_tokens": 103273024.0,
      "step": 46265
    },
    {
      "entropy": 5.68167142868042,
      "epoch": 5.196551917785141,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026493406203300106,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2292763665318489,
      "num_tokens": 103284511.0,
      "step": 46270
    },
    {
      "entropy": 5.731519174575806,
      "epoch": 5.197113494693098,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026489395367227785,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23507935106754302,
      "num_tokens": 103296787.0,
      "step": 46275
    },
    {
      "entropy": 5.7486762523651125,
      "epoch": 5.197675071601056,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000264853845633335,
      "loss": 4.862,
      "mean_token_accuracy": 0.23705264031887055,
      "num_tokens": 103308422.0,
      "step": 46280
    },
    {
      "entropy": 5.666660928726197,
      "epoch": 5.198236648509013,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002648137379174497,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23508990108966826,
      "num_tokens": 103319592.0,
      "step": 46285
    },
    {
      "entropy": 5.657373428344727,
      "epoch": 5.1987982254169705,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002647736305258989,
      "loss": 4.865,
      "mean_token_accuracy": 0.23680053949356078,
      "num_tokens": 103330702.0,
      "step": 46290
    },
    {
      "entropy": 5.623896217346191,
      "epoch": 5.199359802324929,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002647335234599597,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23566691130399703,
      "num_tokens": 103342339.0,
      "step": 46295
    },
    {
      "entropy": 5.647460079193115,
      "epoch": 5.199921379232886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026469341672090906,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23885817229747772,
      "num_tokens": 103353604.0,
      "step": 46300
    },
    {
      "entropy": 5.693013143539429,
      "epoch": 5.200482956140844,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026465331031002404,
      "loss": 4.8586,
      "mean_token_accuracy": 0.24005049467086792,
      "num_tokens": 103363713.0,
      "step": 46305
    },
    {
      "entropy": 5.819362783432007,
      "epoch": 5.201044533048801,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026461320422858166,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2247081518173218,
      "num_tokens": 103374966.0,
      "step": 46310
    },
    {
      "entropy": 5.73968768119812,
      "epoch": 5.201606109956758,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002645730984778589,
      "loss": 4.9133,
      "mean_token_accuracy": 0.23311241418123246,
      "num_tokens": 103387025.0,
      "step": 46315
    },
    {
      "entropy": 5.765611791610718,
      "epoch": 5.202167686864716,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002645329930591327,
      "loss": 4.9583,
      "mean_token_accuracy": 0.23159278482198714,
      "num_tokens": 103397947.0,
      "step": 46320
    },
    {
      "entropy": 5.695612955093384,
      "epoch": 5.202729263772674,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026449288797368006,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2330719769001007,
      "num_tokens": 103409874.0,
      "step": 46325
    },
    {
      "entropy": 5.660548734664917,
      "epoch": 5.203290840680631,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026445278322277796,
      "loss": 4.939,
      "mean_token_accuracy": 0.231119067966938,
      "num_tokens": 103423484.0,
      "step": 46330
    },
    {
      "entropy": 5.723313570022583,
      "epoch": 5.203852417588589,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002644126788077032,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23438322693109512,
      "num_tokens": 103434933.0,
      "step": 46335
    },
    {
      "entropy": 5.740327501296997,
      "epoch": 5.204413994496546,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026437257472973296,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22778406441211702,
      "num_tokens": 103447111.0,
      "step": 46340
    },
    {
      "entropy": 5.690903377532959,
      "epoch": 5.204975571404503,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026433247099014406,
      "loss": 4.8306,
      "mean_token_accuracy": 0.24367947578430177,
      "num_tokens": 103457627.0,
      "step": 46345
    },
    {
      "entropy": 5.681720352172851,
      "epoch": 5.2055371483124615,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002642923675902135,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23668881952762605,
      "num_tokens": 103468047.0,
      "step": 46350
    },
    {
      "entropy": 5.63530478477478,
      "epoch": 5.206098725220419,
      "grad_norm": 1.28125,
      "learning_rate": 0.000264252264531218,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2375578001141548,
      "num_tokens": 103479671.0,
      "step": 46355
    },
    {
      "entropy": 5.659372091293335,
      "epoch": 5.206660302128377,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026421216181443457,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23085007518529893,
      "num_tokens": 103490565.0,
      "step": 46360
    },
    {
      "entropy": 5.722633218765258,
      "epoch": 5.207221879036334,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002641720594411401,
      "loss": 4.8837,
      "mean_token_accuracy": 0.23902569264173507,
      "num_tokens": 103500915.0,
      "step": 46365
    },
    {
      "entropy": 5.745269155502319,
      "epoch": 5.207783455944291,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026413195741261145,
      "loss": 4.9295,
      "mean_token_accuracy": 0.236953441798687,
      "num_tokens": 103511739.0,
      "step": 46370
    },
    {
      "entropy": 5.739796257019043,
      "epoch": 5.208345032852249,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002640918557301256,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22144089192152022,
      "num_tokens": 103521957.0,
      "step": 46375
    },
    {
      "entropy": 5.698322677612305,
      "epoch": 5.208906609760207,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002640517543949592,
      "loss": 4.9308,
      "mean_token_accuracy": 0.2354379951953888,
      "num_tokens": 103532511.0,
      "step": 46380
    },
    {
      "entropy": 5.731065177917481,
      "epoch": 5.209468186668164,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002640116534083892,
      "loss": 4.9175,
      "mean_token_accuracy": 0.24048219621181488,
      "num_tokens": 103544484.0,
      "step": 46385
    },
    {
      "entropy": 5.731211757659912,
      "epoch": 5.210029763576122,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026397155277169246,
      "loss": 4.8814,
      "mean_token_accuracy": 0.23475283533334732,
      "num_tokens": 103555020.0,
      "step": 46390
    },
    {
      "entropy": 5.701137447357178,
      "epoch": 5.210591340484079,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002639314524861457,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2329340472817421,
      "num_tokens": 103565331.0,
      "step": 46395
    },
    {
      "entropy": 5.738254356384277,
      "epoch": 5.211152917392037,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002638913525530258,
      "loss": 4.9411,
      "mean_token_accuracy": 0.23996510207653046,
      "num_tokens": 103576970.0,
      "step": 46400
    },
    {
      "entropy": 5.717150545120239,
      "epoch": 5.2117144942999944,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026385125297360946,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22716268450021743,
      "num_tokens": 103587981.0,
      "step": 46405
    },
    {
      "entropy": 5.7298726558685305,
      "epoch": 5.212276071207952,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026381115374917363,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23643561005592345,
      "num_tokens": 103599508.0,
      "step": 46410
    },
    {
      "entropy": 5.7057137966156,
      "epoch": 5.21283764811591,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026377105488099503,
      "loss": 4.8557,
      "mean_token_accuracy": 0.23506652265787126,
      "num_tokens": 103609982.0,
      "step": 46415
    },
    {
      "entropy": 5.659122943878174,
      "epoch": 5.213399225023867,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026373095637035037,
      "loss": 4.8659,
      "mean_token_accuracy": 0.24030200093984605,
      "num_tokens": 103620419.0,
      "step": 46420
    },
    {
      "entropy": 5.637861490249634,
      "epoch": 5.213960801931824,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026369085821851645,
      "loss": 4.8411,
      "mean_token_accuracy": 0.24087354838848113,
      "num_tokens": 103630512.0,
      "step": 46425
    },
    {
      "entropy": 5.7438570022583,
      "epoch": 5.214522378839782,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026365076042677,
      "loss": 4.9525,
      "mean_token_accuracy": 0.23865618854761123,
      "num_tokens": 103641596.0,
      "step": 46430
    },
    {
      "entropy": 5.734738397598266,
      "epoch": 5.2150839557477395,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002636106629963877,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23853284865617752,
      "num_tokens": 103654291.0,
      "step": 46435
    },
    {
      "entropy": 5.615873622894287,
      "epoch": 5.215645532655698,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002635705659286463,
      "loss": 4.7452,
      "mean_token_accuracy": 0.24583536237478257,
      "num_tokens": 103665843.0,
      "step": 46440
    },
    {
      "entropy": 5.771135997772217,
      "epoch": 5.216207109563655,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026353046922482256,
      "loss": 5.0459,
      "mean_token_accuracy": 0.22921413630247117,
      "num_tokens": 103676682.0,
      "step": 46445
    },
    {
      "entropy": 5.709071826934815,
      "epoch": 5.216768686471612,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002634903728861931,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2274070754647255,
      "num_tokens": 103687189.0,
      "step": 46450
    },
    {
      "entropy": 5.697632503509522,
      "epoch": 5.21733026337957,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002634502769140346,
      "loss": 4.8798,
      "mean_token_accuracy": 0.24004128724336624,
      "num_tokens": 103698332.0,
      "step": 46455
    },
    {
      "entropy": 5.704958868026734,
      "epoch": 5.217891840287527,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026341018130962384,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23309043049812317,
      "num_tokens": 103708652.0,
      "step": 46460
    },
    {
      "entropy": 5.723262977600098,
      "epoch": 5.218453417195485,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026337008607423733,
      "loss": 4.9099,
      "mean_token_accuracy": 0.23372831791639329,
      "num_tokens": 103719275.0,
      "step": 46465
    },
    {
      "entropy": 5.67871994972229,
      "epoch": 5.219014994103443,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026332999120915184,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23820821493864058,
      "num_tokens": 103730865.0,
      "step": 46470
    },
    {
      "entropy": 5.727453231811523,
      "epoch": 5.2195765710114,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026328989671564395,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22471209168434142,
      "num_tokens": 103743784.0,
      "step": 46475
    },
    {
      "entropy": 5.761459684371948,
      "epoch": 5.220138147919357,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002632498025949902,
      "loss": 5.0151,
      "mean_token_accuracy": 0.2299495279788971,
      "num_tokens": 103755580.0,
      "step": 46480
    },
    {
      "entropy": 5.748398160934448,
      "epoch": 5.220699724827315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002632097088484674,
      "loss": 5.0302,
      "mean_token_accuracy": 0.22365171760320662,
      "num_tokens": 103767296.0,
      "step": 46485
    },
    {
      "entropy": 5.714146518707276,
      "epoch": 5.2212613017352725,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026316961547735203,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23652060329914093,
      "num_tokens": 103778176.0,
      "step": 46490
    },
    {
      "entropy": 5.797995758056641,
      "epoch": 5.221822878643231,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026312952248292067,
      "loss": 4.994,
      "mean_token_accuracy": 0.23562780767679214,
      "num_tokens": 103791499.0,
      "step": 46495
    },
    {
      "entropy": 5.727673578262329,
      "epoch": 5.222384455551188,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002630894298664499,
      "loss": 4.9509,
      "mean_token_accuracy": 0.23547192811965942,
      "num_tokens": 103803102.0,
      "step": 46500
    },
    {
      "entropy": 5.733958721160889,
      "epoch": 5.222946032459145,
      "grad_norm": 1.25,
      "learning_rate": 0.00026304933762921637,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23624562174081803,
      "num_tokens": 103814273.0,
      "step": 46505
    },
    {
      "entropy": 5.667802858352661,
      "epoch": 5.223507609367103,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002630092457724965,
      "loss": 4.7493,
      "mean_token_accuracy": 0.2471777766942978,
      "num_tokens": 103823713.0,
      "step": 46510
    },
    {
      "entropy": 5.654237604141235,
      "epoch": 5.22406918627506,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002629691542975669,
      "loss": 4.8017,
      "mean_token_accuracy": 0.24062236696481704,
      "num_tokens": 103834244.0,
      "step": 46515
    },
    {
      "entropy": 5.582267475128174,
      "epoch": 5.224630763183018,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002629290632057041,
      "loss": 4.7486,
      "mean_token_accuracy": 0.24826951771974565,
      "num_tokens": 103844561.0,
      "step": 46520
    },
    {
      "entropy": 5.728285121917724,
      "epoch": 5.225192340090976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002628889724981846,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22834345698356628,
      "num_tokens": 103856812.0,
      "step": 46525
    },
    {
      "entropy": 5.804863214492798,
      "epoch": 5.225753916998933,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002628488821762849,
      "loss": 5.0235,
      "mean_token_accuracy": 0.22302734106779099,
      "num_tokens": 103870062.0,
      "step": 46530
    },
    {
      "entropy": 5.726810121536255,
      "epoch": 5.22631549390689,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026280879224128165,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22889400720596315,
      "num_tokens": 103880763.0,
      "step": 46535
    },
    {
      "entropy": 5.72319188117981,
      "epoch": 5.226877070814848,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000262768702694451,
      "loss": 4.9213,
      "mean_token_accuracy": 0.23071005940437317,
      "num_tokens": 103892528.0,
      "step": 46540
    },
    {
      "entropy": 5.710497665405273,
      "epoch": 5.227438647722805,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002627286135370697,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23307046443223953,
      "num_tokens": 103903589.0,
      "step": 46545
    },
    {
      "entropy": 5.670304250717163,
      "epoch": 5.2280002246307635,
      "grad_norm": 1.28125,
      "learning_rate": 0.000262688524770414,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23097388148307801,
      "num_tokens": 103914192.0,
      "step": 46550
    },
    {
      "entropy": 5.692320060729981,
      "epoch": 5.228561801538721,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026264843639576055,
      "loss": 4.817,
      "mean_token_accuracy": 0.23694709092378616,
      "num_tokens": 103924649.0,
      "step": 46555
    },
    {
      "entropy": 5.619986486434937,
      "epoch": 5.229123378446678,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002626083484143856,
      "loss": 4.8232,
      "mean_token_accuracy": 0.24171436578035355,
      "num_tokens": 103935834.0,
      "step": 46560
    },
    {
      "entropy": 5.600392055511475,
      "epoch": 5.229684955354636,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002625682608275657,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23191910088062287,
      "num_tokens": 103947007.0,
      "step": 46565
    },
    {
      "entropy": 5.709244346618652,
      "epoch": 5.230246532262593,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002625281736365772,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2350715920329094,
      "num_tokens": 103958424.0,
      "step": 46570
    },
    {
      "entropy": 5.778601741790771,
      "epoch": 5.2308081091705505,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026248808684269646,
      "loss": 4.9458,
      "mean_token_accuracy": 0.23455306142568588,
      "num_tokens": 103969289.0,
      "step": 46575
    },
    {
      "entropy": 5.821851587295532,
      "epoch": 5.231369686078509,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002624480004472,
      "loss": 4.9759,
      "mean_token_accuracy": 0.23103559613227845,
      "num_tokens": 103981081.0,
      "step": 46580
    },
    {
      "entropy": 5.749802303314209,
      "epoch": 5.231931262986466,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026240791445136384,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22816327214241028,
      "num_tokens": 103990904.0,
      "step": 46585
    },
    {
      "entropy": 5.682088565826416,
      "epoch": 5.232492839894424,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002623678288564647,
      "loss": 4.916,
      "mean_token_accuracy": 0.23746413737535477,
      "num_tokens": 104001535.0,
      "step": 46590
    },
    {
      "entropy": 5.653249216079712,
      "epoch": 5.233054416802381,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002623277436637788,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23161215484142303,
      "num_tokens": 104012840.0,
      "step": 46595
    },
    {
      "entropy": 5.706270980834961,
      "epoch": 5.233615993710338,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002622876588745825,
      "loss": 4.8155,
      "mean_token_accuracy": 0.24273701012134552,
      "num_tokens": 104024035.0,
      "step": 46600
    },
    {
      "entropy": 5.740480422973633,
      "epoch": 5.2341775706182965,
      "grad_norm": 1.171875,
      "learning_rate": 0.000262247574490152,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2416738525032997,
      "num_tokens": 104035391.0,
      "step": 46605
    },
    {
      "entropy": 5.740904331207275,
      "epoch": 5.234739147526254,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002622074905117637,
      "loss": 4.9111,
      "mean_token_accuracy": 0.23687534183263778,
      "num_tokens": 104046645.0,
      "step": 46610
    },
    {
      "entropy": 5.659625101089477,
      "epoch": 5.235300724434211,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026216740694069387,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23693876564502717,
      "num_tokens": 104057706.0,
      "step": 46615
    },
    {
      "entropy": 5.646159601211548,
      "epoch": 5.235862301342169,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002621273237782188,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2336444854736328,
      "num_tokens": 104068076.0,
      "step": 46620
    },
    {
      "entropy": 5.679137897491455,
      "epoch": 5.236423878250126,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002620872410256146,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22901256680488585,
      "num_tokens": 104078116.0,
      "step": 46625
    },
    {
      "entropy": 5.645619630813599,
      "epoch": 5.236985455158084,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002620471586841577,
      "loss": 4.7383,
      "mean_token_accuracy": 0.24708303660154343,
      "num_tokens": 104089543.0,
      "step": 46630
    },
    {
      "entropy": 5.659222078323364,
      "epoch": 5.2375470320660416,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002620070767551244,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2407239556312561,
      "num_tokens": 104100824.0,
      "step": 46635
    },
    {
      "entropy": 5.652384328842163,
      "epoch": 5.238108608973999,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002619669952397908,
      "loss": 4.914,
      "mean_token_accuracy": 0.23547839522361755,
      "num_tokens": 104112410.0,
      "step": 46640
    },
    {
      "entropy": 5.660143852233887,
      "epoch": 5.238670185881957,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002619269141394331,
      "loss": 4.926,
      "mean_token_accuracy": 0.23291048109531404,
      "num_tokens": 104124299.0,
      "step": 46645
    },
    {
      "entropy": 5.817063522338867,
      "epoch": 5.239231762789914,
      "grad_norm": 1.25,
      "learning_rate": 0.0002618868334553275,
      "loss": 4.9936,
      "mean_token_accuracy": 0.2250005915760994,
      "num_tokens": 104136266.0,
      "step": 46650
    },
    {
      "entropy": 5.7402002811431885,
      "epoch": 5.239793339697871,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026184675318875023,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2288266807794571,
      "num_tokens": 104146449.0,
      "step": 46655
    },
    {
      "entropy": 5.6562518119812015,
      "epoch": 5.240354916605829,
      "grad_norm": 1.25,
      "learning_rate": 0.00026180667334097733,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23727083206176758,
      "num_tokens": 104156959.0,
      "step": 46660
    },
    {
      "entropy": 5.648704528808594,
      "epoch": 5.240916493513787,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002617665939132852,
      "loss": 4.839,
      "mean_token_accuracy": 0.23321563750505447,
      "num_tokens": 104167834.0,
      "step": 46665
    },
    {
      "entropy": 5.754955148696899,
      "epoch": 5.241478070421744,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002617265149069498,
      "loss": 4.9635,
      "mean_token_accuracy": 0.23238086104393005,
      "num_tokens": 104180035.0,
      "step": 46670
    },
    {
      "entropy": 5.823697328567505,
      "epoch": 5.242039647329702,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002616864363232474,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2325715959072113,
      "num_tokens": 104192960.0,
      "step": 46675
    },
    {
      "entropy": 5.769395112991333,
      "epoch": 5.242601224237659,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026164635816345395,
      "loss": 4.9575,
      "mean_token_accuracy": 0.23100510537624358,
      "num_tokens": 104204120.0,
      "step": 46680
    },
    {
      "entropy": 5.675906801223755,
      "epoch": 5.243162801145617,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002616062804288457,
      "loss": 4.888,
      "mean_token_accuracy": 0.23598155826330186,
      "num_tokens": 104215222.0,
      "step": 46685
    },
    {
      "entropy": 5.698899173736573,
      "epoch": 5.2437243780535745,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002615662031206986,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2341357111930847,
      "num_tokens": 104227208.0,
      "step": 46690
    },
    {
      "entropy": 5.753890132904052,
      "epoch": 5.244285954961532,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002615261262402887,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22758207619190216,
      "num_tokens": 104236836.0,
      "step": 46695
    },
    {
      "entropy": 5.720966148376465,
      "epoch": 5.24484753186949,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026148604978889237,
      "loss": 4.9069,
      "mean_token_accuracy": 0.23551147282123566,
      "num_tokens": 104248950.0,
      "step": 46700
    },
    {
      "entropy": 5.736207818984985,
      "epoch": 5.245409108777447,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026144597376778533,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2301898717880249,
      "num_tokens": 104260179.0,
      "step": 46705
    },
    {
      "entropy": 5.722194862365723,
      "epoch": 5.245970685685404,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002614058981782438,
      "loss": 5.0085,
      "mean_token_accuracy": 0.22630532532930375,
      "num_tokens": 104273704.0,
      "step": 46710
    },
    {
      "entropy": 5.689619350433349,
      "epoch": 5.246532262593362,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026136582302154377,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23863385766744613,
      "num_tokens": 104283851.0,
      "step": 46715
    },
    {
      "entropy": 5.713491296768188,
      "epoch": 5.24709383950132,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026132574829896123,
      "loss": 4.8812,
      "mean_token_accuracy": 0.24484591335058212,
      "num_tokens": 104296064.0,
      "step": 46720
    },
    {
      "entropy": 5.770611715316773,
      "epoch": 5.247655416409277,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002612856740117721,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22080347537994385,
      "num_tokens": 104309016.0,
      "step": 46725
    },
    {
      "entropy": 5.7211456298828125,
      "epoch": 5.248216993317235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026124560016125236,
      "loss": 4.9246,
      "mean_token_accuracy": 0.2347528338432312,
      "num_tokens": 104320598.0,
      "step": 46730
    },
    {
      "entropy": 5.76572585105896,
      "epoch": 5.248778570225192,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002612055267486782,
      "loss": 4.9417,
      "mean_token_accuracy": 0.22773464024066925,
      "num_tokens": 104332055.0,
      "step": 46735
    },
    {
      "entropy": 5.732954788208008,
      "epoch": 5.24934014713315,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026116545377532536,
      "loss": 4.9248,
      "mean_token_accuracy": 0.23482657074928284,
      "num_tokens": 104343879.0,
      "step": 46740
    },
    {
      "entropy": 5.680763626098633,
      "epoch": 5.249901724041107,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002611253812424698,
      "loss": 4.912,
      "mean_token_accuracy": 0.2319685474038124,
      "num_tokens": 104354653.0,
      "step": 46745
    },
    {
      "entropy": 5.757514619827271,
      "epoch": 5.250463300949065,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002610853091513876,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2352791830897331,
      "num_tokens": 104365273.0,
      "step": 46750
    },
    {
      "entropy": 5.774101066589355,
      "epoch": 5.251024877857023,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026104523750335445,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22510050535202025,
      "num_tokens": 104375990.0,
      "step": 46755
    },
    {
      "entropy": 5.697412729263306,
      "epoch": 5.25158645476498,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002610051662996464,
      "loss": 4.8177,
      "mean_token_accuracy": 0.24094783663749694,
      "num_tokens": 104385975.0,
      "step": 46760
    },
    {
      "entropy": 5.75560884475708,
      "epoch": 5.252148031672937,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026096509554153914,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2379390776157379,
      "num_tokens": 104396053.0,
      "step": 46765
    },
    {
      "entropy": 5.63915638923645,
      "epoch": 5.252709608580895,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002609250252303088,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2365503951907158,
      "num_tokens": 104407874.0,
      "step": 46770
    },
    {
      "entropy": 5.6835693359375,
      "epoch": 5.2532711854888525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002608849553672311,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2415129080414772,
      "num_tokens": 104418527.0,
      "step": 46775
    },
    {
      "entropy": 5.683479118347168,
      "epoch": 5.253832762396811,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026084488595358197,
      "loss": 4.8987,
      "mean_token_accuracy": 0.24040063619613647,
      "num_tokens": 104430490.0,
      "step": 46780
    },
    {
      "entropy": 5.7373175621032715,
      "epoch": 5.254394339304768,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002608048169906371,
      "loss": 4.8449,
      "mean_token_accuracy": 0.24023679196834563,
      "num_tokens": 104441767.0,
      "step": 46785
    },
    {
      "entropy": 5.658791780471802,
      "epoch": 5.254955916212725,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026076474847967236,
      "loss": 4.9018,
      "mean_token_accuracy": 0.24007322937250136,
      "num_tokens": 104452720.0,
      "step": 46790
    },
    {
      "entropy": 5.7363299369812015,
      "epoch": 5.255517493120683,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026072468042196356,
      "loss": 4.9347,
      "mean_token_accuracy": 0.23236274421215058,
      "num_tokens": 104463829.0,
      "step": 46795
    },
    {
      "entropy": 5.720104074478149,
      "epoch": 5.25607907002864,
      "grad_norm": 1.390625,
      "learning_rate": 0.00026068461281878645,
      "loss": 4.875,
      "mean_token_accuracy": 0.2359563246369362,
      "num_tokens": 104474830.0,
      "step": 46800
    },
    {
      "entropy": 5.691752624511719,
      "epoch": 5.256640646936598,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026064454567141683,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23568983227014542,
      "num_tokens": 104485518.0,
      "step": 46805
    },
    {
      "entropy": 5.771925640106201,
      "epoch": 5.257202223844556,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002606044789811304,
      "loss": 4.9531,
      "mean_token_accuracy": 0.2309253916144371,
      "num_tokens": 104495562.0,
      "step": 46810
    },
    {
      "entropy": 5.720928812026978,
      "epoch": 5.257763800752513,
      "grad_norm": 1.265625,
      "learning_rate": 0.000260564412749203,
      "loss": 4.8658,
      "mean_token_accuracy": 0.23923284858465194,
      "num_tokens": 104506057.0,
      "step": 46815
    },
    {
      "entropy": 5.659759855270385,
      "epoch": 5.258325377660471,
      "grad_norm": 1.125,
      "learning_rate": 0.00026052434697691035,
      "loss": 4.869,
      "mean_token_accuracy": 0.24442353397607802,
      "num_tokens": 104518269.0,
      "step": 46820
    },
    {
      "entropy": 5.6745843410491945,
      "epoch": 5.258886954568428,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000260484281665528,
      "loss": 4.8497,
      "mean_token_accuracy": 0.2337751343846321,
      "num_tokens": 104528469.0,
      "step": 46825
    },
    {
      "entropy": 5.718021011352539,
      "epoch": 5.2594485314763855,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002604442168163318,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2279845356941223,
      "num_tokens": 104539079.0,
      "step": 46830
    },
    {
      "entropy": 5.688871145248413,
      "epoch": 5.260010108384344,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002604041524305974,
      "loss": 5.0042,
      "mean_token_accuracy": 0.23231087476015091,
      "num_tokens": 104551235.0,
      "step": 46835
    },
    {
      "entropy": 5.847911930084228,
      "epoch": 5.260571685292301,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002603640885096004,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22903861105442047,
      "num_tokens": 104563911.0,
      "step": 46840
    },
    {
      "entropy": 5.723000001907349,
      "epoch": 5.261133262200258,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002603240250546165,
      "loss": 4.94,
      "mean_token_accuracy": 0.23478449881076813,
      "num_tokens": 104574877.0,
      "step": 46845
    },
    {
      "entropy": 5.740295171737671,
      "epoch": 5.261694839108216,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026028396206692134,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22897557467222213,
      "num_tokens": 104585209.0,
      "step": 46850
    },
    {
      "entropy": 5.645459413528442,
      "epoch": 5.262256416016173,
      "grad_norm": 1.25,
      "learning_rate": 0.00026024389954779054,
      "loss": 4.8684,
      "mean_token_accuracy": 0.24123035222291947,
      "num_tokens": 104596213.0,
      "step": 46855
    },
    {
      "entropy": 5.681660318374634,
      "epoch": 5.2628179929241306,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002602038374984997,
      "loss": 4.8906,
      "mean_token_accuracy": 0.23549729436635972,
      "num_tokens": 104606813.0,
      "step": 46860
    },
    {
      "entropy": 5.742633962631226,
      "epoch": 5.263379569832089,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026016377592032446,
      "loss": 4.9462,
      "mean_token_accuracy": 0.23536335676908493,
      "num_tokens": 104619040.0,
      "step": 46865
    },
    {
      "entropy": 5.602248811721802,
      "epoch": 5.263941146740046,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002601237148145403,
      "loss": 4.7571,
      "mean_token_accuracy": 0.24091448336839677,
      "num_tokens": 104629879.0,
      "step": 46870
    },
    {
      "entropy": 5.6634605884552,
      "epoch": 5.264502723648004,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026008365418242284,
      "loss": 4.8797,
      "mean_token_accuracy": 0.24062736332416534,
      "num_tokens": 104641408.0,
      "step": 46875
    },
    {
      "entropy": 5.813330936431885,
      "epoch": 5.265064300555961,
      "grad_norm": 1.25,
      "learning_rate": 0.0002600435940252476,
      "loss": 5.0227,
      "mean_token_accuracy": 0.22836972922086715,
      "num_tokens": 104653537.0,
      "step": 46880
    },
    {
      "entropy": 5.658177900314331,
      "epoch": 5.265625877463918,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002600035343442902,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2438049152493477,
      "num_tokens": 104665499.0,
      "step": 46885
    },
    {
      "entropy": 5.6947520732879635,
      "epoch": 5.2661874543718765,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025996347514082603,
      "loss": 4.9462,
      "mean_token_accuracy": 0.23193206042051315,
      "num_tokens": 104676964.0,
      "step": 46890
    },
    {
      "entropy": 5.773726320266723,
      "epoch": 5.266749031279834,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025992341641613065,
      "loss": 4.9619,
      "mean_token_accuracy": 0.23134741932153702,
      "num_tokens": 104688853.0,
      "step": 46895
    },
    {
      "entropy": 5.680579948425293,
      "epoch": 5.267310608187791,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002598833581714795,
      "loss": 4.8629,
      "mean_token_accuracy": 0.23620384633541108,
      "num_tokens": 104699723.0,
      "step": 46900
    },
    {
      "entropy": 5.729286050796508,
      "epoch": 5.267872185095749,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025984330040814817,
      "loss": 4.9127,
      "mean_token_accuracy": 0.23567592203617097,
      "num_tokens": 104711370.0,
      "step": 46905
    },
    {
      "entropy": 5.75636396408081,
      "epoch": 5.268433762003706,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002598032431274119,
      "loss": 4.997,
      "mean_token_accuracy": 0.22620311826467515,
      "num_tokens": 104722939.0,
      "step": 46910
    },
    {
      "entropy": 5.740830898284912,
      "epoch": 5.2689953389116635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025976318633054636,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22910384386777877,
      "num_tokens": 104735433.0,
      "step": 46915
    },
    {
      "entropy": 5.73405385017395,
      "epoch": 5.269556915819622,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002597231300188269,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2360968232154846,
      "num_tokens": 104748325.0,
      "step": 46920
    },
    {
      "entropy": 5.743597555160522,
      "epoch": 5.270118492727579,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025968307419352884,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2318265601992607,
      "num_tokens": 104758745.0,
      "step": 46925
    },
    {
      "entropy": 5.674802684783936,
      "epoch": 5.270680069635537,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025964301885592766,
      "loss": 4.808,
      "mean_token_accuracy": 0.23921433538198472,
      "num_tokens": 104769424.0,
      "step": 46930
    },
    {
      "entropy": 5.618935489654541,
      "epoch": 5.271241646543494,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002596029640072988,
      "loss": 4.8335,
      "mean_token_accuracy": 0.24153956025838852,
      "num_tokens": 104779913.0,
      "step": 46935
    },
    {
      "entropy": 5.665738534927368,
      "epoch": 5.271803223451451,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002595629096489174,
      "loss": 4.916,
      "mean_token_accuracy": 0.23039576560258865,
      "num_tokens": 104793066.0,
      "step": 46940
    },
    {
      "entropy": 5.679321193695069,
      "epoch": 5.2723648003594095,
      "grad_norm": 1.296875,
      "learning_rate": 0.000259522855782059,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2324199855327606,
      "num_tokens": 104803641.0,
      "step": 46945
    },
    {
      "entropy": 5.8137799263000485,
      "epoch": 5.272926377267367,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002594828024079989,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22271984517574311,
      "num_tokens": 104816025.0,
      "step": 46950
    },
    {
      "entropy": 5.785797739028931,
      "epoch": 5.273487954175324,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025944274952801233,
      "loss": 4.991,
      "mean_token_accuracy": 0.23009854853153228,
      "num_tokens": 104828188.0,
      "step": 46955
    },
    {
      "entropy": 5.720520114898681,
      "epoch": 5.274049531083282,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002594026971433747,
      "loss": 4.941,
      "mean_token_accuracy": 0.2288806676864624,
      "num_tokens": 104838276.0,
      "step": 46960
    },
    {
      "entropy": 5.719373893737793,
      "epoch": 5.274611107991239,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025936264525536126,
      "loss": 4.9335,
      "mean_token_accuracy": 0.23375141024589538,
      "num_tokens": 104848295.0,
      "step": 46965
    },
    {
      "entropy": 5.796766948699951,
      "epoch": 5.275172684899197,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002593225938652472,
      "loss": 5.0353,
      "mean_token_accuracy": 0.22695463448762893,
      "num_tokens": 104857857.0,
      "step": 46970
    },
    {
      "entropy": 5.696336555480957,
      "epoch": 5.2757342618071545,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002592825429743079,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23092270791530609,
      "num_tokens": 104867819.0,
      "step": 46975
    },
    {
      "entropy": 5.797442436218262,
      "epoch": 5.276295838715112,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002592424925838185,
      "loss": 4.9291,
      "mean_token_accuracy": 0.23119294941425322,
      "num_tokens": 104879034.0,
      "step": 46980
    },
    {
      "entropy": 5.773262405395508,
      "epoch": 5.27685741562307,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002592024426950541,
      "loss": 5.0218,
      "mean_token_accuracy": 0.22336367666721343,
      "num_tokens": 104890245.0,
      "step": 46985
    },
    {
      "entropy": 5.702397775650025,
      "epoch": 5.277418992531027,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025916239330929026,
      "loss": 4.9385,
      "mean_token_accuracy": 0.23150516897439957,
      "num_tokens": 104900805.0,
      "step": 46990
    },
    {
      "entropy": 5.672099018096924,
      "epoch": 5.277980569438984,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025912234442780185,
      "loss": 4.8231,
      "mean_token_accuracy": 0.24136356115341187,
      "num_tokens": 104911636.0,
      "step": 46995
    },
    {
      "entropy": 5.740815019607544,
      "epoch": 5.278542146346942,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025908229605186423,
      "loss": 4.9167,
      "mean_token_accuracy": 0.23425903916358948,
      "num_tokens": 104922253.0,
      "step": 47000
    },
    {
      "entropy": 5.645400333404541,
      "epoch": 5.2791037232549,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025904224818275247,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2381330505013466,
      "num_tokens": 104933274.0,
      "step": 47005
    },
    {
      "entropy": 5.745932245254517,
      "epoch": 5.279665300162858,
      "grad_norm": 1.25,
      "learning_rate": 0.00025900220082174165,
      "loss": 4.9686,
      "mean_token_accuracy": 0.22317835837602615,
      "num_tokens": 104943931.0,
      "step": 47010
    },
    {
      "entropy": 5.673333597183228,
      "epoch": 5.280226877070815,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025896215397010696,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2291715994477272,
      "num_tokens": 104955080.0,
      "step": 47015
    },
    {
      "entropy": 5.747134733200073,
      "epoch": 5.280788453978772,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002589221076291236,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23815740495920182,
      "num_tokens": 104967232.0,
      "step": 47020
    },
    {
      "entropy": 5.733242750167847,
      "epoch": 5.28135003088673,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025888206180006647,
      "loss": 4.965,
      "mean_token_accuracy": 0.23028009682893752,
      "num_tokens": 104979462.0,
      "step": 47025
    },
    {
      "entropy": 5.693036651611328,
      "epoch": 5.2819116077946875,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002588420164842108,
      "loss": 4.951,
      "mean_token_accuracy": 0.22820139527320862,
      "num_tokens": 104990567.0,
      "step": 47030
    },
    {
      "entropy": 5.612577104568482,
      "epoch": 5.282473184702645,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002588019716828316,
      "loss": 4.7657,
      "mean_token_accuracy": 0.24198840856552123,
      "num_tokens": 105001965.0,
      "step": 47035
    },
    {
      "entropy": 5.70015959739685,
      "epoch": 5.283034761610603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025876192739720383,
      "loss": 4.7976,
      "mean_token_accuracy": 0.24152784496545793,
      "num_tokens": 105013349.0,
      "step": 47040
    },
    {
      "entropy": 5.704971742630005,
      "epoch": 5.28359633851856,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025872188362860264,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23151450157165526,
      "num_tokens": 105025009.0,
      "step": 47045
    },
    {
      "entropy": 5.638602828979492,
      "epoch": 5.284157915426517,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025868184037830305,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23999378830194473,
      "num_tokens": 105035474.0,
      "step": 47050
    },
    {
      "entropy": 5.742486429214478,
      "epoch": 5.284719492334475,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002586417976475798,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2302134230732918,
      "num_tokens": 105047310.0,
      "step": 47055
    },
    {
      "entropy": 5.632612562179565,
      "epoch": 5.285281069242433,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025860175543770817,
      "loss": 4.789,
      "mean_token_accuracy": 0.23504245281219482,
      "num_tokens": 105057198.0,
      "step": 47060
    },
    {
      "entropy": 5.748014450073242,
      "epoch": 5.285842646150391,
      "grad_norm": 1.25,
      "learning_rate": 0.0002585617137499631,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23555559515953065,
      "num_tokens": 105067885.0,
      "step": 47065
    },
    {
      "entropy": 5.7350921630859375,
      "epoch": 5.286404223058348,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002585216725856193,
      "loss": 4.9502,
      "mean_token_accuracy": 0.23022558838129042,
      "num_tokens": 105078917.0,
      "step": 47070
    },
    {
      "entropy": 5.745331573486328,
      "epoch": 5.286965799966305,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002584816319459519,
      "loss": 5.01,
      "mean_token_accuracy": 0.22591779232025147,
      "num_tokens": 105089425.0,
      "step": 47075
    },
    {
      "entropy": 5.7991071224212645,
      "epoch": 5.287527376874263,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002584415918322358,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23495145440101622,
      "num_tokens": 105100043.0,
      "step": 47080
    },
    {
      "entropy": 5.671721601486206,
      "epoch": 5.28808895378222,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002584015522457458,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23632175475358963,
      "num_tokens": 105110317.0,
      "step": 47085
    },
    {
      "entropy": 5.72740683555603,
      "epoch": 5.288650530690178,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025836151318775675,
      "loss": 4.9237,
      "mean_token_accuracy": 0.22999488264322282,
      "num_tokens": 105120688.0,
      "step": 47090
    },
    {
      "entropy": 5.812247228622437,
      "epoch": 5.289212107598136,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025832147465954365,
      "loss": 5.0155,
      "mean_token_accuracy": 0.22893061190843583,
      "num_tokens": 105130883.0,
      "step": 47095
    },
    {
      "entropy": 5.799921655654908,
      "epoch": 5.289773684506093,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025828143666238124,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2298721820116043,
      "num_tokens": 105141203.0,
      "step": 47100
    },
    {
      "entropy": 5.7272943496704105,
      "epoch": 5.29033526141405,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025824139919754436,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2384709507226944,
      "num_tokens": 105152419.0,
      "step": 47105
    },
    {
      "entropy": 5.798831224441528,
      "epoch": 5.290896838322008,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002582013622663079,
      "loss": 5.0429,
      "mean_token_accuracy": 0.22754748314619064,
      "num_tokens": 105164242.0,
      "step": 47110
    },
    {
      "entropy": 5.751965522766113,
      "epoch": 5.2914584152299655,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002581613258699465,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2333269789814949,
      "num_tokens": 105174008.0,
      "step": 47115
    },
    {
      "entropy": 5.654055976867676,
      "epoch": 5.292019992137924,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000258121290009735,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2299996316432953,
      "num_tokens": 105184880.0,
      "step": 47120
    },
    {
      "entropy": 5.750131750106812,
      "epoch": 5.292581569045881,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002580812546869481,
      "loss": 4.8931,
      "mean_token_accuracy": 0.23032632172107698,
      "num_tokens": 105196492.0,
      "step": 47125
    },
    {
      "entropy": 5.661539602279663,
      "epoch": 5.293143145953838,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002580412199028607,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23665878623723985,
      "num_tokens": 105206227.0,
      "step": 47130
    },
    {
      "entropy": 5.816286182403564,
      "epoch": 5.293704722861796,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025800118565874754,
      "loss": 5.0731,
      "mean_token_accuracy": 0.22788214534521103,
      "num_tokens": 105217877.0,
      "step": 47135
    },
    {
      "entropy": 5.724438762664795,
      "epoch": 5.294266299769753,
      "grad_norm": 1.25,
      "learning_rate": 0.0002579611519558831,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23757777363061905,
      "num_tokens": 105229124.0,
      "step": 47140
    },
    {
      "entropy": 5.640496253967285,
      "epoch": 5.294827876677711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025792111879554224,
      "loss": 4.7861,
      "mean_token_accuracy": 0.24926097840070724,
      "num_tokens": 105240952.0,
      "step": 47145
    },
    {
      "entropy": 5.697187614440918,
      "epoch": 5.295389453585669,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002578810861789996,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23841699808835984,
      "num_tokens": 105252594.0,
      "step": 47150
    },
    {
      "entropy": 5.791427707672119,
      "epoch": 5.295951030493626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002578410541075298,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22798853665590285,
      "num_tokens": 105263457.0,
      "step": 47155
    },
    {
      "entropy": 5.710557699203491,
      "epoch": 5.296512607401584,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025780102258240745,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2367543175816536,
      "num_tokens": 105275455.0,
      "step": 47160
    },
    {
      "entropy": 5.7393639087677,
      "epoch": 5.297074184309541,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002577609916049072,
      "loss": 4.938,
      "mean_token_accuracy": 0.23281036913394929,
      "num_tokens": 105285643.0,
      "step": 47165
    },
    {
      "entropy": 5.722599458694458,
      "epoch": 5.2976357612174985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002577209611763037,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2344611868262291,
      "num_tokens": 105297322.0,
      "step": 47170
    },
    {
      "entropy": 5.6814593315124515,
      "epoch": 5.298197338125457,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002576809312978715,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23547836542129516,
      "num_tokens": 105307977.0,
      "step": 47175
    },
    {
      "entropy": 5.711939525604248,
      "epoch": 5.298758915033414,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002576409019708852,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2398606076836586,
      "num_tokens": 105318461.0,
      "step": 47180
    },
    {
      "entropy": 5.802626276016236,
      "epoch": 5.299320491941371,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002576008731966193,
      "loss": 5.0232,
      "mean_token_accuracy": 0.22479719817638397,
      "num_tokens": 105329494.0,
      "step": 47185
    },
    {
      "entropy": 5.576816272735596,
      "epoch": 5.299882068849329,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00025756084497634827,
      "loss": 4.7488,
      "mean_token_accuracy": 0.24345605671405793,
      "num_tokens": 105341475.0,
      "step": 47190
    },
    {
      "entropy": 5.720276594161987,
      "epoch": 5.300443645757286,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002575208173113468,
      "loss": 4.837,
      "mean_token_accuracy": 0.24000584781169892,
      "num_tokens": 105352075.0,
      "step": 47195
    },
    {
      "entropy": 5.745607137680054,
      "epoch": 5.301005222665244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002574807902028891,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22956327050924302,
      "num_tokens": 105363801.0,
      "step": 47200
    },
    {
      "entropy": 5.751192808151245,
      "epoch": 5.301566799573202,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025744076365225,
      "loss": 4.9227,
      "mean_token_accuracy": 0.23767729699611664,
      "num_tokens": 105374821.0,
      "step": 47205
    },
    {
      "entropy": 5.709401798248291,
      "epoch": 5.302128376481159,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025740073766070376,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23862674683332444,
      "num_tokens": 105385890.0,
      "step": 47210
    },
    {
      "entropy": 5.752326250076294,
      "epoch": 5.302689953389117,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002573607122295249,
      "loss": 4.971,
      "mean_token_accuracy": 0.22537703514099122,
      "num_tokens": 105396943.0,
      "step": 47215
    },
    {
      "entropy": 5.773328256607056,
      "epoch": 5.303251530297074,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002573206873599878,
      "loss": 4.9748,
      "mean_token_accuracy": 0.23207196444272996,
      "num_tokens": 105407223.0,
      "step": 47220
    },
    {
      "entropy": 5.7715612888336185,
      "epoch": 5.303813107205031,
      "grad_norm": 1.25,
      "learning_rate": 0.00025728066305336685,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23320509642362594,
      "num_tokens": 105417714.0,
      "step": 47225
    },
    {
      "entropy": 5.824811697006226,
      "epoch": 5.3043746841129895,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002572406393109364,
      "loss": 5.059,
      "mean_token_accuracy": 0.23214223533868789,
      "num_tokens": 105429035.0,
      "step": 47230
    },
    {
      "entropy": 5.708117723464966,
      "epoch": 5.304936261020947,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002572006161339709,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2284708634018898,
      "num_tokens": 105440347.0,
      "step": 47235
    },
    {
      "entropy": 5.789874505996704,
      "epoch": 5.305497837928904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002571605935237448,
      "loss": 5.0075,
      "mean_token_accuracy": 0.23104092925786973,
      "num_tokens": 105451265.0,
      "step": 47240
    },
    {
      "entropy": 5.776071405410766,
      "epoch": 5.306059414836862,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002571205714815322,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22551586627960205,
      "num_tokens": 105462034.0,
      "step": 47245
    },
    {
      "entropy": 5.754523086547851,
      "epoch": 5.306620991744819,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002570805500086076,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2375489130616188,
      "num_tokens": 105472733.0,
      "step": 47250
    },
    {
      "entropy": 5.754774522781372,
      "epoch": 5.307182568652777,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002570405291062452,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22886655181646348,
      "num_tokens": 105483152.0,
      "step": 47255
    },
    {
      "entropy": 5.639927721023559,
      "epoch": 5.307744145560735,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002570005087757193,
      "loss": 4.8051,
      "mean_token_accuracy": 0.2411996141076088,
      "num_tokens": 105493541.0,
      "step": 47260
    },
    {
      "entropy": 5.67252368927002,
      "epoch": 5.308305722468692,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002569604890183043,
      "loss": 4.9231,
      "mean_token_accuracy": 0.23651246428489686,
      "num_tokens": 105505550.0,
      "step": 47265
    },
    {
      "entropy": 5.65726523399353,
      "epoch": 5.30886729937665,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002569204698352741,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23658327460289003,
      "num_tokens": 105518460.0,
      "step": 47270
    },
    {
      "entropy": 5.713668012619019,
      "epoch": 5.309428876284607,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002568804512279033,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23584527373313904,
      "num_tokens": 105529074.0,
      "step": 47275
    },
    {
      "entropy": 5.735877609252929,
      "epoch": 5.309990453192564,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025684043319746595,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23271605670452117,
      "num_tokens": 105540501.0,
      "step": 47280
    },
    {
      "entropy": 5.71440954208374,
      "epoch": 5.3105520301005225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025680041574523626,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23498334288597106,
      "num_tokens": 105551158.0,
      "step": 47285
    },
    {
      "entropy": 5.693095874786377,
      "epoch": 5.31111360700848,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002567603988724884,
      "loss": 4.8656,
      "mean_token_accuracy": 0.24733310639858247,
      "num_tokens": 105563219.0,
      "step": 47290
    },
    {
      "entropy": 5.674000453948975,
      "epoch": 5.311675183916437,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002567203825804966,
      "loss": 4.9074,
      "mean_token_accuracy": 0.23836369812488556,
      "num_tokens": 105574846.0,
      "step": 47295
    },
    {
      "entropy": 5.695641565322876,
      "epoch": 5.312236760824395,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002566803668705347,
      "loss": 4.8648,
      "mean_token_accuracy": 0.24041241258382798,
      "num_tokens": 105585848.0,
      "step": 47300
    },
    {
      "entropy": 5.750316476821899,
      "epoch": 5.312798337732352,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002566403517438772,
      "loss": 4.9496,
      "mean_token_accuracy": 0.23601290732622146,
      "num_tokens": 105596757.0,
      "step": 47305
    },
    {
      "entropy": 5.697372913360596,
      "epoch": 5.31335991464031,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025660033720179795,
      "loss": 5.0407,
      "mean_token_accuracy": 0.22765031903982164,
      "num_tokens": 105609363.0,
      "step": 47310
    },
    {
      "entropy": 5.6921971321105955,
      "epoch": 5.3139214915482675,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025656032324557107,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23428680151700973,
      "num_tokens": 105620909.0,
      "step": 47315
    },
    {
      "entropy": 5.638175439834595,
      "epoch": 5.314483068456225,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00025652030987647066,
      "loss": 4.7771,
      "mean_token_accuracy": 0.24100423008203506,
      "num_tokens": 105631134.0,
      "step": 47320
    },
    {
      "entropy": 5.639823913574219,
      "epoch": 5.315044645364183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002564802970957708,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23858678340911865,
      "num_tokens": 105641979.0,
      "step": 47325
    },
    {
      "entropy": 5.700850105285644,
      "epoch": 5.31560622227214,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025644028490474544,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23187279999256133,
      "num_tokens": 105652320.0,
      "step": 47330
    },
    {
      "entropy": 5.715745353698731,
      "epoch": 5.316167799180098,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025640027330466866,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22605280727148055,
      "num_tokens": 105663278.0,
      "step": 47335
    },
    {
      "entropy": 5.71765103340149,
      "epoch": 5.316729376088055,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025636026229681433,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2351987138390541,
      "num_tokens": 105673930.0,
      "step": 47340
    },
    {
      "entropy": 5.680312204360962,
      "epoch": 5.317290952996013,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002563202518824564,
      "loss": 4.8562,
      "mean_token_accuracy": 0.23844639360904693,
      "num_tokens": 105684617.0,
      "step": 47345
    },
    {
      "entropy": 5.653462839126587,
      "epoch": 5.317852529903971,
      "grad_norm": 1.296875,
      "learning_rate": 0.000256280242062869,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2323695182800293,
      "num_tokens": 105695278.0,
      "step": 47350
    },
    {
      "entropy": 5.725755405426026,
      "epoch": 5.318414106811928,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002562402328393259,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22796403467655182,
      "num_tokens": 105706026.0,
      "step": 47355
    },
    {
      "entropy": 5.703262853622436,
      "epoch": 5.318975683719885,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025620022421310115,
      "loss": 4.8563,
      "mean_token_accuracy": 0.23747212886810304,
      "num_tokens": 105717141.0,
      "step": 47360
    },
    {
      "entropy": 5.702901983261109,
      "epoch": 5.319537260627843,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025616021618546847,
      "loss": 4.8911,
      "mean_token_accuracy": 0.23702707886695862,
      "num_tokens": 105729240.0,
      "step": 47365
    },
    {
      "entropy": 5.710912227630615,
      "epoch": 5.3200988375358005,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025612020875770186,
      "loss": 4.887,
      "mean_token_accuracy": 0.23456826955080032,
      "num_tokens": 105741344.0,
      "step": 47370
    },
    {
      "entropy": 5.741385650634766,
      "epoch": 5.320660414443758,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002560802019310751,
      "loss": 4.9046,
      "mean_token_accuracy": 0.24122727215290068,
      "num_tokens": 105752750.0,
      "step": 47375
    },
    {
      "entropy": 5.730955934524536,
      "epoch": 5.321221991351716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025604019570686203,
      "loss": 4.9469,
      "mean_token_accuracy": 0.23569077849388123,
      "num_tokens": 105764677.0,
      "step": 47380
    },
    {
      "entropy": 5.748482322692871,
      "epoch": 5.321783568259673,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002560001900863365,
      "loss": 4.8868,
      "mean_token_accuracy": 0.229813414812088,
      "num_tokens": 105775649.0,
      "step": 47385
    },
    {
      "entropy": 5.748031044006348,
      "epoch": 5.322345145167631,
      "grad_norm": 1.25,
      "learning_rate": 0.0002559601850707723,
      "loss": 4.9075,
      "mean_token_accuracy": 0.23734910786151886,
      "num_tokens": 105785605.0,
      "step": 47390
    },
    {
      "entropy": 5.7284753799438475,
      "epoch": 5.322906722075588,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002559201806614432,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2299339309334755,
      "num_tokens": 105795887.0,
      "step": 47395
    },
    {
      "entropy": 5.664316320419312,
      "epoch": 5.323468298983546,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025588017685962297,
      "loss": 4.8257,
      "mean_token_accuracy": 0.24059315919876098,
      "num_tokens": 105806129.0,
      "step": 47400
    },
    {
      "entropy": 5.758799266815186,
      "epoch": 5.324029875891504,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025584017366658525,
      "loss": 4.9457,
      "mean_token_accuracy": 0.23814863264560698,
      "num_tokens": 105817048.0,
      "step": 47405
    },
    {
      "entropy": 5.730309534072876,
      "epoch": 5.324591452799461,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025580017108360386,
      "loss": 4.9871,
      "mean_token_accuracy": 0.232316692173481,
      "num_tokens": 105829847.0,
      "step": 47410
    },
    {
      "entropy": 5.727626466751099,
      "epoch": 5.325153029707418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002557601691119524,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2410579577088356,
      "num_tokens": 105839797.0,
      "step": 47415
    },
    {
      "entropy": 5.734302759170532,
      "epoch": 5.325714606615376,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025572016775290475,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22856686264276505,
      "num_tokens": 105851004.0,
      "step": 47420
    },
    {
      "entropy": 5.707690668106079,
      "epoch": 5.326276183523333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025568016700773435,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22706363946199418,
      "num_tokens": 105861558.0,
      "step": 47425
    },
    {
      "entropy": 5.681494808197021,
      "epoch": 5.326837760431291,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025564016687771495,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23760860413312912,
      "num_tokens": 105872705.0,
      "step": 47430
    },
    {
      "entropy": 5.665523433685303,
      "epoch": 5.327399337339249,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025560016736412015,
      "loss": 4.9423,
      "mean_token_accuracy": 0.2302128180861473,
      "num_tokens": 105884426.0,
      "step": 47435
    },
    {
      "entropy": 5.679527616500854,
      "epoch": 5.327960914247206,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025556016846822353,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2337416872382164,
      "num_tokens": 105896141.0,
      "step": 47440
    },
    {
      "entropy": 5.689730072021485,
      "epoch": 5.328522491155164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025552017019129874,
      "loss": 4.8455,
      "mean_token_accuracy": 0.24515432715415955,
      "num_tokens": 105907216.0,
      "step": 47445
    },
    {
      "entropy": 5.7901500225067135,
      "epoch": 5.329084068063121,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025548017253461923,
      "loss": 4.9844,
      "mean_token_accuracy": 0.23528571128845216,
      "num_tokens": 105918011.0,
      "step": 47450
    },
    {
      "entropy": 5.801064205169678,
      "epoch": 5.3296456449710785,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002554401754994586,
      "loss": 4.9792,
      "mean_token_accuracy": 0.23217263221740722,
      "num_tokens": 105929173.0,
      "step": 47455
    },
    {
      "entropy": 5.746858263015747,
      "epoch": 5.330207221879037,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002554001790870904,
      "loss": 4.9555,
      "mean_token_accuracy": 0.23242285549640657,
      "num_tokens": 105940825.0,
      "step": 47460
    },
    {
      "entropy": 5.682340812683106,
      "epoch": 5.330768798786994,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025536018329878814,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23968354165554046,
      "num_tokens": 105951335.0,
      "step": 47465
    },
    {
      "entropy": 5.726014709472656,
      "epoch": 5.331330375694951,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002553201881358252,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22530122250318527,
      "num_tokens": 105963529.0,
      "step": 47470
    },
    {
      "entropy": 5.835592031478882,
      "epoch": 5.331891952602909,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025528019359947514,
      "loss": 5.0117,
      "mean_token_accuracy": 0.227929051220417,
      "num_tokens": 105975386.0,
      "step": 47475
    },
    {
      "entropy": 5.702443552017212,
      "epoch": 5.332453529510866,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002552401996910113,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23179688453674316,
      "num_tokens": 105985710.0,
      "step": 47480
    },
    {
      "entropy": 5.654328680038452,
      "epoch": 5.333015106418824,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002552002064117072,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2389496758580208,
      "num_tokens": 105997216.0,
      "step": 47485
    },
    {
      "entropy": 5.752251243591308,
      "epoch": 5.333576683326782,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002551602137628361,
      "loss": 4.8842,
      "mean_token_accuracy": 0.24069187343120574,
      "num_tokens": 106008111.0,
      "step": 47490
    },
    {
      "entropy": 5.732543897628784,
      "epoch": 5.334138260234739,
      "grad_norm": 1.375,
      "learning_rate": 0.00025512022174567164,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2420314446091652,
      "num_tokens": 106018732.0,
      "step": 47495
    },
    {
      "entropy": 5.665557622909546,
      "epoch": 5.334699837142697,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025508023036148697,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23666500002145768,
      "num_tokens": 106029740.0,
      "step": 47500
    },
    {
      "entropy": 5.656563472747803,
      "epoch": 5.335261414050654,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025504023961155553,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2446555495262146,
      "num_tokens": 106040553.0,
      "step": 47505
    },
    {
      "entropy": 5.710263299942016,
      "epoch": 5.3358229909586115,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025500024949715055,
      "loss": 4.9341,
      "mean_token_accuracy": 0.23215208649635316,
      "num_tokens": 106051667.0,
      "step": 47510
    },
    {
      "entropy": 5.7838328838348385,
      "epoch": 5.33638456786657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025496026001954543,
      "loss": 5.0034,
      "mean_token_accuracy": 0.23478787392377853,
      "num_tokens": 106062492.0,
      "step": 47515
    },
    {
      "entropy": 5.787406396865845,
      "epoch": 5.336946144774527,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025492027118001345,
      "loss": 4.9296,
      "mean_token_accuracy": 0.2338593378663063,
      "num_tokens": 106072645.0,
      "step": 47520
    },
    {
      "entropy": 5.64267053604126,
      "epoch": 5.337507721682485,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002548802829798277,
      "loss": 4.8451,
      "mean_token_accuracy": 0.2417842447757721,
      "num_tokens": 106084115.0,
      "step": 47525
    },
    {
      "entropy": 5.668855762481689,
      "epoch": 5.338069298590442,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002548402954202616,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2377066820859909,
      "num_tokens": 106094093.0,
      "step": 47530
    },
    {
      "entropy": 5.720921564102173,
      "epoch": 5.338630875498399,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002548003085025883,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23707081973552704,
      "num_tokens": 106104452.0,
      "step": 47535
    },
    {
      "entropy": 5.7080176830291744,
      "epoch": 5.339192452406357,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025476032222808106,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2312936395406723,
      "num_tokens": 106114540.0,
      "step": 47540
    },
    {
      "entropy": 5.7398889541625975,
      "epoch": 5.339754029314315,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025472033659801297,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23394485116004943,
      "num_tokens": 106124981.0,
      "step": 47545
    },
    {
      "entropy": 5.705216360092163,
      "epoch": 5.340315606222272,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002546803516136572,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22563489824533461,
      "num_tokens": 106135037.0,
      "step": 47550
    },
    {
      "entropy": 5.770456314086914,
      "epoch": 5.34087718313023,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025464036727628695,
      "loss": 4.9814,
      "mean_token_accuracy": 0.23070569038391114,
      "num_tokens": 106145515.0,
      "step": 47555
    },
    {
      "entropy": 5.7377259731292725,
      "epoch": 5.341438760038187,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025460038358717527,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22993164509534836,
      "num_tokens": 106157246.0,
      "step": 47560
    },
    {
      "entropy": 5.734280681610107,
      "epoch": 5.342000336946144,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002545604005475953,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22916661500930785,
      "num_tokens": 106167889.0,
      "step": 47565
    },
    {
      "entropy": 5.661777305603027,
      "epoch": 5.3425619138541025,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002545204181588201,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2379866674542427,
      "num_tokens": 106178330.0,
      "step": 47570
    },
    {
      "entropy": 5.715252542495728,
      "epoch": 5.34312349076206,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025448043642212274,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2347126454114914,
      "num_tokens": 106188995.0,
      "step": 47575
    },
    {
      "entropy": 5.750619840621948,
      "epoch": 5.343685067670018,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002544404553387763,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23143173456192018,
      "num_tokens": 106199271.0,
      "step": 47580
    },
    {
      "entropy": 5.684042596817017,
      "epoch": 5.344246644577975,
      "grad_norm": 1.25,
      "learning_rate": 0.00025440047491005367,
      "loss": 4.9228,
      "mean_token_accuracy": 0.23526255935430526,
      "num_tokens": 106211495.0,
      "step": 47585
    },
    {
      "entropy": 5.673867702484131,
      "epoch": 5.344808221485932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025436049513722795,
      "loss": 4.863,
      "mean_token_accuracy": 0.2382035344839096,
      "num_tokens": 106223437.0,
      "step": 47590
    },
    {
      "entropy": 5.751700353622437,
      "epoch": 5.34536979839389,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025432051602157203,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22355954945087433,
      "num_tokens": 106233456.0,
      "step": 47595
    },
    {
      "entropy": 5.703354120254517,
      "epoch": 5.345931375301848,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002542805375643589,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2391288846731186,
      "num_tokens": 106244559.0,
      "step": 47600
    },
    {
      "entropy": 5.649822139739991,
      "epoch": 5.346492952209805,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002542405597668615,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23667985200881958,
      "num_tokens": 106254427.0,
      "step": 47605
    },
    {
      "entropy": 5.676586627960205,
      "epoch": 5.347054529117763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025420058263035276,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23624218106269837,
      "num_tokens": 106264616.0,
      "step": 47610
    },
    {
      "entropy": 5.707436037063599,
      "epoch": 5.34761610602572,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002541606061561055,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2418051078915596,
      "num_tokens": 106275103.0,
      "step": 47615
    },
    {
      "entropy": 5.777983570098877,
      "epoch": 5.348177682933677,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002541206303453927,
      "loss": 4.9904,
      "mean_token_accuracy": 0.23245765417814254,
      "num_tokens": 106286004.0,
      "step": 47620
    },
    {
      "entropy": 5.75753436088562,
      "epoch": 5.3487392598416355,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000254080655199487,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2374343231320381,
      "num_tokens": 106297161.0,
      "step": 47625
    },
    {
      "entropy": 5.74769024848938,
      "epoch": 5.349300836749593,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025404068071966133,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2245124861598015,
      "num_tokens": 106309330.0,
      "step": 47630
    },
    {
      "entropy": 5.751378870010376,
      "epoch": 5.349862413657551,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025400070690718865,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22665065228939058,
      "num_tokens": 106319801.0,
      "step": 47635
    },
    {
      "entropy": 5.723150777816772,
      "epoch": 5.350423990565508,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025396073376334155,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2262982651591301,
      "num_tokens": 106331328.0,
      "step": 47640
    },
    {
      "entropy": 5.734500551223755,
      "epoch": 5.350985567473465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025392076128939283,
      "loss": 4.9085,
      "mean_token_accuracy": 0.23595277965068817,
      "num_tokens": 106343740.0,
      "step": 47645
    },
    {
      "entropy": 5.696189594268799,
      "epoch": 5.351547144381423,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002538807894866153,
      "loss": 4.8497,
      "mean_token_accuracy": 0.24480486959218978,
      "num_tokens": 106354620.0,
      "step": 47650
    },
    {
      "entropy": 5.684376573562622,
      "epoch": 5.3521087212893805,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025384081835628163,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2357694238424301,
      "num_tokens": 106366247.0,
      "step": 47655
    },
    {
      "entropy": 5.741050291061401,
      "epoch": 5.352670298197338,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002538008478996645,
      "loss": 4.9504,
      "mean_token_accuracy": 0.23564403355121613,
      "num_tokens": 106377045.0,
      "step": 47660
    },
    {
      "entropy": 5.6562933921813965,
      "epoch": 5.353231875105296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002537608781180366,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2374303326010704,
      "num_tokens": 106389070.0,
      "step": 47665
    },
    {
      "entropy": 5.676293468475341,
      "epoch": 5.353793452013253,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002537209090126704,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23570389300584793,
      "num_tokens": 106401292.0,
      "step": 47670
    },
    {
      "entropy": 5.730082941055298,
      "epoch": 5.35435502892121,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025368094058483885,
      "loss": 4.8749,
      "mean_token_accuracy": 0.23601098507642745,
      "num_tokens": 106413045.0,
      "step": 47675
    },
    {
      "entropy": 5.719054317474365,
      "epoch": 5.354916605829168,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002536409728358145,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23968146741390228,
      "num_tokens": 106424655.0,
      "step": 47680
    },
    {
      "entropy": 5.762643957138062,
      "epoch": 5.355478182737126,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002536010057668698,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2220989152789116,
      "num_tokens": 106436544.0,
      "step": 47685
    },
    {
      "entropy": 5.716799116134643,
      "epoch": 5.356039759645084,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002535610393792773,
      "loss": 4.9087,
      "mean_token_accuracy": 0.23231343775987626,
      "num_tokens": 106446936.0,
      "step": 47690
    },
    {
      "entropy": 5.7718141078948975,
      "epoch": 5.356601336553041,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025352107367430975,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2329880863428116,
      "num_tokens": 106457328.0,
      "step": 47695
    },
    {
      "entropy": 5.703288602828979,
      "epoch": 5.357162913460998,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025348110865323946,
      "loss": 4.9328,
      "mean_token_accuracy": 0.23533077239990235,
      "num_tokens": 106468037.0,
      "step": 47700
    },
    {
      "entropy": 5.708434152603149,
      "epoch": 5.357724490368956,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002534411443173389,
      "loss": 4.9154,
      "mean_token_accuracy": 0.23614398539066314,
      "num_tokens": 106479406.0,
      "step": 47705
    },
    {
      "entropy": 5.591104125976562,
      "epoch": 5.3582860672769135,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002534011806678809,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22998458594083787,
      "num_tokens": 106491142.0,
      "step": 47710
    },
    {
      "entropy": 5.673634481430054,
      "epoch": 5.358847644184872,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002533612177061376,
      "loss": 4.92,
      "mean_token_accuracy": 0.23602895736694335,
      "num_tokens": 106502743.0,
      "step": 47715
    },
    {
      "entropy": 5.711542367935181,
      "epoch": 5.359409221092829,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025332125543338146,
      "loss": 4.9289,
      "mean_token_accuracy": 0.23778961598873138,
      "num_tokens": 106513765.0,
      "step": 47720
    },
    {
      "entropy": 5.659247732162475,
      "epoch": 5.359970798000786,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000253281293850885,
      "loss": 4.786,
      "mean_token_accuracy": 0.24379458129405976,
      "num_tokens": 106524154.0,
      "step": 47725
    },
    {
      "entropy": 5.698378944396973,
      "epoch": 5.360532374908744,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025324133295992056,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2351692572236061,
      "num_tokens": 106535946.0,
      "step": 47730
    },
    {
      "entropy": 5.77576174736023,
      "epoch": 5.361093951816701,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002532013727617605,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2300814002752304,
      "num_tokens": 106546352.0,
      "step": 47735
    },
    {
      "entropy": 5.724182605743408,
      "epoch": 5.361655528724659,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002531614132576772,
      "loss": 4.8333,
      "mean_token_accuracy": 0.24127038866281508,
      "num_tokens": 106557640.0,
      "step": 47740
    },
    {
      "entropy": 5.719255542755127,
      "epoch": 5.362217105632617,
      "grad_norm": 1.359375,
      "learning_rate": 0.000253121454448943,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22949947863817216,
      "num_tokens": 106567762.0,
      "step": 47745
    },
    {
      "entropy": 5.708317518234253,
      "epoch": 5.362778682540574,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025308149633683014,
      "loss": 4.8923,
      "mean_token_accuracy": 0.24470712542533873,
      "num_tokens": 106579589.0,
      "step": 47750
    },
    {
      "entropy": 5.701757287979126,
      "epoch": 5.363340259448531,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000253041538922611,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23402509987354278,
      "num_tokens": 106590122.0,
      "step": 47755
    },
    {
      "entropy": 5.782387924194336,
      "epoch": 5.363901836356489,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025300158220755776,
      "loss": 4.9704,
      "mean_token_accuracy": 0.23302506953477858,
      "num_tokens": 106601269.0,
      "step": 47760
    },
    {
      "entropy": 5.792614459991455,
      "epoch": 5.364463413264446,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025296162619294267,
      "loss": 5.0608,
      "mean_token_accuracy": 0.22506362497806548,
      "num_tokens": 106612462.0,
      "step": 47765
    },
    {
      "entropy": 5.741956996917724,
      "epoch": 5.3650249901724045,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002529216708800379,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2377462476491928,
      "num_tokens": 106623333.0,
      "step": 47770
    },
    {
      "entropy": 5.686005640029907,
      "epoch": 5.365586567080362,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025288171627011564,
      "loss": 4.8585,
      "mean_token_accuracy": 0.24078427404165267,
      "num_tokens": 106634505.0,
      "step": 47775
    },
    {
      "entropy": 5.777945137023925,
      "epoch": 5.366148143988319,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025284176236444817,
      "loss": 4.9847,
      "mean_token_accuracy": 0.23384389877319336,
      "num_tokens": 106645764.0,
      "step": 47780
    },
    {
      "entropy": 5.659556531906128,
      "epoch": 5.366709720896277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025280180916430764,
      "loss": 4.8533,
      "mean_token_accuracy": 0.23767328560352324,
      "num_tokens": 106657484.0,
      "step": 47785
    },
    {
      "entropy": 5.662928581237793,
      "epoch": 5.367271297804234,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025276185667096605,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22865820229053496,
      "num_tokens": 106669038.0,
      "step": 47790
    },
    {
      "entropy": 5.757769346237183,
      "epoch": 5.3678328747121915,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025272190488569564,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23814859390258789,
      "num_tokens": 106679447.0,
      "step": 47795
    },
    {
      "entropy": 5.737098169326782,
      "epoch": 5.36839445162015,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025268195380976835,
      "loss": 4.9314,
      "mean_token_accuracy": 0.2261977642774582,
      "num_tokens": 106689596.0,
      "step": 47800
    },
    {
      "entropy": 5.735416555404663,
      "epoch": 5.368956028528107,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002526420034444564,
      "loss": 4.9083,
      "mean_token_accuracy": 0.243391515314579,
      "num_tokens": 106701825.0,
      "step": 47805
    },
    {
      "entropy": 5.79840579032898,
      "epoch": 5.369517605436064,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025260205379103157,
      "loss": 5.0815,
      "mean_token_accuracy": 0.22399531602859496,
      "num_tokens": 106713451.0,
      "step": 47810
    },
    {
      "entropy": 5.746904373168945,
      "epoch": 5.370079182344022,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00025256210485076616,
      "loss": 5.012,
      "mean_token_accuracy": 0.22547446489334105,
      "num_tokens": 106724732.0,
      "step": 47815
    },
    {
      "entropy": 5.765582609176636,
      "epoch": 5.370640759251979,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000252522156624932,
      "loss": 4.9543,
      "mean_token_accuracy": 0.23957640826702117,
      "num_tokens": 106736533.0,
      "step": 47820
    },
    {
      "entropy": 5.680222749710083,
      "epoch": 5.3712023361599375,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025248220911480105,
      "loss": 4.8489,
      "mean_token_accuracy": 0.24016108512878417,
      "num_tokens": 106747454.0,
      "step": 47825
    },
    {
      "entropy": 5.643669462203979,
      "epoch": 5.371763913067895,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025244226232164533,
      "loss": 4.8588,
      "mean_token_accuracy": 0.24413769990205764,
      "num_tokens": 106757833.0,
      "step": 47830
    },
    {
      "entropy": 5.610504102706909,
      "epoch": 5.372325489975852,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025240231624673674,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23201136738061906,
      "num_tokens": 106768646.0,
      "step": 47835
    },
    {
      "entropy": 5.730157709121704,
      "epoch": 5.37288706688381,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002523623708913472,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23714499771595002,
      "num_tokens": 106780434.0,
      "step": 47840
    },
    {
      "entropy": 5.706421899795532,
      "epoch": 5.373448643791767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002523224262567485,
      "loss": 4.9285,
      "mean_token_accuracy": 0.23951153606176376,
      "num_tokens": 106792372.0,
      "step": 47845
    },
    {
      "entropy": 5.720449352264405,
      "epoch": 5.3740102206997244,
      "grad_norm": 1.4375,
      "learning_rate": 0.00025228248234421247,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22814246267080307,
      "num_tokens": 106802251.0,
      "step": 47850
    },
    {
      "entropy": 5.703924703598022,
      "epoch": 5.3745717976076826,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002522425391550111,
      "loss": 4.949,
      "mean_token_accuracy": 0.23544256538152694,
      "num_tokens": 106813818.0,
      "step": 47855
    },
    {
      "entropy": 5.7594523429870605,
      "epoch": 5.37513337451564,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025220259669041613,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22515311986207961,
      "num_tokens": 106824096.0,
      "step": 47860
    },
    {
      "entropy": 5.787222194671631,
      "epoch": 5.375694951423597,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002521626549516993,
      "loss": 5.0448,
      "mean_token_accuracy": 0.22576398700475692,
      "num_tokens": 106836037.0,
      "step": 47865
    },
    {
      "entropy": 5.829408168792725,
      "epoch": 5.376256528331555,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002521227139401324,
      "loss": 5.0223,
      "mean_token_accuracy": 0.23107749670743943,
      "num_tokens": 106846370.0,
      "step": 47870
    },
    {
      "entropy": 5.770985555648804,
      "epoch": 5.376818105239512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025208277365698715,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2266801342368126,
      "num_tokens": 106857945.0,
      "step": 47875
    },
    {
      "entropy": 5.721414232254029,
      "epoch": 5.37737968214747,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002520428341035353,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23641197681427,
      "num_tokens": 106867899.0,
      "step": 47880
    },
    {
      "entropy": 5.68513445854187,
      "epoch": 5.377941259055428,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002520028952810485,
      "loss": 4.9164,
      "mean_token_accuracy": 0.23056080490350722,
      "num_tokens": 106879305.0,
      "step": 47885
    },
    {
      "entropy": 5.749175214767456,
      "epoch": 5.378502835963385,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025196295719079835,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2319471389055252,
      "num_tokens": 106891944.0,
      "step": 47890
    },
    {
      "entropy": 5.6951581001281735,
      "epoch": 5.379064412871343,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002519230198340567,
      "loss": 4.9178,
      "mean_token_accuracy": 0.2362222909927368,
      "num_tokens": 106903420.0,
      "step": 47895
    },
    {
      "entropy": 5.725127267837524,
      "epoch": 5.3796259897793,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025188308321209504,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2345975011587143,
      "num_tokens": 106914838.0,
      "step": 47900
    },
    {
      "entropy": 5.620047760009766,
      "epoch": 5.380187566687258,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025184314732618496,
      "loss": 4.8231,
      "mean_token_accuracy": 0.24288619607686995,
      "num_tokens": 106926519.0,
      "step": 47905
    },
    {
      "entropy": 5.693844509124756,
      "epoch": 5.3807491435952155,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000251803212177598,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2340455859899521,
      "num_tokens": 106938303.0,
      "step": 47910
    },
    {
      "entropy": 5.703704071044922,
      "epoch": 5.381310720503173,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002517632777676058,
      "loss": 4.9444,
      "mean_token_accuracy": 0.23705532103776933,
      "num_tokens": 106950120.0,
      "step": 47915
    },
    {
      "entropy": 5.750429153442383,
      "epoch": 5.381872297411131,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025172334409747976,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22746725231409073,
      "num_tokens": 106960998.0,
      "step": 47920
    },
    {
      "entropy": 5.7049497127532955,
      "epoch": 5.382433874319088,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002516834111684915,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23530831038951874,
      "num_tokens": 106972640.0,
      "step": 47925
    },
    {
      "entropy": 5.720781564712524,
      "epoch": 5.382995451227045,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002516434789819126,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2332780361175537,
      "num_tokens": 106985300.0,
      "step": 47930
    },
    {
      "entropy": 5.728969669342041,
      "epoch": 5.383557028135003,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002516035475390143,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22399232238531114,
      "num_tokens": 106996074.0,
      "step": 47935
    },
    {
      "entropy": 5.747810125350952,
      "epoch": 5.384118605042961,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025156361684106814,
      "loss": 4.94,
      "mean_token_accuracy": 0.23601100742816924,
      "num_tokens": 107006493.0,
      "step": 47940
    },
    {
      "entropy": 5.695468187332153,
      "epoch": 5.384680181950918,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002515236868893455,
      "loss": 4.8752,
      "mean_token_accuracy": 0.230063097178936,
      "num_tokens": 107018739.0,
      "step": 47945
    },
    {
      "entropy": 5.638602781295776,
      "epoch": 5.385241758858876,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025148375768511776,
      "loss": 4.7907,
      "mean_token_accuracy": 0.24771356731653213,
      "num_tokens": 107028762.0,
      "step": 47950
    },
    {
      "entropy": 5.737870359420777,
      "epoch": 5.385803335766833,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002514438292296563,
      "loss": 4.9381,
      "mean_token_accuracy": 0.23012017458677292,
      "num_tokens": 107039609.0,
      "step": 47955
    },
    {
      "entropy": 5.634198570251465,
      "epoch": 5.386364912674791,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025140390152423247,
      "loss": 4.8381,
      "mean_token_accuracy": 0.24350180774927138,
      "num_tokens": 107051871.0,
      "step": 47960
    },
    {
      "entropy": 5.688655376434326,
      "epoch": 5.3869264895827484,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002513639745701176,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2331802576780319,
      "num_tokens": 107063560.0,
      "step": 47965
    },
    {
      "entropy": 5.732557344436645,
      "epoch": 5.387488066490706,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025132404836858294,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2365995854139328,
      "num_tokens": 107074466.0,
      "step": 47970
    },
    {
      "entropy": 5.73053240776062,
      "epoch": 5.388049643398664,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002512841229208997,
      "loss": 4.9195,
      "mean_token_accuracy": 0.24256702661514282,
      "num_tokens": 107085470.0,
      "step": 47975
    },
    {
      "entropy": 5.7322532653808596,
      "epoch": 5.388611220306621,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002512441982283392,
      "loss": 4.9817,
      "mean_token_accuracy": 0.23579345643520355,
      "num_tokens": 107098179.0,
      "step": 47980
    },
    {
      "entropy": 5.670466756820678,
      "epoch": 5.389172797214578,
      "grad_norm": 1.25,
      "learning_rate": 0.0002512042742921727,
      "loss": 4.912,
      "mean_token_accuracy": 0.23699742555618286,
      "num_tokens": 107108941.0,
      "step": 47985
    },
    {
      "entropy": 5.725223064422607,
      "epoch": 5.389734374122536,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002511643511136712,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2311322271823883,
      "num_tokens": 107120019.0,
      "step": 47990
    },
    {
      "entropy": 5.755124902725219,
      "epoch": 5.3902959510304935,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025112442869410607,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2268099695444107,
      "num_tokens": 107130254.0,
      "step": 47995
    },
    {
      "entropy": 5.67406702041626,
      "epoch": 5.390857527938451,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025108450703474844,
      "loss": 4.8197,
      "mean_token_accuracy": 0.2455747604370117,
      "num_tokens": 107140625.0,
      "step": 48000
    },
    {
      "epoch": 5.390857527938451,
      "eval_entropy": 5.415910740503799,
      "eval_loss": 5.045635223388672,
      "eval_mean_token_accuracy": 0.23461757304551892,
      "eval_num_tokens": 107140625.0,
      "eval_runtime": 18.4859,
      "eval_samples_per_second": 1677.654,
      "eval_steps_per_second": 209.727,
      "step": 48000
    },
    {
      "entropy": 5.707392454147339,
      "epoch": 5.391419104846409,
      "grad_norm": 1.453125,
      "learning_rate": 0.00025104458613686936,
      "loss": 4.968,
      "mean_token_accuracy": 0.2259875774383545,
      "num_tokens": 107150318.0,
      "step": 48005
    },
    {
      "entropy": 5.6435800075531,
      "epoch": 5.391980681754366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00025100466600174,
      "loss": 4.8995,
      "mean_token_accuracy": 0.23202821165323256,
      "num_tokens": 107162831.0,
      "step": 48010
    },
    {
      "entropy": 5.775879716873169,
      "epoch": 5.392542258662324,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002509647466306313,
      "loss": 4.9438,
      "mean_token_accuracy": 0.23699920922517775,
      "num_tokens": 107174775.0,
      "step": 48015
    },
    {
      "entropy": 5.769490098953247,
      "epoch": 5.393103835570281,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025092482802481443,
      "loss": 4.9915,
      "mean_token_accuracy": 0.22403928488492966,
      "num_tokens": 107185644.0,
      "step": 48020
    },
    {
      "entropy": 5.721153783798218,
      "epoch": 5.393665412478239,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002508849101855604,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23660936057567597,
      "num_tokens": 107196669.0,
      "step": 48025
    },
    {
      "entropy": 5.754781007766724,
      "epoch": 5.394226989386197,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002508449931141401,
      "loss": 4.9162,
      "mean_token_accuracy": 0.24142121970653535,
      "num_tokens": 107207625.0,
      "step": 48030
    },
    {
      "entropy": 5.760581731796265,
      "epoch": 5.394788566294154,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025080507681182464,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2295905113220215,
      "num_tokens": 107218925.0,
      "step": 48035
    },
    {
      "entropy": 5.738750839233399,
      "epoch": 5.395350143202111,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002507651612798849,
      "loss": 4.9332,
      "mean_token_accuracy": 0.23790261298418044,
      "num_tokens": 107231669.0,
      "step": 48040
    },
    {
      "entropy": 5.690502071380616,
      "epoch": 5.395911720110069,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025072524651959183,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2322612300515175,
      "num_tokens": 107242922.0,
      "step": 48045
    },
    {
      "entropy": 5.722575378417969,
      "epoch": 5.3964732970180265,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002506853325322164,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22555407583713533,
      "num_tokens": 107254713.0,
      "step": 48050
    },
    {
      "entropy": 5.747110748291016,
      "epoch": 5.397034873925985,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002506454193190293,
      "loss": 4.9424,
      "mean_token_accuracy": 0.2333752766251564,
      "num_tokens": 107266202.0,
      "step": 48055
    },
    {
      "entropy": 5.712117719650268,
      "epoch": 5.397596450833942,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025060550688130157,
      "loss": 4.9189,
      "mean_token_accuracy": 0.23440591543912886,
      "num_tokens": 107277914.0,
      "step": 48060
    },
    {
      "entropy": 5.749688959121704,
      "epoch": 5.398158027741899,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025056559522030383,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22845654040575028,
      "num_tokens": 107289592.0,
      "step": 48065
    },
    {
      "entropy": 5.616835260391236,
      "epoch": 5.398719604649857,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002505256843373071,
      "loss": 4.8277,
      "mean_token_accuracy": 0.24173953980207444,
      "num_tokens": 107301428.0,
      "step": 48070
    },
    {
      "entropy": 5.658674526214599,
      "epoch": 5.399281181557814,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002504857742335821,
      "loss": 4.9226,
      "mean_token_accuracy": 0.23472503423690796,
      "num_tokens": 107313651.0,
      "step": 48075
    },
    {
      "entropy": 5.748832082748413,
      "epoch": 5.3998427584657716,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00025044586491039956,
      "loss": 4.955,
      "mean_token_accuracy": 0.22806933522224426,
      "num_tokens": 107324375.0,
      "step": 48080
    },
    {
      "entropy": 5.723131322860718,
      "epoch": 5.40040433537373,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002504059563690302,
      "loss": 4.899,
      "mean_token_accuracy": 0.23916201442480087,
      "num_tokens": 107335445.0,
      "step": 48085
    },
    {
      "entropy": 5.700506687164307,
      "epoch": 5.400965912281687,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002503660486107447,
      "loss": 4.8727,
      "mean_token_accuracy": 0.2393421471118927,
      "num_tokens": 107346384.0,
      "step": 48090
    },
    {
      "entropy": 5.682506561279297,
      "epoch": 5.401527489189645,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002503261416368138,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22903032302856446,
      "num_tokens": 107356188.0,
      "step": 48095
    },
    {
      "entropy": 5.657235765457154,
      "epoch": 5.402089066097602,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002502862354485081,
      "loss": 4.818,
      "mean_token_accuracy": 0.24359550029039384,
      "num_tokens": 107366370.0,
      "step": 48100
    },
    {
      "entropy": 5.687361383438111,
      "epoch": 5.402650643005559,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025024633004709824,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22864104956388473,
      "num_tokens": 107376975.0,
      "step": 48105
    },
    {
      "entropy": 5.732682704925537,
      "epoch": 5.4032122199135175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025020642543385486,
      "loss": 4.995,
      "mean_token_accuracy": 0.2251165345311165,
      "num_tokens": 107388980.0,
      "step": 48110
    },
    {
      "entropy": 5.775548648834229,
      "epoch": 5.403773796821475,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00025016652161004847,
      "loss": 4.9734,
      "mean_token_accuracy": 0.23247991651296615,
      "num_tokens": 107399299.0,
      "step": 48115
    },
    {
      "entropy": 5.708317470550537,
      "epoch": 5.404335373729432,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002501266185769497,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22831321358680726,
      "num_tokens": 107410438.0,
      "step": 48120
    },
    {
      "entropy": 5.722907447814942,
      "epoch": 5.40489695063739,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025008671633582904,
      "loss": 4.9451,
      "mean_token_accuracy": 0.23779220134019852,
      "num_tokens": 107422597.0,
      "step": 48125
    },
    {
      "entropy": 5.731387329101563,
      "epoch": 5.405458527545347,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000250046814887957,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23355186879634857,
      "num_tokens": 107432762.0,
      "step": 48130
    },
    {
      "entropy": 5.703008508682251,
      "epoch": 5.4060201044533045,
      "grad_norm": 1.25,
      "learning_rate": 0.0002500069142346039,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23120340406894685,
      "num_tokens": 107443644.0,
      "step": 48135
    },
    {
      "entropy": 5.748010969161987,
      "epoch": 5.406581681361263,
      "grad_norm": 1.25,
      "learning_rate": 0.0002499670143770405,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22502927482128143,
      "num_tokens": 107454752.0,
      "step": 48140
    },
    {
      "entropy": 5.677725172042846,
      "epoch": 5.40714325826922,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002499271153165371,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22801931202411652,
      "num_tokens": 107464829.0,
      "step": 48145
    },
    {
      "entropy": 5.799708032608033,
      "epoch": 5.407704835177178,
      "grad_norm": 1.34375,
      "learning_rate": 0.000249887217054364,
      "loss": 4.9985,
      "mean_token_accuracy": 0.22896209359169006,
      "num_tokens": 107475214.0,
      "step": 48150
    },
    {
      "entropy": 5.802849388122558,
      "epoch": 5.408266412085135,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002498473195917917,
      "loss": 5.0312,
      "mean_token_accuracy": 0.22570584267377852,
      "num_tokens": 107486693.0,
      "step": 48155
    },
    {
      "entropy": 5.777779769897461,
      "epoch": 5.408827988993092,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002498074229300905,
      "loss": 4.916,
      "mean_token_accuracy": 0.23150861114263535,
      "num_tokens": 107498347.0,
      "step": 48160
    },
    {
      "entropy": 5.77620120048523,
      "epoch": 5.4093895659010505,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002497675270705307,
      "loss": 4.9723,
      "mean_token_accuracy": 0.22873859703540803,
      "num_tokens": 107507525.0,
      "step": 48165
    },
    {
      "entropy": 5.7008686542510985,
      "epoch": 5.409951142809008,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002497276320143826,
      "loss": 4.908,
      "mean_token_accuracy": 0.2355155572295189,
      "num_tokens": 107519196.0,
      "step": 48170
    },
    {
      "entropy": 5.756192588806153,
      "epoch": 5.410512719716965,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024968773776291647,
      "loss": 4.9777,
      "mean_token_accuracy": 0.23389951437711715,
      "num_tokens": 107532740.0,
      "step": 48175
    },
    {
      "entropy": 5.679021167755127,
      "epoch": 5.411074296624923,
      "grad_norm": 1.125,
      "learning_rate": 0.0002496478443174026,
      "loss": 4.8462,
      "mean_token_accuracy": 0.23695618957281112,
      "num_tokens": 107544098.0,
      "step": 48180
    },
    {
      "entropy": 5.8006500720977785,
      "epoch": 5.41163587353288,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002496079516791113,
      "loss": 5.0192,
      "mean_token_accuracy": 0.23670049607753754,
      "num_tokens": 107555560.0,
      "step": 48185
    },
    {
      "entropy": 5.66933126449585,
      "epoch": 5.412197450440837,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002495680598493126,
      "loss": 4.844,
      "mean_token_accuracy": 0.2377375215291977,
      "num_tokens": 107565761.0,
      "step": 48190
    },
    {
      "entropy": 5.721105241775513,
      "epoch": 5.4127590273487955,
      "grad_norm": 1.25,
      "learning_rate": 0.0002495281688292768,
      "loss": 4.9599,
      "mean_token_accuracy": 0.23345747888088225,
      "num_tokens": 107576742.0,
      "step": 48195
    },
    {
      "entropy": 5.744028472900391,
      "epoch": 5.413320604256753,
      "grad_norm": 1.25,
      "learning_rate": 0.00024948827862027385,
      "loss": 4.8945,
      "mean_token_accuracy": 0.24080401957035064,
      "num_tokens": 107588810.0,
      "step": 48200
    },
    {
      "entropy": 5.734835386276245,
      "epoch": 5.413882181164711,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002494483892235741,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23237758874893188,
      "num_tokens": 107599392.0,
      "step": 48205
    },
    {
      "entropy": 5.7681702136993405,
      "epoch": 5.414443758072668,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002494085006404474,
      "loss": 4.9033,
      "mean_token_accuracy": 0.24120962470769883,
      "num_tokens": 107608288.0,
      "step": 48210
    },
    {
      "entropy": 5.71226806640625,
      "epoch": 5.415005334980625,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002493686128721641,
      "loss": 4.99,
      "mean_token_accuracy": 0.23165810257196426,
      "num_tokens": 107620747.0,
      "step": 48215
    },
    {
      "entropy": 5.849596738815308,
      "epoch": 5.415566911888583,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00024932872591999414,
      "loss": 4.9882,
      "mean_token_accuracy": 0.23062735050916672,
      "num_tokens": 107632754.0,
      "step": 48220
    },
    {
      "entropy": 5.652824401855469,
      "epoch": 5.416128488796541,
      "grad_norm": 1.203125,
      "learning_rate": 0.00024928883978520743,
      "loss": 4.825,
      "mean_token_accuracy": 0.2363985225558281,
      "num_tokens": 107643326.0,
      "step": 48225
    },
    {
      "entropy": 5.718243646621704,
      "epoch": 5.416690065704498,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024924895446907407,
      "loss": 4.9007,
      "mean_token_accuracy": 0.23218321204185485,
      "num_tokens": 107654474.0,
      "step": 48230
    },
    {
      "entropy": 5.6511438369750975,
      "epoch": 5.417251642612456,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000249209069972864,
      "loss": 4.8213,
      "mean_token_accuracy": 0.23941780179738997,
      "num_tokens": 107665110.0,
      "step": 48235
    },
    {
      "entropy": 5.758889198303223,
      "epoch": 5.417813219520413,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00024916918629784714,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2263076826930046,
      "num_tokens": 107675290.0,
      "step": 48240
    },
    {
      "entropy": 5.705904006958008,
      "epoch": 5.418374796428371,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002491293034452933,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23437079936265945,
      "num_tokens": 107685202.0,
      "step": 48245
    },
    {
      "entropy": 5.674900007247925,
      "epoch": 5.4189363733363285,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002490894214164726,
      "loss": 4.981,
      "mean_token_accuracy": 0.22569597661495208,
      "num_tokens": 107695459.0,
      "step": 48250
    },
    {
      "entropy": 5.690354108810425,
      "epoch": 5.419497950244286,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002490495402126548,
      "loss": 4.8952,
      "mean_token_accuracy": 0.23932674825191497,
      "num_tokens": 107706776.0,
      "step": 48255
    },
    {
      "entropy": 5.781277084350586,
      "epoch": 5.420059527152244,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002490096598351096,
      "loss": 4.88,
      "mean_token_accuracy": 0.24142919927835466,
      "num_tokens": 107717969.0,
      "step": 48260
    },
    {
      "entropy": 5.7447985172271725,
      "epoch": 5.420621104060201,
      "grad_norm": 1.328125,
      "learning_rate": 0.00024896978028510696,
      "loss": 4.976,
      "mean_token_accuracy": 0.2284802943468094,
      "num_tokens": 107727812.0,
      "step": 48265
    },
    {
      "entropy": 5.746131181716919,
      "epoch": 5.421182680968158,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002489299015639167,
      "loss": 4.963,
      "mean_token_accuracy": 0.2334570750594139,
      "num_tokens": 107737823.0,
      "step": 48270
    },
    {
      "entropy": 5.740817594528198,
      "epoch": 5.421744257876116,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002488900236728083,
      "loss": 4.9321,
      "mean_token_accuracy": 0.242582468688488,
      "num_tokens": 107747835.0,
      "step": 48275
    },
    {
      "entropy": 5.683602666854858,
      "epoch": 5.422305834784074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002488501466130517,
      "loss": 4.8711,
      "mean_token_accuracy": 0.24302262514829637,
      "num_tokens": 107759167.0,
      "step": 48280
    },
    {
      "entropy": 5.695745420455933,
      "epoch": 5.422867411692032,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00024881027038591664,
      "loss": 4.8991,
      "mean_token_accuracy": 0.24117893874645233,
      "num_tokens": 107770130.0,
      "step": 48285
    },
    {
      "entropy": 5.71949028968811,
      "epoch": 5.423428988599989,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024877039499267273,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22910294979810714,
      "num_tokens": 107780923.0,
      "step": 48290
    },
    {
      "entropy": 5.641154718399048,
      "epoch": 5.423990565507946,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002487305204345896,
      "loss": 4.8835,
      "mean_token_accuracy": 0.23750228732824324,
      "num_tokens": 107792417.0,
      "step": 48295
    },
    {
      "entropy": 5.79011116027832,
      "epoch": 5.424552142415904,
      "grad_norm": 1.296875,
      "learning_rate": 0.00024869064671293687,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22980955839157105,
      "num_tokens": 107802124.0,
      "step": 48300
    },
    {
      "entropy": 5.760301971435547,
      "epoch": 5.425113719323861,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024865077382898423,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23773365765810012,
      "num_tokens": 107814025.0,
      "step": 48305
    },
    {
      "entropy": 5.751373720169068,
      "epoch": 5.425675296231819,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00024861090178400106,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22753774672746657,
      "num_tokens": 107825271.0,
      "step": 48310
    },
    {
      "entropy": 5.715556859970093,
      "epoch": 5.426236873139777,
      "grad_norm": 1.21875,
      "learning_rate": 0.00024857103057925707,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23738120198249818,
      "num_tokens": 107836419.0,
      "step": 48315
    },
    {
      "entropy": 5.729059267044067,
      "epoch": 5.426798450047734,
      "grad_norm": 1.203125,
      "learning_rate": 0.00024853116021602173,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22907237708568573,
      "num_tokens": 107847700.0,
      "step": 48320
    },
    {
      "entropy": 5.686341285705566,
      "epoch": 5.427360026955691,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002484912906955645,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23404186218976974,
      "num_tokens": 107858870.0,
      "step": 48325
    },
    {
      "entropy": 5.8539941787719725,
      "epoch": 5.427921603863649,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002484514220191549,
      "loss": 5.1842,
      "mean_token_accuracy": 0.2182514876127243,
      "num_tokens": 107870414.0,
      "step": 48330
    },
    {
      "entropy": 5.745848178863525,
      "epoch": 5.4284831807716065,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002484115541880623,
      "loss": 5.0029,
      "mean_token_accuracy": 0.22322814911603928,
      "num_tokens": 107880798.0,
      "step": 48335
    },
    {
      "entropy": 5.666528940200806,
      "epoch": 5.429044757679565,
      "grad_norm": 1.515625,
      "learning_rate": 0.00024837168720355605,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2490301713347435,
      "num_tokens": 107891332.0,
      "step": 48340
    },
    {
      "entropy": 5.7056801319122314,
      "epoch": 5.429606334587522,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002483318210669057,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23908441066741942,
      "num_tokens": 107902579.0,
      "step": 48345
    },
    {
      "entropy": 5.695467376708985,
      "epoch": 5.430167911495479,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002482919557793805,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22278932183980943,
      "num_tokens": 107912921.0,
      "step": 48350
    },
    {
      "entropy": 5.698290729522705,
      "epoch": 5.430729488403437,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024825209134224965,
      "loss": 4.809,
      "mean_token_accuracy": 0.24034191519021988,
      "num_tokens": 107923767.0,
      "step": 48355
    },
    {
      "entropy": 5.723067903518677,
      "epoch": 5.431291065311394,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002482122277567827,
      "loss": 4.9741,
      "mean_token_accuracy": 0.22363787591457368,
      "num_tokens": 107935905.0,
      "step": 48360
    },
    {
      "entropy": 5.544680547714234,
      "epoch": 5.431852642219352,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024817236502424885,
      "loss": 4.7699,
      "mean_token_accuracy": 0.26325169056653974,
      "num_tokens": 107947235.0,
      "step": 48365
    },
    {
      "entropy": 5.72051591873169,
      "epoch": 5.43241421912731,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002481325031459172,
      "loss": 4.8485,
      "mean_token_accuracy": 0.23437022566795349,
      "num_tokens": 107959323.0,
      "step": 48370
    },
    {
      "entropy": 5.73471360206604,
      "epoch": 5.432975796035267,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024809264212305714,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22934456169605255,
      "num_tokens": 107970790.0,
      "step": 48375
    },
    {
      "entropy": 5.70734281539917,
      "epoch": 5.433537372943224,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002480527819569378,
      "loss": 4.9378,
      "mean_token_accuracy": 0.23570555001497268,
      "num_tokens": 107982122.0,
      "step": 48380
    },
    {
      "entropy": 5.713508462905883,
      "epoch": 5.434098949851182,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002480129226488283,
      "loss": 4.8995,
      "mean_token_accuracy": 0.24135760068893433,
      "num_tokens": 107993093.0,
      "step": 48385
    },
    {
      "entropy": 5.674458169937134,
      "epoch": 5.4346605267591395,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00024797306419999777,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23344823867082595,
      "num_tokens": 108004805.0,
      "step": 48390
    },
    {
      "entropy": 5.7550726413726805,
      "epoch": 5.435222103667098,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002479332066117154,
      "loss": 4.9367,
      "mean_token_accuracy": 0.23119449466466904,
      "num_tokens": 108016515.0,
      "step": 48395
    },
    {
      "entropy": 5.620275402069092,
      "epoch": 5.435783680575055,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002478933498852503,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23814865648746492,
      "num_tokens": 108027439.0,
      "step": 48400
    },
    {
      "entropy": 5.80205225944519,
      "epoch": 5.436345257483012,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002478534940218714,
      "loss": 4.9815,
      "mean_token_accuracy": 0.23481914401054382,
      "num_tokens": 108038373.0,
      "step": 48405
    },
    {
      "entropy": 5.771289205551147,
      "epoch": 5.43690683439097,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00024781363902284776,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2325449228286743,
      "num_tokens": 108048336.0,
      "step": 48410
    },
    {
      "entropy": 5.755316257476807,
      "epoch": 5.437468411298927,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002477737848894484,
      "loss": 4.9931,
      "mean_token_accuracy": 0.23738003969192506,
      "num_tokens": 108059189.0,
      "step": 48415
    },
    {
      "entropy": 5.699724817276001,
      "epoch": 5.4380299882068845,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002477339316229423,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2321525424718857,
      "num_tokens": 108071579.0,
      "step": 48420
    },
    {
      "entropy": 5.765399599075318,
      "epoch": 5.438591565114843,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002476940792245983,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2268492728471756,
      "num_tokens": 108082709.0,
      "step": 48425
    },
    {
      "entropy": 5.6923235893249515,
      "epoch": 5.4391531420228,
      "grad_norm": 1.234375,
      "learning_rate": 0.00024765422769568553,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23772634863853453,
      "num_tokens": 108094139.0,
      "step": 48430
    },
    {
      "entropy": 5.702487134933472,
      "epoch": 5.439714718930758,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002476143770374727,
      "loss": 4.8722,
      "mean_token_accuracy": 0.24174801558256148,
      "num_tokens": 108105352.0,
      "step": 48435
    },
    {
      "entropy": 5.647349786758423,
      "epoch": 5.440276295838715,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024757452725122876,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23365949392318724,
      "num_tokens": 108118313.0,
      "step": 48440
    },
    {
      "entropy": 5.715411758422851,
      "epoch": 5.440837872746672,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002475346783382225,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2339803859591484,
      "num_tokens": 108129002.0,
      "step": 48445
    },
    {
      "entropy": 5.753759574890137,
      "epoch": 5.4413994496546305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024749483029972276,
      "loss": 5.0038,
      "mean_token_accuracy": 0.22520774900913237,
      "num_tokens": 108141071.0,
      "step": 48450
    },
    {
      "entropy": 5.771226215362549,
      "epoch": 5.441961026562588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002474549831369983,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2334357276558876,
      "num_tokens": 108152923.0,
      "step": 48455
    },
    {
      "entropy": 5.749312114715576,
      "epoch": 5.442522603470545,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00024741513685131773,
      "loss": 4.9497,
      "mean_token_accuracy": 0.23042097836732864,
      "num_tokens": 108164645.0,
      "step": 48460
    },
    {
      "entropy": 5.674823760986328,
      "epoch": 5.443084180378503,
      "grad_norm": 1.359375,
      "learning_rate": 0.00024737529144395,
      "loss": 4.817,
      "mean_token_accuracy": 0.23793592303991318,
      "num_tokens": 108174547.0,
      "step": 48465
    },
    {
      "entropy": 5.730569839477539,
      "epoch": 5.44364575728646,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002473354469161637,
      "loss": 4.9215,
      "mean_token_accuracy": 0.23646779954433442,
      "num_tokens": 108186322.0,
      "step": 48470
    },
    {
      "entropy": 5.689666318893432,
      "epoch": 5.444207334194418,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002472956032692275,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23942870348691941,
      "num_tokens": 108196771.0,
      "step": 48475
    },
    {
      "entropy": 5.73966007232666,
      "epoch": 5.444768911102376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002472557605044101,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22975764125585557,
      "num_tokens": 108208222.0,
      "step": 48480
    },
    {
      "entropy": 5.636793708801269,
      "epoch": 5.445330488010333,
      "grad_norm": 1.25,
      "learning_rate": 0.00024721591862297994,
      "loss": 4.8199,
      "mean_token_accuracy": 0.25453889816999437,
      "num_tokens": 108218572.0,
      "step": 48485
    },
    {
      "entropy": 5.726640081405639,
      "epoch": 5.445892064918291,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00024717607762620575,
      "loss": 4.954,
      "mean_token_accuracy": 0.2331942304968834,
      "num_tokens": 108230458.0,
      "step": 48490
    },
    {
      "entropy": 5.78500337600708,
      "epoch": 5.446453641826248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002471362375153559,
      "loss": 5.0097,
      "mean_token_accuracy": 0.22748877257108688,
      "num_tokens": 108243840.0,
      "step": 48495
    },
    {
      "entropy": 5.73282527923584,
      "epoch": 5.447015218734205,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024709639829169924,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22547635287046433,
      "num_tokens": 108255390.0,
      "step": 48500
    },
    {
      "entropy": 5.63403639793396,
      "epoch": 5.4475767956421635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024705655995650407,
      "loss": 4.8461,
      "mean_token_accuracy": 0.24345284402370454,
      "num_tokens": 108267166.0,
      "step": 48505
    },
    {
      "entropy": 5.7366396427154545,
      "epoch": 5.448138372550121,
      "grad_norm": 1.25,
      "learning_rate": 0.0002470167225110388,
      "loss": 4.941,
      "mean_token_accuracy": 0.23460599184036254,
      "num_tokens": 108278353.0,
      "step": 48510
    },
    {
      "entropy": 5.751259088516235,
      "epoch": 5.448699949458078,
      "grad_norm": 1.296875,
      "learning_rate": 0.000246976885956572,
      "loss": 4.9607,
      "mean_token_accuracy": 0.22968632429838182,
      "num_tokens": 108290239.0,
      "step": 48515
    },
    {
      "entropy": 5.657074928283691,
      "epoch": 5.449261526366036,
      "grad_norm": 1.1875,
      "learning_rate": 0.000246937050294372,
      "loss": 4.8704,
      "mean_token_accuracy": 0.24270341843366622,
      "num_tokens": 108301866.0,
      "step": 48520
    },
    {
      "entropy": 5.65687403678894,
      "epoch": 5.449823103273993,
      "grad_norm": 1.21875,
      "learning_rate": 0.00024689721552570714,
      "loss": 4.9077,
      "mean_token_accuracy": 0.23399832099676132,
      "num_tokens": 108314514.0,
      "step": 48525
    },
    {
      "entropy": 5.778806543350219,
      "epoch": 5.450384680181951,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002468573816518459,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2384499952197075,
      "num_tokens": 108325018.0,
      "step": 48530
    },
    {
      "entropy": 5.682415342330932,
      "epoch": 5.4509462570899085,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002468175486740565,
      "loss": 4.8606,
      "mean_token_accuracy": 0.24540432244539262,
      "num_tokens": 108336195.0,
      "step": 48535
    },
    {
      "entropy": 5.644238615036011,
      "epoch": 5.451507833997866,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002467777165936073,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23559585213661194,
      "num_tokens": 108348169.0,
      "step": 48540
    },
    {
      "entropy": 5.688355112075806,
      "epoch": 5.452069410905824,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002467378854117666,
      "loss": 4.9277,
      "mean_token_accuracy": 0.2320377543568611,
      "num_tokens": 108359168.0,
      "step": 48545
    },
    {
      "entropy": 5.718879270553589,
      "epoch": 5.452630987813781,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002466980551298026,
      "loss": 4.9172,
      "mean_token_accuracy": 0.23563073724508285,
      "num_tokens": 108370615.0,
      "step": 48550
    },
    {
      "entropy": 5.768033504486084,
      "epoch": 5.453192564721738,
      "grad_norm": 1.359375,
      "learning_rate": 0.00024665822574898344,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23071707785129547,
      "num_tokens": 108380487.0,
      "step": 48555
    },
    {
      "entropy": 5.703965902328491,
      "epoch": 5.453754141629696,
      "grad_norm": 1.265625,
      "learning_rate": 0.00024661839727057737,
      "loss": 4.8294,
      "mean_token_accuracy": 0.2399798169732094,
      "num_tokens": 108392221.0,
      "step": 48560
    },
    {
      "entropy": 5.798436307907105,
      "epoch": 5.454315718537654,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024657856969585257,
      "loss": 4.928,
      "mean_token_accuracy": 0.23235781937837602,
      "num_tokens": 108404524.0,
      "step": 48565
    },
    {
      "entropy": 5.759130620956421,
      "epoch": 5.454877295445611,
      "grad_norm": 1.21875,
      "learning_rate": 0.000246538743026077,
      "loss": 4.9963,
      "mean_token_accuracy": 0.22232633084058762,
      "num_tokens": 108416088.0,
      "step": 48570
    },
    {
      "entropy": 5.760550880432129,
      "epoch": 5.455438872353569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000246498917262519,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22920591831207277,
      "num_tokens": 108428413.0,
      "step": 48575
    },
    {
      "entropy": 5.716575765609742,
      "epoch": 5.456000449261526,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002464590924064465,
      "loss": 4.8676,
      "mean_token_accuracy": 0.2399131253361702,
      "num_tokens": 108439505.0,
      "step": 48580
    },
    {
      "entropy": 5.696085405349732,
      "epoch": 5.456562026169484,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024641926845912765,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22550782561302185,
      "num_tokens": 108452849.0,
      "step": 48585
    },
    {
      "entropy": 5.69395170211792,
      "epoch": 5.4571236030774415,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00024637944542183035,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23569288700819016,
      "num_tokens": 108463859.0,
      "step": 48590
    },
    {
      "entropy": 5.687915992736817,
      "epoch": 5.457685179985399,
      "grad_norm": 1.3125,
      "learning_rate": 0.00024633962329582255,
      "loss": 4.783,
      "mean_token_accuracy": 0.2398453876376152,
      "num_tokens": 108474972.0,
      "step": 48595
    },
    {
      "entropy": 5.684831333160401,
      "epoch": 5.458246756893357,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002462998020823723,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2365328699350357,
      "num_tokens": 108485490.0,
      "step": 48600
    },
    {
      "entropy": 5.726432037353516,
      "epoch": 5.458808333801314,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024625998178274747,
      "loss": 4.9637,
      "mean_token_accuracy": 0.23069851845502853,
      "num_tokens": 108496460.0,
      "step": 48605
    },
    {
      "entropy": 5.706678199768066,
      "epoch": 5.459369910709271,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002462201623982159,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23629885613918306,
      "num_tokens": 108508269.0,
      "step": 48610
    },
    {
      "entropy": 5.751873397827149,
      "epoch": 5.459931487617229,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002461803439300456,
      "loss": 4.9407,
      "mean_token_accuracy": 0.23062913864850998,
      "num_tokens": 108519035.0,
      "step": 48615
    },
    {
      "entropy": 5.66952486038208,
      "epoch": 5.460493064525187,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002461405263795044,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23553071767091752,
      "num_tokens": 108529646.0,
      "step": 48620
    },
    {
      "entropy": 5.771422052383423,
      "epoch": 5.461054641433145,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002461007097478599,
      "loss": 5.0274,
      "mean_token_accuracy": 0.228683765232563,
      "num_tokens": 108540888.0,
      "step": 48625
    },
    {
      "entropy": 5.585369396209717,
      "epoch": 5.461616218341102,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002460608940363801,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24295783489942552,
      "num_tokens": 108552933.0,
      "step": 48630
    },
    {
      "entropy": 5.6865016460418705,
      "epoch": 5.462177795249059,
      "grad_norm": 1.5,
      "learning_rate": 0.00024602107924633266,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23379723131656646,
      "num_tokens": 108564213.0,
      "step": 48635
    },
    {
      "entropy": 5.74531102180481,
      "epoch": 5.462739372157017,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002459812653789851,
      "loss": 4.9715,
      "mean_token_accuracy": 0.23658567816019058,
      "num_tokens": 108575184.0,
      "step": 48640
    },
    {
      "entropy": 5.667637014389038,
      "epoch": 5.463300949064974,
      "grad_norm": 1.25,
      "learning_rate": 0.0002459414524356055,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23412189334630967,
      "num_tokens": 108585226.0,
      "step": 48645
    },
    {
      "entropy": 5.651174688339234,
      "epoch": 5.463862525972932,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002459016404174613,
      "loss": 4.7662,
      "mean_token_accuracy": 0.2524492651224136,
      "num_tokens": 108595758.0,
      "step": 48650
    },
    {
      "entropy": 5.776463270187378,
      "epoch": 5.46442410288089,
      "grad_norm": 1.296875,
      "learning_rate": 0.00024586182932582014,
      "loss": 4.8771,
      "mean_token_accuracy": 0.23556325286626817,
      "num_tokens": 108606866.0,
      "step": 48655
    },
    {
      "entropy": 5.6865715980529785,
      "epoch": 5.464985679788847,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002458220191619496,
      "loss": 4.9341,
      "mean_token_accuracy": 0.23495759218931198,
      "num_tokens": 108618307.0,
      "step": 48660
    },
    {
      "entropy": 5.7285308837890625,
      "epoch": 5.465547256696805,
      "grad_norm": 1.328125,
      "learning_rate": 0.00024578220992711736,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23687992990016937,
      "num_tokens": 108628669.0,
      "step": 48665
    },
    {
      "entropy": 5.739324855804443,
      "epoch": 5.466108833604762,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002457424016225909,
      "loss": 4.8832,
      "mean_token_accuracy": 0.24091757237911224,
      "num_tokens": 108638943.0,
      "step": 48670
    },
    {
      "entropy": 5.688745355606079,
      "epoch": 5.4666704105127195,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002457025942496376,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23668971061706542,
      "num_tokens": 108649177.0,
      "step": 48675
    },
    {
      "entropy": 5.701318264007568,
      "epoch": 5.467231987420678,
      "grad_norm": 1.15625,
      "learning_rate": 0.00024566278780952514,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23118384033441544,
      "num_tokens": 108660387.0,
      "step": 48680
    },
    {
      "entropy": 5.7291319370269775,
      "epoch": 5.467793564328635,
      "grad_norm": 1.25,
      "learning_rate": 0.0002456229823035209,
      "loss": 4.8527,
      "mean_token_accuracy": 0.24297750294208526,
      "num_tokens": 108671327.0,
      "step": 48685
    },
    {
      "entropy": 5.6988372802734375,
      "epoch": 5.468355141236592,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024558317773289227,
      "loss": 4.898,
      "mean_token_accuracy": 0.23166432827711106,
      "num_tokens": 108681901.0,
      "step": 48690
    },
    {
      "entropy": 5.708074426651001,
      "epoch": 5.46891671814455,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002455433740989067,
      "loss": 4.9621,
      "mean_token_accuracy": 0.234345880150795,
      "num_tokens": 108693205.0,
      "step": 48695
    },
    {
      "entropy": 5.722608232498169,
      "epoch": 5.469478295052507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002455035714028315,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2401943698525429,
      "num_tokens": 108703651.0,
      "step": 48700
    },
    {
      "entropy": 5.6537196159362795,
      "epoch": 5.470039871960465,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002454637696459341,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2383439838886261,
      "num_tokens": 108715811.0,
      "step": 48705
    },
    {
      "entropy": 5.6866003513336185,
      "epoch": 5.470601448868423,
      "grad_norm": 1.15625,
      "learning_rate": 0.00024542396882948163,
      "loss": 4.9309,
      "mean_token_accuracy": 0.2294653221964836,
      "num_tokens": 108727479.0,
      "step": 48710
    },
    {
      "entropy": 5.768385744094848,
      "epoch": 5.47116302577638,
      "grad_norm": 1.234375,
      "learning_rate": 0.00024538416895474144,
      "loss": 5.0202,
      "mean_token_accuracy": 0.22636246383190156,
      "num_tokens": 108739083.0,
      "step": 48715
    },
    {
      "entropy": 5.8041605949401855,
      "epoch": 5.471724602684338,
      "grad_norm": 1.203125,
      "learning_rate": 0.00024534437002298083,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22646221667528152,
      "num_tokens": 108751596.0,
      "step": 48720
    },
    {
      "entropy": 5.791163206100464,
      "epoch": 5.472286179592295,
      "grad_norm": 1.25,
      "learning_rate": 0.000245304572035467,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2374052330851555,
      "num_tokens": 108762856.0,
      "step": 48725
    },
    {
      "entropy": 5.68000054359436,
      "epoch": 5.4728477565002525,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002452647749934671,
      "loss": 4.8742,
      "mean_token_accuracy": 0.24059663712978363,
      "num_tokens": 108773763.0,
      "step": 48730
    },
    {
      "entropy": 5.583693170547486,
      "epoch": 5.473409333408211,
      "grad_norm": 1.203125,
      "learning_rate": 0.00024522497889824837,
      "loss": 4.859,
      "mean_token_accuracy": 0.23228552043437958,
      "num_tokens": 108784803.0,
      "step": 48735
    },
    {
      "entropy": 5.6805534839630125,
      "epoch": 5.473970910316168,
      "grad_norm": 1.25,
      "learning_rate": 0.0002451851837510778,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2343878522515297,
      "num_tokens": 108795417.0,
      "step": 48740
    },
    {
      "entropy": 5.694190168380738,
      "epoch": 5.474532487224125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002451453895532226,
      "loss": 4.912,
      "mean_token_accuracy": 0.233713960647583,
      "num_tokens": 108806382.0,
      "step": 48745
    },
    {
      "entropy": 5.730755138397217,
      "epoch": 5.475094064132083,
      "grad_norm": 1.25,
      "learning_rate": 0.0002451055963059496,
      "loss": 4.846,
      "mean_token_accuracy": 0.24621084481477737,
      "num_tokens": 108816857.0,
      "step": 48750
    },
    {
      "entropy": 5.641189908981323,
      "epoch": 5.47565564104004,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024506580401052616,
      "loss": 4.823,
      "mean_token_accuracy": 0.23270409256219865,
      "num_tokens": 108828359.0,
      "step": 48755
    },
    {
      "entropy": 5.752550077438355,
      "epoch": 5.4762172179479975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002450260126682191,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2326003834605217,
      "num_tokens": 108839520.0,
      "step": 48760
    },
    {
      "entropy": 5.761789321899414,
      "epoch": 5.476778794855956,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024498622228029545,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2352812558412552,
      "num_tokens": 108850529.0,
      "step": 48765
    },
    {
      "entropy": 5.716484498977661,
      "epoch": 5.477340371763913,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002449464328480221,
      "loss": 4.9408,
      "mean_token_accuracy": 0.23975188434123992,
      "num_tokens": 108863452.0,
      "step": 48770
    },
    {
      "entropy": 5.6695023536682125,
      "epoch": 5.477901948671871,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000244906644372666,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2349812716245651,
      "num_tokens": 108875050.0,
      "step": 48775
    },
    {
      "entropy": 5.714032125473023,
      "epoch": 5.478463525579828,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000244866856855494,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2416660502552986,
      "num_tokens": 108886386.0,
      "step": 48780
    },
    {
      "entropy": 5.734966993331909,
      "epoch": 5.479025102487785,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024482707029777284,
      "loss": 4.9864,
      "mean_token_accuracy": 0.23315643072128295,
      "num_tokens": 108899474.0,
      "step": 48785
    },
    {
      "entropy": 5.670153427124023,
      "epoch": 5.4795866793957435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00024478728470076963,
      "loss": 4.8394,
      "mean_token_accuracy": 0.24778898060321808,
      "num_tokens": 108911714.0,
      "step": 48790
    },
    {
      "entropy": 5.721865940093994,
      "epoch": 5.480148256303701,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002447475000657509,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23304520696401596,
      "num_tokens": 108922500.0,
      "step": 48795
    },
    {
      "entropy": 5.702246522903442,
      "epoch": 5.480709833211658,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002447077163939836,
      "loss": 4.95,
      "mean_token_accuracy": 0.22725351452827453,
      "num_tokens": 108933805.0,
      "step": 48800
    },
    {
      "entropy": 5.772807455062866,
      "epoch": 5.481271410119616,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00024466793368673427,
      "loss": 5.0088,
      "mean_token_accuracy": 0.22436127364635466,
      "num_tokens": 108945538.0,
      "step": 48805
    },
    {
      "entropy": 5.731902933120727,
      "epoch": 5.481832987027573,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002446281519452698,
      "loss": 4.9494,
      "mean_token_accuracy": 0.23289889842271805,
      "num_tokens": 108957696.0,
      "step": 48810
    },
    {
      "entropy": 5.702057886123657,
      "epoch": 5.482394563935531,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002445883711708567,
      "loss": 4.8306,
      "mean_token_accuracy": 0.24038033932447433,
      "num_tokens": 108967930.0,
      "step": 48815
    },
    {
      "entropy": 5.723030757904053,
      "epoch": 5.482956140843489,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00024454859136476155,
      "loss": 4.9057,
      "mean_token_accuracy": 0.2396693229675293,
      "num_tokens": 108978058.0,
      "step": 48820
    },
    {
      "entropy": 5.683976459503174,
      "epoch": 5.483517717751446,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024450881252825117,
      "loss": 4.8809,
      "mean_token_accuracy": 0.24174241572618485,
      "num_tokens": 108988854.0,
      "step": 48825
    },
    {
      "entropy": 5.696445560455322,
      "epoch": 5.484079294659404,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000244469034662592,
      "loss": 4.9343,
      "mean_token_accuracy": 0.23448040187358857,
      "num_tokens": 109001060.0,
      "step": 48830
    },
    {
      "entropy": 5.702075862884522,
      "epoch": 5.484640871567361,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002444292577690506,
      "loss": 4.9025,
      "mean_token_accuracy": 0.23300651758909224,
      "num_tokens": 109011943.0,
      "step": 48835
    },
    {
      "entropy": 5.691203260421753,
      "epoch": 5.485202448475318,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00024438948184889354,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2355933129787445,
      "num_tokens": 109022511.0,
      "step": 48840
    },
    {
      "entropy": 5.63370304107666,
      "epoch": 5.4857640253832765,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002443497069033872,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23351497054100037,
      "num_tokens": 109033757.0,
      "step": 48845
    },
    {
      "entropy": 5.626951122283936,
      "epoch": 5.486325602291234,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002443099329337981,
      "loss": 4.7846,
      "mean_token_accuracy": 0.2444995626807213,
      "num_tokens": 109046075.0,
      "step": 48850
    },
    {
      "entropy": 5.67520694732666,
      "epoch": 5.486887179199192,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002442701599413927,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22720606327056886,
      "num_tokens": 109056768.0,
      "step": 48855
    },
    {
      "entropy": 5.730797004699707,
      "epoch": 5.487448756107149,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00024423038792743715,
      "loss": 4.96,
      "mean_token_accuracy": 0.232691989839077,
      "num_tokens": 109068929.0,
      "step": 48860
    },
    {
      "entropy": 5.775974893569947,
      "epoch": 5.488010333015106,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002441906168931981,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2328963488340378,
      "num_tokens": 109080613.0,
      "step": 48865
    },
    {
      "entropy": 5.708809852600098,
      "epoch": 5.488571909923064,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002441508468399417,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23531643450260162,
      "num_tokens": 109092742.0,
      "step": 48870
    },
    {
      "entropy": 5.648669624328614,
      "epoch": 5.4891334868310215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00024411107776893437,
      "loss": 4.8492,
      "mean_token_accuracy": 0.2348180055618286,
      "num_tokens": 109103846.0,
      "step": 48875
    },
    {
      "entropy": 5.703899765014649,
      "epoch": 5.489695063738979,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002440713096814423,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22952632308006288,
      "num_tokens": 109115492.0,
      "step": 48880
    },
    {
      "entropy": 5.755914402008057,
      "epoch": 5.490256640646937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00024403154257873172,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22978106439113616,
      "num_tokens": 109126528.0,
      "step": 48885
    },
    {
      "entropy": 5.7725544452667235,
      "epoch": 5.490818217554894,
      "grad_norm": 1.296875,
      "learning_rate": 0.00024399177646206884,
      "loss": 4.9574,
      "mean_token_accuracy": 0.23405147045850755,
      "num_tokens": 109138506.0,
      "step": 48890
    },
    {
      "entropy": 5.737675189971924,
      "epoch": 5.491379794462851,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024395201133271978,
      "loss": 4.9527,
      "mean_token_accuracy": 0.23600469380617142,
      "num_tokens": 109150380.0,
      "step": 48895
    },
    {
      "entropy": 5.801615524291992,
      "epoch": 5.491941371370809,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024391224719195077,
      "loss": 5.0042,
      "mean_token_accuracy": 0.23525922894477844,
      "num_tokens": 109162369.0,
      "step": 48900
    },
    {
      "entropy": 5.706682872772217,
      "epoch": 5.492502948278767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024387248404102779,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2384227067232132,
      "num_tokens": 109173872.0,
      "step": 48905
    },
    {
      "entropy": 5.7254801273345945,
      "epoch": 5.493064525186725,
      "grad_norm": 1.3125,
      "learning_rate": 0.00024383272188121708,
      "loss": 4.979,
      "mean_token_accuracy": 0.22701365351676941,
      "num_tokens": 109184316.0,
      "step": 48910
    },
    {
      "entropy": 5.725991296768188,
      "epoch": 5.493626102094682,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002437929607137846,
      "loss": 4.9488,
      "mean_token_accuracy": 0.23428297340869902,
      "num_tokens": 109196345.0,
      "step": 48915
    },
    {
      "entropy": 5.699605321884155,
      "epoch": 5.494187679002639,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024375320053999627,
      "loss": 4.8261,
      "mean_token_accuracy": 0.24319795072078704,
      "num_tokens": 109206858.0,
      "step": 48920
    },
    {
      "entropy": 5.738527631759643,
      "epoch": 5.494749255910597,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024371344136111823,
      "loss": 4.9003,
      "mean_token_accuracy": 0.23248763084411622,
      "num_tokens": 109216866.0,
      "step": 48925
    },
    {
      "entropy": 5.67579493522644,
      "epoch": 5.4953108328185545,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002436736831784162,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23938054740428924,
      "num_tokens": 109226623.0,
      "step": 48930
    },
    {
      "entropy": 5.766329193115235,
      "epoch": 5.495872409726512,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024363392599315636,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22334299087524415,
      "num_tokens": 109237560.0,
      "step": 48935
    },
    {
      "entropy": 5.719032621383667,
      "epoch": 5.49643398663447,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024359416980660454,
      "loss": 4.9421,
      "mean_token_accuracy": 0.23049750328063964,
      "num_tokens": 109249010.0,
      "step": 48940
    },
    {
      "entropy": 5.732293605804443,
      "epoch": 5.496995563542427,
      "grad_norm": 1.234375,
      "learning_rate": 0.00024355441462002648,
      "loss": 4.9578,
      "mean_token_accuracy": 0.23340585082769394,
      "num_tokens": 109260328.0,
      "step": 48945
    },
    {
      "entropy": 5.742058277130127,
      "epoch": 5.497557140450384,
      "grad_norm": 1.25,
      "learning_rate": 0.00024351466043468807,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23538212180137635,
      "num_tokens": 109270896.0,
      "step": 48950
    },
    {
      "entropy": 5.809207534790039,
      "epoch": 5.498118717358342,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024347490725185506,
      "loss": 5.0631,
      "mean_token_accuracy": 0.22514744102954865,
      "num_tokens": 109282259.0,
      "step": 48955
    },
    {
      "entropy": 5.65314998626709,
      "epoch": 5.4986802942663,
      "grad_norm": 1.203125,
      "learning_rate": 0.00024343515507279325,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23689794689416885,
      "num_tokens": 109293871.0,
      "step": 48960
    },
    {
      "entropy": 5.73141884803772,
      "epoch": 5.499241871174258,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002433954038987683,
      "loss": 4.921,
      "mean_token_accuracy": 0.22993701845407485,
      "num_tokens": 109305881.0,
      "step": 48965
    },
    {
      "entropy": 5.740840673446655,
      "epoch": 5.499803448082215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00024335565373104594,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23643287420272827,
      "num_tokens": 109317088.0,
      "step": 48970
    },
    {
      "entropy": 5.764564561843872,
      "epoch": 5.500365024990172,
      "grad_norm": 1.140625,
      "learning_rate": 0.00024331590457089193,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2339499205350876,
      "num_tokens": 109328379.0,
      "step": 48975
    },
    {
      "entropy": 5.672082328796387,
      "epoch": 5.50092660189813,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024327615641957175,
      "loss": 4.8298,
      "mean_token_accuracy": 0.2421851247549057,
      "num_tokens": 109339177.0,
      "step": 48980
    },
    {
      "entropy": 5.721300792694092,
      "epoch": 5.501488178806087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024323640927835106,
      "loss": 4.9316,
      "mean_token_accuracy": 0.23803888261318207,
      "num_tokens": 109350725.0,
      "step": 48985
    },
    {
      "entropy": 5.702010774612427,
      "epoch": 5.5020497557140455,
      "grad_norm": 1.25,
      "learning_rate": 0.00024319666314849542,
      "loss": 4.8963,
      "mean_token_accuracy": 0.24064189940690994,
      "num_tokens": 109361814.0,
      "step": 48990
    },
    {
      "entropy": 5.742294788360596,
      "epoch": 5.502611332622003,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00024315691803127037,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2485820919275284,
      "num_tokens": 109372920.0,
      "step": 48995
    },
    {
      "entropy": 5.789233112335205,
      "epoch": 5.50317290952996,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002431171739279413,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22334277033805847,
      "num_tokens": 109384415.0,
      "step": 49000
    },
    {
      "entropy": 5.757606792449951,
      "epoch": 5.503734486437917,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024307743083977384,
      "loss": 4.9997,
      "mean_token_accuracy": 0.23627754151821137,
      "num_tokens": 109394696.0,
      "step": 49005
    },
    {
      "entropy": 5.773317098617554,
      "epoch": 5.504296063345875,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024303768876803345,
      "loss": 4.9697,
      "mean_token_accuracy": 0.23688600659370423,
      "num_tokens": 109405513.0,
      "step": 49010
    },
    {
      "entropy": 5.649604320526123,
      "epoch": 5.5048576402538325,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00024299794771398536,
      "loss": 4.908,
      "mean_token_accuracy": 0.2341262310743332,
      "num_tokens": 109416693.0,
      "step": 49015
    },
    {
      "entropy": 5.7220845222473145,
      "epoch": 5.505419217161791,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024295820767889504,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21828473657369613,
      "num_tokens": 109428128.0,
      "step": 49020
    },
    {
      "entropy": 5.742692518234253,
      "epoch": 5.505980794069748,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024291846866402784,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2342733770608902,
      "num_tokens": 109441068.0,
      "step": 49025
    },
    {
      "entropy": 5.862460803985596,
      "epoch": 5.506542370977705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00024287873067064904,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2302213951945305,
      "num_tokens": 109452352.0,
      "step": 49030
    },
    {
      "entropy": 5.656011629104614,
      "epoch": 5.507103947885663,
      "grad_norm": 1.34375,
      "learning_rate": 0.00024283899370002393,
      "loss": 4.892,
      "mean_token_accuracy": 0.22801269739866256,
      "num_tokens": 109463742.0,
      "step": 49035
    },
    {
      "entropy": 5.714546203613281,
      "epoch": 5.50766552479362,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024279925775341765,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22586680948734283,
      "num_tokens": 109474521.0,
      "step": 49040
    },
    {
      "entropy": 5.726767444610596,
      "epoch": 5.5082271017015785,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00024275952283209557,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22622006386518478,
      "num_tokens": 109486207.0,
      "step": 49045
    },
    {
      "entropy": 5.736765670776367,
      "epoch": 5.508788678609536,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024271978893732277,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2342637374997139,
      "num_tokens": 109497509.0,
      "step": 49050
    },
    {
      "entropy": 5.749752998352051,
      "epoch": 5.509350255517493,
      "grad_norm": 1.3125,
      "learning_rate": 0.00024268005607036442,
      "loss": 4.9513,
      "mean_token_accuracy": 0.235555736720562,
      "num_tokens": 109509850.0,
      "step": 49055
    },
    {
      "entropy": 5.841482925415039,
      "epoch": 5.509911832425451,
      "grad_norm": 1.546875,
      "learning_rate": 0.00024264032423248555,
      "loss": 4.9776,
      "mean_token_accuracy": 0.23428122699260712,
      "num_tokens": 109521098.0,
      "step": 49060
    },
    {
      "entropy": 5.7131383419036865,
      "epoch": 5.510473409333408,
      "grad_norm": 1.25,
      "learning_rate": 0.0002426005934249514,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22751832902431487,
      "num_tokens": 109532563.0,
      "step": 49065
    },
    {
      "entropy": 5.726970958709717,
      "epoch": 5.5110349862413655,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002425608636490269,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22786740362644195,
      "num_tokens": 109543793.0,
      "step": 49070
    },
    {
      "entropy": 5.771748208999634,
      "epoch": 5.511596563149324,
      "grad_norm": 1.25,
      "learning_rate": 0.00024252113490597692,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2240043729543686,
      "num_tokens": 109554062.0,
      "step": 49075
    },
    {
      "entropy": 5.703523206710815,
      "epoch": 5.512158140057281,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002424814071970668,
      "loss": 4.8974,
      "mean_token_accuracy": 0.23398059457540513,
      "num_tokens": 109564404.0,
      "step": 49080
    },
    {
      "entropy": 5.8639702796936035,
      "epoch": 5.512719716965238,
      "grad_norm": 1.4375,
      "learning_rate": 0.00024244168052356125,
      "loss": 5.0985,
      "mean_token_accuracy": 0.22550399005413055,
      "num_tokens": 109575080.0,
      "step": 49085
    },
    {
      "entropy": 5.82400689125061,
      "epoch": 5.513281293873196,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024240195488672516,
      "loss": 4.969,
      "mean_token_accuracy": 0.2365937650203705,
      "num_tokens": 109586653.0,
      "step": 49090
    },
    {
      "entropy": 5.718645763397217,
      "epoch": 5.513842870781153,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024236223028782357,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22733087539672853,
      "num_tokens": 109598399.0,
      "step": 49095
    },
    {
      "entropy": 5.798913764953613,
      "epoch": 5.514404447689111,
      "grad_norm": 1.171875,
      "learning_rate": 0.00024232250672812116,
      "loss": 5.0926,
      "mean_token_accuracy": 0.22013463973999023,
      "num_tokens": 109610946.0,
      "step": 49100
    },
    {
      "entropy": 5.6665589809417725,
      "epoch": 5.514966024597069,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024228278420888283,
      "loss": 4.7731,
      "mean_token_accuracy": 0.25134028792381286,
      "num_tokens": 109621838.0,
      "step": 49105
    },
    {
      "entropy": 5.732748556137085,
      "epoch": 5.515527601505026,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002422430627313733,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22956553250551223,
      "num_tokens": 109633897.0,
      "step": 49110
    },
    {
      "entropy": 5.789325428009033,
      "epoch": 5.516089178412984,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00024220334229685736,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2273125022649765,
      "num_tokens": 109646012.0,
      "step": 49115
    },
    {
      "entropy": 5.692488384246826,
      "epoch": 5.516650755320941,
      "grad_norm": 1.375,
      "learning_rate": 0.00024216362290659977,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23598089665174485,
      "num_tokens": 109658188.0,
      "step": 49120
    },
    {
      "entropy": 5.655441474914551,
      "epoch": 5.517212332228898,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024212390456186512,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23748117834329605,
      "num_tokens": 109669219.0,
      "step": 49125
    },
    {
      "entropy": 5.770650863647461,
      "epoch": 5.5177739091368565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00024208418726391813,
      "loss": 5.0699,
      "mean_token_accuracy": 0.22672647833824158,
      "num_tokens": 109681040.0,
      "step": 49130
    },
    {
      "entropy": 5.745171308517456,
      "epoch": 5.518335486044814,
      "grad_norm": 1.265625,
      "learning_rate": 0.00024204447101402337,
      "loss": 4.918,
      "mean_token_accuracy": 0.24088890105485916,
      "num_tokens": 109692123.0,
      "step": 49135
    },
    {
      "entropy": 5.711923456192016,
      "epoch": 5.518897062952771,
      "grad_norm": 1.25,
      "learning_rate": 0.00024200475581344545,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23634764850139617,
      "num_tokens": 109702900.0,
      "step": 49140
    },
    {
      "entropy": 5.690035772323609,
      "epoch": 5.519458639860729,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002419650416634488,
      "loss": 4.928,
      "mean_token_accuracy": 0.23051147311925888,
      "num_tokens": 109713301.0,
      "step": 49145
    },
    {
      "entropy": 5.6668988227844235,
      "epoch": 5.520020216768686,
      "grad_norm": 1.234375,
      "learning_rate": 0.00024192532856529814,
      "loss": 4.9004,
      "mean_token_accuracy": 0.23581877797842027,
      "num_tokens": 109725272.0,
      "step": 49150
    },
    {
      "entropy": 5.740903091430664,
      "epoch": 5.520581793676644,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024188561652025787,
      "loss": 4.9123,
      "mean_token_accuracy": 0.2362104132771492,
      "num_tokens": 109736154.0,
      "step": 49155
    },
    {
      "entropy": 5.741043758392334,
      "epoch": 5.521143370584602,
      "grad_norm": 1.265625,
      "learning_rate": 0.00024184590552959234,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2316289633512497,
      "num_tokens": 109748403.0,
      "step": 49160
    },
    {
      "entropy": 5.701460075378418,
      "epoch": 5.521704947492559,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00024180619559456608,
      "loss": 4.8121,
      "mean_token_accuracy": 0.24437153935432435,
      "num_tokens": 109758232.0,
      "step": 49165
    },
    {
      "entropy": 5.746933317184448,
      "epoch": 5.522266524400517,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002417664867164434,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22850574553012848,
      "num_tokens": 109770198.0,
      "step": 49170
    },
    {
      "entropy": 5.715861511230469,
      "epoch": 5.522828101308474,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002417267788964887,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23636020421981813,
      "num_tokens": 109782564.0,
      "step": 49175
    },
    {
      "entropy": 5.656738996505737,
      "epoch": 5.523389678216432,
      "grad_norm": 1.25,
      "learning_rate": 0.00024168707213596626,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23929931074380875,
      "num_tokens": 109793186.0,
      "step": 49180
    },
    {
      "entropy": 5.6244340419769285,
      "epoch": 5.5239512551243894,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024164736643614033,
      "loss": 4.833,
      "mean_token_accuracy": 0.23525693267583847,
      "num_tokens": 109804799.0,
      "step": 49185
    },
    {
      "entropy": 5.6548666000366214,
      "epoch": 5.524512832032347,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002416076617982752,
      "loss": 4.7792,
      "mean_token_accuracy": 0.23889702558517456,
      "num_tokens": 109816031.0,
      "step": 49190
    },
    {
      "entropy": 5.807084798812866,
      "epoch": 5.525074408940304,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024156795822363508,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22567394971847535,
      "num_tokens": 109826952.0,
      "step": 49195
    },
    {
      "entropy": 5.773458051681518,
      "epoch": 5.525635985848262,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00024152825571348418,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2300048366189003,
      "num_tokens": 109839045.0,
      "step": 49200
    },
    {
      "entropy": 5.7518057346344,
      "epoch": 5.526197562756219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024148855426908656,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2299994185566902,
      "num_tokens": 109849812.0,
      "step": 49205
    },
    {
      "entropy": 5.712730789184571,
      "epoch": 5.526759139664177,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002414488538917064,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23674538731575012,
      "num_tokens": 109861368.0,
      "step": 49210
    },
    {
      "entropy": 5.707191705703735,
      "epoch": 5.5273207165721345,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002414091545826077,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22887323349714278,
      "num_tokens": 109872388.0,
      "step": 49215
    },
    {
      "entropy": 5.720988178253174,
      "epoch": 5.527882293480092,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00024136945634305445,
      "loss": 4.9379,
      "mean_token_accuracy": 0.23186415135860444,
      "num_tokens": 109883301.0,
      "step": 49220
    },
    {
      "entropy": 5.644603538513183,
      "epoch": 5.52844387038805,
      "grad_norm": 1.21875,
      "learning_rate": 0.00024132975917431088,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23537037670612335,
      "num_tokens": 109894823.0,
      "step": 49225
    },
    {
      "entropy": 5.72249207496643,
      "epoch": 5.529005447296007,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024129006307764084,
      "loss": 5.0036,
      "mean_token_accuracy": 0.22911134064197541,
      "num_tokens": 109905946.0,
      "step": 49230
    },
    {
      "entropy": 5.755976247787475,
      "epoch": 5.529567024203965,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00024125036805430829,
      "loss": 4.9582,
      "mean_token_accuracy": 0.23166203647851943,
      "num_tokens": 109917369.0,
      "step": 49235
    },
    {
      "entropy": 5.747470903396606,
      "epoch": 5.530128601111922,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000241210674105577,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2334508255124092,
      "num_tokens": 109929001.0,
      "step": 49240
    },
    {
      "entropy": 5.774543714523316,
      "epoch": 5.53069017801988,
      "grad_norm": 1.390625,
      "learning_rate": 0.00024117098123271107,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22463954240083694,
      "num_tokens": 109941557.0,
      "step": 49245
    },
    {
      "entropy": 5.775048732757568,
      "epoch": 5.531251754927838,
      "grad_norm": 1.28125,
      "learning_rate": 0.00024113128943697415,
      "loss": 4.9341,
      "mean_token_accuracy": 0.23500818610191346,
      "num_tokens": 109951603.0,
      "step": 49250
    },
    {
      "entropy": 5.741413879394531,
      "epoch": 5.531813331835795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002410915987196301,
      "loss": 4.9387,
      "mean_token_accuracy": 0.23238314390182496,
      "num_tokens": 109963037.0,
      "step": 49255
    },
    {
      "entropy": 5.719691514968872,
      "epoch": 5.532374908743752,
      "grad_norm": 1.1875,
      "learning_rate": 0.00024105190908194275,
      "loss": 4.8724,
      "mean_token_accuracy": 0.23940070271492003,
      "num_tokens": 109974267.0,
      "step": 49260
    },
    {
      "entropy": 5.672064399719238,
      "epoch": 5.53293648565171,
      "grad_norm": 1.328125,
      "learning_rate": 0.00024101222052517573,
      "loss": 4.9009,
      "mean_token_accuracy": 0.24051993936300278,
      "num_tokens": 109985591.0,
      "step": 49265
    },
    {
      "entropy": 5.6176190853118895,
      "epoch": 5.5334980625596675,
      "grad_norm": 1.1875,
      "learning_rate": 0.00024097253305059287,
      "loss": 4.8001,
      "mean_token_accuracy": 0.24163561165332795,
      "num_tokens": 109997219.0,
      "step": 49270
    },
    {
      "entropy": 5.776381349563598,
      "epoch": 5.534059639467625,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002409328466594577,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22662074714899064,
      "num_tokens": 110006775.0,
      "step": 49275
    },
    {
      "entropy": 5.754631042480469,
      "epoch": 5.534621216375583,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002408931613530339,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2379388615489006,
      "num_tokens": 110018162.0,
      "step": 49280
    },
    {
      "entropy": 5.843792724609375,
      "epoch": 5.53518279328354,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002408534771325851,
      "loss": 5.0326,
      "mean_token_accuracy": 0.22095028012990953,
      "num_tokens": 110029960.0,
      "step": 49285
    },
    {
      "entropy": 5.732512950897217,
      "epoch": 5.535744370191498,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002408137939993747,
      "loss": 4.9367,
      "mean_token_accuracy": 0.2301919937133789,
      "num_tokens": 110042043.0,
      "step": 49290
    },
    {
      "entropy": 5.6245745658874515,
      "epoch": 5.536305947099455,
      "grad_norm": 1.25,
      "learning_rate": 0.00024077411195466648,
      "loss": 4.928,
      "mean_token_accuracy": 0.22802761048078538,
      "num_tokens": 110053094.0,
      "step": 49295
    },
    {
      "entropy": 5.615874433517456,
      "epoch": 5.5368675240074126,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002407344309997238,
      "loss": 4.8213,
      "mean_token_accuracy": 0.24183914810419083,
      "num_tokens": 110063555.0,
      "step": 49300
    },
    {
      "entropy": 5.766332769393921,
      "epoch": 5.537429100915371,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024069475113581014,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22946658730506897,
      "num_tokens": 110075089.0,
      "step": 49305
    },
    {
      "entropy": 5.775435733795166,
      "epoch": 5.537990677823328,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00024065507236418888,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2270097851753235,
      "num_tokens": 110086170.0,
      "step": 49310
    },
    {
      "entropy": 5.687372446060181,
      "epoch": 5.538552254731285,
      "grad_norm": 1.3125,
      "learning_rate": 0.00024061539468612348,
      "loss": 4.8218,
      "mean_token_accuracy": 0.24224018007516862,
      "num_tokens": 110096251.0,
      "step": 49315
    },
    {
      "entropy": 5.635742712020874,
      "epoch": 5.539113831639243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024057571810287722,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23498299866914749,
      "num_tokens": 110107625.0,
      "step": 49320
    },
    {
      "entropy": 5.652885055541992,
      "epoch": 5.5396754085472,
      "grad_norm": 1.265625,
      "learning_rate": 0.00024053604261571338,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23270214945077897,
      "num_tokens": 110116929.0,
      "step": 49325
    },
    {
      "entropy": 5.702238082885742,
      "epoch": 5.540236985455158,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024049636822589537,
      "loss": 4.8803,
      "mean_token_accuracy": 0.23369839936494827,
      "num_tokens": 110126975.0,
      "step": 49330
    },
    {
      "entropy": 5.774954509735108,
      "epoch": 5.540798562363116,
      "grad_norm": 1.25,
      "learning_rate": 0.0002404566949346864,
      "loss": 4.9111,
      "mean_token_accuracy": 0.23670997172594072,
      "num_tokens": 110137817.0,
      "step": 49335
    },
    {
      "entropy": 5.790497732162476,
      "epoch": 5.541360139271073,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002404170227433497,
      "loss": 5.055,
      "mean_token_accuracy": 0.22907251715660096,
      "num_tokens": 110149897.0,
      "step": 49340
    },
    {
      "entropy": 5.723130607604981,
      "epoch": 5.541921716179031,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00024037735165314834,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23371192067861557,
      "num_tokens": 110161291.0,
      "step": 49345
    },
    {
      "entropy": 5.832889366149902,
      "epoch": 5.542483293086988,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024033768166534553,
      "loss": 4.971,
      "mean_token_accuracy": 0.2311115473508835,
      "num_tokens": 110172670.0,
      "step": 49350
    },
    {
      "entropy": 5.813878154754638,
      "epoch": 5.5430448699949455,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00024029801278120438,
      "loss": 5.0123,
      "mean_token_accuracy": 0.23088107258081436,
      "num_tokens": 110183165.0,
      "step": 49355
    },
    {
      "entropy": 5.694322633743286,
      "epoch": 5.543606446902904,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002402583450019879,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22834409177303314,
      "num_tokens": 110194177.0,
      "step": 49360
    },
    {
      "entropy": 5.658424949645996,
      "epoch": 5.544168023810861,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00024021867832895927,
      "loss": 4.8137,
      "mean_token_accuracy": 0.2437702387571335,
      "num_tokens": 110204641.0,
      "step": 49365
    },
    {
      "entropy": 5.776859951019287,
      "epoch": 5.544729600718819,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002401790127633814,
      "loss": 5.0916,
      "mean_token_accuracy": 0.2253921240568161,
      "num_tokens": 110216580.0,
      "step": 49370
    },
    {
      "entropy": 5.708100509643555,
      "epoch": 5.545291177626776,
      "grad_norm": 1.234375,
      "learning_rate": 0.00024013934830651728,
      "loss": 4.9762,
      "mean_token_accuracy": 0.23006606251001357,
      "num_tokens": 110228731.0,
      "step": 49375
    },
    {
      "entropy": 5.761509418487549,
      "epoch": 5.545852754534733,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00024009968495962986,
      "loss": 4.9114,
      "mean_token_accuracy": 0.23989674299955369,
      "num_tokens": 110241308.0,
      "step": 49380
    },
    {
      "entropy": 5.62846269607544,
      "epoch": 5.546414331442691,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00024006002272398199,
      "loss": 4.8764,
      "mean_token_accuracy": 0.23452382683753967,
      "num_tokens": 110253759.0,
      "step": 49385
    },
    {
      "entropy": 5.723520994186401,
      "epoch": 5.546975908350649,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002400203616008365,
      "loss": 4.946,
      "mean_token_accuracy": 0.24013960212469102,
      "num_tokens": 110265193.0,
      "step": 49390
    },
    {
      "entropy": 5.724983882904053,
      "epoch": 5.547537485258606,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002399807015914563,
      "loss": 4.878,
      "mean_token_accuracy": 0.232897487282753,
      "num_tokens": 110275777.0,
      "step": 49395
    },
    {
      "entropy": 5.736044454574585,
      "epoch": 5.548099062166564,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002399410426971041,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23959595412015916,
      "num_tokens": 110286537.0,
      "step": 49400
    },
    {
      "entropy": 5.729160785675049,
      "epoch": 5.548660639074521,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023990138491904273,
      "loss": 4.909,
      "mean_token_accuracy": 0.2364452823996544,
      "num_tokens": 110297188.0,
      "step": 49405
    },
    {
      "entropy": 5.664326858520508,
      "epoch": 5.5492222159824784,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023986172825853492,
      "loss": 4.795,
      "mean_token_accuracy": 0.24618472903966904,
      "num_tokens": 110306355.0,
      "step": 49410
    },
    {
      "entropy": 5.671721744537353,
      "epoch": 5.5497837928904366,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023982207271684326,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2312536671757698,
      "num_tokens": 110317657.0,
      "step": 49415
    },
    {
      "entropy": 5.685136175155639,
      "epoch": 5.550345369798394,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002397824182952305,
      "loss": 4.8674,
      "mean_token_accuracy": 0.24336303025484085,
      "num_tokens": 110327665.0,
      "step": 49420
    },
    {
      "entropy": 5.632123565673828,
      "epoch": 5.550906946706352,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00023974276499495918,
      "loss": 4.8416,
      "mean_token_accuracy": 0.24014101475477218,
      "num_tokens": 110338054.0,
      "step": 49425
    },
    {
      "entropy": 5.716318607330322,
      "epoch": 5.551468523614309,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023970311281729185,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22799790203571318,
      "num_tokens": 110349360.0,
      "step": 49430
    },
    {
      "entropy": 5.622117567062378,
      "epoch": 5.552030100522266,
      "grad_norm": 1.25,
      "learning_rate": 0.000239663461763491,
      "loss": 4.7914,
      "mean_token_accuracy": 0.23993350118398665,
      "num_tokens": 110359336.0,
      "step": 49435
    },
    {
      "entropy": 5.686192560195923,
      "epoch": 5.552591677430224,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002396238118348194,
      "loss": 4.8499,
      "mean_token_accuracy": 0.23479283452033997,
      "num_tokens": 110371808.0,
      "step": 49440
    },
    {
      "entropy": 5.75822229385376,
      "epoch": 5.553153254338182,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023958416303253933,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22708627730607986,
      "num_tokens": 110383697.0,
      "step": 49445
    },
    {
      "entropy": 5.660908126831055,
      "epoch": 5.553714831246139,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00023954451535791323,
      "loss": 4.8702,
      "mean_token_accuracy": 0.24283770769834517,
      "num_tokens": 110395310.0,
      "step": 49450
    },
    {
      "entropy": 5.710464334487915,
      "epoch": 5.554276408154097,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023950486881220353,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23797009736299515,
      "num_tokens": 110407140.0,
      "step": 49455
    },
    {
      "entropy": 5.75835599899292,
      "epoch": 5.554837985062054,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023946522339667254,
      "loss": 4.8678,
      "mean_token_accuracy": 0.24487511962652206,
      "num_tokens": 110418213.0,
      "step": 49460
    },
    {
      "entropy": 5.72408356666565,
      "epoch": 5.555399561970011,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002394255791125826,
      "loss": 4.8791,
      "mean_token_accuracy": 0.24113782942295076,
      "num_tokens": 110428868.0,
      "step": 49465
    },
    {
      "entropy": 5.786887979507446,
      "epoch": 5.5559611388779695,
      "grad_norm": 1.421875,
      "learning_rate": 0.000239385935961196,
      "loss": 4.9912,
      "mean_token_accuracy": 0.23907381892204285,
      "num_tokens": 110440156.0,
      "step": 49470
    },
    {
      "entropy": 5.727372694015503,
      "epoch": 5.556522715785927,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023934629394377512,
      "loss": 4.9069,
      "mean_token_accuracy": 0.23623932898044586,
      "num_tokens": 110450889.0,
      "step": 49475
    },
    {
      "entropy": 5.750430679321289,
      "epoch": 5.557084292693885,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000239306653061582,
      "loss": 4.9604,
      "mean_token_accuracy": 0.23708945363759995,
      "num_tokens": 110463557.0,
      "step": 49480
    },
    {
      "entropy": 5.7519556999206545,
      "epoch": 5.557645869601842,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00023926701331587885,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23436813801527023,
      "num_tokens": 110473944.0,
      "step": 49485
    },
    {
      "entropy": 5.766938495635986,
      "epoch": 5.558207446509799,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023922737470792794,
      "loss": 4.9521,
      "mean_token_accuracy": 0.23557764440774917,
      "num_tokens": 110484999.0,
      "step": 49490
    },
    {
      "entropy": 5.856929683685303,
      "epoch": 5.558769023417757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002391877372389913,
      "loss": 5.0555,
      "mean_token_accuracy": 0.22857159972190857,
      "num_tokens": 110498669.0,
      "step": 49495
    },
    {
      "entropy": 5.804921007156372,
      "epoch": 5.559330600325715,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023914810091033096,
      "loss": 5.0303,
      "mean_token_accuracy": 0.22396834045648575,
      "num_tokens": 110511183.0,
      "step": 49500
    },
    {
      "entropy": 5.776377534866333,
      "epoch": 5.559892177233672,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002391084657232089,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2300248086452484,
      "num_tokens": 110521886.0,
      "step": 49505
    },
    {
      "entropy": 5.712280225753784,
      "epoch": 5.56045375414163,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023906883167888727,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23418981283903123,
      "num_tokens": 110532413.0,
      "step": 49510
    },
    {
      "entropy": 5.726885986328125,
      "epoch": 5.561015331049587,
      "grad_norm": 1.078125,
      "learning_rate": 0.000239029198778628,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22650038748979567,
      "num_tokens": 110544900.0,
      "step": 49515
    },
    {
      "entropy": 5.68768424987793,
      "epoch": 5.561576907957544,
      "grad_norm": 1.515625,
      "learning_rate": 0.000238989567023693,
      "loss": 4.9438,
      "mean_token_accuracy": 0.23005007207393646,
      "num_tokens": 110556811.0,
      "step": 49520
    },
    {
      "entropy": 5.685465288162232,
      "epoch": 5.562138484865502,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023894993641534418,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23939232677221298,
      "num_tokens": 110567688.0,
      "step": 49525
    },
    {
      "entropy": 5.724913454055786,
      "epoch": 5.56270006177346,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00023891030695484333,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23346307575702668,
      "num_tokens": 110579821.0,
      "step": 49530
    },
    {
      "entropy": 5.658068704605102,
      "epoch": 5.563261638681418,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002388706786434523,
      "loss": 4.836,
      "mean_token_accuracy": 0.23171228617429734,
      "num_tokens": 110589897.0,
      "step": 49535
    },
    {
      "entropy": 5.748286628723145,
      "epoch": 5.563823215589375,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002388310514824329,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2336263656616211,
      "num_tokens": 110600826.0,
      "step": 49540
    },
    {
      "entropy": 5.7037159442901615,
      "epoch": 5.564384792497332,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023879142547304677,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2281272068619728,
      "num_tokens": 110611138.0,
      "step": 49545
    },
    {
      "entropy": 5.73829026222229,
      "epoch": 5.56494636940529,
      "grad_norm": 1.25,
      "learning_rate": 0.00023875180061655576,
      "loss": 4.9383,
      "mean_token_accuracy": 0.23758969753980635,
      "num_tokens": 110622967.0,
      "step": 49550
    },
    {
      "entropy": 5.783356380462647,
      "epoch": 5.5655079463132475,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023871217691422147,
      "loss": 4.9277,
      "mean_token_accuracy": 0.23133252561092377,
      "num_tokens": 110635062.0,
      "step": 49555
    },
    {
      "entropy": 5.706760883331299,
      "epoch": 5.566069523221206,
      "grad_norm": 1.296875,
      "learning_rate": 0.00023867255436730555,
      "loss": 4.8877,
      "mean_token_accuracy": 0.24174720495939256,
      "num_tokens": 110646061.0,
      "step": 49560
    },
    {
      "entropy": 5.6979225158691404,
      "epoch": 5.566631100129163,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002386329329770696,
      "loss": 4.8879,
      "mean_token_accuracy": 0.230607707798481,
      "num_tokens": 110658233.0,
      "step": 49565
    },
    {
      "entropy": 5.78012318611145,
      "epoch": 5.56719267703712,
      "grad_norm": 1.296875,
      "learning_rate": 0.00023859331274477514,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23028271198272704,
      "num_tokens": 110669253.0,
      "step": 49570
    },
    {
      "entropy": 5.654448843002319,
      "epoch": 5.567754253945077,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023855369367168378,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2372177228331566,
      "num_tokens": 110679386.0,
      "step": 49575
    },
    {
      "entropy": 5.725543594360351,
      "epoch": 5.568315830853035,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023851407575905682,
      "loss": 4.9177,
      "mean_token_accuracy": 0.23597621619701387,
      "num_tokens": 110690517.0,
      "step": 49580
    },
    {
      "entropy": 5.685705280303955,
      "epoch": 5.568877407760993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002384744590081559,
      "loss": 4.9622,
      "mean_token_accuracy": 0.23246621191501618,
      "num_tokens": 110701023.0,
      "step": 49585
    },
    {
      "entropy": 5.724173879623413,
      "epoch": 5.569438984668951,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002384348434202424,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23473188430070877,
      "num_tokens": 110711530.0,
      "step": 49590
    },
    {
      "entropy": 5.737039232254029,
      "epoch": 5.570000561576908,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023839522899657767,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2326465591788292,
      "num_tokens": 110723201.0,
      "step": 49595
    },
    {
      "entropy": 5.63750114440918,
      "epoch": 5.570562138484865,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023835561573842302,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2379717156291008,
      "num_tokens": 110733872.0,
      "step": 49600
    },
    {
      "entropy": 5.734098196029663,
      "epoch": 5.571123715392823,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002383160036470398,
      "loss": 4.976,
      "mean_token_accuracy": 0.2259795218706131,
      "num_tokens": 110745038.0,
      "step": 49605
    },
    {
      "entropy": 5.677025318145752,
      "epoch": 5.5716852923007805,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023827639272368923,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2296934112906456,
      "num_tokens": 110756002.0,
      "step": 49610
    },
    {
      "entropy": 5.740366506576538,
      "epoch": 5.572246869208739,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023823678296963253,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23249689489603043,
      "num_tokens": 110766462.0,
      "step": 49615
    },
    {
      "entropy": 5.737279176712036,
      "epoch": 5.572808446116696,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023819717438613097,
      "loss": 4.9434,
      "mean_token_accuracy": 0.23031264692544937,
      "num_tokens": 110777764.0,
      "step": 49620
    },
    {
      "entropy": 5.728754711151123,
      "epoch": 5.573370023024653,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023815756697444568,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23790262043476104,
      "num_tokens": 110788790.0,
      "step": 49625
    },
    {
      "entropy": 5.665294170379639,
      "epoch": 5.573931599932611,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023811796073583774,
      "loss": 4.8622,
      "mean_token_accuracy": 0.2392094537615776,
      "num_tokens": 110800178.0,
      "step": 49630
    },
    {
      "entropy": 5.711123466491699,
      "epoch": 5.574493176840568,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023807835567156815,
      "loss": 4.9139,
      "mean_token_accuracy": 0.235559318959713,
      "num_tokens": 110811037.0,
      "step": 49635
    },
    {
      "entropy": 5.781418943405152,
      "epoch": 5.5750547537485255,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023803875178289813,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2309546008706093,
      "num_tokens": 110821368.0,
      "step": 49640
    },
    {
      "entropy": 5.716587972640991,
      "epoch": 5.575616330656484,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023799914907108856,
      "loss": 4.897,
      "mean_token_accuracy": 0.23180558979511262,
      "num_tokens": 110832831.0,
      "step": 49645
    },
    {
      "entropy": 5.693026113510132,
      "epoch": 5.576177907564441,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023795954753740034,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23473577052354813,
      "num_tokens": 110844280.0,
      "step": 49650
    },
    {
      "entropy": 5.655135250091552,
      "epoch": 5.576739484472398,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023791994718309455,
      "loss": 4.8353,
      "mean_token_accuracy": 0.24018216580152513,
      "num_tokens": 110854882.0,
      "step": 49655
    },
    {
      "entropy": 5.776859092712402,
      "epoch": 5.577301061380356,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023788034800943203,
      "loss": 4.9743,
      "mean_token_accuracy": 0.23797797560691833,
      "num_tokens": 110865929.0,
      "step": 49660
    },
    {
      "entropy": 5.705296564102173,
      "epoch": 5.577862638288313,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023784075001767369,
      "loss": 4.8957,
      "mean_token_accuracy": 0.23580513745546341,
      "num_tokens": 110875996.0,
      "step": 49665
    },
    {
      "entropy": 5.764061832427979,
      "epoch": 5.5784242151962715,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023780115320908024,
      "loss": 4.9683,
      "mean_token_accuracy": 0.23033206909894943,
      "num_tokens": 110887862.0,
      "step": 49670
    },
    {
      "entropy": 5.705309057235718,
      "epoch": 5.578985792104229,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023776155758491253,
      "loss": 4.8226,
      "mean_token_accuracy": 0.2413991317152977,
      "num_tokens": 110899340.0,
      "step": 49675
    },
    {
      "entropy": 5.774061870574951,
      "epoch": 5.579547369012186,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002377219631464312,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22527080923318862,
      "num_tokens": 110910141.0,
      "step": 49680
    },
    {
      "entropy": 5.706602382659912,
      "epoch": 5.580108945920144,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000237682369894897,
      "loss": 4.9255,
      "mean_token_accuracy": 0.23477713316679,
      "num_tokens": 110920488.0,
      "step": 49685
    },
    {
      "entropy": 5.806616878509521,
      "epoch": 5.580670522828101,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023764277783157068,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22821013480424882,
      "num_tokens": 110933074.0,
      "step": 49690
    },
    {
      "entropy": 5.778977251052856,
      "epoch": 5.5812320997360585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023760318695771282,
      "loss": 4.96,
      "mean_token_accuracy": 0.23426734358072282,
      "num_tokens": 110945251.0,
      "step": 49695
    },
    {
      "entropy": 5.66997971534729,
      "epoch": 5.581793676644017,
      "grad_norm": 1.375,
      "learning_rate": 0.00023756359727458397,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2360146760940552,
      "num_tokens": 110955742.0,
      "step": 49700
    },
    {
      "entropy": 5.617361259460449,
      "epoch": 5.582355253551974,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002375240087834447,
      "loss": 4.7962,
      "mean_token_accuracy": 0.2379898101091385,
      "num_tokens": 110965395.0,
      "step": 49705
    },
    {
      "entropy": 5.60694317817688,
      "epoch": 5.582916830459931,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023748442148555554,
      "loss": 4.8019,
      "mean_token_accuracy": 0.2405146598815918,
      "num_tokens": 110976770.0,
      "step": 49710
    },
    {
      "entropy": 5.7100182056427,
      "epoch": 5.583478407367889,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023744483538217694,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22797396928071975,
      "num_tokens": 110987997.0,
      "step": 49715
    },
    {
      "entropy": 5.806688261032105,
      "epoch": 5.584039984275846,
      "grad_norm": 1.375,
      "learning_rate": 0.00023740525047456936,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2236882522702217,
      "num_tokens": 110998969.0,
      "step": 49720
    },
    {
      "entropy": 5.733681106567383,
      "epoch": 5.5846015611838045,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023736566676399307,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23762478679418564,
      "num_tokens": 111010481.0,
      "step": 49725
    },
    {
      "entropy": 5.708830881118774,
      "epoch": 5.585163138091762,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002373260842517086,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2354792445898056,
      "num_tokens": 111020355.0,
      "step": 49730
    },
    {
      "entropy": 5.73932375907898,
      "epoch": 5.585724714999719,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002372865029389763,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23365700095891953,
      "num_tokens": 111030044.0,
      "step": 49735
    },
    {
      "entropy": 5.718300676345825,
      "epoch": 5.586286291907677,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023724692282705635,
      "loss": 5.0077,
      "mean_token_accuracy": 0.23245724588632583,
      "num_tokens": 111041988.0,
      "step": 49740
    },
    {
      "entropy": 5.7295488834381105,
      "epoch": 5.586847868815634,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000237207343917209,
      "loss": 4.9147,
      "mean_token_accuracy": 0.2282541200518608,
      "num_tokens": 111052841.0,
      "step": 49745
    },
    {
      "entropy": 5.703019142150879,
      "epoch": 5.587409445723592,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023716776621069446,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22866038978099823,
      "num_tokens": 111063892.0,
      "step": 49750
    },
    {
      "entropy": 5.690382385253907,
      "epoch": 5.5879710226315495,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002371281897087729,
      "loss": 4.8927,
      "mean_token_accuracy": 0.24076860398054123,
      "num_tokens": 111074859.0,
      "step": 49755
    },
    {
      "entropy": 5.75516791343689,
      "epoch": 5.588532599539507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023708861441270452,
      "loss": 4.8836,
      "mean_token_accuracy": 0.23699159920215607,
      "num_tokens": 111087345.0,
      "step": 49760
    },
    {
      "entropy": 5.685900497436523,
      "epoch": 5.589094176447465,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023704904032374936,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23966031521558762,
      "num_tokens": 111098698.0,
      "step": 49765
    },
    {
      "entropy": 5.707938480377197,
      "epoch": 5.589655753355422,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00023700946744316747,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2445186898112297,
      "num_tokens": 111108793.0,
      "step": 49770
    },
    {
      "entropy": 5.782810258865356,
      "epoch": 5.590217330263379,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002369698957722189,
      "loss": 4.9693,
      "mean_token_accuracy": 0.23038095086812974,
      "num_tokens": 111120933.0,
      "step": 49775
    },
    {
      "entropy": 5.752042484283447,
      "epoch": 5.590778907171337,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023693032531216352,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23938625752925874,
      "num_tokens": 111132257.0,
      "step": 49780
    },
    {
      "entropy": 5.769192838668824,
      "epoch": 5.591340484079295,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023689075606426143,
      "loss": 4.924,
      "mean_token_accuracy": 0.23459392935037612,
      "num_tokens": 111143976.0,
      "step": 49785
    },
    {
      "entropy": 5.69423508644104,
      "epoch": 5.591902060987252,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002368511880297724,
      "loss": 4.8302,
      "mean_token_accuracy": 0.2447735846042633,
      "num_tokens": 111155404.0,
      "step": 49790
    },
    {
      "entropy": 5.6589759349822994,
      "epoch": 5.59246363789521,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023681162120995624,
      "loss": 4.868,
      "mean_token_accuracy": 0.23467596173286437,
      "num_tokens": 111166565.0,
      "step": 49795
    },
    {
      "entropy": 5.654398488998413,
      "epoch": 5.593025214803167,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023677205560607294,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22774346619844438,
      "num_tokens": 111177433.0,
      "step": 49800
    },
    {
      "entropy": 5.701624631881714,
      "epoch": 5.593586791711125,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023673249121938228,
      "loss": 4.9,
      "mean_token_accuracy": 0.2304779827594757,
      "num_tokens": 111189839.0,
      "step": 49805
    },
    {
      "entropy": 5.752690076828003,
      "epoch": 5.5941483686190825,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00023669292805114395,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2343818485736847,
      "num_tokens": 111200395.0,
      "step": 49810
    },
    {
      "entropy": 5.852237129211426,
      "epoch": 5.59470994552704,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023665336610261762,
      "loss": 5.0216,
      "mean_token_accuracy": 0.22929861694574355,
      "num_tokens": 111212741.0,
      "step": 49815
    },
    {
      "entropy": 5.67987289428711,
      "epoch": 5.595271522434998,
      "grad_norm": 1.203125,
      "learning_rate": 0.000236613805375063,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22574403285980224,
      "num_tokens": 111224448.0,
      "step": 49820
    },
    {
      "entropy": 5.7904421329498295,
      "epoch": 5.595833099342955,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002365742458697397,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23549492806196212,
      "num_tokens": 111236022.0,
      "step": 49825
    },
    {
      "entropy": 5.668386793136596,
      "epoch": 5.596394676250912,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023653468758790726,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23888513892889024,
      "num_tokens": 111246735.0,
      "step": 49830
    },
    {
      "entropy": 5.64355435371399,
      "epoch": 5.59695625315887,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023649513053082528,
      "loss": 4.843,
      "mean_token_accuracy": 0.24346702694892883,
      "num_tokens": 111258102.0,
      "step": 49835
    },
    {
      "entropy": 5.703329277038574,
      "epoch": 5.597517830066828,
      "grad_norm": 1.25,
      "learning_rate": 0.0002364555746997533,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2439573436975479,
      "num_tokens": 111268935.0,
      "step": 49840
    },
    {
      "entropy": 5.670661973953247,
      "epoch": 5.598079406974785,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023641602009595087,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23793557584285735,
      "num_tokens": 111281234.0,
      "step": 49845
    },
    {
      "entropy": 5.723091268539429,
      "epoch": 5.598640983882743,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002363764667206772,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23679340332746507,
      "num_tokens": 111293176.0,
      "step": 49850
    },
    {
      "entropy": 5.797414541244507,
      "epoch": 5.5992025607907,
      "grad_norm": 1.25,
      "learning_rate": 0.00023633691457519191,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2348275899887085,
      "num_tokens": 111304378.0,
      "step": 49855
    },
    {
      "entropy": 5.745563220977783,
      "epoch": 5.599764137698658,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023629736366075416,
      "loss": 4.8863,
      "mean_token_accuracy": 0.24140945971012115,
      "num_tokens": 111316045.0,
      "step": 49860
    },
    {
      "entropy": 5.698120832443237,
      "epoch": 5.600325714606615,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023625781397862333,
      "loss": 4.8997,
      "mean_token_accuracy": 0.24097406417131423,
      "num_tokens": 111326576.0,
      "step": 49865
    },
    {
      "entropy": 5.67765736579895,
      "epoch": 5.600887291514573,
      "grad_norm": 1.25,
      "learning_rate": 0.00023621826553005883,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2379505455493927,
      "num_tokens": 111337649.0,
      "step": 49870
    },
    {
      "entropy": 5.662923288345337,
      "epoch": 5.601448868422531,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023617871831631976,
      "loss": 4.8549,
      "mean_token_accuracy": 0.24190095365047454,
      "num_tokens": 111348656.0,
      "step": 49875
    },
    {
      "entropy": 5.724891805648804,
      "epoch": 5.602010445330488,
      "grad_norm": 1.34375,
      "learning_rate": 0.00023613917233866538,
      "loss": 4.9174,
      "mean_token_accuracy": 0.23892705887556076,
      "num_tokens": 111359737.0,
      "step": 49880
    },
    {
      "entropy": 5.7180455207824705,
      "epoch": 5.602572022238445,
      "grad_norm": 1.25,
      "learning_rate": 0.0002360996275983548,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2310969963669777,
      "num_tokens": 111372074.0,
      "step": 49885
    },
    {
      "entropy": 5.7387590408325195,
      "epoch": 5.603133599146403,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002360600840966472,
      "loss": 4.9521,
      "mean_token_accuracy": 0.23155470043420792,
      "num_tokens": 111383996.0,
      "step": 49890
    },
    {
      "entropy": 5.757465982437134,
      "epoch": 5.6036951760543605,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002360205418348016,
      "loss": 4.9414,
      "mean_token_accuracy": 0.23625704348087312,
      "num_tokens": 111394781.0,
      "step": 49895
    },
    {
      "entropy": 5.729977750778199,
      "epoch": 5.604256752962318,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00023598100081407704,
      "loss": 4.8568,
      "mean_token_accuracy": 0.2394014775753021,
      "num_tokens": 111406056.0,
      "step": 49900
    },
    {
      "entropy": 5.720129823684692,
      "epoch": 5.604818329870276,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023594146103573255,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22746681421995163,
      "num_tokens": 111417020.0,
      "step": 49905
    },
    {
      "entropy": 5.67790994644165,
      "epoch": 5.605379906778233,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002359019225010271,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22729936987161636,
      "num_tokens": 111428085.0,
      "step": 49910
    },
    {
      "entropy": 5.744340991973877,
      "epoch": 5.605941483686191,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002358623852112196,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22909461408853532,
      "num_tokens": 111438016.0,
      "step": 49915
    },
    {
      "entropy": 5.683776092529297,
      "epoch": 5.606503060594148,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023582284916756898,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23243512362241744,
      "num_tokens": 111449040.0,
      "step": 49920
    },
    {
      "entropy": 5.645930767059326,
      "epoch": 5.607064637502106,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000235783314371334,
      "loss": 4.806,
      "mean_token_accuracy": 0.24080441147089005,
      "num_tokens": 111459732.0,
      "step": 49925
    },
    {
      "entropy": 5.741466236114502,
      "epoch": 5.607626214410064,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023574378082377351,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2423311233520508,
      "num_tokens": 111470564.0,
      "step": 49930
    },
    {
      "entropy": 5.755554914474487,
      "epoch": 5.608187791318021,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002357042485261463,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22906013429164887,
      "num_tokens": 111481839.0,
      "step": 49935
    },
    {
      "entropy": 5.635622835159301,
      "epoch": 5.608749368225979,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002356647174797109,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23695996403694153,
      "num_tokens": 111492616.0,
      "step": 49940
    },
    {
      "entropy": 5.674778699874878,
      "epoch": 5.609310945133936,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023562518768572628,
      "loss": 4.9091,
      "mean_token_accuracy": 0.23753556460142136,
      "num_tokens": 111503707.0,
      "step": 49945
    },
    {
      "entropy": 5.7569764137268065,
      "epoch": 5.6098725220418935,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023558565914545094,
      "loss": 4.9986,
      "mean_token_accuracy": 0.22620685547590255,
      "num_tokens": 111513395.0,
      "step": 49950
    },
    {
      "entropy": 5.735737276077271,
      "epoch": 5.610434098949852,
      "grad_norm": 1.25,
      "learning_rate": 0.00023554613186014352,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23413964360952377,
      "num_tokens": 111523300.0,
      "step": 49955
    },
    {
      "entropy": 5.7132871627807615,
      "epoch": 5.610995675857809,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023550660583106253,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23876589089632033,
      "num_tokens": 111534510.0,
      "step": 49960
    },
    {
      "entropy": 5.878841876983643,
      "epoch": 5.611557252765766,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023546708105946657,
      "loss": 5.0807,
      "mean_token_accuracy": 0.22076600193977355,
      "num_tokens": 111546351.0,
      "step": 49965
    },
    {
      "entropy": 5.758942747116089,
      "epoch": 5.612118829673724,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023542755754661404,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23170113563537598,
      "num_tokens": 111557972.0,
      "step": 49970
    },
    {
      "entropy": 5.679187774658203,
      "epoch": 5.612680406581681,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002353880352937634,
      "loss": 4.8658,
      "mean_token_accuracy": 0.23907694220542908,
      "num_tokens": 111569710.0,
      "step": 49975
    },
    {
      "entropy": 5.783315181732178,
      "epoch": 5.6132419834896385,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023534851430217314,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2285304620862007,
      "num_tokens": 111580506.0,
      "step": 49980
    },
    {
      "entropy": 5.718686819076538,
      "epoch": 5.613803560397597,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023530899457310157,
      "loss": 4.8489,
      "mean_token_accuracy": 0.23621753305196763,
      "num_tokens": 111590866.0,
      "step": 49985
    },
    {
      "entropy": 5.744917631149292,
      "epoch": 5.614365137305554,
      "grad_norm": 1.375,
      "learning_rate": 0.00023526947610780696,
      "loss": 4.9037,
      "mean_token_accuracy": 0.23939224630594252,
      "num_tokens": 111600770.0,
      "step": 49990
    },
    {
      "entropy": 5.677889013290406,
      "epoch": 5.614926714213512,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023522995890754768,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2292979896068573,
      "num_tokens": 111612232.0,
      "step": 49995
    },
    {
      "entropy": 5.742619466781616,
      "epoch": 5.615488291121469,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002351904429735819,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23284255862236022,
      "num_tokens": 111622542.0,
      "step": 50000
    },
    {
      "entropy": 5.707755994796753,
      "epoch": 5.616049868029426,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002351509283071679,
      "loss": 4.9006,
      "mean_token_accuracy": 0.23179979324340821,
      "num_tokens": 111635991.0,
      "step": 50005
    },
    {
      "entropy": 5.688301944732666,
      "epoch": 5.6166114449373845,
      "grad_norm": 1.375,
      "learning_rate": 0.0002351114149095636,
      "loss": 4.7977,
      "mean_token_accuracy": 0.24634284228086473,
      "num_tokens": 111646232.0,
      "step": 50010
    },
    {
      "entropy": 5.694799375534058,
      "epoch": 5.617173021845342,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00023507190278202755,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23872286528348924,
      "num_tokens": 111657132.0,
      "step": 50015
    },
    {
      "entropy": 5.693377542495727,
      "epoch": 5.617734598753299,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00023503239192581755,
      "loss": 4.8598,
      "mean_token_accuracy": 0.24090895652770997,
      "num_tokens": 111667817.0,
      "step": 50020
    },
    {
      "entropy": 5.749580717086792,
      "epoch": 5.618296175661257,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002349928823421917,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2322448268532753,
      "num_tokens": 111679899.0,
      "step": 50025
    },
    {
      "entropy": 5.6867108821868895,
      "epoch": 5.618857752569214,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023495337403240795,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22926224321126937,
      "num_tokens": 111691303.0,
      "step": 50030
    },
    {
      "entropy": 5.827323579788208,
      "epoch": 5.6194193294771715,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023491386699772432,
      "loss": 4.9879,
      "mean_token_accuracy": 0.2378131628036499,
      "num_tokens": 111701726.0,
      "step": 50035
    },
    {
      "entropy": 5.678784513473511,
      "epoch": 5.61998090638513,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002348743612393987,
      "loss": 4.8295,
      "mean_token_accuracy": 0.24334892630577087,
      "num_tokens": 111712100.0,
      "step": 50040
    },
    {
      "entropy": 5.763339376449585,
      "epoch": 5.620542483293087,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023483485675868899,
      "loss": 4.9843,
      "mean_token_accuracy": 0.22481646239757538,
      "num_tokens": 111723396.0,
      "step": 50045
    },
    {
      "entropy": 5.668117427825928,
      "epoch": 5.621104060201045,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023479535355685306,
      "loss": 4.874,
      "mean_token_accuracy": 0.23167621195316315,
      "num_tokens": 111733870.0,
      "step": 50050
    },
    {
      "entropy": 5.701257514953613,
      "epoch": 5.621665637109002,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023475585163514862,
      "loss": 4.9294,
      "mean_token_accuracy": 0.23733251243829728,
      "num_tokens": 111745581.0,
      "step": 50055
    },
    {
      "entropy": 5.6857929706573485,
      "epoch": 5.622227214016959,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002347163509948335,
      "loss": 4.8779,
      "mean_token_accuracy": 0.24331656098365784,
      "num_tokens": 111756284.0,
      "step": 50060
    },
    {
      "entropy": 5.757672691345215,
      "epoch": 5.6227887909249175,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023467685163716534,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2328497737646103,
      "num_tokens": 111768018.0,
      "step": 50065
    },
    {
      "entropy": 5.78818359375,
      "epoch": 5.623350367832875,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023463735356340194,
      "loss": 5.0141,
      "mean_token_accuracy": 0.22986658662557602,
      "num_tokens": 111779455.0,
      "step": 50070
    },
    {
      "entropy": 5.788697242736816,
      "epoch": 5.623911944740832,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023459785677480078,
      "loss": 4.8683,
      "mean_token_accuracy": 0.24792924225330354,
      "num_tokens": 111790474.0,
      "step": 50075
    },
    {
      "entropy": 5.7673403263092045,
      "epoch": 5.62447352164879,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002345583612726196,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22888614535331725,
      "num_tokens": 111801392.0,
      "step": 50080
    },
    {
      "entropy": 5.640330600738525,
      "epoch": 5.625035098556747,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023451886705811576,
      "loss": 4.7915,
      "mean_token_accuracy": 0.24001013487577438,
      "num_tokens": 111812872.0,
      "step": 50085
    },
    {
      "entropy": 5.781858730316162,
      "epoch": 5.625596675464704,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000234479374132547,
      "loss": 5.0632,
      "mean_token_accuracy": 0.22704575657844545,
      "num_tokens": 111824538.0,
      "step": 50090
    },
    {
      "entropy": 5.702319526672364,
      "epoch": 5.6261582523726625,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023443988249717063,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2381973847746849,
      "num_tokens": 111835070.0,
      "step": 50095
    },
    {
      "entropy": 5.746301078796387,
      "epoch": 5.62671982928062,
      "grad_norm": 1.25,
      "learning_rate": 0.00023440039215324415,
      "loss": 4.8756,
      "mean_token_accuracy": 0.23614386916160585,
      "num_tokens": 111845919.0,
      "step": 50100
    },
    {
      "entropy": 5.677224445343017,
      "epoch": 5.627281406188578,
      "grad_norm": 1.40625,
      "learning_rate": 0.00023436090310202497,
      "loss": 4.7793,
      "mean_token_accuracy": 0.24312132298946382,
      "num_tokens": 111855996.0,
      "step": 50105
    },
    {
      "entropy": 5.78469557762146,
      "epoch": 5.627842983096535,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023432141534477036,
      "loss": 5.0624,
      "mean_token_accuracy": 0.22283658683300017,
      "num_tokens": 111867323.0,
      "step": 50110
    },
    {
      "entropy": 5.725283145904541,
      "epoch": 5.628404560004492,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023428192888273768,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23424901813268661,
      "num_tokens": 111878070.0,
      "step": 50115
    },
    {
      "entropy": 5.778795337677002,
      "epoch": 5.62896613691245,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023424244371718418,
      "loss": 4.9833,
      "mean_token_accuracy": 0.22744714170694352,
      "num_tokens": 111890163.0,
      "step": 50120
    },
    {
      "entropy": 5.75582365989685,
      "epoch": 5.629527713820408,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023420295984936707,
      "loss": 4.9678,
      "mean_token_accuracy": 0.23563210368156434,
      "num_tokens": 111901329.0,
      "step": 50125
    },
    {
      "entropy": 5.7186026096344,
      "epoch": 5.630089290728366,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023416347728054355,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23637309670448303,
      "num_tokens": 111911737.0,
      "step": 50130
    },
    {
      "entropy": 5.7683202743530275,
      "epoch": 5.630650867636323,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023412399601197072,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2246601939201355,
      "num_tokens": 111923375.0,
      "step": 50135
    },
    {
      "entropy": 5.667181062698364,
      "epoch": 5.63121244454428,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023408451604490577,
      "loss": 4.861,
      "mean_token_accuracy": 0.23173041343688966,
      "num_tokens": 111935304.0,
      "step": 50140
    },
    {
      "entropy": 5.759236621856689,
      "epoch": 5.631774021452238,
      "grad_norm": 1.25,
      "learning_rate": 0.00023404503738060566,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2288876459002495,
      "num_tokens": 111946773.0,
      "step": 50145
    },
    {
      "entropy": 5.7505090713500975,
      "epoch": 5.6323355983601955,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00023400556002032747,
      "loss": 4.964,
      "mean_token_accuracy": 0.22910052090883254,
      "num_tokens": 111958963.0,
      "step": 50150
    },
    {
      "entropy": 5.838738536834716,
      "epoch": 5.632897175268153,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000233966083965328,
      "loss": 5.0037,
      "mean_token_accuracy": 0.23985014706850052,
      "num_tokens": 111971249.0,
      "step": 50155
    },
    {
      "entropy": 5.7532764911651615,
      "epoch": 5.633458752176111,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023392660921686448,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2331301376223564,
      "num_tokens": 111982783.0,
      "step": 50160
    },
    {
      "entropy": 5.626570510864258,
      "epoch": 5.634020329084068,
      "grad_norm": 1.390625,
      "learning_rate": 0.00023388713577619357,
      "loss": 4.762,
      "mean_token_accuracy": 0.24394776672124863,
      "num_tokens": 111993653.0,
      "step": 50165
    },
    {
      "entropy": 5.712830018997193,
      "epoch": 5.634581905992025,
      "grad_norm": 1.390625,
      "learning_rate": 0.00023384766364457227,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23973798006772995,
      "num_tokens": 112004332.0,
      "step": 50170
    },
    {
      "entropy": 5.728810167312622,
      "epoch": 5.635143482899983,
      "grad_norm": 1.390625,
      "learning_rate": 0.00023380819282325732,
      "loss": 4.9221,
      "mean_token_accuracy": 0.22789032906293868,
      "num_tokens": 112015480.0,
      "step": 50175
    },
    {
      "entropy": 5.812261009216309,
      "epoch": 5.635705059807941,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023376872331350553,
      "loss": 4.9493,
      "mean_token_accuracy": 0.231537364423275,
      "num_tokens": 112026437.0,
      "step": 50180
    },
    {
      "entropy": 5.739535522460938,
      "epoch": 5.636266636715899,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002337292551165735,
      "loss": 4.8545,
      "mean_token_accuracy": 0.23922946006059648,
      "num_tokens": 112038183.0,
      "step": 50185
    },
    {
      "entropy": 5.690198659896851,
      "epoch": 5.636828213623856,
      "grad_norm": 1.25,
      "learning_rate": 0.000233689788233718,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2237753301858902,
      "num_tokens": 112050466.0,
      "step": 50190
    },
    {
      "entropy": 5.700616073608399,
      "epoch": 5.637389790531813,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023365032266619567,
      "loss": 4.9116,
      "mean_token_accuracy": 0.23163291513919831,
      "num_tokens": 112060818.0,
      "step": 50195
    },
    {
      "entropy": 5.708398246765137,
      "epoch": 5.637951367439771,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023361085841526314,
      "loss": 4.8813,
      "mean_token_accuracy": 0.24028423875570298,
      "num_tokens": 112070980.0,
      "step": 50200
    },
    {
      "entropy": 5.749102163314819,
      "epoch": 5.638512944347728,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023357139548217692,
      "loss": 4.926,
      "mean_token_accuracy": 0.23434635996818542,
      "num_tokens": 112082725.0,
      "step": 50205
    },
    {
      "entropy": 5.607955026626587,
      "epoch": 5.639074521255686,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023353193386819353,
      "loss": 4.7596,
      "mean_token_accuracy": 0.24297629296779633,
      "num_tokens": 112093900.0,
      "step": 50210
    },
    {
      "entropy": 5.657453727722168,
      "epoch": 5.639636098163644,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00023349247357456942,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2343771606683731,
      "num_tokens": 112104626.0,
      "step": 50215
    },
    {
      "entropy": 5.765619945526123,
      "epoch": 5.640197675071601,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023345301460256113,
      "loss": 4.933,
      "mean_token_accuracy": 0.23468345403671265,
      "num_tokens": 112116383.0,
      "step": 50220
    },
    {
      "entropy": 5.638074684143066,
      "epoch": 5.640759251979558,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002334135569534248,
      "loss": 4.7802,
      "mean_token_accuracy": 0.2447135865688324,
      "num_tokens": 112126645.0,
      "step": 50225
    },
    {
      "entropy": 5.742459774017334,
      "epoch": 5.641320828887516,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023337410062841706,
      "loss": 4.948,
      "mean_token_accuracy": 0.22940537035465242,
      "num_tokens": 112138740.0,
      "step": 50230
    },
    {
      "entropy": 5.6976173400878904,
      "epoch": 5.6418824057954735,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023333464562879408,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23891365379095078,
      "num_tokens": 112150037.0,
      "step": 50235
    },
    {
      "entropy": 5.74933443069458,
      "epoch": 5.642443982703432,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00023329519195581218,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2392736241221428,
      "num_tokens": 112161006.0,
      "step": 50240
    },
    {
      "entropy": 5.695224857330322,
      "epoch": 5.643005559611389,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023325573961072754,
      "loss": 4.82,
      "mean_token_accuracy": 0.2396562486886978,
      "num_tokens": 112172571.0,
      "step": 50245
    },
    {
      "entropy": 5.618524265289307,
      "epoch": 5.643567136519346,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00023321628859479632,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23852080404758452,
      "num_tokens": 112183745.0,
      "step": 50250
    },
    {
      "entropy": 5.717742776870727,
      "epoch": 5.644128713427304,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002331768389092747,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23344624936580657,
      "num_tokens": 112195200.0,
      "step": 50255
    },
    {
      "entropy": 5.795098924636841,
      "epoch": 5.644690290335261,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023313739055541878,
      "loss": 5.058,
      "mean_token_accuracy": 0.22395531535148622,
      "num_tokens": 112207295.0,
      "step": 50260
    },
    {
      "entropy": 5.743473386764526,
      "epoch": 5.645251867243219,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002330979435344845,
      "loss": 4.9052,
      "mean_token_accuracy": 0.23232258707284928,
      "num_tokens": 112218346.0,
      "step": 50265
    },
    {
      "entropy": 5.696404838562012,
      "epoch": 5.645813444151177,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00023305849784772799,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23193161934614182,
      "num_tokens": 112228876.0,
      "step": 50270
    },
    {
      "entropy": 5.6903787612915036,
      "epoch": 5.646375021059134,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002330190534964051,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23272258192300796,
      "num_tokens": 112240643.0,
      "step": 50275
    },
    {
      "entropy": 5.657125091552734,
      "epoch": 5.646936597967091,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023297961048177195,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23254963010549545,
      "num_tokens": 112251861.0,
      "step": 50280
    },
    {
      "entropy": 5.778654956817627,
      "epoch": 5.647498174875049,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023294016880508428,
      "loss": 4.95,
      "mean_token_accuracy": 0.23436036109924316,
      "num_tokens": 112262948.0,
      "step": 50285
    },
    {
      "entropy": 5.78067569732666,
      "epoch": 5.6480597517830065,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023290072846759784,
      "loss": 5.0138,
      "mean_token_accuracy": 0.23068304657936095,
      "num_tokens": 112274332.0,
      "step": 50290
    },
    {
      "entropy": 5.753297185897827,
      "epoch": 5.648621328690965,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023286128947056867,
      "loss": 4.9607,
      "mean_token_accuracy": 0.23218244761228563,
      "num_tokens": 112286241.0,
      "step": 50295
    },
    {
      "entropy": 5.650755214691162,
      "epoch": 5.649182905598922,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023282185181525223,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2368913248181343,
      "num_tokens": 112297179.0,
      "step": 50300
    },
    {
      "entropy": 5.719104528427124,
      "epoch": 5.649744482506879,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023278241550290453,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23819683641195297,
      "num_tokens": 112307951.0,
      "step": 50305
    },
    {
      "entropy": 5.676811647415161,
      "epoch": 5.650306059414837,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00023274298053478106,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2372742012143135,
      "num_tokens": 112319376.0,
      "step": 50310
    },
    {
      "entropy": 5.803207492828369,
      "epoch": 5.650867636322794,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023270354691213742,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2361740365624428,
      "num_tokens": 112330777.0,
      "step": 50315
    },
    {
      "entropy": 5.789609813690186,
      "epoch": 5.651429213230752,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023266411463622927,
      "loss": 4.9567,
      "mean_token_accuracy": 0.23626240342855453,
      "num_tokens": 112340551.0,
      "step": 50320
    },
    {
      "entropy": 5.752246332168579,
      "epoch": 5.65199079013871,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002326246837083122,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2373642459511757,
      "num_tokens": 112353012.0,
      "step": 50325
    },
    {
      "entropy": 5.697169256210327,
      "epoch": 5.652552367046667,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002325852541296416,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23754272013902664,
      "num_tokens": 112364031.0,
      "step": 50330
    },
    {
      "entropy": 5.737769651412964,
      "epoch": 5.653113943954625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023254582590147288,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23844321817159653,
      "num_tokens": 112375421.0,
      "step": 50335
    },
    {
      "entropy": 5.717795276641846,
      "epoch": 5.653675520862582,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002325063990250616,
      "loss": 4.9696,
      "mean_token_accuracy": 0.23410693407058716,
      "num_tokens": 112388023.0,
      "step": 50340
    },
    {
      "entropy": 5.673612833023071,
      "epoch": 5.654237097770539,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023246697350166306,
      "loss": 4.8514,
      "mean_token_accuracy": 0.24276581555604934,
      "num_tokens": 112397464.0,
      "step": 50345
    },
    {
      "entropy": 5.739718723297119,
      "epoch": 5.6547986746784975,
      "grad_norm": 1.25,
      "learning_rate": 0.00023242754933253256,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22874320298433304,
      "num_tokens": 112408252.0,
      "step": 50350
    },
    {
      "entropy": 5.66951322555542,
      "epoch": 5.655360251586455,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002323881265189255,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23456604331731795,
      "num_tokens": 112419043.0,
      "step": 50355
    },
    {
      "entropy": 5.700084781646728,
      "epoch": 5.655921828494412,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00023234870506209693,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23498287945985794,
      "num_tokens": 112429488.0,
      "step": 50360
    },
    {
      "entropy": 5.712156867980957,
      "epoch": 5.65648340540237,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002323092849633021,
      "loss": 4.925,
      "mean_token_accuracy": 0.23004397004842758,
      "num_tokens": 112441937.0,
      "step": 50365
    },
    {
      "entropy": 5.736015701293946,
      "epoch": 5.657044982310327,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023226986622379615,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23085020780563353,
      "num_tokens": 112453100.0,
      "step": 50370
    },
    {
      "entropy": 5.804655694961548,
      "epoch": 5.657606559218285,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00023223044884483436,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23970020115375518,
      "num_tokens": 112462691.0,
      "step": 50375
    },
    {
      "entropy": 5.649366188049316,
      "epoch": 5.658168136126243,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023219103282767157,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23751309514045715,
      "num_tokens": 112473390.0,
      "step": 50380
    },
    {
      "entropy": 5.678043079376221,
      "epoch": 5.6587297130342,
      "grad_norm": 1.28125,
      "learning_rate": 0.000232151618173563,
      "loss": 4.8639,
      "mean_token_accuracy": 0.24236710816621782,
      "num_tokens": 112482976.0,
      "step": 50385
    },
    {
      "entropy": 5.764587497711181,
      "epoch": 5.659291289942158,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002321122048837635,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2322361022233963,
      "num_tokens": 112493565.0,
      "step": 50390
    },
    {
      "entropy": 5.718734312057495,
      "epoch": 5.659852866850115,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023207279295952798,
      "loss": 4.9735,
      "mean_token_accuracy": 0.23204863369464873,
      "num_tokens": 112504794.0,
      "step": 50395
    },
    {
      "entropy": 5.675027894973755,
      "epoch": 5.660414443758072,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023203338240211142,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23892010301351546,
      "num_tokens": 112515561.0,
      "step": 50400
    },
    {
      "entropy": 5.775420475006103,
      "epoch": 5.6609760206660305,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002319939732127686,
      "loss": 4.9189,
      "mean_token_accuracy": 0.24194641709327697,
      "num_tokens": 112525534.0,
      "step": 50405
    },
    {
      "entropy": 5.685073661804199,
      "epoch": 5.661537597573988,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023195456539275433,
      "loss": 4.8633,
      "mean_token_accuracy": 0.24127948880195618,
      "num_tokens": 112536790.0,
      "step": 50410
    },
    {
      "entropy": 5.758313465118408,
      "epoch": 5.662099174481945,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023191515894332345,
      "loss": 4.9802,
      "mean_token_accuracy": 0.23345784544944764,
      "num_tokens": 112548613.0,
      "step": 50415
    },
    {
      "entropy": 5.7835516929626465,
      "epoch": 5.662660751389903,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023187575386573057,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23748953491449357,
      "num_tokens": 112558709.0,
      "step": 50420
    },
    {
      "entropy": 5.747518920898438,
      "epoch": 5.66322232829786,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023183635016123044,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2328342616558075,
      "num_tokens": 112569850.0,
      "step": 50425
    },
    {
      "entropy": 5.645466995239258,
      "epoch": 5.663783905205818,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002317969478310777,
      "loss": 4.823,
      "mean_token_accuracy": 0.2396102160215378,
      "num_tokens": 112580842.0,
      "step": 50430
    },
    {
      "entropy": 5.631681108474732,
      "epoch": 5.6643454821137755,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023175754687652684,
      "loss": 4.7572,
      "mean_token_accuracy": 0.24136146008968354,
      "num_tokens": 112590879.0,
      "step": 50435
    },
    {
      "entropy": 5.6464110851287845,
      "epoch": 5.664907059021733,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002317181472988325,
      "loss": 4.8137,
      "mean_token_accuracy": 0.24255583733320235,
      "num_tokens": 112602327.0,
      "step": 50440
    },
    {
      "entropy": 5.627379417419434,
      "epoch": 5.665468635929691,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023167874909924902,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2351918488740921,
      "num_tokens": 112613406.0,
      "step": 50445
    },
    {
      "entropy": 5.6666277885437015,
      "epoch": 5.666030212837648,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002316393522790311,
      "loss": 4.8739,
      "mean_token_accuracy": 0.23830221742391586,
      "num_tokens": 112624286.0,
      "step": 50450
    },
    {
      "entropy": 5.751467275619507,
      "epoch": 5.666591789745605,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00023159995683943297,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2400374159216881,
      "num_tokens": 112635097.0,
      "step": 50455
    },
    {
      "entropy": 5.757274580001831,
      "epoch": 5.667153366653563,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023156056278170905,
      "loss": 4.9308,
      "mean_token_accuracy": 0.23757594227790832,
      "num_tokens": 112645994.0,
      "step": 50460
    },
    {
      "entropy": 5.798316287994385,
      "epoch": 5.667714943561521,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023152117010711372,
      "loss": 4.9724,
      "mean_token_accuracy": 0.23255922943353652,
      "num_tokens": 112657902.0,
      "step": 50465
    },
    {
      "entropy": 5.699656534194946,
      "epoch": 5.668276520469478,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023148177881690112,
      "loss": 4.8598,
      "mean_token_accuracy": 0.24834474176168442,
      "num_tokens": 112668495.0,
      "step": 50470
    },
    {
      "entropy": 5.695471715927124,
      "epoch": 5.668838097377436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002314423889123256,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2275857925415039,
      "num_tokens": 112679314.0,
      "step": 50475
    },
    {
      "entropy": 5.728737878799438,
      "epoch": 5.669399674285393,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023140300039464125,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23504095375537873,
      "num_tokens": 112689114.0,
      "step": 50480
    },
    {
      "entropy": 5.7291234016418455,
      "epoch": 5.669961251193351,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023136361326510236,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23293666243553163,
      "num_tokens": 112701548.0,
      "step": 50485
    },
    {
      "entropy": 5.7751795768737795,
      "epoch": 5.6705228281013085,
      "grad_norm": 1.375,
      "learning_rate": 0.0002313242275249629,
      "loss": 4.9501,
      "mean_token_accuracy": 0.23291421681642532,
      "num_tokens": 112714576.0,
      "step": 50490
    },
    {
      "entropy": 5.655209732055664,
      "epoch": 5.671084405009266,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023128484317547698,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23914948850870132,
      "num_tokens": 112725835.0,
      "step": 50495
    },
    {
      "entropy": 5.713798999786377,
      "epoch": 5.671645981917224,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023124546021789866,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23647824972867965,
      "num_tokens": 112737525.0,
      "step": 50500
    },
    {
      "entropy": 5.751344823837281,
      "epoch": 5.672207558825181,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002312060786534818,
      "loss": 4.9899,
      "mean_token_accuracy": 0.22899970710277556,
      "num_tokens": 112750608.0,
      "step": 50505
    },
    {
      "entropy": 5.7826879024505615,
      "epoch": 5.672769135733139,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023116669848348043,
      "loss": 4.9888,
      "mean_token_accuracy": 0.23533567935228347,
      "num_tokens": 112761721.0,
      "step": 50510
    },
    {
      "entropy": 5.747591829299926,
      "epoch": 5.673330712641096,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023112731970914823,
      "loss": 4.9862,
      "mean_token_accuracy": 0.23282515853643418,
      "num_tokens": 112772797.0,
      "step": 50515
    },
    {
      "entropy": 5.723374319076538,
      "epoch": 5.673892289549054,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002310879423317393,
      "loss": 4.9279,
      "mean_token_accuracy": 0.23788437843322754,
      "num_tokens": 112784433.0,
      "step": 50520
    },
    {
      "entropy": 5.726518011093139,
      "epoch": 5.674453866457012,
      "grad_norm": 1.359375,
      "learning_rate": 0.00023104856635250732,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2293500304222107,
      "num_tokens": 112795340.0,
      "step": 50525
    },
    {
      "entropy": 5.706187629699707,
      "epoch": 5.675015443364969,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023100919177270607,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22994197010993958,
      "num_tokens": 112806779.0,
      "step": 50530
    },
    {
      "entropy": 5.757265615463257,
      "epoch": 5.675577020272926,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023096981859358918,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22886714488267898,
      "num_tokens": 112816844.0,
      "step": 50535
    },
    {
      "entropy": 5.655700635910034,
      "epoch": 5.676138597180884,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023093044681641035,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23949967920780182,
      "num_tokens": 112827647.0,
      "step": 50540
    },
    {
      "entropy": 5.611314821243286,
      "epoch": 5.676700174088841,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002308910764424232,
      "loss": 4.7952,
      "mean_token_accuracy": 0.23988854438066481,
      "num_tokens": 112838658.0,
      "step": 50545
    },
    {
      "entropy": 5.6977497100830075,
      "epoch": 5.677261750996799,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002308517074728812,
      "loss": 4.8444,
      "mean_token_accuracy": 0.24748685210943222,
      "num_tokens": 112850093.0,
      "step": 50550
    },
    {
      "entropy": 5.604024839401245,
      "epoch": 5.677823327904757,
      "grad_norm": 1.265625,
      "learning_rate": 0.00023081233990903798,
      "loss": 4.8271,
      "mean_token_accuracy": 0.240070903301239,
      "num_tokens": 112860269.0,
      "step": 50555
    },
    {
      "entropy": 5.6812458515167235,
      "epoch": 5.678384904812714,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023077297375214708,
      "loss": 4.8172,
      "mean_token_accuracy": 0.24324440509080886,
      "num_tokens": 112869904.0,
      "step": 50560
    },
    {
      "entropy": 5.724668979644775,
      "epoch": 5.678946481720672,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002307336090034618,
      "loss": 4.8861,
      "mean_token_accuracy": 0.24194006323814393,
      "num_tokens": 112881384.0,
      "step": 50565
    },
    {
      "entropy": 5.692864847183228,
      "epoch": 5.679508058628629,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023069424566423557,
      "loss": 4.8915,
      "mean_token_accuracy": 0.23645198345184326,
      "num_tokens": 112891873.0,
      "step": 50570
    },
    {
      "entropy": 5.681383228302002,
      "epoch": 5.6800696355365865,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023065488373572173,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23456970155239104,
      "num_tokens": 112903579.0,
      "step": 50575
    },
    {
      "entropy": 5.642356824874878,
      "epoch": 5.680631212444545,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023061552321917368,
      "loss": 4.8747,
      "mean_token_accuracy": 0.240111643075943,
      "num_tokens": 112913798.0,
      "step": 50580
    },
    {
      "entropy": 5.716984796524048,
      "epoch": 5.681192789352502,
      "grad_norm": 1.3125,
      "learning_rate": 0.00023057616411584453,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23400239944458007,
      "num_tokens": 112924442.0,
      "step": 50585
    },
    {
      "entropy": 5.681754207611084,
      "epoch": 5.681754366260459,
      "grad_norm": 1.21875,
      "learning_rate": 0.00023053680642698743,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23701150417327882,
      "num_tokens": 112935347.0,
      "step": 50590
    },
    {
      "entropy": 5.664907503128052,
      "epoch": 5.682315943168417,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00023049745015385576,
      "loss": 4.8303,
      "mean_token_accuracy": 0.23475640416145324,
      "num_tokens": 112945727.0,
      "step": 50595
    },
    {
      "entropy": 5.734347009658814,
      "epoch": 5.682877520076374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00023045809529770252,
      "loss": 4.8568,
      "mean_token_accuracy": 0.24015027582645415,
      "num_tokens": 112957998.0,
      "step": 50600
    },
    {
      "entropy": 5.590861988067627,
      "epoch": 5.683439096984332,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00023041874185978086,
      "loss": 4.7754,
      "mean_token_accuracy": 0.24103261679410934,
      "num_tokens": 112969660.0,
      "step": 50605
    },
    {
      "entropy": 5.705480289459229,
      "epoch": 5.68400067389229,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002303793898413437,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23486198782920836,
      "num_tokens": 112980996.0,
      "step": 50610
    },
    {
      "entropy": 5.702691268920899,
      "epoch": 5.684562250800247,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002303400392436441,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23709216862916946,
      "num_tokens": 112991200.0,
      "step": 50615
    },
    {
      "entropy": 5.736744070053101,
      "epoch": 5.685123827708205,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002303006900679349,
      "loss": 4.953,
      "mean_token_accuracy": 0.23822615295648575,
      "num_tokens": 113003097.0,
      "step": 50620
    },
    {
      "entropy": 5.649279928207397,
      "epoch": 5.685685404616162,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00023026134231546907,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2353222295641899,
      "num_tokens": 113013767.0,
      "step": 50625
    },
    {
      "entropy": 5.734450101852417,
      "epoch": 5.6862469815241194,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00023022199598749955,
      "loss": 4.941,
      "mean_token_accuracy": 0.23229118287563325,
      "num_tokens": 113024731.0,
      "step": 50630
    },
    {
      "entropy": 5.6291725635528564,
      "epoch": 5.6868085584320776,
      "grad_norm": 1.234375,
      "learning_rate": 0.00023018265108527903,
      "loss": 4.8116,
      "mean_token_accuracy": 0.24215140640735627,
      "num_tokens": 113036289.0,
      "step": 50635
    },
    {
      "entropy": 5.721045398712159,
      "epoch": 5.687370135340035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00023014330761006018,
      "loss": 4.9226,
      "mean_token_accuracy": 0.23030027151107788,
      "num_tokens": 113047968.0,
      "step": 50640
    },
    {
      "entropy": 5.658940982818604,
      "epoch": 5.687931712247992,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023010396556309592,
      "loss": 4.8288,
      "mean_token_accuracy": 0.24293868541717528,
      "num_tokens": 113059687.0,
      "step": 50645
    },
    {
      "entropy": 5.730331945419311,
      "epoch": 5.68849328915595,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002300646249456388,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23387132883071898,
      "num_tokens": 113071094.0,
      "step": 50650
    },
    {
      "entropy": 5.763234710693359,
      "epoch": 5.689054866063907,
      "grad_norm": 1.296875,
      "learning_rate": 0.00023002528575894138,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2459324538707733,
      "num_tokens": 113081600.0,
      "step": 50655
    },
    {
      "entropy": 5.690684604644775,
      "epoch": 5.6896164429718645,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00022998594800425627,
      "loss": 4.7551,
      "mean_token_accuracy": 0.2421037659049034,
      "num_tokens": 113092176.0,
      "step": 50660
    },
    {
      "entropy": 5.706968498229981,
      "epoch": 5.690178019879823,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022994661168283608,
      "loss": 4.9201,
      "mean_token_accuracy": 0.23314590007066727,
      "num_tokens": 113102745.0,
      "step": 50665
    },
    {
      "entropy": 5.767347717285157,
      "epoch": 5.69073959678778,
      "grad_norm": 1.25,
      "learning_rate": 0.00022990727679593332,
      "loss": 5.0117,
      "mean_token_accuracy": 0.22893679440021514,
      "num_tokens": 113115977.0,
      "step": 50670
    },
    {
      "entropy": 5.736521577835083,
      "epoch": 5.691301173695738,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022986794334480033,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2333666741847992,
      "num_tokens": 113127089.0,
      "step": 50675
    },
    {
      "entropy": 5.73823184967041,
      "epoch": 5.691862750603695,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002298286113306895,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2411332756280899,
      "num_tokens": 113137100.0,
      "step": 50680
    },
    {
      "entropy": 5.726071739196778,
      "epoch": 5.692424327511652,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00022978928075485322,
      "loss": 4.9431,
      "mean_token_accuracy": 0.23766030371189117,
      "num_tokens": 113148636.0,
      "step": 50685
    },
    {
      "entropy": 5.8310315132141115,
      "epoch": 5.6929859044196105,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002297499516185438,
      "loss": 5.0778,
      "mean_token_accuracy": 0.21774666458368303,
      "num_tokens": 113160806.0,
      "step": 50690
    },
    {
      "entropy": 5.782692003250122,
      "epoch": 5.693547481327568,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00022971062392301335,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22955155074596406,
      "num_tokens": 113172317.0,
      "step": 50695
    },
    {
      "entropy": 5.743481731414795,
      "epoch": 5.694109058235526,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022967129766951429,
      "loss": 4.9311,
      "mean_token_accuracy": 0.23573919236660004,
      "num_tokens": 113183325.0,
      "step": 50700
    },
    {
      "entropy": 5.707510948181152,
      "epoch": 5.694670635143483,
      "grad_norm": 1.25,
      "learning_rate": 0.0002296319728592987,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22872983366250993,
      "num_tokens": 113194897.0,
      "step": 50705
    },
    {
      "entropy": 5.7277624130249025,
      "epoch": 5.69523221205144,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002295926494936187,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2330588221549988,
      "num_tokens": 113205879.0,
      "step": 50710
    },
    {
      "entropy": 5.7628730773925785,
      "epoch": 5.695793788959398,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022955332757372633,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2296076536178589,
      "num_tokens": 113217101.0,
      "step": 50715
    },
    {
      "entropy": 5.73251895904541,
      "epoch": 5.696355365867356,
      "grad_norm": 1.234375,
      "learning_rate": 0.00022951400710087368,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2350664883852005,
      "num_tokens": 113228264.0,
      "step": 50720
    },
    {
      "entropy": 5.708191156387329,
      "epoch": 5.696916942775313,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002294746880763126,
      "loss": 4.9137,
      "mean_token_accuracy": 0.2292372390627861,
      "num_tokens": 113239246.0,
      "step": 50725
    },
    {
      "entropy": 5.720658254623413,
      "epoch": 5.697478519683271,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022943537050129504,
      "loss": 4.9568,
      "mean_token_accuracy": 0.23555784076452255,
      "num_tokens": 113251761.0,
      "step": 50730
    },
    {
      "entropy": 5.697655200958252,
      "epoch": 5.698040096591228,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002293960543770731,
      "loss": 4.8733,
      "mean_token_accuracy": 0.23578172028064728,
      "num_tokens": 113262642.0,
      "step": 50735
    },
    {
      "entropy": 5.683955240249634,
      "epoch": 5.698601673499185,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022935673970489847,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2424854815006256,
      "num_tokens": 113272632.0,
      "step": 50740
    },
    {
      "entropy": 5.66049747467041,
      "epoch": 5.699163250407143,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022931742648602294,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23743731826543807,
      "num_tokens": 113283187.0,
      "step": 50745
    },
    {
      "entropy": 5.637461948394775,
      "epoch": 5.699724827315101,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022927811472169823,
      "loss": 4.8667,
      "mean_token_accuracy": 0.23801787942647934,
      "num_tokens": 113293195.0,
      "step": 50750
    },
    {
      "entropy": 5.6595440864562985,
      "epoch": 5.700286404223059,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022923880441317606,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22878505885601044,
      "num_tokens": 113304238.0,
      "step": 50755
    },
    {
      "entropy": 5.709531211853028,
      "epoch": 5.700847981131016,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002291994955617082,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23763333708047868,
      "num_tokens": 113314688.0,
      "step": 50760
    },
    {
      "entropy": 5.728743410110473,
      "epoch": 5.701409558038973,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002291601881685461,
      "loss": 4.9043,
      "mean_token_accuracy": 0.23556324541568757,
      "num_tokens": 113326402.0,
      "step": 50765
    },
    {
      "entropy": 5.739669513702393,
      "epoch": 5.701971134946931,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022912088223494138,
      "loss": 4.917,
      "mean_token_accuracy": 0.2291679561138153,
      "num_tokens": 113338265.0,
      "step": 50770
    },
    {
      "entropy": 5.700257730484009,
      "epoch": 5.7025327118548885,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022908157776214556,
      "loss": 4.8202,
      "mean_token_accuracy": 0.2370728522539139,
      "num_tokens": 113348079.0,
      "step": 50775
    },
    {
      "entropy": 5.703463411331176,
      "epoch": 5.703094288762846,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022904227475141015,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23977322429418563,
      "num_tokens": 113359545.0,
      "step": 50780
    },
    {
      "entropy": 5.6960530281066895,
      "epoch": 5.703655865670804,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022900297320398654,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23828950822353362,
      "num_tokens": 113370336.0,
      "step": 50785
    },
    {
      "entropy": 5.647885036468506,
      "epoch": 5.704217442578761,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022896367312112608,
      "loss": 4.7925,
      "mean_token_accuracy": 0.24210480749607086,
      "num_tokens": 113380850.0,
      "step": 50790
    },
    {
      "entropy": 5.73498764038086,
      "epoch": 5.704779019486718,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022892437450408015,
      "loss": 4.9312,
      "mean_token_accuracy": 0.23006769865751267,
      "num_tokens": 113392018.0,
      "step": 50795
    },
    {
      "entropy": 5.792138290405274,
      "epoch": 5.705340596394676,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022888507735409997,
      "loss": 4.9749,
      "mean_token_accuracy": 0.23282122910022734,
      "num_tokens": 113403024.0,
      "step": 50800
    },
    {
      "entropy": 5.72588357925415,
      "epoch": 5.705902173302634,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022884578167243675,
      "loss": 4.8501,
      "mean_token_accuracy": 0.24029065817594528,
      "num_tokens": 113413602.0,
      "step": 50805
    },
    {
      "entropy": 5.679409313201904,
      "epoch": 5.706463750210592,
      "grad_norm": 1.4375,
      "learning_rate": 0.00022880648746034183,
      "loss": 4.8775,
      "mean_token_accuracy": 0.24290339797735214,
      "num_tokens": 113423577.0,
      "step": 50810
    },
    {
      "entropy": 5.770303440093994,
      "epoch": 5.707025327118549,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022876719471906626,
      "loss": 4.9439,
      "mean_token_accuracy": 0.23438520431518556,
      "num_tokens": 113435282.0,
      "step": 50815
    },
    {
      "entropy": 5.649136209487915,
      "epoch": 5.707586904026506,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022872790344986117,
      "loss": 4.8085,
      "mean_token_accuracy": 0.2377283826470375,
      "num_tokens": 113445950.0,
      "step": 50820
    },
    {
      "entropy": 5.694232082366943,
      "epoch": 5.708148480934464,
      "grad_norm": 1.234375,
      "learning_rate": 0.00022868861365397752,
      "loss": 4.9228,
      "mean_token_accuracy": 0.23301026821136475,
      "num_tokens": 113457825.0,
      "step": 50825
    },
    {
      "entropy": 5.717355537414551,
      "epoch": 5.7087100578424215,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022864932533266642,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23372165709733964,
      "num_tokens": 113467890.0,
      "step": 50830
    },
    {
      "entropy": 5.681355810165405,
      "epoch": 5.709271634750379,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00022861003848717877,
      "loss": 4.9687,
      "mean_token_accuracy": 0.23363482803106309,
      "num_tokens": 113480140.0,
      "step": 50835
    },
    {
      "entropy": 5.756179761886597,
      "epoch": 5.709833211658337,
      "grad_norm": 1.21875,
      "learning_rate": 0.00022857075311876542,
      "loss": 4.9402,
      "mean_token_accuracy": 0.23382414430379866,
      "num_tokens": 113492888.0,
      "step": 50840
    },
    {
      "entropy": 5.793625688552856,
      "epoch": 5.710394788566294,
      "grad_norm": 1.4375,
      "learning_rate": 0.00022853146922867734,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22772236764431,
      "num_tokens": 113505087.0,
      "step": 50845
    },
    {
      "entropy": 5.709452438354492,
      "epoch": 5.710956365474251,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002284921868181653,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22849286198616028,
      "num_tokens": 113516351.0,
      "step": 50850
    },
    {
      "entropy": 5.755077123641968,
      "epoch": 5.711517942382209,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022845290588848012,
      "loss": 5.0197,
      "mean_token_accuracy": 0.23004274517297746,
      "num_tokens": 113528959.0,
      "step": 50855
    },
    {
      "entropy": 5.670355224609375,
      "epoch": 5.7120795192901666,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022841362644087244,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23854512870311737,
      "num_tokens": 113539445.0,
      "step": 50860
    },
    {
      "entropy": 5.855929565429688,
      "epoch": 5.712641096198125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022837434847659295,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2311660885810852,
      "num_tokens": 113551293.0,
      "step": 50865
    },
    {
      "entropy": 5.694243335723877,
      "epoch": 5.713202673106082,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002283350719968923,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23835580050945282,
      "num_tokens": 113560838.0,
      "step": 50870
    },
    {
      "entropy": 5.691450357437134,
      "epoch": 5.713764250014039,
      "grad_norm": 1.3125,
      "learning_rate": 0.000228295797003021,
      "loss": 4.8343,
      "mean_token_accuracy": 0.24291236996650695,
      "num_tokens": 113571682.0,
      "step": 50875
    },
    {
      "entropy": 5.665078067779541,
      "epoch": 5.714325826921997,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022825652349622966,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23749197721481324,
      "num_tokens": 113581612.0,
      "step": 50880
    },
    {
      "entropy": 5.766972351074219,
      "epoch": 5.714887403829954,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022821725147776877,
      "loss": 4.9756,
      "mean_token_accuracy": 0.227487014234066,
      "num_tokens": 113592674.0,
      "step": 50885
    },
    {
      "entropy": 5.731192111968994,
      "epoch": 5.7154489807379125,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022817798094888875,
      "loss": 4.972,
      "mean_token_accuracy": 0.22953000664710999,
      "num_tokens": 113603752.0,
      "step": 50890
    },
    {
      "entropy": 5.7134596824646,
      "epoch": 5.71601055764587,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022813871191083995,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23878082782030105,
      "num_tokens": 113614939.0,
      "step": 50895
    },
    {
      "entropy": 5.786653518676758,
      "epoch": 5.716572134553827,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022809944436487278,
      "loss": 4.9244,
      "mean_token_accuracy": 0.23448525220155716,
      "num_tokens": 113626153.0,
      "step": 50900
    },
    {
      "entropy": 5.7516765117645265,
      "epoch": 5.717133711461785,
      "grad_norm": 1.34375,
      "learning_rate": 0.00022806017831223746,
      "loss": 4.89,
      "mean_token_accuracy": 0.23453885465860366,
      "num_tokens": 113635846.0,
      "step": 50905
    },
    {
      "entropy": 5.6281475067138675,
      "epoch": 5.717695288369742,
      "grad_norm": 1.25,
      "learning_rate": 0.00022802091375418422,
      "loss": 4.8374,
      "mean_token_accuracy": 0.24016272276639938,
      "num_tokens": 113647168.0,
      "step": 50910
    },
    {
      "entropy": 5.677828359603882,
      "epoch": 5.7182568652776995,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022798165069196335,
      "loss": 4.8627,
      "mean_token_accuracy": 0.24503183662891387,
      "num_tokens": 113658613.0,
      "step": 50915
    },
    {
      "entropy": 5.747236251831055,
      "epoch": 5.718818442185658,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00022794238912682497,
      "loss": 4.9689,
      "mean_token_accuracy": 0.23431316763162613,
      "num_tokens": 113670491.0,
      "step": 50920
    },
    {
      "entropy": 5.757941865921021,
      "epoch": 5.719380019093615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022790312906001919,
      "loss": 4.8873,
      "mean_token_accuracy": 0.2378758504986763,
      "num_tokens": 113681321.0,
      "step": 50925
    },
    {
      "entropy": 5.68265380859375,
      "epoch": 5.719941596001572,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022786387049279606,
      "loss": 4.9028,
      "mean_token_accuracy": 0.23716887682676316,
      "num_tokens": 113692428.0,
      "step": 50930
    },
    {
      "entropy": 5.715895652770996,
      "epoch": 5.72050317290953,
      "grad_norm": 1.390625,
      "learning_rate": 0.00022782461342640554,
      "loss": 4.8583,
      "mean_token_accuracy": 0.24839462339878082,
      "num_tokens": 113703056.0,
      "step": 50935
    },
    {
      "entropy": 5.673115396499634,
      "epoch": 5.721064749817487,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00022778535786209758,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23744018524885177,
      "num_tokens": 113713545.0,
      "step": 50940
    },
    {
      "entropy": 5.694749879837036,
      "epoch": 5.7216263267254455,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002277461038011222,
      "loss": 4.8905,
      "mean_token_accuracy": 0.231969653069973,
      "num_tokens": 113724952.0,
      "step": 50945
    },
    {
      "entropy": 5.776150321960449,
      "epoch": 5.722187903633403,
      "grad_norm": 1.375,
      "learning_rate": 0.00022770685124472907,
      "loss": 5.0202,
      "mean_token_accuracy": 0.22907583862543107,
      "num_tokens": 113736378.0,
      "step": 50950
    },
    {
      "entropy": 5.634954595565796,
      "epoch": 5.72274948054136,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022766760019416826,
      "loss": 4.888,
      "mean_token_accuracy": 0.23531108498573303,
      "num_tokens": 113748192.0,
      "step": 50955
    },
    {
      "entropy": 5.70733871459961,
      "epoch": 5.723311057449318,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022762835065068943,
      "loss": 4.925,
      "mean_token_accuracy": 0.23235076516866685,
      "num_tokens": 113759448.0,
      "step": 50960
    },
    {
      "entropy": 5.694953203201294,
      "epoch": 5.723872634357275,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022758910261554228,
      "loss": 4.8649,
      "mean_token_accuracy": 0.23622241765260696,
      "num_tokens": 113771216.0,
      "step": 50965
    },
    {
      "entropy": 5.6786778450012205,
      "epoch": 5.724434211265232,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022754985608997648,
      "loss": 4.8471,
      "mean_token_accuracy": 0.23592765629291534,
      "num_tokens": 113782978.0,
      "step": 50970
    },
    {
      "entropy": 5.641299247741699,
      "epoch": 5.7249957881731905,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022751061107524167,
      "loss": 4.8149,
      "mean_token_accuracy": 0.24246489405632018,
      "num_tokens": 113793197.0,
      "step": 50975
    },
    {
      "entropy": 5.734564399719238,
      "epoch": 5.725557365081148,
      "grad_norm": 1.375,
      "learning_rate": 0.00022747136757258742,
      "loss": 4.9206,
      "mean_token_accuracy": 0.23861124068498613,
      "num_tokens": 113804336.0,
      "step": 50980
    },
    {
      "entropy": 5.7260126113891605,
      "epoch": 5.726118941989105,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002274321255832632,
      "loss": 4.9218,
      "mean_token_accuracy": 0.23462932407855988,
      "num_tokens": 113814789.0,
      "step": 50985
    },
    {
      "entropy": 5.811304569244385,
      "epoch": 5.726680518897063,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022739288510851863,
      "loss": 4.9576,
      "mean_token_accuracy": 0.23123427033424376,
      "num_tokens": 113825741.0,
      "step": 50990
    },
    {
      "entropy": 5.769088888168335,
      "epoch": 5.72724209580502,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00022735364614960302,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2296169027686119,
      "num_tokens": 113837099.0,
      "step": 50995
    },
    {
      "entropy": 5.750071954727173,
      "epoch": 5.727803672712978,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022731440870776582,
      "loss": 4.9101,
      "mean_token_accuracy": 0.23746568858623504,
      "num_tokens": 113848448.0,
      "step": 51000
    },
    {
      "epoch": 5.727803672712978,
      "eval_entropy": 5.475889913009435,
      "eval_loss": 5.040252685546875,
      "eval_mean_token_accuracy": 0.23537496654934142,
      "eval_num_tokens": 113848448.0,
      "eval_runtime": 18.4787,
      "eval_samples_per_second": 1678.309,
      "eval_steps_per_second": 209.809,
      "step": 51000
    },
    {
      "entropy": 5.770274448394775,
      "epoch": 5.728365249620936,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022727517278425637,
      "loss": 4.955,
      "mean_token_accuracy": 0.22983956784009935,
      "num_tokens": 113860858.0,
      "step": 51005
    },
    {
      "entropy": 5.770592308044433,
      "epoch": 5.728926826528893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022723593838032391,
      "loss": 4.9251,
      "mean_token_accuracy": 0.23666827529668807,
      "num_tokens": 113872531.0,
      "step": 51010
    },
    {
      "entropy": 5.717377138137818,
      "epoch": 5.729488403436851,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022719670549721765,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2341795727610588,
      "num_tokens": 113882868.0,
      "step": 51015
    },
    {
      "entropy": 5.679789447784424,
      "epoch": 5.730049980344808,
      "grad_norm": 1.171875,
      "learning_rate": 0.00022715747413618677,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2236080914735794,
      "num_tokens": 113895002.0,
      "step": 51020
    },
    {
      "entropy": 5.678621435165406,
      "epoch": 5.730611557252765,
      "grad_norm": 1.453125,
      "learning_rate": 0.00022711824429848054,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2403668373823166,
      "num_tokens": 113904471.0,
      "step": 51025
    },
    {
      "entropy": 5.751374244689941,
      "epoch": 5.7311731341607235,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000227079015985348,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22947370857000352,
      "num_tokens": 113915945.0,
      "step": 51030
    },
    {
      "entropy": 5.796186304092407,
      "epoch": 5.731734711068681,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002270397891980382,
      "loss": 4.9861,
      "mean_token_accuracy": 0.22770701199769974,
      "num_tokens": 113926471.0,
      "step": 51035
    },
    {
      "entropy": 5.734240007400513,
      "epoch": 5.732296287976638,
      "grad_norm": 1.234375,
      "learning_rate": 0.00022700056393780012,
      "loss": 4.8942,
      "mean_token_accuracy": 0.226161827147007,
      "num_tokens": 113936720.0,
      "step": 51040
    },
    {
      "entropy": 5.790862512588501,
      "epoch": 5.732857864884596,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022696134020588267,
      "loss": 4.9908,
      "mean_token_accuracy": 0.23027998208999634,
      "num_tokens": 113947753.0,
      "step": 51045
    },
    {
      "entropy": 5.66494722366333,
      "epoch": 5.733419441792553,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002269221180035348,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23927133083343505,
      "num_tokens": 113958642.0,
      "step": 51050
    },
    {
      "entropy": 5.633037662506103,
      "epoch": 5.733981018700511,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022688289733200522,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2378491148352623,
      "num_tokens": 113969574.0,
      "step": 51055
    },
    {
      "entropy": 5.707807540893555,
      "epoch": 5.734542595608469,
      "grad_norm": 1.28125,
      "learning_rate": 0.000226843678192543,
      "loss": 4.8852,
      "mean_token_accuracy": 0.24184823483228685,
      "num_tokens": 113979336.0,
      "step": 51060
    },
    {
      "entropy": 5.705325889587402,
      "epoch": 5.735104172516426,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002268044605863967,
      "loss": 4.9006,
      "mean_token_accuracy": 0.233249232172966,
      "num_tokens": 113989789.0,
      "step": 51065
    },
    {
      "entropy": 5.787096691131592,
      "epoch": 5.735665749424384,
      "grad_norm": 1.40625,
      "learning_rate": 0.00022676524451481505,
      "loss": 4.9262,
      "mean_token_accuracy": 0.23854009211063384,
      "num_tokens": 114001362.0,
      "step": 51070
    },
    {
      "entropy": 5.761281967163086,
      "epoch": 5.736227326332341,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022672602997904674,
      "loss": 4.9672,
      "mean_token_accuracy": 0.23396296799182892,
      "num_tokens": 114012178.0,
      "step": 51075
    },
    {
      "entropy": 5.705800104141235,
      "epoch": 5.736788903240299,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022668681698034032,
      "loss": 4.8791,
      "mean_token_accuracy": 0.23695298582315444,
      "num_tokens": 114022294.0,
      "step": 51080
    },
    {
      "entropy": 5.779377555847168,
      "epoch": 5.737350480148256,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022664760551994445,
      "loss": 5.0232,
      "mean_token_accuracy": 0.22725771963596345,
      "num_tokens": 114033483.0,
      "step": 51085
    },
    {
      "entropy": 5.763644790649414,
      "epoch": 5.737912057056214,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00022660839559910756,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23104379773139955,
      "num_tokens": 114044553.0,
      "step": 51090
    },
    {
      "entropy": 5.800331354141235,
      "epoch": 5.738473633964172,
      "grad_norm": 1.296875,
      "learning_rate": 0.000226569187219078,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22464509308338165,
      "num_tokens": 114055041.0,
      "step": 51095
    },
    {
      "entropy": 5.735109758377075,
      "epoch": 5.739035210872129,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002265299803811044,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23323934376239777,
      "num_tokens": 114066507.0,
      "step": 51100
    },
    {
      "entropy": 5.691771125793457,
      "epoch": 5.739596787780086,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022649077508643507,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23188407272100447,
      "num_tokens": 114078509.0,
      "step": 51105
    },
    {
      "entropy": 5.729889154434204,
      "epoch": 5.740158364688044,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022645157133631823,
      "loss": 4.9817,
      "mean_token_accuracy": 0.23065504431724548,
      "num_tokens": 114089722.0,
      "step": 51110
    },
    {
      "entropy": 5.636091613769532,
      "epoch": 5.7407199415960015,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022641236913200215,
      "loss": 4.8498,
      "mean_token_accuracy": 0.23310959637165068,
      "num_tokens": 114101557.0,
      "step": 51115
    },
    {
      "entropy": 5.768421363830567,
      "epoch": 5.741281518503959,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022637316847473515,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2346772313117981,
      "num_tokens": 114113146.0,
      "step": 51120
    },
    {
      "entropy": 5.786077785491943,
      "epoch": 5.741843095411917,
      "grad_norm": 1.171875,
      "learning_rate": 0.00022633396936576534,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2296988323330879,
      "num_tokens": 114124255.0,
      "step": 51125
    },
    {
      "entropy": 5.729560375213623,
      "epoch": 5.742404672319874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022629477180634072,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2323276862502098,
      "num_tokens": 114134676.0,
      "step": 51130
    },
    {
      "entropy": 5.745161628723144,
      "epoch": 5.742966249227832,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002262555757977095,
      "loss": 4.995,
      "mean_token_accuracy": 0.22887725830078126,
      "num_tokens": 114146512.0,
      "step": 51135
    },
    {
      "entropy": 5.749880599975586,
      "epoch": 5.743527826135789,
      "grad_norm": 1.125,
      "learning_rate": 0.00022621638134111966,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23991550952196122,
      "num_tokens": 114157993.0,
      "step": 51140
    },
    {
      "entropy": 5.61900520324707,
      "epoch": 5.744089403043747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002261771884378192,
      "loss": 4.7513,
      "mean_token_accuracy": 0.24519666135311127,
      "num_tokens": 114169972.0,
      "step": 51145
    },
    {
      "entropy": 5.747770309448242,
      "epoch": 5.744650979951705,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002261379970890559,
      "loss": 5.0644,
      "mean_token_accuracy": 0.22115389853715897,
      "num_tokens": 114180938.0,
      "step": 51150
    },
    {
      "entropy": 5.724523115158081,
      "epoch": 5.745212556859662,
      "grad_norm": 1.46875,
      "learning_rate": 0.00022609880729607785,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2305856838822365,
      "num_tokens": 114191580.0,
      "step": 51155
    },
    {
      "entropy": 5.770151090621948,
      "epoch": 5.745774133767619,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022605961906013266,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2402611717581749,
      "num_tokens": 114203093.0,
      "step": 51160
    },
    {
      "entropy": 5.708196020126342,
      "epoch": 5.746335710675577,
      "grad_norm": 1.25,
      "learning_rate": 0.0002260204323824681,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23073617070913316,
      "num_tokens": 114213994.0,
      "step": 51165
    },
    {
      "entropy": 5.7011494636535645,
      "epoch": 5.7468972875835345,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022598124726433205,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2353825479745865,
      "num_tokens": 114225779.0,
      "step": 51170
    },
    {
      "entropy": 5.718143653869629,
      "epoch": 5.747458864491492,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022594206370697212,
      "loss": 4.8491,
      "mean_token_accuracy": 0.24450392425060272,
      "num_tokens": 114237515.0,
      "step": 51175
    },
    {
      "entropy": 5.717296743392945,
      "epoch": 5.74802044139945,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00022590288171163593,
      "loss": 4.9083,
      "mean_token_accuracy": 0.24148720055818557,
      "num_tokens": 114248075.0,
      "step": 51180
    },
    {
      "entropy": 5.749168395996094,
      "epoch": 5.748582018307407,
      "grad_norm": 1.28125,
      "learning_rate": 0.000225863701279571,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2326269954442978,
      "num_tokens": 114258863.0,
      "step": 51185
    },
    {
      "entropy": 5.722118234634399,
      "epoch": 5.749143595215365,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002258245224120249,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2313338503241539,
      "num_tokens": 114269567.0,
      "step": 51190
    },
    {
      "entropy": 5.706813955307007,
      "epoch": 5.749705172123322,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022578534511024506,
      "loss": 4.9394,
      "mean_token_accuracy": 0.2344011053442955,
      "num_tokens": 114280633.0,
      "step": 51195
    },
    {
      "entropy": 5.859527444839477,
      "epoch": 5.7502667490312795,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00022574616937547887,
      "loss": 5.0794,
      "mean_token_accuracy": 0.22497980296611786,
      "num_tokens": 114293015.0,
      "step": 51200
    },
    {
      "entropy": 5.805923986434936,
      "epoch": 5.750828325939238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022570699520897382,
      "loss": 5.005,
      "mean_token_accuracy": 0.22699466198682786,
      "num_tokens": 114304036.0,
      "step": 51205
    },
    {
      "entropy": 5.799118709564209,
      "epoch": 5.751389902847195,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002256678226119771,
      "loss": 4.917,
      "mean_token_accuracy": 0.24173789769411086,
      "num_tokens": 114315294.0,
      "step": 51210
    },
    {
      "entropy": 5.826858329772949,
      "epoch": 5.751951479755152,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022562865158573615,
      "loss": 4.9762,
      "mean_token_accuracy": 0.23340684175491333,
      "num_tokens": 114326788.0,
      "step": 51215
    },
    {
      "entropy": 5.56150107383728,
      "epoch": 5.75251305666311,
      "grad_norm": 1.234375,
      "learning_rate": 0.000225589482131498,
      "loss": 4.7161,
      "mean_token_accuracy": 0.24324504286050797,
      "num_tokens": 114336873.0,
      "step": 51220
    },
    {
      "entropy": 5.618921756744385,
      "epoch": 5.753074633571067,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00022555031425050993,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23629596680402756,
      "num_tokens": 114346743.0,
      "step": 51225
    },
    {
      "entropy": 5.742931270599366,
      "epoch": 5.753636210479025,
      "grad_norm": 1.15625,
      "learning_rate": 0.000225511147944019,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2278559535741806,
      "num_tokens": 114358794.0,
      "step": 51230
    },
    {
      "entropy": 5.79579758644104,
      "epoch": 5.754197787386983,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022547198321327223,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22118086814880372,
      "num_tokens": 114369697.0,
      "step": 51235
    },
    {
      "entropy": 5.750486516952515,
      "epoch": 5.75475936429494,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002254328200595168,
      "loss": 4.8545,
      "mean_token_accuracy": 0.24123610854148864,
      "num_tokens": 114380219.0,
      "step": 51240
    },
    {
      "entropy": 5.706678104400635,
      "epoch": 5.755320941202898,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022539365848399967,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23811500370502472,
      "num_tokens": 114390561.0,
      "step": 51245
    },
    {
      "entropy": 5.6831340312957765,
      "epoch": 5.755882518110855,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022535449848796763,
      "loss": 4.83,
      "mean_token_accuracy": 0.23894305974245073,
      "num_tokens": 114400996.0,
      "step": 51250
    },
    {
      "entropy": 5.694957304000854,
      "epoch": 5.7564440950188125,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022531534007266764,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2324714407324791,
      "num_tokens": 114414176.0,
      "step": 51255
    },
    {
      "entropy": 5.661417245864868,
      "epoch": 5.757005671926771,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022527618323934652,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2373937889933586,
      "num_tokens": 114424070.0,
      "step": 51260
    },
    {
      "entropy": 5.707091331481934,
      "epoch": 5.757567248834728,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022523702798925093,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23352333456277846,
      "num_tokens": 114435634.0,
      "step": 51265
    },
    {
      "entropy": 5.878308343887329,
      "epoch": 5.758128825742686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022519787432362777,
      "loss": 5.0441,
      "mean_token_accuracy": 0.22665376514196395,
      "num_tokens": 114446675.0,
      "step": 51270
    },
    {
      "entropy": 5.779688978195191,
      "epoch": 5.758690402650643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00022515872224372353,
      "loss": 4.9923,
      "mean_token_accuracy": 0.23192495703697205,
      "num_tokens": 114458876.0,
      "step": 51275
    },
    {
      "entropy": 5.730689859390258,
      "epoch": 5.7592519795586,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002251195717507849,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22757302224636078,
      "num_tokens": 114469733.0,
      "step": 51280
    },
    {
      "entropy": 5.713296890258789,
      "epoch": 5.7598135564665585,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022508042284605852,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2326044976711273,
      "num_tokens": 114481413.0,
      "step": 51285
    },
    {
      "entropy": 5.730577182769776,
      "epoch": 5.760375133374516,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002250412755307908,
      "loss": 4.9628,
      "mean_token_accuracy": 0.23911195993423462,
      "num_tokens": 114492705.0,
      "step": 51290
    },
    {
      "entropy": 5.814316701889038,
      "epoch": 5.760936710282473,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022500212980622818,
      "loss": 5.0168,
      "mean_token_accuracy": 0.22756853252649306,
      "num_tokens": 114504477.0,
      "step": 51295
    },
    {
      "entropy": 5.763852643966675,
      "epoch": 5.761498287190431,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00022496298567361724,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23538397550582885,
      "num_tokens": 114515070.0,
      "step": 51300
    },
    {
      "entropy": 5.661612033843994,
      "epoch": 5.762059864098388,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022492384313420416,
      "loss": 4.82,
      "mean_token_accuracy": 0.23650692999362946,
      "num_tokens": 114525219.0,
      "step": 51305
    },
    {
      "entropy": 5.734990072250366,
      "epoch": 5.762621441006345,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022488470218923534,
      "loss": 4.902,
      "mean_token_accuracy": 0.23315127193927765,
      "num_tokens": 114536435.0,
      "step": 51310
    },
    {
      "entropy": 5.656395339965821,
      "epoch": 5.7631830179143035,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022484556283995706,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24230635315179824,
      "num_tokens": 114547048.0,
      "step": 51315
    },
    {
      "entropy": 5.667955160140991,
      "epoch": 5.763744594822261,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002248064250876155,
      "loss": 4.8448,
      "mean_token_accuracy": 0.24293564558029174,
      "num_tokens": 114557926.0,
      "step": 51320
    },
    {
      "entropy": 5.613663578033448,
      "epoch": 5.764306171730219,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022476728893345687,
      "loss": 4.8387,
      "mean_token_accuracy": 0.2315140560269356,
      "num_tokens": 114568811.0,
      "step": 51325
    },
    {
      "entropy": 5.738299894332886,
      "epoch": 5.764867748638176,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022472815437872724,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23833830058574676,
      "num_tokens": 114579668.0,
      "step": 51330
    },
    {
      "entropy": 5.736721324920654,
      "epoch": 5.765429325546133,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022468902142467267,
      "loss": 4.8662,
      "mean_token_accuracy": 0.245122466981411,
      "num_tokens": 114590368.0,
      "step": 51335
    },
    {
      "entropy": 5.705243444442749,
      "epoch": 5.765990902454091,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022464989007253922,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2470136970281601,
      "num_tokens": 114602004.0,
      "step": 51340
    },
    {
      "entropy": 5.737397289276123,
      "epoch": 5.766552479362049,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002246107603235727,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22928195148706437,
      "num_tokens": 114613068.0,
      "step": 51345
    },
    {
      "entropy": 5.6933496475219725,
      "epoch": 5.767114056270006,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00022457163217901916,
      "loss": 4.9053,
      "mean_token_accuracy": 0.233658304810524,
      "num_tokens": 114623408.0,
      "step": 51350
    },
    {
      "entropy": 5.752229404449463,
      "epoch": 5.767675633177964,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002245325056401244,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23519800454378129,
      "num_tokens": 114636416.0,
      "step": 51355
    },
    {
      "entropy": 5.84016604423523,
      "epoch": 5.768237210085921,
      "grad_norm": 1.390625,
      "learning_rate": 0.00022449338070813424,
      "loss": 4.996,
      "mean_token_accuracy": 0.23347003310918807,
      "num_tokens": 114648582.0,
      "step": 51360
    },
    {
      "entropy": 5.724435472488404,
      "epoch": 5.768798786993878,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022445425738429442,
      "loss": 4.8312,
      "mean_token_accuracy": 0.2408464714884758,
      "num_tokens": 114659491.0,
      "step": 51365
    },
    {
      "entropy": 5.6683272361755375,
      "epoch": 5.7693603639018365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00022441513566985068,
      "loss": 4.8546,
      "mean_token_accuracy": 0.2349257081747055,
      "num_tokens": 114671377.0,
      "step": 51370
    },
    {
      "entropy": 5.7554339408874515,
      "epoch": 5.769921940809794,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00022437601556604865,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2371269643306732,
      "num_tokens": 114682352.0,
      "step": 51375
    },
    {
      "entropy": 5.778681707382202,
      "epoch": 5.770483517717752,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022433689707413373,
      "loss": 4.9876,
      "mean_token_accuracy": 0.23300490975379945,
      "num_tokens": 114693449.0,
      "step": 51380
    },
    {
      "entropy": 5.736691570281982,
      "epoch": 5.771045094625709,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022429778019535186,
      "loss": 4.8515,
      "mean_token_accuracy": 0.24486321955919266,
      "num_tokens": 114705344.0,
      "step": 51385
    },
    {
      "entropy": 5.635157632827759,
      "epoch": 5.771606671533666,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022425866493094827,
      "loss": 4.8289,
      "mean_token_accuracy": 0.2306130051612854,
      "num_tokens": 114715416.0,
      "step": 51390
    },
    {
      "entropy": 5.721509552001953,
      "epoch": 5.772168248441624,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022421955128216854,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23617562055587768,
      "num_tokens": 114727697.0,
      "step": 51395
    },
    {
      "entropy": 5.731263113021851,
      "epoch": 5.772729825349582,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002241804392502579,
      "loss": 4.93,
      "mean_token_accuracy": 0.23053082078695297,
      "num_tokens": 114739157.0,
      "step": 51400
    },
    {
      "entropy": 5.714481544494629,
      "epoch": 5.773291402257539,
      "grad_norm": 1.21875,
      "learning_rate": 0.00022414132883646183,
      "loss": 4.9393,
      "mean_token_accuracy": 0.23238437622785568,
      "num_tokens": 114750380.0,
      "step": 51405
    },
    {
      "entropy": 5.760295486450195,
      "epoch": 5.773852979165497,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022410222004202557,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23954906314611435,
      "num_tokens": 114761355.0,
      "step": 51410
    },
    {
      "entropy": 5.714566278457641,
      "epoch": 5.774414556073454,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002240631128681943,
      "loss": 4.9525,
      "mean_token_accuracy": 0.2265983536839485,
      "num_tokens": 114772336.0,
      "step": 51415
    },
    {
      "entropy": 5.62793664932251,
      "epoch": 5.774976132981411,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022402400731621336,
      "loss": 4.7979,
      "mean_token_accuracy": 0.2443908527493477,
      "num_tokens": 114782751.0,
      "step": 51420
    },
    {
      "entropy": 5.730263519287109,
      "epoch": 5.775537709889369,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002239849033873278,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22567692548036575,
      "num_tokens": 114793301.0,
      "step": 51425
    },
    {
      "entropy": 5.682091188430786,
      "epoch": 5.776099286797327,
      "grad_norm": 1.234375,
      "learning_rate": 0.00022394580108278263,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2247866839170456,
      "num_tokens": 114804697.0,
      "step": 51430
    },
    {
      "entropy": 5.642593574523926,
      "epoch": 5.776660863705285,
      "grad_norm": 2.078125,
      "learning_rate": 0.000223906700403823,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24931965619325638,
      "num_tokens": 114814994.0,
      "step": 51435
    },
    {
      "entropy": 5.707507801055908,
      "epoch": 5.777222440613242,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022386760135169388,
      "loss": 4.8817,
      "mean_token_accuracy": 0.23722273260354995,
      "num_tokens": 114825843.0,
      "step": 51440
    },
    {
      "entropy": 5.740588855743408,
      "epoch": 5.777784017521199,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00022382850392764015,
      "loss": 4.9369,
      "mean_token_accuracy": 0.2402213215827942,
      "num_tokens": 114836936.0,
      "step": 51445
    },
    {
      "entropy": 5.672879552841186,
      "epoch": 5.778345594429157,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022378940813290666,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2406810387969017,
      "num_tokens": 114848783.0,
      "step": 51450
    },
    {
      "entropy": 5.751715850830078,
      "epoch": 5.7789071713371145,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022375031396873818,
      "loss": 4.9288,
      "mean_token_accuracy": 0.24397983849048616,
      "num_tokens": 114860424.0,
      "step": 51455
    },
    {
      "entropy": 5.705823707580566,
      "epoch": 5.779468748245073,
      "grad_norm": 1.421875,
      "learning_rate": 0.00022371122143637973,
      "loss": 4.8896,
      "mean_token_accuracy": 0.23260789066553117,
      "num_tokens": 114870706.0,
      "step": 51460
    },
    {
      "entropy": 5.7574462890625,
      "epoch": 5.78003032515303,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022367213053707587,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22663760781288148,
      "num_tokens": 114880401.0,
      "step": 51465
    },
    {
      "entropy": 5.734952831268311,
      "epoch": 5.780591902060987,
      "grad_norm": 1.390625,
      "learning_rate": 0.00022363304127207123,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2380670949816704,
      "num_tokens": 114891276.0,
      "step": 51470
    },
    {
      "entropy": 5.705136632919311,
      "epoch": 5.781153478968945,
      "grad_norm": 1.34375,
      "learning_rate": 0.00022359395364261048,
      "loss": 4.8423,
      "mean_token_accuracy": 0.24252893328666686,
      "num_tokens": 114902545.0,
      "step": 51475
    },
    {
      "entropy": 5.7601830005645756,
      "epoch": 5.781715055876902,
      "grad_norm": 1.25,
      "learning_rate": 0.00022355486764993822,
      "loss": 4.9218,
      "mean_token_accuracy": 0.23173656314611435,
      "num_tokens": 114913366.0,
      "step": 51480
    },
    {
      "entropy": 5.673405361175537,
      "epoch": 5.78227663278486,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00022351578329529893,
      "loss": 4.8178,
      "mean_token_accuracy": 0.2403123676776886,
      "num_tokens": 114924213.0,
      "step": 51485
    },
    {
      "entropy": 5.788849687576294,
      "epoch": 5.782838209692818,
      "grad_norm": 1.359375,
      "learning_rate": 0.000223476700579937,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22480319887399675,
      "num_tokens": 114935156.0,
      "step": 51490
    },
    {
      "entropy": 5.691527843475342,
      "epoch": 5.783399786600775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000223437619505097,
      "loss": 4.9916,
      "mean_token_accuracy": 0.22785030901432038,
      "num_tokens": 114946526.0,
      "step": 51495
    },
    {
      "entropy": 5.7652122497558596,
      "epoch": 5.783961363508732,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00022339854007202316,
      "loss": 4.9607,
      "mean_token_accuracy": 0.23385505378246307,
      "num_tokens": 114958211.0,
      "step": 51500
    },
    {
      "entropy": 5.699865865707397,
      "epoch": 5.78452294041669,
      "grad_norm": 1.25,
      "learning_rate": 0.0002233594622819598,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23616321980953217,
      "num_tokens": 114969172.0,
      "step": 51505
    },
    {
      "entropy": 5.789093112945556,
      "epoch": 5.7850845173246475,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022332038613615118,
      "loss": 5.0266,
      "mean_token_accuracy": 0.23039894700050353,
      "num_tokens": 114981198.0,
      "step": 51510
    },
    {
      "entropy": 5.774710750579834,
      "epoch": 5.785646094232606,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00022328131163584154,
      "loss": 4.8901,
      "mean_token_accuracy": 0.240580278635025,
      "num_tokens": 114992937.0,
      "step": 51515
    },
    {
      "entropy": 5.781042289733887,
      "epoch": 5.786207671140563,
      "grad_norm": 1.3125,
      "learning_rate": 0.000223242238782275,
      "loss": 4.9427,
      "mean_token_accuracy": 0.23374650478363038,
      "num_tokens": 115003833.0,
      "step": 51520
    },
    {
      "entropy": 5.775661563873291,
      "epoch": 5.78676924804852,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00022320316757669557,
      "loss": 4.9717,
      "mean_token_accuracy": 0.2331201732158661,
      "num_tokens": 115014516.0,
      "step": 51525
    },
    {
      "entropy": 5.7559229850769045,
      "epoch": 5.787330824956478,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022316409802034743,
      "loss": 4.9456,
      "mean_token_accuracy": 0.2325296863913536,
      "num_tokens": 115025965.0,
      "step": 51530
    },
    {
      "entropy": 5.672036504745483,
      "epoch": 5.787892401864435,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002231250301144745,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23173330575227738,
      "num_tokens": 115037094.0,
      "step": 51535
    },
    {
      "entropy": 5.646825122833252,
      "epoch": 5.7884539787723925,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002230859638603208,
      "loss": 4.8563,
      "mean_token_accuracy": 0.23868732452392577,
      "num_tokens": 115047956.0,
      "step": 51540
    },
    {
      "entropy": 5.790056419372559,
      "epoch": 5.789015555680351,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022304689925913007,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2348768398165703,
      "num_tokens": 115061123.0,
      "step": 51545
    },
    {
      "entropy": 5.7664307117462155,
      "epoch": 5.789577132588308,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022300783631214627,
      "loss": 4.9665,
      "mean_token_accuracy": 0.23530775755643846,
      "num_tokens": 115072379.0,
      "step": 51550
    },
    {
      "entropy": 5.762107801437378,
      "epoch": 5.790138709496265,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022296877502061307,
      "loss": 4.9558,
      "mean_token_accuracy": 0.23761643022298812,
      "num_tokens": 115083623.0,
      "step": 51555
    },
    {
      "entropy": 5.728553915023804,
      "epoch": 5.790700286404223,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022292971538577427,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2307760611176491,
      "num_tokens": 115094927.0,
      "step": 51560
    },
    {
      "entropy": 5.704002571105957,
      "epoch": 5.79126186331218,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022289065740887354,
      "loss": 4.8465,
      "mean_token_accuracy": 0.23318640589714051,
      "num_tokens": 115104848.0,
      "step": 51565
    },
    {
      "entropy": 5.714755535125732,
      "epoch": 5.7918234402201385,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00022285160109115448,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2404559686779976,
      "num_tokens": 115115602.0,
      "step": 51570
    },
    {
      "entropy": 5.72626485824585,
      "epoch": 5.792385017128096,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00022281254643386072,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2416672945022583,
      "num_tokens": 115126589.0,
      "step": 51575
    },
    {
      "entropy": 5.709738063812256,
      "epoch": 5.792946594036053,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022277349343823567,
      "loss": 4.8154,
      "mean_token_accuracy": 0.24362429976463318,
      "num_tokens": 115136220.0,
      "step": 51580
    },
    {
      "entropy": 5.67528977394104,
      "epoch": 5.793508170944011,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022273444210552286,
      "loss": 4.819,
      "mean_token_accuracy": 0.24249829351902008,
      "num_tokens": 115147420.0,
      "step": 51585
    },
    {
      "entropy": 5.589626598358154,
      "epoch": 5.794069747851968,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022269539243696568,
      "loss": 4.7107,
      "mean_token_accuracy": 0.24641807973384858,
      "num_tokens": 115157573.0,
      "step": 51590
    },
    {
      "entropy": 5.684301519393921,
      "epoch": 5.7946313247599255,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00022265634443380743,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23576277643442153,
      "num_tokens": 115167454.0,
      "step": 51595
    },
    {
      "entropy": 5.72449426651001,
      "epoch": 5.795192901667884,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00022261729809729152,
      "loss": 4.8399,
      "mean_token_accuracy": 0.24254648983478547,
      "num_tokens": 115177590.0,
      "step": 51600
    },
    {
      "entropy": 5.635267162322998,
      "epoch": 5.795754478575841,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022257825342866116,
      "loss": 4.8564,
      "mean_token_accuracy": 0.23885189592838288,
      "num_tokens": 115188596.0,
      "step": 51605
    },
    {
      "entropy": 5.74009485244751,
      "epoch": 5.796316055483798,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002225392104291596,
      "loss": 4.947,
      "mean_token_accuracy": 0.2357650652527809,
      "num_tokens": 115200695.0,
      "step": 51610
    },
    {
      "entropy": 5.782014799118042,
      "epoch": 5.796877632391756,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002225001691000299,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2302143707871437,
      "num_tokens": 115212122.0,
      "step": 51615
    },
    {
      "entropy": 5.765129852294922,
      "epoch": 5.797439209299713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002224611294425151,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23749928325414657,
      "num_tokens": 115223969.0,
      "step": 51620
    },
    {
      "entropy": 5.65385251045227,
      "epoch": 5.7980007862076715,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002224220914578584,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23750011771917343,
      "num_tokens": 115236119.0,
      "step": 51625
    },
    {
      "entropy": 5.684053707122803,
      "epoch": 5.798562363115629,
      "grad_norm": 1.375,
      "learning_rate": 0.00022238305514730267,
      "loss": 4.893,
      "mean_token_accuracy": 0.23006889671087266,
      "num_tokens": 115247369.0,
      "step": 51630
    },
    {
      "entropy": 5.752643156051636,
      "epoch": 5.799123940023586,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022234402051209086,
      "loss": 5.0135,
      "mean_token_accuracy": 0.23262586444616318,
      "num_tokens": 115258145.0,
      "step": 51635
    },
    {
      "entropy": 5.725398588180542,
      "epoch": 5.799685516931544,
      "grad_norm": 1.5,
      "learning_rate": 0.00022230498755346583,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23959910422563552,
      "num_tokens": 115270377.0,
      "step": 51640
    },
    {
      "entropy": 5.724246835708618,
      "epoch": 5.800247093839501,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022226595627267048,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23569829761981964,
      "num_tokens": 115282164.0,
      "step": 51645
    },
    {
      "entropy": 5.80714168548584,
      "epoch": 5.800808670747459,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022222692667094756,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2310805857181549,
      "num_tokens": 115293262.0,
      "step": 51650
    },
    {
      "entropy": 5.703454256057739,
      "epoch": 5.8013702476554165,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022218789874953976,
      "loss": 4.8329,
      "mean_token_accuracy": 0.23970262408256532,
      "num_tokens": 115303078.0,
      "step": 51655
    },
    {
      "entropy": 5.775246429443359,
      "epoch": 5.801931824563374,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022214887250968974,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2309309020638466,
      "num_tokens": 115314785.0,
      "step": 51660
    },
    {
      "entropy": 5.757782125473023,
      "epoch": 5.802493401471332,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002221098479526401,
      "loss": 4.9338,
      "mean_token_accuracy": 0.23755278438329697,
      "num_tokens": 115327139.0,
      "step": 51665
    },
    {
      "entropy": 5.797751951217651,
      "epoch": 5.803054978379289,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00022207082507963327,
      "loss": 4.9269,
      "mean_token_accuracy": 0.23152697086334229,
      "num_tokens": 115337511.0,
      "step": 51670
    },
    {
      "entropy": 5.567119789123535,
      "epoch": 5.803616555287246,
      "grad_norm": 1.21875,
      "learning_rate": 0.00022203180389191208,
      "loss": 4.7726,
      "mean_token_accuracy": 0.23950911015272142,
      "num_tokens": 115347583.0,
      "step": 51675
    },
    {
      "entropy": 5.631263875961304,
      "epoch": 5.804178132195204,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022199278439071873,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23832450658082963,
      "num_tokens": 115358407.0,
      "step": 51680
    },
    {
      "entropy": 5.7484677791595455,
      "epoch": 5.804739709103162,
      "grad_norm": 1.21875,
      "learning_rate": 0.00022195376657729578,
      "loss": 4.9864,
      "mean_token_accuracy": 0.228875233232975,
      "num_tokens": 115371470.0,
      "step": 51685
    },
    {
      "entropy": 5.73886251449585,
      "epoch": 5.805301286011119,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022191475045288533,
      "loss": 4.8842,
      "mean_token_accuracy": 0.24220335930585862,
      "num_tokens": 115382017.0,
      "step": 51690
    },
    {
      "entropy": 5.7706798076629635,
      "epoch": 5.805862862919077,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022187573601872994,
      "loss": 4.9273,
      "mean_token_accuracy": 0.23720477223396302,
      "num_tokens": 115393222.0,
      "step": 51695
    },
    {
      "entropy": 5.764036226272583,
      "epoch": 5.806424439827034,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00022183672327607168,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2313613012433052,
      "num_tokens": 115404527.0,
      "step": 51700
    },
    {
      "entropy": 5.767170715332031,
      "epoch": 5.806986016734992,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022179771222615268,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23276442736387254,
      "num_tokens": 115416575.0,
      "step": 51705
    },
    {
      "entropy": 5.79401273727417,
      "epoch": 5.8075475936429495,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022175870287021528,
      "loss": 4.973,
      "mean_token_accuracy": 0.23538828343153,
      "num_tokens": 115428006.0,
      "step": 51710
    },
    {
      "entropy": 5.702192258834839,
      "epoch": 5.808109170550907,
      "grad_norm": 1.171875,
      "learning_rate": 0.00022171969520950143,
      "loss": 4.883,
      "mean_token_accuracy": 0.2358553722500801,
      "num_tokens": 115438962.0,
      "step": 51715
    },
    {
      "entropy": 5.660125207901001,
      "epoch": 5.808670747458865,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022168068924525308,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2309230461716652,
      "num_tokens": 115450246.0,
      "step": 51720
    },
    {
      "entropy": 5.709968376159668,
      "epoch": 5.809232324366822,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00022164168497871227,
      "loss": 4.8675,
      "mean_token_accuracy": 0.23970804661512374,
      "num_tokens": 115461292.0,
      "step": 51725
    },
    {
      "entropy": 5.762299823760986,
      "epoch": 5.809793901274779,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022160268241112096,
      "loss": 4.9398,
      "mean_token_accuracy": 0.23448225408792495,
      "num_tokens": 115473193.0,
      "step": 51730
    },
    {
      "entropy": 5.738122081756591,
      "epoch": 5.810355478182737,
      "grad_norm": 1.25,
      "learning_rate": 0.0002215636815437209,
      "loss": 4.8768,
      "mean_token_accuracy": 0.24377911686897277,
      "num_tokens": 115483515.0,
      "step": 51735
    },
    {
      "entropy": 5.674794483184814,
      "epoch": 5.810917055090695,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00022152468237775387,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2295625999569893,
      "num_tokens": 115493915.0,
      "step": 51740
    },
    {
      "entropy": 5.805453491210938,
      "epoch": 5.811478631998652,
      "grad_norm": 1.15625,
      "learning_rate": 0.00022148568491446181,
      "loss": 5.0576,
      "mean_token_accuracy": 0.232759690284729,
      "num_tokens": 115507423.0,
      "step": 51745
    },
    {
      "entropy": 5.6781086921691895,
      "epoch": 5.81204020890661,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022144668915508625,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23663035333156585,
      "num_tokens": 115518648.0,
      "step": 51750
    },
    {
      "entropy": 5.727983903884888,
      "epoch": 5.812601785814567,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00022140769510086894,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2378190651535988,
      "num_tokens": 115529898.0,
      "step": 51755
    },
    {
      "entropy": 5.765877723693848,
      "epoch": 5.813163362722525,
      "grad_norm": 1.375,
      "learning_rate": 0.0002213687027530514,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23788727819919586,
      "num_tokens": 115538823.0,
      "step": 51760
    },
    {
      "entropy": 5.752447032928467,
      "epoch": 5.813724939630482,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002213297121128751,
      "loss": 4.964,
      "mean_token_accuracy": 0.23960501849651336,
      "num_tokens": 115548983.0,
      "step": 51765
    },
    {
      "entropy": 5.707715606689453,
      "epoch": 5.81428651653844,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002212907231815816,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2339496210217476,
      "num_tokens": 115559716.0,
      "step": 51770
    },
    {
      "entropy": 5.73499755859375,
      "epoch": 5.814848093446398,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022125173596041232,
      "loss": 4.9367,
      "mean_token_accuracy": 0.23936388939619063,
      "num_tokens": 115570797.0,
      "step": 51775
    },
    {
      "entropy": 5.755304574966431,
      "epoch": 5.815409670354355,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022121275045060857,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22895717471837998,
      "num_tokens": 115582701.0,
      "step": 51780
    },
    {
      "entropy": 5.754820680618286,
      "epoch": 5.815971247262313,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022117376665341178,
      "loss": 4.8806,
      "mean_token_accuracy": 0.2413788065314293,
      "num_tokens": 115594072.0,
      "step": 51785
    },
    {
      "entropy": 5.828810024261474,
      "epoch": 5.81653282417027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00022113478457006308,
      "loss": 4.9574,
      "mean_token_accuracy": 0.232246296107769,
      "num_tokens": 115605063.0,
      "step": 51790
    },
    {
      "entropy": 5.742793941497803,
      "epoch": 5.8170944010782275,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022109580420180374,
      "loss": 4.8809,
      "mean_token_accuracy": 0.24036475867033005,
      "num_tokens": 115616445.0,
      "step": 51795
    },
    {
      "entropy": 5.707271099090576,
      "epoch": 5.817655977986185,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022105682554987493,
      "loss": 4.9253,
      "mean_token_accuracy": 0.23030034452676773,
      "num_tokens": 115627289.0,
      "step": 51800
    },
    {
      "entropy": 5.721210527420044,
      "epoch": 5.818217554894143,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002210178486155177,
      "loss": 4.9938,
      "mean_token_accuracy": 0.23105955868959427,
      "num_tokens": 115638609.0,
      "step": 51805
    },
    {
      "entropy": 5.810652017593384,
      "epoch": 5.8187791318021,
      "grad_norm": 1.234375,
      "learning_rate": 0.00022097887339997312,
      "loss": 4.9844,
      "mean_token_accuracy": 0.23313872665166854,
      "num_tokens": 115649496.0,
      "step": 51810
    },
    {
      "entropy": 5.718262434005737,
      "epoch": 5.819340708710058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00022093989990448203,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2433892771601677,
      "num_tokens": 115660467.0,
      "step": 51815
    },
    {
      "entropy": 5.690379428863525,
      "epoch": 5.819902285618015,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022090092813028563,
      "loss": 4.8326,
      "mean_token_accuracy": 0.23293564915657045,
      "num_tokens": 115670795.0,
      "step": 51820
    },
    {
      "entropy": 5.715802001953125,
      "epoch": 5.820463862525973,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00022086195807862464,
      "loss": 4.9646,
      "mean_token_accuracy": 0.23638884723186493,
      "num_tokens": 115682058.0,
      "step": 51825
    },
    {
      "entropy": 5.677052974700928,
      "epoch": 5.821025439433931,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022082298975073988,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22966777980327607,
      "num_tokens": 115693492.0,
      "step": 51830
    },
    {
      "entropy": 5.731157922744751,
      "epoch": 5.821587016341888,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002207840231478722,
      "loss": 4.8857,
      "mean_token_accuracy": 0.238027361035347,
      "num_tokens": 115706461.0,
      "step": 51835
    },
    {
      "entropy": 5.843856477737427,
      "epoch": 5.822148593249846,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022074505827126222,
      "loss": 5.0226,
      "mean_token_accuracy": 0.23177628517150878,
      "num_tokens": 115718798.0,
      "step": 51840
    },
    {
      "entropy": 5.803911590576172,
      "epoch": 5.822710170157803,
      "grad_norm": 1.3125,
      "learning_rate": 0.00022070609512215062,
      "loss": 4.923,
      "mean_token_accuracy": 0.23103905022144317,
      "num_tokens": 115730032.0,
      "step": 51845
    },
    {
      "entropy": 5.7361180782318115,
      "epoch": 5.8232717470657605,
      "grad_norm": 1.234375,
      "learning_rate": 0.000220667133701778,
      "loss": 4.8769,
      "mean_token_accuracy": 0.24436476230621337,
      "num_tokens": 115741676.0,
      "step": 51850
    },
    {
      "entropy": 5.7384514808654785,
      "epoch": 5.823833323973719,
      "grad_norm": 1.328125,
      "learning_rate": 0.00022062817401138492,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23658860623836517,
      "num_tokens": 115751726.0,
      "step": 51855
    },
    {
      "entropy": 5.690083980560303,
      "epoch": 5.824394900881676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022058921605221189,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23652833551168442,
      "num_tokens": 115763413.0,
      "step": 51860
    },
    {
      "entropy": 5.670993232727051,
      "epoch": 5.824956477789633,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022055025982549924,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23511035442352296,
      "num_tokens": 115774335.0,
      "step": 51865
    },
    {
      "entropy": 5.708130264282227,
      "epoch": 5.825518054697591,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022051130533248754,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23713887184858323,
      "num_tokens": 115785645.0,
      "step": 51870
    },
    {
      "entropy": 5.7154176235198975,
      "epoch": 5.826079631605548,
      "grad_norm": 1.390625,
      "learning_rate": 0.000220472352574417,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2424347534775734,
      "num_tokens": 115797640.0,
      "step": 51875
    },
    {
      "entropy": 5.725360679626465,
      "epoch": 5.8266412085135055,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00022043340155252788,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23936520218849183,
      "num_tokens": 115808724.0,
      "step": 51880
    },
    {
      "entropy": 5.656923055648804,
      "epoch": 5.827202785421464,
      "grad_norm": 1.28125,
      "learning_rate": 0.00022039445226806033,
      "loss": 4.8522,
      "mean_token_accuracy": 0.24645356684923173,
      "num_tokens": 115819003.0,
      "step": 51885
    },
    {
      "entropy": 5.748033046722412,
      "epoch": 5.827764362329421,
      "grad_norm": 1.25,
      "learning_rate": 0.00022035550472225467,
      "loss": 4.8824,
      "mean_token_accuracy": 0.24260275512933732,
      "num_tokens": 115830630.0,
      "step": 51890
    },
    {
      "entropy": 5.6986291885375975,
      "epoch": 5.828325939237379,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00022031655891635094,
      "loss": 4.9926,
      "mean_token_accuracy": 0.228540301322937,
      "num_tokens": 115840899.0,
      "step": 51895
    },
    {
      "entropy": 5.696767997741699,
      "epoch": 5.828887516145336,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002202776148515892,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22343092411756516,
      "num_tokens": 115852947.0,
      "step": 51900
    },
    {
      "entropy": 5.712741327285767,
      "epoch": 5.829449093053293,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00022023867252920942,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23813854455947875,
      "num_tokens": 115865019.0,
      "step": 51905
    },
    {
      "entropy": 5.749134349822998,
      "epoch": 5.8300106699612515,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00022019973195045155,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2310641437768936,
      "num_tokens": 115876158.0,
      "step": 51910
    },
    {
      "entropy": 5.753592491149902,
      "epoch": 5.830572246869209,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022016079311655546,
      "loss": 4.9414,
      "mean_token_accuracy": 0.23597004264593124,
      "num_tokens": 115887037.0,
      "step": 51915
    },
    {
      "entropy": 5.728525543212891,
      "epoch": 5.831133823777166,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022012185602876094,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22401365935802459,
      "num_tokens": 115897940.0,
      "step": 51920
    },
    {
      "entropy": 5.705820560455322,
      "epoch": 5.831695400685124,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022008292068830787,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23455483317375184,
      "num_tokens": 115909107.0,
      "step": 51925
    },
    {
      "entropy": 5.752883338928223,
      "epoch": 5.832256977593081,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00022004398709643587,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23506007343530655,
      "num_tokens": 115919938.0,
      "step": 51930
    },
    {
      "entropy": 5.712595462799072,
      "epoch": 5.8328185545010385,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022000505525438462,
      "loss": 4.9026,
      "mean_token_accuracy": 0.23786091357469558,
      "num_tokens": 115931334.0,
      "step": 51935
    },
    {
      "entropy": 5.745816707611084,
      "epoch": 5.833380131408997,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021996612516339382,
      "loss": 4.8555,
      "mean_token_accuracy": 0.23347305208444596,
      "num_tokens": 115942283.0,
      "step": 51940
    },
    {
      "entropy": 5.73284797668457,
      "epoch": 5.833941708316954,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002199271968247029,
      "loss": 4.8627,
      "mean_token_accuracy": 0.23979801535606385,
      "num_tokens": 115952719.0,
      "step": 51945
    },
    {
      "entropy": 5.738713359832763,
      "epoch": 5.834503285224912,
      "grad_norm": 1.25,
      "learning_rate": 0.00021988827023955145,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2359532043337822,
      "num_tokens": 115964542.0,
      "step": 51950
    },
    {
      "entropy": 5.735721397399902,
      "epoch": 5.835064862132869,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021984934540917873,
      "loss": 4.9189,
      "mean_token_accuracy": 0.23784370422363282,
      "num_tokens": 115974915.0,
      "step": 51955
    },
    {
      "entropy": 5.721465301513672,
      "epoch": 5.835626439040826,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00021981042233482423,
      "loss": 4.8942,
      "mean_token_accuracy": 0.23493358343839646,
      "num_tokens": 115986277.0,
      "step": 51960
    },
    {
      "entropy": 5.661626195907592,
      "epoch": 5.8361880159487844,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021977150101772737,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23280384093523027,
      "num_tokens": 115997412.0,
      "step": 51965
    },
    {
      "entropy": 5.740731430053711,
      "epoch": 5.836749592856742,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021973258145912728,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22273778319358825,
      "num_tokens": 116008795.0,
      "step": 51970
    },
    {
      "entropy": 5.747291374206543,
      "epoch": 5.8373111697647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021969366366026327,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23693937659263611,
      "num_tokens": 116020900.0,
      "step": 51975
    },
    {
      "entropy": 5.777532577514648,
      "epoch": 5.837872746672657,
      "grad_norm": 1.375,
      "learning_rate": 0.00021965474762237447,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23831623941659927,
      "num_tokens": 116032617.0,
      "step": 51980
    },
    {
      "entropy": 5.7942126274108885,
      "epoch": 5.838434323580614,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002196158333467,
      "loss": 4.9139,
      "mean_token_accuracy": 0.2390740066766739,
      "num_tokens": 116042717.0,
      "step": 51985
    },
    {
      "entropy": 5.7137126445770265,
      "epoch": 5.838995900488571,
      "grad_norm": 1.25,
      "learning_rate": 0.00021957692083447884,
      "loss": 4.8952,
      "mean_token_accuracy": 0.23702953159809112,
      "num_tokens": 116055112.0,
      "step": 51990
    },
    {
      "entropy": 5.668834686279297,
      "epoch": 5.8395574773965295,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00021953801008694996,
      "loss": 4.8668,
      "mean_token_accuracy": 0.24470579177141188,
      "num_tokens": 116065452.0,
      "step": 51995
    },
    {
      "entropy": 5.789943885803223,
      "epoch": 5.840119054304487,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021949910110535242,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23187409937381745,
      "num_tokens": 116075293.0,
      "step": 52000
    },
    {
      "entropy": 5.801426124572754,
      "epoch": 5.840680631212445,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021946019389092503,
      "loss": 4.9903,
      "mean_token_accuracy": 0.23343756049871445,
      "num_tokens": 116088734.0,
      "step": 52005
    },
    {
      "entropy": 5.735320615768432,
      "epoch": 5.841242208120402,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002194212884449066,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23057173192501068,
      "num_tokens": 116099270.0,
      "step": 52010
    },
    {
      "entropy": 5.662947177886963,
      "epoch": 5.841803785028359,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002193823847685359,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2429459035396576,
      "num_tokens": 116110160.0,
      "step": 52015
    },
    {
      "entropy": 5.811154890060425,
      "epoch": 5.842365361936317,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021934348286305167,
      "loss": 4.9125,
      "mean_token_accuracy": 0.24146943390369416,
      "num_tokens": 116121267.0,
      "step": 52020
    },
    {
      "entropy": 5.696161890029908,
      "epoch": 5.842926938844275,
      "grad_norm": 1.125,
      "learning_rate": 0.00021930458272969243,
      "loss": 4.9013,
      "mean_token_accuracy": 0.24049097746610643,
      "num_tokens": 116133050.0,
      "step": 52025
    },
    {
      "entropy": 5.751045417785645,
      "epoch": 5.843488515752233,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00021926568436969685,
      "loss": 4.9904,
      "mean_token_accuracy": 0.23584576249122619,
      "num_tokens": 116144688.0,
      "step": 52030
    },
    {
      "entropy": 5.718641710281372,
      "epoch": 5.84405009266019,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002192267877843036,
      "loss": 4.8849,
      "mean_token_accuracy": 0.24614461064338683,
      "num_tokens": 116155641.0,
      "step": 52035
    },
    {
      "entropy": 5.674165725708008,
      "epoch": 5.844611669568147,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021918789297475106,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23947302848100663,
      "num_tokens": 116166023.0,
      "step": 52040
    },
    {
      "entropy": 5.655766391754151,
      "epoch": 5.845173246476105,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00021914899994227767,
      "loss": 4.7984,
      "mean_token_accuracy": 0.24083804190158845,
      "num_tokens": 116176875.0,
      "step": 52045
    },
    {
      "entropy": 5.849447011947632,
      "epoch": 5.8457348233840625,
      "grad_norm": 1.25,
      "learning_rate": 0.00021911010868812174,
      "loss": 5.069,
      "mean_token_accuracy": 0.22750101387500762,
      "num_tokens": 116188187.0,
      "step": 52050
    },
    {
      "entropy": 5.771682453155518,
      "epoch": 5.84629640029202,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021907121921352162,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2249685913324356,
      "num_tokens": 116199790.0,
      "step": 52055
    },
    {
      "entropy": 5.7202404022216795,
      "epoch": 5.846857977199978,
      "grad_norm": 1.140625,
      "learning_rate": 0.00021903233151971562,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23557813614606857,
      "num_tokens": 116211628.0,
      "step": 52060
    },
    {
      "entropy": 5.808785009384155,
      "epoch": 5.847419554107935,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021899344560794183,
      "loss": 4.9796,
      "mean_token_accuracy": 0.23865836560726167,
      "num_tokens": 116223041.0,
      "step": 52065
    },
    {
      "entropy": 5.769030475616455,
      "epoch": 5.847981131015892,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021895456147943848,
      "loss": 4.9274,
      "mean_token_accuracy": 0.235689440369606,
      "num_tokens": 116233926.0,
      "step": 52070
    },
    {
      "entropy": 5.71126389503479,
      "epoch": 5.84854270792385,
      "grad_norm": 1.328125,
      "learning_rate": 0.00021891567913544363,
      "loss": 4.9028,
      "mean_token_accuracy": 0.23094141185283662,
      "num_tokens": 116244069.0,
      "step": 52075
    },
    {
      "entropy": 5.705572509765625,
      "epoch": 5.8491042848318076,
      "grad_norm": 1.453125,
      "learning_rate": 0.00021887679857719535,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2346939280629158,
      "num_tokens": 116253622.0,
      "step": 52080
    },
    {
      "entropy": 5.685622596740723,
      "epoch": 5.849665861739766,
      "grad_norm": 1.1875,
      "learning_rate": 0.00021883791980593153,
      "loss": 4.869,
      "mean_token_accuracy": 0.24023250490427017,
      "num_tokens": 116265295.0,
      "step": 52085
    },
    {
      "entropy": 5.782687950134277,
      "epoch": 5.850227438647723,
      "grad_norm": 1.21875,
      "learning_rate": 0.00021879904282289008,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23947916477918624,
      "num_tokens": 116276600.0,
      "step": 52090
    },
    {
      "entropy": 5.726242876052856,
      "epoch": 5.85078901555568,
      "grad_norm": 1.4375,
      "learning_rate": 0.000218760167629309,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23977265805006026,
      "num_tokens": 116287678.0,
      "step": 52095
    },
    {
      "entropy": 5.79821629524231,
      "epoch": 5.851350592463638,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002187212942264259,
      "loss": 5.0188,
      "mean_token_accuracy": 0.22560801953077317,
      "num_tokens": 116300221.0,
      "step": 52100
    },
    {
      "entropy": 5.779354047775269,
      "epoch": 5.851912169371595,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002186824226154786,
      "loss": 5.0353,
      "mean_token_accuracy": 0.22581297010183335,
      "num_tokens": 116312484.0,
      "step": 52105
    },
    {
      "entropy": 5.736456632614136,
      "epoch": 5.852473746279553,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00021864355279770494,
      "loss": 4.9059,
      "mean_token_accuracy": 0.24561125338077544,
      "num_tokens": 116323416.0,
      "step": 52110
    },
    {
      "entropy": 5.7634063243865965,
      "epoch": 5.853035323187511,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002186046847743423,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22999758571386336,
      "num_tokens": 116334621.0,
      "step": 52115
    },
    {
      "entropy": 5.753727960586548,
      "epoch": 5.853596900095468,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021856581854662845,
      "loss": 4.9502,
      "mean_token_accuracy": 0.23830899298191072,
      "num_tokens": 116346375.0,
      "step": 52120
    },
    {
      "entropy": 5.813920021057129,
      "epoch": 5.854158477003425,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021852695411580083,
      "loss": 4.916,
      "mean_token_accuracy": 0.23910043090581895,
      "num_tokens": 116357377.0,
      "step": 52125
    },
    {
      "entropy": 5.802077865600586,
      "epoch": 5.854720053911383,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021848809148309684,
      "loss": 4.9232,
      "mean_token_accuracy": 0.23174482583999634,
      "num_tokens": 116368290.0,
      "step": 52130
    },
    {
      "entropy": 5.735279178619384,
      "epoch": 5.8552816308193405,
      "grad_norm": 1.21875,
      "learning_rate": 0.00021844923064975397,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22993292063474655,
      "num_tokens": 116379479.0,
      "step": 52135
    },
    {
      "entropy": 5.766698503494263,
      "epoch": 5.855843207727299,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00021841037161700943,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22728366106748582,
      "num_tokens": 116389769.0,
      "step": 52140
    },
    {
      "entropy": 5.808904361724854,
      "epoch": 5.856404784635256,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021837151438610074,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22813086360692977,
      "num_tokens": 116400072.0,
      "step": 52145
    },
    {
      "entropy": 5.864442825317383,
      "epoch": 5.856966361543213,
      "grad_norm": 1.40625,
      "learning_rate": 0.00021833265895826492,
      "loss": 5.0247,
      "mean_token_accuracy": 0.22602507174015046,
      "num_tokens": 116410531.0,
      "step": 52150
    },
    {
      "entropy": 5.750176668167114,
      "epoch": 5.857527938451171,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021829380533473924,
      "loss": 4.8573,
      "mean_token_accuracy": 0.24178213775157928,
      "num_tokens": 116421558.0,
      "step": 52155
    },
    {
      "entropy": 5.758966159820557,
      "epoch": 5.858089515359128,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021825495351676083,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23700737506151198,
      "num_tokens": 116432596.0,
      "step": 52160
    },
    {
      "entropy": 5.747045230865479,
      "epoch": 5.8586510922670865,
      "grad_norm": 1.40625,
      "learning_rate": 0.00021821610350556664,
      "loss": 4.9507,
      "mean_token_accuracy": 0.23714938312768935,
      "num_tokens": 116442870.0,
      "step": 52165
    },
    {
      "entropy": 5.658872175216675,
      "epoch": 5.859212669175044,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002181772553023938,
      "loss": 4.8043,
      "mean_token_accuracy": 0.23817805647850038,
      "num_tokens": 116453844.0,
      "step": 52170
    },
    {
      "entropy": 5.7583229541778564,
      "epoch": 5.859774246083001,
      "grad_norm": 1.375,
      "learning_rate": 0.00021813840890847904,
      "loss": 4.9088,
      "mean_token_accuracy": 0.24293164759874344,
      "num_tokens": 116465058.0,
      "step": 52175
    },
    {
      "entropy": 5.781431531906128,
      "epoch": 5.860335822990958,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002180995643250595,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2321265459060669,
      "num_tokens": 116476347.0,
      "step": 52180
    },
    {
      "entropy": 5.737598896026611,
      "epoch": 5.860897399898916,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002180607215533719,
      "loss": 4.8943,
      "mean_token_accuracy": 0.24192076325416564,
      "num_tokens": 116487792.0,
      "step": 52185
    },
    {
      "entropy": 5.66140079498291,
      "epoch": 5.8614589768068734,
      "grad_norm": 1.453125,
      "learning_rate": 0.00021802188059465304,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23779982477426528,
      "num_tokens": 116498617.0,
      "step": 52190
    },
    {
      "entropy": 5.754077482223511,
      "epoch": 5.8620205537148315,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002179830414501396,
      "loss": 4.9911,
      "mean_token_accuracy": 0.23520013093948364,
      "num_tokens": 116508717.0,
      "step": 52195
    },
    {
      "entropy": 5.687806510925293,
      "epoch": 5.862582130622789,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00021794420412106825,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2405487045645714,
      "num_tokens": 116520418.0,
      "step": 52200
    },
    {
      "entropy": 5.743371200561524,
      "epoch": 5.863143707530746,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021790536860867555,
      "loss": 4.9113,
      "mean_token_accuracy": 0.23379310071468354,
      "num_tokens": 116531137.0,
      "step": 52205
    },
    {
      "entropy": 5.7739356517791744,
      "epoch": 5.863705284438704,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000217866534914198,
      "loss": 5.0209,
      "mean_token_accuracy": 0.23389223963022232,
      "num_tokens": 116544070.0,
      "step": 52210
    },
    {
      "entropy": 5.756543350219727,
      "epoch": 5.864266861346661,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021782770303887218,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22889900654554368,
      "num_tokens": 116555433.0,
      "step": 52215
    },
    {
      "entropy": 5.717589473724365,
      "epoch": 5.864828438254619,
      "grad_norm": 1.4375,
      "learning_rate": 0.00021778887298393447,
      "loss": 4.8602,
      "mean_token_accuracy": 0.23395618349313735,
      "num_tokens": 116565985.0,
      "step": 52220
    },
    {
      "entropy": 5.791185712814331,
      "epoch": 5.865390015162577,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002177500447506212,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22771365642547609,
      "num_tokens": 116577958.0,
      "step": 52225
    },
    {
      "entropy": 5.682462120056153,
      "epoch": 5.865951592070534,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021771121834016877,
      "loss": 4.8083,
      "mean_token_accuracy": 0.2387933686375618,
      "num_tokens": 116588249.0,
      "step": 52230
    },
    {
      "entropy": 5.691770696640015,
      "epoch": 5.866513168978492,
      "grad_norm": 1.21875,
      "learning_rate": 0.00021767239375381338,
      "loss": 4.8795,
      "mean_token_accuracy": 0.24014303535223008,
      "num_tokens": 116599951.0,
      "step": 52235
    },
    {
      "entropy": 5.691661596298218,
      "epoch": 5.867074745886449,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021763357099279114,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23479262888431549,
      "num_tokens": 116611439.0,
      "step": 52240
    },
    {
      "entropy": 5.721334743499756,
      "epoch": 5.867636322794406,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00021759475005833818,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2360316962003708,
      "num_tokens": 116622611.0,
      "step": 52245
    },
    {
      "entropy": 5.726314115524292,
      "epoch": 5.8681978997023645,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00021755593095169074,
      "loss": 4.9023,
      "mean_token_accuracy": 0.23252425342798233,
      "num_tokens": 116635193.0,
      "step": 52250
    },
    {
      "entropy": 5.748399066925049,
      "epoch": 5.868759476610322,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002175171136740847,
      "loss": 4.979,
      "mean_token_accuracy": 0.23060130327939987,
      "num_tokens": 116647121.0,
      "step": 52255
    },
    {
      "entropy": 5.725357007980347,
      "epoch": 5.869321053518279,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021747829822675612,
      "loss": 4.8892,
      "mean_token_accuracy": 0.24192488044500352,
      "num_tokens": 116658642.0,
      "step": 52260
    },
    {
      "entropy": 5.761767244338989,
      "epoch": 5.869882630426237,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021743948461094083,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23351812213659287,
      "num_tokens": 116669723.0,
      "step": 52265
    },
    {
      "entropy": 5.792566967010498,
      "epoch": 5.870444207334194,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021740067282787463,
      "loss": 4.9175,
      "mean_token_accuracy": 0.23772228956222535,
      "num_tokens": 116681542.0,
      "step": 52270
    },
    {
      "entropy": 5.698757982254028,
      "epoch": 5.871005784242152,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021736186287879338,
      "loss": 4.866,
      "mean_token_accuracy": 0.2417410284280777,
      "num_tokens": 116692166.0,
      "step": 52275
    },
    {
      "entropy": 5.809066486358643,
      "epoch": 5.87156736115011,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002173230547649327,
      "loss": 5.0419,
      "mean_token_accuracy": 0.2305294692516327,
      "num_tokens": 116703945.0,
      "step": 52280
    },
    {
      "entropy": 5.800346088409424,
      "epoch": 5.872128938058067,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00021728424848752842,
      "loss": 4.97,
      "mean_token_accuracy": 0.22213521003723144,
      "num_tokens": 116714207.0,
      "step": 52285
    },
    {
      "entropy": 5.6677717685699465,
      "epoch": 5.872690514966025,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021724544404781598,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23674795627593995,
      "num_tokens": 116724488.0,
      "step": 52290
    },
    {
      "entropy": 5.702688121795655,
      "epoch": 5.873252091873982,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00021720664144703105,
      "loss": 4.8397,
      "mean_token_accuracy": 0.24259539991617202,
      "num_tokens": 116735358.0,
      "step": 52295
    },
    {
      "entropy": 5.718857955932617,
      "epoch": 5.873813668781939,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021716784068640904,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23032758235931397,
      "num_tokens": 116746724.0,
      "step": 52300
    },
    {
      "entropy": 5.70931658744812,
      "epoch": 5.874375245689897,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021712904176718545,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23813352584838868,
      "num_tokens": 116758072.0,
      "step": 52305
    },
    {
      "entropy": 5.769978141784668,
      "epoch": 5.874936822597855,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021709024469059564,
      "loss": 4.9195,
      "mean_token_accuracy": 0.2363759383559227,
      "num_tokens": 116769871.0,
      "step": 52310
    },
    {
      "entropy": 5.75670599937439,
      "epoch": 5.875498399505812,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021705144945787487,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2315710797905922,
      "num_tokens": 116780601.0,
      "step": 52315
    },
    {
      "entropy": 5.840106010437012,
      "epoch": 5.87605997641377,
      "grad_norm": 1.390625,
      "learning_rate": 0.00021701265607025828,
      "loss": 4.9863,
      "mean_token_accuracy": 0.23302415013313293,
      "num_tokens": 116792873.0,
      "step": 52320
    },
    {
      "entropy": 5.69204421043396,
      "epoch": 5.876621553321727,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021697386452898133,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23467378318309784,
      "num_tokens": 116804573.0,
      "step": 52325
    },
    {
      "entropy": 5.73369460105896,
      "epoch": 5.877183130229685,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00021693507483527908,
      "loss": 4.9338,
      "mean_token_accuracy": 0.23306455761194228,
      "num_tokens": 116816325.0,
      "step": 52330
    },
    {
      "entropy": 5.693931674957275,
      "epoch": 5.8777447071376425,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021689628699038648,
      "loss": 4.8808,
      "mean_token_accuracy": 0.24271277487277984,
      "num_tokens": 116828084.0,
      "step": 52335
    },
    {
      "entropy": 5.802163362503052,
      "epoch": 5.8783062840456,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021685750099553874,
      "loss": 4.9663,
      "mean_token_accuracy": 0.23294569700956344,
      "num_tokens": 116838701.0,
      "step": 52340
    },
    {
      "entropy": 5.719314908981323,
      "epoch": 5.878867860953558,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00021681871685197068,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2348212018609047,
      "num_tokens": 116849844.0,
      "step": 52345
    },
    {
      "entropy": 5.70814266204834,
      "epoch": 5.879429437861515,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021677993456091727,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23602585643529891,
      "num_tokens": 116860587.0,
      "step": 52350
    },
    {
      "entropy": 5.774470710754395,
      "epoch": 5.879991014769473,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021674115412361324,
      "loss": 5.0668,
      "mean_token_accuracy": 0.23141684383153915,
      "num_tokens": 116872758.0,
      "step": 52355
    },
    {
      "entropy": 5.72239122390747,
      "epoch": 5.88055259167743,
      "grad_norm": 1.25,
      "learning_rate": 0.00021670237554129347,
      "loss": 4.832,
      "mean_token_accuracy": 0.23986322283744813,
      "num_tokens": 116884382.0,
      "step": 52360
    },
    {
      "entropy": 5.725979661941528,
      "epoch": 5.881114168585388,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002166635988151927,
      "loss": 4.9471,
      "mean_token_accuracy": 0.23459414690732955,
      "num_tokens": 116897192.0,
      "step": 52365
    },
    {
      "entropy": 5.7964925289154055,
      "epoch": 5.881675745493345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002166248239465456,
      "loss": 5.0754,
      "mean_token_accuracy": 0.2258848249912262,
      "num_tokens": 116908898.0,
      "step": 52370
    },
    {
      "entropy": 5.728056573867798,
      "epoch": 5.882237322401303,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021658605093658674,
      "loss": 4.8716,
      "mean_token_accuracy": 0.24496365934610367,
      "num_tokens": 116919328.0,
      "step": 52375
    },
    {
      "entropy": 5.784745693206787,
      "epoch": 5.88279889930926,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021654727978655066,
      "loss": 4.989,
      "mean_token_accuracy": 0.22999406158924102,
      "num_tokens": 116930802.0,
      "step": 52380
    },
    {
      "entropy": 5.697006320953369,
      "epoch": 5.883360476217218,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00021650851049767185,
      "loss": 4.8883,
      "mean_token_accuracy": 0.236913001537323,
      "num_tokens": 116941890.0,
      "step": 52385
    },
    {
      "entropy": 5.816300964355468,
      "epoch": 5.8839220531251755,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021646974307118467,
      "loss": 5.0314,
      "mean_token_accuracy": 0.22872553318738936,
      "num_tokens": 116953938.0,
      "step": 52390
    },
    {
      "entropy": 5.720765161514282,
      "epoch": 5.884483630033133,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021643097750832365,
      "loss": 4.8598,
      "mean_token_accuracy": 0.231689690053463,
      "num_tokens": 116964451.0,
      "step": 52395
    },
    {
      "entropy": 5.829744815826416,
      "epoch": 5.885045206941091,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00021639221381032303,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22779752016067506,
      "num_tokens": 116975571.0,
      "step": 52400
    },
    {
      "entropy": 5.835440254211425,
      "epoch": 5.885606783849048,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021635345197841712,
      "loss": 4.9707,
      "mean_token_accuracy": 0.23649227917194365,
      "num_tokens": 116986652.0,
      "step": 52405
    },
    {
      "entropy": 5.782286167144775,
      "epoch": 5.886168360757006,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002163146920138399,
      "loss": 4.9787,
      "mean_token_accuracy": 0.23253896385431289,
      "num_tokens": 116997591.0,
      "step": 52410
    },
    {
      "entropy": 5.691976165771484,
      "epoch": 5.886729937664963,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021627593391782574,
      "loss": 4.856,
      "mean_token_accuracy": 0.24421856701374053,
      "num_tokens": 117009083.0,
      "step": 52415
    },
    {
      "entropy": 5.673316240310669,
      "epoch": 5.8872915145729205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021623717769160862,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2369464084506035,
      "num_tokens": 117020622.0,
      "step": 52420
    },
    {
      "entropy": 5.741157579421997,
      "epoch": 5.887853091480879,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021619842333642252,
      "loss": 4.9005,
      "mean_token_accuracy": 0.23993195295333863,
      "num_tokens": 117031611.0,
      "step": 52425
    },
    {
      "entropy": 5.839941453933716,
      "epoch": 5.888414668388836,
      "grad_norm": 1.1875,
      "learning_rate": 0.00021615967085350147,
      "loss": 5.0472,
      "mean_token_accuracy": 0.2358932226896286,
      "num_tokens": 117044021.0,
      "step": 52430
    },
    {
      "entropy": 5.659596490859985,
      "epoch": 5.888976245296793,
      "grad_norm": 1.328125,
      "learning_rate": 0.00021612092024407925,
      "loss": 4.7726,
      "mean_token_accuracy": 0.23950278013944626,
      "num_tokens": 117054431.0,
      "step": 52435
    },
    {
      "entropy": 5.753574562072754,
      "epoch": 5.889537822204751,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00021608217150938987,
      "loss": 4.942,
      "mean_token_accuracy": 0.23660970479249954,
      "num_tokens": 117066162.0,
      "step": 52440
    },
    {
      "entropy": 5.727737379074097,
      "epoch": 5.890099399112708,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002160434246506669,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23604738861322402,
      "num_tokens": 117077510.0,
      "step": 52445
    },
    {
      "entropy": 5.713465309143066,
      "epoch": 5.890660976020666,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002160046796691442,
      "loss": 4.8628,
      "mean_token_accuracy": 0.2342410609126091,
      "num_tokens": 117088564.0,
      "step": 52450
    },
    {
      "entropy": 5.708724069595337,
      "epoch": 5.891222552928624,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002159659365660554,
      "loss": 4.8711,
      "mean_token_accuracy": 0.23686991333961488,
      "num_tokens": 117099606.0,
      "step": 52455
    },
    {
      "entropy": 5.722377824783325,
      "epoch": 5.891784129836581,
      "grad_norm": 1.453125,
      "learning_rate": 0.000215927195342634,
      "loss": 4.9291,
      "mean_token_accuracy": 0.23485891073942183,
      "num_tokens": 117109240.0,
      "step": 52460
    },
    {
      "entropy": 5.67761459350586,
      "epoch": 5.892345706744539,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021588845600011364,
      "loss": 4.8481,
      "mean_token_accuracy": 0.24339757710695267,
      "num_tokens": 117120192.0,
      "step": 52465
    },
    {
      "entropy": 5.672073316574097,
      "epoch": 5.892907283652496,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002158497185397278,
      "loss": 4.8008,
      "mean_token_accuracy": 0.2398560091853142,
      "num_tokens": 117130959.0,
      "step": 52470
    },
    {
      "entropy": 5.769021463394165,
      "epoch": 5.8934688605604535,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021581098296270985,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23745909631252288,
      "num_tokens": 117141745.0,
      "step": 52475
    },
    {
      "entropy": 5.720073986053467,
      "epoch": 5.894030437468412,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021577224927029322,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23118699491024017,
      "num_tokens": 117153715.0,
      "step": 52480
    },
    {
      "entropy": 5.643171072006226,
      "epoch": 5.894592014376369,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021573351746371106,
      "loss": 4.7837,
      "mean_token_accuracy": 0.25209723263978956,
      "num_tokens": 117163934.0,
      "step": 52485
    },
    {
      "entropy": 5.749965810775757,
      "epoch": 5.895153591284326,
      "grad_norm": 1.203125,
      "learning_rate": 0.00021569478754419674,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22740051448345183,
      "num_tokens": 117174515.0,
      "step": 52490
    },
    {
      "entropy": 5.680897045135498,
      "epoch": 5.895715168192284,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00021565605951298335,
      "loss": 4.8425,
      "mean_token_accuracy": 0.23441292494535446,
      "num_tokens": 117184942.0,
      "step": 52495
    },
    {
      "entropy": 5.670980167388916,
      "epoch": 5.896276745100241,
      "grad_norm": 1.34375,
      "learning_rate": 0.000215617333371304,
      "loss": 4.8724,
      "mean_token_accuracy": 0.2298915296792984,
      "num_tokens": 117195653.0,
      "step": 52500
    },
    {
      "entropy": 5.68382511138916,
      "epoch": 5.896838322008199,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021557860912039184,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22897339910268782,
      "num_tokens": 117206078.0,
      "step": 52505
    },
    {
      "entropy": 5.704847145080566,
      "epoch": 5.897399898916157,
      "grad_norm": 1.359375,
      "learning_rate": 0.00021553988676147984,
      "loss": 4.9494,
      "mean_token_accuracy": 0.23210353255271912,
      "num_tokens": 117218051.0,
      "step": 52510
    },
    {
      "entropy": 5.731188535690308,
      "epoch": 5.897961475824114,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002155011662958009,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23949020206928254,
      "num_tokens": 117229397.0,
      "step": 52515
    },
    {
      "entropy": 5.7462625980377195,
      "epoch": 5.898523052732072,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021546244772458785,
      "loss": 4.9514,
      "mean_token_accuracy": 0.23027456402778626,
      "num_tokens": 117240392.0,
      "step": 52520
    },
    {
      "entropy": 5.679622602462769,
      "epoch": 5.899084629640029,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021542373104907364,
      "loss": 4.8602,
      "mean_token_accuracy": 0.23966781049966812,
      "num_tokens": 117251726.0,
      "step": 52525
    },
    {
      "entropy": 5.726957845687866,
      "epoch": 5.899646206547986,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021538501627049085,
      "loss": 4.9009,
      "mean_token_accuracy": 0.23984215557575225,
      "num_tokens": 117264833.0,
      "step": 52530
    },
    {
      "entropy": 5.739059591293335,
      "epoch": 5.9002077834559445,
      "grad_norm": 1.34375,
      "learning_rate": 0.00021534630339007223,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22929173707962036,
      "num_tokens": 117275704.0,
      "step": 52535
    },
    {
      "entropy": 5.721734142303466,
      "epoch": 5.900769360363902,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002153075924090505,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22813452333211898,
      "num_tokens": 117286904.0,
      "step": 52540
    },
    {
      "entropy": 5.705776071548462,
      "epoch": 5.90133093727186,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00021526888332865817,
      "loss": 4.8837,
      "mean_token_accuracy": 0.23269150406122208,
      "num_tokens": 117297882.0,
      "step": 52545
    },
    {
      "entropy": 5.700602579116821,
      "epoch": 5.901892514179817,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021523017615012786,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23514512926340103,
      "num_tokens": 117310012.0,
      "step": 52550
    },
    {
      "entropy": 5.690243291854858,
      "epoch": 5.902454091087774,
      "grad_norm": 1.359375,
      "learning_rate": 0.00021519147087469182,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23792518377304078,
      "num_tokens": 117320755.0,
      "step": 52555
    },
    {
      "entropy": 5.707155656814575,
      "epoch": 5.9030156679957315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021515276750358258,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22799110859632493,
      "num_tokens": 117332429.0,
      "step": 52560
    },
    {
      "entropy": 5.7404297351837155,
      "epoch": 5.90357724490369,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021511406603803247,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23514537811279296,
      "num_tokens": 117343293.0,
      "step": 52565
    },
    {
      "entropy": 5.737614393234253,
      "epoch": 5.904138821811647,
      "grad_norm": 1.1875,
      "learning_rate": 0.00021507536647927361,
      "loss": 4.8639,
      "mean_token_accuracy": 0.24591924846172333,
      "num_tokens": 117353529.0,
      "step": 52570
    },
    {
      "entropy": 5.686261558532715,
      "epoch": 5.904700398719605,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002150366688285384,
      "loss": 4.8724,
      "mean_token_accuracy": 0.23522313088178634,
      "num_tokens": 117363961.0,
      "step": 52575
    },
    {
      "entropy": 5.727719974517822,
      "epoch": 5.905261975627562,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021499797308705893,
      "loss": 4.9598,
      "mean_token_accuracy": 0.23158173710107804,
      "num_tokens": 117375892.0,
      "step": 52580
    },
    {
      "entropy": 5.733008956909179,
      "epoch": 5.905823552535519,
      "grad_norm": 1.21875,
      "learning_rate": 0.00021495927925606722,
      "loss": 4.9471,
      "mean_token_accuracy": 0.23651719987392425,
      "num_tokens": 117387697.0,
      "step": 52585
    },
    {
      "entropy": 5.798114156723022,
      "epoch": 5.9063851294434775,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021492058733679542,
      "loss": 4.963,
      "mean_token_accuracy": 0.23371609449386596,
      "num_tokens": 117399273.0,
      "step": 52590
    },
    {
      "entropy": 5.665938997268677,
      "epoch": 5.906946706351435,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021488189733047535,
      "loss": 4.885,
      "mean_token_accuracy": 0.2309234172105789,
      "num_tokens": 117411293.0,
      "step": 52595
    },
    {
      "entropy": 5.729496574401855,
      "epoch": 5.907508283259393,
      "grad_norm": 1.34375,
      "learning_rate": 0.000214843209238339,
      "loss": 4.9623,
      "mean_token_accuracy": 0.23015014082193375,
      "num_tokens": 117421339.0,
      "step": 52600
    },
    {
      "entropy": 5.67974967956543,
      "epoch": 5.90806986016735,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00021480452306161813,
      "loss": 4.8466,
      "mean_token_accuracy": 0.24395378977060317,
      "num_tokens": 117431702.0,
      "step": 52605
    },
    {
      "entropy": 5.734515762329101,
      "epoch": 5.908631437075307,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021476583880154464,
      "loss": 4.916,
      "mean_token_accuracy": 0.23579398393630982,
      "num_tokens": 117442486.0,
      "step": 52610
    },
    {
      "entropy": 5.797694206237793,
      "epoch": 5.909193013983265,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021472715645935026,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22324419766664505,
      "num_tokens": 117454038.0,
      "step": 52615
    },
    {
      "entropy": 5.787585306167602,
      "epoch": 5.909754590891223,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00021468847603626657,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23299750685691833,
      "num_tokens": 117464401.0,
      "step": 52620
    },
    {
      "entropy": 5.646312808990478,
      "epoch": 5.91031616779918,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002146497975335252,
      "loss": 4.787,
      "mean_token_accuracy": 0.23845719993114473,
      "num_tokens": 117475472.0,
      "step": 52625
    },
    {
      "entropy": 5.719171619415283,
      "epoch": 5.910877744707138,
      "grad_norm": 1.390625,
      "learning_rate": 0.00021461112095235764,
      "loss": 4.8895,
      "mean_token_accuracy": 0.23554143756628038,
      "num_tokens": 117487298.0,
      "step": 52630
    },
    {
      "entropy": 5.674502801895142,
      "epoch": 5.911439321615095,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021457244629399547,
      "loss": 4.8897,
      "mean_token_accuracy": 0.23091792613267897,
      "num_tokens": 117497562.0,
      "step": 52635
    },
    {
      "entropy": 5.680526447296143,
      "epoch": 5.912000898523052,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021453377355967003,
      "loss": 4.7966,
      "mean_token_accuracy": 0.24778589606285095,
      "num_tokens": 117509352.0,
      "step": 52640
    },
    {
      "entropy": 5.822624206542969,
      "epoch": 5.91256247543101,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021449510275061264,
      "loss": 5.0606,
      "mean_token_accuracy": 0.23271507769823074,
      "num_tokens": 117521296.0,
      "step": 52645
    },
    {
      "entropy": 5.702304458618164,
      "epoch": 5.913124052338968,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021445643386805468,
      "loss": 4.8933,
      "mean_token_accuracy": 0.23327337801456452,
      "num_tokens": 117532006.0,
      "step": 52650
    },
    {
      "entropy": 5.709241962432861,
      "epoch": 5.913685629246926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002144177669132274,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23589629381895066,
      "num_tokens": 117543534.0,
      "step": 52655
    },
    {
      "entropy": 5.710425567626953,
      "epoch": 5.914247206154883,
      "grad_norm": 1.25,
      "learning_rate": 0.00021437910188736188,
      "loss": 4.8616,
      "mean_token_accuracy": 0.24137571454048157,
      "num_tokens": 117554943.0,
      "step": 52660
    },
    {
      "entropy": 5.644263458251953,
      "epoch": 5.91480878306284,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021434043879168925,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2406824931502342,
      "num_tokens": 117565326.0,
      "step": 52665
    },
    {
      "entropy": 5.712321090698242,
      "epoch": 5.915370359970798,
      "grad_norm": 1.328125,
      "learning_rate": 0.00021430177762744057,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22635235041379928,
      "num_tokens": 117576360.0,
      "step": 52670
    },
    {
      "entropy": 5.785477495193481,
      "epoch": 5.9159319368787555,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021426311839584688,
      "loss": 4.9806,
      "mean_token_accuracy": 0.23411643505096436,
      "num_tokens": 117587542.0,
      "step": 52675
    },
    {
      "entropy": 5.743315744400024,
      "epoch": 5.916493513786713,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00021422446109813892,
      "loss": 4.8836,
      "mean_token_accuracy": 0.23064443469047546,
      "num_tokens": 117597805.0,
      "step": 52680
    },
    {
      "entropy": 5.677922677993775,
      "epoch": 5.917055090694671,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002141858057355478,
      "loss": 4.8307,
      "mean_token_accuracy": 0.24102818369865417,
      "num_tokens": 117609670.0,
      "step": 52685
    },
    {
      "entropy": 5.655228328704834,
      "epoch": 5.917616667602628,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002141471523093042,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22741994112730027,
      "num_tokens": 117620765.0,
      "step": 52690
    },
    {
      "entropy": 5.694312572479248,
      "epoch": 5.918178244510585,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021410850082063888,
      "loss": 4.847,
      "mean_token_accuracy": 0.24110493212938308,
      "num_tokens": 117631104.0,
      "step": 52695
    },
    {
      "entropy": 5.839631795883179,
      "epoch": 5.918739821418543,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002140698512707825,
      "loss": 5.0001,
      "mean_token_accuracy": 0.23599439710378647,
      "num_tokens": 117642547.0,
      "step": 52700
    },
    {
      "entropy": 5.770058584213257,
      "epoch": 5.919301398326501,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00021403120366096572,
      "loss": 4.9707,
      "mean_token_accuracy": 0.23948506116867066,
      "num_tokens": 117654432.0,
      "step": 52705
    },
    {
      "entropy": 5.648664093017578,
      "epoch": 5.919862975234459,
      "grad_norm": 1.34375,
      "learning_rate": 0.00021399255799241901,
      "loss": 4.8602,
      "mean_token_accuracy": 0.24215614795684814,
      "num_tokens": 117664493.0,
      "step": 52710
    },
    {
      "entropy": 5.7201556205749515,
      "epoch": 5.920424552142416,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00021395391426637288,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22766531109809876,
      "num_tokens": 117676182.0,
      "step": 52715
    },
    {
      "entropy": 5.7415323734283445,
      "epoch": 5.920986129050373,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002139152724840579,
      "loss": 4.9964,
      "mean_token_accuracy": 0.23913909643888473,
      "num_tokens": 117687107.0,
      "step": 52720
    },
    {
      "entropy": 5.770110845565796,
      "epoch": 5.921547705958331,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021387663264670425,
      "loss": 4.8689,
      "mean_token_accuracy": 0.24090705513954164,
      "num_tokens": 117698789.0,
      "step": 52725
    },
    {
      "entropy": 5.740184164047241,
      "epoch": 5.9221092828662885,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021383799475554238,
      "loss": 4.9251,
      "mean_token_accuracy": 0.23293698877096175,
      "num_tokens": 117709883.0,
      "step": 52730
    },
    {
      "entropy": 5.689295053482056,
      "epoch": 5.922670859774247,
      "grad_norm": 1.25,
      "learning_rate": 0.0002137993588118025,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2344183787703514,
      "num_tokens": 117720708.0,
      "step": 52735
    },
    {
      "entropy": 5.7837474822998045,
      "epoch": 5.923232436682204,
      "grad_norm": 1.203125,
      "learning_rate": 0.00021376072481671477,
      "loss": 4.923,
      "mean_token_accuracy": 0.23177866339683534,
      "num_tokens": 117731437.0,
      "step": 52740
    },
    {
      "entropy": 5.722924470901489,
      "epoch": 5.923794013590161,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002137220927715093,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2394036367535591,
      "num_tokens": 117741520.0,
      "step": 52745
    },
    {
      "entropy": 5.704571008682251,
      "epoch": 5.924355590498118,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000213683462677416,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23320562839508058,
      "num_tokens": 117753264.0,
      "step": 52750
    },
    {
      "entropy": 5.781167221069336,
      "epoch": 5.924917167406076,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021364483453566519,
      "loss": 4.989,
      "mean_token_accuracy": 0.2253457397222519,
      "num_tokens": 117763841.0,
      "step": 52755
    },
    {
      "entropy": 5.6851959228515625,
      "epoch": 5.9254787443140335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00021360620834748662,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22773790061473848,
      "num_tokens": 117774868.0,
      "step": 52760
    },
    {
      "entropy": 5.715021705627441,
      "epoch": 5.926040321221992,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002135675841141102,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2250316932797432,
      "num_tokens": 117784464.0,
      "step": 52765
    },
    {
      "entropy": 5.738483095169068,
      "epoch": 5.926601898129949,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00021352896183676574,
      "loss": 4.8977,
      "mean_token_accuracy": 0.23746179789304733,
      "num_tokens": 117796012.0,
      "step": 52770
    },
    {
      "entropy": 5.679723691940308,
      "epoch": 5.927163475037906,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021349034151668295,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22948145270347595,
      "num_tokens": 117807421.0,
      "step": 52775
    },
    {
      "entropy": 5.703344488143921,
      "epoch": 5.927725051945864,
      "grad_norm": 1.390625,
      "learning_rate": 0.00021345172315509153,
      "loss": 4.8764,
      "mean_token_accuracy": 0.24708468466997147,
      "num_tokens": 117818077.0,
      "step": 52780
    },
    {
      "entropy": 5.681015205383301,
      "epoch": 5.928286628853821,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021341310675322112,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2326932817697525,
      "num_tokens": 117829037.0,
      "step": 52785
    },
    {
      "entropy": 5.768882083892822,
      "epoch": 5.9288482057617795,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021337449231230128,
      "loss": 4.8611,
      "mean_token_accuracy": 0.24102332293987275,
      "num_tokens": 117840096.0,
      "step": 52790
    },
    {
      "entropy": 5.734170579910279,
      "epoch": 5.929409782669737,
      "grad_norm": 1.21875,
      "learning_rate": 0.00021333587983356157,
      "loss": 4.9307,
      "mean_token_accuracy": 0.23139936178922654,
      "num_tokens": 117852233.0,
      "step": 52795
    },
    {
      "entropy": 5.734435796737671,
      "epoch": 5.929971359577694,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021329726931823128,
      "loss": 4.8538,
      "mean_token_accuracy": 0.23524200767278672,
      "num_tokens": 117862496.0,
      "step": 52800
    },
    {
      "entropy": 5.703002119064331,
      "epoch": 5.930532936485652,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002132586607675399,
      "loss": 4.8357,
      "mean_token_accuracy": 0.24484531730413436,
      "num_tokens": 117873347.0,
      "step": 52805
    },
    {
      "entropy": 5.708284378051758,
      "epoch": 5.931094513393609,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021322005418271674,
      "loss": 4.925,
      "mean_token_accuracy": 0.2347681552171707,
      "num_tokens": 117885662.0,
      "step": 52810
    },
    {
      "entropy": 5.723811531066895,
      "epoch": 5.9316560903015665,
      "grad_norm": 1.25,
      "learning_rate": 0.00021318144956499098,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23793541938066481,
      "num_tokens": 117896047.0,
      "step": 52815
    },
    {
      "entropy": 5.718330383300781,
      "epoch": 5.932217667209525,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021314284691559182,
      "loss": 4.8432,
      "mean_token_accuracy": 0.2349206790328026,
      "num_tokens": 117908018.0,
      "step": 52820
    },
    {
      "entropy": 5.686875867843628,
      "epoch": 5.932779244117482,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002131042462357483,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2357866495847702,
      "num_tokens": 117919628.0,
      "step": 52825
    },
    {
      "entropy": 5.65454888343811,
      "epoch": 5.933340821025439,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00021306564752668979,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2414078637957573,
      "num_tokens": 117930297.0,
      "step": 52830
    },
    {
      "entropy": 5.705776309967041,
      "epoch": 5.933902397933397,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021302705078964497,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23025428503751755,
      "num_tokens": 117941705.0,
      "step": 52835
    },
    {
      "entropy": 5.702276420593262,
      "epoch": 5.934463974841354,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021298845602584293,
      "loss": 4.9144,
      "mean_token_accuracy": 0.23813132494688033,
      "num_tokens": 117953892.0,
      "step": 52840
    },
    {
      "entropy": 5.850148773193359,
      "epoch": 5.9350255517493125,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002129498632365125,
      "loss": 5.0524,
      "mean_token_accuracy": 0.22422221004962922,
      "num_tokens": 117964534.0,
      "step": 52845
    },
    {
      "entropy": 5.7396543502807615,
      "epoch": 5.93558712865727,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002129112724228825,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23529710173606871,
      "num_tokens": 117976124.0,
      "step": 52850
    },
    {
      "entropy": 5.719940948486328,
      "epoch": 5.936148705565227,
      "grad_norm": 1.40625,
      "learning_rate": 0.00021287268358618168,
      "loss": 4.8794,
      "mean_token_accuracy": 0.2367027148604393,
      "num_tokens": 117987920.0,
      "step": 52855
    },
    {
      "entropy": 5.834096527099609,
      "epoch": 5.936710282473185,
      "grad_norm": 1.25,
      "learning_rate": 0.00021283409672763864,
      "loss": 5.0088,
      "mean_token_accuracy": 0.23125898689031602,
      "num_tokens": 117999144.0,
      "step": 52860
    },
    {
      "entropy": 5.7523205280303955,
      "epoch": 5.937271859381142,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002127955118484821,
      "loss": 4.9263,
      "mean_token_accuracy": 0.23024341762065886,
      "num_tokens": 118011008.0,
      "step": 52865
    },
    {
      "entropy": 5.707960605621338,
      "epoch": 5.937833436289099,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00021275692894994058,
      "loss": 4.8764,
      "mean_token_accuracy": 0.24139794409275056,
      "num_tokens": 118021717.0,
      "step": 52870
    },
    {
      "entropy": 5.732465839385986,
      "epoch": 5.9383950131970575,
      "grad_norm": 1.46875,
      "learning_rate": 0.00021271834803324264,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23069979399442672,
      "num_tokens": 118031804.0,
      "step": 52875
    },
    {
      "entropy": 5.660793018341065,
      "epoch": 5.938956590105015,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002126797690996166,
      "loss": 4.8241,
      "mean_token_accuracy": 0.24743328243494034,
      "num_tokens": 118043540.0,
      "step": 52880
    },
    {
      "entropy": 5.70119423866272,
      "epoch": 5.939518167012972,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00021264119215029092,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23211296647787094,
      "num_tokens": 118054207.0,
      "step": 52885
    },
    {
      "entropy": 5.70995979309082,
      "epoch": 5.94007974392093,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021260261718649388,
      "loss": 4.935,
      "mean_token_accuracy": 0.23148657530546188,
      "num_tokens": 118065236.0,
      "step": 52890
    },
    {
      "entropy": 5.734076499938965,
      "epoch": 5.940641320828887,
      "grad_norm": 1.328125,
      "learning_rate": 0.00021256404420945358,
      "loss": 4.9684,
      "mean_token_accuracy": 0.23221544921398163,
      "num_tokens": 118075609.0,
      "step": 52895
    },
    {
      "entropy": 5.719460535049438,
      "epoch": 5.941202897736845,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002125254732203984,
      "loss": 4.862,
      "mean_token_accuracy": 0.23419974595308304,
      "num_tokens": 118086217.0,
      "step": 52900
    },
    {
      "entropy": 5.666848993301391,
      "epoch": 5.941764474644803,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021248690422055643,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2347588375210762,
      "num_tokens": 118097216.0,
      "step": 52905
    },
    {
      "entropy": 5.782560586929321,
      "epoch": 5.94232605155276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00021244833721115563,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2340303972363472,
      "num_tokens": 118108712.0,
      "step": 52910
    },
    {
      "entropy": 5.765071964263916,
      "epoch": 5.942887628460718,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002124097721934241,
      "loss": 4.9959,
      "mean_token_accuracy": 0.22728253304958343,
      "num_tokens": 118119510.0,
      "step": 52915
    },
    {
      "entropy": 5.6835509777069095,
      "epoch": 5.943449205368675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021237120916858967,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2255554676055908,
      "num_tokens": 118130991.0,
      "step": 52920
    },
    {
      "entropy": 5.787114953994751,
      "epoch": 5.944010782276633,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021233264813788022,
      "loss": 4.927,
      "mean_token_accuracy": 0.2461271271109581,
      "num_tokens": 118141939.0,
      "step": 52925
    },
    {
      "entropy": 5.669981622695923,
      "epoch": 5.9445723591845905,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002122940891025235,
      "loss": 4.833,
      "mean_token_accuracy": 0.23776401281356813,
      "num_tokens": 118153648.0,
      "step": 52930
    },
    {
      "entropy": 5.71578950881958,
      "epoch": 5.945133936092548,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021225553206374737,
      "loss": 4.9116,
      "mean_token_accuracy": 0.23264535069465636,
      "num_tokens": 118163461.0,
      "step": 52935
    },
    {
      "entropy": 5.655777359008789,
      "epoch": 5.945695513000506,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021221697702277943,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23980215638875962,
      "num_tokens": 118175614.0,
      "step": 52940
    },
    {
      "entropy": 5.7059485912323,
      "epoch": 5.946257089908463,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002121784239808473,
      "loss": 4.9063,
      "mean_token_accuracy": 0.24439604729413986,
      "num_tokens": 118186620.0,
      "step": 52945
    },
    {
      "entropy": 5.698924493789673,
      "epoch": 5.94681866681642,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021213987293917846,
      "loss": 4.8675,
      "mean_token_accuracy": 0.23746623545885087,
      "num_tokens": 118197581.0,
      "step": 52950
    },
    {
      "entropy": 5.695282793045044,
      "epoch": 5.947380243724378,
      "grad_norm": 1.421875,
      "learning_rate": 0.00021210132389900055,
      "loss": 4.8112,
      "mean_token_accuracy": 0.24053244441747665,
      "num_tokens": 118208227.0,
      "step": 52955
    },
    {
      "entropy": 5.760551643371582,
      "epoch": 5.947941820632336,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021206277686154078,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22909981906414031,
      "num_tokens": 118219182.0,
      "step": 52960
    },
    {
      "entropy": 5.698824024200439,
      "epoch": 5.948503397540293,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021202423182802656,
      "loss": 4.854,
      "mean_token_accuracy": 0.24101511985063553,
      "num_tokens": 118230073.0,
      "step": 52965
    },
    {
      "entropy": 5.7318689823150635,
      "epoch": 5.949064974448251,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021198568879968527,
      "loss": 4.867,
      "mean_token_accuracy": 0.23552431613206865,
      "num_tokens": 118239702.0,
      "step": 52970
    },
    {
      "entropy": 5.683316612243653,
      "epoch": 5.949626551356208,
      "grad_norm": 1.375,
      "learning_rate": 0.00021194714777774415,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23609186261892318,
      "num_tokens": 118250725.0,
      "step": 52975
    },
    {
      "entropy": 5.693480587005615,
      "epoch": 5.950188128264166,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002119086087634303,
      "loss": 4.9061,
      "mean_token_accuracy": 0.23836590200662613,
      "num_tokens": 118262035.0,
      "step": 52980
    },
    {
      "entropy": 5.726250457763672,
      "epoch": 5.950749705172123,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021187007175797078,
      "loss": 4.8898,
      "mean_token_accuracy": 0.24404179453849792,
      "num_tokens": 118271074.0,
      "step": 52985
    },
    {
      "entropy": 5.7349388122558596,
      "epoch": 5.951311282080081,
      "grad_norm": 1.375,
      "learning_rate": 0.0002118315367625927,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23276246190071107,
      "num_tokens": 118280557.0,
      "step": 52990
    },
    {
      "entropy": 5.717606258392334,
      "epoch": 5.951872858988039,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021179300377852296,
      "loss": 4.9368,
      "mean_token_accuracy": 0.23306964933872223,
      "num_tokens": 118291504.0,
      "step": 52995
    },
    {
      "entropy": 5.653397130966186,
      "epoch": 5.952434435895996,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002117544728069885,
      "loss": 4.8291,
      "mean_token_accuracy": 0.2396787628531456,
      "num_tokens": 118302744.0,
      "step": 53000
    },
    {
      "entropy": 5.658467054367065,
      "epoch": 5.952996012803953,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002117159438492161,
      "loss": 4.8765,
      "mean_token_accuracy": 0.23599570691585542,
      "num_tokens": 118313954.0,
      "step": 53005
    },
    {
      "entropy": 5.720455312728882,
      "epoch": 5.953557589711911,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00021167741690643268,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23911405503749847,
      "num_tokens": 118325487.0,
      "step": 53010
    },
    {
      "entropy": 5.728354358673096,
      "epoch": 5.9541191666198685,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002116388919798648,
      "loss": 4.9404,
      "mean_token_accuracy": 0.23288240730762483,
      "num_tokens": 118336214.0,
      "step": 53015
    },
    {
      "entropy": 5.755486583709716,
      "epoch": 5.954680743527826,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021160036907073921,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23116549849510193,
      "num_tokens": 118346867.0,
      "step": 53020
    },
    {
      "entropy": 5.740579080581665,
      "epoch": 5.955242320435784,
      "grad_norm": 1.390625,
      "learning_rate": 0.00021156184818028247,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23798996210098267,
      "num_tokens": 118357466.0,
      "step": 53025
    },
    {
      "entropy": 5.686578607559204,
      "epoch": 5.955803897343741,
      "grad_norm": 1.34375,
      "learning_rate": 0.00021152332930972107,
      "loss": 4.837,
      "mean_token_accuracy": 0.23407596796751023,
      "num_tokens": 118367579.0,
      "step": 53030
    },
    {
      "entropy": 5.783419370651245,
      "epoch": 5.956365474251699,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021148481246028146,
      "loss": 4.9917,
      "mean_token_accuracy": 0.22378055453300477,
      "num_tokens": 118379212.0,
      "step": 53035
    },
    {
      "entropy": 5.67373948097229,
      "epoch": 5.956927051159656,
      "grad_norm": 1.25,
      "learning_rate": 0.00021144629763318997,
      "loss": 4.868,
      "mean_token_accuracy": 0.23819223195314407,
      "num_tokens": 118390089.0,
      "step": 53040
    },
    {
      "entropy": 5.713241767883301,
      "epoch": 5.957488628067614,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002114077848296731,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2392009511590004,
      "num_tokens": 118400904.0,
      "step": 53045
    },
    {
      "entropy": 5.709656572341919,
      "epoch": 5.958050204975572,
      "grad_norm": 1.171875,
      "learning_rate": 0.00021136927405095707,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2412265956401825,
      "num_tokens": 118413359.0,
      "step": 53050
    },
    {
      "entropy": 5.698308706283569,
      "epoch": 5.958611781883529,
      "grad_norm": 1.359375,
      "learning_rate": 0.000211330765298268,
      "loss": 4.8671,
      "mean_token_accuracy": 0.2381548449397087,
      "num_tokens": 118423890.0,
      "step": 53055
    },
    {
      "entropy": 5.704334354400634,
      "epoch": 5.959173358791486,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002112922585728321,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2330963209271431,
      "num_tokens": 118434937.0,
      "step": 53060
    },
    {
      "entropy": 5.654147005081176,
      "epoch": 5.959734935699444,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002112537538758753,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23782020509243013,
      "num_tokens": 118446417.0,
      "step": 53065
    },
    {
      "entropy": 5.752660322189331,
      "epoch": 5.9602965126074015,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021121525120862384,
      "loss": 4.9038,
      "mean_token_accuracy": 0.240373332798481,
      "num_tokens": 118457963.0,
      "step": 53070
    },
    {
      "entropy": 5.784177637100219,
      "epoch": 5.960858089515359,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00021117675057230335,
      "loss": 4.9028,
      "mean_token_accuracy": 0.23713889718055725,
      "num_tokens": 118468202.0,
      "step": 53075
    },
    {
      "entropy": 5.7377269744873045,
      "epoch": 5.961419666423317,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021113825196813997,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23581595569849015,
      "num_tokens": 118479819.0,
      "step": 53080
    },
    {
      "entropy": 5.706850385665893,
      "epoch": 5.961981243331274,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021109975539735943,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2417992353439331,
      "num_tokens": 118490491.0,
      "step": 53085
    },
    {
      "entropy": 5.6656488418579105,
      "epoch": 5.962542820239232,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021106126086118743,
      "loss": 4.887,
      "mean_token_accuracy": 0.2386616736650467,
      "num_tokens": 118501341.0,
      "step": 53090
    },
    {
      "entropy": 5.7137779712677,
      "epoch": 5.963104397147189,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021102276836084972,
      "loss": 4.8909,
      "mean_token_accuracy": 0.23841462284326553,
      "num_tokens": 118512604.0,
      "step": 53095
    },
    {
      "entropy": 5.682020998001098,
      "epoch": 5.9636659740551465,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00021098427789757184,
      "loss": 4.8919,
      "mean_token_accuracy": 0.238909812271595,
      "num_tokens": 118523550.0,
      "step": 53100
    },
    {
      "entropy": 5.661528873443603,
      "epoch": 5.964227550963105,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002109457894725794,
      "loss": 4.8247,
      "mean_token_accuracy": 0.2401431009173393,
      "num_tokens": 118534009.0,
      "step": 53105
    },
    {
      "entropy": 5.6692461490631105,
      "epoch": 5.964789127871062,
      "grad_norm": 1.359375,
      "learning_rate": 0.00021090730308709777,
      "loss": 4.8556,
      "mean_token_accuracy": 0.2342862144112587,
      "num_tokens": 118545177.0,
      "step": 53110
    },
    {
      "entropy": 5.6875158786773685,
      "epoch": 5.96535070477902,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021086881874235252,
      "loss": 4.9255,
      "mean_token_accuracy": 0.22951716929674149,
      "num_tokens": 118555823.0,
      "step": 53115
    },
    {
      "entropy": 5.735443878173828,
      "epoch": 5.965912281686977,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021083033643956906,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22762267291545868,
      "num_tokens": 118567319.0,
      "step": 53120
    },
    {
      "entropy": 5.690855646133423,
      "epoch": 5.966473858594934,
      "grad_norm": 1.265625,
      "learning_rate": 0.00021079185617997248,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23630268424749373,
      "num_tokens": 118577456.0,
      "step": 53125
    },
    {
      "entropy": 5.674044418334961,
      "epoch": 5.9670354355028925,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00021075337796478815,
      "loss": 4.7844,
      "mean_token_accuracy": 0.25125612616539,
      "num_tokens": 118588203.0,
      "step": 53130
    },
    {
      "entropy": 5.833264398574829,
      "epoch": 5.96759701241085,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021071490179524116,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2363317146897316,
      "num_tokens": 118600913.0,
      "step": 53135
    },
    {
      "entropy": 5.678606271743774,
      "epoch": 5.968158589318807,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00021067642767255663,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2402414634823799,
      "num_tokens": 118612491.0,
      "step": 53140
    },
    {
      "entropy": 5.706014204025268,
      "epoch": 5.968720166226765,
      "grad_norm": 1.125,
      "learning_rate": 0.0002106379555979595,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2347464606165886,
      "num_tokens": 118624612.0,
      "step": 53145
    },
    {
      "entropy": 5.747680187225342,
      "epoch": 5.969281743134722,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00021059948557267493,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23535252064466478,
      "num_tokens": 118634418.0,
      "step": 53150
    },
    {
      "entropy": 5.61766676902771,
      "epoch": 5.9698433200426795,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021056101759792773,
      "loss": 4.7932,
      "mean_token_accuracy": 0.24152573198080063,
      "num_tokens": 118644760.0,
      "step": 53155
    },
    {
      "entropy": 5.78153715133667,
      "epoch": 5.970404896950638,
      "grad_norm": 1.28125,
      "learning_rate": 0.00021052255167494272,
      "loss": 5.0475,
      "mean_token_accuracy": 0.23031560629606246,
      "num_tokens": 118655949.0,
      "step": 53160
    },
    {
      "entropy": 5.845862674713135,
      "epoch": 5.970966473858595,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00021048408780494467,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2302673041820526,
      "num_tokens": 118666633.0,
      "step": 53165
    },
    {
      "entropy": 5.826344299316406,
      "epoch": 5.971528050766553,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00021044562598915824,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22779872566461562,
      "num_tokens": 118677268.0,
      "step": 53170
    },
    {
      "entropy": 5.74085841178894,
      "epoch": 5.97208962767451,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002104071662288081,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23506923615932465,
      "num_tokens": 118689446.0,
      "step": 53175
    },
    {
      "entropy": 5.720685958862305,
      "epoch": 5.972651204582467,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00021036870852511892,
      "loss": 4.8915,
      "mean_token_accuracy": 0.24192604422569275,
      "num_tokens": 118700766.0,
      "step": 53180
    },
    {
      "entropy": 5.628468227386475,
      "epoch": 5.9732127814904254,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000210330252879315,
      "loss": 4.7466,
      "mean_token_accuracy": 0.24120355397462845,
      "num_tokens": 118711176.0,
      "step": 53185
    },
    {
      "entropy": 5.747271633148193,
      "epoch": 5.973774358398383,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00021029179929262098,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22685966789722442,
      "num_tokens": 118721332.0,
      "step": 53190
    },
    {
      "entropy": 5.700373363494873,
      "epoch": 5.97433593530634,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021025334776626122,
      "loss": 4.8633,
      "mean_token_accuracy": 0.24191809743642806,
      "num_tokens": 118734459.0,
      "step": 53195
    },
    {
      "entropy": 5.781235122680664,
      "epoch": 5.974897512214298,
      "grad_norm": 1.25,
      "learning_rate": 0.00021021489830145995,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23212846219539643,
      "num_tokens": 118745387.0,
      "step": 53200
    },
    {
      "entropy": 5.79088282585144,
      "epoch": 5.975459089122255,
      "grad_norm": 1.25,
      "learning_rate": 0.00021017645089944142,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22441236525774003,
      "num_tokens": 118756685.0,
      "step": 53205
    },
    {
      "entropy": 5.794254350662231,
      "epoch": 5.976020666030212,
      "grad_norm": 1.296875,
      "learning_rate": 0.00021013800556142986,
      "loss": 5.0124,
      "mean_token_accuracy": 0.23776061236858367,
      "num_tokens": 118768892.0,
      "step": 53210
    },
    {
      "entropy": 5.75575819015503,
      "epoch": 5.9765822429381705,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021009956228864934,
      "loss": 5.0361,
      "mean_token_accuracy": 0.22825842052698136,
      "num_tokens": 118780165.0,
      "step": 53215
    },
    {
      "entropy": 5.78332257270813,
      "epoch": 5.977143819846128,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002100611210823239,
      "loss": 4.9629,
      "mean_token_accuracy": 0.22908055186271667,
      "num_tokens": 118790067.0,
      "step": 53220
    },
    {
      "entropy": 5.766598510742187,
      "epoch": 5.977705396754086,
      "grad_norm": 1.3125,
      "learning_rate": 0.00021002268194367764,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22853018045425416,
      "num_tokens": 118800887.0,
      "step": 53225
    },
    {
      "entropy": 5.637189769744873,
      "epoch": 5.978266973662043,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002099842448739343,
      "loss": 4.7705,
      "mean_token_accuracy": 0.2518832042813301,
      "num_tokens": 118812068.0,
      "step": 53230
    },
    {
      "entropy": 5.751052761077881,
      "epoch": 5.97882855057,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00020994580987431787,
      "loss": 4.9602,
      "mean_token_accuracy": 0.23705634027719497,
      "num_tokens": 118823556.0,
      "step": 53235
    },
    {
      "entropy": 5.794046640396118,
      "epoch": 5.979390127477958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00020990737694605204,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23095279783010483,
      "num_tokens": 118833981.0,
      "step": 53240
    },
    {
      "entropy": 5.866893577575683,
      "epoch": 5.979951704385916,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002098689460903606,
      "loss": 5.0236,
      "mean_token_accuracy": 0.23509865254163742,
      "num_tokens": 118846039.0,
      "step": 53245
    },
    {
      "entropy": 5.803691434860229,
      "epoch": 5.980513281293873,
      "grad_norm": 1.34375,
      "learning_rate": 0.00020983051730846724,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2322816401720047,
      "num_tokens": 118856653.0,
      "step": 53250
    },
    {
      "entropy": 5.825254774093628,
      "epoch": 5.981074858201831,
      "grad_norm": 1.4375,
      "learning_rate": 0.00020979209060159533,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2262425899505615,
      "num_tokens": 118868198.0,
      "step": 53255
    },
    {
      "entropy": 5.72312364578247,
      "epoch": 5.981636435109788,
      "grad_norm": 1.390625,
      "learning_rate": 0.00020975366597096863,
      "loss": 4.8357,
      "mean_token_accuracy": 0.23126110285520554,
      "num_tokens": 118877184.0,
      "step": 53260
    },
    {
      "entropy": 5.713934755325317,
      "epoch": 5.982198012017745,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002097152434178105,
      "loss": 4.9148,
      "mean_token_accuracy": 0.24240826964378356,
      "num_tokens": 118888823.0,
      "step": 53265
    },
    {
      "entropy": 5.735449123382568,
      "epoch": 5.9827595889257035,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00020967682294334439,
      "loss": 4.9392,
      "mean_token_accuracy": 0.23522940427064895,
      "num_tokens": 118899811.0,
      "step": 53270
    },
    {
      "entropy": 5.746751403808593,
      "epoch": 5.983321165833661,
      "grad_norm": 1.234375,
      "learning_rate": 0.00020963840454879358,
      "loss": 4.9186,
      "mean_token_accuracy": 0.23562588542699814,
      "num_tokens": 118912010.0,
      "step": 53275
    },
    {
      "entropy": 5.674895000457764,
      "epoch": 5.983882742741619,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002095999882353813,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2302503541111946,
      "num_tokens": 118924267.0,
      "step": 53280
    },
    {
      "entropy": 5.690632152557373,
      "epoch": 5.984444319649576,
      "grad_norm": 1.28125,
      "learning_rate": 0.00020956157400433073,
      "loss": 4.8651,
      "mean_token_accuracy": 0.2441067099571228,
      "num_tokens": 118935210.0,
      "step": 53285
    },
    {
      "entropy": 5.744733238220215,
      "epoch": 5.985005896557533,
      "grad_norm": 1.265625,
      "learning_rate": 0.000209523161856865,
      "loss": 4.9008,
      "mean_token_accuracy": 0.23939574062824248,
      "num_tokens": 118947491.0,
      "step": 53290
    },
    {
      "entropy": 5.718619441986084,
      "epoch": 5.985567473465491,
      "grad_norm": 1.28125,
      "learning_rate": 0.00020948475179420723,
      "loss": 4.9532,
      "mean_token_accuracy": 0.23631006181240083,
      "num_tokens": 118958268.0,
      "step": 53295
    },
    {
      "entropy": 5.742892408370972,
      "epoch": 5.986129050373449,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00020944634381758038,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22967873215675355,
      "num_tokens": 118972281.0,
      "step": 53300
    },
    {
      "entropy": 5.7663819789886475,
      "epoch": 5.986690627281407,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020940793792820735,
      "loss": 4.9698,
      "mean_token_accuracy": 0.22922347486019135,
      "num_tokens": 118983396.0,
      "step": 53305
    },
    {
      "entropy": 5.745295524597168,
      "epoch": 5.987252204189364,
      "grad_norm": 1.21875,
      "learning_rate": 0.00020936953412731103,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22336857616901398,
      "num_tokens": 118995965.0,
      "step": 53310
    },
    {
      "entropy": 5.719042730331421,
      "epoch": 5.987813781097321,
      "grad_norm": 1.28125,
      "learning_rate": 0.00020933113241611423,
      "loss": 4.8325,
      "mean_token_accuracy": 0.24641311913728714,
      "num_tokens": 119007074.0,
      "step": 53315
    },
    {
      "entropy": 5.699840927124024,
      "epoch": 5.988375358005279,
      "grad_norm": 1.25,
      "learning_rate": 0.0002092927327958396,
      "loss": 4.8889,
      "mean_token_accuracy": 0.23527536392211915,
      "num_tokens": 119017603.0,
      "step": 53320
    },
    {
      "entropy": 5.707300138473511,
      "epoch": 5.988936934913236,
      "grad_norm": 1.28125,
      "learning_rate": 0.00020925433526770986,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23844069689512254,
      "num_tokens": 119029340.0,
      "step": 53325
    },
    {
      "entropy": 5.7784055233001705,
      "epoch": 5.989498511821194,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020921593983294745,
      "loss": 4.9383,
      "mean_token_accuracy": 0.23545684069395065,
      "num_tokens": 119039369.0,
      "step": 53330
    },
    {
      "entropy": 5.799510812759399,
      "epoch": 5.990060088729152,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002091775464927752,
      "loss": 4.9775,
      "mean_token_accuracy": 0.23019755631685257,
      "num_tokens": 119051355.0,
      "step": 53335
    },
    {
      "entropy": 5.720141124725342,
      "epoch": 5.990621665637109,
      "grad_norm": 1.234375,
      "learning_rate": 0.00020913915524841536,
      "loss": 4.9183,
      "mean_token_accuracy": 0.2334730863571167,
      "num_tokens": 119062079.0,
      "step": 53340
    },
    {
      "entropy": 5.690565299987793,
      "epoch": 5.991183242545066,
      "grad_norm": 1.234375,
      "learning_rate": 0.00020910076610109035,
      "loss": 4.8577,
      "mean_token_accuracy": 0.23730973154306412,
      "num_tokens": 119073876.0,
      "step": 53345
    },
    {
      "entropy": 5.7689807415008545,
      "epoch": 5.991744819453024,
      "grad_norm": 1.25,
      "learning_rate": 0.00020906237905202253,
      "loss": 4.8763,
      "mean_token_accuracy": 0.24052936732769012,
      "num_tokens": 119085989.0,
      "step": 53350
    },
    {
      "entropy": 5.720239114761353,
      "epoch": 5.9923063963609815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00020902399410243405,
      "loss": 4.9465,
      "mean_token_accuracy": 0.23320111632347107,
      "num_tokens": 119099119.0,
      "step": 53355
    },
    {
      "entropy": 5.777137231826782,
      "epoch": 5.99286797326894,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002089856112535473,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22844015955924987,
      "num_tokens": 119109557.0,
      "step": 53360
    },
    {
      "entropy": 5.786613321304321,
      "epoch": 5.993429550176897,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00020894723050658421,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2272211268544197,
      "num_tokens": 119121082.0,
      "step": 53365
    },
    {
      "entropy": 5.766149473190308,
      "epoch": 5.993991127084854,
      "grad_norm": 1.15625,
      "learning_rate": 0.00020890885186276693,
      "loss": 4.9072,
      "mean_token_accuracy": 0.230231311917305,
      "num_tokens": 119133973.0,
      "step": 53370
    },
    {
      "entropy": 5.716275691986084,
      "epoch": 5.994552703992812,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002088704753233175,
      "loss": 4.9321,
      "mean_token_accuracy": 0.23075779229402543,
      "num_tokens": 119146165.0,
      "step": 53375
    },
    {
      "entropy": 5.692575836181641,
      "epoch": 5.995114280900769,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002088321008894577,
      "loss": 4.8294,
      "mean_token_accuracy": 0.24367811381816865,
      "num_tokens": 119155898.0,
      "step": 53380
    },
    {
      "entropy": 5.73818998336792,
      "epoch": 5.995675857808727,
      "grad_norm": 1.421875,
      "learning_rate": 0.00020879372856240952,
      "loss": 4.957,
      "mean_token_accuracy": 0.23513842672109603,
      "num_tokens": 119165787.0,
      "step": 53385
    },
    {
      "entropy": 5.766420078277588,
      "epoch": 5.996237434716685,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002087553583433947,
      "loss": 5.011,
      "mean_token_accuracy": 0.2261477917432785,
      "num_tokens": 119176315.0,
      "step": 53390
    },
    {
      "entropy": 5.767163038253784,
      "epoch": 5.996799011624642,
      "grad_norm": 1.109375,
      "learning_rate": 0.00020871699023363493,
      "loss": 4.8994,
      "mean_token_accuracy": 0.23742036521434784,
      "num_tokens": 119188544.0,
      "step": 53395
    },
    {
      "entropy": 5.816067838668824,
      "epoch": 5.997360588532599,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002086786242343519,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2259610339999199,
      "num_tokens": 119200656.0,
      "step": 53400
    },
    {
      "entropy": 5.751685905456543,
      "epoch": 5.997922165440557,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020864026034676718,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2395101383328438,
      "num_tokens": 119211778.0,
      "step": 53405
    },
    {
      "entropy": 5.746191883087159,
      "epoch": 5.9984837423485144,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020860189857210238,
      "loss": 4.9554,
      "mean_token_accuracy": 0.24038513004779816,
      "num_tokens": 119222932.0,
      "step": 53410
    },
    {
      "entropy": 5.728448057174683,
      "epoch": 5.9990453192564726,
      "grad_norm": 1.359375,
      "learning_rate": 0.00020856353891157882,
      "loss": 4.8558,
      "mean_token_accuracy": 0.2374526798725128,
      "num_tokens": 119233529.0,
      "step": 53415
    },
    {
      "entropy": 5.642176723480224,
      "epoch": 5.99960689616443,
      "grad_norm": 1.15625,
      "learning_rate": 0.00020852518136641802,
      "loss": 4.8667,
      "mean_token_accuracy": 0.23863068968057632,
      "num_tokens": 119246014.0,
      "step": 53420
    },
    {
      "entropy": 5.688486999935574,
      "epoch": 6.000112315381592,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020848682593784117,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2393997824854321,
      "num_tokens": 119255274.0,
      "step": 53425
    },
    {
      "entropy": 5.684983921051026,
      "epoch": 6.000673892289549,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002084484726270696,
      "loss": 4.8815,
      "mean_token_accuracy": 0.2370675325393677,
      "num_tokens": 119267475.0,
      "step": 53430
    },
    {
      "entropy": 5.6871874809265135,
      "epoch": 6.001235469197507,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00020841012143532446,
      "loss": 4.8408,
      "mean_token_accuracy": 0.24056372791528702,
      "num_tokens": 119278555.0,
      "step": 53435
    },
    {
      "entropy": 5.6776376247406,
      "epoch": 6.001797046105464,
      "grad_norm": 1.375,
      "learning_rate": 0.00020837177236382687,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23725394308567047,
      "num_tokens": 119289578.0,
      "step": 53440
    },
    {
      "entropy": 5.641351318359375,
      "epoch": 6.002358623013421,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002083334254137979,
      "loss": 4.8655,
      "mean_token_accuracy": 0.24138548076152802,
      "num_tokens": 119301459.0,
      "step": 53445
    },
    {
      "entropy": 5.720589065551758,
      "epoch": 6.0029201999213795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002082950805864585,
      "loss": 4.9603,
      "mean_token_accuracy": 0.23140944838523864,
      "num_tokens": 119311891.0,
      "step": 53450
    },
    {
      "entropy": 5.728235101699829,
      "epoch": 6.003481776829337,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00020825673788302967,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2357611283659935,
      "num_tokens": 119322534.0,
      "step": 53455
    },
    {
      "entropy": 5.648205327987671,
      "epoch": 6.004043353737294,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002082183973047322,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23687904179096222,
      "num_tokens": 119332979.0,
      "step": 53460
    },
    {
      "entropy": 5.723893308639527,
      "epoch": 6.004604930645252,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002081800588527868,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2406383588910103,
      "num_tokens": 119344382.0,
      "step": 53465
    },
    {
      "entropy": 5.690012836456299,
      "epoch": 6.005166507553209,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020814172252841417,
      "loss": 4.7736,
      "mean_token_accuracy": 0.25169801712036133,
      "num_tokens": 119355032.0,
      "step": 53470
    },
    {
      "entropy": 5.734738397598266,
      "epoch": 6.005728084461167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002081033883328351,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2393154114484787,
      "num_tokens": 119367483.0,
      "step": 53475
    },
    {
      "entropy": 5.788532447814942,
      "epoch": 6.006289661369125,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00020806505626727012,
      "loss": 4.9653,
      "mean_token_accuracy": 0.23523579090833663,
      "num_tokens": 119377923.0,
      "step": 53480
    },
    {
      "entropy": 5.693386077880859,
      "epoch": 6.006851238277082,
      "grad_norm": 1.40625,
      "learning_rate": 0.00020802672633293967,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23875776827335357,
      "num_tokens": 119387829.0,
      "step": 53485
    },
    {
      "entropy": 5.725048828125,
      "epoch": 6.00741281518504,
      "grad_norm": 1.234375,
      "learning_rate": 0.00020798839853106434,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22994726300239562,
      "num_tokens": 119399778.0,
      "step": 53490
    },
    {
      "entropy": 5.670852613449097,
      "epoch": 6.007974392092997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00020795007286286427,
      "loss": 4.7679,
      "mean_token_accuracy": 0.24489304125308992,
      "num_tokens": 119411594.0,
      "step": 53495
    },
    {
      "entropy": 5.751654481887817,
      "epoch": 6.008535969000954,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00020791174932955993,
      "loss": 4.8576,
      "mean_token_accuracy": 0.23979846388101578,
      "num_tokens": 119422094.0,
      "step": 53500
    },
    {
      "entropy": 5.759393930435181,
      "epoch": 6.009097545908912,
      "grad_norm": 1.25,
      "learning_rate": 0.00020787342793237147,
      "loss": 4.9055,
      "mean_token_accuracy": 0.23871396332979203,
      "num_tokens": 119432744.0,
      "step": 53505
    },
    {
      "entropy": 5.7846497058868405,
      "epoch": 6.00965912281687,
      "grad_norm": 1.375,
      "learning_rate": 0.00020783510867251908,
      "loss": 4.929,
      "mean_token_accuracy": 0.24219223856925964,
      "num_tokens": 119443581.0,
      "step": 53510
    },
    {
      "entropy": 5.72072205543518,
      "epoch": 6.010220699724828,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00020779679155122295,
      "loss": 4.8683,
      "mean_token_accuracy": 0.24020744860172272,
      "num_tokens": 119454776.0,
      "step": 53515
    },
    {
      "entropy": 5.664573001861572,
      "epoch": 6.010782276632785,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00020775847656970302,
      "loss": 4.838,
      "mean_token_accuracy": 0.2362663835287094,
      "num_tokens": 119464972.0,
      "step": 53520
    },
    {
      "entropy": 5.664440774917603,
      "epoch": 6.011343853540742,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002077201637291793,
      "loss": 4.8048,
      "mean_token_accuracy": 0.23767081648111343,
      "num_tokens": 119475989.0,
      "step": 53525
    },
    {
      "entropy": 5.742244482040405,
      "epoch": 6.0119054304487,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002076818530308716,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2353823661804199,
      "num_tokens": 119487421.0,
      "step": 53530
    },
    {
      "entropy": 5.783859825134277,
      "epoch": 6.0124670073566575,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00020764354447599988,
      "loss": 4.8966,
      "mean_token_accuracy": 0.24662993401288985,
      "num_tokens": 119499432.0,
      "step": 53535
    },
    {
      "entropy": 5.761072206497192,
      "epoch": 6.013028584264615,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002076052380657838,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2354169338941574,
      "num_tokens": 119509976.0,
      "step": 53540
    },
    {
      "entropy": 5.752656316757202,
      "epoch": 6.013590161172573,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00020756693380144297,
      "loss": 4.9172,
      "mean_token_accuracy": 0.23593052625656127,
      "num_tokens": 119520882.0,
      "step": 53545
    },
    {
      "entropy": 5.727467966079712,
      "epoch": 6.01415173808053,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020752863168419723,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2371185302734375,
      "num_tokens": 119531924.0,
      "step": 53550
    },
    {
      "entropy": 5.7352461338043215,
      "epoch": 6.014713314988487,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000207490331715266,
      "loss": 4.9196,
      "mean_token_accuracy": 0.23450703620910646,
      "num_tokens": 119542550.0,
      "step": 53555
    },
    {
      "entropy": 5.6628851890563965,
      "epoch": 6.015274891896445,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00020745203389586886,
      "loss": 4.8192,
      "mean_token_accuracy": 0.24277208745479584,
      "num_tokens": 119554787.0,
      "step": 53560
    },
    {
      "entropy": 5.729035568237305,
      "epoch": 6.015836468804403,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00020741373822722508,
      "loss": 4.9578,
      "mean_token_accuracy": 0.23391581773757936,
      "num_tokens": 119564951.0,
      "step": 53565
    },
    {
      "entropy": 5.67842116355896,
      "epoch": 6.016398045712361,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00020737544471055415,
      "loss": 4.7972,
      "mean_token_accuracy": 0.2363053470849991,
      "num_tokens": 119575621.0,
      "step": 53570
    },
    {
      "entropy": 5.711249208450317,
      "epoch": 6.016959622620318,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002073371533470752,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23904842436313628,
      "num_tokens": 119585711.0,
      "step": 53575
    },
    {
      "entropy": 5.814452075958252,
      "epoch": 6.017521199528275,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002072988641380076,
      "loss": 5.0384,
      "mean_token_accuracy": 0.22775165140628814,
      "num_tokens": 119596637.0,
      "step": 53580
    },
    {
      "entropy": 5.693995952606201,
      "epoch": 6.018082776436233,
      "grad_norm": 1.296875,
      "learning_rate": 0.00020726057708457036,
      "loss": 4.8445,
      "mean_token_accuracy": 0.23975272625684738,
      "num_tokens": 119608223.0,
      "step": 53585
    },
    {
      "entropy": 5.807613229751587,
      "epoch": 6.0186443533441905,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020722229218798267,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22581028640270234,
      "num_tokens": 119619917.0,
      "step": 53590
    },
    {
      "entropy": 5.704498147964477,
      "epoch": 6.019205930252148,
      "grad_norm": 1.328125,
      "learning_rate": 0.00020718400944946347,
      "loss": 4.8434,
      "mean_token_accuracy": 0.24026448726654054,
      "num_tokens": 119630163.0,
      "step": 53595
    },
    {
      "entropy": 5.705027770996094,
      "epoch": 6.019767507160106,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020714572887023174,
      "loss": 4.845,
      "mean_token_accuracy": 0.23928627669811248,
      "num_tokens": 119640077.0,
      "step": 53600
    },
    {
      "entropy": 5.75603609085083,
      "epoch": 6.020329084068063,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002071074504515063,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22802338898181915,
      "num_tokens": 119649548.0,
      "step": 53605
    },
    {
      "entropy": 5.760705423355103,
      "epoch": 6.020890660976021,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000207069174194506,
      "loss": 4.8752,
      "mean_token_accuracy": 0.24322169721126558,
      "num_tokens": 119661600.0,
      "step": 53610
    },
    {
      "entropy": 5.7566139698028564,
      "epoch": 6.021452237883978,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002070309001004494,
      "loss": 4.9066,
      "mean_token_accuracy": 0.24125340431928635,
      "num_tokens": 119672340.0,
      "step": 53615
    },
    {
      "entropy": 5.744892930984497,
      "epoch": 6.0220138147919355,
      "grad_norm": 1.328125,
      "learning_rate": 0.00020699262817055542,
      "loss": 4.8818,
      "mean_token_accuracy": 0.23681631535291672,
      "num_tokens": 119683641.0,
      "step": 53620
    },
    {
      "entropy": 5.726670408248902,
      "epoch": 6.022575391699894,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00020695435840604255,
      "loss": 4.8847,
      "mean_token_accuracy": 0.23790471702814103,
      "num_tokens": 119695373.0,
      "step": 53625
    },
    {
      "entropy": 5.719177341461181,
      "epoch": 6.023136968607851,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002069160908081293,
      "loss": 5.0057,
      "mean_token_accuracy": 0.23568787872791291,
      "num_tokens": 119706337.0,
      "step": 53630
    },
    {
      "entropy": 5.771531677246093,
      "epoch": 6.023698545515808,
      "grad_norm": 1.390625,
      "learning_rate": 0.00020687782537803407,
      "loss": 4.9608,
      "mean_token_accuracy": 0.23203803598880768,
      "num_tokens": 119717545.0,
      "step": 53635
    },
    {
      "entropy": 5.789220905303955,
      "epoch": 6.024260122423766,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00020683956211697537,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2341507777571678,
      "num_tokens": 119727699.0,
      "step": 53640
    },
    {
      "entropy": 5.774601030349731,
      "epoch": 6.024821699331723,
      "grad_norm": 1.359375,
      "learning_rate": 0.00020680130102617138,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23624784350395203,
      "num_tokens": 119738342.0,
      "step": 53645
    },
    {
      "entropy": 5.764207410812378,
      "epoch": 6.025383276239681,
      "grad_norm": 1.34375,
      "learning_rate": 0.00020676304210684045,
      "loss": 4.9147,
      "mean_token_accuracy": 0.23539799004793166,
      "num_tokens": 119749384.0,
      "step": 53650
    },
    {
      "entropy": 5.703004693984985,
      "epoch": 6.025944853147639,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00020672478536020068,
      "loss": 4.8831,
      "mean_token_accuracy": 0.2355142667889595,
      "num_tokens": 119759146.0,
      "step": 53655
    },
    {
      "entropy": 5.669059610366821,
      "epoch": 6.026506430055596,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002066865307874703,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23940826058387757,
      "num_tokens": 119770466.0,
      "step": 53660
    },
    {
      "entropy": 5.708546304702759,
      "epoch": 6.027068006963554,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002066482783898672,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22992478907108307,
      "num_tokens": 119781676.0,
      "step": 53665
    },
    {
      "entropy": 5.660578346252441,
      "epoch": 6.027629583871511,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002066100281686095,
      "loss": 4.7725,
      "mean_token_accuracy": 0.24605517238378524,
      "num_tokens": 119792581.0,
      "step": 53670
    },
    {
      "entropy": 5.702785301208496,
      "epoch": 6.0281911607794685,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000206571780124915,
      "loss": 4.9303,
      "mean_token_accuracy": 0.23629987686872483,
      "num_tokens": 119803748.0,
      "step": 53675
    },
    {
      "entropy": 5.797177124023437,
      "epoch": 6.028752737687427,
      "grad_norm": 1.34375,
      "learning_rate": 0.00020653353426000154,
      "loss": 4.9865,
      "mean_token_accuracy": 0.23138130009174346,
      "num_tokens": 119814203.0,
      "step": 53680
    },
    {
      "entropy": 5.690945672988891,
      "epoch": 6.029314314595384,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00020649529057508698,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23727195411920549,
      "num_tokens": 119826600.0,
      "step": 53685
    },
    {
      "entropy": 5.694852018356324,
      "epoch": 6.029875891503341,
      "grad_norm": 1.25,
      "learning_rate": 0.00020645704907138875,
      "loss": 4.8152,
      "mean_token_accuracy": 0.23906038850545883,
      "num_tokens": 119837815.0,
      "step": 53690
    },
    {
      "entropy": 5.783730983734131,
      "epoch": 6.030437468411299,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020641880975012477,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2290049359202385,
      "num_tokens": 119849741.0,
      "step": 53695
    },
    {
      "entropy": 5.794666719436646,
      "epoch": 6.030999045319256,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020638057261251254,
      "loss": 4.9465,
      "mean_token_accuracy": 0.23571833223104477,
      "num_tokens": 119860543.0,
      "step": 53700
    },
    {
      "entropy": 5.7451457500457765,
      "epoch": 6.0315606222272145,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00020634233765976946,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23788866996765137,
      "num_tokens": 119870513.0,
      "step": 53705
    },
    {
      "entropy": 5.672198295593262,
      "epoch": 6.032122199135172,
      "grad_norm": 1.390625,
      "learning_rate": 0.00020630410489311298,
      "loss": 4.8341,
      "mean_token_accuracy": 0.23900327682495118,
      "num_tokens": 119881295.0,
      "step": 53710
    },
    {
      "entropy": 5.744524431228638,
      "epoch": 6.032683776043129,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00020626587431376043,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22823336273431777,
      "num_tokens": 119893305.0,
      "step": 53715
    },
    {
      "entropy": 5.801928091049194,
      "epoch": 6.033245352951087,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002062276459229291,
      "loss": 4.9816,
      "mean_token_accuracy": 0.23228537142276764,
      "num_tokens": 119904038.0,
      "step": 53720
    },
    {
      "entropy": 5.802273893356324,
      "epoch": 6.033806929859044,
      "grad_norm": 1.21875,
      "learning_rate": 0.00020618941972183613,
      "loss": 4.9411,
      "mean_token_accuracy": 0.236823108792305,
      "num_tokens": 119915393.0,
      "step": 53725
    },
    {
      "entropy": 5.817698955535889,
      "epoch": 6.034368506767001,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002061511957116988,
      "loss": 5.0505,
      "mean_token_accuracy": 0.23439426422119142,
      "num_tokens": 119927406.0,
      "step": 53730
    },
    {
      "entropy": 5.693911027908325,
      "epoch": 6.0349300836749595,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00020611297389373413,
      "loss": 4.8447,
      "mean_token_accuracy": 0.23643926829099654,
      "num_tokens": 119938506.0,
      "step": 53735
    },
    {
      "entropy": 5.811730861663818,
      "epoch": 6.035491660582917,
      "grad_norm": 1.359375,
      "learning_rate": 0.00020607475426915906,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22662542760372162,
      "num_tokens": 119949322.0,
      "step": 53740
    },
    {
      "entropy": 5.707510471343994,
      "epoch": 6.036053237490874,
      "grad_norm": 1.25,
      "learning_rate": 0.00020603653683919055,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23199950456619262,
      "num_tokens": 119960714.0,
      "step": 53745
    },
    {
      "entropy": 5.705034065246582,
      "epoch": 6.036614814398832,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00020599832160504545,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23263156414031982,
      "num_tokens": 119971489.0,
      "step": 53750
    },
    {
      "entropy": 5.796267366409301,
      "epoch": 6.037176391306789,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00020596010856794057,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2350244089961052,
      "num_tokens": 119981520.0,
      "step": 53755
    },
    {
      "entropy": 5.7587385177612305,
      "epoch": 6.037737968214747,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00020592189772909248,
      "loss": 4.8497,
      "mean_token_accuracy": 0.24278274327516555,
      "num_tokens": 119992374.0,
      "step": 53760
    },
    {
      "entropy": 5.7127612113952635,
      "epoch": 6.038299545122705,
      "grad_norm": 1.21875,
      "learning_rate": 0.00020588368908971806,
      "loss": 4.9261,
      "mean_token_accuracy": 0.23997856080532073,
      "num_tokens": 120004262.0,
      "step": 53765
    },
    {
      "entropy": 5.73455924987793,
      "epoch": 6.038861122030662,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020584548265103382,
      "loss": 4.8646,
      "mean_token_accuracy": 0.23908915221691132,
      "num_tokens": 120015798.0,
      "step": 53770
    },
    {
      "entropy": 5.766662836074829,
      "epoch": 6.03942269893862,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00020580727841425624,
      "loss": 4.9751,
      "mean_token_accuracy": 0.23613746613264083,
      "num_tokens": 120027998.0,
      "step": 53775
    },
    {
      "entropy": 5.781463384628296,
      "epoch": 6.039984275846577,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00020576907638060167,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2218191683292389,
      "num_tokens": 120040592.0,
      "step": 53780
    },
    {
      "entropy": 5.692752552032471,
      "epoch": 6.040545852754534,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00020573087655128658,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2320569559931755,
      "num_tokens": 120051762.0,
      "step": 53785
    },
    {
      "entropy": 5.709880685806274,
      "epoch": 6.0411074296624925,
      "grad_norm": 1.375,
      "learning_rate": 0.0002056926789275273,
      "loss": 4.8333,
      "mean_token_accuracy": 0.24152581840753556,
      "num_tokens": 120061832.0,
      "step": 53790
    },
    {
      "entropy": 5.719084024429321,
      "epoch": 6.04166900657045,
      "grad_norm": 1.421875,
      "learning_rate": 0.00020565448351053995,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23519118279218673,
      "num_tokens": 120071446.0,
      "step": 53795
    },
    {
      "entropy": 5.704668855667114,
      "epoch": 6.042230583478408,
      "grad_norm": 1.328125,
      "learning_rate": 0.00020561629030154077,
      "loss": 4.8942,
      "mean_token_accuracy": 0.24346245527267457,
      "num_tokens": 120082376.0,
      "step": 53800
    },
    {
      "entropy": 5.73470778465271,
      "epoch": 6.042792160386365,
      "grad_norm": 1.375,
      "learning_rate": 0.00020557809930174582,
      "loss": 4.8312,
      "mean_token_accuracy": 0.23969689905643463,
      "num_tokens": 120093222.0,
      "step": 53805
    },
    {
      "entropy": 5.81372184753418,
      "epoch": 6.043353737294322,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002055399105123711,
      "loss": 5.0315,
      "mean_token_accuracy": 0.22320061326026916,
      "num_tokens": 120105336.0,
      "step": 53810
    },
    {
      "entropy": 5.728002071380615,
      "epoch": 6.04391531420228,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00020550172393463256,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2405652955174446,
      "num_tokens": 120115989.0,
      "step": 53815
    },
    {
      "entropy": 5.6649188041687015,
      "epoch": 6.044476891110238,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002054635395697461,
      "loss": 4.8669,
      "mean_token_accuracy": 0.24357237219810485,
      "num_tokens": 120127825.0,
      "step": 53820
    },
    {
      "entropy": 5.738976621627808,
      "epoch": 6.045038468018195,
      "grad_norm": 1.359375,
      "learning_rate": 0.00020542535741892754,
      "loss": 4.8343,
      "mean_token_accuracy": 0.24674847722053528,
      "num_tokens": 120138620.0,
      "step": 53825
    },
    {
      "entropy": 5.742858791351319,
      "epoch": 6.045600044926153,
      "grad_norm": 1.15625,
      "learning_rate": 0.00020538717748339248,
      "loss": 4.9574,
      "mean_token_accuracy": 0.23275481909513474,
      "num_tokens": 120149909.0,
      "step": 53830
    },
    {
      "entropy": 5.772373485565185,
      "epoch": 6.04616162183411,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020534899976435672,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2364571362733841,
      "num_tokens": 120162991.0,
      "step": 53835
    },
    {
      "entropy": 5.799155521392822,
      "epoch": 6.046723198742067,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002053108242630359,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23276612311601638,
      "num_tokens": 120174459.0,
      "step": 53840
    },
    {
      "entropy": 5.783517360687256,
      "epoch": 6.047284775650025,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00020527265098064547,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2392899513244629,
      "num_tokens": 120186494.0,
      "step": 53845
    },
    {
      "entropy": 5.7326771259307865,
      "epoch": 6.047846352557983,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020523447991840078,
      "loss": 4.8515,
      "mean_token_accuracy": 0.24183088093996047,
      "num_tokens": 120197360.0,
      "step": 53850
    },
    {
      "entropy": 5.655851888656616,
      "epoch": 6.048407929465941,
      "grad_norm": 1.203125,
      "learning_rate": 0.00020519631107751736,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23293246626853942,
      "num_tokens": 120209787.0,
      "step": 53855
    },
    {
      "entropy": 5.768480587005615,
      "epoch": 6.048969506373898,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002051581444592105,
      "loss": 4.9381,
      "mean_token_accuracy": 0.23924298733472824,
      "num_tokens": 120220937.0,
      "step": 53860
    },
    {
      "entropy": 5.692521572113037,
      "epoch": 6.049531083281855,
      "grad_norm": 1.34375,
      "learning_rate": 0.00020511998006469533,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2405211180448532,
      "num_tokens": 120231484.0,
      "step": 53865
    },
    {
      "entropy": 5.694277620315551,
      "epoch": 6.050092660189813,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002050818178951871,
      "loss": 4.8614,
      "mean_token_accuracy": 0.24559331089258193,
      "num_tokens": 120242801.0,
      "step": 53870
    },
    {
      "entropy": 5.758042430877685,
      "epoch": 6.0506542370977705,
      "grad_norm": 1.359375,
      "learning_rate": 0.000205043657951901,
      "loss": 4.8948,
      "mean_token_accuracy": 0.23870894610881804,
      "num_tokens": 120253515.0,
      "step": 53875
    },
    {
      "entropy": 5.770544624328613,
      "epoch": 6.051215814005728,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002050055002360519,
      "loss": 4.9551,
      "mean_token_accuracy": 0.23634896278381348,
      "num_tokens": 120264699.0,
      "step": 53880
    },
    {
      "entropy": 5.662475633621216,
      "epoch": 6.051777390913686,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002049673447488548,
      "loss": 4.8138,
      "mean_token_accuracy": 0.23750353008508682,
      "num_tokens": 120275645.0,
      "step": 53885
    },
    {
      "entropy": 5.784109401702881,
      "epoch": 6.052338967821643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00020492919149152467,
      "loss": 4.9497,
      "mean_token_accuracy": 0.23775530606508255,
      "num_tokens": 120287613.0,
      "step": 53890
    },
    {
      "entropy": 5.71296796798706,
      "epoch": 6.052900544729601,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00020489104046527624,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23846097886562348,
      "num_tokens": 120298879.0,
      "step": 53895
    },
    {
      "entropy": 5.762665271759033,
      "epoch": 6.053462121637558,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00020485289167132426,
      "loss": 4.9824,
      "mean_token_accuracy": 0.23443907052278518,
      "num_tokens": 120311889.0,
      "step": 53900
    },
    {
      "entropy": 5.752219772338867,
      "epoch": 6.054023698545516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00020481474511088333,
      "loss": 4.8835,
      "mean_token_accuracy": 0.24066879749298095,
      "num_tokens": 120324337.0,
      "step": 53905
    },
    {
      "entropy": 5.847218179702759,
      "epoch": 6.054585275453474,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002047766007851682,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2357998237013817,
      "num_tokens": 120337128.0,
      "step": 53910
    },
    {
      "entropy": 5.690852355957031,
      "epoch": 6.055146852361431,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00020473845869539337,
      "loss": 4.885,
      "mean_token_accuracy": 0.24447523355484008,
      "num_tokens": 120349880.0,
      "step": 53915
    },
    {
      "entropy": 5.712619924545288,
      "epoch": 6.055708429269388,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002047003188427733,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23826459497213365,
      "num_tokens": 120361673.0,
      "step": 53920
    },
    {
      "entropy": 5.6436169147491455,
      "epoch": 6.056270006177346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020466218122852225,
      "loss": 4.8339,
      "mean_token_accuracy": 0.24032874405384064,
      "num_tokens": 120372788.0,
      "step": 53925
    },
    {
      "entropy": 5.635148429870606,
      "epoch": 6.0568315830853034,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002046240458538547,
      "loss": 4.8587,
      "mean_token_accuracy": 0.24375026524066926,
      "num_tokens": 120384130.0,
      "step": 53930
    },
    {
      "entropy": 5.647089242935181,
      "epoch": 6.057393159993261,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002045859127199848,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2437271848320961,
      "num_tokens": 120395679.0,
      "step": 53935
    },
    {
      "entropy": 5.776175260543823,
      "epoch": 6.057954736901219,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00020454778182812673,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2321244776248932,
      "num_tokens": 120407458.0,
      "step": 53940
    },
    {
      "entropy": 5.7007087707519535,
      "epoch": 6.058516313809176,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002045096531794946,
      "loss": 4.81,
      "mean_token_accuracy": 0.24831959158182143,
      "num_tokens": 120418113.0,
      "step": 53945
    },
    {
      "entropy": 5.73693356513977,
      "epoch": 6.059077890717134,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002044715267753025,
      "loss": 4.8898,
      "mean_token_accuracy": 0.24327505826950074,
      "num_tokens": 120428676.0,
      "step": 53950
    },
    {
      "entropy": 5.738609170913696,
      "epoch": 6.059639467625091,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002044334026167643,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23275148421525954,
      "num_tokens": 120439586.0,
      "step": 53955
    },
    {
      "entropy": 5.748054885864258,
      "epoch": 6.0602010445330485,
      "grad_norm": 1.21875,
      "learning_rate": 0.000204395280705094,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23903768211603166,
      "num_tokens": 120450953.0,
      "step": 53960
    },
    {
      "entropy": 5.765005207061767,
      "epoch": 6.060762621441007,
      "grad_norm": 1.296875,
      "learning_rate": 0.00020435716104150525,
      "loss": 4.895,
      "mean_token_accuracy": 0.23919399380683898,
      "num_tokens": 120462804.0,
      "step": 53965
    },
    {
      "entropy": 5.769513511657715,
      "epoch": 6.061324198348964,
      "grad_norm": 1.375,
      "learning_rate": 0.00020431904362721198,
      "loss": 4.8917,
      "mean_token_accuracy": 0.24010637104511262,
      "num_tokens": 120473861.0,
      "step": 53970
    },
    {
      "entropy": 5.79967622756958,
      "epoch": 6.061885775256921,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00020428092846342766,
      "loss": 4.9491,
      "mean_token_accuracy": 0.2338499128818512,
      "num_tokens": 120484446.0,
      "step": 53975
    },
    {
      "entropy": 5.7066113471984865,
      "epoch": 6.062447352164879,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002042428155513661,
      "loss": 4.8619,
      "mean_token_accuracy": 0.24034639745950698,
      "num_tokens": 120496383.0,
      "step": 53980
    },
    {
      "entropy": 5.6274817943572994,
      "epoch": 6.063008929072836,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020420470489224075,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2428272321820259,
      "num_tokens": 120506855.0,
      "step": 53985
    },
    {
      "entropy": 5.701873731613159,
      "epoch": 6.0635705059807945,
      "grad_norm": 1.375,
      "learning_rate": 0.000204166596487265,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2366675078868866,
      "num_tokens": 120518746.0,
      "step": 53990
    },
    {
      "entropy": 5.750253057479858,
      "epoch": 6.064132082888752,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002041284903376523,
      "loss": 4.996,
      "mean_token_accuracy": 0.22835625261068343,
      "num_tokens": 120529843.0,
      "step": 53995
    },
    {
      "entropy": 5.781793212890625,
      "epoch": 6.064693659796709,
      "grad_norm": 1.265625,
      "learning_rate": 0.00020409038644461598,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2389568269252777,
      "num_tokens": 120541158.0,
      "step": 54000
    },
    {
      "epoch": 6.064693659796709,
      "eval_entropy": 5.479851139000712,
      "eval_loss": 5.038351058959961,
      "eval_mean_token_accuracy": 0.23652274418898026,
      "eval_num_tokens": 120541158.0,
      "eval_runtime": 18.2607,
      "eval_samples_per_second": 1698.35,
      "eval_steps_per_second": 212.314,
      "step": 54000
    }
  ],
  "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": 2.801651832096768e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}