zho-hans-100mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.096771046727149,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.490693664550781,
      "epoch": 0.0009755145839430299,
      "grad_norm": 30.5,
      "learning_rate": 2e-06,
      "loss": 13.5237,
      "mean_token_accuracy": 0.01689293053932488,
      "num_tokens": 17120.0,
      "step": 5
    },
    {
      "entropy": 4.48514404296875,
      "epoch": 0.0019510291678860598,
      "grad_norm": 30.875,
      "learning_rate": 4.5e-06,
      "loss": 13.421,
      "mean_token_accuracy": 0.018032675329595804,
      "num_tokens": 34182.0,
      "step": 10
    },
    {
      "entropy": 4.931977891921997,
      "epoch": 0.0029265437518290896,
      "grad_norm": 36.5,
      "learning_rate": 7e-06,
      "loss": 13.0743,
      "mean_token_accuracy": 0.012429539114236832,
      "num_tokens": 51205.0,
      "step": 15
    },
    {
      "entropy": 6.637991285324096,
      "epoch": 0.0039020583357721197,
      "grad_norm": 29.125,
      "learning_rate": 9.5e-06,
      "loss": 12.1322,
      "mean_token_accuracy": 0.008806258952245117,
      "num_tokens": 67722.0,
      "step": 20
    },
    {
      "entropy": 9.740472221374512,
      "epoch": 0.00487757291971515,
      "grad_norm": 10.8125,
      "learning_rate": 1.2e-05,
      "loss": 11.0704,
      "mean_token_accuracy": 0.007681116322055459,
      "num_tokens": 84230.0,
      "step": 25
    },
    {
      "entropy": 10.419333457946777,
      "epoch": 0.005853087503658179,
      "grad_norm": 3.46875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.7923,
      "mean_token_accuracy": 0.011759244091808795,
      "num_tokens": 100967.0,
      "step": 30
    },
    {
      "entropy": 10.586559581756593,
      "epoch": 0.00682860208760121,
      "grad_norm": 3.84375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.5501,
      "mean_token_accuracy": 0.061883827298879625,
      "num_tokens": 117450.0,
      "step": 35
    },
    {
      "entropy": 10.333465480804444,
      "epoch": 0.007804116671544239,
      "grad_norm": 4.09375,
      "learning_rate": 1.95e-05,
      "loss": 10.2155,
      "mean_token_accuracy": 0.07655484415590763,
      "num_tokens": 135328.0,
      "step": 40
    },
    {
      "entropy": 10.046442413330078,
      "epoch": 0.00877963125548727,
      "grad_norm": 1.7265625,
      "learning_rate": 2.2e-05,
      "loss": 10.0231,
      "mean_token_accuracy": 0.07796396911144257,
      "num_tokens": 152571.0,
      "step": 45
    },
    {
      "entropy": 10.165653514862061,
      "epoch": 0.0097551458394303,
      "grad_norm": 1.84375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.8849,
      "mean_token_accuracy": 0.08077023848891259,
      "num_tokens": 170216.0,
      "step": 50
    },
    {
      "entropy": 10.36428689956665,
      "epoch": 0.010730660423373329,
      "grad_norm": 1.7265625,
      "learning_rate": 2.7e-05,
      "loss": 9.7759,
      "mean_token_accuracy": 0.0794611006975174,
      "num_tokens": 186513.0,
      "step": 55
    },
    {
      "entropy": 10.221637153625489,
      "epoch": 0.011706175007316359,
      "grad_norm": 1.375,
      "learning_rate": 2.95e-05,
      "loss": 9.7436,
      "mean_token_accuracy": 0.07498926073312759,
      "num_tokens": 202600.0,
      "step": 60
    },
    {
      "entropy": 10.093542098999023,
      "epoch": 0.01268168959125939,
      "grad_norm": 1.953125,
      "learning_rate": 3.2e-05,
      "loss": 9.6498,
      "mean_token_accuracy": 0.08184118494391442,
      "num_tokens": 220270.0,
      "step": 65
    },
    {
      "entropy": 10.226324462890625,
      "epoch": 0.01365720417520242,
      "grad_norm": 1.7578125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.5928,
      "mean_token_accuracy": 0.0831097610294819,
      "num_tokens": 236810.0,
      "step": 70
    },
    {
      "entropy": 10.119788265228271,
      "epoch": 0.014632718759145449,
      "grad_norm": 1.2109375,
      "learning_rate": 3.7e-05,
      "loss": 9.5686,
      "mean_token_accuracy": 0.07497851252555847,
      "num_tokens": 252559.0,
      "step": 75
    },
    {
      "entropy": 10.039910221099854,
      "epoch": 0.015608233343088479,
      "grad_norm": 1.140625,
      "learning_rate": 3.95e-05,
      "loss": 9.4446,
      "mean_token_accuracy": 0.08544577062129974,
      "num_tokens": 268507.0,
      "step": 80
    },
    {
      "entropy": 10.107688331604004,
      "epoch": 0.01658374792703151,
      "grad_norm": 1.3671875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.4306,
      "mean_token_accuracy": 0.08484108000993729,
      "num_tokens": 284776.0,
      "step": 85
    },
    {
      "entropy": 10.037123680114746,
      "epoch": 0.01755926251097454,
      "grad_norm": 1.09375,
      "learning_rate": 4.45e-05,
      "loss": 9.3856,
      "mean_token_accuracy": 0.08017717525362969,
      "num_tokens": 300937.0,
      "step": 90
    },
    {
      "entropy": 10.079811477661133,
      "epoch": 0.01853477709491757,
      "grad_norm": 1.2578125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.2975,
      "mean_token_accuracy": 0.0832655370235443,
      "num_tokens": 317464.0,
      "step": 95
    },
    {
      "entropy": 9.896672248840332,
      "epoch": 0.0195102916788606,
      "grad_norm": 1.3828125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.1675,
      "mean_token_accuracy": 0.08660522624850273,
      "num_tokens": 334022.0,
      "step": 100
    },
    {
      "entropy": 9.873467636108398,
      "epoch": 0.02048580626280363,
      "grad_norm": 1.078125,
      "learning_rate": 5.2e-05,
      "loss": 9.1306,
      "mean_token_accuracy": 0.08643193244934082,
      "num_tokens": 350545.0,
      "step": 105
    },
    {
      "entropy": 9.815353202819825,
      "epoch": 0.021461320846746658,
      "grad_norm": 0.921875,
      "learning_rate": 5.45e-05,
      "loss": 9.0525,
      "mean_token_accuracy": 0.0924916334450245,
      "num_tokens": 368878.0,
      "step": 110
    },
    {
      "entropy": 9.754250431060791,
      "epoch": 0.022436835430689687,
      "grad_norm": 0.96484375,
      "learning_rate": 5.7e-05,
      "loss": 9.034,
      "mean_token_accuracy": 0.08547283411026001,
      "num_tokens": 388381.0,
      "step": 115
    },
    {
      "entropy": 9.797440242767333,
      "epoch": 0.023412350014632717,
      "grad_norm": 1.0859375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.9648,
      "mean_token_accuracy": 0.07924327179789543,
      "num_tokens": 403727.0,
      "step": 120
    },
    {
      "entropy": 9.65400218963623,
      "epoch": 0.02438786459857575,
      "grad_norm": 0.90625,
      "learning_rate": 6.2e-05,
      "loss": 8.9086,
      "mean_token_accuracy": 0.08146455883979797,
      "num_tokens": 419046.0,
      "step": 125
    },
    {
      "entropy": 9.547494602203368,
      "epoch": 0.02536337918251878,
      "grad_norm": 2.015625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.805,
      "mean_token_accuracy": 0.08776633739471436,
      "num_tokens": 435186.0,
      "step": 130
    },
    {
      "entropy": 9.431383323669433,
      "epoch": 0.02633889376646181,
      "grad_norm": 0.890625,
      "learning_rate": 6.7e-05,
      "loss": 8.7668,
      "mean_token_accuracy": 0.0854455754160881,
      "num_tokens": 451552.0,
      "step": 135
    },
    {
      "entropy": 9.342393016815185,
      "epoch": 0.02731440835040484,
      "grad_norm": 1.25,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.7312,
      "mean_token_accuracy": 0.08847364038228989,
      "num_tokens": 468006.0,
      "step": 140
    },
    {
      "entropy": 9.34831943511963,
      "epoch": 0.02828992293434787,
      "grad_norm": 0.91015625,
      "learning_rate": 7.2e-05,
      "loss": 8.6737,
      "mean_token_accuracy": 0.08393737226724625,
      "num_tokens": 484381.0,
      "step": 145
    },
    {
      "entropy": 9.162296390533447,
      "epoch": 0.029265437518290898,
      "grad_norm": 1.3046875,
      "learning_rate": 7.45e-05,
      "loss": 8.68,
      "mean_token_accuracy": 0.0858243778347969,
      "num_tokens": 501309.0,
      "step": 150
    },
    {
      "entropy": 9.130147743225098,
      "epoch": 0.030240952102233928,
      "grad_norm": 0.9140625,
      "learning_rate": 7.7e-05,
      "loss": 8.6269,
      "mean_token_accuracy": 0.08063350766897201,
      "num_tokens": 518290.0,
      "step": 155
    },
    {
      "entropy": 9.100608634948731,
      "epoch": 0.031216466686176957,
      "grad_norm": 0.765625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.5909,
      "mean_token_accuracy": 0.08441304340958596,
      "num_tokens": 533700.0,
      "step": 160
    },
    {
      "entropy": 8.903340816497803,
      "epoch": 0.03219198127011999,
      "grad_norm": 0.72265625,
      "learning_rate": 8.2e-05,
      "loss": 8.4616,
      "mean_token_accuracy": 0.09071405306458473,
      "num_tokens": 548935.0,
      "step": 165
    },
    {
      "entropy": 8.944788551330566,
      "epoch": 0.03316749585406302,
      "grad_norm": 0.6640625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.5225,
      "mean_token_accuracy": 0.08350471183657646,
      "num_tokens": 565869.0,
      "step": 170
    },
    {
      "entropy": 8.891490459442139,
      "epoch": 0.03414301043800605,
      "grad_norm": 1.0234375,
      "learning_rate": 8.7e-05,
      "loss": 8.5397,
      "mean_token_accuracy": 0.0859730139374733,
      "num_tokens": 583065.0,
      "step": 175
    },
    {
      "entropy": 8.781160163879395,
      "epoch": 0.03511852502194908,
      "grad_norm": 1.1171875,
      "learning_rate": 8.95e-05,
      "loss": 8.5389,
      "mean_token_accuracy": 0.08567977696657181,
      "num_tokens": 599875.0,
      "step": 180
    },
    {
      "entropy": 8.857763767242432,
      "epoch": 0.03609403960589211,
      "grad_norm": 0.77734375,
      "learning_rate": 9.2e-05,
      "loss": 8.502,
      "mean_token_accuracy": 0.08799432292580604,
      "num_tokens": 616296.0,
      "step": 185
    },
    {
      "entropy": 8.845306587219238,
      "epoch": 0.03706955418983514,
      "grad_norm": 0.86328125,
      "learning_rate": 9.45e-05,
      "loss": 8.4613,
      "mean_token_accuracy": 0.08114315047860146,
      "num_tokens": 632548.0,
      "step": 190
    },
    {
      "entropy": 8.763722133636474,
      "epoch": 0.03804506877377817,
      "grad_norm": 0.9375,
      "learning_rate": 9.7e-05,
      "loss": 8.4714,
      "mean_token_accuracy": 0.08705383539199829,
      "num_tokens": 650395.0,
      "step": 195
    },
    {
      "entropy": 8.692799091339111,
      "epoch": 0.0390205833577212,
      "grad_norm": 0.7578125,
      "learning_rate": 9.95e-05,
      "loss": 8.4872,
      "mean_token_accuracy": 0.09539326205849648,
      "num_tokens": 668453.0,
      "step": 200
    },
    {
      "entropy": 8.626627922058105,
      "epoch": 0.03999609794166423,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000102,
      "loss": 8.3857,
      "mean_token_accuracy": 0.08746680319309234,
      "num_tokens": 683914.0,
      "step": 205
    },
    {
      "entropy": 8.74559440612793,
      "epoch": 0.04097161252560726,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.4933,
      "mean_token_accuracy": 0.08703801408410072,
      "num_tokens": 700480.0,
      "step": 210
    },
    {
      "entropy": 8.703277587890625,
      "epoch": 0.041947127109550286,
      "grad_norm": 0.796875,
      "learning_rate": 0.000107,
      "loss": 8.472,
      "mean_token_accuracy": 0.08402965739369392,
      "num_tokens": 716604.0,
      "step": 215
    },
    {
      "entropy": 8.640272426605225,
      "epoch": 0.042922641693493316,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001095,
      "loss": 8.4605,
      "mean_token_accuracy": 0.08894563764333725,
      "num_tokens": 732697.0,
      "step": 220
    },
    {
      "entropy": 8.721662902832032,
      "epoch": 0.043898156277436345,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000112,
      "loss": 8.5299,
      "mean_token_accuracy": 0.08261891677975655,
      "num_tokens": 749072.0,
      "step": 225
    },
    {
      "entropy": 8.660103130340577,
      "epoch": 0.044873670861379375,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0001145,
      "loss": 8.4681,
      "mean_token_accuracy": 0.08579774647951126,
      "num_tokens": 767826.0,
      "step": 230
    },
    {
      "entropy": 8.619204425811768,
      "epoch": 0.045849185445322405,
      "grad_norm": 1.46875,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.3936,
      "mean_token_accuracy": 0.0902378536760807,
      "num_tokens": 786863.0,
      "step": 235
    },
    {
      "entropy": 8.532190132141114,
      "epoch": 0.046824700029265434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.3713,
      "mean_token_accuracy": 0.10228274986147881,
      "num_tokens": 803925.0,
      "step": 240
    },
    {
      "entropy": 8.610904693603516,
      "epoch": 0.04780021461320847,
      "grad_norm": 0.890625,
      "learning_rate": 0.000122,
      "loss": 8.4531,
      "mean_token_accuracy": 0.09021551758050919,
      "num_tokens": 822088.0,
      "step": 245
    },
    {
      "entropy": 8.60029354095459,
      "epoch": 0.0487757291971515,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001245,
      "loss": 8.453,
      "mean_token_accuracy": 0.08811493217945099,
      "num_tokens": 837991.0,
      "step": 250
    },
    {
      "entropy": 8.63068675994873,
      "epoch": 0.04975124378109453,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000127,
      "loss": 8.3579,
      "mean_token_accuracy": 0.09977038279175758,
      "num_tokens": 855593.0,
      "step": 255
    },
    {
      "entropy": 8.618537425994873,
      "epoch": 0.05072675836503756,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001295,
      "loss": 8.4751,
      "mean_token_accuracy": 0.08795029670000076,
      "num_tokens": 873696.0,
      "step": 260
    },
    {
      "entropy": 8.56423215866089,
      "epoch": 0.05170227294898059,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000132,
      "loss": 8.3553,
      "mean_token_accuracy": 0.09531263113021851,
      "num_tokens": 890023.0,
      "step": 265
    },
    {
      "entropy": 8.592487812042236,
      "epoch": 0.05267778753292362,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.4957,
      "mean_token_accuracy": 0.086310463398695,
      "num_tokens": 906492.0,
      "step": 270
    },
    {
      "entropy": 8.64133062362671,
      "epoch": 0.05365330211686665,
      "grad_norm": 1.0,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.434,
      "mean_token_accuracy": 0.08785242363810539,
      "num_tokens": 921837.0,
      "step": 275
    },
    {
      "entropy": 8.58052043914795,
      "epoch": 0.05462881670080968,
      "grad_norm": 1.234375,
      "learning_rate": 0.0001395,
      "loss": 8.377,
      "mean_token_accuracy": 0.09836494624614715,
      "num_tokens": 938771.0,
      "step": 280
    },
    {
      "entropy": 8.553132820129395,
      "epoch": 0.05560433128475271,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.3995,
      "mean_token_accuracy": 0.08714988306164742,
      "num_tokens": 954562.0,
      "step": 285
    },
    {
      "entropy": 8.662264537811279,
      "epoch": 0.05657984586869574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0001445,
      "loss": 8.343,
      "mean_token_accuracy": 0.09556876793503762,
      "num_tokens": 973700.0,
      "step": 290
    },
    {
      "entropy": 8.443976783752442,
      "epoch": 0.05755536045263877,
      "grad_norm": 0.75,
      "learning_rate": 0.000147,
      "loss": 8.3889,
      "mean_token_accuracy": 0.08947161361575126,
      "num_tokens": 990688.0,
      "step": 295
    },
    {
      "entropy": 8.586378479003907,
      "epoch": 0.058530875036581796,
      "grad_norm": 0.734375,
      "learning_rate": 0.0001495,
      "loss": 8.4625,
      "mean_token_accuracy": 0.08533673211932183,
      "num_tokens": 1008144.0,
      "step": 300
    },
    {
      "entropy": 8.544894409179687,
      "epoch": 0.059506389620524826,
      "grad_norm": 0.765625,
      "learning_rate": 0.000152,
      "loss": 8.4238,
      "mean_token_accuracy": 0.08675077632069587,
      "num_tokens": 1027379.0,
      "step": 305
    },
    {
      "entropy": 8.509546852111816,
      "epoch": 0.060481904204467855,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00015450000000000001,
      "loss": 8.3964,
      "mean_token_accuracy": 0.09509446248412132,
      "num_tokens": 1043999.0,
      "step": 310
    },
    {
      "entropy": 8.641741275787354,
      "epoch": 0.061457418788410885,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000157,
      "loss": 8.4003,
      "mean_token_accuracy": 0.08781646862626076,
      "num_tokens": 1062237.0,
      "step": 315
    },
    {
      "entropy": 8.513209438323974,
      "epoch": 0.062432933372353915,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0001595,
      "loss": 8.3329,
      "mean_token_accuracy": 0.09510535821318626,
      "num_tokens": 1078092.0,
      "step": 320
    },
    {
      "entropy": 8.517814922332764,
      "epoch": 0.06340844795629695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000162,
      "loss": 8.389,
      "mean_token_accuracy": 0.09345002174377441,
      "num_tokens": 1095124.0,
      "step": 325
    },
    {
      "entropy": 8.54736795425415,
      "epoch": 0.06438396254023998,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00016450000000000001,
      "loss": 8.3712,
      "mean_token_accuracy": 0.09042947217822075,
      "num_tokens": 1112775.0,
      "step": 330
    },
    {
      "entropy": 8.515040588378906,
      "epoch": 0.06535947712418301,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00016700000000000002,
      "loss": 8.3097,
      "mean_token_accuracy": 0.09608379825949669,
      "num_tokens": 1129726.0,
      "step": 335
    },
    {
      "entropy": 8.47906036376953,
      "epoch": 0.06633499170812604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00016950000000000003,
      "loss": 8.3571,
      "mean_token_accuracy": 0.09651407599449158,
      "num_tokens": 1144034.0,
      "step": 340
    },
    {
      "entropy": 8.513033390045166,
      "epoch": 0.06731050629206907,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00017199999999999998,
      "loss": 8.3097,
      "mean_token_accuracy": 0.0950991727411747,
      "num_tokens": 1159633.0,
      "step": 345
    },
    {
      "entropy": 8.47611141204834,
      "epoch": 0.0682860208760121,
      "grad_norm": 0.90625,
      "learning_rate": 0.00017449999999999999,
      "loss": 8.2604,
      "mean_token_accuracy": 0.09262074902653694,
      "num_tokens": 1175350.0,
      "step": 350
    },
    {
      "entropy": 8.391336727142335,
      "epoch": 0.06926153545995513,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000177,
      "loss": 8.188,
      "mean_token_accuracy": 0.09829908162355423,
      "num_tokens": 1193121.0,
      "step": 355
    },
    {
      "entropy": 8.497109603881835,
      "epoch": 0.07023705004389816,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0001795,
      "loss": 8.3481,
      "mean_token_accuracy": 0.09156165719032287,
      "num_tokens": 1209896.0,
      "step": 360
    },
    {
      "entropy": 8.4661789894104,
      "epoch": 0.07121256462784119,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000182,
      "loss": 8.3448,
      "mean_token_accuracy": 0.08826714232563973,
      "num_tokens": 1227011.0,
      "step": 365
    },
    {
      "entropy": 8.423711013793945,
      "epoch": 0.07218807921178422,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001845,
      "loss": 8.263,
      "mean_token_accuracy": 0.09375467002391816,
      "num_tokens": 1241595.0,
      "step": 370
    },
    {
      "entropy": 8.45018196105957,
      "epoch": 0.07316359379572725,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000187,
      "loss": 8.2753,
      "mean_token_accuracy": 0.08940806537866593,
      "num_tokens": 1256919.0,
      "step": 375
    },
    {
      "entropy": 8.34131474494934,
      "epoch": 0.07413910837967028,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0001895,
      "loss": 8.267,
      "mean_token_accuracy": 0.10843634009361267,
      "num_tokens": 1274344.0,
      "step": 380
    },
    {
      "entropy": 8.417487525939942,
      "epoch": 0.0751146229636133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000192,
      "loss": 8.2498,
      "mean_token_accuracy": 0.09864807277917861,
      "num_tokens": 1290252.0,
      "step": 385
    },
    {
      "entropy": 8.479813480377198,
      "epoch": 0.07609013754755634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001945,
      "loss": 8.3017,
      "mean_token_accuracy": 0.09188697189092636,
      "num_tokens": 1306987.0,
      "step": 390
    },
    {
      "entropy": 8.460020828247071,
      "epoch": 0.07706565213149937,
      "grad_norm": 1.578125,
      "learning_rate": 0.00019700000000000002,
      "loss": 8.3412,
      "mean_token_accuracy": 0.0870269738137722,
      "num_tokens": 1323031.0,
      "step": 395
    },
    {
      "entropy": 8.439099597930909,
      "epoch": 0.0780411667154424,
      "grad_norm": 1.078125,
      "learning_rate": 0.00019950000000000002,
      "loss": 8.2752,
      "mean_token_accuracy": 0.09002957195043564,
      "num_tokens": 1339592.0,
      "step": 400
    },
    {
      "entropy": 8.53223638534546,
      "epoch": 0.07901668129938542,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000202,
      "loss": 8.247,
      "mean_token_accuracy": 0.09930683076381683,
      "num_tokens": 1358131.0,
      "step": 405
    },
    {
      "entropy": 8.311397790908813,
      "epoch": 0.07999219588332845,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00020449999999999998,
      "loss": 8.1704,
      "mean_token_accuracy": 0.10103706270456314,
      "num_tokens": 1375705.0,
      "step": 410
    },
    {
      "entropy": 8.206231212615966,
      "epoch": 0.08096771046727148,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000207,
      "loss": 8.2102,
      "mean_token_accuracy": 0.0998042032122612,
      "num_tokens": 1392736.0,
      "step": 415
    },
    {
      "entropy": 8.423250007629395,
      "epoch": 0.08194322505121451,
      "grad_norm": 0.828125,
      "learning_rate": 0.0002095,
      "loss": 8.2514,
      "mean_token_accuracy": 0.09237718880176544,
      "num_tokens": 1408536.0,
      "step": 420
    },
    {
      "entropy": 8.314301776885987,
      "epoch": 0.08291873963515754,
      "grad_norm": 0.875,
      "learning_rate": 0.000212,
      "loss": 8.2075,
      "mean_token_accuracy": 0.09687253460288048,
      "num_tokens": 1425423.0,
      "step": 425
    },
    {
      "entropy": 8.404624748229981,
      "epoch": 0.08389425421910057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002145,
      "loss": 8.2317,
      "mean_token_accuracy": 0.09046228304505348,
      "num_tokens": 1442420.0,
      "step": 430
    },
    {
      "entropy": 8.372211837768555,
      "epoch": 0.0848697688030436,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00021700000000000002,
      "loss": 8.162,
      "mean_token_accuracy": 0.0996602401137352,
      "num_tokens": 1458835.0,
      "step": 435
    },
    {
      "entropy": 8.417982482910157,
      "epoch": 0.08584528338698663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002195,
      "loss": 8.2796,
      "mean_token_accuracy": 0.08996990695595741,
      "num_tokens": 1475563.0,
      "step": 440
    },
    {
      "entropy": 8.326233100891113,
      "epoch": 0.08682079797092966,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000222,
      "loss": 8.2409,
      "mean_token_accuracy": 0.10384741872549057,
      "num_tokens": 1492687.0,
      "step": 445
    },
    {
      "entropy": 8.444601440429688,
      "epoch": 0.08779631255487269,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002245,
      "loss": 8.2404,
      "mean_token_accuracy": 0.09174655601382256,
      "num_tokens": 1508394.0,
      "step": 450
    },
    {
      "entropy": 8.261253166198731,
      "epoch": 0.08877182713881572,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00022700000000000002,
      "loss": 8.2181,
      "mean_token_accuracy": 0.10230232402682304,
      "num_tokens": 1527112.0,
      "step": 455
    },
    {
      "entropy": 8.32946844100952,
      "epoch": 0.08974734172275875,
      "grad_norm": 1.359375,
      "learning_rate": 0.00022950000000000002,
      "loss": 8.2507,
      "mean_token_accuracy": 0.0994840294122696,
      "num_tokens": 1545401.0,
      "step": 460
    },
    {
      "entropy": 8.400847578048706,
      "epoch": 0.09072285630670178,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023200000000000003,
      "loss": 8.1099,
      "mean_token_accuracy": 0.09926176220178604,
      "num_tokens": 1561621.0,
      "step": 465
    },
    {
      "entropy": 8.35053997039795,
      "epoch": 0.09169837089064481,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00023449999999999998,
      "loss": 8.2428,
      "mean_token_accuracy": 0.09348660483956336,
      "num_tokens": 1579225.0,
      "step": 470
    },
    {
      "entropy": 8.344873905181885,
      "epoch": 0.09267388547458784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000237,
      "loss": 8.1339,
      "mean_token_accuracy": 0.09716273322701455,
      "num_tokens": 1596125.0,
      "step": 475
    },
    {
      "entropy": 8.30934934616089,
      "epoch": 0.09364940005853087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002395,
      "loss": 8.2589,
      "mean_token_accuracy": 0.0886540800333023,
      "num_tokens": 1613451.0,
      "step": 480
    },
    {
      "entropy": 8.344436740875244,
      "epoch": 0.09462491464247391,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000242,
      "loss": 8.2056,
      "mean_token_accuracy": 0.09579794630408286,
      "num_tokens": 1630929.0,
      "step": 485
    },
    {
      "entropy": 8.309423828125,
      "epoch": 0.09560042922641694,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002445,
      "loss": 8.2009,
      "mean_token_accuracy": 0.0951028324663639,
      "num_tokens": 1647379.0,
      "step": 490
    },
    {
      "entropy": 8.183022069931031,
      "epoch": 0.09657594381035997,
      "grad_norm": 1.015625,
      "learning_rate": 0.000247,
      "loss": 7.9719,
      "mean_token_accuracy": 0.11050360500812531,
      "num_tokens": 1664415.0,
      "step": 495
    },
    {
      "entropy": 8.176769638061524,
      "epoch": 0.097551458394303,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002495,
      "loss": 8.0551,
      "mean_token_accuracy": 0.10060225948691368,
      "num_tokens": 1679556.0,
      "step": 500
    },
    {
      "entropy": 8.15018105506897,
      "epoch": 0.09852697297824603,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000252,
      "loss": 7.9955,
      "mean_token_accuracy": 0.10563117042183875,
      "num_tokens": 1696727.0,
      "step": 505
    },
    {
      "entropy": 8.249298286437988,
      "epoch": 0.09950248756218906,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002545,
      "loss": 8.0332,
      "mean_token_accuracy": 0.11021090745925903,
      "num_tokens": 1713918.0,
      "step": 510
    },
    {
      "entropy": 8.252583122253418,
      "epoch": 0.10047800214613209,
      "grad_norm": 1.109375,
      "learning_rate": 0.000257,
      "loss": 8.2011,
      "mean_token_accuracy": 0.0905926614999771,
      "num_tokens": 1729581.0,
      "step": 515
    },
    {
      "entropy": 8.262310695648193,
      "epoch": 0.10145351673007512,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002595,
      "loss": 8.0441,
      "mean_token_accuracy": 0.09327254816889763,
      "num_tokens": 1746367.0,
      "step": 520
    },
    {
      "entropy": 8.325894927978515,
      "epoch": 0.10242903131401815,
      "grad_norm": 1.234375,
      "learning_rate": 0.000262,
      "loss": 8.1618,
      "mean_token_accuracy": 0.09271856918931007,
      "num_tokens": 1765009.0,
      "step": 525
    },
    {
      "entropy": 8.254261589050293,
      "epoch": 0.10340454589796118,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026450000000000003,
      "loss": 8.1649,
      "mean_token_accuracy": 0.09503179416060448,
      "num_tokens": 1782113.0,
      "step": 530
    },
    {
      "entropy": 8.215056705474854,
      "epoch": 0.10438006048190421,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00026700000000000004,
      "loss": 8.0514,
      "mean_token_accuracy": 0.09871876537799835,
      "num_tokens": 1798301.0,
      "step": 535
    },
    {
      "entropy": 8.213224983215332,
      "epoch": 0.10535557506584724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026950000000000005,
      "loss": 8.086,
      "mean_token_accuracy": 0.09899957552552223,
      "num_tokens": 1815059.0,
      "step": 540
    },
    {
      "entropy": 8.175263071060181,
      "epoch": 0.10633108964979027,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.9918,
      "mean_token_accuracy": 0.10212236493825913,
      "num_tokens": 1831699.0,
      "step": 545
    },
    {
      "entropy": 8.153118419647218,
      "epoch": 0.1073066042337333,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002745,
      "loss": 8.066,
      "mean_token_accuracy": 0.09812919944524764,
      "num_tokens": 1847993.0,
      "step": 550
    },
    {
      "entropy": 8.155826520919799,
      "epoch": 0.10828211881767633,
      "grad_norm": 1.140625,
      "learning_rate": 0.000277,
      "loss": 7.8979,
      "mean_token_accuracy": 0.11081337481737137,
      "num_tokens": 1864753.0,
      "step": 555
    },
    {
      "entropy": 8.19793872833252,
      "epoch": 0.10925763340161936,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002795,
      "loss": 8.0736,
      "mean_token_accuracy": 0.10154270082712173,
      "num_tokens": 1882154.0,
      "step": 560
    },
    {
      "entropy": 8.117816829681397,
      "epoch": 0.11023314798556239,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.9647,
      "mean_token_accuracy": 0.10528781563043595,
      "num_tokens": 1899683.0,
      "step": 565
    },
    {
      "entropy": 8.081390380859375,
      "epoch": 0.11120866256950541,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002845,
      "loss": 7.9668,
      "mean_token_accuracy": 0.10001010969281196,
      "num_tokens": 1918489.0,
      "step": 570
    },
    {
      "entropy": 8.130353212356567,
      "epoch": 0.11218417715344844,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000287,
      "loss": 8.011,
      "mean_token_accuracy": 0.09865634441375733,
      "num_tokens": 1935609.0,
      "step": 575
    },
    {
      "entropy": 8.109779262542725,
      "epoch": 0.11315969173739147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002895,
      "loss": 8.0395,
      "mean_token_accuracy": 0.09440108239650727,
      "num_tokens": 1951311.0,
      "step": 580
    },
    {
      "entropy": 8.174573421478271,
      "epoch": 0.1141352063213345,
      "grad_norm": 1.140625,
      "learning_rate": 0.000292,
      "loss": 8.1061,
      "mean_token_accuracy": 0.0926342748105526,
      "num_tokens": 1968908.0,
      "step": 585
    },
    {
      "entropy": 8.177379894256593,
      "epoch": 0.11511072090527753,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002945,
      "loss": 7.9996,
      "mean_token_accuracy": 0.09846640899777412,
      "num_tokens": 1983807.0,
      "step": 590
    },
    {
      "entropy": 8.253669786453248,
      "epoch": 0.11608623548922056,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000297,
      "loss": 8.051,
      "mean_token_accuracy": 0.09801550693809986,
      "num_tokens": 2000219.0,
      "step": 595
    },
    {
      "entropy": 8.11186671257019,
      "epoch": 0.11706175007316359,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002995,
      "loss": 7.9772,
      "mean_token_accuracy": 0.10445283949375153,
      "num_tokens": 2017360.0,
      "step": 600
    },
    {
      "entropy": 8.043530797958374,
      "epoch": 0.11803726465710662,
      "grad_norm": 1.015625,
      "learning_rate": 0.000302,
      "loss": 7.9765,
      "mean_token_accuracy": 0.10461995676159859,
      "num_tokens": 2034953.0,
      "step": 605
    },
    {
      "entropy": 8.068679714202881,
      "epoch": 0.11901277924104965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003045,
      "loss": 7.9205,
      "mean_token_accuracy": 0.10359138026833534,
      "num_tokens": 2051551.0,
      "step": 610
    },
    {
      "entropy": 8.153577280044555,
      "epoch": 0.11998829382499268,
      "grad_norm": 0.828125,
      "learning_rate": 0.000307,
      "loss": 7.9171,
      "mean_token_accuracy": 0.10637792795896531,
      "num_tokens": 2067188.0,
      "step": 615
    },
    {
      "entropy": 8.009168004989624,
      "epoch": 0.12096380840893571,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003095,
      "loss": 7.9537,
      "mean_token_accuracy": 0.10119322836399078,
      "num_tokens": 2082569.0,
      "step": 620
    },
    {
      "entropy": 8.181542825698852,
      "epoch": 0.12193932299287874,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000312,
      "loss": 8.024,
      "mean_token_accuracy": 0.09720693081617356,
      "num_tokens": 2099446.0,
      "step": 625
    },
    {
      "entropy": 8.081124019622802,
      "epoch": 0.12291483757682177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003145,
      "loss": 7.9494,
      "mean_token_accuracy": 0.10721178874373435,
      "num_tokens": 2116526.0,
      "step": 630
    },
    {
      "entropy": 8.138178634643555,
      "epoch": 0.1238903521607648,
      "grad_norm": 0.859375,
      "learning_rate": 0.000317,
      "loss": 7.9943,
      "mean_token_accuracy": 0.09462400078773499,
      "num_tokens": 2134624.0,
      "step": 635
    },
    {
      "entropy": 7.9849425792694095,
      "epoch": 0.12486586674470783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003195,
      "loss": 7.8401,
      "mean_token_accuracy": 0.10716697350144386,
      "num_tokens": 2152872.0,
      "step": 640
    },
    {
      "entropy": 8.098246574401855,
      "epoch": 0.12584138132865086,
      "grad_norm": 0.921875,
      "learning_rate": 0.000322,
      "loss": 7.9826,
      "mean_token_accuracy": 0.09792573973536492,
      "num_tokens": 2169480.0,
      "step": 645
    },
    {
      "entropy": 8.041435432434081,
      "epoch": 0.1268168959125939,
      "grad_norm": 0.859375,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.8415,
      "mean_token_accuracy": 0.11095677688717842,
      "num_tokens": 2186054.0,
      "step": 650
    },
    {
      "entropy": 8.054678249359132,
      "epoch": 0.12779241049653692,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.9727,
      "mean_token_accuracy": 0.0941509373486042,
      "num_tokens": 2202320.0,
      "step": 655
    },
    {
      "entropy": 7.981766939163208,
      "epoch": 0.12876792508047996,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.8641,
      "mean_token_accuracy": 0.10060839653015137,
      "num_tokens": 2218817.0,
      "step": 660
    },
    {
      "entropy": 8.093716478347778,
      "epoch": 0.12974343966442298,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033200000000000005,
      "loss": 8.0126,
      "mean_token_accuracy": 0.09739565700292588,
      "num_tokens": 2234845.0,
      "step": 665
    },
    {
      "entropy": 8.075230646133424,
      "epoch": 0.13071895424836602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.9454,
      "mean_token_accuracy": 0.09940807446837426,
      "num_tokens": 2251904.0,
      "step": 670
    },
    {
      "entropy": 7.942912292480469,
      "epoch": 0.13169446883230904,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000337,
      "loss": 7.9409,
      "mean_token_accuracy": 0.10191294699907302,
      "num_tokens": 2268343.0,
      "step": 675
    },
    {
      "entropy": 8.05466251373291,
      "epoch": 0.13266998341625208,
      "grad_norm": 1.25,
      "learning_rate": 0.0003395,
      "loss": 7.9356,
      "mean_token_accuracy": 0.09942738711833954,
      "num_tokens": 2284524.0,
      "step": 680
    },
    {
      "entropy": 7.954032278060913,
      "epoch": 0.1336454980001951,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000342,
      "loss": 7.8617,
      "mean_token_accuracy": 0.10729206651449204,
      "num_tokens": 2302350.0,
      "step": 685
    },
    {
      "entropy": 8.02203860282898,
      "epoch": 0.13462101258413814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.9402,
      "mean_token_accuracy": 0.09717295095324516,
      "num_tokens": 2319953.0,
      "step": 690
    },
    {
      "entropy": 8.024251461029053,
      "epoch": 0.13559652716808115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000347,
      "loss": 7.8139,
      "mean_token_accuracy": 0.10286439582705498,
      "num_tokens": 2336698.0,
      "step": 695
    },
    {
      "entropy": 7.851852083206177,
      "epoch": 0.1365720417520242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003495,
      "loss": 7.7683,
      "mean_token_accuracy": 0.10691581144928933,
      "num_tokens": 2353986.0,
      "step": 700
    },
    {
      "entropy": 7.9333963871002195,
      "epoch": 0.1375475563359672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000352,
      "loss": 7.8256,
      "mean_token_accuracy": 0.10445680022239685,
      "num_tokens": 2370320.0,
      "step": 705
    },
    {
      "entropy": 7.971817111968994,
      "epoch": 0.13852307091991026,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003545,
      "loss": 7.8781,
      "mean_token_accuracy": 0.0994373232126236,
      "num_tokens": 2385453.0,
      "step": 710
    },
    {
      "entropy": 7.98693323135376,
      "epoch": 0.13949858550385327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000357,
      "loss": 7.8666,
      "mean_token_accuracy": 0.10084751695394516,
      "num_tokens": 2401545.0,
      "step": 715
    },
    {
      "entropy": 7.911950635910034,
      "epoch": 0.14047410008779632,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003595,
      "loss": 7.8597,
      "mean_token_accuracy": 0.10480492040514947,
      "num_tokens": 2418414.0,
      "step": 720
    },
    {
      "entropy": 7.9676766872406,
      "epoch": 0.14144961467173933,
      "grad_norm": 1.0,
      "learning_rate": 0.000362,
      "loss": 7.922,
      "mean_token_accuracy": 0.10325973257422447,
      "num_tokens": 2434973.0,
      "step": 725
    },
    {
      "entropy": 7.979449939727783,
      "epoch": 0.14242512925568238,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003645,
      "loss": 7.9006,
      "mean_token_accuracy": 0.09724714756011962,
      "num_tokens": 2451053.0,
      "step": 730
    },
    {
      "entropy": 7.946141576766967,
      "epoch": 0.1434006438396254,
      "grad_norm": 0.875,
      "learning_rate": 0.000367,
      "loss": 7.8747,
      "mean_token_accuracy": 0.09853564202785492,
      "num_tokens": 2467280.0,
      "step": 735
    },
    {
      "entropy": 7.954542589187622,
      "epoch": 0.14437615842356843,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003695,
      "loss": 7.8025,
      "mean_token_accuracy": 0.10932159945368766,
      "num_tokens": 2484588.0,
      "step": 740
    },
    {
      "entropy": 7.866448926925659,
      "epoch": 0.14535167300751145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000372,
      "loss": 7.8115,
      "mean_token_accuracy": 0.1057325579226017,
      "num_tokens": 2501200.0,
      "step": 745
    },
    {
      "entropy": 7.84429702758789,
      "epoch": 0.1463271875914545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003745,
      "loss": 7.7671,
      "mean_token_accuracy": 0.11158974543213844,
      "num_tokens": 2517199.0,
      "step": 750
    },
    {
      "entropy": 7.894868183135986,
      "epoch": 0.1473027021753975,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000377,
      "loss": 7.8305,
      "mean_token_accuracy": 0.09836434125900269,
      "num_tokens": 2534824.0,
      "step": 755
    },
    {
      "entropy": 7.745544052124023,
      "epoch": 0.14827821675934055,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003795,
      "loss": 7.7139,
      "mean_token_accuracy": 0.10689699798822402,
      "num_tokens": 2550842.0,
      "step": 760
    },
    {
      "entropy": 7.920146465301514,
      "epoch": 0.14925373134328357,
      "grad_norm": 1.046875,
      "learning_rate": 0.000382,
      "loss": 7.6577,
      "mean_token_accuracy": 0.10967442616820336,
      "num_tokens": 2567878.0,
      "step": 765
    },
    {
      "entropy": 7.803998613357544,
      "epoch": 0.1502292459272266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003845,
      "loss": 7.7583,
      "mean_token_accuracy": 0.10285873413085937,
      "num_tokens": 2584148.0,
      "step": 770
    },
    {
      "entropy": 7.811223030090332,
      "epoch": 0.15120476051116966,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.7677,
      "mean_token_accuracy": 0.10665770620107651,
      "num_tokens": 2600067.0,
      "step": 775
    },
    {
      "entropy": 7.864786195755005,
      "epoch": 0.15218027509511267,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.747,
      "mean_token_accuracy": 0.10223184376955033,
      "num_tokens": 2615746.0,
      "step": 780
    },
    {
      "entropy": 7.87882251739502,
      "epoch": 0.15315578967905571,
      "grad_norm": 0.859375,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.7799,
      "mean_token_accuracy": 0.10211106389760971,
      "num_tokens": 2631641.0,
      "step": 785
    },
    {
      "entropy": 7.835959625244141,
      "epoch": 0.15413130426299873,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.7807,
      "mean_token_accuracy": 0.10545132532715798,
      "num_tokens": 2647015.0,
      "step": 790
    },
    {
      "entropy": 7.828943157196045,
      "epoch": 0.15510681884694177,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.6984,
      "mean_token_accuracy": 0.10220426917076111,
      "num_tokens": 2664045.0,
      "step": 795
    },
    {
      "entropy": 7.740712451934814,
      "epoch": 0.1560823334308848,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003995,
      "loss": 7.6759,
      "mean_token_accuracy": 0.10618537291884422,
      "num_tokens": 2680774.0,
      "step": 800
    },
    {
      "entropy": 7.804680252075196,
      "epoch": 0.15705784801482783,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000402,
      "loss": 7.7804,
      "mean_token_accuracy": 0.10261463448405266,
      "num_tokens": 2697719.0,
      "step": 805
    },
    {
      "entropy": 7.877059888839722,
      "epoch": 0.15803336259877085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004045,
      "loss": 7.7416,
      "mean_token_accuracy": 0.09565807282924652,
      "num_tokens": 2714518.0,
      "step": 810
    },
    {
      "entropy": 7.821208715438843,
      "epoch": 0.1590088771827139,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.7914,
      "mean_token_accuracy": 0.10448245778679847,
      "num_tokens": 2729288.0,
      "step": 815
    },
    {
      "entropy": 7.8001782417297365,
      "epoch": 0.1599843917666569,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004095,
      "loss": 7.782,
      "mean_token_accuracy": 0.10158958658576012,
      "num_tokens": 2746782.0,
      "step": 820
    },
    {
      "entropy": 7.77209358215332,
      "epoch": 0.16095990635059995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000412,
      "loss": 7.8125,
      "mean_token_accuracy": 0.09743853360414505,
      "num_tokens": 2763159.0,
      "step": 825
    },
    {
      "entropy": 7.774577522277832,
      "epoch": 0.16193542093454297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004145,
      "loss": 7.7575,
      "mean_token_accuracy": 0.10294459313154221,
      "num_tokens": 2780338.0,
      "step": 830
    },
    {
      "entropy": 7.727686357498169,
      "epoch": 0.162910935518486,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000417,
      "loss": 7.6655,
      "mean_token_accuracy": 0.1120273858308792,
      "num_tokens": 2796214.0,
      "step": 835
    },
    {
      "entropy": 7.858849382400512,
      "epoch": 0.16388645010242903,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004195,
      "loss": 7.7812,
      "mean_token_accuracy": 0.09670486152172089,
      "num_tokens": 2812745.0,
      "step": 840
    },
    {
      "entropy": 7.7591468334198,
      "epoch": 0.16486196468637207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000422,
      "loss": 7.6485,
      "mean_token_accuracy": 0.1039587639272213,
      "num_tokens": 2828288.0,
      "step": 845
    },
    {
      "entropy": 7.783154487609863,
      "epoch": 0.16583747927031509,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004245,
      "loss": 7.7887,
      "mean_token_accuracy": 0.09705574363470078,
      "num_tokens": 2843618.0,
      "step": 850
    },
    {
      "entropy": 7.673272943496704,
      "epoch": 0.16681299385425813,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000427,
      "loss": 7.5843,
      "mean_token_accuracy": 0.11242094039916992,
      "num_tokens": 2860659.0,
      "step": 855
    },
    {
      "entropy": 7.708390426635742,
      "epoch": 0.16778850843820114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004295,
      "loss": 7.6903,
      "mean_token_accuracy": 0.09876248240470886,
      "num_tokens": 2877482.0,
      "step": 860
    },
    {
      "entropy": 7.789376258850098,
      "epoch": 0.1687640230221442,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000432,
      "loss": 7.7691,
      "mean_token_accuracy": 0.09691327959299087,
      "num_tokens": 2895868.0,
      "step": 865
    },
    {
      "entropy": 7.818999862670898,
      "epoch": 0.1697395376060872,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004345,
      "loss": 7.7386,
      "mean_token_accuracy": 0.1005122035741806,
      "num_tokens": 2913103.0,
      "step": 870
    },
    {
      "entropy": 7.735997247695923,
      "epoch": 0.17071505219003025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000437,
      "loss": 7.6454,
      "mean_token_accuracy": 0.107300965487957,
      "num_tokens": 2928291.0,
      "step": 875
    },
    {
      "entropy": 7.669623756408692,
      "epoch": 0.17169056677397326,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004395,
      "loss": 7.7114,
      "mean_token_accuracy": 0.10503170862793923,
      "num_tokens": 2945115.0,
      "step": 880
    },
    {
      "entropy": 7.688317728042603,
      "epoch": 0.1726660813579163,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000442,
      "loss": 7.6097,
      "mean_token_accuracy": 0.10320904403924942,
      "num_tokens": 2962869.0,
      "step": 885
    },
    {
      "entropy": 7.76020359992981,
      "epoch": 0.17364159594185932,
      "grad_norm": 0.875,
      "learning_rate": 0.0004445,
      "loss": 7.7197,
      "mean_token_accuracy": 0.10083551406860351,
      "num_tokens": 2979298.0,
      "step": 890
    },
    {
      "entropy": 7.708668375015259,
      "epoch": 0.17461711052580237,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000447,
      "loss": 7.6131,
      "mean_token_accuracy": 0.11331969946622848,
      "num_tokens": 2996146.0,
      "step": 895
    },
    {
      "entropy": 7.756977510452271,
      "epoch": 0.17559262510974538,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.7027,
      "mean_token_accuracy": 0.10274371355772019,
      "num_tokens": 3013075.0,
      "step": 900
    },
    {
      "entropy": 7.691476202011108,
      "epoch": 0.17656813969368843,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.6613,
      "mean_token_accuracy": 0.10025065019726753,
      "num_tokens": 3029494.0,
      "step": 905
    },
    {
      "entropy": 7.6594343185424805,
      "epoch": 0.17754365427763144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.6761,
      "mean_token_accuracy": 0.1022347241640091,
      "num_tokens": 3046333.0,
      "step": 910
    },
    {
      "entropy": 7.692041826248169,
      "epoch": 0.17851916886157448,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.6515,
      "mean_token_accuracy": 0.09957996159791946,
      "num_tokens": 3062714.0,
      "step": 915
    },
    {
      "entropy": 7.757144641876221,
      "epoch": 0.1794946834455175,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.6434,
      "mean_token_accuracy": 0.10665524601936341,
      "num_tokens": 3080711.0,
      "step": 920
    },
    {
      "entropy": 7.669725751876831,
      "epoch": 0.18047019802946054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000462,
      "loss": 7.7298,
      "mean_token_accuracy": 0.09781264886260033,
      "num_tokens": 3096885.0,
      "step": 925
    },
    {
      "entropy": 7.704789543151856,
      "epoch": 0.18144571261340356,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004645,
      "loss": 7.6261,
      "mean_token_accuracy": 0.10874809324741364,
      "num_tokens": 3114536.0,
      "step": 930
    },
    {
      "entropy": 7.586754035949707,
      "epoch": 0.1824212271973466,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000467,
      "loss": 7.577,
      "mean_token_accuracy": 0.1104214146733284,
      "num_tokens": 3131921.0,
      "step": 935
    },
    {
      "entropy": 7.68097333908081,
      "epoch": 0.18339674178128962,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004695,
      "loss": 7.7225,
      "mean_token_accuracy": 0.09999383538961411,
      "num_tokens": 3148064.0,
      "step": 940
    },
    {
      "entropy": 7.627009773254395,
      "epoch": 0.18437225636523266,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000472,
      "loss": 7.5334,
      "mean_token_accuracy": 0.11227804869413376,
      "num_tokens": 3165822.0,
      "step": 945
    },
    {
      "entropy": 7.600102329254151,
      "epoch": 0.18534777094917568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745,
      "loss": 7.6362,
      "mean_token_accuracy": 0.10948333218693733,
      "num_tokens": 3183286.0,
      "step": 950
    },
    {
      "entropy": 7.680917978286743,
      "epoch": 0.18632328553311872,
      "grad_norm": 1.09375,
      "learning_rate": 0.000477,
      "loss": 7.6665,
      "mean_token_accuracy": 0.09648622497916222,
      "num_tokens": 3198522.0,
      "step": 955
    },
    {
      "entropy": 7.6501914978027346,
      "epoch": 0.18729880011706174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004795,
      "loss": 7.6896,
      "mean_token_accuracy": 0.10254479497671128,
      "num_tokens": 3218422.0,
      "step": 960
    },
    {
      "entropy": 7.663035202026367,
      "epoch": 0.18827431470100478,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000482,
      "loss": 7.7127,
      "mean_token_accuracy": 0.09760246202349662,
      "num_tokens": 3236407.0,
      "step": 965
    },
    {
      "entropy": 7.628622388839721,
      "epoch": 0.18924982928494782,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004845,
      "loss": 7.6424,
      "mean_token_accuracy": 0.10239362493157386,
      "num_tokens": 3252740.0,
      "step": 970
    },
    {
      "entropy": 7.645857572555542,
      "epoch": 0.19022534386889084,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000487,
      "loss": 7.5999,
      "mean_token_accuracy": 0.10109758228063584,
      "num_tokens": 3269874.0,
      "step": 975
    },
    {
      "entropy": 7.596111202239991,
      "epoch": 0.19120085845283388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895,
      "loss": 7.6328,
      "mean_token_accuracy": 0.10229030176997185,
      "num_tokens": 3286515.0,
      "step": 980
    },
    {
      "entropy": 7.522288608551025,
      "epoch": 0.1921763730367769,
      "grad_norm": 1.625,
      "learning_rate": 0.000492,
      "loss": 7.4299,
      "mean_token_accuracy": 0.12200271040201187,
      "num_tokens": 3303700.0,
      "step": 985
    },
    {
      "entropy": 7.578052568435669,
      "epoch": 0.19315188762071994,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945,
      "loss": 7.6204,
      "mean_token_accuracy": 0.10630818232893943,
      "num_tokens": 3319768.0,
      "step": 990
    },
    {
      "entropy": 7.581722211837769,
      "epoch": 0.19412740220466296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000497,
      "loss": 7.5208,
      "mean_token_accuracy": 0.11017290130257607,
      "num_tokens": 3336555.0,
      "step": 995
    },
    {
      "entropy": 7.546011447906494,
      "epoch": 0.195102916788606,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995,
      "loss": 7.5242,
      "mean_token_accuracy": 0.1080077975988388,
      "num_tokens": 3353273.0,
      "step": 1000
    },
    {
      "entropy": 7.5461585521698,
      "epoch": 0.19607843137254902,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999999929644166,
      "loss": 7.4324,
      "mean_token_accuracy": 0.11648088321089745,
      "num_tokens": 3369408.0,
      "step": 1005
    },
    {
      "entropy": 7.518179798126221,
      "epoch": 0.19705394595649206,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999999643823598,
      "loss": 7.4768,
      "mean_token_accuracy": 0.11048399433493614,
      "num_tokens": 3385834.0,
      "step": 1010
    },
    {
      "entropy": 7.5287184715271,
      "epoch": 0.19802946054043508,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999999138141084,
      "loss": 7.5934,
      "mean_token_accuracy": 0.10513608753681183,
      "num_tokens": 3401407.0,
      "step": 1015
    },
    {
      "entropy": 7.622691631317139,
      "epoch": 0.19900497512437812,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999998412596673,
      "loss": 7.5929,
      "mean_token_accuracy": 0.09907566010951996,
      "num_tokens": 3418256.0,
      "step": 1020
    },
    {
      "entropy": 7.530287981033325,
      "epoch": 0.19998048970832114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999997467190435,
      "loss": 7.5624,
      "mean_token_accuracy": 0.10766850039362907,
      "num_tokens": 3433932.0,
      "step": 1025
    },
    {
      "entropy": 7.536963415145874,
      "epoch": 0.20095600429226418,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999996301922465,
      "loss": 7.5978,
      "mean_token_accuracy": 0.09948049932718277,
      "num_tokens": 3451723.0,
      "step": 1030
    },
    {
      "entropy": 7.599093818664551,
      "epoch": 0.2019315188762072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999994916792876,
      "loss": 7.5778,
      "mean_token_accuracy": 0.10314931347966194,
      "num_tokens": 3469170.0,
      "step": 1035
    },
    {
      "entropy": 7.5477721214294435,
      "epoch": 0.20290703346015024,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999993311801802,
      "loss": 7.5969,
      "mean_token_accuracy": 0.10023527815937996,
      "num_tokens": 3485282.0,
      "step": 1040
    },
    {
      "entropy": 7.546366262435913,
      "epoch": 0.20388254804409325,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00049999914869494,
      "loss": 7.5739,
      "mean_token_accuracy": 0.09581654444336891,
      "num_tokens": 3502206.0,
      "step": 1045
    },
    {
      "entropy": 7.614881753921509,
      "epoch": 0.2048580626280363,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499998944223585,
      "loss": 7.5152,
      "mean_token_accuracy": 0.10448609367012977,
      "num_tokens": 3520384.0,
      "step": 1050
    },
    {
      "entropy": 7.466865873336792,
      "epoch": 0.2058335772119793,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999987177661351,
      "loss": 7.5128,
      "mean_token_accuracy": 0.10680526867508888,
      "num_tokens": 3536314.0,
      "step": 1055
    },
    {
      "entropy": 7.491249132156372,
      "epoch": 0.20680909179592236,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999984693226124,
      "loss": 7.4862,
      "mean_token_accuracy": 0.11341608539223672,
      "num_tokens": 3551510.0,
      "step": 1060
    },
    {
      "entropy": 7.450982570648193,
      "epoch": 0.20778460637986537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999981988930411,
      "loss": 7.4799,
      "mean_token_accuracy": 0.10830849185585975,
      "num_tokens": 3568515.0,
      "step": 1065
    },
    {
      "entropy": 7.528032445907593,
      "epoch": 0.20876012096380842,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999979064774478,
      "loss": 7.524,
      "mean_token_accuracy": 0.10810093954205513,
      "num_tokens": 3585183.0,
      "step": 1070
    },
    {
      "entropy": 7.467760324478149,
      "epoch": 0.20973563554775143,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000499997592075861,
      "loss": 7.4664,
      "mean_token_accuracy": 0.11346053704619408,
      "num_tokens": 3601898.0,
      "step": 1075
    },
    {
      "entropy": 7.559082460403443,
      "epoch": 0.21071115013169447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999972556883113,
      "loss": 7.528,
      "mean_token_accuracy": 0.10768525674939156,
      "num_tokens": 3616699.0,
      "step": 1080
    },
    {
      "entropy": 7.52727255821228,
      "epoch": 0.2116866647156375,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999968973148318,
      "loss": 7.5395,
      "mean_token_accuracy": 0.10629705637693405,
      "num_tokens": 3634030.0,
      "step": 1085
    },
    {
      "entropy": 7.566522407531738,
      "epoch": 0.21266217929958053,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999965169554573,
      "loss": 7.5379,
      "mean_token_accuracy": 0.10067117214202881,
      "num_tokens": 3650495.0,
      "step": 1090
    },
    {
      "entropy": 7.514937686920166,
      "epoch": 0.21363769388352355,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999961146102251,
      "loss": 7.4935,
      "mean_token_accuracy": 0.10824285820126534,
      "num_tokens": 3666926.0,
      "step": 1095
    },
    {
      "entropy": 7.427997159957886,
      "epoch": 0.2146132084674666,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004999956902791746,
      "loss": 7.4484,
      "mean_token_accuracy": 0.11637519970536232,
      "num_tokens": 3683061.0,
      "step": 1100
    },
    {
      "entropy": 7.449176025390625,
      "epoch": 0.2155887230514096,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004999952439623469,
      "loss": 7.5252,
      "mean_token_accuracy": 0.10821306332945824,
      "num_tokens": 3703013.0,
      "step": 1105
    },
    {
      "entropy": 7.500231170654297,
      "epoch": 0.21656423763535265,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499994775659786,
      "loss": 7.3686,
      "mean_token_accuracy": 0.1167745478451252,
      "num_tokens": 3719885.0,
      "step": 1110
    },
    {
      "entropy": 7.491282320022583,
      "epoch": 0.21753975221929567,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999942853715375,
      "loss": 7.5083,
      "mean_token_accuracy": 0.10426270142197609,
      "num_tokens": 3735122.0,
      "step": 1115
    },
    {
      "entropy": 7.507296085357666,
      "epoch": 0.2185152668032387,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999937730976494,
      "loss": 7.4958,
      "mean_token_accuracy": 0.10152617171406746,
      "num_tokens": 3751627.0,
      "step": 1120
    },
    {
      "entropy": 7.388977718353272,
      "epoch": 0.21949078138718173,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999932388381716,
      "loss": 7.5081,
      "mean_token_accuracy": 0.10370905101299285,
      "num_tokens": 3768165.0,
      "step": 1125
    },
    {
      "entropy": 7.584489440917968,
      "epoch": 0.22046629597112477,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999926825931564,
      "loss": 7.5383,
      "mean_token_accuracy": 0.10599100366234779,
      "num_tokens": 3784245.0,
      "step": 1130
    },
    {
      "entropy": 7.3387915134429935,
      "epoch": 0.2214418105550678,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999921043626582,
      "loss": 7.2752,
      "mean_token_accuracy": 0.11657643169164658,
      "num_tokens": 3802057.0,
      "step": 1135
    },
    {
      "entropy": 7.452451610565186,
      "epoch": 0.22241732513901083,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999915041467333,
      "loss": 7.475,
      "mean_token_accuracy": 0.1047409564256668,
      "num_tokens": 3817104.0,
      "step": 1140
    },
    {
      "entropy": 7.472776222229004,
      "epoch": 0.22339283972295385,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004999908819454407,
      "loss": 7.4666,
      "mean_token_accuracy": 0.10979759097099304,
      "num_tokens": 3833037.0,
      "step": 1145
    },
    {
      "entropy": 7.396123790740967,
      "epoch": 0.2243683543068969,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999902377588408,
      "loss": 7.3222,
      "mean_token_accuracy": 0.11395804211497307,
      "num_tokens": 3849634.0,
      "step": 1150
    },
    {
      "entropy": 7.454972934722901,
      "epoch": 0.2253438688908399,
      "grad_norm": 0.875,
      "learning_rate": 0.000499989571586997,
      "loss": 7.5012,
      "mean_token_accuracy": 0.1047336332499981,
      "num_tokens": 3865949.0,
      "step": 1155
    },
    {
      "entropy": 7.396062326431275,
      "epoch": 0.22631938347478295,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999888834299739,
      "loss": 7.403,
      "mean_token_accuracy": 0.11621749475598335,
      "num_tokens": 3884646.0,
      "step": 1160
    },
    {
      "entropy": 7.504799795150757,
      "epoch": 0.227294898058726,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999881732878391,
      "loss": 7.478,
      "mean_token_accuracy": 0.10949801355600357,
      "num_tokens": 3900740.0,
      "step": 1165
    },
    {
      "entropy": 7.4035848617553714,
      "epoch": 0.228270412642669,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999874411606618,
      "loss": 7.4124,
      "mean_token_accuracy": 0.11259354799985885,
      "num_tokens": 3918619.0,
      "step": 1170
    },
    {
      "entropy": 7.401635408401489,
      "epoch": 0.22924592722661205,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999866870485137,
      "loss": 7.4045,
      "mean_token_accuracy": 0.10750121995806694,
      "num_tokens": 3938161.0,
      "step": 1175
    },
    {
      "entropy": 7.498355913162231,
      "epoch": 0.23022144181055507,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999859109514684,
      "loss": 7.453,
      "mean_token_accuracy": 0.11029804646968841,
      "num_tokens": 3953582.0,
      "step": 1180
    },
    {
      "entropy": 7.408897161483765,
      "epoch": 0.2311969563944981,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999851128696017,
      "loss": 7.4261,
      "mean_token_accuracy": 0.10325548648834229,
      "num_tokens": 3969747.0,
      "step": 1185
    },
    {
      "entropy": 7.495371007919312,
      "epoch": 0.23217247097844113,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999842928029917,
      "loss": 7.4663,
      "mean_token_accuracy": 0.10589183792471886,
      "num_tokens": 3984646.0,
      "step": 1190
    },
    {
      "entropy": 7.412689352035523,
      "epoch": 0.23314798556238417,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999834507517184,
      "loss": 7.4473,
      "mean_token_accuracy": 0.10245015770196915,
      "num_tokens": 4003108.0,
      "step": 1195
    },
    {
      "entropy": 7.400495195388794,
      "epoch": 0.23412350014632718,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999825867158642,
      "loss": 7.5222,
      "mean_token_accuracy": 0.10239234641194343,
      "num_tokens": 4019944.0,
      "step": 1200
    },
    {
      "entropy": 7.547888946533203,
      "epoch": 0.23509901473027023,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004999817006955134,
      "loss": 7.4009,
      "mean_token_accuracy": 0.1101340301334858,
      "num_tokens": 4037617.0,
      "step": 1205
    },
    {
      "entropy": 7.400909948348999,
      "epoch": 0.23607452931421324,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999807926907527,
      "loss": 7.4013,
      "mean_token_accuracy": 0.10389823541045189,
      "num_tokens": 4053483.0,
      "step": 1210
    },
    {
      "entropy": 7.413779258728027,
      "epoch": 0.2370500438981563,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999798627016707,
      "loss": 7.4679,
      "mean_token_accuracy": 0.11093057245016098,
      "num_tokens": 4070165.0,
      "step": 1215
    },
    {
      "entropy": 7.4767746925354,
      "epoch": 0.2380255584820993,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999789107283586,
      "loss": 7.4501,
      "mean_token_accuracy": 0.10087850391864776,
      "num_tokens": 4086208.0,
      "step": 1220
    },
    {
      "entropy": 7.351885414123535,
      "epoch": 0.23900107306604235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499977936770909,
      "loss": 7.4011,
      "mean_token_accuracy": 0.10884249210357666,
      "num_tokens": 4101790.0,
      "step": 1225
    },
    {
      "entropy": 7.422788381576538,
      "epoch": 0.23997658764998536,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999769408294173,
      "loss": 7.3379,
      "mean_token_accuracy": 0.115897548943758,
      "num_tokens": 4117633.0,
      "step": 1230
    },
    {
      "entropy": 7.380688333511353,
      "epoch": 0.2409521022339284,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999759229039807,
      "loss": 7.4042,
      "mean_token_accuracy": 0.11269243955612182,
      "num_tokens": 4132634.0,
      "step": 1235
    },
    {
      "entropy": 7.398476409912109,
      "epoch": 0.24192761681787142,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999748829946988,
      "loss": 7.4414,
      "mean_token_accuracy": 0.10720136612653733,
      "num_tokens": 4150702.0,
      "step": 1240
    },
    {
      "entropy": 7.424192476272583,
      "epoch": 0.24290313140181446,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999738211016732,
      "loss": 7.3182,
      "mean_token_accuracy": 0.12101202383637429,
      "num_tokens": 4168483.0,
      "step": 1245
    },
    {
      "entropy": 7.216082954406739,
      "epoch": 0.24387864598575748,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999727372250076,
      "loss": 7.2769,
      "mean_token_accuracy": 0.11471722945570946,
      "num_tokens": 4186199.0,
      "step": 1250
    },
    {
      "entropy": 7.420723295211792,
      "epoch": 0.24485416056970052,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999716313648079,
      "loss": 7.4253,
      "mean_token_accuracy": 0.10887003168463708,
      "num_tokens": 4202303.0,
      "step": 1255
    },
    {
      "entropy": 7.443055725097656,
      "epoch": 0.24582967515364354,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999705035211822,
      "loss": 7.4559,
      "mean_token_accuracy": 0.10704999566078185,
      "num_tokens": 4218541.0,
      "step": 1260
    },
    {
      "entropy": 7.378991985321045,
      "epoch": 0.24680518973758658,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999693536942406,
      "loss": 7.425,
      "mean_token_accuracy": 0.10749276131391525,
      "num_tokens": 4234025.0,
      "step": 1265
    },
    {
      "entropy": 7.444829225540161,
      "epoch": 0.2477807043215296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999681818840957,
      "loss": 7.428,
      "mean_token_accuracy": 0.10735326111316681,
      "num_tokens": 4250798.0,
      "step": 1270
    },
    {
      "entropy": 7.360717630386352,
      "epoch": 0.24875621890547264,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499966988090862,
      "loss": 7.4303,
      "mean_token_accuracy": 0.10274397060275078,
      "num_tokens": 4266453.0,
      "step": 1275
    },
    {
      "entropy": 7.5464013576507565,
      "epoch": 0.24973173348941566,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999657723146558,
      "loss": 7.3895,
      "mean_token_accuracy": 0.1079128511250019,
      "num_tokens": 4281825.0,
      "step": 1280
    },
    {
      "entropy": 7.169006109237671,
      "epoch": 0.2507072480733587,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999645345555962,
      "loss": 7.2298,
      "mean_token_accuracy": 0.12292835414409638,
      "num_tokens": 4298821.0,
      "step": 1285
    },
    {
      "entropy": 7.373716640472412,
      "epoch": 0.2516827626573017,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999632748138041,
      "loss": 7.2614,
      "mean_token_accuracy": 0.11685901060700417,
      "num_tokens": 4314076.0,
      "step": 1290
    },
    {
      "entropy": 7.306162357330322,
      "epoch": 0.25265827724124473,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999619930894025,
      "loss": 7.257,
      "mean_token_accuracy": 0.11965253800153733,
      "num_tokens": 4330909.0,
      "step": 1295
    },
    {
      "entropy": 7.387978792190552,
      "epoch": 0.2536337918251878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999606893825168,
      "loss": 7.3761,
      "mean_token_accuracy": 0.11097977533936501,
      "num_tokens": 4346927.0,
      "step": 1300
    },
    {
      "entropy": 7.355130195617676,
      "epoch": 0.2546093064091308,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999593636932742,
      "loss": 7.4053,
      "mean_token_accuracy": 0.10671861544251442,
      "num_tokens": 4364956.0,
      "step": 1305
    },
    {
      "entropy": 7.356976699829102,
      "epoch": 0.25558482099307384,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999580160218044,
      "loss": 7.3358,
      "mean_token_accuracy": 0.11136947125196457,
      "num_tokens": 4383014.0,
      "step": 1310
    },
    {
      "entropy": 7.2942369937896725,
      "epoch": 0.25656033557701685,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999566463682391,
      "loss": 7.4324,
      "mean_token_accuracy": 0.11078611016273499,
      "num_tokens": 4401050.0,
      "step": 1315
    },
    {
      "entropy": 7.312287998199463,
      "epoch": 0.2575358501609599,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999552547327119,
      "loss": 7.3142,
      "mean_token_accuracy": 0.10984900295734405,
      "num_tokens": 4415453.0,
      "step": 1320
    },
    {
      "entropy": 7.337052583694458,
      "epoch": 0.25851136474490294,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999538411153591,
      "loss": 7.2751,
      "mean_token_accuracy": 0.11757619455456733,
      "num_tokens": 4433107.0,
      "step": 1325
    },
    {
      "entropy": 7.333510684967041,
      "epoch": 0.25948687932884595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999524055163187,
      "loss": 7.3852,
      "mean_token_accuracy": 0.10584537237882614,
      "num_tokens": 4449440.0,
      "step": 1330
    },
    {
      "entropy": 7.452435064315796,
      "epoch": 0.26046239391278897,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499950947935731,
      "loss": 7.3637,
      "mean_token_accuracy": 0.10850140005350113,
      "num_tokens": 4465327.0,
      "step": 1335
    },
    {
      "entropy": 7.347502660751343,
      "epoch": 0.26143790849673204,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999494683737384,
      "loss": 7.3954,
      "mean_token_accuracy": 0.10589608028531075,
      "num_tokens": 4481518.0,
      "step": 1340
    },
    {
      "entropy": 7.41256275177002,
      "epoch": 0.26241342308067506,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999479668304854,
      "loss": 7.4012,
      "mean_token_accuracy": 0.10691455230116845,
      "num_tokens": 4497632.0,
      "step": 1345
    },
    {
      "entropy": 7.192644643783569,
      "epoch": 0.2633889376646181,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999464433061188,
      "loss": 7.164,
      "mean_token_accuracy": 0.1220066249370575,
      "num_tokens": 4513790.0,
      "step": 1350
    },
    {
      "entropy": 7.298626232147217,
      "epoch": 0.26436445224856114,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999448978007876,
      "loss": 7.2856,
      "mean_token_accuracy": 0.11468039527535438,
      "num_tokens": 4530802.0,
      "step": 1355
    },
    {
      "entropy": 7.398645114898682,
      "epoch": 0.26533996683250416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999433303146426,
      "loss": 7.3514,
      "mean_token_accuracy": 0.1049342192709446,
      "num_tokens": 4546208.0,
      "step": 1360
    },
    {
      "entropy": 7.333724117279052,
      "epoch": 0.2663154814164472,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004999417408478371,
      "loss": 7.3977,
      "mean_token_accuracy": 0.10778557956218719,
      "num_tokens": 4563753.0,
      "step": 1365
    },
    {
      "entropy": 7.251411771774292,
      "epoch": 0.2672909960003902,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999401294005264,
      "loss": 7.2841,
      "mean_token_accuracy": 0.11472156941890717,
      "num_tokens": 4580274.0,
      "step": 1370
    },
    {
      "entropy": 7.316902112960816,
      "epoch": 0.26826651058433326,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499938495972868,
      "loss": 7.3648,
      "mean_token_accuracy": 0.11064173728227615,
      "num_tokens": 4596085.0,
      "step": 1375
    },
    {
      "entropy": 7.4154500484466555,
      "epoch": 0.2692420251682763,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999368405650214,
      "loss": 7.3904,
      "mean_token_accuracy": 0.10575067326426506,
      "num_tokens": 4613258.0,
      "step": 1380
    },
    {
      "entropy": 7.198368501663208,
      "epoch": 0.2702175397522193,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999351631771485,
      "loss": 7.2203,
      "mean_token_accuracy": 0.116555605083704,
      "num_tokens": 4630128.0,
      "step": 1385
    },
    {
      "entropy": 7.378967189788819,
      "epoch": 0.2711930543361623,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999334638094131,
      "loss": 7.2537,
      "mean_token_accuracy": 0.11374187767505646,
      "num_tokens": 4646181.0,
      "step": 1390
    },
    {
      "entropy": 7.308010864257812,
      "epoch": 0.2721685689201054,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999317424619812,
      "loss": 7.3158,
      "mean_token_accuracy": 0.1069622166454792,
      "num_tokens": 4661804.0,
      "step": 1395
    },
    {
      "entropy": 7.26732611656189,
      "epoch": 0.2731440835040484,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999299991350212,
      "loss": 7.2252,
      "mean_token_accuracy": 0.1148923471570015,
      "num_tokens": 4678163.0,
      "step": 1400
    },
    {
      "entropy": 7.2540120601654055,
      "epoch": 0.2741195980879914,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999282338287034,
      "loss": 7.2136,
      "mean_token_accuracy": 0.11390212774276734,
      "num_tokens": 4695469.0,
      "step": 1405
    },
    {
      "entropy": 7.369423818588257,
      "epoch": 0.2750951126719344,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999264465432001,
      "loss": 7.35,
      "mean_token_accuracy": 0.10951042175292969,
      "num_tokens": 4711599.0,
      "step": 1410
    },
    {
      "entropy": 7.328134298324585,
      "epoch": 0.2760706272558775,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999246372786862,
      "loss": 7.3479,
      "mean_token_accuracy": 0.10849231481552124,
      "num_tokens": 4728381.0,
      "step": 1415
    },
    {
      "entropy": 7.239855241775513,
      "epoch": 0.2770461418398205,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999228060353385,
      "loss": 7.3519,
      "mean_token_accuracy": 0.10967302173376084,
      "num_tokens": 4745348.0,
      "step": 1420
    },
    {
      "entropy": 7.360491466522217,
      "epoch": 0.27802165642376353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999209528133357,
      "loss": 7.3547,
      "mean_token_accuracy": 0.10041134282946587,
      "num_tokens": 4763011.0,
      "step": 1425
    },
    {
      "entropy": 7.366155385971069,
      "epoch": 0.27899717100770655,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499919077612859,
      "loss": 7.2941,
      "mean_token_accuracy": 0.11149064376950264,
      "num_tokens": 4779462.0,
      "step": 1430
    },
    {
      "entropy": 7.137221622467041,
      "epoch": 0.2799726855916496,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999171804340917,
      "loss": 7.2561,
      "mean_token_accuracy": 0.11627396196126938,
      "num_tokens": 4796844.0,
      "step": 1435
    },
    {
      "entropy": 7.344420099258423,
      "epoch": 0.28094820017559263,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999152612772191,
      "loss": 7.2756,
      "mean_token_accuracy": 0.11215969026088715,
      "num_tokens": 4812851.0,
      "step": 1440
    },
    {
      "entropy": 7.175891160964966,
      "epoch": 0.28192371475953565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999133201424288,
      "loss": 7.2121,
      "mean_token_accuracy": 0.11662106513977051,
      "num_tokens": 4830144.0,
      "step": 1445
    },
    {
      "entropy": 7.3987799167633055,
      "epoch": 0.28289922934347866,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999113570299105,
      "loss": 7.3119,
      "mean_token_accuracy": 0.10317229107022285,
      "num_tokens": 4846981.0,
      "step": 1450
    },
    {
      "entropy": 7.359449481964111,
      "epoch": 0.28387474392742174,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004999093719398561,
      "loss": 7.456,
      "mean_token_accuracy": 0.10354385450482369,
      "num_tokens": 4864404.0,
      "step": 1455
    },
    {
      "entropy": 7.347408819198608,
      "epoch": 0.28485025851136475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999073648724592,
      "loss": 7.3405,
      "mean_token_accuracy": 0.10498289689421654,
      "num_tokens": 4882004.0,
      "step": 1460
    },
    {
      "entropy": 7.3070391654968265,
      "epoch": 0.28582577309530777,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999053358279164,
      "loss": 7.281,
      "mean_token_accuracy": 0.10592853873968125,
      "num_tokens": 4899444.0,
      "step": 1465
    },
    {
      "entropy": 7.365157556533814,
      "epoch": 0.2868012876792508,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999032848064256,
      "loss": 7.2762,
      "mean_token_accuracy": 0.10759353488683701,
      "num_tokens": 4917214.0,
      "step": 1470
    },
    {
      "entropy": 7.192684745788574,
      "epoch": 0.28777680226319385,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999012118081875,
      "loss": 7.3657,
      "mean_token_accuracy": 0.1031188927590847,
      "num_tokens": 4934534.0,
      "step": 1475
    },
    {
      "entropy": 7.404789686203003,
      "epoch": 0.28875231684713687,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998991168334045,
      "loss": 7.2727,
      "mean_token_accuracy": 0.10799477323889732,
      "num_tokens": 4950740.0,
      "step": 1480
    },
    {
      "entropy": 7.260487079620361,
      "epoch": 0.2897278314310799,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998969998822814,
      "loss": 7.3667,
      "mean_token_accuracy": 0.1046581894159317,
      "num_tokens": 4966720.0,
      "step": 1485
    },
    {
      "entropy": 7.4018322944641115,
      "epoch": 0.2907033460150229,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998948609550249,
      "loss": 7.2424,
      "mean_token_accuracy": 0.10568741858005523,
      "num_tokens": 4982195.0,
      "step": 1490
    },
    {
      "entropy": 7.231603527069092,
      "epoch": 0.29167886059896597,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998927000518442,
      "loss": 7.3653,
      "mean_token_accuracy": 0.10721067860722541,
      "num_tokens": 4997487.0,
      "step": 1495
    },
    {
      "entropy": 7.307701826095581,
      "epoch": 0.292654375182909,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998905171729505,
      "loss": 7.247,
      "mean_token_accuracy": 0.10762926042079926,
      "num_tokens": 5012913.0,
      "step": 1500
    },
    {
      "entropy": 7.192307758331299,
      "epoch": 0.293629889766852,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998883123185568,
      "loss": 7.223,
      "mean_token_accuracy": 0.11950238272547722,
      "num_tokens": 5031462.0,
      "step": 1505
    },
    {
      "entropy": 7.29890809059143,
      "epoch": 0.294605404350795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998860854888788,
      "loss": 7.2008,
      "mean_token_accuracy": 0.11756352931261063,
      "num_tokens": 5047354.0,
      "step": 1510
    },
    {
      "entropy": 7.216483163833618,
      "epoch": 0.2955809189347381,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004998838366841341,
      "loss": 7.2785,
      "mean_token_accuracy": 0.11354635432362556,
      "num_tokens": 5065268.0,
      "step": 1515
    },
    {
      "entropy": 7.290679788589477,
      "epoch": 0.2965564335186811,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998815659045423,
      "loss": 7.2672,
      "mean_token_accuracy": 0.10739901438355445,
      "num_tokens": 5081066.0,
      "step": 1520
    },
    {
      "entropy": 7.26018328666687,
      "epoch": 0.2975319481026241,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998792731503254,
      "loss": 7.2949,
      "mean_token_accuracy": 0.11383327096700668,
      "num_tokens": 5099440.0,
      "step": 1525
    },
    {
      "entropy": 7.2476519584655765,
      "epoch": 0.29850746268656714,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998769584217074,
      "loss": 7.2505,
      "mean_token_accuracy": 0.11950337737798691,
      "num_tokens": 5116573.0,
      "step": 1530
    },
    {
      "entropy": 7.100352048873901,
      "epoch": 0.2994829772705102,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998746217189145,
      "loss": 7.0171,
      "mean_token_accuracy": 0.1289096012711525,
      "num_tokens": 5132920.0,
      "step": 1535
    },
    {
      "entropy": 7.349333477020264,
      "epoch": 0.3004584918544532,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499872263042175,
      "loss": 7.2819,
      "mean_token_accuracy": 0.10944682136178016,
      "num_tokens": 5150681.0,
      "step": 1540
    },
    {
      "entropy": 7.147453355789184,
      "epoch": 0.30143400643839624,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998698823917196,
      "loss": 7.0497,
      "mean_token_accuracy": 0.12986519634723664,
      "num_tokens": 5167473.0,
      "step": 1545
    },
    {
      "entropy": 7.240125036239624,
      "epoch": 0.3024095210223393,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004998674797677805,
      "loss": 7.3297,
      "mean_token_accuracy": 0.10615943893790245,
      "num_tokens": 5183847.0,
      "step": 1550
    },
    {
      "entropy": 7.296258497238159,
      "epoch": 0.3033850356062823,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998650551705929,
      "loss": 7.2784,
      "mean_token_accuracy": 0.10746836736798286,
      "num_tokens": 5200019.0,
      "step": 1555
    },
    {
      "entropy": 7.196804428100586,
      "epoch": 0.30436055019022534,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998626086003934,
      "loss": 7.1546,
      "mean_token_accuracy": 0.11324814632534981,
      "num_tokens": 5217245.0,
      "step": 1560
    },
    {
      "entropy": 7.238556861877441,
      "epoch": 0.30533606477416836,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004998601400574213,
      "loss": 7.0515,
      "mean_token_accuracy": 0.12279537171125413,
      "num_tokens": 5234822.0,
      "step": 1565
    },
    {
      "entropy": 7.139337778091431,
      "epoch": 0.30631157935811143,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004998576495419177,
      "loss": 7.2222,
      "mean_token_accuracy": 0.10850641578435898,
      "num_tokens": 5251282.0,
      "step": 1570
    },
    {
      "entropy": 7.156428527832031,
      "epoch": 0.30728709394205445,
      "grad_norm": 0.78125,
      "learning_rate": 0.000499855137054126,
      "loss": 7.1658,
      "mean_token_accuracy": 0.11120741963386535,
      "num_tokens": 5269630.0,
      "step": 1575
    },
    {
      "entropy": 7.260046768188476,
      "epoch": 0.30826260852599746,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998526025942917,
      "loss": 7.2444,
      "mean_token_accuracy": 0.1078572191298008,
      "num_tokens": 5287106.0,
      "step": 1580
    },
    {
      "entropy": 7.320893049240112,
      "epoch": 0.3092381231099405,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998500461626625,
      "loss": 7.2631,
      "mean_token_accuracy": 0.1063196413218975,
      "num_tokens": 5303387.0,
      "step": 1585
    },
    {
      "entropy": 7.289645624160767,
      "epoch": 0.31021363769388355,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000499847467759488,
      "loss": 7.2598,
      "mean_token_accuracy": 0.10823246538639068,
      "num_tokens": 5320482.0,
      "step": 1590
    },
    {
      "entropy": 7.193205499649048,
      "epoch": 0.31118915227782656,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998448673850205,
      "loss": 7.2105,
      "mean_token_accuracy": 0.11387670114636421,
      "num_tokens": 5336420.0,
      "step": 1595
    },
    {
      "entropy": 7.176516199111939,
      "epoch": 0.3121646668617696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998422450395139,
      "loss": 7.1626,
      "mean_token_accuracy": 0.10995283350348473,
      "num_tokens": 5353464.0,
      "step": 1600
    },
    {
      "entropy": 7.260991001129151,
      "epoch": 0.3131401814457126,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998396007232244,
      "loss": 7.2239,
      "mean_token_accuracy": 0.11493458598852158,
      "num_tokens": 5372607.0,
      "step": 1605
    },
    {
      "entropy": 7.189897060394287,
      "epoch": 0.31411569602965567,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998369344364105,
      "loss": 7.2321,
      "mean_token_accuracy": 0.11545405685901641,
      "num_tokens": 5390134.0,
      "step": 1610
    },
    {
      "entropy": 7.215340948104858,
      "epoch": 0.3150912106135987,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998342461793328,
      "loss": 7.1547,
      "mean_token_accuracy": 0.11492555439472199,
      "num_tokens": 5407083.0,
      "step": 1615
    },
    {
      "entropy": 7.0581450939178465,
      "epoch": 0.3160667251975417,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998315359522538,
      "loss": 7.0481,
      "mean_token_accuracy": 0.12926709651947021,
      "num_tokens": 5424617.0,
      "step": 1620
    },
    {
      "entropy": 7.196384525299072,
      "epoch": 0.3170422397814847,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998288037554383,
      "loss": 7.1915,
      "mean_token_accuracy": 0.11906242743134499,
      "num_tokens": 5440913.0,
      "step": 1625
    },
    {
      "entropy": 7.212459516525269,
      "epoch": 0.3180177543654278,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998260495891535,
      "loss": 7.2101,
      "mean_token_accuracy": 0.1167129822075367,
      "num_tokens": 5457157.0,
      "step": 1630
    },
    {
      "entropy": 7.159144496917724,
      "epoch": 0.3189932689493708,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004998232734536685,
      "loss": 7.1375,
      "mean_token_accuracy": 0.11458515524864196,
      "num_tokens": 5474520.0,
      "step": 1635
    },
    {
      "entropy": 7.263358926773071,
      "epoch": 0.3199687835333138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998204753492544,
      "loss": 7.2081,
      "mean_token_accuracy": 0.11552096009254456,
      "num_tokens": 5490599.0,
      "step": 1640
    },
    {
      "entropy": 7.2186919212341305,
      "epoch": 0.32094429811725683,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998176552761848,
      "loss": 7.28,
      "mean_token_accuracy": 0.10384325012564659,
      "num_tokens": 5506714.0,
      "step": 1645
    },
    {
      "entropy": 7.2394510269165036,
      "epoch": 0.3219198127011999,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998148132347352,
      "loss": 7.2283,
      "mean_token_accuracy": 0.10727055072784424,
      "num_tokens": 5522894.0,
      "step": 1650
    },
    {
      "entropy": 7.196140766143799,
      "epoch": 0.3228953272851429,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998119492251832,
      "loss": 7.1119,
      "mean_token_accuracy": 0.11906631961464882,
      "num_tokens": 5540054.0,
      "step": 1655
    },
    {
      "entropy": 7.2405791759490965,
      "epoch": 0.32387084186908593,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998090632478088,
      "loss": 7.2668,
      "mean_token_accuracy": 0.10199896171689034,
      "num_tokens": 5556189.0,
      "step": 1660
    },
    {
      "entropy": 7.2902802467346195,
      "epoch": 0.32484635645302895,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499806155302894,
      "loss": 7.2854,
      "mean_token_accuracy": 0.10600126460194588,
      "num_tokens": 5572562.0,
      "step": 1665
    },
    {
      "entropy": 7.28464298248291,
      "epoch": 0.325821871036972,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499803225390723,
      "loss": 7.2758,
      "mean_token_accuracy": 0.1067048043012619,
      "num_tokens": 5588992.0,
      "step": 1670
    },
    {
      "entropy": 7.256864261627197,
      "epoch": 0.32679738562091504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998002735115819,
      "loss": 7.19,
      "mean_token_accuracy": 0.11200586631894112,
      "num_tokens": 5604143.0,
      "step": 1675
    },
    {
      "entropy": 7.225049877166748,
      "epoch": 0.32777290020485805,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997972996657593,
      "loss": 7.301,
      "mean_token_accuracy": 0.10044858381152152,
      "num_tokens": 5620887.0,
      "step": 1680
    },
    {
      "entropy": 7.232311201095581,
      "epoch": 0.32874841478880107,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997943038535459,
      "loss": 7.2432,
      "mean_token_accuracy": 0.11001040786504745,
      "num_tokens": 5637156.0,
      "step": 1685
    },
    {
      "entropy": 7.249419164657593,
      "epoch": 0.32972392937274414,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004997912860752341,
      "loss": 7.1909,
      "mean_token_accuracy": 0.10798757523298264,
      "num_tokens": 5654024.0,
      "step": 1690
    },
    {
      "entropy": 7.1812872886657715,
      "epoch": 0.33069944395668716,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997882463311192,
      "loss": 7.1462,
      "mean_token_accuracy": 0.11451933011412621,
      "num_tokens": 5671538.0,
      "step": 1695
    },
    {
      "entropy": 7.190932369232177,
      "epoch": 0.33167495854063017,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997851846214979,
      "loss": 7.1927,
      "mean_token_accuracy": 0.11635877266526222,
      "num_tokens": 5688788.0,
      "step": 1700
    },
    {
      "entropy": 7.17593240737915,
      "epoch": 0.3326504731245732,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004997821009466695,
      "loss": 7.125,
      "mean_token_accuracy": 0.11613107174634933,
      "num_tokens": 5706066.0,
      "step": 1705
    },
    {
      "entropy": 7.102512502670288,
      "epoch": 0.33362598770851626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997789953069354,
      "loss": 7.1113,
      "mean_token_accuracy": 0.11301351860165595,
      "num_tokens": 5722100.0,
      "step": 1710
    },
    {
      "entropy": 7.258580112457276,
      "epoch": 0.3346015022924593,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004997758677025989,
      "loss": 7.2387,
      "mean_token_accuracy": 0.10814922899007798,
      "num_tokens": 5739068.0,
      "step": 1715
    },
    {
      "entropy": 7.199310684204102,
      "epoch": 0.3355770168764023,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997727181339658,
      "loss": 7.2005,
      "mean_token_accuracy": 0.10906586721539498,
      "num_tokens": 5756147.0,
      "step": 1720
    },
    {
      "entropy": 7.21035852432251,
      "epoch": 0.3365525314603453,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997695466013438,
      "loss": 7.1651,
      "mean_token_accuracy": 0.1126245193183422,
      "num_tokens": 5772157.0,
      "step": 1725
    },
    {
      "entropy": 7.2826355457305905,
      "epoch": 0.3375280460442884,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997663531050427,
      "loss": 7.2089,
      "mean_token_accuracy": 0.10928624793887139,
      "num_tokens": 5788128.0,
      "step": 1730
    },
    {
      "entropy": 7.09017162322998,
      "epoch": 0.3385035606282314,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997631376453749,
      "loss": 7.186,
      "mean_token_accuracy": 0.1129564292728901,
      "num_tokens": 5804793.0,
      "step": 1735
    },
    {
      "entropy": 7.164242076873779,
      "epoch": 0.3394790752121744,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997599002226541,
      "loss": 7.1177,
      "mean_token_accuracy": 0.11402822509407998,
      "num_tokens": 5822409.0,
      "step": 1740
    },
    {
      "entropy": 7.144210338592529,
      "epoch": 0.3404545897961175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499756640837197,
      "loss": 7.1508,
      "mean_token_accuracy": 0.10974019914865493,
      "num_tokens": 5839023.0,
      "step": 1745
    },
    {
      "entropy": 7.194488573074341,
      "epoch": 0.3414301043800605,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499753359489322,
      "loss": 7.165,
      "mean_token_accuracy": 0.10810371488332748,
      "num_tokens": 5855804.0,
      "step": 1750
    },
    {
      "entropy": 7.167374801635742,
      "epoch": 0.3424056189640035,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997500561793498,
      "loss": 7.18,
      "mean_token_accuracy": 0.11254183501005173,
      "num_tokens": 5872372.0,
      "step": 1755
    },
    {
      "entropy": 7.24317741394043,
      "epoch": 0.3433811335479465,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997467309076031,
      "loss": 7.2269,
      "mean_token_accuracy": 0.1065982572734356,
      "num_tokens": 5888666.0,
      "step": 1760
    },
    {
      "entropy": 7.082931280136108,
      "epoch": 0.3443566481318896,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997433836744069,
      "loss": 7.1613,
      "mean_token_accuracy": 0.11193820387125016,
      "num_tokens": 5905577.0,
      "step": 1765
    },
    {
      "entropy": 7.1858104228973385,
      "epoch": 0.3453321627158326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997400144800882,
      "loss": 7.1613,
      "mean_token_accuracy": 0.11466953307390212,
      "num_tokens": 5922570.0,
      "step": 1770
    },
    {
      "entropy": 7.035737466812134,
      "epoch": 0.34630767729977563,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997366233249762,
      "loss": 7.0891,
      "mean_token_accuracy": 0.12224273383617401,
      "num_tokens": 5938995.0,
      "step": 1775
    },
    {
      "entropy": 7.134057426452637,
      "epoch": 0.34728319188371864,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997332102094024,
      "loss": 7.102,
      "mean_token_accuracy": 0.11878659799695016,
      "num_tokens": 5954869.0,
      "step": 1780
    },
    {
      "entropy": 7.30258355140686,
      "epoch": 0.3482587064676617,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997297751337004,
      "loss": 7.2144,
      "mean_token_accuracy": 0.1139085553586483,
      "num_tokens": 5972750.0,
      "step": 1785
    },
    {
      "entropy": 7.116592693328857,
      "epoch": 0.34923422105160473,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997263180982055,
      "loss": 7.1292,
      "mean_token_accuracy": 0.11655831038951874,
      "num_tokens": 5989004.0,
      "step": 1790
    },
    {
      "entropy": 7.091948843002319,
      "epoch": 0.35020973563554775,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997228391032559,
      "loss": 7.1509,
      "mean_token_accuracy": 0.11381514072418213,
      "num_tokens": 6004749.0,
      "step": 1795
    },
    {
      "entropy": 7.1248828887939455,
      "epoch": 0.35118525021949076,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997193381491913,
      "loss": 7.0421,
      "mean_token_accuracy": 0.11761141419410706,
      "num_tokens": 6022235.0,
      "step": 1800
    },
    {
      "entropy": 7.1107800006866455,
      "epoch": 0.35216076480343383,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499715815236354,
      "loss": 7.1905,
      "mean_token_accuracy": 0.10815553218126298,
      "num_tokens": 6039194.0,
      "step": 1805
    },
    {
      "entropy": 7.217487049102783,
      "epoch": 0.35313627938737685,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000499712270365088,
      "loss": 7.0662,
      "mean_token_accuracy": 0.11293599009513855,
      "num_tokens": 6055376.0,
      "step": 1810
    },
    {
      "entropy": 7.159701347351074,
      "epoch": 0.35411179397131987,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997087035357398,
      "loss": 7.1209,
      "mean_token_accuracy": 0.11384262070059777,
      "num_tokens": 6071105.0,
      "step": 1815
    },
    {
      "entropy": 7.057770109176635,
      "epoch": 0.3550873085552629,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997051147486582,
      "loss": 7.0503,
      "mean_token_accuracy": 0.11704391539096833,
      "num_tokens": 6087964.0,
      "step": 1820
    },
    {
      "entropy": 7.12920560836792,
      "epoch": 0.35606282313920595,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004997015040041935,
      "loss": 7.1108,
      "mean_token_accuracy": 0.1092826284468174,
      "num_tokens": 6105245.0,
      "step": 1825
    },
    {
      "entropy": 7.146176195144653,
      "epoch": 0.35703833772314897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996978713026986,
      "loss": 7.082,
      "mean_token_accuracy": 0.12028156518936158,
      "num_tokens": 6121736.0,
      "step": 1830
    },
    {
      "entropy": 7.127839231491089,
      "epoch": 0.358013852307092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996942166445287,
      "loss": 7.13,
      "mean_token_accuracy": 0.11404056623578071,
      "num_tokens": 6137751.0,
      "step": 1835
    },
    {
      "entropy": 7.152217197418213,
      "epoch": 0.358989366891035,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996905400300407,
      "loss": 7.1837,
      "mean_token_accuracy": 0.11267100870609284,
      "num_tokens": 6154655.0,
      "step": 1840
    },
    {
      "entropy": 7.18564395904541,
      "epoch": 0.35996488147497807,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499686841459594,
      "loss": 7.1864,
      "mean_token_accuracy": 0.10949580073356628,
      "num_tokens": 6172230.0,
      "step": 1845
    },
    {
      "entropy": 7.117382955551148,
      "epoch": 0.3609403960589211,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996831209335497,
      "loss": 7.1217,
      "mean_token_accuracy": 0.10805677026510238,
      "num_tokens": 6188520.0,
      "step": 1850
    },
    {
      "entropy": 7.1415510177612305,
      "epoch": 0.3619159106428641,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996793784522719,
      "loss": 7.1543,
      "mean_token_accuracy": 0.10972608849406243,
      "num_tokens": 6205555.0,
      "step": 1855
    },
    {
      "entropy": 7.2085896015167235,
      "epoch": 0.3628914252268071,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996756140161259,
      "loss": 7.141,
      "mean_token_accuracy": 0.11507897600531578,
      "num_tokens": 6223548.0,
      "step": 1860
    },
    {
      "entropy": 7.129433488845825,
      "epoch": 0.3638669398107502,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996718276254797,
      "loss": 7.1486,
      "mean_token_accuracy": 0.10784759968519211,
      "num_tokens": 6239266.0,
      "step": 1865
    },
    {
      "entropy": 7.1826282978057865,
      "epoch": 0.3648424543946932,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996680192807032,
      "loss": 7.1199,
      "mean_token_accuracy": 0.11224561482667923,
      "num_tokens": 6255020.0,
      "step": 1870
    },
    {
      "entropy": 7.139209032058716,
      "epoch": 0.3658179689786362,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996641889821686,
      "loss": 7.053,
      "mean_token_accuracy": 0.11049313396215439,
      "num_tokens": 6272774.0,
      "step": 1875
    },
    {
      "entropy": 7.10287446975708,
      "epoch": 0.36679348356257924,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996603367302502,
      "loss": 7.0757,
      "mean_token_accuracy": 0.11213366091251373,
      "num_tokens": 6290478.0,
      "step": 1880
    },
    {
      "entropy": 7.12583212852478,
      "epoch": 0.3677689981465223,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004996564625253244,
      "loss": 7.1468,
      "mean_token_accuracy": 0.11182805076241494,
      "num_tokens": 6308003.0,
      "step": 1885
    },
    {
      "entropy": 7.0979126453399655,
      "epoch": 0.3687445127304653,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996525663677697,
      "loss": 7.1043,
      "mean_token_accuracy": 0.10848119333386422,
      "num_tokens": 6325748.0,
      "step": 1890
    },
    {
      "entropy": 7.112027978897094,
      "epoch": 0.36972002731440834,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499648648257967,
      "loss": 7.0532,
      "mean_token_accuracy": 0.117766385525465,
      "num_tokens": 6344031.0,
      "step": 1895
    },
    {
      "entropy": 7.047448968887329,
      "epoch": 0.37069554189835136,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499644708196299,
      "loss": 7.1157,
      "mean_token_accuracy": 0.11333042085170746,
      "num_tokens": 6359821.0,
      "step": 1900
    },
    {
      "entropy": 7.05104513168335,
      "epoch": 0.3716710564822944,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996407461831507,
      "loss": 6.9577,
      "mean_token_accuracy": 0.12579699233174324,
      "num_tokens": 6376464.0,
      "step": 1905
    },
    {
      "entropy": 7.115139818191528,
      "epoch": 0.37264657106623744,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996367622189094,
      "loss": 7.1305,
      "mean_token_accuracy": 0.10904230326414108,
      "num_tokens": 6392415.0,
      "step": 1910
    },
    {
      "entropy": 7.063139200210571,
      "epoch": 0.37362208565018046,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996327563039643,
      "loss": 7.1325,
      "mean_token_accuracy": 0.1099869281053543,
      "num_tokens": 6409151.0,
      "step": 1915
    },
    {
      "entropy": 7.200403642654419,
      "epoch": 0.3745976002341235,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004996287284387068,
      "loss": 7.0745,
      "mean_token_accuracy": 0.11576901897788047,
      "num_tokens": 6426997.0,
      "step": 1920
    },
    {
      "entropy": 7.096438884735107,
      "epoch": 0.37557311481806654,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996246786235306,
      "loss": 7.0819,
      "mean_token_accuracy": 0.11352694556117057,
      "num_tokens": 6442871.0,
      "step": 1925
    },
    {
      "entropy": 7.026518487930298,
      "epoch": 0.37654862940200956,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996206068588314,
      "loss": 7.0813,
      "mean_token_accuracy": 0.1129517525434494,
      "num_tokens": 6459677.0,
      "step": 1930
    },
    {
      "entropy": 7.2005476474761965,
      "epoch": 0.3775241439859526,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996165131450071,
      "loss": 7.2063,
      "mean_token_accuracy": 0.10632368177175522,
      "num_tokens": 6477239.0,
      "step": 1935
    },
    {
      "entropy": 7.1526548862457275,
      "epoch": 0.37849965856989565,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996123974824576,
      "loss": 7.0799,
      "mean_token_accuracy": 0.11786233857274056,
      "num_tokens": 6493277.0,
      "step": 1940
    },
    {
      "entropy": 7.055672645568848,
      "epoch": 0.37947517315383866,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996082598715852,
      "loss": 7.0894,
      "mean_token_accuracy": 0.11412486806511879,
      "num_tokens": 6509204.0,
      "step": 1945
    },
    {
      "entropy": 7.012831974029541,
      "epoch": 0.3804506877377817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499604100312794,
      "loss": 6.9451,
      "mean_token_accuracy": 0.12578009217977523,
      "num_tokens": 6525486.0,
      "step": 1950
    },
    {
      "entropy": 7.067043399810791,
      "epoch": 0.3814262023217247,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004995999188064908,
      "loss": 7.1636,
      "mean_token_accuracy": 0.10759316012263298,
      "num_tokens": 6542227.0,
      "step": 1955
    },
    {
      "entropy": 7.145499515533447,
      "epoch": 0.38240171690566777,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499595715353084,
      "loss": 7.2149,
      "mean_token_accuracy": 0.10373876541852951,
      "num_tokens": 6558951.0,
      "step": 1960
    },
    {
      "entropy": 7.20166802406311,
      "epoch": 0.3833772314896108,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995914899529844,
      "loss": 7.1288,
      "mean_token_accuracy": 0.11313018873333931,
      "num_tokens": 6574696.0,
      "step": 1965
    },
    {
      "entropy": 7.211101913452149,
      "epoch": 0.3843527460735538,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995872426066048,
      "loss": 7.1401,
      "mean_token_accuracy": 0.10959764048457146,
      "num_tokens": 6591628.0,
      "step": 1970
    },
    {
      "entropy": 7.087547159194946,
      "epoch": 0.3853282606574968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995829733143602,
      "loss": 7.0457,
      "mean_token_accuracy": 0.11506836861371994,
      "num_tokens": 6608107.0,
      "step": 1975
    },
    {
      "entropy": 7.08004412651062,
      "epoch": 0.3863037752414399,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499578682076668,
      "loss": 7.1628,
      "mean_token_accuracy": 0.11517509818077087,
      "num_tokens": 6626699.0,
      "step": 1980
    },
    {
      "entropy": 7.229003095626831,
      "epoch": 0.3872792898253829,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995743688939474,
      "loss": 7.1814,
      "mean_token_accuracy": 0.10261408910155297,
      "num_tokens": 6644540.0,
      "step": 1985
    },
    {
      "entropy": 7.120494174957275,
      "epoch": 0.3882548044093259,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004995700337666199,
      "loss": 7.1216,
      "mean_token_accuracy": 0.10787840634584427,
      "num_tokens": 6660653.0,
      "step": 1990
    },
    {
      "entropy": 7.160486888885498,
      "epoch": 0.38923031899326893,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499565676695109,
      "loss": 7.1258,
      "mean_token_accuracy": 0.10839822441339493,
      "num_tokens": 6677553.0,
      "step": 1995
    },
    {
      "entropy": 7.107635259628296,
      "epoch": 0.390205833577212,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995612976798406,
      "loss": 7.1104,
      "mean_token_accuracy": 0.10860190391540528,
      "num_tokens": 6695274.0,
      "step": 2000
    },
    {
      "entropy": 7.074828433990478,
      "epoch": 0.391181348161155,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995568967212425,
      "loss": 7.0548,
      "mean_token_accuracy": 0.11277033910155296,
      "num_tokens": 6711166.0,
      "step": 2005
    },
    {
      "entropy": 7.0729128360748295,
      "epoch": 0.39215686274509803,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995524738197448,
      "loss": 7.0148,
      "mean_token_accuracy": 0.11710212528705596,
      "num_tokens": 6728827.0,
      "step": 2010
    },
    {
      "entropy": 6.975606203079224,
      "epoch": 0.39313237732904105,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995480289757796,
      "loss": 7.0034,
      "mean_token_accuracy": 0.12345981225371361,
      "num_tokens": 6745901.0,
      "step": 2015
    },
    {
      "entropy": 7.042986583709717,
      "epoch": 0.3941078919129841,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004995435621897815,
      "loss": 6.9828,
      "mean_token_accuracy": 0.120838313549757,
      "num_tokens": 6762176.0,
      "step": 2020
    },
    {
      "entropy": 7.105779600143433,
      "epoch": 0.39508340649692714,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995390734621867,
      "loss": 7.1412,
      "mean_token_accuracy": 0.10852285772562027,
      "num_tokens": 6779322.0,
      "step": 2025
    },
    {
      "entropy": 7.158405876159668,
      "epoch": 0.39605892108087015,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995345627934338,
      "loss": 7.0704,
      "mean_token_accuracy": 0.11450749263167381,
      "num_tokens": 6794293.0,
      "step": 2030
    },
    {
      "entropy": 6.862678384780883,
      "epoch": 0.39703443566481317,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995300301839637,
      "loss": 6.8987,
      "mean_token_accuracy": 0.13570844605565072,
      "num_tokens": 6812880.0,
      "step": 2035
    },
    {
      "entropy": 7.121017122268677,
      "epoch": 0.39800995024875624,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995254756342194,
      "loss": 7.1007,
      "mean_token_accuracy": 0.11062633246183395,
      "num_tokens": 6829230.0,
      "step": 2040
    },
    {
      "entropy": 7.110508584976197,
      "epoch": 0.39898546483269925,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995208991446458,
      "loss": 7.0967,
      "mean_token_accuracy": 0.11192586347460746,
      "num_tokens": 6846787.0,
      "step": 2045
    },
    {
      "entropy": 7.028483963012695,
      "epoch": 0.39996097941664227,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995163007156902,
      "loss": 6.9704,
      "mean_token_accuracy": 0.11334318965673447,
      "num_tokens": 6862990.0,
      "step": 2050
    },
    {
      "entropy": 7.138707399368286,
      "epoch": 0.4009364940005853,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995116803478018,
      "loss": 7.1291,
      "mean_token_accuracy": 0.10785363912582398,
      "num_tokens": 6879170.0,
      "step": 2055
    },
    {
      "entropy": 7.040028429031372,
      "epoch": 0.40191200858452836,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995070380414321,
      "loss": 7.1042,
      "mean_token_accuracy": 0.10964220613241196,
      "num_tokens": 6895635.0,
      "step": 2060
    },
    {
      "entropy": 7.082418775558471,
      "epoch": 0.4028875231684714,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499502373797035,
      "loss": 7.0193,
      "mean_token_accuracy": 0.11688834205269813,
      "num_tokens": 6912836.0,
      "step": 2065
    },
    {
      "entropy": 7.042549419403076,
      "epoch": 0.4038630377524144,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499497687615066,
      "loss": 7.0527,
      "mean_token_accuracy": 0.11874325275421142,
      "num_tokens": 6929507.0,
      "step": 2070
    },
    {
      "entropy": 7.097054290771484,
      "epoch": 0.4048385523363574,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499492979495983,
      "loss": 7.0704,
      "mean_token_accuracy": 0.11026768907904624,
      "num_tokens": 6946527.0,
      "step": 2075
    },
    {
      "entropy": 7.113293838500977,
      "epoch": 0.4058140669203005,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994882494402462,
      "loss": 7.0686,
      "mean_token_accuracy": 0.11713291183114052,
      "num_tokens": 6962582.0,
      "step": 2080
    },
    {
      "entropy": 7.083294153213501,
      "epoch": 0.4067895815042435,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994834974483178,
      "loss": 7.0635,
      "mean_token_accuracy": 0.11450244709849358,
      "num_tokens": 6979389.0,
      "step": 2085
    },
    {
      "entropy": 6.981550216674805,
      "epoch": 0.4077650960881865,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004994787235206621,
      "loss": 7.041,
      "mean_token_accuracy": 0.11269214749336243,
      "num_tokens": 6996637.0,
      "step": 2090
    },
    {
      "entropy": 7.18500280380249,
      "epoch": 0.4087406106721295,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004994739276577456,
      "loss": 7.1306,
      "mean_token_accuracy": 0.10788496881723404,
      "num_tokens": 7013013.0,
      "step": 2095
    },
    {
      "entropy": 6.985776424407959,
      "epoch": 0.4097161252560726,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004994691098600369,
      "loss": 7.0873,
      "mean_token_accuracy": 0.11788748800754548,
      "num_tokens": 7031476.0,
      "step": 2100
    },
    {
      "entropy": 7.15326042175293,
      "epoch": 0.4106916398400156,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994642701280068,
      "loss": 7.0002,
      "mean_token_accuracy": 0.11540462523698806,
      "num_tokens": 7047424.0,
      "step": 2105
    },
    {
      "entropy": 7.044336938858033,
      "epoch": 0.4116671544239586,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004994594084621282,
      "loss": 7.0826,
      "mean_token_accuracy": 0.1102508321404457,
      "num_tokens": 7063793.0,
      "step": 2110
    },
    {
      "entropy": 7.192307329177856,
      "epoch": 0.41264266900790164,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994545248628762,
      "loss": 7.1316,
      "mean_token_accuracy": 0.10800800696015359,
      "num_tokens": 7079026.0,
      "step": 2115
    },
    {
      "entropy": 7.062437677383423,
      "epoch": 0.4136181835918447,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499449619330728,
      "loss": 7.0481,
      "mean_token_accuracy": 0.11414429694414138,
      "num_tokens": 7097524.0,
      "step": 2120
    },
    {
      "entropy": 7.104547214508057,
      "epoch": 0.41459369817578773,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499444691866163,
      "loss": 7.1283,
      "mean_token_accuracy": 0.11073413416743279,
      "num_tokens": 7115514.0,
      "step": 2125
    },
    {
      "entropy": 7.113216829299927,
      "epoch": 0.41556921275973074,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994397424696625,
      "loss": 7.0699,
      "mean_token_accuracy": 0.11289520859718323,
      "num_tokens": 7132589.0,
      "step": 2130
    },
    {
      "entropy": 7.062982511520386,
      "epoch": 0.4165447273436738,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994347711417105,
      "loss": 7.0817,
      "mean_token_accuracy": 0.1103174589574337,
      "num_tokens": 7149416.0,
      "step": 2135
    },
    {
      "entropy": 7.003421640396118,
      "epoch": 0.41752024192761683,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994297778827924,
      "loss": 7.0104,
      "mean_token_accuracy": 0.11797065064311027,
      "num_tokens": 7165763.0,
      "step": 2140
    },
    {
      "entropy": 7.055677080154419,
      "epoch": 0.41849575651155985,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994247626933964,
      "loss": 6.9991,
      "mean_token_accuracy": 0.12271674796938896,
      "num_tokens": 7183131.0,
      "step": 2145
    },
    {
      "entropy": 6.981582450866699,
      "epoch": 0.41947127109550286,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994197255740124,
      "loss": 7.0142,
      "mean_token_accuracy": 0.12010570019483566,
      "num_tokens": 7200467.0,
      "step": 2150
    },
    {
      "entropy": 7.1045958518981935,
      "epoch": 0.42044678567944593,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994146665251327,
      "loss": 7.0218,
      "mean_token_accuracy": 0.1165754571557045,
      "num_tokens": 7217311.0,
      "step": 2155
    },
    {
      "entropy": 6.998889541625976,
      "epoch": 0.42142230026338895,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004994095855472516,
      "loss": 7.0347,
      "mean_token_accuracy": 0.1161852665245533,
      "num_tokens": 7235244.0,
      "step": 2160
    },
    {
      "entropy": 7.119374418258667,
      "epoch": 0.42239781484733196,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004994044826408656,
      "loss": 7.065,
      "mean_token_accuracy": 0.11025517135858535,
      "num_tokens": 7252235.0,
      "step": 2165
    },
    {
      "entropy": 7.022988224029541,
      "epoch": 0.423373329431275,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993993578064733,
      "loss": 7.0747,
      "mean_token_accuracy": 0.110476765781641,
      "num_tokens": 7269367.0,
      "step": 2170
    },
    {
      "entropy": 7.157013368606568,
      "epoch": 0.42434884401521805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993942110445756,
      "loss": 7.0479,
      "mean_token_accuracy": 0.11119917035102844,
      "num_tokens": 7287198.0,
      "step": 2175
    },
    {
      "entropy": 7.0396891117095945,
      "epoch": 0.42532435859916107,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993890423556753,
      "loss": 6.9786,
      "mean_token_accuracy": 0.11426467075943947,
      "num_tokens": 7302356.0,
      "step": 2180
    },
    {
      "entropy": 7.056444549560547,
      "epoch": 0.4262998731831041,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993838517402775,
      "loss": 6.9778,
      "mean_token_accuracy": 0.12020173147320748,
      "num_tokens": 7318530.0,
      "step": 2185
    },
    {
      "entropy": 7.040367078781128,
      "epoch": 0.4272753877670471,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004993786391988894,
      "loss": 6.9622,
      "mean_token_accuracy": 0.11859990432858467,
      "num_tokens": 7335195.0,
      "step": 2190
    },
    {
      "entropy": 6.926948690414429,
      "epoch": 0.42825090235099017,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993734047320206,
      "loss": 6.9732,
      "mean_token_accuracy": 0.11311818957328797,
      "num_tokens": 7352233.0,
      "step": 2195
    },
    {
      "entropy": 7.1166057109832765,
      "epoch": 0.4292264169349332,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993681483401821,
      "loss": 7.0317,
      "mean_token_accuracy": 0.11632806286215783,
      "num_tokens": 7368569.0,
      "step": 2200
    },
    {
      "entropy": 6.972989654541015,
      "epoch": 0.4302019315188762,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993628700238879,
      "loss": 7.0812,
      "mean_token_accuracy": 0.1123574584722519,
      "num_tokens": 7384669.0,
      "step": 2205
    },
    {
      "entropy": 7.17704405784607,
      "epoch": 0.4311774461028192,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993575697836537,
      "loss": 7.084,
      "mean_token_accuracy": 0.10872181057929993,
      "num_tokens": 7401833.0,
      "step": 2210
    },
    {
      "entropy": 7.023128604888916,
      "epoch": 0.4321529606867623,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993522476199975,
      "loss": 6.9775,
      "mean_token_accuracy": 0.11380465403199196,
      "num_tokens": 7418663.0,
      "step": 2215
    },
    {
      "entropy": 7.082082462310791,
      "epoch": 0.4331284752707053,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004993469035334392,
      "loss": 6.9614,
      "mean_token_accuracy": 0.12375471666455269,
      "num_tokens": 7434701.0,
      "step": 2220
    },
    {
      "entropy": 6.968666362762451,
      "epoch": 0.4341039898546483,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499341537524501,
      "loss": 7.0388,
      "mean_token_accuracy": 0.10865203589200974,
      "num_tokens": 7450154.0,
      "step": 2225
    },
    {
      "entropy": 6.962765455245972,
      "epoch": 0.43507950443859134,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993361495937074,
      "loss": 6.8572,
      "mean_token_accuracy": 0.13519329726696014,
      "num_tokens": 7468110.0,
      "step": 2230
    },
    {
      "entropy": 7.0003479480743405,
      "epoch": 0.4360550190225344,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993307397415848,
      "loss": 6.9826,
      "mean_token_accuracy": 0.11327130421996116,
      "num_tokens": 7482962.0,
      "step": 2235
    },
    {
      "entropy": 7.152976131439209,
      "epoch": 0.4370305336064774,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004993253079686619,
      "loss": 7.1131,
      "mean_token_accuracy": 0.10820467174053192,
      "num_tokens": 7499116.0,
      "step": 2240
    },
    {
      "entropy": 7.034345626831055,
      "epoch": 0.43800604819042044,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993198542754692,
      "loss": 7.0649,
      "mean_token_accuracy": 0.11201289296150208,
      "num_tokens": 7515765.0,
      "step": 2245
    },
    {
      "entropy": 7.100940895080567,
      "epoch": 0.43898156277436345,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993143786625401,
      "loss": 7.0562,
      "mean_token_accuracy": 0.11279803812503815,
      "num_tokens": 7532074.0,
      "step": 2250
    },
    {
      "entropy": 6.856002759933472,
      "epoch": 0.4399570773583065,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993088811304092,
      "loss": 6.891,
      "mean_token_accuracy": 0.12257369831204415,
      "num_tokens": 7547704.0,
      "step": 2255
    },
    {
      "entropy": 7.078932285308838,
      "epoch": 0.44093259194224954,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993033616796139,
      "loss": 7.0064,
      "mean_token_accuracy": 0.11650566309690476,
      "num_tokens": 7563317.0,
      "step": 2260
    },
    {
      "entropy": 6.949041366577148,
      "epoch": 0.44190810652619256,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004992978203106936,
      "loss": 7.0034,
      "mean_token_accuracy": 0.1107703149318695,
      "num_tokens": 7580770.0,
      "step": 2265
    },
    {
      "entropy": 7.0201255798339846,
      "epoch": 0.4428836211101356,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004992922570241896,
      "loss": 6.9751,
      "mean_token_accuracy": 0.1167850024998188,
      "num_tokens": 7598181.0,
      "step": 2270
    },
    {
      "entropy": 7.028388261795044,
      "epoch": 0.44385913569407864,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992866718206457,
      "loss": 6.9768,
      "mean_token_accuracy": 0.11980046629905701,
      "num_tokens": 7613739.0,
      "step": 2275
    },
    {
      "entropy": 7.049353408813476,
      "epoch": 0.44483465027802166,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992810647006077,
      "loss": 7.0595,
      "mean_token_accuracy": 0.11202531084418296,
      "num_tokens": 7629297.0,
      "step": 2280
    },
    {
      "entropy": 7.0988686084747314,
      "epoch": 0.4458101648619647,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992754356646234,
      "loss": 7.0359,
      "mean_token_accuracy": 0.11156322211027145,
      "num_tokens": 7645273.0,
      "step": 2285
    },
    {
      "entropy": 7.062013387680054,
      "epoch": 0.4467856794459077,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004992697847132427,
      "loss": 6.9982,
      "mean_token_accuracy": 0.1103116363286972,
      "num_tokens": 7661296.0,
      "step": 2290
    },
    {
      "entropy": 6.91034460067749,
      "epoch": 0.44776119402985076,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000499264111847018,
      "loss": 6.997,
      "mean_token_accuracy": 0.11417637616395951,
      "num_tokens": 7678658.0,
      "step": 2295
    },
    {
      "entropy": 7.085509729385376,
      "epoch": 0.4487367086137938,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004992584170665037,
      "loss": 7.0253,
      "mean_token_accuracy": 0.11035038605332374,
      "num_tokens": 7693678.0,
      "step": 2300
    },
    {
      "entropy": 7.064776134490967,
      "epoch": 0.4497122231977368,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499252700372256,
      "loss": 7.0226,
      "mean_token_accuracy": 0.11261648908257485,
      "num_tokens": 7709526.0,
      "step": 2305
    },
    {
      "entropy": 7.004364538192749,
      "epoch": 0.4506877377816798,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992469617648337,
      "loss": 6.9237,
      "mean_token_accuracy": 0.11412907987833024,
      "num_tokens": 7725055.0,
      "step": 2310
    },
    {
      "entropy": 6.889954853057861,
      "epoch": 0.4516632523656229,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992412012447976,
      "loss": 6.8284,
      "mean_token_accuracy": 0.132039175927639,
      "num_tokens": 7742047.0,
      "step": 2315
    },
    {
      "entropy": 7.015885591506958,
      "epoch": 0.4526387669495659,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992354188127105,
      "loss": 6.9909,
      "mean_token_accuracy": 0.11513673812150955,
      "num_tokens": 7758283.0,
      "step": 2320
    },
    {
      "entropy": 6.970900583267212,
      "epoch": 0.4536142815335089,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992296144691375,
      "loss": 6.9359,
      "mean_token_accuracy": 0.1174176998436451,
      "num_tokens": 7774660.0,
      "step": 2325
    },
    {
      "entropy": 6.923218393325806,
      "epoch": 0.454589796117452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992237882146456,
      "loss": 6.9417,
      "mean_token_accuracy": 0.11896992549300194,
      "num_tokens": 7792147.0,
      "step": 2330
    },
    {
      "entropy": 6.939533996582031,
      "epoch": 0.455565310701395,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992179400498044,
      "loss": 6.8937,
      "mean_token_accuracy": 0.11782030612230301,
      "num_tokens": 7809021.0,
      "step": 2335
    },
    {
      "entropy": 7.026896572113037,
      "epoch": 0.456540825285338,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992120699751852,
      "loss": 6.962,
      "mean_token_accuracy": 0.1151209183037281,
      "num_tokens": 7825074.0,
      "step": 2340
    },
    {
      "entropy": 6.8574577331542965,
      "epoch": 0.45751633986928103,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992061779913616,
      "loss": 6.8863,
      "mean_token_accuracy": 0.12193938046693802,
      "num_tokens": 7843413.0,
      "step": 2345
    },
    {
      "entropy": 7.073199272155762,
      "epoch": 0.4584918544532241,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992002640989094,
      "loss": 6.9867,
      "mean_token_accuracy": 0.10821845382452011,
      "num_tokens": 7859874.0,
      "step": 2350
    },
    {
      "entropy": 6.867405271530151,
      "epoch": 0.4594673690371671,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004991943282984065,
      "loss": 6.9132,
      "mean_token_accuracy": 0.1207984820008278,
      "num_tokens": 7877302.0,
      "step": 2355
    },
    {
      "entropy": 7.034324979782104,
      "epoch": 0.46044288362111013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991883705904328,
      "loss": 7.0444,
      "mean_token_accuracy": 0.11169995591044427,
      "num_tokens": 7893207.0,
      "step": 2360
    },
    {
      "entropy": 6.87566819190979,
      "epoch": 0.46141839820505315,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991823909755705,
      "loss": 6.8706,
      "mean_token_accuracy": 0.1258732944726944,
      "num_tokens": 7910239.0,
      "step": 2365
    },
    {
      "entropy": 7.0774555683135985,
      "epoch": 0.4623939127889962,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991763894544041,
      "loss": 7.0115,
      "mean_token_accuracy": 0.11220350340008736,
      "num_tokens": 7928030.0,
      "step": 2370
    },
    {
      "entropy": 6.952159023284912,
      "epoch": 0.46336942737293924,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991703660275199,
      "loss": 6.9563,
      "mean_token_accuracy": 0.11582406088709832,
      "num_tokens": 7943653.0,
      "step": 2375
    },
    {
      "entropy": 7.076309013366699,
      "epoch": 0.46434494195688225,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991643206955064,
      "loss": 7.033,
      "mean_token_accuracy": 0.1133937492966652,
      "num_tokens": 7960564.0,
      "step": 2380
    },
    {
      "entropy": 7.016517114639282,
      "epoch": 0.46532045654082527,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004991582534589544,
      "loss": 6.9864,
      "mean_token_accuracy": 0.11657843217253686,
      "num_tokens": 7978045.0,
      "step": 2385
    },
    {
      "entropy": 7.005275440216065,
      "epoch": 0.46629597112476834,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991521643184567,
      "loss": 7.0214,
      "mean_token_accuracy": 0.11196061298251152,
      "num_tokens": 7995074.0,
      "step": 2390
    },
    {
      "entropy": 7.063782405853272,
      "epoch": 0.46727148570871135,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991460532746087,
      "loss": 7.0212,
      "mean_token_accuracy": 0.10701687708497047,
      "num_tokens": 8011800.0,
      "step": 2395
    },
    {
      "entropy": 6.993447494506836,
      "epoch": 0.46824700029265437,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499139920328007,
      "loss": 6.9919,
      "mean_token_accuracy": 0.11320114657282829,
      "num_tokens": 8028090.0,
      "step": 2400
    },
    {
      "entropy": 7.05373911857605,
      "epoch": 0.4692225148765974,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991337654792512,
      "loss": 6.9645,
      "mean_token_accuracy": 0.11767757162451745,
      "num_tokens": 8044140.0,
      "step": 2405
    },
    {
      "entropy": 6.914071464538575,
      "epoch": 0.47019802946054046,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991275887289427,
      "loss": 6.8981,
      "mean_token_accuracy": 0.11747554391622543,
      "num_tokens": 8060030.0,
      "step": 2410
    },
    {
      "entropy": 6.969658994674683,
      "epoch": 0.47117354404448347,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991213900776849,
      "loss": 6.945,
      "mean_token_accuracy": 0.11483549773693084,
      "num_tokens": 8075445.0,
      "step": 2415
    },
    {
      "entropy": 7.0339874744415285,
      "epoch": 0.4721490586284265,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991151695260838,
      "loss": 6.9459,
      "mean_token_accuracy": 0.11431458592414856,
      "num_tokens": 8093250.0,
      "step": 2420
    },
    {
      "entropy": 6.8671637058258055,
      "epoch": 0.4731245732123695,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499108927074747,
      "loss": 6.8974,
      "mean_token_accuracy": 0.12990258485078812,
      "num_tokens": 8109615.0,
      "step": 2425
    },
    {
      "entropy": 6.953899288177491,
      "epoch": 0.4741000877963126,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991026627242846,
      "loss": 6.9952,
      "mean_token_accuracy": 0.11256581172347069,
      "num_tokens": 8126591.0,
      "step": 2430
    },
    {
      "entropy": 6.994833993911743,
      "epoch": 0.4750756023802556,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990963764753088,
      "loss": 6.9212,
      "mean_token_accuracy": 0.11618381217122078,
      "num_tokens": 8141360.0,
      "step": 2435
    },
    {
      "entropy": 7.006396913528443,
      "epoch": 0.4760511169641986,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990900683284337,
      "loss": 7.0176,
      "mean_token_accuracy": 0.11735672503709793,
      "num_tokens": 8159688.0,
      "step": 2440
    },
    {
      "entropy": 6.978948783874512,
      "epoch": 0.4770266315481416,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990837382842758,
      "loss": 6.9747,
      "mean_token_accuracy": 0.11554915532469749,
      "num_tokens": 8176887.0,
      "step": 2445
    },
    {
      "entropy": 6.814501571655273,
      "epoch": 0.4780021461320847,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990773863434537,
      "loss": 6.7836,
      "mean_token_accuracy": 0.12544326335191727,
      "num_tokens": 8193372.0,
      "step": 2450
    },
    {
      "entropy": 6.993631887435913,
      "epoch": 0.4789776607160277,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004990710125065879,
      "loss": 7.0062,
      "mean_token_accuracy": 0.11460533291101456,
      "num_tokens": 8209713.0,
      "step": 2455
    },
    {
      "entropy": 7.021953868865967,
      "epoch": 0.4799531752999707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990646167743015,
      "loss": 7.0021,
      "mean_token_accuracy": 0.11365641131997109,
      "num_tokens": 8226359.0,
      "step": 2460
    },
    {
      "entropy": 6.913775777816772,
      "epoch": 0.48092868988391374,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990581991472193,
      "loss": 6.9722,
      "mean_token_accuracy": 0.11557035371661187,
      "num_tokens": 8243182.0,
      "step": 2465
    },
    {
      "entropy": 7.020796918869019,
      "epoch": 0.4819042044678568,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990517596259684,
      "loss": 6.9688,
      "mean_token_accuracy": 0.11420403867959976,
      "num_tokens": 8259185.0,
      "step": 2470
    },
    {
      "entropy": 7.048989534378052,
      "epoch": 0.4828797190517998,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990452982111781,
      "loss": 7.0372,
      "mean_token_accuracy": 0.10982453152537346,
      "num_tokens": 8275941.0,
      "step": 2475
    },
    {
      "entropy": 6.9889849662780765,
      "epoch": 0.48385523363574284,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990388149034797,
      "loss": 6.9699,
      "mean_token_accuracy": 0.11409663334488869,
      "num_tokens": 8292535.0,
      "step": 2480
    },
    {
      "entropy": 6.946880102157593,
      "epoch": 0.48483074821968586,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990323097035069,
      "loss": 6.929,
      "mean_token_accuracy": 0.12251158505678177,
      "num_tokens": 8310562.0,
      "step": 2485
    },
    {
      "entropy": 6.985591983795166,
      "epoch": 0.48580626280362893,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004990257826118952,
      "loss": 6.9546,
      "mean_token_accuracy": 0.10929256454110145,
      "num_tokens": 8327241.0,
      "step": 2490
    },
    {
      "entropy": 6.895032739639282,
      "epoch": 0.48678177738757195,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004990192336292824,
      "loss": 6.9035,
      "mean_token_accuracy": 0.11808636114001274,
      "num_tokens": 8343516.0,
      "step": 2495
    },
    {
      "entropy": 6.827773857116699,
      "epoch": 0.48775729197151496,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990126627563087,
      "loss": 6.7639,
      "mean_token_accuracy": 0.12457770556211471,
      "num_tokens": 8361809.0,
      "step": 2500
    },
    {
      "entropy": 6.819586992263794,
      "epoch": 0.488732806555458,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990060699936159,
      "loss": 6.8297,
      "mean_token_accuracy": 0.12708571404218674,
      "num_tokens": 8378632.0,
      "step": 2505
    },
    {
      "entropy": 6.9895045280456545,
      "epoch": 0.48970832113940105,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989994553418483,
      "loss": 6.985,
      "mean_token_accuracy": 0.11090396344661713,
      "num_tokens": 8396201.0,
      "step": 2510
    },
    {
      "entropy": 6.991191768646241,
      "epoch": 0.49068383572334406,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004989928188016522,
      "loss": 6.8719,
      "mean_token_accuracy": 0.12096665650606156,
      "num_tokens": 8412851.0,
      "step": 2515
    },
    {
      "entropy": 6.712438106536865,
      "epoch": 0.4916593503072871,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989861603736762,
      "loss": 6.7015,
      "mean_token_accuracy": 0.13208603635430335,
      "num_tokens": 8430026.0,
      "step": 2520
    },
    {
      "entropy": 6.962678480148315,
      "epoch": 0.49263486489123015,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989794800585711,
      "loss": 6.9305,
      "mean_token_accuracy": 0.11510602608323098,
      "num_tokens": 8446696.0,
      "step": 2525
    },
    {
      "entropy": 7.062203788757325,
      "epoch": 0.49361037947517317,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989727778569893,
      "loss": 7.0043,
      "mean_token_accuracy": 0.11590273976325989,
      "num_tokens": 8464827.0,
      "step": 2530
    },
    {
      "entropy": 6.9866313457489015,
      "epoch": 0.4945858940591162,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989660537695859,
      "loss": 6.9741,
      "mean_token_accuracy": 0.11345633417367935,
      "num_tokens": 8480185.0,
      "step": 2535
    },
    {
      "entropy": 6.862218570709229,
      "epoch": 0.4955614086430592,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498959307797018,
      "loss": 6.8559,
      "mean_token_accuracy": 0.126577777415514,
      "num_tokens": 8497511.0,
      "step": 2540
    },
    {
      "entropy": 6.9351707935333256,
      "epoch": 0.49653692322700227,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989525399399447,
      "loss": 6.9272,
      "mean_token_accuracy": 0.11688981875777245,
      "num_tokens": 8514394.0,
      "step": 2545
    },
    {
      "entropy": 6.986079406738281,
      "epoch": 0.4975124378109453,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004989457501990275,
      "loss": 6.9237,
      "mean_token_accuracy": 0.11213519722223282,
      "num_tokens": 8531659.0,
      "step": 2550
    },
    {
      "entropy": 6.9696861743927006,
      "epoch": 0.4984879523948883,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004989389385749297,
      "loss": 6.972,
      "mean_token_accuracy": 0.11650464758276939,
      "num_tokens": 8547986.0,
      "step": 2555
    },
    {
      "entropy": 6.967264032363891,
      "epoch": 0.4994634669788313,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004989321050683169,
      "loss": 6.9337,
      "mean_token_accuracy": 0.11631564348936081,
      "num_tokens": 8565021.0,
      "step": 2560
    },
    {
      "entropy": 6.929176139831543,
      "epoch": 0.5004389815627743,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498925249679857,
      "loss": 6.9337,
      "mean_token_accuracy": 0.11361033692955971,
      "num_tokens": 8582135.0,
      "step": 2565
    },
    {
      "entropy": 6.898085260391236,
      "epoch": 0.5014144961467174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989183724102196,
      "loss": 6.9669,
      "mean_token_accuracy": 0.11791260913014412,
      "num_tokens": 8599946.0,
      "step": 2570
    },
    {
      "entropy": 7.006285619735718,
      "epoch": 0.5023900107306605,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989114732600771,
      "loss": 6.9814,
      "mean_token_accuracy": 0.11188279166817665,
      "num_tokens": 8615831.0,
      "step": 2575
    },
    {
      "entropy": 6.918506193161011,
      "epoch": 0.5033655253146034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989045522301033,
      "loss": 6.877,
      "mean_token_accuracy": 0.11590302884578704,
      "num_tokens": 8631886.0,
      "step": 2580
    },
    {
      "entropy": 6.969996213912964,
      "epoch": 0.5043410398985465,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988976093209747,
      "loss": 6.922,
      "mean_token_accuracy": 0.11348427534103393,
      "num_tokens": 8647520.0,
      "step": 2585
    },
    {
      "entropy": 6.89022126197815,
      "epoch": 0.5053165544824895,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988906445333698,
      "loss": 6.8236,
      "mean_token_accuracy": 0.12668280601501464,
      "num_tokens": 8663995.0,
      "step": 2590
    },
    {
      "entropy": 6.906827592849732,
      "epoch": 0.5062920690664325,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988836578679691,
      "loss": 6.9132,
      "mean_token_accuracy": 0.11209813728928567,
      "num_tokens": 8680201.0,
      "step": 2595
    },
    {
      "entropy": 6.970617389678955,
      "epoch": 0.5072675836503756,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004988766493254552,
      "loss": 6.9094,
      "mean_token_accuracy": 0.12233168333768844,
      "num_tokens": 8696978.0,
      "step": 2600
    },
    {
      "entropy": 6.8369691371917725,
      "epoch": 0.5082430982343186,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498869618906513,
      "loss": 6.9549,
      "mean_token_accuracy": 0.11839013695716857,
      "num_tokens": 8713335.0,
      "step": 2605
    },
    {
      "entropy": 6.996060132980347,
      "epoch": 0.5092186128182616,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988625666118296,
      "loss": 6.9367,
      "mean_token_accuracy": 0.11863038241863251,
      "num_tokens": 8730722.0,
      "step": 2610
    },
    {
      "entropy": 6.884844493865967,
      "epoch": 0.5101941274022047,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498855492442094,
      "loss": 6.8577,
      "mean_token_accuracy": 0.11674225553870202,
      "num_tokens": 8747430.0,
      "step": 2615
    },
    {
      "entropy": 6.927949571609497,
      "epoch": 0.5111696419861477,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988483963979975,
      "loss": 6.9499,
      "mean_token_accuracy": 0.1175406351685524,
      "num_tokens": 8764817.0,
      "step": 2620
    },
    {
      "entropy": 6.976333093643189,
      "epoch": 0.5121451565700907,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988412784802336,
      "loss": 6.9705,
      "mean_token_accuracy": 0.11199398934841157,
      "num_tokens": 8780412.0,
      "step": 2625
    },
    {
      "entropy": 7.039850282669067,
      "epoch": 0.5131206711540337,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988341386894977,
      "loss": 7.0343,
      "mean_token_accuracy": 0.11212069839239121,
      "num_tokens": 8795526.0,
      "step": 2630
    },
    {
      "entropy": 6.928880023956299,
      "epoch": 0.5140961857379768,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004988269770264876,
      "loss": 6.8226,
      "mean_token_accuracy": 0.1236750140786171,
      "num_tokens": 8813568.0,
      "step": 2635
    },
    {
      "entropy": 7.019903326034546,
      "epoch": 0.5150717003219198,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004988197934919029,
      "loss": 6.9877,
      "mean_token_accuracy": 0.11166180521249772,
      "num_tokens": 8831807.0,
      "step": 2640
    },
    {
      "entropy": 6.995202112197876,
      "epoch": 0.5160472149058628,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004988125880864458,
      "loss": 6.9858,
      "mean_token_accuracy": 0.11477422341704369,
      "num_tokens": 8848875.0,
      "step": 2645
    },
    {
      "entropy": 6.766485023498535,
      "epoch": 0.5170227294898059,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004988053608108203,
      "loss": 6.6849,
      "mean_token_accuracy": 0.12702686265110968,
      "num_tokens": 8866317.0,
      "step": 2650
    },
    {
      "entropy": 6.885069274902344,
      "epoch": 0.517998244073749,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987981116657325,
      "loss": 6.8418,
      "mean_token_accuracy": 0.11617801636457444,
      "num_tokens": 8882543.0,
      "step": 2655
    },
    {
      "entropy": 6.942510557174683,
      "epoch": 0.5189737586576919,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987908406518909,
      "loss": 6.8619,
      "mean_token_accuracy": 0.11533142924308777,
      "num_tokens": 8898640.0,
      "step": 2660
    },
    {
      "entropy": 6.9259865283966064,
      "epoch": 0.519949273241635,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498783547770006,
      "loss": 6.9629,
      "mean_token_accuracy": 0.11299280896782875,
      "num_tokens": 8914649.0,
      "step": 2665
    },
    {
      "entropy": 6.984437799453735,
      "epoch": 0.5209247878255779,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987762330207903,
      "loss": 6.9722,
      "mean_token_accuracy": 0.11515414044260978,
      "num_tokens": 8931585.0,
      "step": 2670
    },
    {
      "entropy": 6.820010805130005,
      "epoch": 0.521900302409521,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987688964049587,
      "loss": 6.8595,
      "mean_token_accuracy": 0.12085901573300362,
      "num_tokens": 8948827.0,
      "step": 2675
    },
    {
      "entropy": 6.8946051597595215,
      "epoch": 0.5228758169934641,
      "grad_norm": 0.765625,
      "learning_rate": 0.000498761537923228,
      "loss": 6.8342,
      "mean_token_accuracy": 0.1254794180393219,
      "num_tokens": 8965680.0,
      "step": 2680
    },
    {
      "entropy": 7.019130516052246,
      "epoch": 0.523851331577407,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987541575763175,
      "loss": 6.9597,
      "mean_token_accuracy": 0.10987325012683868,
      "num_tokens": 8982813.0,
      "step": 2685
    },
    {
      "entropy": 6.912060880661011,
      "epoch": 0.5248268461613501,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498746755364948,
      "loss": 6.9388,
      "mean_token_accuracy": 0.11895869076251983,
      "num_tokens": 9000497.0,
      "step": 2690
    },
    {
      "entropy": 6.953133487701416,
      "epoch": 0.5258023607452932,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004987393312898431,
      "loss": 6.8675,
      "mean_token_accuracy": 0.11721265986561775,
      "num_tokens": 9016195.0,
      "step": 2695
    },
    {
      "entropy": 6.867087030410767,
      "epoch": 0.5267778753292361,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987318853517282,
      "loss": 6.8329,
      "mean_token_accuracy": 0.12630052193999292,
      "num_tokens": 9033295.0,
      "step": 2700
    },
    {
      "entropy": 6.92639045715332,
      "epoch": 0.5277533899131792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987244175513308,
      "loss": 6.9239,
      "mean_token_accuracy": 0.11546016260981559,
      "num_tokens": 9050725.0,
      "step": 2705
    },
    {
      "entropy": 6.946275138854981,
      "epoch": 0.5287289044971223,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004987169278893807,
      "loss": 6.8693,
      "mean_token_accuracy": 0.11885097920894623,
      "num_tokens": 9067842.0,
      "step": 2710
    },
    {
      "entropy": 6.9172258377075195,
      "epoch": 0.5297044190810652,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987094163666098,
      "loss": 6.8994,
      "mean_token_accuracy": 0.1129316121339798,
      "num_tokens": 9084602.0,
      "step": 2715
    },
    {
      "entropy": 6.8367022514343265,
      "epoch": 0.5306799336650083,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498701882983752,
      "loss": 6.8249,
      "mean_token_accuracy": 0.1250123456120491,
      "num_tokens": 9100530.0,
      "step": 2720
    },
    {
      "entropy": 6.962347030639648,
      "epoch": 0.5316554482489513,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986943277415435,
      "loss": 6.9101,
      "mean_token_accuracy": 0.12127360850572586,
      "num_tokens": 9117304.0,
      "step": 2725
    },
    {
      "entropy": 6.802248811721801,
      "epoch": 0.5326309628328944,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004986867506407225,
      "loss": 6.8584,
      "mean_token_accuracy": 0.13018720969557762,
      "num_tokens": 9133974.0,
      "step": 2730
    },
    {
      "entropy": 6.919925498962402,
      "epoch": 0.5336064774168374,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986791516820295,
      "loss": 6.8217,
      "mean_token_accuracy": 0.11861017867922782,
      "num_tokens": 9150600.0,
      "step": 2735
    },
    {
      "entropy": 6.799954748153686,
      "epoch": 0.5345819920007804,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000498671530866207,
      "loss": 6.8359,
      "mean_token_accuracy": 0.11661089509725571,
      "num_tokens": 9167449.0,
      "step": 2740
    },
    {
      "entropy": 6.932155656814575,
      "epoch": 0.5355575065847235,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986638881939997,
      "loss": 6.7617,
      "mean_token_accuracy": 0.12293269336223603,
      "num_tokens": 9184570.0,
      "step": 2745
    },
    {
      "entropy": 6.8632814407348635,
      "epoch": 0.5365330211686665,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986562236661544,
      "loss": 6.9606,
      "mean_token_accuracy": 0.1100731372833252,
      "num_tokens": 9200537.0,
      "step": 2750
    },
    {
      "entropy": 6.9976402759552006,
      "epoch": 0.5375085357526095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00049864853728342,
      "loss": 6.8563,
      "mean_token_accuracy": 0.1256386786699295,
      "num_tokens": 9217252.0,
      "step": 2755
    },
    {
      "entropy": 6.891844272613525,
      "epoch": 0.5384840503365526,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986408290465476,
      "loss": 6.9017,
      "mean_token_accuracy": 0.11649853512644767,
      "num_tokens": 9234108.0,
      "step": 2760
    },
    {
      "entropy": 6.9134743213653564,
      "epoch": 0.5394595649204955,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986330989562906,
      "loss": 6.8092,
      "mean_token_accuracy": 0.12046945318579674,
      "num_tokens": 9251017.0,
      "step": 2765
    },
    {
      "entropy": 6.7959212303161625,
      "epoch": 0.5404350795044386,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986253470134042,
      "loss": 6.7755,
      "mean_token_accuracy": 0.1264660306274891,
      "num_tokens": 9267922.0,
      "step": 2770
    },
    {
      "entropy": 6.800236415863037,
      "epoch": 0.5414105940883817,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986175732186459,
      "loss": 6.7688,
      "mean_token_accuracy": 0.13141302540898323,
      "num_tokens": 9284954.0,
      "step": 2775
    },
    {
      "entropy": 6.9570934772491455,
      "epoch": 0.5423861086723246,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004986097775727752,
      "loss": 6.883,
      "mean_token_accuracy": 0.12054598703980446,
      "num_tokens": 9303306.0,
      "step": 2780
    },
    {
      "entropy": 6.789645624160767,
      "epoch": 0.5433616232562677,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004986019600765541,
      "loss": 6.7905,
      "mean_token_accuracy": 0.12379697784781456,
      "num_tokens": 9319754.0,
      "step": 2785
    },
    {
      "entropy": 6.855822896957397,
      "epoch": 0.5443371378402108,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985941207307462,
      "loss": 6.8849,
      "mean_token_accuracy": 0.11971316859126091,
      "num_tokens": 9335736.0,
      "step": 2790
    },
    {
      "entropy": 6.860150671005249,
      "epoch": 0.5453126524241537,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985862595361179,
      "loss": 6.8503,
      "mean_token_accuracy": 0.12303772121667862,
      "num_tokens": 9350435.0,
      "step": 2795
    },
    {
      "entropy": 6.943622159957886,
      "epoch": 0.5462881670080968,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985783764934372,
      "loss": 6.9081,
      "mean_token_accuracy": 0.11459068953990936,
      "num_tokens": 9367613.0,
      "step": 2800
    },
    {
      "entropy": 6.865189838409424,
      "epoch": 0.5472636815920398,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985704716034744,
      "loss": 6.8042,
      "mean_token_accuracy": 0.12057539001107216,
      "num_tokens": 9385590.0,
      "step": 2805
    },
    {
      "entropy": 6.917732000350952,
      "epoch": 0.5482391961759828,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004985625448670019,
      "loss": 6.934,
      "mean_token_accuracy": 0.11029709205031395,
      "num_tokens": 9403286.0,
      "step": 2810
    },
    {
      "entropy": 6.978166198730468,
      "epoch": 0.5492147107599259,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985545962847943,
      "loss": 6.9123,
      "mean_token_accuracy": 0.11426826044917107,
      "num_tokens": 9419880.0,
      "step": 2815
    },
    {
      "entropy": 6.894027137756348,
      "epoch": 0.5501902253438689,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985466258576283,
      "loss": 6.8613,
      "mean_token_accuracy": 0.11296845152974129,
      "num_tokens": 9436579.0,
      "step": 2820
    },
    {
      "entropy": 6.968469572067261,
      "epoch": 0.5511657399278119,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985386335862829,
      "loss": 6.9412,
      "mean_token_accuracy": 0.11159636527299881,
      "num_tokens": 9452802.0,
      "step": 2825
    },
    {
      "entropy": 6.914433288574219,
      "epoch": 0.552141254511755,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985306194715387,
      "loss": 6.9498,
      "mean_token_accuracy": 0.11330530717968941,
      "num_tokens": 9468297.0,
      "step": 2830
    },
    {
      "entropy": 6.955451250076294,
      "epoch": 0.553116769095698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985225835141792,
      "loss": 6.9513,
      "mean_token_accuracy": 0.11002258956432343,
      "num_tokens": 9484954.0,
      "step": 2835
    },
    {
      "entropy": 7.008716344833374,
      "epoch": 0.554092283679641,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985145257149895,
      "loss": 6.9883,
      "mean_token_accuracy": 0.11100693419575691,
      "num_tokens": 9502211.0,
      "step": 2840
    },
    {
      "entropy": 6.928589630126953,
      "epoch": 0.555067798263584,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004985064460747569,
      "loss": 6.8805,
      "mean_token_accuracy": 0.11256838291883468,
      "num_tokens": 9518103.0,
      "step": 2845
    },
    {
      "entropy": 6.83609938621521,
      "epoch": 0.5560433128475271,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498498344594271,
      "loss": 6.8144,
      "mean_token_accuracy": 0.11466094553470611,
      "num_tokens": 9534155.0,
      "step": 2850
    },
    {
      "entropy": 6.833941411972046,
      "epoch": 0.5570188274314701,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984902212743234,
      "loss": 6.7947,
      "mean_token_accuracy": 0.12757234200835227,
      "num_tokens": 9549972.0,
      "step": 2855
    },
    {
      "entropy": 6.80628137588501,
      "epoch": 0.5579943420154131,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984820761157079,
      "loss": 6.7243,
      "mean_token_accuracy": 0.13037320226430893,
      "num_tokens": 9567306.0,
      "step": 2860
    },
    {
      "entropy": 6.9516716480255125,
      "epoch": 0.5589698565993562,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984739091192205,
      "loss": 6.9277,
      "mean_token_accuracy": 0.1152263417840004,
      "num_tokens": 9583488.0,
      "step": 2865
    },
    {
      "entropy": 6.741307401657105,
      "epoch": 0.5599453711832992,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984657202856592,
      "loss": 6.7223,
      "mean_token_accuracy": 0.13048833832144738,
      "num_tokens": 9600030.0,
      "step": 2870
    },
    {
      "entropy": 6.8341851234436035,
      "epoch": 0.5609208857672422,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498457509615824,
      "loss": 6.8515,
      "mean_token_accuracy": 0.12074055820703507,
      "num_tokens": 9615791.0,
      "step": 2875
    },
    {
      "entropy": 6.995968627929687,
      "epoch": 0.5618964003511853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984492771105175,
      "loss": 6.9204,
      "mean_token_accuracy": 0.11317621320486068,
      "num_tokens": 9630955.0,
      "step": 2880
    },
    {
      "entropy": 6.782270812988282,
      "epoch": 0.5628719149351282,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984410227705439,
      "loss": 6.8268,
      "mean_token_accuracy": 0.1273679867386818,
      "num_tokens": 9647743.0,
      "step": 2885
    },
    {
      "entropy": 6.832012748718261,
      "epoch": 0.5638474295190713,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984327465967101,
      "loss": 6.7574,
      "mean_token_accuracy": 0.12513990476727485,
      "num_tokens": 9665495.0,
      "step": 2890
    },
    {
      "entropy": 6.813667488098145,
      "epoch": 0.5648229441030144,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004984244485898246,
      "loss": 6.7867,
      "mean_token_accuracy": 0.12143838852643966,
      "num_tokens": 9683489.0,
      "step": 2895
    },
    {
      "entropy": 6.702491760253906,
      "epoch": 0.5657984586869573,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004984161287506982,
      "loss": 6.7186,
      "mean_token_accuracy": 0.14046730175614358,
      "num_tokens": 9700491.0,
      "step": 2900
    },
    {
      "entropy": 6.982992839813233,
      "epoch": 0.5667739732709004,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498407787080144,
      "loss": 6.8942,
      "mean_token_accuracy": 0.11420398429036141,
      "num_tokens": 9715694.0,
      "step": 2905
    },
    {
      "entropy": 6.88687162399292,
      "epoch": 0.5677494878548435,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498399423578977,
      "loss": 6.9429,
      "mean_token_accuracy": 0.11011143624782563,
      "num_tokens": 9731317.0,
      "step": 2910
    },
    {
      "entropy": 7.040346956253051,
      "epoch": 0.5687250024387864,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983910382480146,
      "loss": 6.925,
      "mean_token_accuracy": 0.11294664517045021,
      "num_tokens": 9747928.0,
      "step": 2915
    },
    {
      "entropy": 6.869180822372437,
      "epoch": 0.5697005170227295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983826310880762,
      "loss": 6.8392,
      "mean_token_accuracy": 0.11410067304968834,
      "num_tokens": 9765454.0,
      "step": 2920
    },
    {
      "entropy": 6.957724380493164,
      "epoch": 0.5706760316066726,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983742020999832,
      "loss": 6.909,
      "mean_token_accuracy": 0.11039541885256768,
      "num_tokens": 9782255.0,
      "step": 2925
    },
    {
      "entropy": 6.745351552963257,
      "epoch": 0.5716515461906155,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983657512845593,
      "loss": 6.7201,
      "mean_token_accuracy": 0.13016780316829682,
      "num_tokens": 9798680.0,
      "step": 2930
    },
    {
      "entropy": 6.912977790832519,
      "epoch": 0.5726270607745586,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983572786426302,
      "loss": 6.8858,
      "mean_token_accuracy": 0.116475909948349,
      "num_tokens": 9813955.0,
      "step": 2935
    },
    {
      "entropy": 6.743524360656738,
      "epoch": 0.5736025753585016,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983487841750239,
      "loss": 6.7628,
      "mean_token_accuracy": 0.12516291812062263,
      "num_tokens": 9830692.0,
      "step": 2940
    },
    {
      "entropy": 6.908202314376831,
      "epoch": 0.5745780899424446,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983402678825705,
      "loss": 6.8886,
      "mean_token_accuracy": 0.11570696383714676,
      "num_tokens": 9848825.0,
      "step": 2945
    },
    {
      "entropy": 6.849818515777588,
      "epoch": 0.5755536045263877,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983317297661021,
      "loss": 6.8759,
      "mean_token_accuracy": 0.11652199402451516,
      "num_tokens": 9865802.0,
      "step": 2950
    },
    {
      "entropy": 6.935366773605347,
      "epoch": 0.5765291191103307,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004983231698264531,
      "loss": 6.9181,
      "mean_token_accuracy": 0.11485676914453506,
      "num_tokens": 9883540.0,
      "step": 2955
    },
    {
      "entropy": 6.903877782821655,
      "epoch": 0.5775046336942737,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983145880644597,
      "loss": 6.8614,
      "mean_token_accuracy": 0.11511279791593551,
      "num_tokens": 9897724.0,
      "step": 2960
    },
    {
      "entropy": 6.9072425365448,
      "epoch": 0.5784801482782168,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983059844809608,
      "loss": 6.8037,
      "mean_token_accuracy": 0.12381555512547493,
      "num_tokens": 9914710.0,
      "step": 2965
    },
    {
      "entropy": 6.824421691894531,
      "epoch": 0.5794556628621598,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004982973590767969,
      "loss": 6.9211,
      "mean_token_accuracy": 0.11980470642447472,
      "num_tokens": 9932362.0,
      "step": 2970
    },
    {
      "entropy": 6.857925033569336,
      "epoch": 0.5804311774461028,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498288711852811,
      "loss": 6.7182,
      "mean_token_accuracy": 0.12366797849535942,
      "num_tokens": 9949409.0,
      "step": 2975
    },
    {
      "entropy": 6.862184047698975,
      "epoch": 0.5814066920300458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498280042809848,
      "loss": 6.8736,
      "mean_token_accuracy": 0.11821404024958611,
      "num_tokens": 9967467.0,
      "step": 2980
    },
    {
      "entropy": 6.876873397827149,
      "epoch": 0.5823822066139889,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982713519487549,
      "loss": 6.7982,
      "mean_token_accuracy": 0.12363352626562119,
      "num_tokens": 9983082.0,
      "step": 2985
    },
    {
      "entropy": 6.845055437088012,
      "epoch": 0.5833577211979319,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982626392703812,
      "loss": 6.8781,
      "mean_token_accuracy": 0.11882330104708672,
      "num_tokens": 10000046.0,
      "step": 2990
    },
    {
      "entropy": 6.910937643051147,
      "epoch": 0.5843332357818749,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498253904775578,
      "loss": 6.8151,
      "mean_token_accuracy": 0.11773023009300232,
      "num_tokens": 10014674.0,
      "step": 2995
    },
    {
      "entropy": 6.7271734237670895,
      "epoch": 0.585308750365818,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982451484651988,
      "loss": 6.6986,
      "mean_token_accuracy": 0.12597279995679855,
      "num_tokens": 10030647.0,
      "step": 3000
    },
    {
      "epoch": 0.585308750365818,
      "eval_entropy": 6.672496932747855,
      "eval_loss": 6.871516227722168,
      "eval_mean_token_accuracy": 0.12053095382039408,
      "eval_num_tokens": 10030647.0,
      "eval_runtime": 15.3632,
      "eval_samples_per_second": 1079.005,
      "eval_steps_per_second": 134.933,
      "step": 3000
    },
    {
      "entropy": 6.86599702835083,
      "epoch": 0.586284264949761,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982363703400995,
      "loss": 6.7855,
      "mean_token_accuracy": 0.11965514495968818,
      "num_tokens": 10046468.0,
      "step": 3005
    },
    {
      "entropy": 6.906035280227661,
      "epoch": 0.587259779533704,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004982275704011377,
      "loss": 6.889,
      "mean_token_accuracy": 0.11167991161346436,
      "num_tokens": 10063062.0,
      "step": 3010
    },
    {
      "entropy": 6.829477262496948,
      "epoch": 0.5882352941176471,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982187486491734,
      "loss": 6.8367,
      "mean_token_accuracy": 0.11886958852410316,
      "num_tokens": 10079616.0,
      "step": 3015
    },
    {
      "entropy": 6.7937616348266605,
      "epoch": 0.58921080870159,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982099050850685,
      "loss": 6.808,
      "mean_token_accuracy": 0.11928128674626351,
      "num_tokens": 10096830.0,
      "step": 3020
    },
    {
      "entropy": 6.84074444770813,
      "epoch": 0.5901863232855331,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982010397096871,
      "loss": 6.8439,
      "mean_token_accuracy": 0.12148243859410286,
      "num_tokens": 10114041.0,
      "step": 3025
    },
    {
      "entropy": 6.863445138931274,
      "epoch": 0.5911618378694762,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981921525238957,
      "loss": 6.7635,
      "mean_token_accuracy": 0.1219376154243946,
      "num_tokens": 10130075.0,
      "step": 3030
    },
    {
      "entropy": 6.855212831497193,
      "epoch": 0.5921373524534191,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004981832435285626,
      "loss": 6.9046,
      "mean_token_accuracy": 0.11763010397553444,
      "num_tokens": 10146716.0,
      "step": 3035
    },
    {
      "entropy": 6.9085009574890135,
      "epoch": 0.5931128670373622,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981743127245584,
      "loss": 6.8288,
      "mean_token_accuracy": 0.12004327401518822,
      "num_tokens": 10162116.0,
      "step": 3040
    },
    {
      "entropy": 6.958953905105591,
      "epoch": 0.5940883816213053,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981653601127557,
      "loss": 6.9243,
      "mean_token_accuracy": 0.11546042636036873,
      "num_tokens": 10179326.0,
      "step": 3045
    },
    {
      "entropy": 6.850890684127807,
      "epoch": 0.5950638962052482,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981563856940295,
      "loss": 6.8533,
      "mean_token_accuracy": 0.12064573913812637,
      "num_tokens": 10195278.0,
      "step": 3050
    },
    {
      "entropy": 6.821883344650269,
      "epoch": 0.5960394107891913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981473894692565,
      "loss": 6.7775,
      "mean_token_accuracy": 0.12411454319953918,
      "num_tokens": 10210175.0,
      "step": 3055
    },
    {
      "entropy": 6.876020765304565,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498138371439316,
      "loss": 6.8758,
      "mean_token_accuracy": 0.11442647203803062,
      "num_tokens": 10225600.0,
      "step": 3060
    },
    {
      "entropy": 6.684839057922363,
      "epoch": 0.5979904399570773,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004981293316050891,
      "loss": 6.5734,
      "mean_token_accuracy": 0.14426453411579132,
      "num_tokens": 10243451.0,
      "step": 3065
    },
    {
      "entropy": 6.940177154541016,
      "epoch": 0.5989659545410204,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981202699674592,
      "loss": 6.8593,
      "mean_token_accuracy": 0.11604441031813621,
      "num_tokens": 10260392.0,
      "step": 3070
    },
    {
      "entropy": 6.876484632492065,
      "epoch": 0.5999414691249634,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004981111865273117,
      "loss": 6.8958,
      "mean_token_accuracy": 0.11444505527615548,
      "num_tokens": 10277016.0,
      "step": 3075
    },
    {
      "entropy": 6.766681432723999,
      "epoch": 0.6009169837089064,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004981020812855342,
      "loss": 6.753,
      "mean_token_accuracy": 0.12857064604759216,
      "num_tokens": 10295927.0,
      "step": 3080
    },
    {
      "entropy": 6.862094068527222,
      "epoch": 0.6018924982928495,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980929542430165,
      "loss": 6.8444,
      "mean_token_accuracy": 0.11809502243995666,
      "num_tokens": 10311140.0,
      "step": 3085
    },
    {
      "entropy": 6.857626962661743,
      "epoch": 0.6028680128767925,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004980838054006504,
      "loss": 6.8804,
      "mean_token_accuracy": 0.11959288492798806,
      "num_tokens": 10328496.0,
      "step": 3090
    },
    {
      "entropy": 6.925609111785889,
      "epoch": 0.6038435274607356,
      "grad_norm": 0.875,
      "learning_rate": 0.00049807463475933,
      "loss": 6.8267,
      "mean_token_accuracy": 0.11578144133090973,
      "num_tokens": 10345232.0,
      "step": 3095
    },
    {
      "entropy": 6.927273464202881,
      "epoch": 0.6048190420446786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980654423199513,
      "loss": 6.9633,
      "mean_token_accuracy": 0.11427217870950698,
      "num_tokens": 10361777.0,
      "step": 3100
    },
    {
      "entropy": 6.87502760887146,
      "epoch": 0.6057945566286216,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980562280834126,
      "loss": 6.8915,
      "mean_token_accuracy": 0.12031263709068299,
      "num_tokens": 10377117.0,
      "step": 3105
    },
    {
      "entropy": 6.698544502258301,
      "epoch": 0.6067700712125647,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980469920506142,
      "loss": 6.524,
      "mean_token_accuracy": 0.14353142753243447,
      "num_tokens": 10394506.0,
      "step": 3110
    },
    {
      "entropy": 6.7167610168457035,
      "epoch": 0.6077455857965076,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980377342224588,
      "loss": 6.6573,
      "mean_token_accuracy": 0.1346955791115761,
      "num_tokens": 10410281.0,
      "step": 3115
    },
    {
      "entropy": 6.896825361251831,
      "epoch": 0.6087211003804507,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980284545998509,
      "loss": 6.8756,
      "mean_token_accuracy": 0.11365101113915443,
      "num_tokens": 10425165.0,
      "step": 3120
    },
    {
      "entropy": 6.92686095237732,
      "epoch": 0.6096966149643938,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980191531836973,
      "loss": 6.8321,
      "mean_token_accuracy": 0.11379202753305435,
      "num_tokens": 10441643.0,
      "step": 3125
    },
    {
      "entropy": 6.8630529880523685,
      "epoch": 0.6106721295483367,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980098299749068,
      "loss": 6.8615,
      "mean_token_accuracy": 0.11276723966002464,
      "num_tokens": 10457436.0,
      "step": 3130
    },
    {
      "entropy": 6.902652168273926,
      "epoch": 0.6116476441322798,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980004849743908,
      "loss": 6.8693,
      "mean_token_accuracy": 0.11519403904676437,
      "num_tokens": 10475020.0,
      "step": 3135
    },
    {
      "entropy": 6.64963436126709,
      "epoch": 0.6126231587162229,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497991118183062,
      "loss": 6.648,
      "mean_token_accuracy": 0.13124900534749032,
      "num_tokens": 10492029.0,
      "step": 3140
    },
    {
      "entropy": 6.741730451583862,
      "epoch": 0.6135986733001658,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004979817296018359,
      "loss": 6.6663,
      "mean_token_accuracy": 0.13161050975322724,
      "num_tokens": 10510145.0,
      "step": 3145
    },
    {
      "entropy": 6.892311382293701,
      "epoch": 0.6145741878841089,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979723192316299,
      "loss": 6.9094,
      "mean_token_accuracy": 0.11838359609246255,
      "num_tokens": 10527192.0,
      "step": 3150
    },
    {
      "entropy": 6.915144205093384,
      "epoch": 0.6155497024680519,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979628870733637,
      "loss": 6.8586,
      "mean_token_accuracy": 0.11659516468644142,
      "num_tokens": 10542269.0,
      "step": 3155
    },
    {
      "entropy": 6.840810918807984,
      "epoch": 0.6165252170519949,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979534331279586,
      "loss": 6.8419,
      "mean_token_accuracy": 0.11983768492937089,
      "num_tokens": 10559425.0,
      "step": 3160
    },
    {
      "entropy": 6.905071830749511,
      "epoch": 0.617500731635938,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979439573963388,
      "loss": 6.8527,
      "mean_token_accuracy": 0.11907846704125405,
      "num_tokens": 10576818.0,
      "step": 3165
    },
    {
      "entropy": 6.7245340824127195,
      "epoch": 0.618476246219881,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979344598794299,
      "loss": 6.7534,
      "mean_token_accuracy": 0.12337901741266251,
      "num_tokens": 10593456.0,
      "step": 3170
    },
    {
      "entropy": 6.938669061660766,
      "epoch": 0.619451760803824,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004979249405781602,
      "loss": 6.9348,
      "mean_token_accuracy": 0.11441336721181869,
      "num_tokens": 10609049.0,
      "step": 3175
    },
    {
      "entropy": 6.78485255241394,
      "epoch": 0.6204272753877671,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004979153994934599,
      "loss": 6.7529,
      "mean_token_accuracy": 0.12293577119708061,
      "num_tokens": 10626921.0,
      "step": 3180
    },
    {
      "entropy": 6.843013620376587,
      "epoch": 0.6214027899717101,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979058366262613,
      "loss": 6.825,
      "mean_token_accuracy": 0.11384221240878105,
      "num_tokens": 10643447.0,
      "step": 3185
    },
    {
      "entropy": 6.902149200439453,
      "epoch": 0.6223783045556531,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978962519774987,
      "loss": 6.8907,
      "mean_token_accuracy": 0.11368588134646415,
      "num_tokens": 10659896.0,
      "step": 3190
    },
    {
      "entropy": 6.81766529083252,
      "epoch": 0.6233538191395961,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978866455481087,
      "loss": 6.7501,
      "mean_token_accuracy": 0.12061190605163574,
      "num_tokens": 10675331.0,
      "step": 3195
    },
    {
      "entropy": 6.903246831893921,
      "epoch": 0.6243293337235392,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978770173390301,
      "loss": 6.9067,
      "mean_token_accuracy": 0.11321326643228531,
      "num_tokens": 10692678.0,
      "step": 3200
    },
    {
      "entropy": 6.921315002441406,
      "epoch": 0.6253048483074822,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978673673512039,
      "loss": 6.8339,
      "mean_token_accuracy": 0.11405565664172172,
      "num_tokens": 10708500.0,
      "step": 3205
    },
    {
      "entropy": 6.786546134948731,
      "epoch": 0.6262803628914252,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978576955855727,
      "loss": 6.7677,
      "mean_token_accuracy": 0.11904651448130607,
      "num_tokens": 10725424.0,
      "step": 3210
    },
    {
      "entropy": 6.86941967010498,
      "epoch": 0.6272558774753683,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978480020430818,
      "loss": 6.7885,
      "mean_token_accuracy": 0.12381382882595063,
      "num_tokens": 10740960.0,
      "step": 3215
    },
    {
      "entropy": 6.8420721054077145,
      "epoch": 0.6282313920593113,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978382867246784,
      "loss": 6.8334,
      "mean_token_accuracy": 0.11320647895336151,
      "num_tokens": 10756426.0,
      "step": 3220
    },
    {
      "entropy": 6.777997732162476,
      "epoch": 0.6292069066432543,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978285496313119,
      "loss": 6.7034,
      "mean_token_accuracy": 0.12760446295142175,
      "num_tokens": 10772748.0,
      "step": 3225
    },
    {
      "entropy": 6.7658689498901365,
      "epoch": 0.6301824212271974,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978187907639336,
      "loss": 6.7412,
      "mean_token_accuracy": 0.1257241800427437,
      "num_tokens": 10790024.0,
      "step": 3230
    },
    {
      "entropy": 6.905918931961059,
      "epoch": 0.6311579358111403,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978090101234972,
      "loss": 6.8492,
      "mean_token_accuracy": 0.11303161382675171,
      "num_tokens": 10806280.0,
      "step": 3235
    },
    {
      "entropy": 6.772386360168457,
      "epoch": 0.6321334503950834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977992077109585,
      "loss": 6.729,
      "mean_token_accuracy": 0.12388993799686432,
      "num_tokens": 10822166.0,
      "step": 3240
    },
    {
      "entropy": 6.876743125915527,
      "epoch": 0.6331089649790265,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004977893835272754,
      "loss": 6.8846,
      "mean_token_accuracy": 0.1133842334151268,
      "num_tokens": 10839833.0,
      "step": 3245
    },
    {
      "entropy": 6.860595417022705,
      "epoch": 0.6340844795629694,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004977795375734075,
      "loss": 6.8477,
      "mean_token_accuracy": 0.11659579649567604,
      "num_tokens": 10856803.0,
      "step": 3250
    },
    {
      "entropy": 6.77113790512085,
      "epoch": 0.6350599941469125,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004977696698503174,
      "loss": 6.7105,
      "mean_token_accuracy": 0.12731380090117456,
      "num_tokens": 10874245.0,
      "step": 3255
    },
    {
      "entropy": 6.789052867889405,
      "epoch": 0.6360355087308556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497759780358969,
      "loss": 6.7091,
      "mean_token_accuracy": 0.12589445635676383,
      "num_tokens": 10890380.0,
      "step": 3260
    },
    {
      "entropy": 6.851020336151123,
      "epoch": 0.6370110233147985,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977498691003288,
      "loss": 6.7779,
      "mean_token_accuracy": 0.11898131072521209,
      "num_tokens": 10906927.0,
      "step": 3265
    },
    {
      "entropy": 6.852160549163818,
      "epoch": 0.6379865378987416,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004977399360753654,
      "loss": 6.8409,
      "mean_token_accuracy": 0.11594166457653046,
      "num_tokens": 10922922.0,
      "step": 3270
    },
    {
      "entropy": 6.833810043334961,
      "epoch": 0.6389620524826846,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004977299812850492,
      "loss": 6.8504,
      "mean_token_accuracy": 0.11509638279676437,
      "num_tokens": 10939607.0,
      "step": 3275
    },
    {
      "entropy": 6.772852230072021,
      "epoch": 0.6399375670666276,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977200047303532,
      "loss": 6.7669,
      "mean_token_accuracy": 0.12343585416674614,
      "num_tokens": 10956830.0,
      "step": 3280
    },
    {
      "entropy": 6.791714286804199,
      "epoch": 0.6409130816505707,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497710006412252,
      "loss": 6.7154,
      "mean_token_accuracy": 0.12172173485159873,
      "num_tokens": 10974364.0,
      "step": 3285
    },
    {
      "entropy": 6.812239027023315,
      "epoch": 0.6418885962345137,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004976999863317228,
      "loss": 6.795,
      "mean_token_accuracy": 0.12271926626563072,
      "num_tokens": 10990645.0,
      "step": 3290
    },
    {
      "entropy": 6.862871694564819,
      "epoch": 0.6428641108184567,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976899444897446,
      "loss": 6.7896,
      "mean_token_accuracy": 0.12550718784332277,
      "num_tokens": 11007952.0,
      "step": 3295
    },
    {
      "entropy": 6.634796285629273,
      "epoch": 0.6438396254023998,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976798808872987,
      "loss": 6.6884,
      "mean_token_accuracy": 0.12827200442552567,
      "num_tokens": 11024708.0,
      "step": 3300
    },
    {
      "entropy": 6.88479585647583,
      "epoch": 0.6448151399863428,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976697955253685,
      "loss": 6.8084,
      "mean_token_accuracy": 0.11946654766798019,
      "num_tokens": 11040447.0,
      "step": 3305
    },
    {
      "entropy": 6.841684055328369,
      "epoch": 0.6457906545702858,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976596884049395,
      "loss": 6.8462,
      "mean_token_accuracy": 0.11883102133870124,
      "num_tokens": 11057359.0,
      "step": 3310
    },
    {
      "entropy": 6.753314876556397,
      "epoch": 0.6467661691542289,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976495595269994,
      "loss": 6.6904,
      "mean_token_accuracy": 0.12631044834852218,
      "num_tokens": 11072997.0,
      "step": 3315
    },
    {
      "entropy": 6.865084218978882,
      "epoch": 0.6477416837381719,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976394088925377,
      "loss": 6.9069,
      "mean_token_accuracy": 0.11198164224624634,
      "num_tokens": 11089951.0,
      "step": 3320
    },
    {
      "entropy": 6.768247032165528,
      "epoch": 0.6487171983221149,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976292365025465,
      "loss": 6.6104,
      "mean_token_accuracy": 0.13177356645464897,
      "num_tokens": 11107929.0,
      "step": 3325
    },
    {
      "entropy": 6.632949781417847,
      "epoch": 0.6496927129060579,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976190423580199,
      "loss": 6.5842,
      "mean_token_accuracy": 0.13627510964870454,
      "num_tokens": 11125479.0,
      "step": 3330
    },
    {
      "entropy": 6.7710779190063475,
      "epoch": 0.650668227490001,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976088264599538,
      "loss": 6.8237,
      "mean_token_accuracy": 0.11492152661085128,
      "num_tokens": 11142316.0,
      "step": 3335
    },
    {
      "entropy": 6.904786825180054,
      "epoch": 0.651643742073944,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975985888093467,
      "loss": 6.8861,
      "mean_token_accuracy": 0.11398355588316918,
      "num_tokens": 11158664.0,
      "step": 3340
    },
    {
      "entropy": 6.828014135360718,
      "epoch": 0.652619256657887,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975883294071987,
      "loss": 6.8123,
      "mean_token_accuracy": 0.11526884734630585,
      "num_tokens": 11175732.0,
      "step": 3345
    },
    {
      "entropy": 6.904920244216919,
      "epoch": 0.6535947712418301,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975780482545125,
      "loss": 6.863,
      "mean_token_accuracy": 0.11198565661907196,
      "num_tokens": 11191921.0,
      "step": 3350
    },
    {
      "entropy": 6.714852714538575,
      "epoch": 0.6545702858257731,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975677453522928,
      "loss": 6.6556,
      "mean_token_accuracy": 0.13059833496809006,
      "num_tokens": 11208119.0,
      "step": 3355
    },
    {
      "entropy": 6.890081548690796,
      "epoch": 0.6555458004097161,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004975574207015462,
      "loss": 6.8144,
      "mean_token_accuracy": 0.12255159392952919,
      "num_tokens": 11224501.0,
      "step": 3360
    },
    {
      "entropy": 6.832798290252685,
      "epoch": 0.6565213149936592,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975470743032817,
      "loss": 6.856,
      "mean_token_accuracy": 0.12004402801394462,
      "num_tokens": 11240211.0,
      "step": 3365
    },
    {
      "entropy": 6.811522579193115,
      "epoch": 0.6574968295776021,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975367061585102,
      "loss": 6.7702,
      "mean_token_accuracy": 0.11939697787165642,
      "num_tokens": 11257921.0,
      "step": 3370
    },
    {
      "entropy": 6.829410076141357,
      "epoch": 0.6584723441615452,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975263162682449,
      "loss": 6.748,
      "mean_token_accuracy": 0.12618860676884652,
      "num_tokens": 11274526.0,
      "step": 3375
    },
    {
      "entropy": 6.8541905879974365,
      "epoch": 0.6594478587454883,
      "grad_norm": 0.875,
      "learning_rate": 0.000497515904633501,
      "loss": 6.7425,
      "mean_token_accuracy": 0.11885955855250359,
      "num_tokens": 11290883.0,
      "step": 3380
    },
    {
      "entropy": 6.7890349388122555,
      "epoch": 0.6604233733294312,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975054712552963,
      "loss": 6.7978,
      "mean_token_accuracy": 0.1210212156176567,
      "num_tokens": 11307406.0,
      "step": 3385
    },
    {
      "entropy": 6.711779451370239,
      "epoch": 0.6613988879133743,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974950161346496,
      "loss": 6.6278,
      "mean_token_accuracy": 0.12863511368632316,
      "num_tokens": 11322728.0,
      "step": 3390
    },
    {
      "entropy": 6.667343521118164,
      "epoch": 0.6623744024973174,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497484539272583,
      "loss": 6.618,
      "mean_token_accuracy": 0.13354130387306212,
      "num_tokens": 11338995.0,
      "step": 3395
    },
    {
      "entropy": 6.679088878631592,
      "epoch": 0.6633499170812603,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004974740406701203,
      "loss": 6.6092,
      "mean_token_accuracy": 0.1341503992676735,
      "num_tokens": 11356154.0,
      "step": 3400
    },
    {
      "entropy": 6.7625030994415285,
      "epoch": 0.6643254316652034,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974635203282873,
      "loss": 6.7591,
      "mean_token_accuracy": 0.12193894758820534,
      "num_tokens": 11372755.0,
      "step": 3405
    },
    {
      "entropy": 6.776936674118042,
      "epoch": 0.6653009462491464,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974529782481119,
      "loss": 6.7054,
      "mean_token_accuracy": 0.1295569770038128,
      "num_tokens": 11389424.0,
      "step": 3410
    },
    {
      "entropy": 6.7954916000366214,
      "epoch": 0.6662764608330894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974424144306241,
      "loss": 6.7893,
      "mean_token_accuracy": 0.11836361885070801,
      "num_tokens": 11406515.0,
      "step": 3415
    },
    {
      "entropy": 6.8078454494476315,
      "epoch": 0.6672519754170325,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004974318288768566,
      "loss": 6.7121,
      "mean_token_accuracy": 0.12833340615034103,
      "num_tokens": 11423195.0,
      "step": 3420
    },
    {
      "entropy": 6.751342535018921,
      "epoch": 0.6682274900009755,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974212215878435,
      "loss": 6.7953,
      "mean_token_accuracy": 0.11711516231298447,
      "num_tokens": 11440011.0,
      "step": 3425
    },
    {
      "entropy": 6.875479221343994,
      "epoch": 0.6692030045849185,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974105925646213,
      "loss": 6.7959,
      "mean_token_accuracy": 0.12360327243804932,
      "num_tokens": 11456131.0,
      "step": 3430
    },
    {
      "entropy": 6.715889835357666,
      "epoch": 0.6701785191688616,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973999418082287,
      "loss": 6.7232,
      "mean_token_accuracy": 0.12104954943060875,
      "num_tokens": 11471420.0,
      "step": 3435
    },
    {
      "entropy": 6.777550983428955,
      "epoch": 0.6711540337528046,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973892693197064,
      "loss": 6.6547,
      "mean_token_accuracy": 0.130117579549551,
      "num_tokens": 11488453.0,
      "step": 3440
    },
    {
      "entropy": 6.851372814178466,
      "epoch": 0.6721295483367477,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973785751000974,
      "loss": 6.8277,
      "mean_token_accuracy": 0.11581717506051063,
      "num_tokens": 11505144.0,
      "step": 3445
    },
    {
      "entropy": 6.802788496017456,
      "epoch": 0.6731050629206906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973678591504464,
      "loss": 6.7802,
      "mean_token_accuracy": 0.11815149560570717,
      "num_tokens": 11521249.0,
      "step": 3450
    },
    {
      "entropy": 6.72557954788208,
      "epoch": 0.6740805775046337,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497357121471801,
      "loss": 6.6989,
      "mean_token_accuracy": 0.1251370206475258,
      "num_tokens": 11537977.0,
      "step": 3455
    },
    {
      "entropy": 6.745234966278076,
      "epoch": 0.6750560920885768,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00049734636206521,
      "loss": 6.7736,
      "mean_token_accuracy": 0.11552431136369705,
      "num_tokens": 11556642.0,
      "step": 3460
    },
    {
      "entropy": 6.880000162124634,
      "epoch": 0.6760316066725197,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000497335580931725,
      "loss": 6.9284,
      "mean_token_accuracy": 0.11051849126815796,
      "num_tokens": 11574727.0,
      "step": 3465
    },
    {
      "entropy": 6.7403394222259525,
      "epoch": 0.6770071212564628,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973247780723995,
      "loss": 6.6181,
      "mean_token_accuracy": 0.12882431596517563,
      "num_tokens": 11592982.0,
      "step": 3470
    },
    {
      "entropy": 6.733096504211426,
      "epoch": 0.6779826358404059,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004973139534882891,
      "loss": 6.707,
      "mean_token_accuracy": 0.13039307966828345,
      "num_tokens": 11609250.0,
      "step": 3475
    },
    {
      "entropy": 6.81804895401001,
      "epoch": 0.6789581504243488,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004973031071804515,
      "loss": 6.7337,
      "mean_token_accuracy": 0.11859161406755447,
      "num_tokens": 11626269.0,
      "step": 3480
    },
    {
      "entropy": 6.769158840179443,
      "epoch": 0.6799336650082919,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004972922391499465,
      "loss": 6.7439,
      "mean_token_accuracy": 0.11937217488884926,
      "num_tokens": 11642721.0,
      "step": 3485
    },
    {
      "entropy": 6.699689960479736,
      "epoch": 0.680909179592235,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972813493978361,
      "loss": 6.6696,
      "mean_token_accuracy": 0.1250682942569256,
      "num_tokens": 11659729.0,
      "step": 3490
    },
    {
      "entropy": 6.814701986312866,
      "epoch": 0.6818846941761779,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972704379251845,
      "loss": 6.6954,
      "mean_token_accuracy": 0.11863087266683578,
      "num_tokens": 11675568.0,
      "step": 3495
    },
    {
      "entropy": 6.795450830459595,
      "epoch": 0.682860208760121,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000497259504733058,
      "loss": 6.8084,
      "mean_token_accuracy": 0.12470971420407295,
      "num_tokens": 11692827.0,
      "step": 3500
    },
    {
      "entropy": 6.843821716308594,
      "epoch": 0.683835723344064,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972485498225247,
      "loss": 6.7903,
      "mean_token_accuracy": 0.11543741300702096,
      "num_tokens": 11708730.0,
      "step": 3505
    },
    {
      "entropy": 6.774370622634888,
      "epoch": 0.684811237928007,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972375731946554,
      "loss": 6.746,
      "mean_token_accuracy": 0.11533987671136856,
      "num_tokens": 11725031.0,
      "step": 3510
    },
    {
      "entropy": 6.866035079956054,
      "epoch": 0.6857867525119501,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972265748505224,
      "loss": 6.8399,
      "mean_token_accuracy": 0.1149707593023777,
      "num_tokens": 11742419.0,
      "step": 3515
    },
    {
      "entropy": 6.743462133407593,
      "epoch": 0.686762267095893,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972155547912005,
      "loss": 6.7277,
      "mean_token_accuracy": 0.11798767820000648,
      "num_tokens": 11757474.0,
      "step": 3520
    },
    {
      "entropy": 6.9130033493042,
      "epoch": 0.6877377816798361,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004972045130177666,
      "loss": 6.8168,
      "mean_token_accuracy": 0.11203407943248749,
      "num_tokens": 11774467.0,
      "step": 3525
    },
    {
      "entropy": 6.723356008529663,
      "epoch": 0.6887132962637792,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004971934495312996,
      "loss": 6.6888,
      "mean_token_accuracy": 0.1262807197868824,
      "num_tokens": 11790597.0,
      "step": 3530
    },
    {
      "entropy": 6.799177026748657,
      "epoch": 0.6896888108477222,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971823643328807,
      "loss": 6.7049,
      "mean_token_accuracy": 0.13015776947140695,
      "num_tokens": 11806459.0,
      "step": 3535
    },
    {
      "entropy": 6.7149741649627686,
      "epoch": 0.6906643254316652,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497171257423593,
      "loss": 6.7593,
      "mean_token_accuracy": 0.12130732238292694,
      "num_tokens": 11823410.0,
      "step": 3540
    },
    {
      "entropy": 6.805893850326538,
      "epoch": 0.6916398400156082,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971601288045219,
      "loss": 6.7533,
      "mean_token_accuracy": 0.11716760471463203,
      "num_tokens": 11838684.0,
      "step": 3545
    },
    {
      "entropy": 6.788658952713012,
      "epoch": 0.6926153545995513,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971489784767547,
      "loss": 6.7449,
      "mean_token_accuracy": 0.12425903901457787,
      "num_tokens": 11854324.0,
      "step": 3550
    },
    {
      "entropy": 6.790124988555908,
      "epoch": 0.6935908691834943,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971378064413813,
      "loss": 6.7433,
      "mean_token_accuracy": 0.11819701790809631,
      "num_tokens": 11871992.0,
      "step": 3555
    },
    {
      "entropy": 6.847756290435791,
      "epoch": 0.6945663837674373,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971266126994929,
      "loss": 6.8451,
      "mean_token_accuracy": 0.11089114472270012,
      "num_tokens": 11889195.0,
      "step": 3560
    },
    {
      "entropy": 6.811157512664795,
      "epoch": 0.6955418983513804,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971153972521837,
      "loss": 6.8238,
      "mean_token_accuracy": 0.111813722550869,
      "num_tokens": 11906316.0,
      "step": 3565
    },
    {
      "entropy": 6.844501781463623,
      "epoch": 0.6965174129353234,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004971041601005495,
      "loss": 6.831,
      "mean_token_accuracy": 0.11643264293670655,
      "num_tokens": 11924097.0,
      "step": 3570
    },
    {
      "entropy": 6.88101978302002,
      "epoch": 0.6974929275192664,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004970929012456884,
      "loss": 6.8399,
      "mean_token_accuracy": 0.11809063032269478,
      "num_tokens": 11942344.0,
      "step": 3575
    },
    {
      "entropy": 6.768538522720337,
      "epoch": 0.6984684421032095,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004970816206887005,
      "loss": 6.6634,
      "mean_token_accuracy": 0.13062100857496262,
      "num_tokens": 11959045.0,
      "step": 3580
    },
    {
      "entropy": 6.771565723419189,
      "epoch": 0.6994439566871524,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970703184306881,
      "loss": 6.7075,
      "mean_token_accuracy": 0.1214524269104004,
      "num_tokens": 11976244.0,
      "step": 3585
    },
    {
      "entropy": 6.794751214981079,
      "epoch": 0.7004194712710955,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970589944727558,
      "loss": 6.7433,
      "mean_token_accuracy": 0.11434363499283791,
      "num_tokens": 11993644.0,
      "step": 3590
    },
    {
      "entropy": 6.706568765640259,
      "epoch": 0.7013949858550386,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970476488160098,
      "loss": 6.6535,
      "mean_token_accuracy": 0.1250632919371128,
      "num_tokens": 12010419.0,
      "step": 3595
    },
    {
      "entropy": 6.719916868209839,
      "epoch": 0.7023705004389815,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970362814615592,
      "loss": 6.6751,
      "mean_token_accuracy": 0.12672994658350945,
      "num_tokens": 12027199.0,
      "step": 3600
    },
    {
      "entropy": 6.63888955116272,
      "epoch": 0.7033460150229246,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970248924105144,
      "loss": 6.6744,
      "mean_token_accuracy": 0.12744994685053826,
      "num_tokens": 12044267.0,
      "step": 3605
    },
    {
      "entropy": 6.856422805786133,
      "epoch": 0.7043215296068677,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970134816639884,
      "loss": 6.7216,
      "mean_token_accuracy": 0.11745517179369927,
      "num_tokens": 12062838.0,
      "step": 3610
    },
    {
      "entropy": 6.789428997039795,
      "epoch": 0.7052970441908106,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970020492230963,
      "loss": 6.8131,
      "mean_token_accuracy": 0.1137891374528408,
      "num_tokens": 12079144.0,
      "step": 3615
    },
    {
      "entropy": 6.721288061141967,
      "epoch": 0.7062725587747537,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496990595088955,
      "loss": 6.6601,
      "mean_token_accuracy": 0.1287293776869774,
      "num_tokens": 12096477.0,
      "step": 3620
    },
    {
      "entropy": 6.771053838729858,
      "epoch": 0.7072480733586967,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004969791192626841,
      "loss": 6.7094,
      "mean_token_accuracy": 0.11969477012753486,
      "num_tokens": 12113307.0,
      "step": 3625
    },
    {
      "entropy": 6.788485383987426,
      "epoch": 0.7082235879426397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969676217454048,
      "loss": 6.8546,
      "mean_token_accuracy": 0.11760410889983178,
      "num_tokens": 12131814.0,
      "step": 3630
    },
    {
      "entropy": 6.8928601264953615,
      "epoch": 0.7091991025265828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969561025382405,
      "loss": 6.6821,
      "mean_token_accuracy": 0.11522524803876877,
      "num_tokens": 12147110.0,
      "step": 3635
    },
    {
      "entropy": 6.647023296356201,
      "epoch": 0.7101746171105258,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969445616423171,
      "loss": 6.6499,
      "mean_token_accuracy": 0.12332733720541,
      "num_tokens": 12162954.0,
      "step": 3640
    },
    {
      "entropy": 6.686932945251465,
      "epoch": 0.7111501316944688,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496932999058762,
      "loss": 6.561,
      "mean_token_accuracy": 0.12394139990210533,
      "num_tokens": 12181169.0,
      "step": 3645
    },
    {
      "entropy": 6.857557249069214,
      "epoch": 0.7121256462784119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969214147887053,
      "loss": 6.8856,
      "mean_token_accuracy": 0.11562231406569481,
      "num_tokens": 12198367.0,
      "step": 3650
    },
    {
      "entropy": 6.748447370529175,
      "epoch": 0.7131011608623549,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004969098088332789,
      "loss": 6.6892,
      "mean_token_accuracy": 0.12322335243225098,
      "num_tokens": 12215614.0,
      "step": 3655
    },
    {
      "entropy": 6.765234041213989,
      "epoch": 0.7140766754462979,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968981811936169,
      "loss": 6.6748,
      "mean_token_accuracy": 0.12507233023643494,
      "num_tokens": 12232123.0,
      "step": 3660
    },
    {
      "entropy": 6.738817119598389,
      "epoch": 0.7150521900302409,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968865318708555,
      "loss": 6.7698,
      "mean_token_accuracy": 0.1179160125553608,
      "num_tokens": 12248417.0,
      "step": 3665
    },
    {
      "entropy": 6.805019044876099,
      "epoch": 0.716027704614184,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968748608661328,
      "loss": 6.7203,
      "mean_token_accuracy": 0.11798951253294945,
      "num_tokens": 12264388.0,
      "step": 3670
    },
    {
      "entropy": 6.745490217208863,
      "epoch": 0.717003219198127,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968631681805897,
      "loss": 6.7799,
      "mean_token_accuracy": 0.11337770000100136,
      "num_tokens": 12282157.0,
      "step": 3675
    },
    {
      "entropy": 6.847876071929932,
      "epoch": 0.71797873378207,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968514538153685,
      "loss": 6.7769,
      "mean_token_accuracy": 0.1151386633515358,
      "num_tokens": 12298368.0,
      "step": 3680
    },
    {
      "entropy": 6.6725178241729735,
      "epoch": 0.7189542483660131,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496839717771614,
      "loss": 6.6934,
      "mean_token_accuracy": 0.13087327852845193,
      "num_tokens": 12314860.0,
      "step": 3685
    },
    {
      "entropy": 6.7805140018463135,
      "epoch": 0.7199297629499561,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004968279600504729,
      "loss": 6.7076,
      "mean_token_accuracy": 0.12170645818114281,
      "num_tokens": 12331424.0,
      "step": 3690
    },
    {
      "entropy": 6.811889123916626,
      "epoch": 0.7209052775338991,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496816180653094,
      "loss": 6.827,
      "mean_token_accuracy": 0.11624950245022773,
      "num_tokens": 12348358.0,
      "step": 3695
    },
    {
      "entropy": 6.808113050460816,
      "epoch": 0.7218807921178422,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968043795806285,
      "loss": 6.7202,
      "mean_token_accuracy": 0.12507479041814804,
      "num_tokens": 12364932.0,
      "step": 3700
    },
    {
      "entropy": 6.798572826385498,
      "epoch": 0.7228563067017852,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967925568342297,
      "loss": 6.7884,
      "mean_token_accuracy": 0.11744219660758973,
      "num_tokens": 12381852.0,
      "step": 3705
    },
    {
      "entropy": 6.786315536499023,
      "epoch": 0.7238318212857282,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967807124150527,
      "loss": 6.7779,
      "mean_token_accuracy": 0.11528379395604134,
      "num_tokens": 12397164.0,
      "step": 3710
    },
    {
      "entropy": 6.765109443664551,
      "epoch": 0.7248073358696713,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967688463242549,
      "loss": 6.7384,
      "mean_token_accuracy": 0.12121680527925491,
      "num_tokens": 12412520.0,
      "step": 3715
    },
    {
      "entropy": 6.812240552902222,
      "epoch": 0.7257828504536142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967569585629958,
      "loss": 6.7616,
      "mean_token_accuracy": 0.1241777092218399,
      "num_tokens": 12427039.0,
      "step": 3720
    },
    {
      "entropy": 6.799846935272217,
      "epoch": 0.7267583650375573,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967450491324371,
      "loss": 6.7291,
      "mean_token_accuracy": 0.12012900039553642,
      "num_tokens": 12441903.0,
      "step": 3725
    },
    {
      "entropy": 6.67380952835083,
      "epoch": 0.7277338796215004,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967331180337425,
      "loss": 6.6156,
      "mean_token_accuracy": 0.12848428562283515,
      "num_tokens": 12458883.0,
      "step": 3730
    },
    {
      "entropy": 6.775661325454712,
      "epoch": 0.7287093942054433,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967211652680779,
      "loss": 6.7964,
      "mean_token_accuracy": 0.11972467973828316,
      "num_tokens": 12475347.0,
      "step": 3735
    },
    {
      "entropy": 6.796820592880249,
      "epoch": 0.7296849087893864,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967091908366113,
      "loss": 6.794,
      "mean_token_accuracy": 0.1099328987300396,
      "num_tokens": 12490393.0,
      "step": 3740
    },
    {
      "entropy": 6.849007844924927,
      "epoch": 0.7306604233733295,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966971947405127,
      "loss": 6.766,
      "mean_token_accuracy": 0.12455734834074975,
      "num_tokens": 12505679.0,
      "step": 3745
    },
    {
      "entropy": 6.723217153549195,
      "epoch": 0.7316359379572724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966851769809543,
      "loss": 6.727,
      "mean_token_accuracy": 0.12228133603930473,
      "num_tokens": 12522316.0,
      "step": 3750
    },
    {
      "entropy": 6.838070583343506,
      "epoch": 0.7326114525412155,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004966731375591106,
      "loss": 6.8083,
      "mean_token_accuracy": 0.12010138928890228,
      "num_tokens": 12539638.0,
      "step": 3755
    },
    {
      "entropy": 6.718032646179199,
      "epoch": 0.7335869671251585,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966610764761579,
      "loss": 6.6682,
      "mean_token_accuracy": 0.11844480708241463,
      "num_tokens": 12556296.0,
      "step": 3760
    },
    {
      "entropy": 6.800854253768921,
      "epoch": 0.7345624817091015,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966489937332748,
      "loss": 6.7611,
      "mean_token_accuracy": 0.12091182842850685,
      "num_tokens": 12573248.0,
      "step": 3765
    },
    {
      "entropy": 6.644363212585449,
      "epoch": 0.7355379962930446,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966368893316421,
      "loss": 6.5635,
      "mean_token_accuracy": 0.13609384819865228,
      "num_tokens": 12590918.0,
      "step": 3770
    },
    {
      "entropy": 6.71870756149292,
      "epoch": 0.7365135108769876,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966247632724425,
      "loss": 6.6987,
      "mean_token_accuracy": 0.12216609790921211,
      "num_tokens": 12607037.0,
      "step": 3775
    },
    {
      "entropy": 6.648871898651123,
      "epoch": 0.7374890254609306,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004966126155568609,
      "loss": 6.6567,
      "mean_token_accuracy": 0.12721072807908057,
      "num_tokens": 12623934.0,
      "step": 3780
    },
    {
      "entropy": 6.7633765697479244,
      "epoch": 0.7384645400448737,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004966004461860843,
      "loss": 6.6447,
      "mean_token_accuracy": 0.13108034655451775,
      "num_tokens": 12640607.0,
      "step": 3785
    },
    {
      "entropy": 6.781022024154663,
      "epoch": 0.7394400546288167,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965882551613019,
      "loss": 6.6909,
      "mean_token_accuracy": 0.11714243143796921,
      "num_tokens": 12656671.0,
      "step": 3790
    },
    {
      "entropy": 6.758154678344726,
      "epoch": 0.7404155692127597,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496576042483705,
      "loss": 6.788,
      "mean_token_accuracy": 0.1090017519891262,
      "num_tokens": 12673271.0,
      "step": 3795
    },
    {
      "entropy": 6.618661880493164,
      "epoch": 0.7413910837967027,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004965638081544869,
      "loss": 6.6196,
      "mean_token_accuracy": 0.1349961668252945,
      "num_tokens": 12691005.0,
      "step": 3800
    },
    {
      "entropy": 6.822234678268432,
      "epoch": 0.7423665983806458,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965515521748432,
      "loss": 6.7399,
      "mean_token_accuracy": 0.12078917548060417,
      "num_tokens": 12706170.0,
      "step": 3805
    },
    {
      "entropy": 6.67201795578003,
      "epoch": 0.7433421129645889,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004965392745459714,
      "loss": 6.6351,
      "mean_token_accuracy": 0.1263989694416523,
      "num_tokens": 12722794.0,
      "step": 3810
    },
    {
      "entropy": 6.827039813995361,
      "epoch": 0.7443176275485318,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965269752690713,
      "loss": 6.7652,
      "mean_token_accuracy": 0.11326155588030815,
      "num_tokens": 12739676.0,
      "step": 3815
    },
    {
      "entropy": 6.764903402328491,
      "epoch": 0.7452931421324749,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965146543453446,
      "loss": 6.7859,
      "mean_token_accuracy": 0.11849705651402473,
      "num_tokens": 12756231.0,
      "step": 3820
    },
    {
      "entropy": 6.603110504150391,
      "epoch": 0.746268656716418,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965023117759954,
      "loss": 6.6085,
      "mean_token_accuracy": 0.13924905955791472,
      "num_tokens": 12771877.0,
      "step": 3825
    },
    {
      "entropy": 6.766062164306641,
      "epoch": 0.7472441713003609,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004964899475622299,
      "loss": 6.7226,
      "mean_token_accuracy": 0.11697940677404403,
      "num_tokens": 12790698.0,
      "step": 3830
    },
    {
      "entropy": 6.773321008682251,
      "epoch": 0.748219685884304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496477561705256,
      "loss": 6.7235,
      "mean_token_accuracy": 0.11808153614401817,
      "num_tokens": 12806160.0,
      "step": 3835
    },
    {
      "entropy": 6.81643533706665,
      "epoch": 0.749195200468247,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964651542062841,
      "loss": 6.7772,
      "mean_token_accuracy": 0.10997553169727325,
      "num_tokens": 12822382.0,
      "step": 3840
    },
    {
      "entropy": 6.814971494674682,
      "epoch": 0.75017071505219,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964527250665268,
      "loss": 6.7146,
      "mean_token_accuracy": 0.12292608171701432,
      "num_tokens": 12839421.0,
      "step": 3845
    },
    {
      "entropy": 6.659747457504272,
      "epoch": 0.7511462296361331,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004964402742871983,
      "loss": 6.5371,
      "mean_token_accuracy": 0.13697807043790816,
      "num_tokens": 12856609.0,
      "step": 3850
    },
    {
      "entropy": 6.783792448043823,
      "epoch": 0.752121744220076,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964278018695156,
      "loss": 6.7378,
      "mean_token_accuracy": 0.11603448390960694,
      "num_tokens": 12871659.0,
      "step": 3855
    },
    {
      "entropy": 6.708601188659668,
      "epoch": 0.7530972588040191,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964153078146971,
      "loss": 6.7277,
      "mean_token_accuracy": 0.12079250812530518,
      "num_tokens": 12889029.0,
      "step": 3860
    },
    {
      "entropy": 6.762743425369263,
      "epoch": 0.7540727733879622,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004964027921239639,
      "loss": 6.6984,
      "mean_token_accuracy": 0.1269061081111431,
      "num_tokens": 12905735.0,
      "step": 3865
    },
    {
      "entropy": 6.755882215499878,
      "epoch": 0.7550482879719052,
      "grad_norm": 1.078125,
      "learning_rate": 0.000496390254798539,
      "loss": 6.6822,
      "mean_token_accuracy": 0.11955450028181076,
      "num_tokens": 12921886.0,
      "step": 3870
    },
    {
      "entropy": 6.638718557357788,
      "epoch": 0.7560238025558482,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963776958396475,
      "loss": 6.6226,
      "mean_token_accuracy": 0.12904171869158745,
      "num_tokens": 12938658.0,
      "step": 3875
    },
    {
      "entropy": 6.789564228057861,
      "epoch": 0.7569993171397913,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004963651152485165,
      "loss": 6.7811,
      "mean_token_accuracy": 0.11869219988584519,
      "num_tokens": 12955788.0,
      "step": 3880
    },
    {
      "entropy": 6.671612977981567,
      "epoch": 0.7579748317237343,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963525130263753,
      "loss": 6.6018,
      "mean_token_accuracy": 0.1281113676726818,
      "num_tokens": 12972590.0,
      "step": 3885
    },
    {
      "entropy": 6.844368314743042,
      "epoch": 0.7589503463076773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963398891744557,
      "loss": 6.7342,
      "mean_token_accuracy": 0.11816896200180053,
      "num_tokens": 12989560.0,
      "step": 3890
    },
    {
      "entropy": 6.747540283203125,
      "epoch": 0.7599258608916203,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963272436939908,
      "loss": 6.6824,
      "mean_token_accuracy": 0.12627451717853547,
      "num_tokens": 13005584.0,
      "step": 3895
    },
    {
      "entropy": 6.711453866958618,
      "epoch": 0.7609013754755634,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963145765862166,
      "loss": 6.7416,
      "mean_token_accuracy": 0.11659909412264824,
      "num_tokens": 13021635.0,
      "step": 3900
    },
    {
      "entropy": 6.744056129455567,
      "epoch": 0.7618768900595064,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963018878523706,
      "loss": 6.6667,
      "mean_token_accuracy": 0.12659808695316316,
      "num_tokens": 13037838.0,
      "step": 3905
    },
    {
      "entropy": 6.639765501022339,
      "epoch": 0.7628524046434494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962891774936929,
      "loss": 6.6207,
      "mean_token_accuracy": 0.13163458183407784,
      "num_tokens": 13054596.0,
      "step": 3910
    },
    {
      "entropy": 6.713065338134766,
      "epoch": 0.7638279192273925,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004962764455114255,
      "loss": 6.6785,
      "mean_token_accuracy": 0.1258792035281658,
      "num_tokens": 13070462.0,
      "step": 3915
    },
    {
      "entropy": 6.695505666732788,
      "epoch": 0.7648034338113355,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004962636919068125,
      "loss": 6.6364,
      "mean_token_accuracy": 0.12667165994644164,
      "num_tokens": 13086686.0,
      "step": 3920
    },
    {
      "entropy": 6.746862268447876,
      "epoch": 0.7657789483952785,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962509166811001,
      "loss": 6.7676,
      "mean_token_accuracy": 0.11716409921646118,
      "num_tokens": 13102430.0,
      "step": 3925
    },
    {
      "entropy": 6.753959608078003,
      "epoch": 0.7667544629792216,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962381198355367,
      "loss": 6.6595,
      "mean_token_accuracy": 0.13087640479207038,
      "num_tokens": 13118944.0,
      "step": 3930
    },
    {
      "entropy": 6.763225317001343,
      "epoch": 0.7677299775631645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962253013713726,
      "loss": 6.6942,
      "mean_token_accuracy": 0.12063451632857322,
      "num_tokens": 13136393.0,
      "step": 3935
    },
    {
      "entropy": 6.7614531993865965,
      "epoch": 0.7687054921471076,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962124612898607,
      "loss": 6.7554,
      "mean_token_accuracy": 0.11848993003368377,
      "num_tokens": 13153834.0,
      "step": 3940
    },
    {
      "entropy": 6.693967723846436,
      "epoch": 0.7696810067310507,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961995995922553,
      "loss": 6.6101,
      "mean_token_accuracy": 0.12682552337646485,
      "num_tokens": 13170904.0,
      "step": 3945
    },
    {
      "entropy": 6.751755714416504,
      "epoch": 0.7706565213149936,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961867162798136,
      "loss": 6.7144,
      "mean_token_accuracy": 0.11927454695105552,
      "num_tokens": 13187274.0,
      "step": 3950
    },
    {
      "entropy": 6.654374742507935,
      "epoch": 0.7716320358989367,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961738113537943,
      "loss": 6.6829,
      "mean_token_accuracy": 0.13006692305207251,
      "num_tokens": 13203448.0,
      "step": 3955
    },
    {
      "entropy": 6.764598226547241,
      "epoch": 0.7726075504828798,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961608848154583,
      "loss": 6.6947,
      "mean_token_accuracy": 0.12008918300271035,
      "num_tokens": 13220460.0,
      "step": 3960
    },
    {
      "entropy": 6.719996595382691,
      "epoch": 0.7735830650668227,
      "grad_norm": 0.828125,
      "learning_rate": 0.000496147936666069,
      "loss": 6.6534,
      "mean_token_accuracy": 0.12478170469403267,
      "num_tokens": 13237502.0,
      "step": 3965
    },
    {
      "entropy": 6.73527979850769,
      "epoch": 0.7745585796507658,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961349669068913,
      "loss": 6.6837,
      "mean_token_accuracy": 0.12079736962914467,
      "num_tokens": 13254314.0,
      "step": 3970
    },
    {
      "entropy": 6.780675792694092,
      "epoch": 0.7755340942347088,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496121975539193,
      "loss": 6.7699,
      "mean_token_accuracy": 0.12089243456721306,
      "num_tokens": 13271737.0,
      "step": 3975
    },
    {
      "entropy": 6.816179466247559,
      "epoch": 0.7765096088186518,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004961089625642433,
      "loss": 6.714,
      "mean_token_accuracy": 0.12314795181155205,
      "num_tokens": 13288546.0,
      "step": 3980
    },
    {
      "entropy": 6.670663976669312,
      "epoch": 0.7774851234025949,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960959279833137,
      "loss": 6.6138,
      "mean_token_accuracy": 0.13126650527119638,
      "num_tokens": 13304309.0,
      "step": 3985
    },
    {
      "entropy": 6.708993864059448,
      "epoch": 0.7784606379865379,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960828717976781,
      "loss": 6.6812,
      "mean_token_accuracy": 0.12428322359919548,
      "num_tokens": 13319797.0,
      "step": 3990
    },
    {
      "entropy": 6.625681018829345,
      "epoch": 0.7794361525704809,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960697940086121,
      "loss": 6.5172,
      "mean_token_accuracy": 0.1399451769888401,
      "num_tokens": 13336040.0,
      "step": 3995
    },
    {
      "entropy": 6.804778671264648,
      "epoch": 0.780411667154424,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004960566946173939,
      "loss": 6.6908,
      "mean_token_accuracy": 0.12321212366223336,
      "num_tokens": 13352762.0,
      "step": 4000
    },
    {
      "entropy": 6.7093593120574955,
      "epoch": 0.781387181738367,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004960435736253034,
      "loss": 6.649,
      "mean_token_accuracy": 0.12115823030471802,
      "num_tokens": 13368577.0,
      "step": 4005
    },
    {
      "entropy": 6.755229091644287,
      "epoch": 0.78236269632231,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960304310336226,
      "loss": 6.7906,
      "mean_token_accuracy": 0.11535691171884536,
      "num_tokens": 13384170.0,
      "step": 4010
    },
    {
      "entropy": 6.6579162120819095,
      "epoch": 0.783338210906253,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960172668436359,
      "loss": 6.5882,
      "mean_token_accuracy": 0.13404057025909424,
      "num_tokens": 13400143.0,
      "step": 4015
    },
    {
      "entropy": 6.715267848968506,
      "epoch": 0.7843137254901961,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960040810566295,
      "loss": 6.7071,
      "mean_token_accuracy": 0.1215901404619217,
      "num_tokens": 13416553.0,
      "step": 4020
    },
    {
      "entropy": 6.6722043514251705,
      "epoch": 0.7852892400741391,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959908736738921,
      "loss": 6.6939,
      "mean_token_accuracy": 0.11827299371361732,
      "num_tokens": 13432694.0,
      "step": 4025
    },
    {
      "entropy": 6.760225915908814,
      "epoch": 0.7862647546580821,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959776446967141,
      "loss": 6.7112,
      "mean_token_accuracy": 0.11911704316735268,
      "num_tokens": 13450333.0,
      "step": 4030
    },
    {
      "entropy": 6.5740233898162845,
      "epoch": 0.7872402692420252,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959643941263882,
      "loss": 6.525,
      "mean_token_accuracy": 0.13143547624349594,
      "num_tokens": 13467150.0,
      "step": 4035
    },
    {
      "entropy": 6.703020906448364,
      "epoch": 0.7882157838259682,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959511219642092,
      "loss": 6.6807,
      "mean_token_accuracy": 0.12341507226228714,
      "num_tokens": 13484796.0,
      "step": 4040
    },
    {
      "entropy": 6.726278448104859,
      "epoch": 0.7891912984099112,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959378282114741,
      "loss": 6.6155,
      "mean_token_accuracy": 0.1287731021642685,
      "num_tokens": 13501533.0,
      "step": 4045
    },
    {
      "entropy": 6.759412050247192,
      "epoch": 0.7901668129938543,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959245128694819,
      "loss": 6.7837,
      "mean_token_accuracy": 0.11228923574090004,
      "num_tokens": 13518270.0,
      "step": 4050
    },
    {
      "entropy": 6.682141494750977,
      "epoch": 0.7911423275777972,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004959111759395337,
      "loss": 6.5606,
      "mean_token_accuracy": 0.1318668693304062,
      "num_tokens": 13535356.0,
      "step": 4055
    },
    {
      "entropy": 6.677602863311767,
      "epoch": 0.7921178421617403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958978174229326,
      "loss": 6.6077,
      "mean_token_accuracy": 0.12422248348593712,
      "num_tokens": 13550530.0,
      "step": 4060
    },
    {
      "entropy": 6.6310020923614506,
      "epoch": 0.7930933567456834,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958844373209843,
      "loss": 6.5823,
      "mean_token_accuracy": 0.13396794721484184,
      "num_tokens": 13569492.0,
      "step": 4065
    },
    {
      "entropy": 6.792881155014038,
      "epoch": 0.7940688713296263,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958710356349958,
      "loss": 6.7831,
      "mean_token_accuracy": 0.11514096707105637,
      "num_tokens": 13585096.0,
      "step": 4070
    },
    {
      "entropy": 6.785440587997437,
      "epoch": 0.7950443859135694,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000495857612366277,
      "loss": 6.7457,
      "mean_token_accuracy": 0.11839191764593124,
      "num_tokens": 13600667.0,
      "step": 4075
    },
    {
      "entropy": 6.775422096252441,
      "epoch": 0.7960199004975125,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004958441675161395,
      "loss": 6.7656,
      "mean_token_accuracy": 0.11530844420194626,
      "num_tokens": 13618212.0,
      "step": 4080
    },
    {
      "entropy": 6.762940120697022,
      "epoch": 0.7969954150814554,
      "grad_norm": 0.875,
      "learning_rate": 0.000495830701085897,
      "loss": 6.7017,
      "mean_token_accuracy": 0.12442354559898376,
      "num_tokens": 13634067.0,
      "step": 4085
    },
    {
      "entropy": 6.635333919525147,
      "epoch": 0.7979709296653985,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004958172130768655,
      "loss": 6.6376,
      "mean_token_accuracy": 0.1304635711014271,
      "num_tokens": 13651306.0,
      "step": 4090
    },
    {
      "entropy": 6.757288312911987,
      "epoch": 0.7989464442493416,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004958037034903628,
      "loss": 6.6698,
      "mean_token_accuracy": 0.11854898482561112,
      "num_tokens": 13667563.0,
      "step": 4095
    },
    {
      "entropy": 6.592153167724609,
      "epoch": 0.7999219588332845,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004957901723277093,
      "loss": 6.558,
      "mean_token_accuracy": 0.12907954007387162,
      "num_tokens": 13684857.0,
      "step": 4100
    },
    {
      "entropy": 6.814383029937744,
      "epoch": 0.8008974734172276,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004957766195902271,
      "loss": 6.7663,
      "mean_token_accuracy": 0.11462492272257804,
      "num_tokens": 13700356.0,
      "step": 4105
    },
    {
      "entropy": 6.682872247695923,
      "epoch": 0.8018729880011706,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957630452792404,
      "loss": 6.6146,
      "mean_token_accuracy": 0.130269917845726,
      "num_tokens": 13717146.0,
      "step": 4110
    },
    {
      "entropy": 6.685856723785401,
      "epoch": 0.8028485025851136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957494493960757,
      "loss": 6.6952,
      "mean_token_accuracy": 0.11699162647128106,
      "num_tokens": 13732520.0,
      "step": 4115
    },
    {
      "entropy": 6.8340826511383055,
      "epoch": 0.8038240171690567,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957358319420617,
      "loss": 6.7973,
      "mean_token_accuracy": 0.11386164277791977,
      "num_tokens": 13749135.0,
      "step": 4120
    },
    {
      "entropy": 6.6975304126739506,
      "epoch": 0.8047995317529997,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004957221929185288,
      "loss": 6.5342,
      "mean_token_accuracy": 0.13566584140062332,
      "num_tokens": 13766218.0,
      "step": 4125
    },
    {
      "entropy": 6.727292108535766,
      "epoch": 0.8057750463369427,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004957085323268099,
      "loss": 6.7356,
      "mean_token_accuracy": 0.12132264971733094,
      "num_tokens": 13784421.0,
      "step": 4130
    },
    {
      "entropy": 6.671951866149902,
      "epoch": 0.8067505609208858,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956948501682399,
      "loss": 6.6207,
      "mean_token_accuracy": 0.1273565337061882,
      "num_tokens": 13800054.0,
      "step": 4135
    },
    {
      "entropy": 6.727924585342407,
      "epoch": 0.8077260755048288,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956811464441556,
      "loss": 6.6394,
      "mean_token_accuracy": 0.12454090267419815,
      "num_tokens": 13816433.0,
      "step": 4140
    },
    {
      "entropy": 6.733154535293579,
      "epoch": 0.8087015900887718,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956674211558963,
      "loss": 6.7588,
      "mean_token_accuracy": 0.11669637858867646,
      "num_tokens": 13832441.0,
      "step": 4145
    },
    {
      "entropy": 6.761666011810303,
      "epoch": 0.8096771046727148,
      "grad_norm": 0.796875,
      "learning_rate": 0.000495653674304803,
      "loss": 6.6919,
      "mean_token_accuracy": 0.1238759808242321,
      "num_tokens": 13849012.0,
      "step": 4150
    },
    {
      "entropy": 6.807874488830566,
      "epoch": 0.8106526192566579,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956399058922192,
      "loss": 6.7464,
      "mean_token_accuracy": 0.11355878338217736,
      "num_tokens": 13866486.0,
      "step": 4155
    },
    {
      "entropy": 6.689435148239136,
      "epoch": 0.811628133840601,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049562611591949,
      "loss": 6.6078,
      "mean_token_accuracy": 0.12563450559973716,
      "num_tokens": 13882453.0,
      "step": 4160
    },
    {
      "entropy": 6.723975133895874,
      "epoch": 0.8126036484245439,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004956123043879632,
      "loss": 6.6839,
      "mean_token_accuracy": 0.12428088188171386,
      "num_tokens": 13898090.0,
      "step": 4165
    },
    {
      "entropy": 6.700619554519653,
      "epoch": 0.813579163008487,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955984712989882,
      "loss": 6.6379,
      "mean_token_accuracy": 0.12283719629049301,
      "num_tokens": 13914818.0,
      "step": 4170
    },
    {
      "entropy": 6.788127422332764,
      "epoch": 0.81455467759243,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955846166539169,
      "loss": 6.7339,
      "mean_token_accuracy": 0.116616839915514,
      "num_tokens": 13930690.0,
      "step": 4175
    },
    {
      "entropy": 6.7581048011779785,
      "epoch": 0.815530192176373,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495570740454103,
      "loss": 6.734,
      "mean_token_accuracy": 0.12181395515799523,
      "num_tokens": 13946128.0,
      "step": 4180
    },
    {
      "entropy": 6.705576753616333,
      "epoch": 0.8165057067603161,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955568427009024,
      "loss": 6.6549,
      "mean_token_accuracy": 0.12648576870560646,
      "num_tokens": 13961766.0,
      "step": 4185
    },
    {
      "entropy": 6.638235330581665,
      "epoch": 0.817481221344259,
      "grad_norm": 0.875,
      "learning_rate": 0.0004955429233956733,
      "loss": 6.617,
      "mean_token_accuracy": 0.12915298193693162,
      "num_tokens": 13978632.0,
      "step": 4190
    },
    {
      "entropy": 6.777352380752563,
      "epoch": 0.8184567359282021,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004955289825397757,
      "loss": 6.7014,
      "mean_token_accuracy": 0.11758175566792488,
      "num_tokens": 13995297.0,
      "step": 4195
    },
    {
      "entropy": 6.736868667602539,
      "epoch": 0.8194322505121452,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495515020134572,
      "loss": 6.7148,
      "mean_token_accuracy": 0.12223832905292512,
      "num_tokens": 14010439.0,
      "step": 4200
    },
    {
      "entropy": 6.729992437362671,
      "epoch": 0.8204077650960881,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955010361814264,
      "loss": 6.7032,
      "mean_token_accuracy": 0.12575263455510138,
      "num_tokens": 14028538.0,
      "step": 4205
    },
    {
      "entropy": 6.791779327392578,
      "epoch": 0.8213832796800312,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954870306817054,
      "loss": 6.7353,
      "mean_token_accuracy": 0.1165425144135952,
      "num_tokens": 14045290.0,
      "step": 4210
    },
    {
      "entropy": 6.613925790786743,
      "epoch": 0.8223587942639743,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004954730036367776,
      "loss": 6.6438,
      "mean_token_accuracy": 0.12751975059509277,
      "num_tokens": 14062247.0,
      "step": 4215
    },
    {
      "entropy": 6.710271549224854,
      "epoch": 0.8233343088479173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004954589550480137,
      "loss": 6.6756,
      "mean_token_accuracy": 0.12357802391052246,
      "num_tokens": 14078599.0,
      "step": 4220
    },
    {
      "entropy": 6.686732912063599,
      "epoch": 0.8243098234318603,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004954448849167864,
      "loss": 6.6737,
      "mean_token_accuracy": 0.12241702526807785,
      "num_tokens": 14094686.0,
      "step": 4225
    },
    {
      "entropy": 6.7201306343078615,
      "epoch": 0.8252853380158033,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954307932444707,
      "loss": 6.6754,
      "mean_token_accuracy": 0.11928008943796158,
      "num_tokens": 14110255.0,
      "step": 4230
    },
    {
      "entropy": 6.737729406356811,
      "epoch": 0.8262608525997464,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954166800324434,
      "loss": 6.6859,
      "mean_token_accuracy": 0.11992513462901115,
      "num_tokens": 14127160.0,
      "step": 4235
    },
    {
      "entropy": 6.624211025238037,
      "epoch": 0.8272363671836894,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004954025452820837,
      "loss": 6.5974,
      "mean_token_accuracy": 0.1269684202969074,
      "num_tokens": 14144704.0,
      "step": 4240
    },
    {
      "entropy": 6.758190441131592,
      "epoch": 0.8282118817676324,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495388388994773,
      "loss": 6.68,
      "mean_token_accuracy": 0.12156076654791832,
      "num_tokens": 14162125.0,
      "step": 4245
    },
    {
      "entropy": 6.691038608551025,
      "epoch": 0.8291873963515755,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004953742111718942,
      "loss": 6.5941,
      "mean_token_accuracy": 0.12432070896029472,
      "num_tokens": 14179735.0,
      "step": 4250
    },
    {
      "entropy": 6.646215438842773,
      "epoch": 0.8301629109355185,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000495360011814833,
      "loss": 6.7886,
      "mean_token_accuracy": 0.11768869012594223,
      "num_tokens": 14196964.0,
      "step": 4255
    },
    {
      "entropy": 6.709884071350098,
      "epoch": 0.8311384255194615,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953457909249768,
      "loss": 6.6084,
      "mean_token_accuracy": 0.126287105679512,
      "num_tokens": 14212802.0,
      "step": 4260
    },
    {
      "entropy": 6.685347414016723,
      "epoch": 0.8321139401034046,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953315485037152,
      "loss": 6.6196,
      "mean_token_accuracy": 0.1255659580230713,
      "num_tokens": 14229183.0,
      "step": 4265
    },
    {
      "entropy": 6.733370161056518,
      "epoch": 0.8330894546873476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953172845524401,
      "loss": 6.64,
      "mean_token_accuracy": 0.11907010227441787,
      "num_tokens": 14245859.0,
      "step": 4270
    },
    {
      "entropy": 6.738180160522461,
      "epoch": 0.8340649692712906,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004953029990725452,
      "loss": 6.7143,
      "mean_token_accuracy": 0.12743450328707695,
      "num_tokens": 14261689.0,
      "step": 4275
    },
    {
      "entropy": 6.651930332183838,
      "epoch": 0.8350404838552337,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004952886920654262,
      "loss": 6.5454,
      "mean_token_accuracy": 0.13072821125388145,
      "num_tokens": 14280638.0,
      "step": 4280
    },
    {
      "entropy": 6.6360331058502195,
      "epoch": 0.8360159984391766,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952743635324813,
      "loss": 6.5631,
      "mean_token_accuracy": 0.1354525551199913,
      "num_tokens": 14296595.0,
      "step": 4285
    },
    {
      "entropy": 6.672172546386719,
      "epoch": 0.8369915130231197,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004952600134751109,
      "loss": 6.7387,
      "mean_token_accuracy": 0.121571946144104,
      "num_tokens": 14312384.0,
      "step": 4290
    },
    {
      "entropy": 6.742876815795898,
      "epoch": 0.8379670276070628,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004952456418947168,
      "loss": 6.6543,
      "mean_token_accuracy": 0.12362372428178788,
      "num_tokens": 14327852.0,
      "step": 4295
    },
    {
      "entropy": 6.65139365196228,
      "epoch": 0.8389425421910057,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952312487927037,
      "loss": 6.6669,
      "mean_token_accuracy": 0.11880057752132415,
      "num_tokens": 14343666.0,
      "step": 4300
    },
    {
      "entropy": 6.710541248321533,
      "epoch": 0.8399180567749488,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952168341704778,
      "loss": 6.6282,
      "mean_token_accuracy": 0.1322393834590912,
      "num_tokens": 14359819.0,
      "step": 4305
    },
    {
      "entropy": 6.787149858474732,
      "epoch": 0.8408935713588919,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004952023980294478,
      "loss": 6.7544,
      "mean_token_accuracy": 0.11651138961315155,
      "num_tokens": 14377762.0,
      "step": 4310
    },
    {
      "entropy": 6.663956117630005,
      "epoch": 0.8418690859428348,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951879403710242,
      "loss": 6.6018,
      "mean_token_accuracy": 0.12272122204303741,
      "num_tokens": 14393995.0,
      "step": 4315
    },
    {
      "entropy": 6.766266107559204,
      "epoch": 0.8428446005267779,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004951734611966199,
      "loss": 6.6509,
      "mean_token_accuracy": 0.12611735388636588,
      "num_tokens": 14411959.0,
      "step": 4320
    },
    {
      "entropy": 6.662760257720947,
      "epoch": 0.8438201151107209,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951589605076495,
      "loss": 6.7319,
      "mean_token_accuracy": 0.11751077845692634,
      "num_tokens": 14427659.0,
      "step": 4325
    },
    {
      "entropy": 6.535267877578735,
      "epoch": 0.8447956296946639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951444383055304,
      "loss": 6.4745,
      "mean_token_accuracy": 0.14167806580662728,
      "num_tokens": 14443887.0,
      "step": 4330
    },
    {
      "entropy": 6.740965747833252,
      "epoch": 0.845771144278607,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951298945916812,
      "loss": 6.6686,
      "mean_token_accuracy": 0.12335980385541916,
      "num_tokens": 14461907.0,
      "step": 4335
    },
    {
      "entropy": 6.637248945236206,
      "epoch": 0.84674665886255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004951153293675235,
      "loss": 6.6205,
      "mean_token_accuracy": 0.12218183130025864,
      "num_tokens": 14476912.0,
      "step": 4340
    },
    {
      "entropy": 6.65047025680542,
      "epoch": 0.847722173446493,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004951007426344801,
      "loss": 6.5317,
      "mean_token_accuracy": 0.12912099584937095,
      "num_tokens": 14493499.0,
      "step": 4345
    },
    {
      "entropy": 6.731754732131958,
      "epoch": 0.8486976880304361,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950861343939767,
      "loss": 6.6855,
      "mean_token_accuracy": 0.12522788047790528,
      "num_tokens": 14511266.0,
      "step": 4350
    },
    {
      "entropy": 6.70497465133667,
      "epoch": 0.8496732026143791,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950715046474405,
      "loss": 6.6922,
      "mean_token_accuracy": 0.12075289487838745,
      "num_tokens": 14528930.0,
      "step": 4355
    },
    {
      "entropy": 6.712876987457276,
      "epoch": 0.8506487171983221,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950568533963013,
      "loss": 6.6492,
      "mean_token_accuracy": 0.13299945890903472,
      "num_tokens": 14546637.0,
      "step": 4360
    },
    {
      "entropy": 6.63966965675354,
      "epoch": 0.8516242317822651,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950421806419907,
      "loss": 6.6638,
      "mean_token_accuracy": 0.12255307585000992,
      "num_tokens": 14563332.0,
      "step": 4365
    },
    {
      "entropy": 6.613657522201538,
      "epoch": 0.8525997463662082,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950274863859425,
      "loss": 6.51,
      "mean_token_accuracy": 0.13205826804041862,
      "num_tokens": 14579806.0,
      "step": 4370
    },
    {
      "entropy": 6.743616199493408,
      "epoch": 0.8535752609501512,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950127706295923,
      "loss": 6.729,
      "mean_token_accuracy": 0.11711006313562393,
      "num_tokens": 14597129.0,
      "step": 4375
    },
    {
      "entropy": 6.57832407951355,
      "epoch": 0.8545507755340942,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004949980333743786,
      "loss": 6.4581,
      "mean_token_accuracy": 0.13882744684815407,
      "num_tokens": 14615000.0,
      "step": 4380
    },
    {
      "entropy": 6.6925966262817385,
      "epoch": 0.8555262901180373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494983274621741,
      "loss": 6.6304,
      "mean_token_accuracy": 0.13345332443714142,
      "num_tokens": 14631338.0,
      "step": 4385
    },
    {
      "entropy": 6.66128101348877,
      "epoch": 0.8565018047019803,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949684943731219,
      "loss": 6.6342,
      "mean_token_accuracy": 0.1260090857744217,
      "num_tokens": 14647886.0,
      "step": 4390
    },
    {
      "entropy": 6.7071614265441895,
      "epoch": 0.8574773192859233,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004949536926299654,
      "loss": 6.6093,
      "mean_token_accuracy": 0.12621169164776802,
      "num_tokens": 14665054.0,
      "step": 4395
    },
    {
      "entropy": 6.631485080718994,
      "epoch": 0.8584528338698664,
      "grad_norm": 1.0,
      "learning_rate": 0.000494938869393718,
      "loss": 6.6356,
      "mean_token_accuracy": 0.12456467002630234,
      "num_tokens": 14679413.0,
      "step": 4400
    },
    {
      "entropy": 6.643982172012329,
      "epoch": 0.8594283484538093,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949240246658283,
      "loss": 6.5868,
      "mean_token_accuracy": 0.12808526530861855,
      "num_tokens": 14696562.0,
      "step": 4405
    },
    {
      "entropy": 6.601659822463989,
      "epoch": 0.8604038630377524,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004949091584477465,
      "loss": 6.4997,
      "mean_token_accuracy": 0.13219641298055648,
      "num_tokens": 14713782.0,
      "step": 4410
    },
    {
      "entropy": 6.748961973190307,
      "epoch": 0.8613793776216955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948942707409257,
      "loss": 6.7479,
      "mean_token_accuracy": 0.11928285062313079,
      "num_tokens": 14730139.0,
      "step": 4415
    },
    {
      "entropy": 6.718807744979858,
      "epoch": 0.8623548922056384,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948793615468205,
      "loss": 6.6725,
      "mean_token_accuracy": 0.11546050384640694,
      "num_tokens": 14746830.0,
      "step": 4420
    },
    {
      "entropy": 6.780854558944702,
      "epoch": 0.8633304067895815,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948644308668877,
      "loss": 6.7089,
      "mean_token_accuracy": 0.11925071254372596,
      "num_tokens": 14762410.0,
      "step": 4425
    },
    {
      "entropy": 6.690085983276367,
      "epoch": 0.8643059213735246,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948494787025863,
      "loss": 6.602,
      "mean_token_accuracy": 0.1172464057803154,
      "num_tokens": 14778283.0,
      "step": 4430
    },
    {
      "entropy": 6.590355014801025,
      "epoch": 0.8652814359574675,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948345050553774,
      "loss": 6.5698,
      "mean_token_accuracy": 0.1266999714076519,
      "num_tokens": 14797005.0,
      "step": 4435
    },
    {
      "entropy": 6.742372035980225,
      "epoch": 0.8662569505414106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004948195099267242,
      "loss": 6.6951,
      "mean_token_accuracy": 0.11459987610578537,
      "num_tokens": 14814548.0,
      "step": 4440
    },
    {
      "entropy": 6.718578815460205,
      "epoch": 0.8672324651253536,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004948044933180919,
      "loss": 6.6252,
      "mean_token_accuracy": 0.129029793292284,
      "num_tokens": 14830630.0,
      "step": 4445
    },
    {
      "entropy": 6.67466630935669,
      "epoch": 0.8682079797092966,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494789455230948,
      "loss": 6.6507,
      "mean_token_accuracy": 0.1204644188284874,
      "num_tokens": 14846247.0,
      "step": 4450
    },
    {
      "entropy": 6.7684645652771,
      "epoch": 0.8691834942932397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947743956667617,
      "loss": 6.6806,
      "mean_token_accuracy": 0.12118270620703697,
      "num_tokens": 14864480.0,
      "step": 4455
    },
    {
      "entropy": 6.722702980041504,
      "epoch": 0.8701590088771827,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947593146270049,
      "loss": 6.7171,
      "mean_token_accuracy": 0.12335770800709725,
      "num_tokens": 14881382.0,
      "step": 4460
    },
    {
      "entropy": 6.7464264869689945,
      "epoch": 0.8711345234611257,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494744212113151,
      "loss": 6.6542,
      "mean_token_accuracy": 0.12049593850970268,
      "num_tokens": 14897546.0,
      "step": 4465
    },
    {
      "entropy": 6.628796148300171,
      "epoch": 0.8721100380450688,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947290881266759,
      "loss": 6.6211,
      "mean_token_accuracy": 0.12305195406079292,
      "num_tokens": 14913995.0,
      "step": 4470
    },
    {
      "entropy": 6.712023448944092,
      "epoch": 0.8730855526290118,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947139426690576,
      "loss": 6.6815,
      "mean_token_accuracy": 0.11477718204259872,
      "num_tokens": 14931097.0,
      "step": 4475
    },
    {
      "entropy": 6.605894374847412,
      "epoch": 0.8740610672129548,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004946987757417757,
      "loss": 6.5654,
      "mean_token_accuracy": 0.12948881983757018,
      "num_tokens": 14948010.0,
      "step": 4480
    },
    {
      "entropy": 6.540175247192383,
      "epoch": 0.8750365817968979,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946835873463126,
      "loss": 6.5257,
      "mean_token_accuracy": 0.13895187005400658,
      "num_tokens": 14966940.0,
      "step": 4485
    },
    {
      "entropy": 6.790498352050781,
      "epoch": 0.8760120963808409,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004946683774841522,
      "loss": 6.7268,
      "mean_token_accuracy": 0.11760951206088066,
      "num_tokens": 14983482.0,
      "step": 4490
    },
    {
      "entropy": 6.527669906616211,
      "epoch": 0.876987610964784,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494653146156781,
      "loss": 6.5628,
      "mean_token_accuracy": 0.12887031212449074,
      "num_tokens": 15000709.0,
      "step": 4495
    },
    {
      "entropy": 6.679251337051392,
      "epoch": 0.8779631255487269,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946378933656871,
      "loss": 6.5501,
      "mean_token_accuracy": 0.12868396788835526,
      "num_tokens": 15016584.0,
      "step": 4500
    },
    {
      "entropy": 6.6421661376953125,
      "epoch": 0.87893864013267,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000494622619112361,
      "loss": 6.7092,
      "mean_token_accuracy": 0.12514919564127922,
      "num_tokens": 15033598.0,
      "step": 4505
    },
    {
      "entropy": 6.761171817779541,
      "epoch": 0.879914154716613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946073233982954,
      "loss": 6.6867,
      "mean_token_accuracy": 0.12088168188929557,
      "num_tokens": 15049842.0,
      "step": 4510
    },
    {
      "entropy": 6.599032783508301,
      "epoch": 0.880889669300556,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004945920062249849,
      "loss": 6.4252,
      "mean_token_accuracy": 0.13500453308224677,
      "num_tokens": 15066278.0,
      "step": 4515
    },
    {
      "entropy": 6.666645574569702,
      "epoch": 0.8818651838844991,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945766675939261,
      "loss": 6.651,
      "mean_token_accuracy": 0.1295979917049408,
      "num_tokens": 15083316.0,
      "step": 4520
    },
    {
      "entropy": 6.6491804122924805,
      "epoch": 0.8828406984684422,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945613075066181,
      "loss": 6.6752,
      "mean_token_accuracy": 0.12004767507314681,
      "num_tokens": 15103135.0,
      "step": 4525
    },
    {
      "entropy": 6.6046923160552975,
      "epoch": 0.8838162130523851,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004945459259645614,
      "loss": 6.5078,
      "mean_token_accuracy": 0.1430646613240242,
      "num_tokens": 15122441.0,
      "step": 4530
    },
    {
      "entropy": 6.763333415985107,
      "epoch": 0.8847917276363282,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945305229692594,
      "loss": 6.6417,
      "mean_token_accuracy": 0.12176171094179153,
      "num_tokens": 15139526.0,
      "step": 4535
    },
    {
      "entropy": 6.625238466262817,
      "epoch": 0.8857672422202711,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004945150985222171,
      "loss": 6.6186,
      "mean_token_accuracy": 0.1207355797290802,
      "num_tokens": 15156139.0,
      "step": 4540
    },
    {
      "entropy": 6.620577669143676,
      "epoch": 0.8867427568042142,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944996526249419,
      "loss": 6.5476,
      "mean_token_accuracy": 0.1338853120803833,
      "num_tokens": 15172894.0,
      "step": 4545
    },
    {
      "entropy": 6.642191648483276,
      "epoch": 0.8877182713881573,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944841852789427,
      "loss": 6.6542,
      "mean_token_accuracy": 0.12280394062399864,
      "num_tokens": 15190900.0,
      "step": 4550
    },
    {
      "entropy": 6.744839668273926,
      "epoch": 0.8886937859721002,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944686964857314,
      "loss": 6.6158,
      "mean_token_accuracy": 0.11929288581013679,
      "num_tokens": 15208018.0,
      "step": 4555
    },
    {
      "entropy": 6.662910795211792,
      "epoch": 0.8896693005560433,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004944531862468211,
      "loss": 6.6294,
      "mean_token_accuracy": 0.1297538161277771,
      "num_tokens": 15222781.0,
      "step": 4560
    },
    {
      "entropy": 6.646192836761474,
      "epoch": 0.8906448151399864,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944376545637277,
      "loss": 6.6664,
      "mean_token_accuracy": 0.12433210983872414,
      "num_tokens": 15241405.0,
      "step": 4565
    },
    {
      "entropy": 6.776825666427612,
      "epoch": 0.8916203297239294,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944221014379688,
      "loss": 6.6821,
      "mean_token_accuracy": 0.12513951063156128,
      "num_tokens": 15258340.0,
      "step": 4570
    },
    {
      "entropy": 6.646236419677734,
      "epoch": 0.8925958443078724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004944065268710641,
      "loss": 6.6284,
      "mean_token_accuracy": 0.12477434575557708,
      "num_tokens": 15274655.0,
      "step": 4575
    },
    {
      "entropy": 6.715242528915406,
      "epoch": 0.8935713588918154,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943909308645356,
      "loss": 6.668,
      "mean_token_accuracy": 0.11738622635602951,
      "num_tokens": 15291197.0,
      "step": 4580
    },
    {
      "entropy": 6.716484832763672,
      "epoch": 0.8945468734757585,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004943753134199072,
      "loss": 6.6548,
      "mean_token_accuracy": 0.12307366132736205,
      "num_tokens": 15306879.0,
      "step": 4585
    },
    {
      "entropy": 6.699360990524292,
      "epoch": 0.8955223880597015,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943596745387052,
      "loss": 6.6767,
      "mean_token_accuracy": 0.12045475021004677,
      "num_tokens": 15322880.0,
      "step": 4590
    },
    {
      "entropy": 6.6272620677948,
      "epoch": 0.8964979026436445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943440142224575,
      "loss": 6.5909,
      "mean_token_accuracy": 0.12716587856411934,
      "num_tokens": 15338282.0,
      "step": 4595
    },
    {
      "entropy": 6.708596086502075,
      "epoch": 0.8974734172275876,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943283324726944,
      "loss": 6.6175,
      "mean_token_accuracy": 0.12717477977275848,
      "num_tokens": 15354507.0,
      "step": 4600
    },
    {
      "entropy": 6.688210201263428,
      "epoch": 0.8984489318115306,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943126292909485,
      "loss": 6.639,
      "mean_token_accuracy": 0.12103248238563538,
      "num_tokens": 15373204.0,
      "step": 4605
    },
    {
      "entropy": 6.611228513717651,
      "epoch": 0.8994244463954736,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004942969046787541,
      "loss": 6.4861,
      "mean_token_accuracy": 0.13544998168945313,
      "num_tokens": 15390743.0,
      "step": 4610
    },
    {
      "entropy": 6.654214477539062,
      "epoch": 0.9003999609794167,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004942811586376476,
      "loss": 6.6064,
      "mean_token_accuracy": 0.11817645579576493,
      "num_tokens": 15408565.0,
      "step": 4615
    },
    {
      "entropy": 6.598478651046753,
      "epoch": 0.9013754755633596,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942653911691679,
      "loss": 6.5608,
      "mean_token_accuracy": 0.1310117118060589,
      "num_tokens": 15425747.0,
      "step": 4620
    },
    {
      "entropy": 6.607557106018066,
      "epoch": 0.9023509901473027,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004942496022748557,
      "loss": 6.5764,
      "mean_token_accuracy": 0.12445183992385864,
      "num_tokens": 15441882.0,
      "step": 4625
    },
    {
      "entropy": 6.655602169036865,
      "epoch": 0.9033265047312458,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004942337919562539,
      "loss": 6.5445,
      "mean_token_accuracy": 0.12399753928184509,
      "num_tokens": 15458003.0,
      "step": 4630
    },
    {
      "entropy": 6.4853919506072994,
      "epoch": 0.9043020193151887,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942179602149071,
      "loss": 6.4634,
      "mean_token_accuracy": 0.13799505010247232,
      "num_tokens": 15473798.0,
      "step": 4635
    },
    {
      "entropy": 6.573524332046508,
      "epoch": 0.9052775338991318,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942021070523626,
      "loss": 6.4793,
      "mean_token_accuracy": 0.13835482373833657,
      "num_tokens": 15490620.0,
      "step": 4640
    },
    {
      "entropy": 6.597295808792114,
      "epoch": 0.9062530484830749,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941862324701695,
      "loss": 6.5021,
      "mean_token_accuracy": 0.13603186160326003,
      "num_tokens": 15506777.0,
      "step": 4645
    },
    {
      "entropy": 6.673757934570313,
      "epoch": 0.9072285630670178,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004941703364698789,
      "loss": 6.7191,
      "mean_token_accuracy": 0.11880931332707405,
      "num_tokens": 15523416.0,
      "step": 4650
    },
    {
      "entropy": 6.682160329818726,
      "epoch": 0.9082040776509609,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941544190530442,
      "loss": 6.4728,
      "mean_token_accuracy": 0.13535947129130363,
      "num_tokens": 15541064.0,
      "step": 4655
    },
    {
      "entropy": 6.696257448196411,
      "epoch": 0.909179592234904,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941384802212207,
      "loss": 6.7437,
      "mean_token_accuracy": 0.11826530247926711,
      "num_tokens": 15559391.0,
      "step": 4660
    },
    {
      "entropy": 6.643043565750122,
      "epoch": 0.9101551068188469,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494122519975966,
      "loss": 6.5535,
      "mean_token_accuracy": 0.12430363520979881,
      "num_tokens": 15576842.0,
      "step": 4665
    },
    {
      "entropy": 6.575019645690918,
      "epoch": 0.91113062140279,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004941065383188396,
      "loss": 6.5473,
      "mean_token_accuracy": 0.13056226521730424,
      "num_tokens": 15592442.0,
      "step": 4670
    },
    {
      "entropy": 6.667135190963745,
      "epoch": 0.912106135986733,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940905352514032,
      "loss": 6.6266,
      "mean_token_accuracy": 0.1208216167986393,
      "num_tokens": 15608402.0,
      "step": 4675
    },
    {
      "entropy": 6.6816431999206545,
      "epoch": 0.913081650570676,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940745107752205,
      "loss": 6.5842,
      "mean_token_accuracy": 0.12529784068465233,
      "num_tokens": 15624349.0,
      "step": 4680
    },
    {
      "entropy": 6.626902484893799,
      "epoch": 0.9140571651546191,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940584648918573,
      "loss": 6.6541,
      "mean_token_accuracy": 0.1265459068119526,
      "num_tokens": 15640901.0,
      "step": 4685
    },
    {
      "entropy": 6.767576599121094,
      "epoch": 0.9150326797385621,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004940423976028819,
      "loss": 6.69,
      "mean_token_accuracy": 0.12248190194368362,
      "num_tokens": 15658947.0,
      "step": 4690
    },
    {
      "entropy": 6.744420051574707,
      "epoch": 0.9160081943225051,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940263089098639,
      "loss": 6.7202,
      "mean_token_accuracy": 0.12265089750289918,
      "num_tokens": 15674459.0,
      "step": 4695
    },
    {
      "entropy": 6.664055442810058,
      "epoch": 0.9169837089064482,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004940101988143757,
      "loss": 6.62,
      "mean_token_accuracy": 0.1278279460966587,
      "num_tokens": 15691951.0,
      "step": 4700
    },
    {
      "entropy": 6.686093282699585,
      "epoch": 0.9179592234903912,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004939940673179914,
      "loss": 6.6538,
      "mean_token_accuracy": 0.1175222635269165,
      "num_tokens": 15708539.0,
      "step": 4705
    },
    {
      "entropy": 6.65610294342041,
      "epoch": 0.9189347380743342,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004939779144222873,
      "loss": 6.6309,
      "mean_token_accuracy": 0.12498649656772613,
      "num_tokens": 15727108.0,
      "step": 4710
    },
    {
      "entropy": 6.718616533279419,
      "epoch": 0.9199102526582772,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939617401288419,
      "loss": 6.6062,
      "mean_token_accuracy": 0.1217942401766777,
      "num_tokens": 15743436.0,
      "step": 4715
    },
    {
      "entropy": 6.695496606826782,
      "epoch": 0.9208857672422203,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004939455444392357,
      "loss": 6.6953,
      "mean_token_accuracy": 0.11932194754481315,
      "num_tokens": 15761438.0,
      "step": 4720
    },
    {
      "entropy": 6.690045595169067,
      "epoch": 0.9218612818261633,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939293273550512,
      "loss": 6.5648,
      "mean_token_accuracy": 0.12628737688064576,
      "num_tokens": 15777064.0,
      "step": 4725
    },
    {
      "entropy": 6.682928848266601,
      "epoch": 0.9228367964101063,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004939130888778731,
      "loss": 6.5807,
      "mean_token_accuracy": 0.12611402198672295,
      "num_tokens": 15793346.0,
      "step": 4730
    },
    {
      "entropy": 6.639653205871582,
      "epoch": 0.9238123109940494,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938968290092883,
      "loss": 6.6064,
      "mean_token_accuracy": 0.12027175426483154,
      "num_tokens": 15809557.0,
      "step": 4735
    },
    {
      "entropy": 6.7091147899627686,
      "epoch": 0.9247878255779924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004938805477508853,
      "loss": 6.6439,
      "mean_token_accuracy": 0.11807163804769516,
      "num_tokens": 15826717.0,
      "step": 4740
    },
    {
      "entropy": 6.705831861495971,
      "epoch": 0.9257633401619354,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938642451042554,
      "loss": 6.6154,
      "mean_token_accuracy": 0.12317036241292953,
      "num_tokens": 15842472.0,
      "step": 4745
    },
    {
      "entropy": 6.572211837768554,
      "epoch": 0.9267388547458785,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938479210709915,
      "loss": 6.5228,
      "mean_token_accuracy": 0.13058767169713975,
      "num_tokens": 15860136.0,
      "step": 4750
    },
    {
      "entropy": 6.637489318847656,
      "epoch": 0.9277143693298214,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004938315756526886,
      "loss": 6.5734,
      "mean_token_accuracy": 0.1328073836863041,
      "num_tokens": 15878538.0,
      "step": 4755
    },
    {
      "entropy": 6.648066282272339,
      "epoch": 0.9286898839137645,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938152088509442,
      "loss": 6.6221,
      "mean_token_accuracy": 0.12146447226405144,
      "num_tokens": 15892875.0,
      "step": 4760
    },
    {
      "entropy": 6.611206340789795,
      "epoch": 0.9296653984977076,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937988206673576,
      "loss": 6.4713,
      "mean_token_accuracy": 0.13447628393769265,
      "num_tokens": 15909366.0,
      "step": 4765
    },
    {
      "entropy": 6.547593784332276,
      "epoch": 0.9306409130816505,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937824111035298,
      "loss": 6.5425,
      "mean_token_accuracy": 0.12822133898735047,
      "num_tokens": 15927712.0,
      "step": 4770
    },
    {
      "entropy": 6.752969074249267,
      "epoch": 0.9316164276655936,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937659801610647,
      "loss": 6.678,
      "mean_token_accuracy": 0.11847516149282455,
      "num_tokens": 15943474.0,
      "step": 4775
    },
    {
      "entropy": 6.65536847114563,
      "epoch": 0.9325919422495367,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937495278415677,
      "loss": 6.6363,
      "mean_token_accuracy": 0.11360074579715729,
      "num_tokens": 15959546.0,
      "step": 4780
    },
    {
      "entropy": 6.689901065826416,
      "epoch": 0.9335674568334796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937330541466465,
      "loss": 6.5635,
      "mean_token_accuracy": 0.12114342376589775,
      "num_tokens": 15975669.0,
      "step": 4785
    },
    {
      "entropy": 6.532901382446289,
      "epoch": 0.9345429714174227,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937165590779108,
      "loss": 6.5451,
      "mean_token_accuracy": 0.14050283506512642,
      "num_tokens": 15992898.0,
      "step": 4790
    },
    {
      "entropy": 6.64002423286438,
      "epoch": 0.9355184860013657,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937000426369724,
      "loss": 6.5603,
      "mean_token_accuracy": 0.1222429670393467,
      "num_tokens": 16010102.0,
      "step": 4795
    },
    {
      "entropy": 6.742595386505127,
      "epoch": 0.9364940005853087,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004936835048254453,
      "loss": 6.6468,
      "mean_token_accuracy": 0.11871858611702919,
      "num_tokens": 16027167.0,
      "step": 4800
    },
    {
      "entropy": 6.636006546020508,
      "epoch": 0.9374695151692518,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936669456449455,
      "loss": 6.6107,
      "mean_token_accuracy": 0.12101469859480858,
      "num_tokens": 16042873.0,
      "step": 4805
    },
    {
      "entropy": 6.672233247756958,
      "epoch": 0.9384450297531948,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936503650970912,
      "loss": 6.6561,
      "mean_token_accuracy": 0.12408060356974601,
      "num_tokens": 16060814.0,
      "step": 4810
    },
    {
      "entropy": 6.665853881835938,
      "epoch": 0.9394205443371378,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936337631835024,
      "loss": 6.6179,
      "mean_token_accuracy": 0.11690256521105766,
      "num_tokens": 16078378.0,
      "step": 4815
    },
    {
      "entropy": 6.499120044708252,
      "epoch": 0.9403960589210809,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936171399058017,
      "loss": 6.3969,
      "mean_token_accuracy": 0.13375149294734,
      "num_tokens": 16094601.0,
      "step": 4820
    },
    {
      "entropy": 6.5251641273498535,
      "epoch": 0.9413715735050239,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936004952656131,
      "loss": 6.4788,
      "mean_token_accuracy": 0.13522776439785958,
      "num_tokens": 16113189.0,
      "step": 4825
    },
    {
      "entropy": 6.728594493865967,
      "epoch": 0.9423470880889669,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935838292645632,
      "loss": 6.6979,
      "mean_token_accuracy": 0.11835109367966652,
      "num_tokens": 16131225.0,
      "step": 4830
    },
    {
      "entropy": 6.595848655700683,
      "epoch": 0.9433226026729099,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004935671419042805,
      "loss": 6.5581,
      "mean_token_accuracy": 0.12215998694300652,
      "num_tokens": 16149733.0,
      "step": 4835
    },
    {
      "entropy": 6.639884757995605,
      "epoch": 0.944298117256853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935504331863958,
      "loss": 6.5056,
      "mean_token_accuracy": 0.1268516443669796,
      "num_tokens": 16167499.0,
      "step": 4840
    },
    {
      "entropy": 6.478286075592041,
      "epoch": 0.945273631840796,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004935337031125416,
      "loss": 6.4445,
      "mean_token_accuracy": 0.14081556871533393,
      "num_tokens": 16184858.0,
      "step": 4845
    },
    {
      "entropy": 6.631220436096191,
      "epoch": 0.946249146424739,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004935169516843528,
      "loss": 6.6119,
      "mean_token_accuracy": 0.13074078261852265,
      "num_tokens": 16201824.0,
      "step": 4850
    },
    {
      "entropy": 6.565113925933838,
      "epoch": 0.9472246610086821,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935001789034664,
      "loss": 6.4264,
      "mean_token_accuracy": 0.13433127775788306,
      "num_tokens": 16219551.0,
      "step": 4855
    },
    {
      "entropy": 6.620077466964721,
      "epoch": 0.9482001755926251,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004934833847715212,
      "loss": 6.55,
      "mean_token_accuracy": 0.13258987292647362,
      "num_tokens": 16237447.0,
      "step": 4860
    },
    {
      "entropy": 6.694759798049927,
      "epoch": 0.9491756901765681,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934665692901583,
      "loss": 6.7261,
      "mean_token_accuracy": 0.12079535573720931,
      "num_tokens": 16254377.0,
      "step": 4865
    },
    {
      "entropy": 6.645356273651123,
      "epoch": 0.9501512047605112,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493449732461021,
      "loss": 6.5231,
      "mean_token_accuracy": 0.1307654373347759,
      "num_tokens": 16271669.0,
      "step": 4870
    },
    {
      "entropy": 6.5346602439880375,
      "epoch": 0.9511267193444543,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934328742857543,
      "loss": 6.4996,
      "mean_token_accuracy": 0.1328557774424553,
      "num_tokens": 16287088.0,
      "step": 4875
    },
    {
      "entropy": 6.528139257431031,
      "epoch": 0.9521022339283972,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934159947660057,
      "loss": 6.4668,
      "mean_token_accuracy": 0.1305578663945198,
      "num_tokens": 16303646.0,
      "step": 4880
    },
    {
      "entropy": 6.557377004623413,
      "epoch": 0.9530777485123403,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933990939034244,
      "loss": 6.5271,
      "mean_token_accuracy": 0.12777940258383752,
      "num_tokens": 16320610.0,
      "step": 4885
    },
    {
      "entropy": 6.595474481582642,
      "epoch": 0.9540532630962832,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933821716996621,
      "loss": 6.4809,
      "mean_token_accuracy": 0.13107233196496965,
      "num_tokens": 16337983.0,
      "step": 4890
    },
    {
      "entropy": 6.548955583572388,
      "epoch": 0.9550287776802263,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004933652281563723,
      "loss": 6.5163,
      "mean_token_accuracy": 0.12557137459516526,
      "num_tokens": 16356645.0,
      "step": 4895
    },
    {
      "entropy": 6.621065521240235,
      "epoch": 0.9560042922641694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933482632752107,
      "loss": 6.5462,
      "mean_token_accuracy": 0.12250527516007423,
      "num_tokens": 16371619.0,
      "step": 4900
    },
    {
      "entropy": 6.577734899520874,
      "epoch": 0.9569798068481123,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004933312770578351,
      "loss": 6.5508,
      "mean_token_accuracy": 0.13134991899132728,
      "num_tokens": 16389831.0,
      "step": 4905
    },
    {
      "entropy": 6.735396194458008,
      "epoch": 0.9579553214320554,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004933142695059052,
      "loss": 6.597,
      "mean_token_accuracy": 0.11985352486371995,
      "num_tokens": 16406132.0,
      "step": 4910
    },
    {
      "entropy": 6.620125961303711,
      "epoch": 0.9589308360159985,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004932972406210829,
      "loss": 6.5872,
      "mean_token_accuracy": 0.12829024344682693,
      "num_tokens": 16422442.0,
      "step": 4915
    },
    {
      "entropy": 6.6058845043182375,
      "epoch": 0.9599063505999414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004932801904050323,
      "loss": 6.5189,
      "mean_token_accuracy": 0.12804559618234634,
      "num_tokens": 16440290.0,
      "step": 4920
    },
    {
      "entropy": 6.6799946308135985,
      "epoch": 0.9608818651838845,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004932631188594195,
      "loss": 6.6202,
      "mean_token_accuracy": 0.12769833952188492,
      "num_tokens": 16456904.0,
      "step": 4925
    },
    {
      "entropy": 6.635846090316773,
      "epoch": 0.9618573797678275,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004932460259859127,
      "loss": 6.546,
      "mean_token_accuracy": 0.13174739703536034,
      "num_tokens": 16473777.0,
      "step": 4930
    },
    {
      "entropy": 6.582870244979858,
      "epoch": 0.9628328943517706,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004932289117861819,
      "loss": 6.5032,
      "mean_token_accuracy": 0.12783631086349487,
      "num_tokens": 16490401.0,
      "step": 4935
    },
    {
      "entropy": 6.6530838966369625,
      "epoch": 0.9638084089357136,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932117762618997,
      "loss": 6.5526,
      "mean_token_accuracy": 0.12341334894299508,
      "num_tokens": 16508582.0,
      "step": 4940
    },
    {
      "entropy": 6.667615509033203,
      "epoch": 0.9647839235196566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931946194147404,
      "loss": 6.6012,
      "mean_token_accuracy": 0.12513884902000427,
      "num_tokens": 16523939.0,
      "step": 4945
    },
    {
      "entropy": 6.57700023651123,
      "epoch": 0.9657594381035997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931774412463806,
      "loss": 6.5648,
      "mean_token_accuracy": 0.12426361218094825,
      "num_tokens": 16541139.0,
      "step": 4950
    },
    {
      "entropy": 6.695276880264283,
      "epoch": 0.9667349526875427,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931602417584989,
      "loss": 6.5912,
      "mean_token_accuracy": 0.12199072614312172,
      "num_tokens": 16558351.0,
      "step": 4955
    },
    {
      "entropy": 6.644053411483765,
      "epoch": 0.9677104672714857,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931430209527759,
      "loss": 6.6617,
      "mean_token_accuracy": 0.11836917996406555,
      "num_tokens": 16575777.0,
      "step": 4960
    },
    {
      "entropy": 6.693426942825317,
      "epoch": 0.9686859818554288,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931257788308943,
      "loss": 6.6176,
      "mean_token_accuracy": 0.12615134865045546,
      "num_tokens": 16592490.0,
      "step": 4965
    },
    {
      "entropy": 6.67751898765564,
      "epoch": 0.9696614964393717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493108515394539,
      "loss": 6.615,
      "mean_token_accuracy": 0.12720415964722634,
      "num_tokens": 16610378.0,
      "step": 4970
    },
    {
      "entropy": 6.634885597229004,
      "epoch": 0.9706370110233148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930912306453971,
      "loss": 6.65,
      "mean_token_accuracy": 0.11780654937028885,
      "num_tokens": 16625445.0,
      "step": 4975
    },
    {
      "entropy": 6.753392028808594,
      "epoch": 0.9716125256072579,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930739245851571,
      "loss": 6.6514,
      "mean_token_accuracy": 0.11931768357753754,
      "num_tokens": 16641821.0,
      "step": 4980
    },
    {
      "entropy": 6.6718357563018795,
      "epoch": 0.9725880401912008,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004930565972155106,
      "loss": 6.6054,
      "mean_token_accuracy": 0.12021942287683487,
      "num_tokens": 16658418.0,
      "step": 4985
    },
    {
      "entropy": 6.65066180229187,
      "epoch": 0.9735635547751439,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930392485381506,
      "loss": 6.623,
      "mean_token_accuracy": 0.12545933946967125,
      "num_tokens": 16674218.0,
      "step": 4990
    },
    {
      "entropy": 6.651250171661377,
      "epoch": 0.974539069359087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930218785547724,
      "loss": 6.6249,
      "mean_token_accuracy": 0.1307213805615902,
      "num_tokens": 16692171.0,
      "step": 4995
    },
    {
      "entropy": 6.7233082294464115,
      "epoch": 0.9755145839430299,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930044872670733,
      "loss": 6.6744,
      "mean_token_accuracy": 0.11614948287606239,
      "num_tokens": 16709231.0,
      "step": 5000
    },
    {
      "entropy": 6.667290449142456,
      "epoch": 0.976490098526973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929870746767525,
      "loss": 6.5908,
      "mean_token_accuracy": 0.12379949539899826,
      "num_tokens": 16726281.0,
      "step": 5005
    },
    {
      "entropy": 6.631280612945557,
      "epoch": 0.977465613110916,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929696407855118,
      "loss": 6.6437,
      "mean_token_accuracy": 0.11903728470206261,
      "num_tokens": 16744821.0,
      "step": 5010
    },
    {
      "entropy": 6.607823324203491,
      "epoch": 0.978441127694859,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004929521855950546,
      "loss": 6.4703,
      "mean_token_accuracy": 0.12637271657586097,
      "num_tokens": 16762564.0,
      "step": 5015
    },
    {
      "entropy": 6.569560384750366,
      "epoch": 0.9794166422788021,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929347091070866,
      "loss": 6.4237,
      "mean_token_accuracy": 0.13844026923179625,
      "num_tokens": 16777681.0,
      "step": 5020
    },
    {
      "entropy": 6.668828248977661,
      "epoch": 0.9803921568627451,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929172113233157,
      "loss": 6.6562,
      "mean_token_accuracy": 0.12064324542880059,
      "num_tokens": 16793820.0,
      "step": 5025
    },
    {
      "entropy": 6.707146215438843,
      "epoch": 0.9813676714466881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004928996922454515,
      "loss": 6.6416,
      "mean_token_accuracy": 0.12269999384880066,
      "num_tokens": 16811445.0,
      "step": 5030
    },
    {
      "entropy": 6.63042721748352,
      "epoch": 0.9823431860306312,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928821518752059,
      "loss": 6.6091,
      "mean_token_accuracy": 0.12471438571810722,
      "num_tokens": 16827961.0,
      "step": 5035
    },
    {
      "entropy": 6.682687711715698,
      "epoch": 0.9833187006145742,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928645902142931,
      "loss": 6.5042,
      "mean_token_accuracy": 0.12736130952835084,
      "num_tokens": 16844074.0,
      "step": 5040
    },
    {
      "entropy": 6.579976749420166,
      "epoch": 0.9842942151985172,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000492847007264429,
      "loss": 6.5413,
      "mean_token_accuracy": 0.12310777455568314,
      "num_tokens": 16862004.0,
      "step": 5045
    },
    {
      "entropy": 6.637821006774902,
      "epoch": 0.9852697297824603,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928294030273317,
      "loss": 6.5596,
      "mean_token_accuracy": 0.12315315455198288,
      "num_tokens": 16879738.0,
      "step": 5050
    },
    {
      "entropy": 6.558614778518677,
      "epoch": 0.9862452443664033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004928117775047215,
      "loss": 6.5144,
      "mean_token_accuracy": 0.1313558891415596,
      "num_tokens": 16896565.0,
      "step": 5055
    },
    {
      "entropy": 6.669071912765503,
      "epoch": 0.9872207589503463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004927941306983207,
      "loss": 6.5804,
      "mean_token_accuracy": 0.12495598196983337,
      "num_tokens": 16912443.0,
      "step": 5060
    },
    {
      "entropy": 6.719775247573852,
      "epoch": 0.9881962735342893,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927764626098537,
      "loss": 6.642,
      "mean_token_accuracy": 0.12029725685715675,
      "num_tokens": 16928292.0,
      "step": 5065
    },
    {
      "entropy": 6.566314601898194,
      "epoch": 0.9891717881182324,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492758773241047,
      "loss": 6.608,
      "mean_token_accuracy": 0.12571064084768296,
      "num_tokens": 16945432.0,
      "step": 5070
    },
    {
      "entropy": 6.56963701248169,
      "epoch": 0.9901473027021754,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000492741062593629,
      "loss": 6.431,
      "mean_token_accuracy": 0.1381141483783722,
      "num_tokens": 16962440.0,
      "step": 5075
    },
    {
      "entropy": 6.608655405044556,
      "epoch": 0.9911228172861184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004927233306693305,
      "loss": 6.5485,
      "mean_token_accuracy": 0.12887553945183755,
      "num_tokens": 16979227.0,
      "step": 5080
    },
    {
      "entropy": 6.65983624458313,
      "epoch": 0.9920983318700615,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004927055774698839,
      "loss": 6.6052,
      "mean_token_accuracy": 0.1178313933312893,
      "num_tokens": 16994873.0,
      "step": 5085
    },
    {
      "entropy": 6.703251552581787,
      "epoch": 0.9930738464540045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004926878029970245,
      "loss": 6.6372,
      "mean_token_accuracy": 0.12175166457891465,
      "num_tokens": 17011516.0,
      "step": 5090
    },
    {
      "entropy": 6.6461550235748295,
      "epoch": 0.9940493610379475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926700072524887,
      "loss": 6.5447,
      "mean_token_accuracy": 0.12201424613595009,
      "num_tokens": 17029407.0,
      "step": 5095
    },
    {
      "entropy": 6.602279424667358,
      "epoch": 0.9950248756218906,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004926521902380154,
      "loss": 6.6076,
      "mean_token_accuracy": 0.1184044562280178,
      "num_tokens": 17044451.0,
      "step": 5100
    },
    {
      "entropy": 6.526291084289551,
      "epoch": 0.9960003902058335,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004926343519553461,
      "loss": 6.4621,
      "mean_token_accuracy": 0.1353236570954323,
      "num_tokens": 17060420.0,
      "step": 5105
    },
    {
      "entropy": 6.640697288513183,
      "epoch": 0.9969759047897766,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926164924062234,
      "loss": 6.5813,
      "mean_token_accuracy": 0.12561142593622207,
      "num_tokens": 17076562.0,
      "step": 5110
    },
    {
      "entropy": 6.5548206806182865,
      "epoch": 0.9979514193737197,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004925986115923927,
      "loss": 6.6214,
      "mean_token_accuracy": 0.1372695356607437,
      "num_tokens": 17092441.0,
      "step": 5115
    },
    {
      "entropy": 6.712176752090454,
      "epoch": 0.9989269339576626,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004925807095156013,
      "loss": 6.6069,
      "mean_token_accuracy": 0.12348895519971848,
      "num_tokens": 17109447.0,
      "step": 5120
    },
    {
      "entropy": 6.546150207519531,
      "epoch": 0.9999024485416057,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004925627861775983,
      "loss": 6.4965,
      "mean_token_accuracy": 0.1317557990550995,
      "num_tokens": 17125795.0,
      "step": 5125
    },
    {
      "entropy": 6.632970492045085,
      "epoch": 1.0007804116671544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925448415801353,
      "loss": 6.521,
      "mean_token_accuracy": 0.1260004316767057,
      "num_tokens": 17139062.0,
      "step": 5130
    },
    {
      "entropy": 6.675327491760254,
      "epoch": 1.0017559262510976,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925268757249657,
      "loss": 6.5332,
      "mean_token_accuracy": 0.11733304336667061,
      "num_tokens": 17155465.0,
      "step": 5135
    },
    {
      "entropy": 6.628605699539184,
      "epoch": 1.0027314408350405,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925088886138451,
      "loss": 6.4318,
      "mean_token_accuracy": 0.12678520604968072,
      "num_tokens": 17171284.0,
      "step": 5140
    },
    {
      "entropy": 6.467437505722046,
      "epoch": 1.0037069554189835,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004924908802485311,
      "loss": 6.3604,
      "mean_token_accuracy": 0.13715625777840615,
      "num_tokens": 17190823.0,
      "step": 5145
    },
    {
      "entropy": 6.650004720687866,
      "epoch": 1.0046824700029267,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004924728506307835,
      "loss": 6.4752,
      "mean_token_accuracy": 0.12543649822473527,
      "num_tokens": 17208866.0,
      "step": 5150
    },
    {
      "entropy": 6.643254899978638,
      "epoch": 1.0056579845868696,
      "grad_norm": 1.125,
      "learning_rate": 0.000492454799762364,
      "loss": 6.5642,
      "mean_token_accuracy": 0.12036867663264275,
      "num_tokens": 17225609.0,
      "step": 5155
    },
    {
      "entropy": 6.6838665962219235,
      "epoch": 1.0066334991708126,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924367276450364,
      "loss": 6.5633,
      "mean_token_accuracy": 0.12325121238827705,
      "num_tokens": 17242198.0,
      "step": 5160
    },
    {
      "entropy": 6.5908466339111325,
      "epoch": 1.0076090137547555,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924186342805669,
      "loss": 6.5166,
      "mean_token_accuracy": 0.1237028606235981,
      "num_tokens": 17258712.0,
      "step": 5165
    },
    {
      "entropy": 6.631665945053101,
      "epoch": 1.0085845283386987,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924005196707233,
      "loss": 6.5005,
      "mean_token_accuracy": 0.1302240550518036,
      "num_tokens": 17274790.0,
      "step": 5170
    },
    {
      "entropy": 6.631072187423706,
      "epoch": 1.0095600429226417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923823838172756,
      "loss": 6.4597,
      "mean_token_accuracy": 0.12370958849787712,
      "num_tokens": 17290526.0,
      "step": 5175
    },
    {
      "entropy": 6.619173765182495,
      "epoch": 1.0105355575065846,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923642267219962,
      "loss": 6.5446,
      "mean_token_accuracy": 0.1220270849764347,
      "num_tokens": 17306022.0,
      "step": 5180
    },
    {
      "entropy": 6.626012897491455,
      "epoch": 1.0115110720905278,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923460483866593,
      "loss": 6.5119,
      "mean_token_accuracy": 0.12375197634100914,
      "num_tokens": 17321601.0,
      "step": 5185
    },
    {
      "entropy": 6.507984828948975,
      "epoch": 1.0124865866744708,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923278488130412,
      "loss": 6.3925,
      "mean_token_accuracy": 0.13171566426753997,
      "num_tokens": 17337997.0,
      "step": 5190
    },
    {
      "entropy": 6.637670660018921,
      "epoch": 1.0134621012584137,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004923096280029202,
      "loss": 6.4763,
      "mean_token_accuracy": 0.12316565215587616,
      "num_tokens": 17354809.0,
      "step": 5195
    },
    {
      "entropy": 6.537084913253784,
      "epoch": 1.014437615842357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000492291385958077,
      "loss": 6.3359,
      "mean_token_accuracy": 0.13839974477887154,
      "num_tokens": 17371548.0,
      "step": 5200
    },
    {
      "entropy": 6.5619823932647705,
      "epoch": 1.0154131304262999,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004922731226802939,
      "loss": 6.452,
      "mean_token_accuracy": 0.1287774607539177,
      "num_tokens": 17388891.0,
      "step": 5205
    },
    {
      "entropy": 6.4974089622497555,
      "epoch": 1.0163886450102428,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922548381713556,
      "loss": 6.4235,
      "mean_token_accuracy": 0.13361327797174455,
      "num_tokens": 17407501.0,
      "step": 5210
    },
    {
      "entropy": 6.570951509475708,
      "epoch": 1.017364159594186,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922365324330489,
      "loss": 6.4646,
      "mean_token_accuracy": 0.12979131415486336,
      "num_tokens": 17425541.0,
      "step": 5215
    },
    {
      "entropy": 6.564201211929321,
      "epoch": 1.018339674178129,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922182054671625,
      "loss": 6.4153,
      "mean_token_accuracy": 0.12152216583490372,
      "num_tokens": 17441588.0,
      "step": 5220
    },
    {
      "entropy": 6.567869997024536,
      "epoch": 1.019315188762072,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921998572754872,
      "loss": 6.4176,
      "mean_token_accuracy": 0.12886710688471795,
      "num_tokens": 17458874.0,
      "step": 5225
    },
    {
      "entropy": 6.509510946273804,
      "epoch": 1.0202907033460151,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921814878598159,
      "loss": 6.3845,
      "mean_token_accuracy": 0.1318638265132904,
      "num_tokens": 17475815.0,
      "step": 5230
    },
    {
      "entropy": 6.561601638793945,
      "epoch": 1.021266217929958,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921630972219437,
      "loss": 6.4241,
      "mean_token_accuracy": 0.12493757084012032,
      "num_tokens": 17492068.0,
      "step": 5235
    },
    {
      "entropy": 6.52676739692688,
      "epoch": 1.022241732513901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921446853636676,
      "loss": 6.409,
      "mean_token_accuracy": 0.12581436336040497,
      "num_tokens": 17508156.0,
      "step": 5240
    },
    {
      "entropy": 6.6034191131591795,
      "epoch": 1.023217247097844,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921262522867869,
      "loss": 6.5045,
      "mean_token_accuracy": 0.12258622720837593,
      "num_tokens": 17525374.0,
      "step": 5245
    },
    {
      "entropy": 6.569457960128784,
      "epoch": 1.0241927616817872,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921077979931026,
      "loss": 6.3946,
      "mean_token_accuracy": 0.12949460372328758,
      "num_tokens": 17541453.0,
      "step": 5250
    },
    {
      "entropy": 6.579229068756104,
      "epoch": 1.0251682762657301,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492089322484418,
      "loss": 6.4588,
      "mean_token_accuracy": 0.12597510516643523,
      "num_tokens": 17557981.0,
      "step": 5255
    },
    {
      "entropy": 6.527565383911133,
      "epoch": 1.026143790849673,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004920708257625386,
      "loss": 6.3558,
      "mean_token_accuracy": 0.13429988622665406,
      "num_tokens": 17573094.0,
      "step": 5260
    },
    {
      "entropy": 6.60800986289978,
      "epoch": 1.0271193054336163,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920523078292718,
      "loss": 6.5875,
      "mean_token_accuracy": 0.12211998999118805,
      "num_tokens": 17589933.0,
      "step": 5265
    },
    {
      "entropy": 6.6778264999389645,
      "epoch": 1.0280948200175593,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004920337686864269,
      "loss": 6.5118,
      "mean_token_accuracy": 0.12417216822504998,
      "num_tokens": 17607008.0,
      "step": 5270
    },
    {
      "entropy": 6.6295506954193115,
      "epoch": 1.0290703346015022,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004920152083358158,
      "loss": 6.4801,
      "mean_token_accuracy": 0.13306098207831382,
      "num_tokens": 17624160.0,
      "step": 5275
    },
    {
      "entropy": 6.529311943054199,
      "epoch": 1.0300458491854454,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000491996626779252,
      "loss": 6.4314,
      "mean_token_accuracy": 0.1316734366118908,
      "num_tokens": 17640067.0,
      "step": 5280
    },
    {
      "entropy": 6.621422195434571,
      "epoch": 1.0310213637693884,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004919780240185512,
      "loss": 6.4966,
      "mean_token_accuracy": 0.12465468496084213,
      "num_tokens": 17658278.0,
      "step": 5285
    },
    {
      "entropy": 6.628733062744141,
      "epoch": 1.0319968783533313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004919594000555311,
      "loss": 6.4976,
      "mean_token_accuracy": 0.1205541729927063,
      "num_tokens": 17673878.0,
      "step": 5290
    },
    {
      "entropy": 6.524856662750244,
      "epoch": 1.0329723929372745,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919407548920118,
      "loss": 6.4307,
      "mean_token_accuracy": 0.128728000074625,
      "num_tokens": 17688872.0,
      "step": 5295
    },
    {
      "entropy": 6.606214570999145,
      "epoch": 1.0339479075212175,
      "grad_norm": 0.953125,
      "learning_rate": 0.000491922088529815,
      "loss": 6.4626,
      "mean_token_accuracy": 0.12438188269734382,
      "num_tokens": 17704284.0,
      "step": 5300
    },
    {
      "entropy": 6.490259838104248,
      "epoch": 1.0349234221051604,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004919034009707651,
      "loss": 6.3298,
      "mean_token_accuracy": 0.13744132071733475,
      "num_tokens": 17721296.0,
      "step": 5305
    },
    {
      "entropy": 6.605351066589355,
      "epoch": 1.0358989366891036,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004918846922166877,
      "loss": 6.5343,
      "mean_token_accuracy": 0.1257959745824337,
      "num_tokens": 17738549.0,
      "step": 5310
    },
    {
      "entropy": 6.597479057312012,
      "epoch": 1.0368744512730466,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004918659622694111,
      "loss": 6.4065,
      "mean_token_accuracy": 0.13634497299790382,
      "num_tokens": 17755297.0,
      "step": 5315
    },
    {
      "entropy": 6.505791282653808,
      "epoch": 1.0378499658569895,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004918472111307658,
      "loss": 6.4534,
      "mean_token_accuracy": 0.12901222333312035,
      "num_tokens": 17771459.0,
      "step": 5320
    },
    {
      "entropy": 6.5554612159729,
      "epoch": 1.0388254804409325,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918284388025837,
      "loss": 6.4612,
      "mean_token_accuracy": 0.13127356842160226,
      "num_tokens": 17787009.0,
      "step": 5325
    },
    {
      "entropy": 6.5526814460754395,
      "epoch": 1.0398009950248757,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004918096452866995,
      "loss": 6.4791,
      "mean_token_accuracy": 0.12473830506205559,
      "num_tokens": 17804643.0,
      "step": 5330
    },
    {
      "entropy": 6.615708446502685,
      "epoch": 1.0407765096088186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917908305849493,
      "loss": 6.533,
      "mean_token_accuracy": 0.12203987613320351,
      "num_tokens": 17821104.0,
      "step": 5335
    },
    {
      "entropy": 6.711125469207763,
      "epoch": 1.0417520241927616,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491771994699172,
      "loss": 6.533,
      "mean_token_accuracy": 0.12563105449080467,
      "num_tokens": 17838094.0,
      "step": 5340
    },
    {
      "entropy": 6.446725130081177,
      "epoch": 1.0427275387767048,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917531376312079,
      "loss": 6.3524,
      "mean_token_accuracy": 0.13376612737774848,
      "num_tokens": 17853856.0,
      "step": 5345
    },
    {
      "entropy": 6.554876708984375,
      "epoch": 1.0437030533606477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917342593828997,
      "loss": 6.4893,
      "mean_token_accuracy": 0.12014652118086815,
      "num_tokens": 17870032.0,
      "step": 5350
    },
    {
      "entropy": 6.500568389892578,
      "epoch": 1.0446785679445907,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491715359956092,
      "loss": 6.3645,
      "mean_token_accuracy": 0.13705315738916396,
      "num_tokens": 17886384.0,
      "step": 5355
    },
    {
      "entropy": 6.784595394134522,
      "epoch": 1.0456540825285339,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004916964393526319,
      "loss": 6.5921,
      "mean_token_accuracy": 0.1252087749540806,
      "num_tokens": 17903444.0,
      "step": 5360
    },
    {
      "entropy": 6.5807770729064945,
      "epoch": 1.0466295971124768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004916774975743679,
      "loss": 6.4392,
      "mean_token_accuracy": 0.12736949771642686,
      "num_tokens": 17920042.0,
      "step": 5365
    },
    {
      "entropy": 6.496833658218383,
      "epoch": 1.0476051116964198,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916585346231513,
      "loss": 6.4227,
      "mean_token_accuracy": 0.1430770516395569,
      "num_tokens": 17937115.0,
      "step": 5370
    },
    {
      "entropy": 6.589770030975342,
      "epoch": 1.048580626280363,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916395505008347,
      "loss": 6.4783,
      "mean_token_accuracy": 0.12518419325351715,
      "num_tokens": 17953582.0,
      "step": 5375
    },
    {
      "entropy": 6.526508474349976,
      "epoch": 1.049556140864306,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916205452092735,
      "loss": 6.4672,
      "mean_token_accuracy": 0.12632611617445946,
      "num_tokens": 17970457.0,
      "step": 5380
    },
    {
      "entropy": 6.654922676086426,
      "epoch": 1.050531655448249,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916015187503245,
      "loss": 6.4535,
      "mean_token_accuracy": 0.13003793060779573,
      "num_tokens": 17987168.0,
      "step": 5385
    },
    {
      "entropy": 6.394747495651245,
      "epoch": 1.051507170032192,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004915824711258472,
      "loss": 6.3849,
      "mean_token_accuracy": 0.14045090228319168,
      "num_tokens": 18004964.0,
      "step": 5390
    },
    {
      "entropy": 6.644694137573242,
      "epoch": 1.052482684616135,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004915634023377026,
      "loss": 6.458,
      "mean_token_accuracy": 0.12786104157567024,
      "num_tokens": 18021778.0,
      "step": 5395
    },
    {
      "entropy": 6.386481094360351,
      "epoch": 1.053458199200078,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004915443123877543,
      "loss": 6.3057,
      "mean_token_accuracy": 0.1314587637782097,
      "num_tokens": 18037620.0,
      "step": 5400
    },
    {
      "entropy": 6.614221811294556,
      "epoch": 1.0544337137840212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004915252012778676,
      "loss": 6.4189,
      "mean_token_accuracy": 0.12962789088487625,
      "num_tokens": 18054623.0,
      "step": 5405
    },
    {
      "entropy": 6.504178190231324,
      "epoch": 1.0554092283679641,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915060690099098,
      "loss": 6.4365,
      "mean_token_accuracy": 0.12476388588547707,
      "num_tokens": 18071271.0,
      "step": 5410
    },
    {
      "entropy": 6.571506786346435,
      "epoch": 1.056384742951907,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914869155857506,
      "loss": 6.5073,
      "mean_token_accuracy": 0.12393756732344627,
      "num_tokens": 18087661.0,
      "step": 5415
    },
    {
      "entropy": 6.541922664642334,
      "epoch": 1.05736025753585,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914677410072616,
      "loss": 6.4075,
      "mean_token_accuracy": 0.13256005048751832,
      "num_tokens": 18106116.0,
      "step": 5420
    },
    {
      "entropy": 6.627392196655274,
      "epoch": 1.0583357721197932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914485452763165,
      "loss": 6.5062,
      "mean_token_accuracy": 0.12414982840418816,
      "num_tokens": 18122420.0,
      "step": 5425
    },
    {
      "entropy": 6.617330074310303,
      "epoch": 1.0593112867037362,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491429328394791,
      "loss": 6.5207,
      "mean_token_accuracy": 0.12191471606492996,
      "num_tokens": 18137974.0,
      "step": 5430
    },
    {
      "entropy": 6.533131170272827,
      "epoch": 1.0602868012876792,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004914100903645629,
      "loss": 6.3467,
      "mean_token_accuracy": 0.1365069791674614,
      "num_tokens": 18154711.0,
      "step": 5435
    },
    {
      "entropy": 6.564487743377685,
      "epoch": 1.0612623158716223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491390831187512,
      "loss": 6.515,
      "mean_token_accuracy": 0.1275060720741749,
      "num_tokens": 18172422.0,
      "step": 5440
    },
    {
      "entropy": 6.4970887184143065,
      "epoch": 1.0622378304555653,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004913715508655203,
      "loss": 6.3923,
      "mean_token_accuracy": 0.138767609000206,
      "num_tokens": 18189517.0,
      "step": 5445
    },
    {
      "entropy": 6.569665861129761,
      "epoch": 1.0632133450395083,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913522494004719,
      "loss": 6.464,
      "mean_token_accuracy": 0.11712840795516968,
      "num_tokens": 18206339.0,
      "step": 5450
    },
    {
      "entropy": 6.503337001800537,
      "epoch": 1.0641888596234514,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004913329267942527,
      "loss": 6.3153,
      "mean_token_accuracy": 0.13992415964603425,
      "num_tokens": 18223953.0,
      "step": 5455
    },
    {
      "entropy": 6.652460527420044,
      "epoch": 1.0651643742073944,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491313583048751,
      "loss": 6.4296,
      "mean_token_accuracy": 0.12903519049286843,
      "num_tokens": 18241361.0,
      "step": 5460
    },
    {
      "entropy": 6.453370714187622,
      "epoch": 1.0661398887913374,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912942181658569,
      "loss": 6.4686,
      "mean_token_accuracy": 0.13046226501464844,
      "num_tokens": 18257897.0,
      "step": 5465
    },
    {
      "entropy": 6.644266080856323,
      "epoch": 1.0671154033752805,
      "grad_norm": 1.125,
      "learning_rate": 0.0004912748321474627,
      "loss": 6.4615,
      "mean_token_accuracy": 0.12619092538952828,
      "num_tokens": 18274935.0,
      "step": 5470
    },
    {
      "entropy": 6.494524335861206,
      "epoch": 1.0680909179592235,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912554249954626,
      "loss": 6.5187,
      "mean_token_accuracy": 0.12464958652853966,
      "num_tokens": 18292529.0,
      "step": 5475
    },
    {
      "entropy": 6.696386671066284,
      "epoch": 1.0690664325431665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912359967117532,
      "loss": 6.4545,
      "mean_token_accuracy": 0.12537138313055038,
      "num_tokens": 18310278.0,
      "step": 5480
    },
    {
      "entropy": 6.542104196548462,
      "epoch": 1.0700419471271096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912165472982329,
      "loss": 6.4641,
      "mean_token_accuracy": 0.13138798698782922,
      "num_tokens": 18326972.0,
      "step": 5485
    },
    {
      "entropy": 6.5346049785614015,
      "epoch": 1.0710174617110526,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911970767568022,
      "loss": 6.5672,
      "mean_token_accuracy": 0.1251851089298725,
      "num_tokens": 18345065.0,
      "step": 5490
    },
    {
      "entropy": 6.620600795745849,
      "epoch": 1.0719929762949956,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911775850893637,
      "loss": 6.4888,
      "mean_token_accuracy": 0.12605516836047173,
      "num_tokens": 18363251.0,
      "step": 5495
    },
    {
      "entropy": 6.504895734786987,
      "epoch": 1.0729684908789388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491158072297822,
      "loss": 6.3235,
      "mean_token_accuracy": 0.13353361412882805,
      "num_tokens": 18378650.0,
      "step": 5500
    },
    {
      "entropy": 6.539031505584717,
      "epoch": 1.0739440054628817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911385383840839,
      "loss": 6.484,
      "mean_token_accuracy": 0.1277184709906578,
      "num_tokens": 18394533.0,
      "step": 5505
    },
    {
      "entropy": 6.639795446395874,
      "epoch": 1.0749195200468247,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911189833500582,
      "loss": 6.5248,
      "mean_token_accuracy": 0.12210818007588387,
      "num_tokens": 18410752.0,
      "step": 5510
    },
    {
      "entropy": 6.426144647598266,
      "epoch": 1.0758950346307676,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910994071976556,
      "loss": 6.3162,
      "mean_token_accuracy": 0.13739656060934066,
      "num_tokens": 18428027.0,
      "step": 5515
    },
    {
      "entropy": 6.4942862033844,
      "epoch": 1.0768705492147108,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910798099287894,
      "loss": 6.3808,
      "mean_token_accuracy": 0.131735647469759,
      "num_tokens": 18446380.0,
      "step": 5520
    },
    {
      "entropy": 6.475603485107422,
      "epoch": 1.0778460637986538,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491060191545374,
      "loss": 6.3369,
      "mean_token_accuracy": 0.14313014149665831,
      "num_tokens": 18464628.0,
      "step": 5525
    },
    {
      "entropy": 6.54031343460083,
      "epoch": 1.0788215783825967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910405520493268,
      "loss": 6.4265,
      "mean_token_accuracy": 0.1312645547091961,
      "num_tokens": 18479995.0,
      "step": 5530
    },
    {
      "entropy": 6.575513219833374,
      "epoch": 1.07979709296654,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004910208914425669,
      "loss": 6.5047,
      "mean_token_accuracy": 0.12677802443504332,
      "num_tokens": 18497143.0,
      "step": 5535
    },
    {
      "entropy": 6.598835372924805,
      "epoch": 1.0807726075504829,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004910012097270153,
      "loss": 6.4687,
      "mean_token_accuracy": 0.1258193537592888,
      "num_tokens": 18513428.0,
      "step": 5540
    },
    {
      "entropy": 6.627675485610962,
      "epoch": 1.0817481221344258,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909815069045955,
      "loss": 6.5659,
      "mean_token_accuracy": 0.12378744184970855,
      "num_tokens": 18529475.0,
      "step": 5545
    },
    {
      "entropy": 6.6549944400787355,
      "epoch": 1.082723636718369,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909617829772325,
      "loss": 6.4936,
      "mean_token_accuracy": 0.126087536662817,
      "num_tokens": 18546156.0,
      "step": 5550
    },
    {
      "entropy": 6.402008724212647,
      "epoch": 1.083699151302312,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909420379468537,
      "loss": 6.2711,
      "mean_token_accuracy": 0.15442217737436295,
      "num_tokens": 18562719.0,
      "step": 5555
    },
    {
      "entropy": 6.504349231719971,
      "epoch": 1.084674665886255,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909222718153887,
      "loss": 6.4432,
      "mean_token_accuracy": 0.13431486040353774,
      "num_tokens": 18580123.0,
      "step": 5560
    },
    {
      "entropy": 6.556483173370362,
      "epoch": 1.0856501804701981,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909024845847689,
      "loss": 6.3553,
      "mean_token_accuracy": 0.1313481517136097,
      "num_tokens": 18597003.0,
      "step": 5565
    },
    {
      "entropy": 6.513379955291748,
      "epoch": 1.086625695054141,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004908826762569276,
      "loss": 6.4476,
      "mean_token_accuracy": 0.12424071207642555,
      "num_tokens": 18613812.0,
      "step": 5570
    },
    {
      "entropy": 6.485783529281616,
      "epoch": 1.087601209638084,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004908628468338008,
      "loss": 6.3794,
      "mean_token_accuracy": 0.1415726013481617,
      "num_tokens": 18630333.0,
      "step": 5575
    },
    {
      "entropy": 6.615262269973755,
      "epoch": 1.088576724222027,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490842996317326,
      "loss": 6.4808,
      "mean_token_accuracy": 0.12457754090428352,
      "num_tokens": 18646784.0,
      "step": 5580
    },
    {
      "entropy": 6.544322443008423,
      "epoch": 1.0895522388059702,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908231247094427,
      "loss": 6.4229,
      "mean_token_accuracy": 0.13085711002349854,
      "num_tokens": 18664358.0,
      "step": 5585
    },
    {
      "entropy": 6.479617977142334,
      "epoch": 1.0905277533899131,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908032320120929,
      "loss": 6.3473,
      "mean_token_accuracy": 0.13286124914884567,
      "num_tokens": 18681849.0,
      "step": 5590
    },
    {
      "entropy": 6.574735307693482,
      "epoch": 1.091503267973856,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907833182272204,
      "loss": 6.4805,
      "mean_token_accuracy": 0.12393207326531411,
      "num_tokens": 18697127.0,
      "step": 5595
    },
    {
      "entropy": 6.453974008560181,
      "epoch": 1.0924787825577993,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004907633833567711,
      "loss": 6.4117,
      "mean_token_accuracy": 0.13349757269024848,
      "num_tokens": 18714859.0,
      "step": 5600
    },
    {
      "entropy": 6.584815835952758,
      "epoch": 1.0934542971417422,
      "grad_norm": 1.296875,
      "learning_rate": 0.000490743427402693,
      "loss": 6.4417,
      "mean_token_accuracy": 0.12737256810069084,
      "num_tokens": 18730083.0,
      "step": 5605
    },
    {
      "entropy": 6.497727346420288,
      "epoch": 1.0944298117256852,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004907234503669361,
      "loss": 6.3725,
      "mean_token_accuracy": 0.1302713632583618,
      "num_tokens": 18746694.0,
      "step": 5610
    },
    {
      "entropy": 6.62595763206482,
      "epoch": 1.0954053263096284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004907034522514524,
      "loss": 6.4582,
      "mean_token_accuracy": 0.1301114723086357,
      "num_tokens": 18764333.0,
      "step": 5615
    },
    {
      "entropy": 6.5605308532714846,
      "epoch": 1.0963808408935714,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906834330581963,
      "loss": 6.4501,
      "mean_token_accuracy": 0.12743061780929565,
      "num_tokens": 18780309.0,
      "step": 5620
    },
    {
      "entropy": 6.510411977767944,
      "epoch": 1.0973563554775143,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004906633927891237,
      "loss": 6.4524,
      "mean_token_accuracy": 0.13104955106973648,
      "num_tokens": 18797322.0,
      "step": 5625
    },
    {
      "entropy": 6.555394315719605,
      "epoch": 1.0983318700614575,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490643331446193,
      "loss": 6.4986,
      "mean_token_accuracy": 0.1253698855638504,
      "num_tokens": 18813185.0,
      "step": 5630
    },
    {
      "entropy": 6.4955360889434814,
      "epoch": 1.0993073846454005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906232490313646,
      "loss": 6.3813,
      "mean_token_accuracy": 0.13279257714748383,
      "num_tokens": 18829445.0,
      "step": 5635
    },
    {
      "entropy": 6.557409715652466,
      "epoch": 1.1002828992293434,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004906031455466008,
      "loss": 6.5312,
      "mean_token_accuracy": 0.12552498355507852,
      "num_tokens": 18846229.0,
      "step": 5640
    },
    {
      "entropy": 6.583230781555176,
      "epoch": 1.1012584138132866,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004905830209938659,
      "loss": 6.4913,
      "mean_token_accuracy": 0.123032096773386,
      "num_tokens": 18861904.0,
      "step": 5645
    },
    {
      "entropy": 6.577669429779053,
      "epoch": 1.1022339283972296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905628753751267,
      "loss": 6.5218,
      "mean_token_accuracy": 0.12506726011633873,
      "num_tokens": 18879160.0,
      "step": 5650
    },
    {
      "entropy": 6.653974962234497,
      "epoch": 1.1032094429811725,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905427086923515,
      "loss": 6.4941,
      "mean_token_accuracy": 0.11950188800692559,
      "num_tokens": 18895388.0,
      "step": 5655
    },
    {
      "entropy": 6.522398853302002,
      "epoch": 1.1041849575651157,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490522520947511,
      "loss": 6.4499,
      "mean_token_accuracy": 0.12685119286179541,
      "num_tokens": 18912392.0,
      "step": 5660
    },
    {
      "entropy": 6.598175001144409,
      "epoch": 1.1051604721490587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905023121425779,
      "loss": 6.4297,
      "mean_token_accuracy": 0.12772816196084022,
      "num_tokens": 18928721.0,
      "step": 5665
    },
    {
      "entropy": 6.597598218917847,
      "epoch": 1.1061359867330016,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000490482082279527,
      "loss": 6.4935,
      "mean_token_accuracy": 0.12312249392271042,
      "num_tokens": 18947144.0,
      "step": 5670
    },
    {
      "entropy": 6.551220989227295,
      "epoch": 1.1071115013169446,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490461831360335,
      "loss": 6.4323,
      "mean_token_accuracy": 0.12854888513684273,
      "num_tokens": 18963682.0,
      "step": 5675
    },
    {
      "entropy": 6.547711658477783,
      "epoch": 1.1080870159008878,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004904415593869806,
      "loss": 6.4702,
      "mean_token_accuracy": 0.12626882418990135,
      "num_tokens": 18980812.0,
      "step": 5680
    },
    {
      "entropy": 6.555589008331299,
      "epoch": 1.1090625304848307,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904212663614452,
      "loss": 6.4423,
      "mean_token_accuracy": 0.13077757507562637,
      "num_tokens": 18996816.0,
      "step": 5685
    },
    {
      "entropy": 6.557824230194091,
      "epoch": 1.1100380450687737,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004904009522857111,
      "loss": 6.4487,
      "mean_token_accuracy": 0.12600879296660422,
      "num_tokens": 19012324.0,
      "step": 5690
    },
    {
      "entropy": 6.462041568756104,
      "epoch": 1.1110135596527169,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903806171617638,
      "loss": 6.3001,
      "mean_token_accuracy": 0.14120589420199395,
      "num_tokens": 19030195.0,
      "step": 5695
    },
    {
      "entropy": 6.458101987838745,
      "epoch": 1.1119890742366598,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004903602609915901,
      "loss": 6.3136,
      "mean_token_accuracy": 0.13771101757884024,
      "num_tokens": 19048467.0,
      "step": 5700
    },
    {
      "entropy": 6.489548587799073,
      "epoch": 1.1129645888206028,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903398837771794,
      "loss": 6.4194,
      "mean_token_accuracy": 0.13316495940089226,
      "num_tokens": 19066290.0,
      "step": 5705
    },
    {
      "entropy": 6.41388168334961,
      "epoch": 1.113940103404546,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903194855205227,
      "loss": 6.3346,
      "mean_token_accuracy": 0.13676948994398117,
      "num_tokens": 19083455.0,
      "step": 5710
    },
    {
      "entropy": 6.5682590961456295,
      "epoch": 1.114915617988489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004902990662236132,
      "loss": 6.4092,
      "mean_token_accuracy": 0.12743354290723802,
      "num_tokens": 19100347.0,
      "step": 5715
    },
    {
      "entropy": 6.507794666290283,
      "epoch": 1.1158911325724319,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902786258884463,
      "loss": 6.4522,
      "mean_token_accuracy": 0.12922040596604348,
      "num_tokens": 19117379.0,
      "step": 5720
    },
    {
      "entropy": 6.5778416156768795,
      "epoch": 1.116866647156375,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004902581645170194,
      "loss": 6.3808,
      "mean_token_accuracy": 0.13638311997056007,
      "num_tokens": 19134751.0,
      "step": 5725
    },
    {
      "entropy": 6.545024776458741,
      "epoch": 1.117842161740318,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004902376821113319,
      "loss": 6.4504,
      "mean_token_accuracy": 0.12631849870085715,
      "num_tokens": 19150647.0,
      "step": 5730
    },
    {
      "entropy": 6.480191850662232,
      "epoch": 1.118817676324261,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004902171786733852,
      "loss": 6.4107,
      "mean_token_accuracy": 0.1286068968474865,
      "num_tokens": 19166654.0,
      "step": 5735
    },
    {
      "entropy": 6.523336172103882,
      "epoch": 1.1197931909082042,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901966542051828,
      "loss": 6.4392,
      "mean_token_accuracy": 0.13024272695183753,
      "num_tokens": 19182787.0,
      "step": 5740
    },
    {
      "entropy": 6.512229013442993,
      "epoch": 1.1207687054921471,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004901761087087303,
      "loss": 6.4109,
      "mean_token_accuracy": 0.12890372574329376,
      "num_tokens": 19200218.0,
      "step": 5745
    },
    {
      "entropy": 6.629920387268067,
      "epoch": 1.12174422007609,
      "grad_norm": 0.875,
      "learning_rate": 0.0004901555421860354,
      "loss": 6.497,
      "mean_token_accuracy": 0.12897952422499656,
      "num_tokens": 19217310.0,
      "step": 5750
    },
    {
      "entropy": 6.572010087966919,
      "epoch": 1.1227197346600333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901349546391078,
      "loss": 6.4623,
      "mean_token_accuracy": 0.1257151335477829,
      "num_tokens": 19233293.0,
      "step": 5755
    },
    {
      "entropy": 6.480763053894043,
      "epoch": 1.1236952492439762,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004901143460699592,
      "loss": 6.3645,
      "mean_token_accuracy": 0.14043401777744294,
      "num_tokens": 19250606.0,
      "step": 5760
    },
    {
      "entropy": 6.466905641555786,
      "epoch": 1.1246707638279192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900937164806033,
      "loss": 6.3344,
      "mean_token_accuracy": 0.1392199605703354,
      "num_tokens": 19267848.0,
      "step": 5765
    },
    {
      "entropy": 6.526986503601075,
      "epoch": 1.1256462784118622,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900730658730561,
      "loss": 6.4146,
      "mean_token_accuracy": 0.13172002509236336,
      "num_tokens": 19284317.0,
      "step": 5770
    },
    {
      "entropy": 6.487687492370606,
      "epoch": 1.1266217929958053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900523942493353,
      "loss": 6.3788,
      "mean_token_accuracy": 0.13198275491595268,
      "num_tokens": 19300766.0,
      "step": 5775
    },
    {
      "entropy": 6.520848274230957,
      "epoch": 1.1275973075797483,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004900317016114611,
      "loss": 6.3916,
      "mean_token_accuracy": 0.1295320600271225,
      "num_tokens": 19316966.0,
      "step": 5780
    },
    {
      "entropy": 6.520283985137939,
      "epoch": 1.1285728221636913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900109879614554,
      "loss": 6.463,
      "mean_token_accuracy": 0.12793975546956063,
      "num_tokens": 19334389.0,
      "step": 5785
    },
    {
      "entropy": 6.623872232437134,
      "epoch": 1.1295483367476344,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899902533013423,
      "loss": 6.4834,
      "mean_token_accuracy": 0.12829265296459197,
      "num_tokens": 19350238.0,
      "step": 5790
    },
    {
      "entropy": 6.593935012817383,
      "epoch": 1.1305238513315774,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004899694976331478,
      "loss": 6.426,
      "mean_token_accuracy": 0.12349902465939522,
      "num_tokens": 19367244.0,
      "step": 5795
    },
    {
      "entropy": 6.522238111495971,
      "epoch": 1.1314993659155204,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004899487209589001,
      "loss": 6.4325,
      "mean_token_accuracy": 0.12900010719895363,
      "num_tokens": 19384949.0,
      "step": 5800
    },
    {
      "entropy": 6.630327129364014,
      "epoch": 1.1324748804994635,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004899279232806297,
      "loss": 6.5138,
      "mean_token_accuracy": 0.12496284767985344,
      "num_tokens": 19402149.0,
      "step": 5805
    },
    {
      "entropy": 6.473363924026489,
      "epoch": 1.1334503950834065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899071046003686,
      "loss": 6.3966,
      "mean_token_accuracy": 0.12676906436681748,
      "num_tokens": 19419032.0,
      "step": 5810
    },
    {
      "entropy": 6.552075481414795,
      "epoch": 1.1344259096673495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004898862649201511,
      "loss": 6.4304,
      "mean_token_accuracy": 0.1319265268743038,
      "num_tokens": 19435656.0,
      "step": 5815
    },
    {
      "entropy": 6.435824251174926,
      "epoch": 1.1354014242512926,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898654042420137,
      "loss": 6.3384,
      "mean_token_accuracy": 0.1376349948346615,
      "num_tokens": 19451482.0,
      "step": 5820
    },
    {
      "entropy": 6.593602275848388,
      "epoch": 1.1363769388352356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898445225679948,
      "loss": 6.4966,
      "mean_token_accuracy": 0.12621192038059234,
      "num_tokens": 19467758.0,
      "step": 5825
    },
    {
      "entropy": 6.606624412536621,
      "epoch": 1.1373524534191786,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898236199001347,
      "loss": 6.4958,
      "mean_token_accuracy": 0.11921861469745636,
      "num_tokens": 19483673.0,
      "step": 5830
    },
    {
      "entropy": 6.601768589019775,
      "epoch": 1.1383279680031215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898026962404762,
      "loss": 6.5528,
      "mean_token_accuracy": 0.12128420546650887,
      "num_tokens": 19500171.0,
      "step": 5835
    },
    {
      "entropy": 6.528275823593139,
      "epoch": 1.1393034825870647,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004897817515910639,
      "loss": 6.3793,
      "mean_token_accuracy": 0.1357000634074211,
      "num_tokens": 19516531.0,
      "step": 5840
    },
    {
      "entropy": 6.5497400760650635,
      "epoch": 1.1402789971710077,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004897607859539443,
      "loss": 6.4922,
      "mean_token_accuracy": 0.12091487124562264,
      "num_tokens": 19534289.0,
      "step": 5845
    },
    {
      "entropy": 6.375201559066772,
      "epoch": 1.1412545117549509,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489739799331166,
      "loss": 6.2946,
      "mean_token_accuracy": 0.1370261035859585,
      "num_tokens": 19550190.0,
      "step": 5850
    },
    {
      "entropy": 6.423020458221435,
      "epoch": 1.1422300263388938,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048971879172478,
      "loss": 6.2188,
      "mean_token_accuracy": 0.14327089339494706,
      "num_tokens": 19567774.0,
      "step": 5855
    },
    {
      "entropy": 6.593882656097412,
      "epoch": 1.1432055409228368,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004896977631368389,
      "loss": 6.5077,
      "mean_token_accuracy": 0.12145421728491783,
      "num_tokens": 19583690.0,
      "step": 5860
    },
    {
      "entropy": 6.5250513553619385,
      "epoch": 1.1441810555067797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896767135693976,
      "loss": 6.3459,
      "mean_token_accuracy": 0.14140734896063806,
      "num_tokens": 19600563.0,
      "step": 5865
    },
    {
      "entropy": 6.454621171951294,
      "epoch": 1.145156570090723,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000489655643024513,
      "loss": 6.396,
      "mean_token_accuracy": 0.13324227258563043,
      "num_tokens": 19618043.0,
      "step": 5870
    },
    {
      "entropy": 6.544258689880371,
      "epoch": 1.1461320846746659,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004896345515042439,
      "loss": 6.4622,
      "mean_token_accuracy": 0.13432404026389122,
      "num_tokens": 19634807.0,
      "step": 5875
    },
    {
      "entropy": 6.472620916366577,
      "epoch": 1.1471075992586088,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896134390106515,
      "loss": 6.3119,
      "mean_token_accuracy": 0.1383487582206726,
      "num_tokens": 19650814.0,
      "step": 5880
    },
    {
      "entropy": 6.3629385471344,
      "epoch": 1.148083113842552,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004895923055457987,
      "loss": 6.2487,
      "mean_token_accuracy": 0.1478450946509838,
      "num_tokens": 19666293.0,
      "step": 5885
    },
    {
      "entropy": 6.426418876647949,
      "epoch": 1.149058628426495,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895711511117506,
      "loss": 6.3207,
      "mean_token_accuracy": 0.13418971970677376,
      "num_tokens": 19682549.0,
      "step": 5890
    },
    {
      "entropy": 6.481686401367187,
      "epoch": 1.150034143010438,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004895499757105743,
      "loss": 6.4317,
      "mean_token_accuracy": 0.12693503499031067,
      "num_tokens": 19697776.0,
      "step": 5895
    },
    {
      "entropy": 6.47916955947876,
      "epoch": 1.1510096575943811,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004895287793443391,
      "loss": 6.3608,
      "mean_token_accuracy": 0.1379731446504593,
      "num_tokens": 19715219.0,
      "step": 5900
    },
    {
      "entropy": 6.403602218627929,
      "epoch": 1.151985172178324,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004895075620151162,
      "loss": 6.4056,
      "mean_token_accuracy": 0.13419408798217775,
      "num_tokens": 19731934.0,
      "step": 5905
    },
    {
      "entropy": 6.493890953063965,
      "epoch": 1.152960686762267,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894863237249789,
      "loss": 6.3884,
      "mean_token_accuracy": 0.13869982734322547,
      "num_tokens": 19748933.0,
      "step": 5910
    },
    {
      "entropy": 6.5359241485595705,
      "epoch": 1.1539362013462102,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004894650644760024,
      "loss": 6.4143,
      "mean_token_accuracy": 0.13316623717546464,
      "num_tokens": 19765430.0,
      "step": 5915
    },
    {
      "entropy": 6.523070335388184,
      "epoch": 1.1549117159301532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894437842702642,
      "loss": 6.4744,
      "mean_token_accuracy": 0.1271375022828579,
      "num_tokens": 19783626.0,
      "step": 5920
    },
    {
      "entropy": 6.610344886779785,
      "epoch": 1.1558872305140961,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894224831098437,
      "loss": 6.466,
      "mean_token_accuracy": 0.12498205229640007,
      "num_tokens": 19800269.0,
      "step": 5925
    },
    {
      "entropy": 6.5292316436767575,
      "epoch": 1.156862745098039,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894011609968224,
      "loss": 6.4026,
      "mean_token_accuracy": 0.1277093768119812,
      "num_tokens": 19816837.0,
      "step": 5930
    },
    {
      "entropy": 6.610481071472168,
      "epoch": 1.1578382596819823,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893798179332837,
      "loss": 6.568,
      "mean_token_accuracy": 0.12244465425610543,
      "num_tokens": 19833029.0,
      "step": 5935
    },
    {
      "entropy": 6.460890054702759,
      "epoch": 1.1588137742659252,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004893584539213134,
      "loss": 6.3109,
      "mean_token_accuracy": 0.1429521732032299,
      "num_tokens": 19851600.0,
      "step": 5940
    },
    {
      "entropy": 6.504047298431397,
      "epoch": 1.1597892888498684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004893370689629988,
      "loss": 6.4272,
      "mean_token_accuracy": 0.13042818307876586,
      "num_tokens": 19867152.0,
      "step": 5945
    },
    {
      "entropy": 6.448392391204834,
      "epoch": 1.1607648034338114,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004893156630604298,
      "loss": 6.2763,
      "mean_token_accuracy": 0.14140164107084274,
      "num_tokens": 19884035.0,
      "step": 5950
    },
    {
      "entropy": 6.5272400856018065,
      "epoch": 1.1617403180177543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000489294236215698,
      "loss": 6.4117,
      "mean_token_accuracy": 0.12947047725319863,
      "num_tokens": 19899186.0,
      "step": 5955
    },
    {
      "entropy": 6.5434812068939205,
      "epoch": 1.1627158326016973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004892727884308972,
      "loss": 6.4416,
      "mean_token_accuracy": 0.12471651136875153,
      "num_tokens": 19914162.0,
      "step": 5960
    },
    {
      "entropy": 6.485411167144775,
      "epoch": 1.1636913471856405,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004892513197081233,
      "loss": 6.3729,
      "mean_token_accuracy": 0.1261335827410221,
      "num_tokens": 19931758.0,
      "step": 5965
    },
    {
      "entropy": 6.515101909637451,
      "epoch": 1.1646668617695835,
      "grad_norm": 0.90625,
      "learning_rate": 0.000489229830049474,
      "loss": 6.4645,
      "mean_token_accuracy": 0.1307239666581154,
      "num_tokens": 19949397.0,
      "step": 5970
    },
    {
      "entropy": 6.490616035461426,
      "epoch": 1.1656423763535264,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004892083194570492,
      "loss": 6.3626,
      "mean_token_accuracy": 0.13046735599637033,
      "num_tokens": 19967356.0,
      "step": 5975
    },
    {
      "entropy": 6.499018001556396,
      "epoch": 1.1666178909374696,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004891867879329509,
      "loss": 6.3074,
      "mean_token_accuracy": 0.13772637844085694,
      "num_tokens": 19984343.0,
      "step": 5980
    },
    {
      "entropy": 6.602190446853638,
      "epoch": 1.1675934055214126,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004891652354792831,
      "loss": 6.4823,
      "mean_token_accuracy": 0.12585388123989105,
      "num_tokens": 20001600.0,
      "step": 5985
    },
    {
      "entropy": 6.521233415603637,
      "epoch": 1.1685689201053555,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891436620981517,
      "loss": 6.422,
      "mean_token_accuracy": 0.12636670768260955,
      "num_tokens": 20016816.0,
      "step": 5990
    },
    {
      "entropy": 6.606795406341552,
      "epoch": 1.1695444346892987,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004891220677916649,
      "loss": 6.4387,
      "mean_token_accuracy": 0.1279299572110176,
      "num_tokens": 20033754.0,
      "step": 5995
    },
    {
      "entropy": 6.374530363082886,
      "epoch": 1.1705199492732417,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891004525619327,
      "loss": 6.3627,
      "mean_token_accuracy": 0.13847156763076782,
      "num_tokens": 20051899.0,
      "step": 6000
    },
    {
      "epoch": 1.1705199492732417,
      "eval_entropy": 6.3350197354549955,
      "eval_loss": 6.526286602020264,
      "eval_mean_token_accuracy": 0.13019166634859697,
      "eval_num_tokens": 20051899.0,
      "eval_runtime": 15.2729,
      "eval_samples_per_second": 1085.385,
      "eval_steps_per_second": 135.73,
      "step": 6000
    },
    {
      "entropy": 6.517582988739013,
      "epoch": 1.1714954638571846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890788164110674,
      "loss": 6.3761,
      "mean_token_accuracy": 0.13014244064688682,
      "num_tokens": 20069066.0,
      "step": 6005
    },
    {
      "entropy": 6.522425937652588,
      "epoch": 1.1724709784411278,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004890571593411833,
      "loss": 6.4092,
      "mean_token_accuracy": 0.12778158709406853,
      "num_tokens": 20085482.0,
      "step": 6010
    },
    {
      "entropy": 6.568148612976074,
      "epoch": 1.1734464930250708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890354813543963,
      "loss": 6.471,
      "mean_token_accuracy": 0.129772612452507,
      "num_tokens": 20101591.0,
      "step": 6015
    },
    {
      "entropy": 6.497191953659057,
      "epoch": 1.1744220076090137,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000489013782452825,
      "loss": 6.4095,
      "mean_token_accuracy": 0.12775073796510697,
      "num_tokens": 20118261.0,
      "step": 6020
    },
    {
      "entropy": 6.457738828659058,
      "epoch": 1.1753975221929567,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004889920626385896,
      "loss": 6.3814,
      "mean_token_accuracy": 0.13284848928451537,
      "num_tokens": 20133552.0,
      "step": 6025
    },
    {
      "entropy": 6.361282014846802,
      "epoch": 1.1763730367768999,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004889703219138125,
      "loss": 6.2092,
      "mean_token_accuracy": 0.14395120814442636,
      "num_tokens": 20150237.0,
      "step": 6030
    },
    {
      "entropy": 6.549466991424561,
      "epoch": 1.1773485513608428,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004889485602806181,
      "loss": 6.4372,
      "mean_token_accuracy": 0.13249493837356568,
      "num_tokens": 20166926.0,
      "step": 6035
    },
    {
      "entropy": 6.499787425994873,
      "epoch": 1.1783240659447858,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889267777411331,
      "loss": 6.4602,
      "mean_token_accuracy": 0.13261881247162818,
      "num_tokens": 20184507.0,
      "step": 6040
    },
    {
      "entropy": 6.491583204269409,
      "epoch": 1.179299580528729,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004889049742974858,
      "loss": 6.376,
      "mean_token_accuracy": 0.13774136826395988,
      "num_tokens": 20200647.0,
      "step": 6045
    },
    {
      "entropy": 6.510242033004761,
      "epoch": 1.180275095112672,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888831499518066,
      "loss": 6.4192,
      "mean_token_accuracy": 0.13291724920272827,
      "num_tokens": 20218634.0,
      "step": 6050
    },
    {
      "entropy": 6.516262865066528,
      "epoch": 1.1812506096966149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004888613047062284,
      "loss": 6.4661,
      "mean_token_accuracy": 0.12564768195152282,
      "num_tokens": 20234641.0,
      "step": 6055
    },
    {
      "entropy": 6.547676849365234,
      "epoch": 1.182226124280558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004888394385628857,
      "loss": 6.4521,
      "mean_token_accuracy": 0.11936526522040367,
      "num_tokens": 20251973.0,
      "step": 6060
    },
    {
      "entropy": 6.4963977336883545,
      "epoch": 1.183201638864501,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004888175515239152,
      "loss": 6.3704,
      "mean_token_accuracy": 0.13320100009441377,
      "num_tokens": 20270300.0,
      "step": 6065
    },
    {
      "entropy": 6.48465666770935,
      "epoch": 1.184177153448444,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004887956435914556,
      "loss": 6.411,
      "mean_token_accuracy": 0.13291623890399934,
      "num_tokens": 20286887.0,
      "step": 6070
    },
    {
      "entropy": 6.5485694885253904,
      "epoch": 1.1851526680323872,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887737147676477,
      "loss": 6.468,
      "mean_token_accuracy": 0.12563446313142776,
      "num_tokens": 20302859.0,
      "step": 6075
    },
    {
      "entropy": 6.540319108963013,
      "epoch": 1.1861281826163301,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004887517650546344,
      "loss": 6.4061,
      "mean_token_accuracy": 0.13909591138362884,
      "num_tokens": 20319863.0,
      "step": 6080
    },
    {
      "entropy": 6.49021692276001,
      "epoch": 1.187103697200273,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004887297944545603,
      "loss": 6.4159,
      "mean_token_accuracy": 0.1312604546546936,
      "num_tokens": 20336134.0,
      "step": 6085
    },
    {
      "entropy": 6.397997903823852,
      "epoch": 1.188079211784216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004887078029695725,
      "loss": 6.2829,
      "mean_token_accuracy": 0.14183085262775422,
      "num_tokens": 20353171.0,
      "step": 6090
    },
    {
      "entropy": 6.518901538848877,
      "epoch": 1.1890547263681592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004886857906018198,
      "loss": 6.3733,
      "mean_token_accuracy": 0.13219263553619384,
      "num_tokens": 20368464.0,
      "step": 6095
    },
    {
      "entropy": 6.4712423324584964,
      "epoch": 1.1900302409521022,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004886637573534534,
      "loss": 6.3882,
      "mean_token_accuracy": 0.13088551089167594,
      "num_tokens": 20383842.0,
      "step": 6100
    },
    {
      "entropy": 6.4915977954864506,
      "epoch": 1.1910057555360454,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488641703226626,
      "loss": 6.3672,
      "mean_token_accuracy": 0.13570242896676063,
      "num_tokens": 20399503.0,
      "step": 6105
    },
    {
      "entropy": 6.436387538909912,
      "epoch": 1.1919812701199883,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004886196282234927,
      "loss": 6.3128,
      "mean_token_accuracy": 0.1343281053006649,
      "num_tokens": 20416012.0,
      "step": 6110
    },
    {
      "entropy": 6.574536514282227,
      "epoch": 1.1929567847039313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885975323462109,
      "loss": 6.4346,
      "mean_token_accuracy": 0.12393123134970666,
      "num_tokens": 20432092.0,
      "step": 6115
    },
    {
      "entropy": 6.463175916671753,
      "epoch": 1.1939322992878743,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004885754155969393,
      "loss": 6.3957,
      "mean_token_accuracy": 0.13624346032738685,
      "num_tokens": 20449388.0,
      "step": 6120
    },
    {
      "entropy": 6.457948970794678,
      "epoch": 1.1949078138718174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885532779778394,
      "loss": 6.3517,
      "mean_token_accuracy": 0.13808408454060556,
      "num_tokens": 20464851.0,
      "step": 6125
    },
    {
      "entropy": 6.564535188674927,
      "epoch": 1.1958833284557604,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004885311194910743,
      "loss": 6.4234,
      "mean_token_accuracy": 0.1287480980157852,
      "num_tokens": 20481346.0,
      "step": 6130
    },
    {
      "entropy": 6.513142204284668,
      "epoch": 1.1968588430397034,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885089401388093,
      "loss": 6.4687,
      "mean_token_accuracy": 0.12830491960048676,
      "num_tokens": 20498373.0,
      "step": 6135
    },
    {
      "entropy": 6.503679180145264,
      "epoch": 1.1978343576236465,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004884867399232116,
      "loss": 6.3887,
      "mean_token_accuracy": 0.12817805334925653,
      "num_tokens": 20515375.0,
      "step": 6140
    },
    {
      "entropy": 6.487637662887574,
      "epoch": 1.1988098722075895,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004884645188464505,
      "loss": 6.3992,
      "mean_token_accuracy": 0.12687078565359117,
      "num_tokens": 20531593.0,
      "step": 6145
    },
    {
      "entropy": 6.600306367874145,
      "epoch": 1.1997853867915325,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004884422769106975,
      "loss": 6.3968,
      "mean_token_accuracy": 0.13056534975767137,
      "num_tokens": 20548668.0,
      "step": 6150
    },
    {
      "entropy": 6.580540609359741,
      "epoch": 1.2007609013754756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004884200141181258,
      "loss": 6.4839,
      "mean_token_accuracy": 0.12142401412129403,
      "num_tokens": 20566772.0,
      "step": 6155
    },
    {
      "entropy": 6.471336460113525,
      "epoch": 1.2017364159594186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488397730470911,
      "loss": 6.3745,
      "mean_token_accuracy": 0.13110021874308586,
      "num_tokens": 20582557.0,
      "step": 6160
    },
    {
      "entropy": 6.5070556640625,
      "epoch": 1.2027119305433616,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004883754259712305,
      "loss": 6.4012,
      "mean_token_accuracy": 0.12743562534451486,
      "num_tokens": 20598516.0,
      "step": 6165
    },
    {
      "entropy": 6.479752445220948,
      "epoch": 1.2036874451273047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000488353100621264,
      "loss": 6.3938,
      "mean_token_accuracy": 0.12608765289187432,
      "num_tokens": 20613773.0,
      "step": 6170
    },
    {
      "entropy": 6.4255053997039795,
      "epoch": 1.2046629597112477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004883307544231928,
      "loss": 6.4116,
      "mean_token_accuracy": 0.1298497013747692,
      "num_tokens": 20631141.0,
      "step": 6175
    },
    {
      "entropy": 6.532848978042603,
      "epoch": 1.2056384742951907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004883083873792007,
      "loss": 6.369,
      "mean_token_accuracy": 0.13675401210784913,
      "num_tokens": 20648408.0,
      "step": 6180
    },
    {
      "entropy": 6.469848394393921,
      "epoch": 1.2066139888791336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882859994914731,
      "loss": 6.4291,
      "mean_token_accuracy": 0.1344759866595268,
      "num_tokens": 20666418.0,
      "step": 6185
    },
    {
      "entropy": 6.502883577346802,
      "epoch": 1.2075895034630768,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048826359076219785,
      "loss": 6.333,
      "mean_token_accuracy": 0.13504263162612914,
      "num_tokens": 20682795.0,
      "step": 6190
    },
    {
      "entropy": 6.522675704956055,
      "epoch": 1.2085650180470198,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004882411611935646,
      "loss": 6.4875,
      "mean_token_accuracy": 0.12914504259824752,
      "num_tokens": 20700964.0,
      "step": 6195
    },
    {
      "entropy": 6.49043550491333,
      "epoch": 1.209540532630963,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004882187107877651,
      "loss": 6.3277,
      "mean_token_accuracy": 0.13526419922709465,
      "num_tokens": 20717288.0,
      "step": 6200
    },
    {
      "entropy": 6.500928783416748,
      "epoch": 1.210516047214906,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048819623954699307,
      "loss": 6.3986,
      "mean_token_accuracy": 0.12552920132875442,
      "num_tokens": 20733859.0,
      "step": 6205
    },
    {
      "entropy": 6.4898253917694095,
      "epoch": 1.2114915617988489,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004881737474734444,
      "loss": 6.4532,
      "mean_token_accuracy": 0.13023871183395386,
      "num_tokens": 20751467.0,
      "step": 6210
    },
    {
      "entropy": 6.566570854187011,
      "epoch": 1.2124670763827918,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048815123456931686,
      "loss": 6.4681,
      "mean_token_accuracy": 0.12612128555774688,
      "num_tokens": 20768019.0,
      "step": 6215
    },
    {
      "entropy": 6.532421255111695,
      "epoch": 1.213442590966735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004881287008368104,
      "loss": 6.4737,
      "mean_token_accuracy": 0.12471134960651398,
      "num_tokens": 20785007.0,
      "step": 6220
    },
    {
      "entropy": 6.518846559524536,
      "epoch": 1.214418105550678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048810614627812685,
      "loss": 6.3769,
      "mean_token_accuracy": 0.1312802292406559,
      "num_tokens": 20801693.0,
      "step": 6225
    },
    {
      "entropy": 6.3213340759277346,
      "epoch": 1.215393620134621,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004880835708954702,
      "loss": 6.2264,
      "mean_token_accuracy": 0.14361219108104706,
      "num_tokens": 20819866.0,
      "step": 6230
    },
    {
      "entropy": 6.559712505340576,
      "epoch": 1.2163691347185641,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880609746910465,
      "loss": 6.4248,
      "mean_token_accuracy": 0.13282878175377846,
      "num_tokens": 20836228.0,
      "step": 6235
    },
    {
      "entropy": 6.478572654724121,
      "epoch": 1.217344649302507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048803835766706366,
      "loss": 6.3587,
      "mean_token_accuracy": 0.12742669135332108,
      "num_tokens": 20853489.0,
      "step": 6240
    },
    {
      "entropy": 6.202415800094604,
      "epoch": 1.21832016388645,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048801571982573185,
      "loss": 6.1402,
      "mean_token_accuracy": 0.1569557748734951,
      "num_tokens": 20870627.0,
      "step": 6245
    },
    {
      "entropy": 6.583403253555298,
      "epoch": 1.2192956784703932,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048799306116926306,
      "loss": 6.4301,
      "mean_token_accuracy": 0.13245616033673285,
      "num_tokens": 20887277.0,
      "step": 6250
    },
    {
      "entropy": 6.554194784164428,
      "epoch": 1.2202711930543362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048797038169987153,
      "loss": 6.4042,
      "mean_token_accuracy": 0.12365722507238389,
      "num_tokens": 20902603.0,
      "step": 6255
    },
    {
      "entropy": 6.475480318069458,
      "epoch": 1.2212467076382791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048794768141977317,
      "loss": 6.3802,
      "mean_token_accuracy": 0.13093633353710174,
      "num_tokens": 20918700.0,
      "step": 6260
    },
    {
      "entropy": 6.450595569610596,
      "epoch": 1.2222222222222223,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004879249603311865,
      "loss": 6.3,
      "mean_token_accuracy": 0.13979081362485885,
      "num_tokens": 20936437.0,
      "step": 6265
    },
    {
      "entropy": 6.455233955383301,
      "epoch": 1.2231977368061653,
      "grad_norm": 1.0,
      "learning_rate": 0.00048790221843633154,
      "loss": 6.314,
      "mean_token_accuracy": 0.14386312514543534,
      "num_tokens": 20954984.0,
      "step": 6270
    },
    {
      "entropy": 6.373082160949707,
      "epoch": 1.2241732513901082,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004878794557374306,
      "loss": 6.3341,
      "mean_token_accuracy": 0.1425437279045582,
      "num_tokens": 20971119.0,
      "step": 6275
    },
    {
      "entropy": 6.486831378936768,
      "epoch": 1.2251487659740512,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004878566722367079,
      "loss": 6.3637,
      "mean_token_accuracy": 0.13986518383026122,
      "num_tokens": 20988211.0,
      "step": 6280
    },
    {
      "entropy": 6.477594947814941,
      "epoch": 1.2261242805579944,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004878338679363899,
      "loss": 6.3671,
      "mean_token_accuracy": 0.1345867894589901,
      "num_tokens": 21004240.0,
      "step": 6285
    },
    {
      "entropy": 6.542951965332032,
      "epoch": 1.2270997951419373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004878110428387048,
      "loss": 6.4567,
      "mean_token_accuracy": 0.12588425427675248,
      "num_tokens": 21021977.0,
      "step": 6290
    },
    {
      "entropy": 6.449786520004272,
      "epoch": 1.2280753097258805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048778819694588317,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1377550482749939,
      "num_tokens": 21038422.0,
      "step": 6295
    },
    {
      "entropy": 6.308367586135864,
      "epoch": 1.2290508243098235,
      "grad_norm": 0.875,
      "learning_rate": 0.00048776533026015725,
      "loss": 6.1748,
      "mean_token_accuracy": 0.15767862722277642,
      "num_tokens": 21055148.0,
      "step": 6300
    },
    {
      "entropy": 6.4515214443206785,
      "epoch": 1.2300263388937664,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048774244278376164,
      "loss": 6.4056,
      "mean_token_accuracy": 0.1356904037296772,
      "num_tokens": 21072137.0,
      "step": 6305
    },
    {
      "entropy": 6.4791769027709964,
      "epoch": 1.2310018534777094,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004877195345189326,
      "loss": 6.3403,
      "mean_token_accuracy": 0.13196767270565032,
      "num_tokens": 21089012.0,
      "step": 6310
    },
    {
      "entropy": 6.513841772079468,
      "epoch": 1.2319773680616526,
      "grad_norm": 1.0,
      "learning_rate": 0.00048769660546790896,
      "loss": 6.3538,
      "mean_token_accuracy": 0.13087425231933594,
      "num_tokens": 21105274.0,
      "step": 6315
    },
    {
      "entropy": 6.501277685165405,
      "epoch": 1.2329528826455955,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048767365563293097,
      "loss": 6.4576,
      "mean_token_accuracy": 0.13015715256333352,
      "num_tokens": 21122784.0,
      "step": 6320
    },
    {
      "entropy": 6.506662178039551,
      "epoch": 1.2339283972295385,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004876506850162414,
      "loss": 6.3095,
      "mean_token_accuracy": 0.13502025455236435,
      "num_tokens": 21139829.0,
      "step": 6325
    },
    {
      "entropy": 6.407135629653931,
      "epoch": 1.2349039118134817,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048762769362008473,
      "loss": 6.3021,
      "mean_token_accuracy": 0.1335664927959442,
      "num_tokens": 21155095.0,
      "step": 6330
    },
    {
      "entropy": 6.495771026611328,
      "epoch": 1.2358794263974247,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004876046814467077,
      "loss": 6.4504,
      "mean_token_accuracy": 0.13156534880399703,
      "num_tokens": 21171865.0,
      "step": 6335
    },
    {
      "entropy": 6.607894706726074,
      "epoch": 1.2368549409813676,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487581648498359,
      "loss": 6.43,
      "mean_token_accuracy": 0.12480097189545632,
      "num_tokens": 21188757.0,
      "step": 6340
    },
    {
      "entropy": 6.473166656494141,
      "epoch": 1.2378304555653108,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004875585947772892,
      "loss": 6.3963,
      "mean_token_accuracy": 0.13542184457182885,
      "num_tokens": 21204324.0,
      "step": 6345
    },
    {
      "entropy": 6.44293851852417,
      "epoch": 1.2388059701492538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048753552028575105,
      "loss": 6.3374,
      "mean_token_accuracy": 0.1366824612021446,
      "num_tokens": 21219836.0,
      "step": 6350
    },
    {
      "entropy": 6.487629032135009,
      "epoch": 1.2397814847331967,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004875124250259994,
      "loss": 6.4172,
      "mean_token_accuracy": 0.1237720526754856,
      "num_tokens": 21235848.0,
      "step": 6355
    },
    {
      "entropy": 6.429219341278076,
      "epoch": 1.24075699931714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048748930900029103,
      "loss": 6.3699,
      "mean_token_accuracy": 0.14345087856054306,
      "num_tokens": 21252779.0,
      "step": 6360
    },
    {
      "entropy": 6.460458898544312,
      "epoch": 1.2417325139010829,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048746617221088464,
      "loss": 6.3398,
      "mean_token_accuracy": 0.13419901058077813,
      "num_tokens": 21269004.0,
      "step": 6365
    },
    {
      "entropy": 6.527614402770996,
      "epoch": 1.2427080284850258,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048744301466004114,
      "loss": 6.4422,
      "mean_token_accuracy": 0.13199608027935028,
      "num_tokens": 21285769.0,
      "step": 6370
    },
    {
      "entropy": 6.495567989349365,
      "epoch": 1.2436835430689688,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004874198363500235,
      "loss": 6.4756,
      "mean_token_accuracy": 0.1251590333878994,
      "num_tokens": 21301986.0,
      "step": 6375
    },
    {
      "entropy": 6.646597194671631,
      "epoch": 1.244659057652912,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004873966372830964,
      "loss": 6.5737,
      "mean_token_accuracy": 0.1205462396144867,
      "num_tokens": 21318755.0,
      "step": 6380
    },
    {
      "entropy": 6.322873592376709,
      "epoch": 1.245634572236855,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000487373417461527,
      "loss": 6.2063,
      "mean_token_accuracy": 0.1518035389482975,
      "num_tokens": 21335188.0,
      "step": 6385
    },
    {
      "entropy": 6.485266256332397,
      "epoch": 1.2466100868207979,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048735017688758406,
      "loss": 6.4133,
      "mean_token_accuracy": 0.1303052619099617,
      "num_tokens": 21351976.0,
      "step": 6390
    },
    {
      "entropy": 6.549084901809692,
      "epoch": 1.247585601404741,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004873269155635387,
      "loss": 6.434,
      "mean_token_accuracy": 0.12727439776062965,
      "num_tokens": 21369043.0,
      "step": 6395
    },
    {
      "entropy": 6.509441471099853,
      "epoch": 1.248561115988684,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048730363349166393,
      "loss": 6.3947,
      "mean_token_accuracy": 0.12976840883493423,
      "num_tokens": 21386583.0,
      "step": 6400
    },
    {
      "entropy": 6.566401052474975,
      "epoch": 1.249536630572627,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004872803306742348,
      "loss": 6.3612,
      "mean_token_accuracy": 0.1340057797729969,
      "num_tokens": 21403455.0,
      "step": 6405
    },
    {
      "entropy": 6.414584636688232,
      "epoch": 1.2505121451565702,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872570071135282,
      "loss": 6.4392,
      "mean_token_accuracy": 0.1267557017505169,
      "num_tokens": 21420107.0,
      "step": 6410
    },
    {
      "entropy": 6.575495481491089,
      "epoch": 1.2514876597405131,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048723366281182344,
      "loss": 6.4366,
      "mean_token_accuracy": 0.12359663397073746,
      "num_tokens": 21436629.0,
      "step": 6415
    },
    {
      "entropy": 6.4896917819976805,
      "epoch": 1.252463174324456,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004872102977714016,
      "loss": 6.3879,
      "mean_token_accuracy": 0.13115460202097892,
      "num_tokens": 21454364.0,
      "step": 6420
    },
    {
      "entropy": 6.424359607696533,
      "epoch": 1.2534386889083993,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048718691199454577,
      "loss": 6.3177,
      "mean_token_accuracy": 0.1353354625403881,
      "num_tokens": 21472056.0,
      "step": 6425
    },
    {
      "entropy": 6.537180614471436,
      "epoch": 1.2544142034923422,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871635054835412,
      "loss": 6.4708,
      "mean_token_accuracy": 0.12270632013678551,
      "num_tokens": 21489125.0,
      "step": 6430
    },
    {
      "entropy": 6.519209003448486,
      "epoch": 1.2553897180762852,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004871400782406749,
      "loss": 6.4669,
      "mean_token_accuracy": 0.12481142729520797,
      "num_tokens": 21505772.0,
      "step": 6435
    },
    {
      "entropy": 6.480989694595337,
      "epoch": 1.2563652326602281,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004871166302682363,
      "loss": 6.357,
      "mean_token_accuracy": 0.13269458934664727,
      "num_tokens": 21522628.0,
      "step": 6440
    },
    {
      "entropy": 6.352712774276734,
      "epoch": 1.2573407472441713,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004870931615685166,
      "loss": 6.3393,
      "mean_token_accuracy": 0.13586879894137383,
      "num_tokens": 21540379.0,
      "step": 6445
    },
    {
      "entropy": 6.558028078079223,
      "epoch": 1.2583162618281143,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004870696721438091,
      "loss": 6.4033,
      "mean_token_accuracy": 0.12506309151649475,
      "num_tokens": 21556698.0,
      "step": 6450
    },
    {
      "entropy": 6.510939836502075,
      "epoch": 1.2592917764120575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048704616199640906,
      "loss": 6.335,
      "mean_token_accuracy": 0.13370306864380838,
      "num_tokens": 21573005.0,
      "step": 6455
    },
    {
      "entropy": 6.402139711380005,
      "epoch": 1.2602672909960004,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048702263112861377,
      "loss": 6.3369,
      "mean_token_accuracy": 0.1417384311556816,
      "num_tokens": 21591171.0,
      "step": 6460
    },
    {
      "entropy": 6.534041118621826,
      "epoch": 1.2612428055799434,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004869990795427226,
      "loss": 6.347,
      "mean_token_accuracy": 0.13342896476387978,
      "num_tokens": 21606912.0,
      "step": 6465
    },
    {
      "entropy": 6.516573143005371,
      "epoch": 1.2622183201638864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000486975507241037,
      "loss": 6.4388,
      "mean_token_accuracy": 0.13050373420119285,
      "num_tokens": 21623392.0,
      "step": 6470
    },
    {
      "entropy": 6.528912353515625,
      "epoch": 1.2631938347478295,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004869519142258604,
      "loss": 6.4162,
      "mean_token_accuracy": 0.13276160955429078,
      "num_tokens": 21639706.0,
      "step": 6475
    },
    {
      "entropy": 6.3175435066223145,
      "epoch": 1.2641693493317725,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048692830049949803,
      "loss": 6.2677,
      "mean_token_accuracy": 0.14349026829004288,
      "num_tokens": 21656487.0,
      "step": 6480
    },
    {
      "entropy": 6.515255498886108,
      "epoch": 1.2651448639157157,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004869046660642575,
      "loss": 6.3942,
      "mean_token_accuracy": 0.12897997424006463,
      "num_tokens": 21673950.0,
      "step": 6485
    },
    {
      "entropy": 6.506541109085083,
      "epoch": 1.2661203784996586,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048688101092244826,
      "loss": 6.4831,
      "mean_token_accuracy": 0.1253752291202545,
      "num_tokens": 21692296.0,
      "step": 6490
    },
    {
      "entropy": 6.504839468002319,
      "epoch": 1.2670958930836016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004868573350763818,
      "loss": 6.3646,
      "mean_token_accuracy": 0.13856419846415519,
      "num_tokens": 21710021.0,
      "step": 6495
    },
    {
      "entropy": 6.434336757659912,
      "epoch": 1.2680714076675446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004868336385283716,
      "loss": 6.3506,
      "mean_token_accuracy": 0.1358800359070301,
      "num_tokens": 21726660.0,
      "step": 6500
    },
    {
      "entropy": 6.34053225517273,
      "epoch": 1.2690469222514877,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048680992128073323,
      "loss": 6.2228,
      "mean_token_accuracy": 0.15104396417737007,
      "num_tokens": 21743840.0,
      "step": 6505
    },
    {
      "entropy": 6.512342834472657,
      "epoch": 1.2700224368354307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048678618333578434,
      "loss": 6.3506,
      "mean_token_accuracy": 0.12765319347381593,
      "num_tokens": 21761000.0,
      "step": 6510
    },
    {
      "entropy": 6.48812346458435,
      "epoch": 1.2709979514193737,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867624246958443,
      "loss": 6.4181,
      "mean_token_accuracy": 0.13297146707773208,
      "num_tokens": 21778564.0,
      "step": 6515
    },
    {
      "entropy": 6.51555700302124,
      "epoch": 1.2719734660033168,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004867386453632349,
      "loss": 6.3783,
      "mean_token_accuracy": 0.13109230995178223,
      "num_tokens": 21794780.0,
      "step": 6520
    },
    {
      "entropy": 6.488344478607178,
      "epoch": 1.2729489805872598,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004867148453402798,
      "loss": 6.4153,
      "mean_token_accuracy": 0.12596814185380936,
      "num_tokens": 21810692.0,
      "step": 6525
    },
    {
      "entropy": 6.492967557907105,
      "epoch": 1.2739244951712028,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004866910246293045,
      "loss": 6.3719,
      "mean_token_accuracy": 0.12811711877584459,
      "num_tokens": 21826488.0,
      "step": 6530
    },
    {
      "entropy": 6.4823057651519775,
      "epoch": 1.2749000097551457,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004866671832326368,
      "loss": 6.3934,
      "mean_token_accuracy": 0.13054317310452462,
      "num_tokens": 21842417.0,
      "step": 6535
    },
    {
      "entropy": 6.476323413848877,
      "epoch": 1.275875524339089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004866433211526063,
      "loss": 6.3638,
      "mean_token_accuracy": 0.12986616790294647,
      "num_tokens": 21858496.0,
      "step": 6540
    },
    {
      "entropy": 6.553343343734741,
      "epoch": 1.2768510389230319,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004866194383915448,
      "loss": 6.4227,
      "mean_token_accuracy": 0.12805749475955963,
      "num_tokens": 21874888.0,
      "step": 6545
    },
    {
      "entropy": 6.441355562210083,
      "epoch": 1.277826553506975,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000486595534951786,
      "loss": 6.2749,
      "mean_token_accuracy": 0.13781743869185448,
      "num_tokens": 21891789.0,
      "step": 6550
    },
    {
      "entropy": 6.494747161865234,
      "epoch": 1.278802068090918,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004865716108356656,
      "loss": 6.4038,
      "mean_token_accuracy": 0.13056811764836312,
      "num_tokens": 21907681.0,
      "step": 6555
    },
    {
      "entropy": 6.461083936691284,
      "epoch": 1.279777582674861,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004865476660455215,
      "loss": 6.3647,
      "mean_token_accuracy": 0.13029403761029243,
      "num_tokens": 21925852.0,
      "step": 6560
    },
    {
      "entropy": 6.487649154663086,
      "epoch": 1.280753097258804,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004865237005836935,
      "loss": 6.4205,
      "mean_token_accuracy": 0.13185539320111275,
      "num_tokens": 21942515.0,
      "step": 6565
    },
    {
      "entropy": 6.522689962387085,
      "epoch": 1.2817286118427471,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048649971445252334,
      "loss": 6.384,
      "mean_token_accuracy": 0.133242417126894,
      "num_tokens": 21961183.0,
      "step": 6570
    },
    {
      "entropy": 6.429194402694702,
      "epoch": 1.28270412642669,
      "grad_norm": 1.0,
      "learning_rate": 0.0004864757076543548,
      "loss": 6.4083,
      "mean_token_accuracy": 0.12840839326381684,
      "num_tokens": 21977806.0,
      "step": 6575
    },
    {
      "entropy": 6.419119548797608,
      "epoch": 1.283679641010633,
      "grad_norm": 1.0,
      "learning_rate": 0.00048645168019153384,
      "loss": 6.3432,
      "mean_token_accuracy": 0.1380515404045582,
      "num_tokens": 21994606.0,
      "step": 6580
    },
    {
      "entropy": 6.5930742740631105,
      "epoch": 1.2846551555945762,
      "grad_norm": 1.0,
      "learning_rate": 0.00048642763206640843,
      "loss": 6.5079,
      "mean_token_accuracy": 0.12351433411240578,
      "num_tokens": 22010386.0,
      "step": 6585
    },
    {
      "entropy": 6.397618198394776,
      "epoch": 1.2856306701785192,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004864035632813282,
      "loss": 6.2302,
      "mean_token_accuracy": 0.14473484233021736,
      "num_tokens": 22027310.0,
      "step": 6590
    },
    {
      "entropy": 6.485787200927734,
      "epoch": 1.2866061847624621,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004863794738386454,
      "loss": 6.4002,
      "mean_token_accuracy": 0.12446612939238548,
      "num_tokens": 22043040.0,
      "step": 6595
    },
    {
      "entropy": 6.40571436882019,
      "epoch": 1.287581699346405,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048635536374071366,
      "loss": 6.3153,
      "mean_token_accuracy": 0.1354794830083847,
      "num_tokens": 22059521.0,
      "step": 6600
    },
    {
      "entropy": 6.528609275817871,
      "epoch": 1.2885572139303483,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048633123298988905,
      "loss": 6.3877,
      "mean_token_accuracy": 0.12932892069220542,
      "num_tokens": 22075705.0,
      "step": 6605
    },
    {
      "entropy": 6.491176891326904,
      "epoch": 1.2895327285142912,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004863070815885296,
      "loss": 6.43,
      "mean_token_accuracy": 0.12789785265922546,
      "num_tokens": 22092366.0,
      "step": 6610
    },
    {
      "entropy": 6.507666492462159,
      "epoch": 1.2905082430982344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004862829095389952,
      "loss": 6.4038,
      "mean_token_accuracy": 0.12422064542770386,
      "num_tokens": 22108414.0,
      "step": 6615
    },
    {
      "entropy": 6.539453363418579,
      "epoch": 1.2914837576821774,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048625871684364796,
      "loss": 6.3613,
      "mean_token_accuracy": 0.13155285269021988,
      "num_tokens": 22123975.0,
      "step": 6620
    },
    {
      "entropy": 6.464390468597412,
      "epoch": 1.2924592722661203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004862345035048519,
      "loss": 6.411,
      "mean_token_accuracy": 0.13271818086504936,
      "num_tokens": 22139529.0,
      "step": 6625
    },
    {
      "entropy": 6.497458267211914,
      "epoch": 1.2934347868500633,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004862102695249729,
      "loss": 6.4148,
      "mean_token_accuracy": 0.1320466198027134,
      "num_tokens": 22155680.0,
      "step": 6630
    },
    {
      "entropy": 6.530553817749023,
      "epoch": 1.2944103014340065,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048618601490637913,
      "loss": 6.4829,
      "mean_token_accuracy": 0.12645730078220369,
      "num_tokens": 22172698.0,
      "step": 6635
    },
    {
      "entropy": 6.570889043807983,
      "epoch": 1.2953858160179494,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048616173965144077,
      "loss": 6.408,
      "mean_token_accuracy": 0.12714374735951423,
      "num_tokens": 22189745.0,
      "step": 6640
    },
    {
      "entropy": 6.525473117828369,
      "epoch": 1.2963613306018926,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048613744376252974,
      "loss": 6.4775,
      "mean_token_accuracy": 0.12685465663671494,
      "num_tokens": 22205931.0,
      "step": 6645
    },
    {
      "entropy": 6.427962827682495,
      "epoch": 1.2973368451858356,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004861131272420202,
      "loss": 6.2996,
      "mean_token_accuracy": 0.13532836586236954,
      "num_tokens": 22222232.0,
      "step": 6650
    },
    {
      "entropy": 6.521352863311767,
      "epoch": 1.2983123597697785,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004860887900922883,
      "loss": 6.4006,
      "mean_token_accuracy": 0.13412273079156875,
      "num_tokens": 22240082.0,
      "step": 6655
    },
    {
      "entropy": 6.361205959320069,
      "epoch": 1.2992878743537215,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004860644323157122,
      "loss": 6.3081,
      "mean_token_accuracy": 0.1378593161702156,
      "num_tokens": 22256105.0,
      "step": 6660
    },
    {
      "entropy": 6.487403392791748,
      "epoch": 1.3002633889376647,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004860400539146719,
      "loss": 6.304,
      "mean_token_accuracy": 0.14037666320800782,
      "num_tokens": 22273400.0,
      "step": 6665
    },
    {
      "entropy": 6.418055200576783,
      "epoch": 1.3012389035216076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004860156548915498,
      "loss": 6.2436,
      "mean_token_accuracy": 0.14154618307948114,
      "num_tokens": 22289322.0,
      "step": 6670
    },
    {
      "entropy": 6.439485216140747,
      "epoch": 1.3022144181055506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048599123524872996,
      "loss": 6.311,
      "mean_token_accuracy": 0.13532692044973374,
      "num_tokens": 22306182.0,
      "step": 6675
    },
    {
      "entropy": 6.465490388870239,
      "epoch": 1.3031899326894938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859667949885985,
      "loss": 6.4006,
      "mean_token_accuracy": 0.1307711645960808,
      "num_tokens": 22322761.0,
      "step": 6680
    },
    {
      "entropy": 6.443994569778442,
      "epoch": 1.3041654472734368,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048594233411354386,
      "loss": 6.3779,
      "mean_token_accuracy": 0.13484960198402404,
      "num_tokens": 22340224.0,
      "step": 6685
    },
    {
      "entropy": 6.490327024459839,
      "epoch": 1.3051409618573797,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004859178526259561,
      "loss": 6.2792,
      "mean_token_accuracy": 0.13698548153042794,
      "num_tokens": 22357070.0,
      "step": 6690
    },
    {
      "entropy": 6.417457199096679,
      "epoch": 1.3061164764413227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004858933505282276,
      "loss": 6.3822,
      "mean_token_accuracy": 0.13294109106063842,
      "num_tokens": 22372723.0,
      "step": 6695
    },
    {
      "entropy": 6.493600225448608,
      "epoch": 1.3070919910252659,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004858688278227524,
      "loss": 6.3386,
      "mean_token_accuracy": 0.132009294629097,
      "num_tokens": 22388486.0,
      "step": 6700
    },
    {
      "entropy": 6.528032827377319,
      "epoch": 1.3080675056092088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004858442845119269,
      "loss": 6.4674,
      "mean_token_accuracy": 0.12026055604219436,
      "num_tokens": 22405212.0,
      "step": 6705
    },
    {
      "entropy": 6.493104743957519,
      "epoch": 1.309043020193152,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004858197205981494,
      "loss": 6.4071,
      "mean_token_accuracy": 0.12899642959237098,
      "num_tokens": 22422899.0,
      "step": 6710
    },
    {
      "entropy": 6.489833164215088,
      "epoch": 1.310018534777095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004857951360838202,
      "loss": 6.3782,
      "mean_token_accuracy": 0.13312939926981926,
      "num_tokens": 22439702.0,
      "step": 6715
    },
    {
      "entropy": 6.49627685546875,
      "epoch": 1.310994049361038,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857705309713416,
      "loss": 6.4324,
      "mean_token_accuracy": 0.12790799513459206,
      "num_tokens": 22457725.0,
      "step": 6720
    },
    {
      "entropy": 6.348059177398682,
      "epoch": 1.3119695639449809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004857459052631179,
      "loss": 6.2835,
      "mean_token_accuracy": 0.1422308124601841,
      "num_tokens": 22475607.0,
      "step": 6725
    },
    {
      "entropy": 6.378645420074463,
      "epoch": 1.312945078528924,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048572125896155537,
      "loss": 6.2406,
      "mean_token_accuracy": 0.14349578246474265,
      "num_tokens": 22493818.0,
      "step": 6730
    },
    {
      "entropy": 6.4592865943908695,
      "epoch": 1.313920593112867,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048569659206906254,
      "loss": 6.44,
      "mean_token_accuracy": 0.12873822078108788,
      "num_tokens": 22509803.0,
      "step": 6735
    },
    {
      "entropy": 6.489472436904907,
      "epoch": 1.3148961076968102,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004856719045880496,
      "loss": 6.3834,
      "mean_token_accuracy": 0.13569658920168876,
      "num_tokens": 22525335.0,
      "step": 6740
    },
    {
      "entropy": 6.467029857635498,
      "epoch": 1.3158716222807532,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004856471965209291,
      "loss": 6.4051,
      "mean_token_accuracy": 0.12740842252969742,
      "num_tokens": 22542726.0,
      "step": 6745
    },
    {
      "entropy": 6.419431400299072,
      "epoch": 1.3168471368646961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856224678701152,
      "loss": 6.2436,
      "mean_token_accuracy": 0.14068242236971856,
      "num_tokens": 22559162.0,
      "step": 6750
    },
    {
      "entropy": 6.515065145492554,
      "epoch": 1.317822651448639,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004855977186380245,
      "loss": 6.4168,
      "mean_token_accuracy": 0.12543955370783805,
      "num_tokens": 22575247.0,
      "step": 6755
    },
    {
      "entropy": 6.490365934371948,
      "epoch": 1.3187981660325823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004855729488270752,
      "loss": 6.4207,
      "mean_token_accuracy": 0.1254555068910122,
      "num_tokens": 22591604.0,
      "step": 6760
    },
    {
      "entropy": 6.434759044647217,
      "epoch": 1.3197736806165252,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004855481584396879,
      "loss": 6.365,
      "mean_token_accuracy": 0.13470933511853217,
      "num_tokens": 22608095.0,
      "step": 6765
    },
    {
      "entropy": 6.501376008987426,
      "epoch": 1.3207491952004682,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048552334747828504,
      "loss": 6.4086,
      "mean_token_accuracy": 0.12524789273738862,
      "num_tokens": 22624090.0,
      "step": 6770
    },
    {
      "entropy": 6.55325665473938,
      "epoch": 1.3217247097844114,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048549851594529095,
      "loss": 6.4218,
      "mean_token_accuracy": 0.12992045730352403,
      "num_tokens": 22641593.0,
      "step": 6775
    },
    {
      "entropy": 6.520679950714111,
      "epoch": 1.3227002243683543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048547366384313204,
      "loss": 6.4757,
      "mean_token_accuracy": 0.11954193189740181,
      "num_tokens": 22657601.0,
      "step": 6780
    },
    {
      "entropy": 6.575753450393677,
      "epoch": 1.3236757389522973,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048544879117423683,
      "loss": 6.4954,
      "mean_token_accuracy": 0.12510077878832818,
      "num_tokens": 22674665.0,
      "step": 6785
    },
    {
      "entropy": 6.488282060623169,
      "epoch": 1.3246512535362402,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004854238979410359,
      "loss": 6.3183,
      "mean_token_accuracy": 0.13602573722600936,
      "num_tokens": 22692178.0,
      "step": 6790
    },
    {
      "entropy": 6.440552806854248,
      "epoch": 1.3256267681201834,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004853989841459616,
      "loss": 6.4193,
      "mean_token_accuracy": 0.12655006423592569,
      "num_tokens": 22709318.0,
      "step": 6795
    },
    {
      "entropy": 6.418773794174195,
      "epoch": 1.3266022827041264,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048537404979144837,
      "loss": 6.2839,
      "mean_token_accuracy": 0.1390961676836014,
      "num_tokens": 22725142.0,
      "step": 6800
    },
    {
      "entropy": 6.49222764968872,
      "epoch": 1.3275777972880696,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048534909487993284,
      "loss": 6.3345,
      "mean_token_accuracy": 0.14409189671278,
      "num_tokens": 22741494.0,
      "step": 6805
    },
    {
      "entropy": 6.4279327392578125,
      "epoch": 1.3285533118720125,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048532411941385346,
      "loss": 6.3288,
      "mean_token_accuracy": 0.133332322537899,
      "num_tokens": 22757232.0,
      "step": 6810
    },
    {
      "entropy": 6.572770404815674,
      "epoch": 1.3295288264559555,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048529912339565076,
      "loss": 6.4628,
      "mean_token_accuracy": 0.1213845394551754,
      "num_tokens": 22772846.0,
      "step": 6815
    },
    {
      "entropy": 6.461372804641724,
      "epoch": 1.3305043410398985,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004852741068277672,
      "loss": 6.3718,
      "mean_token_accuracy": 0.13029908165335655,
      "num_tokens": 22789712.0,
      "step": 6820
    },
    {
      "entropy": 6.47736759185791,
      "epoch": 1.3314798556238416,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048524906971264737,
      "loss": 6.3378,
      "mean_token_accuracy": 0.12452191710472107,
      "num_tokens": 22805743.0,
      "step": 6825
    },
    {
      "entropy": 6.430210924148559,
      "epoch": 1.3324553702077846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048522401205273774,
      "loss": 6.3435,
      "mean_token_accuracy": 0.13557306081056594,
      "num_tokens": 22822947.0,
      "step": 6830
    },
    {
      "entropy": 6.390811777114868,
      "epoch": 1.3334308847917276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004851989338504869,
      "loss": 6.2511,
      "mean_token_accuracy": 0.14938539564609526,
      "num_tokens": 22839767.0,
      "step": 6835
    },
    {
      "entropy": 6.446488380432129,
      "epoch": 1.3344063993756707,
      "grad_norm": 1.0,
      "learning_rate": 0.00048517383510834547,
      "loss": 6.3584,
      "mean_token_accuracy": 0.1364709384739399,
      "num_tokens": 22854523.0,
      "step": 6840
    },
    {
      "entropy": 6.396956491470337,
      "epoch": 1.3353819139596137,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004851487158287659,
      "loss": 6.2994,
      "mean_token_accuracy": 0.14119527861475945,
      "num_tokens": 22871580.0,
      "step": 6845
    },
    {
      "entropy": 6.399468564987183,
      "epoch": 1.3363574285435567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048512357601420287,
      "loss": 6.2608,
      "mean_token_accuracy": 0.13131451234221458,
      "num_tokens": 22887123.0,
      "step": 6850
    },
    {
      "entropy": 6.570219087600708,
      "epoch": 1.3373329431274996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850984156671128,
      "loss": 6.4254,
      "mean_token_accuracy": 0.13022970408201218,
      "num_tokens": 22905616.0,
      "step": 6855
    },
    {
      "entropy": 6.4309087753295895,
      "epoch": 1.3383084577114428,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004850732347899544,
      "loss": 6.2929,
      "mean_token_accuracy": 0.13558529764413835,
      "num_tokens": 22921829.0,
      "step": 6860
    },
    {
      "entropy": 6.33880124092102,
      "epoch": 1.3392839722953858,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004850480333851882,
      "loss": 6.2917,
      "mean_token_accuracy": 0.14221074283123017,
      "num_tokens": 22938766.0,
      "step": 6865
    },
    {
      "entropy": 6.419996452331543,
      "epoch": 1.340259486879329,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004850228114552768,
      "loss": 6.3801,
      "mean_token_accuracy": 0.13364962935447694,
      "num_tokens": 22956428.0,
      "step": 6870
    },
    {
      "entropy": 6.554489326477051,
      "epoch": 1.341235001463272,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048499756900268485,
      "loss": 6.4254,
      "mean_token_accuracy": 0.1261457197368145,
      "num_tokens": 22972538.0,
      "step": 6875
    },
    {
      "entropy": 6.585781002044678,
      "epoch": 1.3422105160472149,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004849723060298788,
      "loss": 6.4584,
      "mean_token_accuracy": 0.12874327674508096,
      "num_tokens": 22988015.0,
      "step": 6880
    },
    {
      "entropy": 6.494842386245727,
      "epoch": 1.3431860306311578,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048494702253932745,
      "loss": 6.436,
      "mean_token_accuracy": 0.12491165772080422,
      "num_tokens": 23004486.0,
      "step": 6885
    },
    {
      "entropy": 6.435236549377441,
      "epoch": 1.344161545215101,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048492171853350134,
      "loss": 6.2415,
      "mean_token_accuracy": 0.14194311574101448,
      "num_tokens": 23021583.0,
      "step": 6890
    },
    {
      "entropy": 6.465733909606934,
      "epoch": 1.345137059799044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048489639401487293,
      "loss": 6.3669,
      "mean_token_accuracy": 0.13492885157465934,
      "num_tokens": 23036920.0,
      "step": 6895
    },
    {
      "entropy": 6.3880791664123535,
      "epoch": 1.3461125743829871,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048487104898591716,
      "loss": 6.2465,
      "mean_token_accuracy": 0.14189494028687477,
      "num_tokens": 23054752.0,
      "step": 6900
    },
    {
      "entropy": 6.450745725631714,
      "epoch": 1.34708808896693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004848456834491103,
      "loss": 6.4129,
      "mean_token_accuracy": 0.12622534856200218,
      "num_tokens": 23069861.0,
      "step": 6905
    },
    {
      "entropy": 6.509674835205078,
      "epoch": 1.348063603550873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004848202974069313,
      "loss": 6.4056,
      "mean_token_accuracy": 0.13171408399939538,
      "num_tokens": 23086349.0,
      "step": 6910
    },
    {
      "entropy": 6.410039329528809,
      "epoch": 1.349039118134816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048479489086186055,
      "loss": 6.2735,
      "mean_token_accuracy": 0.13828462585806847,
      "num_tokens": 23103116.0,
      "step": 6915
    },
    {
      "entropy": 6.394093465805054,
      "epoch": 1.3500146327187592,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847694638163809,
      "loss": 6.2942,
      "mean_token_accuracy": 0.13265690729022026,
      "num_tokens": 23120505.0,
      "step": 6920
    },
    {
      "entropy": 6.479782915115356,
      "epoch": 1.3509901473027022,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004847440162729768,
      "loss": 6.3431,
      "mean_token_accuracy": 0.12901878729462624,
      "num_tokens": 23136751.0,
      "step": 6925
    },
    {
      "entropy": 6.378381967544556,
      "epoch": 1.3519656618866451,
      "grad_norm": 1.0,
      "learning_rate": 0.000484718548234135,
      "loss": 6.2566,
      "mean_token_accuracy": 0.1466432988643646,
      "num_tokens": 23153581.0,
      "step": 6930
    },
    {
      "entropy": 6.391969299316406,
      "epoch": 1.3529411764705883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000484693059702344,
      "loss": 6.2655,
      "mean_token_accuracy": 0.13894706442952157,
      "num_tokens": 23169650.0,
      "step": 6935
    },
    {
      "entropy": 6.526889705657959,
      "epoch": 1.3539166910545313,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004846675506800947,
      "loss": 6.4825,
      "mean_token_accuracy": 0.12456277534365653,
      "num_tokens": 23186170.0,
      "step": 6940
    },
    {
      "entropy": 6.4875627040863035,
      "epoch": 1.3548922056384742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048464202116987957,
      "loss": 6.3871,
      "mean_token_accuracy": 0.12761446312069893,
      "num_tokens": 23203567.0,
      "step": 6945
    },
    {
      "entropy": 6.391657972335816,
      "epoch": 1.3558677202224172,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048461647117419335,
      "loss": 6.2925,
      "mean_token_accuracy": 0.14083907455205918,
      "num_tokens": 23220551.0,
      "step": 6950
    },
    {
      "entropy": 6.358842945098877,
      "epoch": 1.3568432348063604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048459090069553256,
      "loss": 6.224,
      "mean_token_accuracy": 0.14023920968174935,
      "num_tokens": 23237390.0,
      "step": 6955
    },
    {
      "entropy": 6.419101667404175,
      "epoch": 1.3578187493903033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048456530973639604,
      "loss": 6.2771,
      "mean_token_accuracy": 0.1402598038315773,
      "num_tokens": 23254644.0,
      "step": 6960
    },
    {
      "entropy": 6.389404964447022,
      "epoch": 1.3587942639742465,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048453969829928437,
      "loss": 6.3022,
      "mean_token_accuracy": 0.13748536109924317,
      "num_tokens": 23273261.0,
      "step": 6965
    },
    {
      "entropy": 6.511208772659302,
      "epoch": 1.3597697785581895,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004845140663867001,
      "loss": 6.4105,
      "mean_token_accuracy": 0.1262500636279583,
      "num_tokens": 23290366.0,
      "step": 6970
    },
    {
      "entropy": 6.516773319244384,
      "epoch": 1.3607452931421324,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000484488414001148,
      "loss": 6.4142,
      "mean_token_accuracy": 0.12467917874455452,
      "num_tokens": 23306629.0,
      "step": 6975
    },
    {
      "entropy": 6.323417472839355,
      "epoch": 1.3617208077260754,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004844627411451348,
      "loss": 6.2516,
      "mean_token_accuracy": 0.14216941446065903,
      "num_tokens": 23322446.0,
      "step": 6980
    },
    {
      "entropy": 6.386747550964356,
      "epoch": 1.3626963223100186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000484437047821169,
      "loss": 6.3123,
      "mean_token_accuracy": 0.13950911462306975,
      "num_tokens": 23340447.0,
      "step": 6985
    },
    {
      "entropy": 6.43338360786438,
      "epoch": 1.3636718368939615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048441133403176137,
      "loss": 6.3256,
      "mean_token_accuracy": 0.13789551705121994,
      "num_tokens": 23356571.0,
      "step": 6990
    },
    {
      "entropy": 6.509999513626099,
      "epoch": 1.3646473514779047,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048438559977942447,
      "loss": 6.4402,
      "mean_token_accuracy": 0.12819159552454948,
      "num_tokens": 23374012.0,
      "step": 6995
    },
    {
      "entropy": 6.464243173599243,
      "epoch": 1.3656228660618477,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048435984506667307,
      "loss": 6.3582,
      "mean_token_accuracy": 0.13450334891676902,
      "num_tokens": 23391011.0,
      "step": 7000
    },
    {
      "entropy": 6.5537965297698975,
      "epoch": 1.3665983806457906,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004843340698960238,
      "loss": 6.4703,
      "mean_token_accuracy": 0.12451658099889755,
      "num_tokens": 23408080.0,
      "step": 7005
    },
    {
      "entropy": 6.284679317474366,
      "epoch": 1.3675738952297336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843082742699953,
      "loss": 6.1236,
      "mean_token_accuracy": 0.15142035484313965,
      "num_tokens": 23424060.0,
      "step": 7010
    },
    {
      "entropy": 6.3996076583862305,
      "epoch": 1.3685494098136768,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004842824581911082,
      "loss": 6.3124,
      "mean_token_accuracy": 0.12868525087833405,
      "num_tokens": 23441900.0,
      "step": 7015
    },
    {
      "entropy": 6.511254501342774,
      "epoch": 1.3695249243976197,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004842566216618852,
      "loss": 6.3511,
      "mean_token_accuracy": 0.13047146275639535,
      "num_tokens": 23457032.0,
      "step": 7020
    },
    {
      "entropy": 6.483934545516968,
      "epoch": 1.3705004389815627,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004842307646848509,
      "loss": 6.4195,
      "mean_token_accuracy": 0.12202546745538712,
      "num_tokens": 23473934.0,
      "step": 7025
    },
    {
      "entropy": 6.585614633560181,
      "epoch": 1.371475953565506,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048420488726253196,
      "loss": 6.4756,
      "mean_token_accuracy": 0.12003122344613075,
      "num_tokens": 23491024.0,
      "step": 7030
    },
    {
      "entropy": 6.443660163879395,
      "epoch": 1.3724514681494488,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004841789893974571,
      "loss": 6.3986,
      "mean_token_accuracy": 0.12741972506046295,
      "num_tokens": 23506763.0,
      "step": 7035
    },
    {
      "entropy": 6.502723932266235,
      "epoch": 1.3734269827333918,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004841530710921569,
      "loss": 6.3757,
      "mean_token_accuracy": 0.1320430353283882,
      "num_tokens": 23523101.0,
      "step": 7040
    },
    {
      "entropy": 6.431183910369873,
      "epoch": 1.3744024973173348,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004841271323491642,
      "loss": 6.2986,
      "mean_token_accuracy": 0.14038016349077226,
      "num_tokens": 23539787.0,
      "step": 7045
    },
    {
      "entropy": 6.428557968139648,
      "epoch": 1.375378011901278,
      "grad_norm": 1.0,
      "learning_rate": 0.0004841011731710133,
      "loss": 6.3811,
      "mean_token_accuracy": 0.13419227674603462,
      "num_tokens": 23555800.0,
      "step": 7050
    },
    {
      "entropy": 6.52983136177063,
      "epoch": 1.376353526485221,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048407519356024107,
      "loss": 6.3488,
      "mean_token_accuracy": 0.1344088777899742,
      "num_tokens": 23572234.0,
      "step": 7055
    },
    {
      "entropy": 6.408143949508667,
      "epoch": 1.377329041069164,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048404919351938597,
      "loss": 6.4302,
      "mean_token_accuracy": 0.12809761613607407,
      "num_tokens": 23588982.0,
      "step": 7060
    },
    {
      "entropy": 6.512300062179565,
      "epoch": 1.378304555653107,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004840231730509889,
      "loss": 6.3042,
      "mean_token_accuracy": 0.1369057908654213,
      "num_tokens": 23606921.0,
      "step": 7065
    },
    {
      "entropy": 6.405620288848877,
      "epoch": 1.37928007023705,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004839971321575922,
      "loss": 6.4225,
      "mean_token_accuracy": 0.12803878337144853,
      "num_tokens": 23625562.0,
      "step": 7070
    },
    {
      "entropy": 6.400554275512695,
      "epoch": 1.380255584820993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004839710708417407,
      "loss": 6.2337,
      "mean_token_accuracy": 0.14147769659757614,
      "num_tokens": 23642586.0,
      "step": 7075
    },
    {
      "entropy": 6.481622552871704,
      "epoch": 1.3812310994049362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004839449891059809,
      "loss": 6.3314,
      "mean_token_accuracy": 0.13251293897628785,
      "num_tokens": 23659245.0,
      "step": 7080
    },
    {
      "entropy": 6.422457218170166,
      "epoch": 1.3822066139888791,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004839188869528615,
      "loss": 6.285,
      "mean_token_accuracy": 0.13858347982168198,
      "num_tokens": 23674609.0,
      "step": 7085
    },
    {
      "entropy": 6.2247467041015625,
      "epoch": 1.3831821285728223,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004838927643849331,
      "loss": 6.1009,
      "mean_token_accuracy": 0.15440160781145096,
      "num_tokens": 23692778.0,
      "step": 7090
    },
    {
      "entropy": 6.446448850631714,
      "epoch": 1.3841576431567653,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048386662140474814,
      "loss": 6.3958,
      "mean_token_accuracy": 0.13582372516393662,
      "num_tokens": 23708855.0,
      "step": 7095
    },
    {
      "entropy": 6.5245849132537845,
      "epoch": 1.3851331577407082,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048384045801486136,
      "loss": 6.4134,
      "mean_token_accuracy": 0.1276310309767723,
      "num_tokens": 23726791.0,
      "step": 7100
    },
    {
      "entropy": 6.470557594299317,
      "epoch": 1.3861086723246512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838142742178295,
      "loss": 6.3572,
      "mean_token_accuracy": 0.12603426575660706,
      "num_tokens": 23743215.0,
      "step": 7105
    },
    {
      "entropy": 6.4538627624511715,
      "epoch": 1.3870841869085944,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048378807001621084,
      "loss": 6.3178,
      "mean_token_accuracy": 0.1267779991030693,
      "num_tokens": 23760814.0,
      "step": 7110
    },
    {
      "entropy": 6.484725284576416,
      "epoch": 1.3880597014925373,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048376184541256616,
      "loss": 6.3857,
      "mean_token_accuracy": 0.13010825887322425,
      "num_tokens": 23776408.0,
      "step": 7115
    },
    {
      "entropy": 6.469448518753052,
      "epoch": 1.3890352160764803,
      "grad_norm": 1.0,
      "learning_rate": 0.000483735600409458,
      "loss": 6.324,
      "mean_token_accuracy": 0.14090776070952415,
      "num_tokens": 23792237.0,
      "step": 7120
    },
    {
      "entropy": 6.365302038192749,
      "epoch": 1.3900107306604235,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000483709335009451,
      "loss": 6.2625,
      "mean_token_accuracy": 0.14033160209655762,
      "num_tokens": 23809006.0,
      "step": 7125
    },
    {
      "entropy": 6.536910104751587,
      "epoch": 1.3909862452443664,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048368304921511153,
      "loss": 6.5016,
      "mean_token_accuracy": 0.12211707159876824,
      "num_tokens": 23825964.0,
      "step": 7130
    },
    {
      "entropy": 6.421416473388672,
      "epoch": 1.3919617598283094,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004836567430290082,
      "loss": 6.2554,
      "mean_token_accuracy": 0.14255294427275658,
      "num_tokens": 23842674.0,
      "step": 7135
    },
    {
      "entropy": 6.339613151550293,
      "epoch": 1.3929372744122523,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004836304164537117,
      "loss": 6.2467,
      "mean_token_accuracy": 0.14281350001692772,
      "num_tokens": 23859414.0,
      "step": 7140
    },
    {
      "entropy": 6.3821573734283445,
      "epoch": 1.3939127889961955,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004836040694917945,
      "loss": 6.3007,
      "mean_token_accuracy": 0.1422678917646408,
      "num_tokens": 23875529.0,
      "step": 7145
    },
    {
      "entropy": 6.433695888519287,
      "epoch": 1.3948883035801385,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000483577702145831,
      "loss": 6.2781,
      "mean_token_accuracy": 0.13871855810284614,
      "num_tokens": 23892442.0,
      "step": 7150
    },
    {
      "entropy": 6.478998613357544,
      "epoch": 1.3958638181640817,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004835513144183979,
      "loss": 6.3844,
      "mean_token_accuracy": 0.1316008187830448,
      "num_tokens": 23908624.0,
      "step": 7155
    },
    {
      "entropy": 6.504911947250366,
      "epoch": 1.3968393327480246,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048352490631207366,
      "loss": 6.3426,
      "mean_token_accuracy": 0.1375097744166851,
      "num_tokens": 23925651.0,
      "step": 7160
    },
    {
      "entropy": 6.518744087219238,
      "epoch": 1.3978148473319676,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048349847782943873,
      "loss": 6.4744,
      "mean_token_accuracy": 0.12570460364222527,
      "num_tokens": 23943114.0,
      "step": 7165
    },
    {
      "entropy": 6.493485069274902,
      "epoch": 1.3987903619159106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004834720289730757,
      "loss": 6.3828,
      "mean_token_accuracy": 0.13075467050075532,
      "num_tokens": 23960712.0,
      "step": 7170
    },
    {
      "entropy": 6.434821701049804,
      "epoch": 1.3997658764998537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048344555974556895,
      "loss": 6.3217,
      "mean_token_accuracy": 0.13932181373238564,
      "num_tokens": 23976832.0,
      "step": 7175
    },
    {
      "entropy": 6.400946998596192,
      "epoch": 1.4007413910837967,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048341907014950503,
      "loss": 6.3155,
      "mean_token_accuracy": 0.138486997038126,
      "num_tokens": 23993323.0,
      "step": 7180
    },
    {
      "entropy": 6.385240268707276,
      "epoch": 1.4017169056677397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004833925601874724,
      "loss": 6.2646,
      "mean_token_accuracy": 0.13700883314013482,
      "num_tokens": 24010164.0,
      "step": 7185
    },
    {
      "entropy": 6.436604452133179,
      "epoch": 1.4026924202516828,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004833660298620615,
      "loss": 6.3221,
      "mean_token_accuracy": 0.13405472934246063,
      "num_tokens": 24025521.0,
      "step": 7190
    },
    {
      "entropy": 6.375053262710571,
      "epoch": 1.4036679348356258,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048333947917586484,
      "loss": 6.2181,
      "mean_token_accuracy": 0.14269407987594604,
      "num_tokens": 24041674.0,
      "step": 7195
    },
    {
      "entropy": 6.4790760517120365,
      "epoch": 1.4046434494195688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048331290813147677,
      "loss": 6.3289,
      "mean_token_accuracy": 0.13267022222280503,
      "num_tokens": 24057175.0,
      "step": 7200
    },
    {
      "entropy": 6.42034330368042,
      "epoch": 1.4056189640035117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048328631673149377,
      "loss": 6.338,
      "mean_token_accuracy": 0.12865512073040009,
      "num_tokens": 24072602.0,
      "step": 7205
    },
    {
      "entropy": 6.513401985168457,
      "epoch": 1.406594478587455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048325970497851425,
      "loss": 6.2276,
      "mean_token_accuracy": 0.13385093733668327,
      "num_tokens": 24089518.0,
      "step": 7210
    },
    {
      "entropy": 6.336940336227417,
      "epoch": 1.4075699931713979,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832330728751386,
      "loss": 6.2546,
      "mean_token_accuracy": 0.14703218787908554,
      "num_tokens": 24106487.0,
      "step": 7215
    },
    {
      "entropy": 6.4063080787658695,
      "epoch": 1.408545507755341,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048320642042396915,
      "loss": 6.3433,
      "mean_token_accuracy": 0.13369949385523797,
      "num_tokens": 24123244.0,
      "step": 7220
    },
    {
      "entropy": 6.467205810546875,
      "epoch": 1.409521022339284,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048317974762761046,
      "loss": 6.3373,
      "mean_token_accuracy": 0.1328465647995472,
      "num_tokens": 24140221.0,
      "step": 7225
    },
    {
      "entropy": 6.440418529510498,
      "epoch": 1.410496536923227,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048315305448866873,
      "loss": 6.3758,
      "mean_token_accuracy": 0.1315639540553093,
      "num_tokens": 24156968.0,
      "step": 7230
    },
    {
      "entropy": 6.52040843963623,
      "epoch": 1.41147205150717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048312634100975245,
      "loss": 6.4441,
      "mean_token_accuracy": 0.12732095196843146,
      "num_tokens": 24172441.0,
      "step": 7235
    },
    {
      "entropy": 6.527959299087525,
      "epoch": 1.412447566091113,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004830996071934719,
      "loss": 6.4743,
      "mean_token_accuracy": 0.1227174662053585,
      "num_tokens": 24190616.0,
      "step": 7240
    },
    {
      "entropy": 6.435986518859863,
      "epoch": 1.413423080675056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048307285304243945,
      "loss": 6.2976,
      "mean_token_accuracy": 0.13679210171103479,
      "num_tokens": 24207195.0,
      "step": 7245
    },
    {
      "entropy": 6.259978199005127,
      "epoch": 1.4143985952589992,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004830460785592692,
      "loss": 6.2135,
      "mean_token_accuracy": 0.14634981080889703,
      "num_tokens": 24224160.0,
      "step": 7250
    },
    {
      "entropy": 6.565676689147949,
      "epoch": 1.4153741098429422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048301928374657783,
      "loss": 6.4343,
      "mean_token_accuracy": 0.12458222806453705,
      "num_tokens": 24239907.0,
      "step": 7255
    },
    {
      "entropy": 6.476371765136719,
      "epoch": 1.4163496244268852,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048299246860698336,
      "loss": 6.4282,
      "mean_token_accuracy": 0.1251731663942337,
      "num_tokens": 24255515.0,
      "step": 7260
    },
    {
      "entropy": 6.480728435516357,
      "epoch": 1.4173251390108281,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048296563314310616,
      "loss": 6.3854,
      "mean_token_accuracy": 0.12934883683919907,
      "num_tokens": 24273002.0,
      "step": 7265
    },
    {
      "entropy": 6.300130176544189,
      "epoch": 1.4183006535947713,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829387773575686,
      "loss": 6.2121,
      "mean_token_accuracy": 0.14960155710577966,
      "num_tokens": 24291622.0,
      "step": 7270
    },
    {
      "entropy": 6.475707483291626,
      "epoch": 1.4192761681787143,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048291190125299474,
      "loss": 6.3383,
      "mean_token_accuracy": 0.12987383604049682,
      "num_tokens": 24309445.0,
      "step": 7275
    },
    {
      "entropy": 6.483824348449707,
      "epoch": 1.4202516827626572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048288500483201083,
      "loss": 6.3418,
      "mean_token_accuracy": 0.1270505115389824,
      "num_tokens": 24324667.0,
      "step": 7280
    },
    {
      "entropy": 6.391285943984985,
      "epoch": 1.4212271973466004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048285808809724523,
      "loss": 6.2679,
      "mean_token_accuracy": 0.1433006040751934,
      "num_tokens": 24340998.0,
      "step": 7285
    },
    {
      "entropy": 6.477695035934448,
      "epoch": 1.4222027119305434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004828311510513281,
      "loss": 6.3497,
      "mean_token_accuracy": 0.12748810797929763,
      "num_tokens": 24357416.0,
      "step": 7290
    },
    {
      "entropy": 6.392409610748291,
      "epoch": 1.4231782265144863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004828041936968916,
      "loss": 6.324,
      "mean_token_accuracy": 0.13989868015050888,
      "num_tokens": 24373837.0,
      "step": 7295
    },
    {
      "entropy": 6.482010793685913,
      "epoch": 1.4241537410984293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048277721603656986,
      "loss": 6.3858,
      "mean_token_accuracy": 0.12799029871821405,
      "num_tokens": 24389777.0,
      "step": 7300
    },
    {
      "entropy": 6.491123914718628,
      "epoch": 1.4251292556823725,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048275021807299916,
      "loss": 6.3479,
      "mean_token_accuracy": 0.13578226640820504,
      "num_tokens": 24407815.0,
      "step": 7305
    },
    {
      "entropy": 6.491016054153443,
      "epoch": 1.4261047702663154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048272319980881756,
      "loss": 6.4196,
      "mean_token_accuracy": 0.12773850932717323,
      "num_tokens": 24422932.0,
      "step": 7310
    },
    {
      "entropy": 6.415225076675415,
      "epoch": 1.4270802848502586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004826961612466653,
      "loss": 6.2755,
      "mean_token_accuracy": 0.13643277436494827,
      "num_tokens": 24440012.0,
      "step": 7315
    },
    {
      "entropy": 6.500392866134644,
      "epoch": 1.4280557994342016,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048266910238918434,
      "loss": 6.3879,
      "mean_token_accuracy": 0.13089654818177224,
      "num_tokens": 24457056.0,
      "step": 7320
    },
    {
      "entropy": 6.447404575347901,
      "epoch": 1.4290313140181445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826420232390188,
      "loss": 6.2403,
      "mean_token_accuracy": 0.13721374720335006,
      "num_tokens": 24474779.0,
      "step": 7325
    },
    {
      "entropy": 6.497050380706787,
      "epoch": 1.4300068286020875,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004826149237988148,
      "loss": 6.3915,
      "mean_token_accuracy": 0.13091783821582795,
      "num_tokens": 24492046.0,
      "step": 7330
    },
    {
      "entropy": 6.45014762878418,
      "epoch": 1.4309823431860307,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048258780407122045,
      "loss": 6.3894,
      "mean_token_accuracy": 0.12623273059725762,
      "num_tokens": 24507972.0,
      "step": 7335
    },
    {
      "entropy": 6.44755277633667,
      "epoch": 1.4319578577699736,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004825606640588858,
      "loss": 6.308,
      "mean_token_accuracy": 0.12877673655748367,
      "num_tokens": 24526350.0,
      "step": 7340
    },
    {
      "entropy": 6.29955153465271,
      "epoch": 1.4329333723539168,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048253350376446265,
      "loss": 6.2462,
      "mean_token_accuracy": 0.14160918518900872,
      "num_tokens": 24543478.0,
      "step": 7345
    },
    {
      "entropy": 6.484018135070801,
      "epoch": 1.4339088869378598,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004825063231906053,
      "loss": 6.3223,
      "mean_token_accuracy": 0.13833547234535218,
      "num_tokens": 24562285.0,
      "step": 7350
    },
    {
      "entropy": 6.45168023109436,
      "epoch": 1.4348844015218027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004824791223399696,
      "loss": 6.3417,
      "mean_token_accuracy": 0.13222036138176918,
      "num_tokens": 24578271.0,
      "step": 7355
    },
    {
      "entropy": 6.44455304145813,
      "epoch": 1.4358599161057457,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048245190121521346,
      "loss": 6.3812,
      "mean_token_accuracy": 0.1321019396185875,
      "num_tokens": 24595107.0,
      "step": 7360
    },
    {
      "entropy": 6.492379808425904,
      "epoch": 1.4368354306896889,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000482424659818997,
      "loss": 6.3731,
      "mean_token_accuracy": 0.12945975586771966,
      "num_tokens": 24611594.0,
      "step": 7365
    },
    {
      "entropy": 6.462053918838501,
      "epoch": 1.4378109452736318,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000482397398153982,
      "loss": 6.3624,
      "mean_token_accuracy": 0.13529488295316697,
      "num_tokens": 24628284.0,
      "step": 7370
    },
    {
      "entropy": 6.414290809631348,
      "epoch": 1.4387864598575748,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048237011622283244,
      "loss": 6.3221,
      "mean_token_accuracy": 0.13338397517800332,
      "num_tokens": 24644305.0,
      "step": 7375
    },
    {
      "entropy": 6.521754932403565,
      "epoch": 1.439761974441518,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048234281402821416,
      "loss": 6.4339,
      "mean_token_accuracy": 0.13221795558929444,
      "num_tokens": 24661314.0,
      "step": 7380
    },
    {
      "entropy": 6.540429925918579,
      "epoch": 1.440737489025461,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048231549157279514,
      "loss": 6.3276,
      "mean_token_accuracy": 0.1319704495370388,
      "num_tokens": 24677175.0,
      "step": 7385
    },
    {
      "entropy": 6.341984558105469,
      "epoch": 1.441713003609404,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004822881488592452,
      "loss": 6.2872,
      "mean_token_accuracy": 0.13454542607069014,
      "num_tokens": 24695227.0,
      "step": 7390
    },
    {
      "entropy": 6.22827639579773,
      "epoch": 1.4426885181933469,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048226078589023606,
      "loss": 6.0875,
      "mean_token_accuracy": 0.15628089457750322,
      "num_tokens": 24712186.0,
      "step": 7395
    },
    {
      "entropy": 6.4143922328948975,
      "epoch": 1.44366403277729,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048223340266844166,
      "loss": 6.3369,
      "mean_token_accuracy": 0.13485684171319007,
      "num_tokens": 24728874.0,
      "step": 7400
    },
    {
      "entropy": 6.3925456523895265,
      "epoch": 1.444639547361233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004822059991965377,
      "loss": 6.3096,
      "mean_token_accuracy": 0.13359904587268828,
      "num_tokens": 24745534.0,
      "step": 7405
    },
    {
      "entropy": 6.514584350585937,
      "epoch": 1.4456150619451762,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048217857547720207,
      "loss": 6.3347,
      "mean_token_accuracy": 0.13158397227525712,
      "num_tokens": 24760872.0,
      "step": 7410
    },
    {
      "entropy": 6.47188367843628,
      "epoch": 1.4465905765291192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004821511315131144,
      "loss": 6.3643,
      "mean_token_accuracy": 0.12627866715192795,
      "num_tokens": 24776265.0,
      "step": 7415
    },
    {
      "entropy": 6.392746257781982,
      "epoch": 1.4475660911130621,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004821236673069564,
      "loss": 6.2798,
      "mean_token_accuracy": 0.13589751943945885,
      "num_tokens": 24791557.0,
      "step": 7420
    },
    {
      "entropy": 6.440803909301758,
      "epoch": 1.448541605697005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048209618286141194,
      "loss": 6.3092,
      "mean_token_accuracy": 0.13457278683781623,
      "num_tokens": 24807444.0,
      "step": 7425
    },
    {
      "entropy": 6.481490135192871,
      "epoch": 1.4495171202809483,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004820686781791665,
      "loss": 6.3905,
      "mean_token_accuracy": 0.12977924644947053,
      "num_tokens": 24823971.0,
      "step": 7430
    },
    {
      "entropy": 6.317101383209229,
      "epoch": 1.4504926348648912,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048204115326290795,
      "loss": 6.218,
      "mean_token_accuracy": 0.1508413754403591,
      "num_tokens": 24841363.0,
      "step": 7435
    },
    {
      "entropy": 6.478225135803223,
      "epoch": 1.4514681494488342,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048201360811532567,
      "loss": 6.355,
      "mean_token_accuracy": 0.13365337625145912,
      "num_tokens": 24858593.0,
      "step": 7440
    },
    {
      "entropy": 6.5106837272644045,
      "epoch": 1.4524436640327774,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004819860427391115,
      "loss": 6.4547,
      "mean_token_accuracy": 0.12866950705647467,
      "num_tokens": 24876303.0,
      "step": 7445
    },
    {
      "entropy": 6.445487976074219,
      "epoch": 1.4534191786167203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004819584571369589,
      "loss": 6.3051,
      "mean_token_accuracy": 0.12703511640429496,
      "num_tokens": 24891455.0,
      "step": 7450
    },
    {
      "entropy": 6.521517705917359,
      "epoch": 1.4543946932006633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004819308513115635,
      "loss": 6.3587,
      "mean_token_accuracy": 0.13204744085669518,
      "num_tokens": 24906598.0,
      "step": 7455
    },
    {
      "entropy": 6.467092418670655,
      "epoch": 1.4553702077846065,
      "grad_norm": 1.0,
      "learning_rate": 0.0004819032252656229,
      "loss": 6.4138,
      "mean_token_accuracy": 0.12688039541244506,
      "num_tokens": 24923299.0,
      "step": 7460
    },
    {
      "entropy": 6.41647686958313,
      "epoch": 1.4563457223685494,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048187557900183643,
      "loss": 6.2707,
      "mean_token_accuracy": 0.14400876238942145,
      "num_tokens": 24940510.0,
      "step": 7465
    },
    {
      "entropy": 6.465256309509277,
      "epoch": 1.4573212369524924,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004818479125229057,
      "loss": 6.347,
      "mean_token_accuracy": 0.13142351061105728,
      "num_tokens": 24957644.0,
      "step": 7470
    },
    {
      "entropy": 6.4078773021697994,
      "epoch": 1.4582967515364356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004818202258315343,
      "loss": 6.2845,
      "mean_token_accuracy": 0.13732921183109284,
      "num_tokens": 24974405.0,
      "step": 7475
    },
    {
      "entropy": 6.494929599761963,
      "epoch": 1.4592722661203785,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004817925189304275,
      "loss": 6.3979,
      "mean_token_accuracy": 0.1276998423039913,
      "num_tokens": 24992294.0,
      "step": 7480
    },
    {
      "entropy": 6.498716926574707,
      "epoch": 1.4602477807043215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048176479182229266,
      "loss": 6.3502,
      "mean_token_accuracy": 0.12538522779941558,
      "num_tokens": 25008177.0,
      "step": 7485
    },
    {
      "entropy": 6.4651984691619875,
      "epoch": 1.4612232952882644,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048173704450983943,
      "loss": 6.4311,
      "mean_token_accuracy": 0.12539587914943695,
      "num_tokens": 25025321.0,
      "step": 7490
    },
    {
      "entropy": 6.5265590190887455,
      "epoch": 1.4621988098722076,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048170927699577894,
      "loss": 6.3731,
      "mean_token_accuracy": 0.13026623651385308,
      "num_tokens": 25041751.0,
      "step": 7495
    },
    {
      "entropy": 6.422158241271973,
      "epoch": 1.4631743244561506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004816814892828247,
      "loss": 6.2819,
      "mean_token_accuracy": 0.13738936185836792,
      "num_tokens": 25058897.0,
      "step": 7500
    },
    {
      "entropy": 6.452862882614136,
      "epoch": 1.4641498390400938,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048165368137369186,
      "loss": 6.4016,
      "mean_token_accuracy": 0.1271025262773037,
      "num_tokens": 25074750.0,
      "step": 7505
    },
    {
      "entropy": 6.458086919784546,
      "epoch": 1.4651253536240367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048162585327109794,
      "loss": 6.3658,
      "mean_token_accuracy": 0.1317321017384529,
      "num_tokens": 25089416.0,
      "step": 7510
    },
    {
      "entropy": 6.4188456535339355,
      "epoch": 1.4661008682079797,
      "grad_norm": 1.03125,
      "learning_rate": 0.000481598004977762,
      "loss": 6.3221,
      "mean_token_accuracy": 0.1374312236905098,
      "num_tokens": 25105543.0,
      "step": 7515
    },
    {
      "entropy": 6.523442506790161,
      "epoch": 1.4670763827919227,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048157013649640523,
      "loss": 6.3957,
      "mean_token_accuracy": 0.12631377950310707,
      "num_tokens": 25124761.0,
      "step": 7520
    },
    {
      "entropy": 6.385325574874878,
      "epoch": 1.4680518973758658,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048154224782975103,
      "loss": 6.3534,
      "mean_token_accuracy": 0.1315249741077423,
      "num_tokens": 25142484.0,
      "step": 7525
    },
    {
      "entropy": 6.45292329788208,
      "epoch": 1.4690274119598088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048151433898052453,
      "loss": 6.244,
      "mean_token_accuracy": 0.14741309657692908,
      "num_tokens": 25159870.0,
      "step": 7530
    },
    {
      "entropy": 6.1635088443756105,
      "epoch": 1.4700029265437518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004814864099514529,
      "loss": 6.1114,
      "mean_token_accuracy": 0.15597472637891768,
      "num_tokens": 25176643.0,
      "step": 7535
    },
    {
      "entropy": 6.406569194793701,
      "epoch": 1.470978441127695,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004814584607452651,
      "loss": 6.3617,
      "mean_token_accuracy": 0.12973567843437195,
      "num_tokens": 25191963.0,
      "step": 7540
    },
    {
      "entropy": 6.483195018768311,
      "epoch": 1.471953955711638,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004814304913646924,
      "loss": 6.4116,
      "mean_token_accuracy": 0.13429397195577622,
      "num_tokens": 25210250.0,
      "step": 7545
    },
    {
      "entropy": 6.429952096939087,
      "epoch": 1.4729294702955809,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048140250181246785,
      "loss": 6.2672,
      "mean_token_accuracy": 0.13618823438882827,
      "num_tokens": 25226951.0,
      "step": 7550
    },
    {
      "entropy": 6.345352268218994,
      "epoch": 1.4739049848795238,
      "grad_norm": 1.0,
      "learning_rate": 0.0004813744920913265,
      "loss": 6.1976,
      "mean_token_accuracy": 0.1409853182733059,
      "num_tokens": 25244055.0,
      "step": 7555
    },
    {
      "entropy": 6.287253093719483,
      "epoch": 1.474880499463467,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048134646220400517,
      "loss": 6.1867,
      "mean_token_accuracy": 0.15438724607229232,
      "num_tokens": 25261408.0,
      "step": 7560
    },
    {
      "entropy": 6.375587081909179,
      "epoch": 1.47585601404741,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048131841215324307,
      "loss": 6.2355,
      "mean_token_accuracy": 0.14037031531333924,
      "num_tokens": 25277452.0,
      "step": 7565
    },
    {
      "entropy": 6.509365129470825,
      "epoch": 1.4768315286313531,
      "grad_norm": 0.984375,
      "learning_rate": 0.000481290341941781,
      "loss": 6.4412,
      "mean_token_accuracy": 0.12742704823613166,
      "num_tokens": 25293414.0,
      "step": 7570
    },
    {
      "entropy": 6.439072704315185,
      "epoch": 1.477807043215296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048126225157236205,
      "loss": 6.3223,
      "mean_token_accuracy": 0.13460597172379493,
      "num_tokens": 25309684.0,
      "step": 7575
    },
    {
      "entropy": 6.497540998458862,
      "epoch": 1.478782557799239,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048123414104773093,
      "loss": 6.3285,
      "mean_token_accuracy": 0.13388363644480705,
      "num_tokens": 25324810.0,
      "step": 7580
    },
    {
      "entropy": 6.575790119171143,
      "epoch": 1.479758072383182,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048120601037063466,
      "loss": 6.4656,
      "mean_token_accuracy": 0.1296125426888466,
      "num_tokens": 25342921.0,
      "step": 7585
    },
    {
      "entropy": 6.452401208877563,
      "epoch": 1.4807335869671252,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004811778595438219,
      "loss": 6.3607,
      "mean_token_accuracy": 0.13164495453238487,
      "num_tokens": 25358770.0,
      "step": 7590
    },
    {
      "entropy": 6.36121096611023,
      "epoch": 1.4817091015510682,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004811496885700436,
      "loss": 6.1981,
      "mean_token_accuracy": 0.1427338883280754,
      "num_tokens": 25375429.0,
      "step": 7595
    },
    {
      "entropy": 6.468779277801514,
      "epoch": 1.4826846161350113,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004811214974520524,
      "loss": 6.3797,
      "mean_token_accuracy": 0.1338053300976753,
      "num_tokens": 25392262.0,
      "step": 7600
    },
    {
      "entropy": 6.409215831756592,
      "epoch": 1.4836601307189543,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004810932861926032,
      "loss": 6.2945,
      "mean_token_accuracy": 0.13999646231532098,
      "num_tokens": 25408663.0,
      "step": 7605
    },
    {
      "entropy": 6.4435430526733395,
      "epoch": 1.4846356453028973,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004810650547944526,
      "loss": 6.3463,
      "mean_token_accuracy": 0.1325049951672554,
      "num_tokens": 25425587.0,
      "step": 7610
    },
    {
      "entropy": 6.27195029258728,
      "epoch": 1.4856111598868402,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004810368032603591,
      "loss": 6.1294,
      "mean_token_accuracy": 0.15280173420906068,
      "num_tokens": 25443506.0,
      "step": 7615
    },
    {
      "entropy": 6.47093186378479,
      "epoch": 1.4865866744707834,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004810085315930837,
      "loss": 6.3328,
      "mean_token_accuracy": 0.13916512802243233,
      "num_tokens": 25462197.0,
      "step": 7620
    },
    {
      "entropy": 6.39477310180664,
      "epoch": 1.4875621890547264,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048098023979538867,
      "loss": 6.3137,
      "mean_token_accuracy": 0.13670894578099252,
      "num_tokens": 25478732.0,
      "step": 7625
    },
    {
      "entropy": 6.4931573390960695,
      "epoch": 1.4885377036386693,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048095192787003886,
      "loss": 6.3806,
      "mean_token_accuracy": 0.12668586522340775,
      "num_tokens": 25494427.0,
      "step": 7630
    },
    {
      "entropy": 6.381348514556885,
      "epoch": 1.4895132182226125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048092359581980063,
      "loss": 6.2608,
      "mean_token_accuracy": 0.13954768925905228,
      "num_tokens": 25511617.0,
      "step": 7635
    },
    {
      "entropy": 6.453156042098999,
      "epoch": 1.4904887328065555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004808952436474425,
      "loss": 6.3867,
      "mean_token_accuracy": 0.1303196594119072,
      "num_tokens": 25529655.0,
      "step": 7640
    },
    {
      "entropy": 6.50951738357544,
      "epoch": 1.4914642473904984,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048086687135573506,
      "loss": 6.3246,
      "mean_token_accuracy": 0.13155678287148476,
      "num_tokens": 25545615.0,
      "step": 7645
    },
    {
      "entropy": 6.239132213592529,
      "epoch": 1.4924397619744414,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048083847894745064,
      "loss": 6.1953,
      "mean_token_accuracy": 0.1467883490025997,
      "num_tokens": 25562097.0,
      "step": 7650
    },
    {
      "entropy": 6.409989547729492,
      "epoch": 1.4934152765583846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048081006642536366,
      "loss": 6.2618,
      "mean_token_accuracy": 0.13890096098184584,
      "num_tokens": 25578013.0,
      "step": 7655
    },
    {
      "entropy": 6.393611335754395,
      "epoch": 1.4943907911423275,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004807816337922506,
      "loss": 6.3039,
      "mean_token_accuracy": 0.13630437031388282,
      "num_tokens": 25594451.0,
      "step": 7660
    },
    {
      "entropy": 6.352360439300537,
      "epoch": 1.4953663057262707,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004807531810508896,
      "loss": 6.222,
      "mean_token_accuracy": 0.1373550534248352,
      "num_tokens": 25611522.0,
      "step": 7665
    },
    {
      "entropy": 6.4538665294647215,
      "epoch": 1.4963418203102137,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004807247082040611,
      "loss": 6.3285,
      "mean_token_accuracy": 0.1364366218447685,
      "num_tokens": 25627755.0,
      "step": 7670
    },
    {
      "entropy": 6.420915412902832,
      "epoch": 1.4973173348941566,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048069621525454733,
      "loss": 6.3883,
      "mean_token_accuracy": 0.13183033168315889,
      "num_tokens": 25645059.0,
      "step": 7675
    },
    {
      "entropy": 6.500278663635254,
      "epoch": 1.4982928494780996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004806677022051325,
      "loss": 6.3827,
      "mean_token_accuracy": 0.13254290372133254,
      "num_tokens": 25662324.0,
      "step": 7680
    },
    {
      "entropy": 6.440519905090332,
      "epoch": 1.4992683640620428,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048063916905860283,
      "loss": 6.3131,
      "mean_token_accuracy": 0.1339655965566635,
      "num_tokens": 25678648.0,
      "step": 7685
    },
    {
      "entropy": 6.358491468429565,
      "epoch": 1.5002438786459857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004806106158177465,
      "loss": 6.2059,
      "mean_token_accuracy": 0.14812275469303132,
      "num_tokens": 25695694.0,
      "step": 7690
    },
    {
      "entropy": 6.3614589214324955,
      "epoch": 1.501219393229929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004805820424853536,
      "loss": 6.2423,
      "mean_token_accuracy": 0.14040227308869363,
      "num_tokens": 25713168.0,
      "step": 7695
    },
    {
      "entropy": 6.37730860710144,
      "epoch": 1.5021949078138719,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004805534490642162,
      "loss": 6.3014,
      "mean_token_accuracy": 0.13387165442109109,
      "num_tokens": 25729763.0,
      "step": 7700
    },
    {
      "entropy": 6.388044643402099,
      "epoch": 1.5031704223978148,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048052483555712834,
      "loss": 6.2851,
      "mean_token_accuracy": 0.13636611253023148,
      "num_tokens": 25747852.0,
      "step": 7705
    },
    {
      "entropy": 6.410606527328492,
      "epoch": 1.5041459369817578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004804962019668861,
      "loss": 6.2762,
      "mean_token_accuracy": 0.14588508754968643,
      "num_tokens": 25764594.0,
      "step": 7710
    },
    {
      "entropy": 6.433447456359863,
      "epoch": 1.5051214515657008,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004804675482962874,
      "loss": 6.2947,
      "mean_token_accuracy": 0.13699788898229598,
      "num_tokens": 25782835.0,
      "step": 7715
    },
    {
      "entropy": 6.448841190338134,
      "epoch": 1.506096966149644,
      "grad_norm": 1.0,
      "learning_rate": 0.0004804388745481322,
      "loss": 6.3158,
      "mean_token_accuracy": 0.13352611735463144,
      "num_tokens": 25798152.0,
      "step": 7720
    },
    {
      "entropy": 6.352209854125976,
      "epoch": 1.5070724807335871,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048041018072522224,
      "loss": 6.2556,
      "mean_token_accuracy": 0.1443852424621582,
      "num_tokens": 25816521.0,
      "step": 7725
    },
    {
      "entropy": 6.433612489700318,
      "epoch": 1.50804799531753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048038146683036163,
      "loss": 6.3062,
      "mean_token_accuracy": 0.12896057739853858,
      "num_tokens": 25832416.0,
      "step": 7730
    },
    {
      "entropy": 6.327046728134155,
      "epoch": 1.509023509901473,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004803527328663561,
      "loss": 6.2173,
      "mean_token_accuracy": 0.14858810380101203,
      "num_tokens": 25850343.0,
      "step": 7735
    },
    {
      "entropy": 6.481621360778808,
      "epoch": 1.509999024485416,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004803239788360134,
      "loss": 6.4143,
      "mean_token_accuracy": 0.13087122961878778,
      "num_tokens": 25868661.0,
      "step": 7740
    },
    {
      "entropy": 6.498782920837402,
      "epoch": 1.510974539069359,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048029520474214316,
      "loss": 6.3835,
      "mean_token_accuracy": 0.12821986451745032,
      "num_tokens": 25886417.0,
      "step": 7745
    },
    {
      "entropy": 6.414138889312744,
      "epoch": 1.5119500536533022,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048026641058755725,
      "loss": 6.2538,
      "mean_token_accuracy": 0.1382353723049164,
      "num_tokens": 25904271.0,
      "step": 7750
    },
    {
      "entropy": 6.39419584274292,
      "epoch": 1.512925568237245,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048023759637506934,
      "loss": 6.2465,
      "mean_token_accuracy": 0.1428878650069237,
      "num_tokens": 25920444.0,
      "step": 7755
    },
    {
      "entropy": 6.412502813339233,
      "epoch": 1.5139010828211883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000480208762107495,
      "loss": 6.3002,
      "mean_token_accuracy": 0.12848196923732758,
      "num_tokens": 25937655.0,
      "step": 7760
    },
    {
      "entropy": 6.3746387481689455,
      "epoch": 1.5148765974051313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004801799077876517,
      "loss": 6.2435,
      "mean_token_accuracy": 0.14026406928896903,
      "num_tokens": 25954242.0,
      "step": 7765
    },
    {
      "entropy": 6.496229219436645,
      "epoch": 1.5158521119890742,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004801510334183591,
      "loss": 6.3869,
      "mean_token_accuracy": 0.129782085865736,
      "num_tokens": 25971343.0,
      "step": 7770
    },
    {
      "entropy": 6.246442604064941,
      "epoch": 1.5168276265730172,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048012213900243877,
      "loss": 6.2394,
      "mean_token_accuracy": 0.1417051337659359,
      "num_tokens": 25988369.0,
      "step": 7775
    },
    {
      "entropy": 6.445706748962403,
      "epoch": 1.5178031411569601,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000480093224542714,
      "loss": 6.3626,
      "mean_token_accuracy": 0.132488614320755,
      "num_tokens": 26005028.0,
      "step": 7780
    },
    {
      "entropy": 6.509111833572388,
      "epoch": 1.5187786557409033,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004800642900420104,
      "loss": 6.3843,
      "mean_token_accuracy": 0.1366946816444397,
      "num_tokens": 26021529.0,
      "step": 7785
    },
    {
      "entropy": 6.461445617675781,
      "epoch": 1.5197541703248465,
      "grad_norm": 1.0,
      "learning_rate": 0.00048003533550315517,
      "loss": 6.2938,
      "mean_token_accuracy": 0.1364104352891445,
      "num_tokens": 26036804.0,
      "step": 7790
    },
    {
      "entropy": 6.324490690231324,
      "epoch": 1.5207296849087895,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048000636092897773,
      "loss": 6.1908,
      "mean_token_accuracy": 0.14553504362702369,
      "num_tokens": 26054532.0,
      "step": 7795
    },
    {
      "entropy": 6.444527912139892,
      "epoch": 1.5217051994927324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004799773663223093,
      "loss": 6.3582,
      "mean_token_accuracy": 0.13262097761034966,
      "num_tokens": 26072600.0,
      "step": 7800
    },
    {
      "entropy": 6.440227890014649,
      "epoch": 1.5226807140766754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004799483516859833,
      "loss": 6.2928,
      "mean_token_accuracy": 0.13754234462976456,
      "num_tokens": 26087129.0,
      "step": 7805
    },
    {
      "entropy": 6.420616006851196,
      "epoch": 1.5236562286606183,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004799193170228348,
      "loss": 6.2966,
      "mean_token_accuracy": 0.13398282527923583,
      "num_tokens": 26102453.0,
      "step": 7810
    },
    {
      "entropy": 6.40049901008606,
      "epoch": 1.5246317432445615,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047989026233570087,
      "loss": 6.3506,
      "mean_token_accuracy": 0.13372695669531823,
      "num_tokens": 26119591.0,
      "step": 7815
    },
    {
      "entropy": 6.417725610733032,
      "epoch": 1.5256072578285045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004798611876274208,
      "loss": 6.2682,
      "mean_token_accuracy": 0.14096331521868705,
      "num_tokens": 26137155.0,
      "step": 7820
    },
    {
      "entropy": 6.36814317703247,
      "epoch": 1.5265827724124477,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047983209290083565,
      "loss": 6.232,
      "mean_token_accuracy": 0.1431899957358837,
      "num_tokens": 26153523.0,
      "step": 7825
    },
    {
      "entropy": 6.323295593261719,
      "epoch": 1.5275582869963906,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004798029781587884,
      "loss": 6.2353,
      "mean_token_accuracy": 0.14281620159745217,
      "num_tokens": 26172233.0,
      "step": 7830
    },
    {
      "entropy": 6.399880886077881,
      "epoch": 1.5285338015803336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047977384340412406,
      "loss": 6.3323,
      "mean_token_accuracy": 0.1359604574739933,
      "num_tokens": 26189702.0,
      "step": 7835
    },
    {
      "entropy": 6.554447507858276,
      "epoch": 1.5295093161642765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004797446886396895,
      "loss": 6.4145,
      "mean_token_accuracy": 0.12603931576013566,
      "num_tokens": 26205986.0,
      "step": 7840
    },
    {
      "entropy": 6.482942008972168,
      "epoch": 1.5304848307482197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047971551386833376,
      "loss": 6.3573,
      "mean_token_accuracy": 0.12595167234539986,
      "num_tokens": 26223667.0,
      "step": 7845
    },
    {
      "entropy": 6.372615003585816,
      "epoch": 1.5314603453321627,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004796863190929076,
      "loss": 6.2588,
      "mean_token_accuracy": 0.1409389615058899,
      "num_tokens": 26240269.0,
      "step": 7850
    },
    {
      "entropy": 6.514329195022583,
      "epoch": 1.5324358599161059,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047965710431626394,
      "loss": 6.3875,
      "mean_token_accuracy": 0.12531698495149612,
      "num_tokens": 26257184.0,
      "step": 7855
    },
    {
      "entropy": 6.445234632492065,
      "epoch": 1.5334113745000488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004796278695412573,
      "loss": 6.2928,
      "mean_token_accuracy": 0.13994459360837935,
      "num_tokens": 26273002.0,
      "step": 7860
    },
    {
      "entropy": 6.536494779586792,
      "epoch": 1.5343868890839918,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004795986147707447,
      "loss": 6.4306,
      "mean_token_accuracy": 0.12913582101464272,
      "num_tokens": 26289734.0,
      "step": 7865
    },
    {
      "entropy": 6.405992031097412,
      "epoch": 1.5353624036679347,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004795693400075846,
      "loss": 6.2724,
      "mean_token_accuracy": 0.1414544902741909,
      "num_tokens": 26307409.0,
      "step": 7870
    },
    {
      "entropy": 6.443688297271729,
      "epoch": 1.5363379182518777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004795400452546377,
      "loss": 6.3625,
      "mean_token_accuracy": 0.1385938972234726,
      "num_tokens": 26324061.0,
      "step": 7875
    },
    {
      "entropy": 6.395519351959228,
      "epoch": 1.537313432835821,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047951073051476657,
      "loss": 6.2122,
      "mean_token_accuracy": 0.14111365303397178,
      "num_tokens": 26340990.0,
      "step": 7880
    },
    {
      "entropy": 6.423655891418457,
      "epoch": 1.538288947419764,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047948139579083573,
      "loss": 6.2749,
      "mean_token_accuracy": 0.1390646852552891,
      "num_tokens": 26357389.0,
      "step": 7885
    },
    {
      "entropy": 6.530328369140625,
      "epoch": 1.539264462003707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004794520410857117,
      "loss": 6.4127,
      "mean_token_accuracy": 0.12353480383753776,
      "num_tokens": 26373325.0,
      "step": 7890
    },
    {
      "entropy": 6.47208080291748,
      "epoch": 1.54023997658765,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004794226664022628,
      "loss": 6.3393,
      "mean_token_accuracy": 0.13314868211746217,
      "num_tokens": 26391098.0,
      "step": 7895
    },
    {
      "entropy": 6.476150417327881,
      "epoch": 1.541215491171593,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047939327174335974,
      "loss": 6.3226,
      "mean_token_accuracy": 0.1302434593439102,
      "num_tokens": 26407806.0,
      "step": 7900
    },
    {
      "entropy": 6.511614084243774,
      "epoch": 1.542191005755536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047936385711187443,
      "loss": 6.4448,
      "mean_token_accuracy": 0.12692206650972365,
      "num_tokens": 26423670.0,
      "step": 7905
    },
    {
      "entropy": 6.448093223571777,
      "epoch": 1.543166520339479,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047933442251068143,
      "loss": 6.3147,
      "mean_token_accuracy": 0.13048485964536666,
      "num_tokens": 26440493.0,
      "step": 7910
    },
    {
      "entropy": 6.431299543380737,
      "epoch": 1.544142034923422,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004793049679426569,
      "loss": 6.3663,
      "mean_token_accuracy": 0.13107406198978425,
      "num_tokens": 26458347.0,
      "step": 7915
    },
    {
      "entropy": 6.419653654098511,
      "epoch": 1.5451175495073652,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047927549341067916,
      "loss": 6.2574,
      "mean_token_accuracy": 0.141037280857563,
      "num_tokens": 26474832.0,
      "step": 7920
    },
    {
      "entropy": 6.430767965316773,
      "epoch": 1.5460930640913082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047924599891762814,
      "loss": 6.3616,
      "mean_token_accuracy": 0.13218679055571556,
      "num_tokens": 26492343.0,
      "step": 7925
    },
    {
      "entropy": 6.459003686904907,
      "epoch": 1.5470685786752512,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004792164844663861,
      "loss": 6.3574,
      "mean_token_accuracy": 0.13504771292209625,
      "num_tokens": 26508677.0,
      "step": 7930
    },
    {
      "entropy": 6.453594732284546,
      "epoch": 1.5480440932591941,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047918695005983705,
      "loss": 6.3586,
      "mean_token_accuracy": 0.13994947746396064,
      "num_tokens": 26524409.0,
      "step": 7935
    },
    {
      "entropy": 6.447396230697632,
      "epoch": 1.5490196078431373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004791573957008669,
      "loss": 6.3461,
      "mean_token_accuracy": 0.1324540063738823,
      "num_tokens": 26540884.0,
      "step": 7940
    },
    {
      "entropy": 6.432569169998169,
      "epoch": 1.5499951224270803,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004791278213923638,
      "loss": 6.2834,
      "mean_token_accuracy": 0.14047082811594008,
      "num_tokens": 26556636.0,
      "step": 7945
    },
    {
      "entropy": 6.432981300354004,
      "epoch": 1.5509706370110234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004790982271372174,
      "loss": 6.3224,
      "mean_token_accuracy": 0.14044027999043465,
      "num_tokens": 26573023.0,
      "step": 7950
    },
    {
      "entropy": 6.47172269821167,
      "epoch": 1.5519461515949664,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047906861293831965,
      "loss": 6.3565,
      "mean_token_accuracy": 0.13355283588171005,
      "num_tokens": 26590524.0,
      "step": 7955
    },
    {
      "entropy": 6.3017035007476805,
      "epoch": 1.5529216661789094,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047903897879856443,
      "loss": 6.0632,
      "mean_token_accuracy": 0.15730378031730652,
      "num_tokens": 26606866.0,
      "step": 7960
    },
    {
      "entropy": 6.279104900360108,
      "epoch": 1.5538971807628523,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047900932472084745,
      "loss": 6.2333,
      "mean_token_accuracy": 0.138857451826334,
      "num_tokens": 26623723.0,
      "step": 7965
    },
    {
      "entropy": 6.50374698638916,
      "epoch": 1.5548726953467953,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004789796507080663,
      "loss": 6.3109,
      "mean_token_accuracy": 0.13011292219161988,
      "num_tokens": 26639980.0,
      "step": 7970
    },
    {
      "entropy": 6.405613803863526,
      "epoch": 1.5558482099307385,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047894995676312066,
      "loss": 6.2941,
      "mean_token_accuracy": 0.14093357771635057,
      "num_tokens": 26657731.0,
      "step": 7975
    },
    {
      "entropy": 6.3346494197845455,
      "epoch": 1.5568237245146817,
      "grad_norm": 0.875,
      "learning_rate": 0.0004789202428889122,
      "loss": 6.3061,
      "mean_token_accuracy": 0.13660609498620033,
      "num_tokens": 26675754.0,
      "step": 7980
    },
    {
      "entropy": 6.496173524856568,
      "epoch": 1.5577992390986246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004788905090883443,
      "loss": 6.3563,
      "mean_token_accuracy": 0.12783240377902985,
      "num_tokens": 26692067.0,
      "step": 7985
    },
    {
      "entropy": 6.248537969589234,
      "epoch": 1.5587747536825676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004788607553643226,
      "loss": 6.1354,
      "mean_token_accuracy": 0.1555745154619217,
      "num_tokens": 26711026.0,
      "step": 7990
    },
    {
      "entropy": 6.33043909072876,
      "epoch": 1.5597502682665105,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047883098171975445,
      "loss": 6.2612,
      "mean_token_accuracy": 0.13581288680434228,
      "num_tokens": 26728463.0,
      "step": 7995
    },
    {
      "entropy": 6.389739274978638,
      "epoch": 1.5607257828504535,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004788011881575492,
      "loss": 6.1808,
      "mean_token_accuracy": 0.13609228506684304,
      "num_tokens": 26745160.0,
      "step": 8000
    },
    {
      "entropy": 6.469283485412598,
      "epoch": 1.5617012974343967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047877137468061825,
      "loss": 6.392,
      "mean_token_accuracy": 0.12988796532154084,
      "num_tokens": 26761782.0,
      "step": 8005
    },
    {
      "entropy": 6.317342233657837,
      "epoch": 1.5626768120183396,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047874154129187474,
      "loss": 6.1424,
      "mean_token_accuracy": 0.14485514685511588,
      "num_tokens": 26779943.0,
      "step": 8010
    },
    {
      "entropy": 6.451746559143066,
      "epoch": 1.5636523266022828,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047871168799423406,
      "loss": 6.3515,
      "mean_token_accuracy": 0.13083617016673088,
      "num_tokens": 26797607.0,
      "step": 8015
    },
    {
      "entropy": 6.513886451721191,
      "epoch": 1.5646278411862258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047868181479061324,
      "loss": 6.402,
      "mean_token_accuracy": 0.1266683228313923,
      "num_tokens": 26814668.0,
      "step": 8020
    },
    {
      "entropy": 6.4552935600280765,
      "epoch": 1.5656033557701687,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004786519216839314,
      "loss": 6.3097,
      "mean_token_accuracy": 0.13623356446623802,
      "num_tokens": 26831590.0,
      "step": 8025
    },
    {
      "entropy": 6.220515775680542,
      "epoch": 1.5665788703541117,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004786220086771096,
      "loss": 6.0777,
      "mean_token_accuracy": 0.1477731131017208,
      "num_tokens": 26848581.0,
      "step": 8030
    },
    {
      "entropy": 6.475474643707275,
      "epoch": 1.5675543849380549,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047859207577307083,
      "loss": 6.3758,
      "mean_token_accuracy": 0.12954323664307593,
      "num_tokens": 26867599.0,
      "step": 8035
    },
    {
      "entropy": 6.362512731552124,
      "epoch": 1.5685298995219978,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047856212297474007,
      "loss": 6.1823,
      "mean_token_accuracy": 0.1465704545378685,
      "num_tokens": 26884555.0,
      "step": 8040
    },
    {
      "entropy": 6.40194091796875,
      "epoch": 1.569505414105941,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047853215028504406,
      "loss": 6.2836,
      "mean_token_accuracy": 0.1429956652224064,
      "num_tokens": 26899988.0,
      "step": 8045
    },
    {
      "entropy": 6.466417598724365,
      "epoch": 1.570480928689884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004785021577069119,
      "loss": 6.392,
      "mean_token_accuracy": 0.13400833755731584,
      "num_tokens": 26917588.0,
      "step": 8050
    },
    {
      "entropy": 6.4591504573822025,
      "epoch": 1.571456443273827,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000478472145243274,
      "loss": 6.28,
      "mean_token_accuracy": 0.13716477081179618,
      "num_tokens": 26934053.0,
      "step": 8055
    },
    {
      "entropy": 6.270146656036377,
      "epoch": 1.57243195785777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047844211289706337,
      "loss": 6.1756,
      "mean_token_accuracy": 0.14785805121064186,
      "num_tokens": 26953870.0,
      "step": 8060
    },
    {
      "entropy": 6.51096510887146,
      "epoch": 1.5734074724417129,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047841206067121447,
      "loss": 6.4309,
      "mean_token_accuracy": 0.1278168261051178,
      "num_tokens": 26971944.0,
      "step": 8065
    },
    {
      "entropy": 6.433555603027344,
      "epoch": 1.574382987025656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000478381988568664,
      "loss": 6.3312,
      "mean_token_accuracy": 0.13435147553682328,
      "num_tokens": 26988723.0,
      "step": 8070
    },
    {
      "entropy": 6.451853227615357,
      "epoch": 1.5753585016095992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783518965923505,
      "loss": 6.3344,
      "mean_token_accuracy": 0.13361834958195687,
      "num_tokens": 27004553.0,
      "step": 8075
    },
    {
      "entropy": 6.434520196914673,
      "epoch": 1.5763340161935422,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004783217847452144,
      "loss": 6.2621,
      "mean_token_accuracy": 0.1380446247756481,
      "num_tokens": 27021241.0,
      "step": 8080
    },
    {
      "entropy": 6.314197683334351,
      "epoch": 1.5773095307774851,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047829165303019816,
      "loss": 6.1935,
      "mean_token_accuracy": 0.14739861860871314,
      "num_tokens": 27038718.0,
      "step": 8085
    },
    {
      "entropy": 6.420303535461426,
      "epoch": 1.578285045361428,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004782615014502462,
      "loss": 6.3271,
      "mean_token_accuracy": 0.13443446382880211,
      "num_tokens": 27055372.0,
      "step": 8090
    },
    {
      "entropy": 6.498059558868408,
      "epoch": 1.579260559945371,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047823133000830467,
      "loss": 6.3519,
      "mean_token_accuracy": 0.12907275408506394,
      "num_tokens": 27070433.0,
      "step": 8095
    },
    {
      "entropy": 6.381244087219239,
      "epoch": 1.5802360745293142,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004782011387073219,
      "loss": 6.2905,
      "mean_token_accuracy": 0.13929340243339539,
      "num_tokens": 27087467.0,
      "step": 8100
    },
    {
      "entropy": 6.445773553848267,
      "epoch": 1.5812115891132572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004781709275502481,
      "loss": 6.3259,
      "mean_token_accuracy": 0.12723207250237464,
      "num_tokens": 27103388.0,
      "step": 8105
    },
    {
      "entropy": 6.512648439407348,
      "epoch": 1.5821871036972004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047814069654003534,
      "loss": 6.4125,
      "mean_token_accuracy": 0.12939900606870652,
      "num_tokens": 27119752.0,
      "step": 8110
    },
    {
      "entropy": 6.439101696014404,
      "epoch": 1.5831626182811434,
      "grad_norm": 1.0,
      "learning_rate": 0.00047811044567963774,
      "loss": 6.297,
      "mean_token_accuracy": 0.1302100419998169,
      "num_tokens": 27136060.0,
      "step": 8115
    },
    {
      "entropy": 6.422860527038575,
      "epoch": 1.5841381328650863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004780801749720113,
      "loss": 6.3203,
      "mean_token_accuracy": 0.1295614145696163,
      "num_tokens": 27153151.0,
      "step": 8120
    },
    {
      "entropy": 6.318335008621216,
      "epoch": 1.5851136474490293,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047804988442011394,
      "loss": 6.2097,
      "mean_token_accuracy": 0.1473698250949383,
      "num_tokens": 27170806.0,
      "step": 8125
    },
    {
      "entropy": 6.471773862838745,
      "epoch": 1.5860891620329722,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004780195740269056,
      "loss": 6.3751,
      "mean_token_accuracy": 0.1285987041890621,
      "num_tokens": 27187082.0,
      "step": 8130
    },
    {
      "entropy": 6.372449922561645,
      "epoch": 1.5870646766169154,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000477989243795348,
      "loss": 6.2192,
      "mean_token_accuracy": 0.14449946954846382,
      "num_tokens": 27204505.0,
      "step": 8135
    },
    {
      "entropy": 6.279645872116089,
      "epoch": 1.5880401912008586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047795889372840493,
      "loss": 6.1497,
      "mean_token_accuracy": 0.1515430510044098,
      "num_tokens": 27219897.0,
      "step": 8140
    },
    {
      "entropy": 6.323049926757813,
      "epoch": 1.5890157057848016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047792852382904223,
      "loss": 6.186,
      "mean_token_accuracy": 0.14364069923758507,
      "num_tokens": 27235295.0,
      "step": 8145
    },
    {
      "entropy": 6.324961090087891,
      "epoch": 1.5899912203687445,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004778981341002274,
      "loss": 6.1995,
      "mean_token_accuracy": 0.1379023849964142,
      "num_tokens": 27253172.0,
      "step": 8150
    },
    {
      "entropy": 6.488235950469971,
      "epoch": 1.5909667349526875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047786772454493,
      "loss": 6.4259,
      "mean_token_accuracy": 0.128464412689209,
      "num_tokens": 27269958.0,
      "step": 8155
    },
    {
      "entropy": 6.377204275131225,
      "epoch": 1.5919422495366304,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047783729516612163,
      "loss": 6.1311,
      "mean_token_accuracy": 0.15012188032269477,
      "num_tokens": 27286402.0,
      "step": 8160
    },
    {
      "entropy": 6.425440645217895,
      "epoch": 1.5929177641205736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047780684596677563,
      "loss": 6.3167,
      "mean_token_accuracy": 0.12361339628696441,
      "num_tokens": 27301257.0,
      "step": 8165
    },
    {
      "entropy": 6.468948268890381,
      "epoch": 1.5938932787045166,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004777763769498676,
      "loss": 6.3413,
      "mean_token_accuracy": 0.12634046226739884,
      "num_tokens": 27318951.0,
      "step": 8170
    },
    {
      "entropy": 6.494255208969117,
      "epoch": 1.5948687932884598,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047774588811837465,
      "loss": 6.3903,
      "mean_token_accuracy": 0.13318117335438728,
      "num_tokens": 27334538.0,
      "step": 8175
    },
    {
      "entropy": 6.455736112594605,
      "epoch": 1.5958443078724027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047771537947527605,
      "loss": 6.3086,
      "mean_token_accuracy": 0.13645913153886796,
      "num_tokens": 27351769.0,
      "step": 8180
    },
    {
      "entropy": 6.4058098793029785,
      "epoch": 1.5968198224563457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047768485102355314,
      "loss": 6.316,
      "mean_token_accuracy": 0.13207341358065605,
      "num_tokens": 27369085.0,
      "step": 8185
    },
    {
      "entropy": 6.50387020111084,
      "epoch": 1.5977953370402886,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047765430276618896,
      "loss": 6.4418,
      "mean_token_accuracy": 0.13068175688385963,
      "num_tokens": 27385601.0,
      "step": 8190
    },
    {
      "entropy": 6.44583568572998,
      "epoch": 1.5987708516242318,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004776237347061685,
      "loss": 6.2733,
      "mean_token_accuracy": 0.134273449331522,
      "num_tokens": 27401973.0,
      "step": 8195
    },
    {
      "entropy": 6.3113452911376955,
      "epoch": 1.5997463662081748,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047759314684647894,
      "loss": 6.2274,
      "mean_token_accuracy": 0.13745418414473534,
      "num_tokens": 27418363.0,
      "step": 8200
    },
    {
      "entropy": 6.258568239212036,
      "epoch": 1.600721880792118,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047756253919010907,
      "loss": 6.1061,
      "mean_token_accuracy": 0.1510195516049862,
      "num_tokens": 27435569.0,
      "step": 8205
    },
    {
      "entropy": 6.393796730041504,
      "epoch": 1.601697395376061,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047753191174004984,
      "loss": 6.2901,
      "mean_token_accuracy": 0.13171105310320855,
      "num_tokens": 27451036.0,
      "step": 8210
    },
    {
      "entropy": 6.388012504577636,
      "epoch": 1.6026729099600039,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000477501264499294,
      "loss": 6.2505,
      "mean_token_accuracy": 0.14273630380630492,
      "num_tokens": 27466425.0,
      "step": 8215
    },
    {
      "entropy": 6.377392864227295,
      "epoch": 1.6036484245439468,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004774705974708364,
      "loss": 6.2802,
      "mean_token_accuracy": 0.13445906341075897,
      "num_tokens": 27481502.0,
      "step": 8220
    },
    {
      "entropy": 6.300930404663086,
      "epoch": 1.6046239391278898,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047743991065767364,
      "loss": 6.2441,
      "mean_token_accuracy": 0.1396934188902378,
      "num_tokens": 27498493.0,
      "step": 8225
    },
    {
      "entropy": 6.507022666931152,
      "epoch": 1.605599453711833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047740920406280423,
      "loss": 6.3269,
      "mean_token_accuracy": 0.13326430320739746,
      "num_tokens": 27514401.0,
      "step": 8230
    },
    {
      "entropy": 6.434348440170288,
      "epoch": 1.6065749682957762,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047737847768922887,
      "loss": 6.3769,
      "mean_token_accuracy": 0.13277071267366408,
      "num_tokens": 27531100.0,
      "step": 8235
    },
    {
      "entropy": 6.523198366165161,
      "epoch": 1.6075504828797191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004773477315399499,
      "loss": 6.4128,
      "mean_token_accuracy": 0.1317256972193718,
      "num_tokens": 27549097.0,
      "step": 8240
    },
    {
      "entropy": 6.501363325119018,
      "epoch": 1.608525997463662,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047731696561797185,
      "loss": 6.3365,
      "mean_token_accuracy": 0.12750333100557326,
      "num_tokens": 27565485.0,
      "step": 8245
    },
    {
      "entropy": 6.349027442932129,
      "epoch": 1.609501512047605,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047728617992630096,
      "loss": 6.2491,
      "mean_token_accuracy": 0.14667394757270813,
      "num_tokens": 27581832.0,
      "step": 8250
    },
    {
      "entropy": 6.313298892974854,
      "epoch": 1.610477026631548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047725537446794547,
      "loss": 6.1942,
      "mean_token_accuracy": 0.1440240904688835,
      "num_tokens": 27598149.0,
      "step": 8255
    },
    {
      "entropy": 6.255907583236694,
      "epoch": 1.6114525412154912,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004772245492459157,
      "loss": 6.1058,
      "mean_token_accuracy": 0.15090591982007026,
      "num_tokens": 27616920.0,
      "step": 8260
    },
    {
      "entropy": 6.45150990486145,
      "epoch": 1.6124280557994342,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047719370426322374,
      "loss": 6.3404,
      "mean_token_accuracy": 0.12934081330895425,
      "num_tokens": 27632702.0,
      "step": 8265
    },
    {
      "entropy": 6.426601123809815,
      "epoch": 1.6134035703833773,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004771628395228836,
      "loss": 6.3586,
      "mean_token_accuracy": 0.1288245402276516,
      "num_tokens": 27650606.0,
      "step": 8270
    },
    {
      "entropy": 6.398304462432861,
      "epoch": 1.6143790849673203,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004771319550279113,
      "loss": 6.2247,
      "mean_token_accuracy": 0.1438526637852192,
      "num_tokens": 27668466.0,
      "step": 8275
    },
    {
      "entropy": 6.414252281188965,
      "epoch": 1.6153545995512633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004771010507813247,
      "loss": 6.2284,
      "mean_token_accuracy": 0.138103149831295,
      "num_tokens": 27685032.0,
      "step": 8280
    },
    {
      "entropy": 6.37470006942749,
      "epoch": 1.6163301141352062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004770701267861439,
      "loss": 6.3126,
      "mean_token_accuracy": 0.13117850869894027,
      "num_tokens": 27700382.0,
      "step": 8285
    },
    {
      "entropy": 6.490540552139282,
      "epoch": 1.6173056287191494,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047703918304539027,
      "loss": 6.3199,
      "mean_token_accuracy": 0.13244666978716851,
      "num_tokens": 27715603.0,
      "step": 8290
    },
    {
      "entropy": 6.361236572265625,
      "epoch": 1.6182811433030924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004770082195620879,
      "loss": 6.2412,
      "mean_token_accuracy": 0.13519721552729608,
      "num_tokens": 27732341.0,
      "step": 8295
    },
    {
      "entropy": 6.451249122619629,
      "epoch": 1.6192566578870355,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004769772363392622,
      "loss": 6.3076,
      "mean_token_accuracy": 0.1390954963862896,
      "num_tokens": 27747731.0,
      "step": 8300
    },
    {
      "entropy": 6.353865432739258,
      "epoch": 1.6202321724709785,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769462333799409,
      "loss": 6.2993,
      "mean_token_accuracy": 0.13447826877236366,
      "num_tokens": 27764873.0,
      "step": 8305
    },
    {
      "entropy": 6.387086677551269,
      "epoch": 1.6212076870549215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047691521068715344,
      "loss": 6.2596,
      "mean_token_accuracy": 0.13489287719130516,
      "num_tokens": 27781684.0,
      "step": 8310
    },
    {
      "entropy": 6.496578979492187,
      "epoch": 1.6221832016388644,
      "grad_norm": 1.125,
      "learning_rate": 0.0004768841682639312,
      "loss": 6.3891,
      "mean_token_accuracy": 0.12069439962506294,
      "num_tokens": 27799596.0,
      "step": 8315
    },
    {
      "entropy": 6.466693639755249,
      "epoch": 1.6231587162228074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004768531061133075,
      "loss": 6.359,
      "mean_token_accuracy": 0.12753675282001495,
      "num_tokens": 27816469.0,
      "step": 8320
    },
    {
      "entropy": 6.520159149169922,
      "epoch": 1.6241342308067506,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004768220242383178,
      "loss": 6.4098,
      "mean_token_accuracy": 0.126971348375082,
      "num_tokens": 27832794.0,
      "step": 8325
    },
    {
      "entropy": 6.32980055809021,
      "epoch": 1.6251097453906937,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047679092264199915,
      "loss": 6.2248,
      "mean_token_accuracy": 0.14176104292273523,
      "num_tokens": 27851019.0,
      "step": 8330
    },
    {
      "entropy": 6.48318395614624,
      "epoch": 1.6260852599746367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047675980132739067,
      "loss": 6.3602,
      "mean_token_accuracy": 0.13410745337605476,
      "num_tokens": 27869377.0,
      "step": 8335
    },
    {
      "entropy": 6.537897062301636,
      "epoch": 1.6270607745585797,
      "grad_norm": 1.0,
      "learning_rate": 0.0004767286602975336,
      "loss": 6.3335,
      "mean_token_accuracy": 0.1380740635097027,
      "num_tokens": 27887196.0,
      "step": 8340
    },
    {
      "entropy": 6.44108362197876,
      "epoch": 1.6280362891425226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004766974995554707,
      "loss": 6.3324,
      "mean_token_accuracy": 0.12853021174669266,
      "num_tokens": 27903536.0,
      "step": 8345
    },
    {
      "entropy": 6.292126226425171,
      "epoch": 1.6290118037264656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004766663191042471,
      "loss": 6.1952,
      "mean_token_accuracy": 0.1439802974462509,
      "num_tokens": 27919081.0,
      "step": 8350
    },
    {
      "entropy": 6.424186277389526,
      "epoch": 1.6299873183104088,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047663511894690955,
      "loss": 6.2707,
      "mean_token_accuracy": 0.1330537110567093,
      "num_tokens": 27934432.0,
      "step": 8355
    },
    {
      "entropy": 6.395192193984985,
      "epoch": 1.6309628328943517,
      "grad_norm": 1.0,
      "learning_rate": 0.0004766038990865068,
      "loss": 6.3138,
      "mean_token_accuracy": 0.13281900584697723,
      "num_tokens": 27950395.0,
      "step": 8360
    },
    {
      "entropy": 6.425386095046997,
      "epoch": 1.631938347478295,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047657265952608953,
      "loss": 6.3406,
      "mean_token_accuracy": 0.13578801676630975,
      "num_tokens": 27968051.0,
      "step": 8365
    },
    {
      "entropy": 6.4332836151123045,
      "epoch": 1.6329138620622379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004765414002687104,
      "loss": 6.1978,
      "mean_token_accuracy": 0.14416124373674394,
      "num_tokens": 27984942.0,
      "step": 8370
    },
    {
      "entropy": 6.4581831932067875,
      "epoch": 1.6338893766461808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004765101213174239,
      "loss": 6.3752,
      "mean_token_accuracy": 0.13114355355501175,
      "num_tokens": 28001861.0,
      "step": 8375
    },
    {
      "entropy": 6.404871416091919,
      "epoch": 1.6348648912301238,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047647882267528663,
      "loss": 6.303,
      "mean_token_accuracy": 0.13203059136867523,
      "num_tokens": 28018115.0,
      "step": 8380
    },
    {
      "entropy": 6.43725266456604,
      "epoch": 1.6358404058140668,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004764475043453569,
      "loss": 6.2681,
      "mean_token_accuracy": 0.1339235506951809,
      "num_tokens": 28033929.0,
      "step": 8385
    },
    {
      "entropy": 6.269834852218628,
      "epoch": 1.63681592039801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004764161663306949,
      "loss": 6.1309,
      "mean_token_accuracy": 0.14604141488671302,
      "num_tokens": 28050876.0,
      "step": 8390
    },
    {
      "entropy": 6.4788166046142575,
      "epoch": 1.6377914349819531,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004763848086343631,
      "loss": 6.3668,
      "mean_token_accuracy": 0.12799584567546846,
      "num_tokens": 28067453.0,
      "step": 8395
    },
    {
      "entropy": 6.455996799468994,
      "epoch": 1.638766949565896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763534312594254,
      "loss": 6.3578,
      "mean_token_accuracy": 0.12765470072627066,
      "num_tokens": 28084463.0,
      "step": 8400
    },
    {
      "entropy": 6.422712135314941,
      "epoch": 1.639742464149839,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763220342089482,
      "loss": 6.2937,
      "mean_token_accuracy": 0.1446809157729149,
      "num_tokens": 28101126.0,
      "step": 8405
    },
    {
      "entropy": 6.431754112243652,
      "epoch": 1.640717978733782,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047629061748599927,
      "loss": 6.32,
      "mean_token_accuracy": 0.13405325263738632,
      "num_tokens": 28117000.0,
      "step": 8410
    },
    {
      "entropy": 6.425334072113037,
      "epoch": 1.641693493317725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047625918109364863,
      "loss": 6.355,
      "mean_token_accuracy": 0.12643185630440712,
      "num_tokens": 28132290.0,
      "step": 8415
    },
    {
      "entropy": 6.409068393707275,
      "epoch": 1.6426690079016681,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047622772503496806,
      "loss": 6.2488,
      "mean_token_accuracy": 0.1315213546156883,
      "num_tokens": 28149884.0,
      "step": 8420
    },
    {
      "entropy": 6.375530624389649,
      "epoch": 1.6436445224856113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047619624931303145,
      "loss": 6.2393,
      "mean_token_accuracy": 0.1452097401022911,
      "num_tokens": 28166678.0,
      "step": 8425
    },
    {
      "entropy": 6.363833570480347,
      "epoch": 1.6446200370695543,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761647539309144,
      "loss": 6.2862,
      "mean_token_accuracy": 0.14252205938100815,
      "num_tokens": 28184921.0,
      "step": 8430
    },
    {
      "entropy": 6.371412849426269,
      "epoch": 1.6455955516534972,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047613323889169457,
      "loss": 6.2768,
      "mean_token_accuracy": 0.13671627566218375,
      "num_tokens": 28203503.0,
      "step": 8435
    },
    {
      "entropy": 6.450017881393433,
      "epoch": 1.6465710662374402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004761017041984515,
      "loss": 6.3213,
      "mean_token_accuracy": 0.137418295443058,
      "num_tokens": 28220542.0,
      "step": 8440
    },
    {
      "entropy": 6.41711459159851,
      "epoch": 1.6475465808213832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760701498542666,
      "loss": 6.2726,
      "mean_token_accuracy": 0.13781270757317543,
      "num_tokens": 28236680.0,
      "step": 8445
    },
    {
      "entropy": 6.309499788284302,
      "epoch": 1.6485220954053263,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047603857586222333,
      "loss": 6.1262,
      "mean_token_accuracy": 0.14610913321375846,
      "num_tokens": 28253690.0,
      "step": 8450
    },
    {
      "entropy": 6.368608283996582,
      "epoch": 1.6494976099892693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047600698222540687,
      "loss": 6.1704,
      "mean_token_accuracy": 0.1435729369521141,
      "num_tokens": 28270528.0,
      "step": 8455
    },
    {
      "entropy": 6.40486216545105,
      "epoch": 1.6504731245732125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004759753689469046,
      "loss": 6.2673,
      "mean_token_accuracy": 0.13488116189837457,
      "num_tokens": 28287470.0,
      "step": 8460
    },
    {
      "entropy": 6.38212366104126,
      "epoch": 1.6514486391571555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004759437360298055,
      "loss": 6.3123,
      "mean_token_accuracy": 0.1358719952404499,
      "num_tokens": 28304295.0,
      "step": 8465
    },
    {
      "entropy": 6.532471990585327,
      "epoch": 1.6524241537410984,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047591208347720073,
      "loss": 6.4026,
      "mean_token_accuracy": 0.12927182167768478,
      "num_tokens": 28319669.0,
      "step": 8470
    },
    {
      "entropy": 6.452768707275391,
      "epoch": 1.6533996683250414,
      "grad_norm": 1.0,
      "learning_rate": 0.00047588041129218323,
      "loss": 6.3759,
      "mean_token_accuracy": 0.13267260938882827,
      "num_tokens": 28336251.0,
      "step": 8475
    },
    {
      "entropy": 6.4818957328796385,
      "epoch": 1.6543751829089843,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047584871947784787,
      "loss": 6.4029,
      "mean_token_accuracy": 0.13248071894049646,
      "num_tokens": 28352916.0,
      "step": 8480
    },
    {
      "entropy": 6.446785974502563,
      "epoch": 1.6553506974929275,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004758170080372915,
      "loss": 6.2732,
      "mean_token_accuracy": 0.12973505780100822,
      "num_tokens": 28369452.0,
      "step": 8485
    },
    {
      "entropy": 6.360740995407104,
      "epoch": 1.6563262120768707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047578527697361277,
      "loss": 6.2438,
      "mean_token_accuracy": 0.14284233301877974,
      "num_tokens": 28384932.0,
      "step": 8490
    },
    {
      "entropy": 6.462392377853393,
      "epoch": 1.6573017266608137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004757535262899124,
      "loss": 6.2916,
      "mean_token_accuracy": 0.1353907398879528,
      "num_tokens": 28400460.0,
      "step": 8495
    },
    {
      "entropy": 6.385865354537964,
      "epoch": 1.6582772412447566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000475721755989293,
      "loss": 6.3243,
      "mean_token_accuracy": 0.130817062407732,
      "num_tokens": 28417883.0,
      "step": 8500
    },
    {
      "entropy": 6.431553745269776,
      "epoch": 1.6592527558286996,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047568996607485895,
      "loss": 6.2669,
      "mean_token_accuracy": 0.13071645870804788,
      "num_tokens": 28434657.0,
      "step": 8505
    },
    {
      "entropy": 6.428481578826904,
      "epoch": 1.6602282704126425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004756581565497167,
      "loss": 6.3601,
      "mean_token_accuracy": 0.13361974507570268,
      "num_tokens": 28452767.0,
      "step": 8510
    },
    {
      "entropy": 6.5523600578308105,
      "epoch": 1.6612037849965857,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047562632741697454,
      "loss": 6.3971,
      "mean_token_accuracy": 0.1276608809828758,
      "num_tokens": 28469470.0,
      "step": 8515
    },
    {
      "entropy": 6.41212649345398,
      "epoch": 1.6621792995805287,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004755944786797427,
      "loss": 6.278,
      "mean_token_accuracy": 0.13593624383211136,
      "num_tokens": 28486767.0,
      "step": 8520
    },
    {
      "entropy": 6.428006553649903,
      "epoch": 1.6631548141644719,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004755626103411333,
      "loss": 6.3419,
      "mean_token_accuracy": 0.13119750693440438,
      "num_tokens": 28504324.0,
      "step": 8525
    },
    {
      "entropy": 6.369665002822876,
      "epoch": 1.6641303287484148,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004755307224042605,
      "loss": 6.1738,
      "mean_token_accuracy": 0.14318314865231513,
      "num_tokens": 28520774.0,
      "step": 8530
    },
    {
      "entropy": 6.334472036361694,
      "epoch": 1.6651058433323578,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004754988148722402,
      "loss": 6.2696,
      "mean_token_accuracy": 0.1376575693488121,
      "num_tokens": 28536754.0,
      "step": 8535
    },
    {
      "entropy": 6.339354610443115,
      "epoch": 1.6660813579163007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004754668877481903,
      "loss": 6.1798,
      "mean_token_accuracy": 0.1440943792462349,
      "num_tokens": 28552866.0,
      "step": 8540
    },
    {
      "entropy": 6.488823127746582,
      "epoch": 1.667056872500244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004754349410352306,
      "loss": 6.3912,
      "mean_token_accuracy": 0.12401873096823693,
      "num_tokens": 28568802.0,
      "step": 8545
    },
    {
      "entropy": 6.409925270080566,
      "epoch": 1.6680323870841869,
      "grad_norm": 1.0,
      "learning_rate": 0.0004754029747364828,
      "loss": 6.2331,
      "mean_token_accuracy": 0.1338075466454029,
      "num_tokens": 28585050.0,
      "step": 8550
    },
    {
      "entropy": 6.401282787322998,
      "epoch": 1.66900790166813,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047537098885507055,
      "loss": 6.2749,
      "mean_token_accuracy": 0.13893224447965621,
      "num_tokens": 28602319.0,
      "step": 8555
    },
    {
      "entropy": 6.456971168518066,
      "epoch": 1.669983416252073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047533898339411945,
      "loss": 6.2949,
      "mean_token_accuracy": 0.1425710864365101,
      "num_tokens": 28620694.0,
      "step": 8560
    },
    {
      "entropy": 6.267441463470459,
      "epoch": 1.670958930836016,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004753069583567569,
      "loss": 6.0801,
      "mean_token_accuracy": 0.1455279178917408,
      "num_tokens": 28638309.0,
      "step": 8565
    },
    {
      "entropy": 6.382388782501221,
      "epoch": 1.671934445419959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752749137461123,
      "loss": 6.2544,
      "mean_token_accuracy": 0.13416807800531388,
      "num_tokens": 28654713.0,
      "step": 8570
    },
    {
      "entropy": 6.429929637908936,
      "epoch": 1.672909960003902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004752428495653169,
      "loss": 6.2749,
      "mean_token_accuracy": 0.13658047690987588,
      "num_tokens": 28671025.0,
      "step": 8575
    },
    {
      "entropy": 6.251125144958496,
      "epoch": 1.673885474587845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047521076581750387,
      "loss": 6.1885,
      "mean_token_accuracy": 0.14111858159303664,
      "num_tokens": 28687230.0,
      "step": 8580
    },
    {
      "entropy": 6.438810873031616,
      "epoch": 1.6748609891717883,
      "grad_norm": 1.0,
      "learning_rate": 0.0004751786625058085,
      "loss": 6.2951,
      "mean_token_accuracy": 0.132166263461113,
      "num_tokens": 28702595.0,
      "step": 8585
    },
    {
      "entropy": 6.381994009017944,
      "epoch": 1.6758365037557312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004751465396333676,
      "loss": 6.2496,
      "mean_token_accuracy": 0.1432796224951744,
      "num_tokens": 28719656.0,
      "step": 8590
    },
    {
      "entropy": 6.153446960449219,
      "epoch": 1.6768120183396742,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047511439720332015,
      "loss": 5.9729,
      "mean_token_accuracy": 0.16458074226975442,
      "num_tokens": 28737289.0,
      "step": 8595
    },
    {
      "entropy": 6.377745819091797,
      "epoch": 1.6777875329236172,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047508223521880704,
      "loss": 6.3337,
      "mean_token_accuracy": 0.13355341106653212,
      "num_tokens": 28752214.0,
      "step": 8600
    },
    {
      "entropy": 6.372812366485595,
      "epoch": 1.6787630475075601,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004750500536829711,
      "loss": 6.2371,
      "mean_token_accuracy": 0.13555908203125,
      "num_tokens": 28769845.0,
      "step": 8605
    },
    {
      "entropy": 6.529575681686401,
      "epoch": 1.6797385620915033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047501785259895676,
      "loss": 6.3404,
      "mean_token_accuracy": 0.12646386176347732,
      "num_tokens": 28786273.0,
      "step": 8610
    },
    {
      "entropy": 6.3848823547363285,
      "epoch": 1.6807140766754463,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004749856319699108,
      "loss": 6.303,
      "mean_token_accuracy": 0.13209845945239068,
      "num_tokens": 28803171.0,
      "step": 8615
    },
    {
      "entropy": 6.3997588634490965,
      "epoch": 1.6816895912593894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004749533917989817,
      "loss": 6.252,
      "mean_token_accuracy": 0.13916246816515923,
      "num_tokens": 28819585.0,
      "step": 8620
    },
    {
      "entropy": 6.302767992019653,
      "epoch": 1.6826651058433324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047492113208931974,
      "loss": 6.1312,
      "mean_token_accuracy": 0.15182471871376038,
      "num_tokens": 28837158.0,
      "step": 8625
    },
    {
      "entropy": 6.431247711181641,
      "epoch": 1.6836406204272754,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047488885284407734,
      "loss": 6.2287,
      "mean_token_accuracy": 0.13558847680687905,
      "num_tokens": 28852804.0,
      "step": 8630
    },
    {
      "entropy": 6.436433553695679,
      "epoch": 1.6846161350112183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004748565540664086,
      "loss": 6.3561,
      "mean_token_accuracy": 0.1313096486032009,
      "num_tokens": 28870336.0,
      "step": 8635
    },
    {
      "entropy": 6.461747407913208,
      "epoch": 1.6855916495951615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004748242357594697,
      "loss": 6.2322,
      "mean_token_accuracy": 0.13683086037635803,
      "num_tokens": 28886197.0,
      "step": 8640
    },
    {
      "entropy": 6.291606044769287,
      "epoch": 1.6865671641791045,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004747918979264186,
      "loss": 6.1809,
      "mean_token_accuracy": 0.1375397302210331,
      "num_tokens": 28902520.0,
      "step": 8645
    },
    {
      "entropy": 6.428701734542846,
      "epoch": 1.6875426787630476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004747595405704154,
      "loss": 6.3021,
      "mean_token_accuracy": 0.13296685218811036,
      "num_tokens": 28918725.0,
      "step": 8650
    },
    {
      "entropy": 6.4789652824401855,
      "epoch": 1.6885181933469906,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004747271636946219,
      "loss": 6.3218,
      "mean_token_accuracy": 0.13231525868177413,
      "num_tokens": 28936001.0,
      "step": 8655
    },
    {
      "entropy": 6.506434965133667,
      "epoch": 1.6894937079309336,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004746947673022018,
      "loss": 6.4168,
      "mean_token_accuracy": 0.1277687855064869,
      "num_tokens": 28953832.0,
      "step": 8660
    },
    {
      "entropy": 6.431854963302612,
      "epoch": 1.6904692225148765,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004746623513963206,
      "loss": 6.2377,
      "mean_token_accuracy": 0.13375520557165146,
      "num_tokens": 28970982.0,
      "step": 8665
    },
    {
      "entropy": 6.36587929725647,
      "epoch": 1.6914447370988195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047462991598014624,
      "loss": 6.2354,
      "mean_token_accuracy": 0.1305945448577404,
      "num_tokens": 28986142.0,
      "step": 8670
    },
    {
      "entropy": 6.374372673034668,
      "epoch": 1.6924202516827627,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047459746105684786,
      "loss": 6.265,
      "mean_token_accuracy": 0.14048944786190987,
      "num_tokens": 29002244.0,
      "step": 8675
    },
    {
      "entropy": 6.364422988891602,
      "epoch": 1.6933957662667058,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047456498662959694,
      "loss": 6.232,
      "mean_token_accuracy": 0.15143098756670953,
      "num_tokens": 29018667.0,
      "step": 8680
    },
    {
      "entropy": 6.452254104614258,
      "epoch": 1.6943712808506488,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004745324927015669,
      "loss": 6.2889,
      "mean_token_accuracy": 0.1306975193321705,
      "num_tokens": 29034203.0,
      "step": 8685
    },
    {
      "entropy": 6.317226505279541,
      "epoch": 1.6953467954345918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004744999792759327,
      "loss": 6.1885,
      "mean_token_accuracy": 0.1390099599957466,
      "num_tokens": 29050181.0,
      "step": 8690
    },
    {
      "entropy": 6.4422258853912355,
      "epoch": 1.6963223100185347,
      "grad_norm": 1.0,
      "learning_rate": 0.0004744674463558717,
      "loss": 6.3341,
      "mean_token_accuracy": 0.13227170556783677,
      "num_tokens": 29066822.0,
      "step": 8695
    },
    {
      "entropy": 6.417118978500366,
      "epoch": 1.6972978246024777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047443489394456256,
      "loss": 6.244,
      "mean_token_accuracy": 0.13796414583921432,
      "num_tokens": 29084148.0,
      "step": 8700
    },
    {
      "entropy": 6.429953336715698,
      "epoch": 1.6982733391864209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047440232204518654,
      "loss": 6.3536,
      "mean_token_accuracy": 0.13077428862452506,
      "num_tokens": 29099811.0,
      "step": 8705
    },
    {
      "entropy": 6.444475746154785,
      "epoch": 1.6992488537703638,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004743697306609262,
      "loss": 6.2689,
      "mean_token_accuracy": 0.13390282690525054,
      "num_tokens": 29115327.0,
      "step": 8710
    },
    {
      "entropy": 6.346184730529785,
      "epoch": 1.700224368354307,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004743371197949664,
      "loss": 6.2632,
      "mean_token_accuracy": 0.13268154188990594,
      "num_tokens": 29132571.0,
      "step": 8715
    },
    {
      "entropy": 6.4102729797363285,
      "epoch": 1.70119988293825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047430448945049363,
      "loss": 6.2658,
      "mean_token_accuracy": 0.13570413663983344,
      "num_tokens": 29149256.0,
      "step": 8720
    },
    {
      "entropy": 6.352029848098755,
      "epoch": 1.702175397522193,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004742718396306966,
      "loss": 6.1939,
      "mean_token_accuracy": 0.14170687273144722,
      "num_tokens": 29166380.0,
      "step": 8725
    },
    {
      "entropy": 6.357316780090332,
      "epoch": 1.703150912106136,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004742391703387656,
      "loss": 6.2361,
      "mean_token_accuracy": 0.14293377250432968,
      "num_tokens": 29183056.0,
      "step": 8730
    },
    {
      "entropy": 6.346543073654175,
      "epoch": 1.7041264266900789,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000474206481577893,
      "loss": 6.2452,
      "mean_token_accuracy": 0.13888408094644547,
      "num_tokens": 29199631.0,
      "step": 8735
    },
    {
      "entropy": 6.440142965316772,
      "epoch": 1.705101941274022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004741737733512729,
      "loss": 6.2702,
      "mean_token_accuracy": 0.13485286980867386,
      "num_tokens": 29215074.0,
      "step": 8740
    },
    {
      "entropy": 6.422340297698975,
      "epoch": 1.7060774558579652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047414104566210157,
      "loss": 6.1804,
      "mean_token_accuracy": 0.1398899160325527,
      "num_tokens": 29232022.0,
      "step": 8745
    },
    {
      "entropy": 6.269046258926392,
      "epoch": 1.7070529704419082,
      "grad_norm": 1.03125,
      "learning_rate": 0.000474108298513577,
      "loss": 6.081,
      "mean_token_accuracy": 0.14734212979674338,
      "num_tokens": 29248040.0,
      "step": 8750
    },
    {
      "entropy": 6.435280656814575,
      "epoch": 1.7080284850258511,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047407553190889916,
      "loss": 6.3226,
      "mean_token_accuracy": 0.13844671323895455,
      "num_tokens": 29264338.0,
      "step": 8755
    },
    {
      "entropy": 6.294973754882813,
      "epoch": 1.709003999609794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004740427458512699,
      "loss": 6.1631,
      "mean_token_accuracy": 0.14153596609830857,
      "num_tokens": 29279483.0,
      "step": 8760
    },
    {
      "entropy": 6.396880722045898,
      "epoch": 1.709979514193737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004740099403438928,
      "loss": 6.2766,
      "mean_token_accuracy": 0.13402382880449296,
      "num_tokens": 29296582.0,
      "step": 8765
    },
    {
      "entropy": 6.446957921981811,
      "epoch": 1.7109550287776802,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047397711538997366,
      "loss": 6.3893,
      "mean_token_accuracy": 0.12600063905119896,
      "num_tokens": 29312772.0,
      "step": 8770
    },
    {
      "entropy": 6.3789294242858885,
      "epoch": 1.7119305433616232,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047394427099272003,
      "loss": 6.2137,
      "mean_token_accuracy": 0.13923029378056526,
      "num_tokens": 29328962.0,
      "step": 8775
    },
    {
      "entropy": 6.424732780456543,
      "epoch": 1.7129060579455664,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047391140715534127,
      "loss": 6.2914,
      "mean_token_accuracy": 0.13055554032325745,
      "num_tokens": 29346299.0,
      "step": 8780
    },
    {
      "entropy": 6.393940782546997,
      "epoch": 1.7138815725295093,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004738785238810487,
      "loss": 6.2199,
      "mean_token_accuracy": 0.14042325392365457,
      "num_tokens": 29362729.0,
      "step": 8785
    },
    {
      "entropy": 6.397880220413208,
      "epoch": 1.7148570871134523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004738456211730556,
      "loss": 6.3376,
      "mean_token_accuracy": 0.13567682653665541,
      "num_tokens": 29380807.0,
      "step": 8790
    },
    {
      "entropy": 6.3701417446136475,
      "epoch": 1.7158326016973953,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004738126990345771,
      "loss": 6.2231,
      "mean_token_accuracy": 0.1403884045779705,
      "num_tokens": 29397261.0,
      "step": 8795
    },
    {
      "entropy": 6.384936237335205,
      "epoch": 1.7168081162813384,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047377975746883025,
      "loss": 6.2674,
      "mean_token_accuracy": 0.13607952371239662,
      "num_tokens": 29413710.0,
      "step": 8800
    },
    {
      "entropy": 6.429215574264527,
      "epoch": 1.7177836308652814,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047374679647903397,
      "loss": 6.211,
      "mean_token_accuracy": 0.13602049127221108,
      "num_tokens": 29430178.0,
      "step": 8805
    },
    {
      "entropy": 6.329242420196533,
      "epoch": 1.7187591454492246,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004737138160684091,
      "loss": 6.3072,
      "mean_token_accuracy": 0.12894970551133156,
      "num_tokens": 29447602.0,
      "step": 8810
    },
    {
      "entropy": 6.465573072433472,
      "epoch": 1.7197346600331676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004736808162401783,
      "loss": 6.3337,
      "mean_token_accuracy": 0.1302557483315468,
      "num_tokens": 29463868.0,
      "step": 8815
    },
    {
      "entropy": 6.4676751613616945,
      "epoch": 1.7207101746171105,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047364779699756625,
      "loss": 6.2672,
      "mean_token_accuracy": 0.1393133282661438,
      "num_tokens": 29480263.0,
      "step": 8820
    },
    {
      "entropy": 6.085441923141479,
      "epoch": 1.7216856892010535,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047361475834379945,
      "loss": 5.955,
      "mean_token_accuracy": 0.17704318314790726,
      "num_tokens": 29498938.0,
      "step": 8825
    },
    {
      "entropy": 6.379257583618164,
      "epoch": 1.7226612037849964,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004735817002821064,
      "loss": 6.2638,
      "mean_token_accuracy": 0.13448890969157218,
      "num_tokens": 29515522.0,
      "step": 8830
    },
    {
      "entropy": 6.349101400375366,
      "epoch": 1.7236367183689396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047354862281571736,
      "loss": 6.2058,
      "mean_token_accuracy": 0.14130807295441628,
      "num_tokens": 29531050.0,
      "step": 8835
    },
    {
      "entropy": 6.285197639465332,
      "epoch": 1.7246122329528828,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004735155259478645,
      "loss": 6.1222,
      "mean_token_accuracy": 0.14642089158296584,
      "num_tokens": 29547432.0,
      "step": 8840
    },
    {
      "entropy": 6.420377635955811,
      "epoch": 1.7255877475368258,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000473482409681782,
      "loss": 6.2301,
      "mean_token_accuracy": 0.13720409646630288,
      "num_tokens": 29564272.0,
      "step": 8845
    },
    {
      "entropy": 6.416154956817627,
      "epoch": 1.7265632621207687,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047344927402070573,
      "loss": 6.2878,
      "mean_token_accuracy": 0.12901435121893884,
      "num_tokens": 29580433.0,
      "step": 8850
    },
    {
      "entropy": 6.2524085521698,
      "epoch": 1.7275387767047117,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734161189678738,
      "loss": 6.1614,
      "mean_token_accuracy": 0.1434662990272045,
      "num_tokens": 29597602.0,
      "step": 8855
    },
    {
      "entropy": 6.310213708877564,
      "epoch": 1.7285142912886546,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047338294452652584,
      "loss": 6.1361,
      "mean_token_accuracy": 0.1447230890393257,
      "num_tokens": 29616313.0,
      "step": 8860
    },
    {
      "entropy": 6.368034076690674,
      "epoch": 1.7294898058725978,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047334975069990363,
      "loss": 6.2206,
      "mean_token_accuracy": 0.14364252015948295,
      "num_tokens": 29633786.0,
      "step": 8865
    },
    {
      "entropy": 6.377895927429199,
      "epoch": 1.7304653204565408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004733165374912507,
      "loss": 6.2955,
      "mean_token_accuracy": 0.1325770601630211,
      "num_tokens": 29650125.0,
      "step": 8870
    },
    {
      "entropy": 6.440638685226441,
      "epoch": 1.731440835040484,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047328330490381255,
      "loss": 6.3142,
      "mean_token_accuracy": 0.1290828451514244,
      "num_tokens": 29667211.0,
      "step": 8875
    },
    {
      "entropy": 6.511521530151367,
      "epoch": 1.732416349624427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047325005294083654,
      "loss": 6.3497,
      "mean_token_accuracy": 0.12736388370394708,
      "num_tokens": 29683864.0,
      "step": 8880
    },
    {
      "entropy": 6.450252485275269,
      "epoch": 1.7333918642083699,
      "grad_norm": 1.0,
      "learning_rate": 0.00047321678160557187,
      "loss": 6.3704,
      "mean_token_accuracy": 0.12571600303053856,
      "num_tokens": 29699275.0,
      "step": 8885
    },
    {
      "entropy": 6.362071418762207,
      "epoch": 1.7343673787923128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047318349090126984,
      "loss": 6.2559,
      "mean_token_accuracy": 0.13488421365618705,
      "num_tokens": 29715773.0,
      "step": 8890
    },
    {
      "entropy": 6.324081993103027,
      "epoch": 1.735342893376256,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047315018083118334,
      "loss": 6.2284,
      "mean_token_accuracy": 0.14374122619628907,
      "num_tokens": 29732383.0,
      "step": 8895
    },
    {
      "entropy": 6.388317966461182,
      "epoch": 1.736318407960199,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004731168513985674,
      "loss": 6.238,
      "mean_token_accuracy": 0.13530940636992456,
      "num_tokens": 29749159.0,
      "step": 8900
    },
    {
      "entropy": 6.339236640930176,
      "epoch": 1.7372939225441422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047308350260667894,
      "loss": 6.1695,
      "mean_token_accuracy": 0.14472566395998002,
      "num_tokens": 29765357.0,
      "step": 8905
    },
    {
      "entropy": 6.35810284614563,
      "epoch": 1.7382694371280851,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004730501344587765,
      "loss": 6.229,
      "mean_token_accuracy": 0.1377372346818447,
      "num_tokens": 29781269.0,
      "step": 8910
    },
    {
      "entropy": 6.451261186599732,
      "epoch": 1.739244951712028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004730167469581208,
      "loss": 6.2888,
      "mean_token_accuracy": 0.13109415769577026,
      "num_tokens": 29798029.0,
      "step": 8915
    },
    {
      "entropy": 6.307126569747925,
      "epoch": 1.740220466295971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004729833401079743,
      "loss": 6.2331,
      "mean_token_accuracy": 0.1362668789923191,
      "num_tokens": 29813794.0,
      "step": 8920
    },
    {
      "entropy": 6.521246480941772,
      "epoch": 1.741195980879914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047294991391160147,
      "loss": 6.3859,
      "mean_token_accuracy": 0.12463150694966316,
      "num_tokens": 29829589.0,
      "step": 8925
    },
    {
      "entropy": 6.420650625228882,
      "epoch": 1.7421714954638572,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004729164683722685,
      "loss": 6.3052,
      "mean_token_accuracy": 0.13732296079397202,
      "num_tokens": 29845861.0,
      "step": 8930
    },
    {
      "entropy": 6.409773254394532,
      "epoch": 1.7431470100478004,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004728830034932437,
      "loss": 6.3223,
      "mean_token_accuracy": 0.13217500671744348,
      "num_tokens": 29861711.0,
      "step": 8935
    },
    {
      "entropy": 6.427531290054321,
      "epoch": 1.7441225246317433,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047284951927779704,
      "loss": 6.3021,
      "mean_token_accuracy": 0.12834896072745322,
      "num_tokens": 29878679.0,
      "step": 8940
    },
    {
      "entropy": 6.427761888504028,
      "epoch": 1.7450980392156863,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047281601572920053,
      "loss": 6.2766,
      "mean_token_accuracy": 0.13590316027402877,
      "num_tokens": 29895016.0,
      "step": 8945
    },
    {
      "entropy": 6.22990050315857,
      "epoch": 1.7460735537996293,
      "grad_norm": 1.015625,
      "learning_rate": 0.000472782492850728,
      "loss": 6.126,
      "mean_token_accuracy": 0.14912073835730552,
      "num_tokens": 29913115.0,
      "step": 8950
    },
    {
      "entropy": 6.302598285675049,
      "epoch": 1.7470490683835722,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004727489506456551,
      "loss": 6.1463,
      "mean_token_accuracy": 0.15219483524560928,
      "num_tokens": 29928954.0,
      "step": 8955
    },
    {
      "entropy": 6.356267499923706,
      "epoch": 1.7480245829675154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047271538911725964,
      "loss": 6.2869,
      "mean_token_accuracy": 0.138744006305933,
      "num_tokens": 29945077.0,
      "step": 8960
    },
    {
      "entropy": 6.4170819282531735,
      "epoch": 1.7490000975514584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000472681808268821,
      "loss": 6.3603,
      "mean_token_accuracy": 0.12218251302838326,
      "num_tokens": 29962480.0,
      "step": 8965
    },
    {
      "entropy": 6.390072202682495,
      "epoch": 1.7499756121354015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047264820810362054,
      "loss": 6.1849,
      "mean_token_accuracy": 0.1447296790778637,
      "num_tokens": 29978761.0,
      "step": 8970
    },
    {
      "entropy": 6.426595020294189,
      "epoch": 1.7509511267193445,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047261458862494165,
      "loss": 6.33,
      "mean_token_accuracy": 0.12977831065654755,
      "num_tokens": 29995836.0,
      "step": 8975
    },
    {
      "entropy": 6.180754613876343,
      "epoch": 1.7519266413032875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004725809498360694,
      "loss": 5.9564,
      "mean_token_accuracy": 0.16993593126535417,
      "num_tokens": 30014221.0,
      "step": 8980
    },
    {
      "entropy": 6.409104251861573,
      "epoch": 1.7529021558872304,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000472547291740291,
      "loss": 6.3769,
      "mean_token_accuracy": 0.13097261041402816,
      "num_tokens": 30030952.0,
      "step": 8985
    },
    {
      "entropy": 6.251207113265991,
      "epoch": 1.7538776704711734,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047251361434089536,
      "loss": 6.0588,
      "mean_token_accuracy": 0.15414456352591516,
      "num_tokens": 30047916.0,
      "step": 8990
    },
    {
      "entropy": 6.3576305389404295,
      "epoch": 1.7548531850551166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004724799176411732,
      "loss": 6.3109,
      "mean_token_accuracy": 0.13206457272171973,
      "num_tokens": 30063238.0,
      "step": 8995
    },
    {
      "entropy": 6.4482990264892575,
      "epoch": 1.7558286996390597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004724462016444174,
      "loss": 6.2546,
      "mean_token_accuracy": 0.13704488873481752,
      "num_tokens": 30079079.0,
      "step": 9000
    },
    {
      "epoch": 1.7558286996390597,
      "eval_entropy": 6.356214998312406,
      "eval_loss": 6.338339805603027,
      "eval_mean_token_accuracy": 0.13798061654230448,
      "eval_num_tokens": 30079079.0,
      "eval_runtime": 15.3126,
      "eval_samples_per_second": 1082.57,
      "eval_steps_per_second": 135.378,
      "step": 9000
    },
    {
      "entropy": 6.396192121505737,
      "epoch": 1.7568042142230027,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047241246635392246,
      "loss": 6.2142,
      "mean_token_accuracy": 0.13900977447628976,
      "num_tokens": 30095165.0,
      "step": 9005
    },
    {
      "entropy": 6.4126386642456055,
      "epoch": 1.7577797288069457,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723787117729848,
      "loss": 6.2891,
      "mean_token_accuracy": 0.13395532295107843,
      "num_tokens": 30111415.0,
      "step": 9010
    },
    {
      "entropy": 6.395002555847168,
      "epoch": 1.7587552433908886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047234493790490297,
      "loss": 6.2758,
      "mean_token_accuracy": 0.14004846289753914,
      "num_tokens": 30127363.0,
      "step": 9015
    },
    {
      "entropy": 6.394078636169434,
      "epoch": 1.7597307579748316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004723111447529771,
      "loss": 6.2421,
      "mean_token_accuracy": 0.13630868196487428,
      "num_tokens": 30144139.0,
      "step": 9020
    },
    {
      "entropy": 6.2268413543701175,
      "epoch": 1.7607062725587748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004722773323205095,
      "loss": 6.066,
      "mean_token_accuracy": 0.15350706949830056,
      "num_tokens": 30160522.0,
      "step": 9025
    },
    {
      "entropy": 6.342193698883056,
      "epoch": 1.761681787142718,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000472243500610804,
      "loss": 6.2436,
      "mean_token_accuracy": 0.14084294214844703,
      "num_tokens": 30178140.0,
      "step": 9030
    },
    {
      "entropy": 6.46108226776123,
      "epoch": 1.762657301726661,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047220964962716675,
      "loss": 6.3267,
      "mean_token_accuracy": 0.1318124406039715,
      "num_tokens": 30194867.0,
      "step": 9035
    },
    {
      "entropy": 6.392329406738281,
      "epoch": 1.7636328163106039,
      "grad_norm": 1.0,
      "learning_rate": 0.00047217577937290527,
      "loss": 6.26,
      "mean_token_accuracy": 0.13636140525341034,
      "num_tokens": 30211966.0,
      "step": 9040
    },
    {
      "entropy": 6.303209829330444,
      "epoch": 1.7646083308945468,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004721418898513294,
      "loss": 6.1196,
      "mean_token_accuracy": 0.1441027581691742,
      "num_tokens": 30229230.0,
      "step": 9045
    },
    {
      "entropy": 6.355041599273681,
      "epoch": 1.7655838454784898,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004721079810657507,
      "loss": 6.1854,
      "mean_token_accuracy": 0.14338366836309432,
      "num_tokens": 30244793.0,
      "step": 9050
    },
    {
      "entropy": 6.364578914642334,
      "epoch": 1.766559360062433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047207405301948264,
      "loss": 6.3079,
      "mean_token_accuracy": 0.13314854800701142,
      "num_tokens": 30260574.0,
      "step": 9055
    },
    {
      "entropy": 6.405673789978027,
      "epoch": 1.767534874646376,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047204010571584045,
      "loss": 6.2006,
      "mean_token_accuracy": 0.13555394485592842,
      "num_tokens": 30277080.0,
      "step": 9060
    },
    {
      "entropy": 6.217748308181763,
      "epoch": 1.7685103892303191,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004720061391581414,
      "loss": 6.0509,
      "mean_token_accuracy": 0.15661883354187012,
      "num_tokens": 30293282.0,
      "step": 9065
    },
    {
      "entropy": 6.341807699203491,
      "epoch": 1.769485903814262,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047197215334970454,
      "loss": 6.2013,
      "mean_token_accuracy": 0.13975218757987024,
      "num_tokens": 30310268.0,
      "step": 9070
    },
    {
      "entropy": 6.348232555389404,
      "epoch": 1.770461418398205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004719381482938509,
      "loss": 6.2348,
      "mean_token_accuracy": 0.1384379617869854,
      "num_tokens": 30327173.0,
      "step": 9075
    },
    {
      "entropy": 6.378496694564819,
      "epoch": 1.771436932982148,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047190412399390335,
      "loss": 6.2838,
      "mean_token_accuracy": 0.14211479425430298,
      "num_tokens": 30343798.0,
      "step": 9080
    },
    {
      "entropy": 6.321821880340576,
      "epoch": 1.772412447566091,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047187008045318653,
      "loss": 6.1587,
      "mean_token_accuracy": 0.14025045111775397,
      "num_tokens": 30359710.0,
      "step": 9085
    },
    {
      "entropy": 6.427680969238281,
      "epoch": 1.7733879621500341,
      "grad_norm": 1.0,
      "learning_rate": 0.0004718360176750271,
      "loss": 6.3067,
      "mean_token_accuracy": 0.1355241619050503,
      "num_tokens": 30376017.0,
      "step": 9090
    },
    {
      "entropy": 6.383937215805053,
      "epoch": 1.7743634767339773,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004718019356627536,
      "loss": 6.1636,
      "mean_token_accuracy": 0.14163192808628083,
      "num_tokens": 30392805.0,
      "step": 9095
    },
    {
      "entropy": 6.399555015563965,
      "epoch": 1.7753389913179203,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004717678344196963,
      "loss": 6.2704,
      "mean_token_accuracy": 0.1337019719183445,
      "num_tokens": 30409165.0,
      "step": 9100
    },
    {
      "entropy": 6.42686095237732,
      "epoch": 1.7763145059018632,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047173371394918764,
      "loss": 6.3011,
      "mean_token_accuracy": 0.13650632947683333,
      "num_tokens": 30425186.0,
      "step": 9105
    },
    {
      "entropy": 6.35317587852478,
      "epoch": 1.7772900204858062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004716995742545615,
      "loss": 6.2669,
      "mean_token_accuracy": 0.1402476504445076,
      "num_tokens": 30442396.0,
      "step": 9110
    },
    {
      "entropy": 6.508467817306519,
      "epoch": 1.7782655350697492,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004716654153391541,
      "loss": 6.4214,
      "mean_token_accuracy": 0.13137849420309067,
      "num_tokens": 30459166.0,
      "step": 9115
    },
    {
      "entropy": 6.31756443977356,
      "epoch": 1.7792410496536923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004716312372063032,
      "loss": 6.1155,
      "mean_token_accuracy": 0.14320407658815384,
      "num_tokens": 30475226.0,
      "step": 9120
    },
    {
      "entropy": 6.433044672012329,
      "epoch": 1.7802165642376353,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004715970398593486,
      "loss": 6.2537,
      "mean_token_accuracy": 0.12901073470711708,
      "num_tokens": 30491649.0,
      "step": 9125
    },
    {
      "entropy": 6.3709314346313475,
      "epoch": 1.7811920788215785,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004715628233016319,
      "loss": 6.2419,
      "mean_token_accuracy": 0.13682734593749046,
      "num_tokens": 30507499.0,
      "step": 9130
    },
    {
      "entropy": 6.318997383117676,
      "epoch": 1.7821675934055214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047152858753649675,
      "loss": 6.214,
      "mean_token_accuracy": 0.13837372213602067,
      "num_tokens": 30524962.0,
      "step": 9135
    },
    {
      "entropy": 6.338379859924316,
      "epoch": 1.7831431079894644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004714943325672884,
      "loss": 6.2046,
      "mean_token_accuracy": 0.14009304493665695,
      "num_tokens": 30540633.0,
      "step": 9140
    },
    {
      "entropy": 6.465652322769165,
      "epoch": 1.7841186225734074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004714600583973542,
      "loss": 6.2912,
      "mean_token_accuracy": 0.1355964496731758,
      "num_tokens": 30558018.0,
      "step": 9145
    },
    {
      "entropy": 6.32857232093811,
      "epoch": 1.7850941371573505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047142576503004336,
      "loss": 6.2294,
      "mean_token_accuracy": 0.14201530367136,
      "num_tokens": 30574870.0,
      "step": 9150
    },
    {
      "entropy": 6.52151894569397,
      "epoch": 1.7860696517412935,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047139145246870686,
      "loss": 6.3362,
      "mean_token_accuracy": 0.12537020146846772,
      "num_tokens": 30590602.0,
      "step": 9155
    },
    {
      "entropy": 6.334890365600586,
      "epoch": 1.7870451663252367,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047135712071669743,
      "loss": 6.1783,
      "mean_token_accuracy": 0.1409427598118782,
      "num_tokens": 30607616.0,
      "step": 9160
    },
    {
      "entropy": 6.457598352432251,
      "epoch": 1.7880206809091796,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047132276977737014,
      "loss": 6.2902,
      "mean_token_accuracy": 0.13430241942405702,
      "num_tokens": 30624288.0,
      "step": 9165
    },
    {
      "entropy": 6.268063354492187,
      "epoch": 1.7889961954931226,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047128839965408136,
      "loss": 6.1179,
      "mean_token_accuracy": 0.15252004340291023,
      "num_tokens": 30640556.0,
      "step": 9170
    },
    {
      "entropy": 6.318844890594482,
      "epoch": 1.7899717100770656,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047125401035018983,
      "loss": 6.2037,
      "mean_token_accuracy": 0.14125955924391748,
      "num_tokens": 30658377.0,
      "step": 9175
    },
    {
      "entropy": 6.411049509048462,
      "epoch": 1.7909472246610085,
      "grad_norm": 1.0,
      "learning_rate": 0.0004712196018690559,
      "loss": 6.3077,
      "mean_token_accuracy": 0.13227883949875832,
      "num_tokens": 30674424.0,
      "step": 9180
    },
    {
      "entropy": 6.458798599243164,
      "epoch": 1.7919227392449517,
      "grad_norm": 1.0,
      "learning_rate": 0.00047118517421404173,
      "loss": 6.3071,
      "mean_token_accuracy": 0.1315547913312912,
      "num_tokens": 30690945.0,
      "step": 9185
    },
    {
      "entropy": 6.497003269195557,
      "epoch": 1.792898253828895,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047115072738851155,
      "loss": 6.2832,
      "mean_token_accuracy": 0.1416326530277729,
      "num_tokens": 30707024.0,
      "step": 9190
    },
    {
      "entropy": 6.439689111709595,
      "epoch": 1.7938737684128379,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047111626139583146,
      "loss": 6.392,
      "mean_token_accuracy": 0.125529957562685,
      "num_tokens": 30723888.0,
      "step": 9195
    },
    {
      "entropy": 6.44202880859375,
      "epoch": 1.7948492829967808,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00047108177623936926,
      "loss": 6.2471,
      "mean_token_accuracy": 0.14019905552268028,
      "num_tokens": 30740721.0,
      "step": 9200
    },
    {
      "entropy": 6.371349859237671,
      "epoch": 1.7958247975807238,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004710472719224947,
      "loss": 6.2163,
      "mean_token_accuracy": 0.14377099499106408,
      "num_tokens": 30758378.0,
      "step": 9205
    },
    {
      "entropy": 6.467862987518311,
      "epoch": 1.7968003121646667,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047101274844857946,
      "loss": 6.2612,
      "mean_token_accuracy": 0.13596218526363374,
      "num_tokens": 30774380.0,
      "step": 9210
    },
    {
      "entropy": 6.403821611404419,
      "epoch": 1.79777582674861,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004709782058209971,
      "loss": 6.2951,
      "mean_token_accuracy": 0.13282383233308792,
      "num_tokens": 30792051.0,
      "step": 9215
    },
    {
      "entropy": 6.378428745269775,
      "epoch": 1.7987513413325529,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004709436440431229,
      "loss": 6.2327,
      "mean_token_accuracy": 0.1404338076710701,
      "num_tokens": 30808892.0,
      "step": 9220
    },
    {
      "entropy": 6.375603532791137,
      "epoch": 1.799726855916496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047090906311833417,
      "loss": 6.2579,
      "mean_token_accuracy": 0.13540627509355546,
      "num_tokens": 30825248.0,
      "step": 9225
    },
    {
      "entropy": 6.3536725521087645,
      "epoch": 1.800702370500439,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047087446305001,
      "loss": 6.2274,
      "mean_token_accuracy": 0.14427687898278235,
      "num_tokens": 30842388.0,
      "step": 9230
    },
    {
      "entropy": 6.315339660644531,
      "epoch": 1.801677885084382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004708398438415314,
      "loss": 6.1904,
      "mean_token_accuracy": 0.13820627257227897,
      "num_tokens": 30859240.0,
      "step": 9235
    },
    {
      "entropy": 6.435899829864502,
      "epoch": 1.802653399668325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004708052054962813,
      "loss": 6.313,
      "mean_token_accuracy": 0.13297298327088355,
      "num_tokens": 30875085.0,
      "step": 9240
    },
    {
      "entropy": 6.490335369110108,
      "epoch": 1.8036289142522681,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047077054801764436,
      "loss": 6.3291,
      "mean_token_accuracy": 0.1347129762172699,
      "num_tokens": 30890913.0,
      "step": 9245
    },
    {
      "entropy": 6.4481894969940186,
      "epoch": 1.804604428836211,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047073587140900714,
      "loss": 6.3288,
      "mean_token_accuracy": 0.1336304672062397,
      "num_tokens": 30906354.0,
      "step": 9250
    },
    {
      "entropy": 6.338286352157593,
      "epoch": 1.8055799434201543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004707011756737583,
      "loss": 6.2406,
      "mean_token_accuracy": 0.13727771639823913,
      "num_tokens": 30923389.0,
      "step": 9255
    },
    {
      "entropy": 6.392793989181518,
      "epoch": 1.8065554580040972,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004706664608152879,
      "loss": 6.2367,
      "mean_token_accuracy": 0.12915365397930145,
      "num_tokens": 30940445.0,
      "step": 9260
    },
    {
      "entropy": 6.382991313934326,
      "epoch": 1.8075309725880402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047063172683698844,
      "loss": 6.2153,
      "mean_token_accuracy": 0.1374957099556923,
      "num_tokens": 30956084.0,
      "step": 9265
    },
    {
      "entropy": 6.395110416412353,
      "epoch": 1.8085064871719831,
      "grad_norm": 1.0,
      "learning_rate": 0.00047059697374225387,
      "loss": 6.327,
      "mean_token_accuracy": 0.13099658265709876,
      "num_tokens": 30973546.0,
      "step": 9270
    },
    {
      "entropy": 6.371123361587524,
      "epoch": 1.809482001755926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004705622015344801,
      "loss": 6.2561,
      "mean_token_accuracy": 0.13769973143935205,
      "num_tokens": 30991713.0,
      "step": 9275
    },
    {
      "entropy": 6.445638275146484,
      "epoch": 1.8104575163398693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000470527410217065,
      "loss": 6.1908,
      "mean_token_accuracy": 0.1394909217953682,
      "num_tokens": 31007682.0,
      "step": 9280
    },
    {
      "entropy": 6.292893028259277,
      "epoch": 1.8114330309238125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004704925997934083,
      "loss": 6.1392,
      "mean_token_accuracy": 0.14404030591249467,
      "num_tokens": 31023238.0,
      "step": 9285
    },
    {
      "entropy": 6.3866489887237545,
      "epoch": 1.8124085455077554,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004704577702669114,
      "loss": 6.2113,
      "mean_token_accuracy": 0.14247310757637024,
      "num_tokens": 31038891.0,
      "step": 9290
    },
    {
      "entropy": 6.391831922531128,
      "epoch": 1.8133840600916984,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047042292164097796,
      "loss": 6.3365,
      "mean_token_accuracy": 0.12980661392211915,
      "num_tokens": 31055718.0,
      "step": 9295
    },
    {
      "entropy": 6.534249019622803,
      "epoch": 1.8143595746756414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000470388053919013,
      "loss": 6.3381,
      "mean_token_accuracy": 0.13200894817709924,
      "num_tokens": 31071574.0,
      "step": 9300
    },
    {
      "entropy": 6.4068928241729735,
      "epoch": 1.8153350892595843,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047035316710442385,
      "loss": 6.3127,
      "mean_token_accuracy": 0.12872020229697229,
      "num_tokens": 31087391.0,
      "step": 9305
    },
    {
      "entropy": 6.29931960105896,
      "epoch": 1.8163106038435275,
      "grad_norm": 1.125,
      "learning_rate": 0.00047031826120061946,
      "loss": 6.2018,
      "mean_token_accuracy": 0.14230246767401694,
      "num_tokens": 31105290.0,
      "step": 9310
    },
    {
      "entropy": 6.441522645950317,
      "epoch": 1.8172861184274705,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047028333621101063,
      "loss": 6.2935,
      "mean_token_accuracy": 0.13652660101652145,
      "num_tokens": 31121229.0,
      "step": 9315
    },
    {
      "entropy": 6.308462524414063,
      "epoch": 1.8182616330114136,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047024839213901027,
      "loss": 6.1207,
      "mean_token_accuracy": 0.14867119193077089,
      "num_tokens": 31139306.0,
      "step": 9320
    },
    {
      "entropy": 6.308221340179443,
      "epoch": 1.8192371475953566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047021342898803296,
      "loss": 6.1558,
      "mean_token_accuracy": 0.14652331918478012,
      "num_tokens": 31155154.0,
      "step": 9325
    },
    {
      "entropy": 6.377928400039673,
      "epoch": 1.8202126621792996,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047017844676149496,
      "loss": 6.3034,
      "mean_token_accuracy": 0.1374237656593323,
      "num_tokens": 31172816.0,
      "step": 9330
    },
    {
      "entropy": 6.506085968017578,
      "epoch": 1.8211881767632425,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047014344546281487,
      "loss": 6.3164,
      "mean_token_accuracy": 0.1299572080373764,
      "num_tokens": 31189152.0,
      "step": 9335
    },
    {
      "entropy": 6.436393785476684,
      "epoch": 1.8221636913471855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047010842509541276,
      "loss": 6.319,
      "mean_token_accuracy": 0.13326427415013314,
      "num_tokens": 31205201.0,
      "step": 9340
    },
    {
      "entropy": 6.394602060317993,
      "epoch": 1.8231392059311287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004700733856627107,
      "loss": 6.2105,
      "mean_token_accuracy": 0.13805087953805922,
      "num_tokens": 31222071.0,
      "step": 9345
    },
    {
      "entropy": 6.402201414108276,
      "epoch": 1.8241147205150718,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004700383271681328,
      "loss": 6.3036,
      "mean_token_accuracy": 0.1320319764316082,
      "num_tokens": 31237717.0,
      "step": 9350
    },
    {
      "entropy": 6.338006162643433,
      "epoch": 1.8250902350990148,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047000324961510455,
      "loss": 6.2801,
      "mean_token_accuracy": 0.13770928755402564,
      "num_tokens": 31255113.0,
      "step": 9355
    },
    {
      "entropy": 6.369744825363159,
      "epoch": 1.8260657496829578,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004699681530070538,
      "loss": 6.2323,
      "mean_token_accuracy": 0.14245341941714287,
      "num_tokens": 31271246.0,
      "step": 9360
    },
    {
      "entropy": 6.517406368255616,
      "epoch": 1.8270412642669007,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046993303734740996,
      "loss": 6.3907,
      "mean_token_accuracy": 0.12413990497589111,
      "num_tokens": 31289079.0,
      "step": 9365
    },
    {
      "entropy": 6.25498104095459,
      "epoch": 1.8280167788508437,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046989790263960454,
      "loss": 6.1024,
      "mean_token_accuracy": 0.15743164122104644,
      "num_tokens": 31305432.0,
      "step": 9370
    },
    {
      "entropy": 6.406301498413086,
      "epoch": 1.8289922934347869,
      "grad_norm": 1.0,
      "learning_rate": 0.00046986274888707066,
      "loss": 6.3139,
      "mean_token_accuracy": 0.13144850060343743,
      "num_tokens": 31321683.0,
      "step": 9375
    },
    {
      "entropy": 6.405181932449341,
      "epoch": 1.8299678080187298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004698275760932435,
      "loss": 6.2661,
      "mean_token_accuracy": 0.13588965833187103,
      "num_tokens": 31337978.0,
      "step": 9380
    },
    {
      "entropy": 6.344358587265015,
      "epoch": 1.830943322602673,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004697923842615599,
      "loss": 6.2201,
      "mean_token_accuracy": 0.14074813947081566,
      "num_tokens": 31354426.0,
      "step": 9385
    },
    {
      "entropy": 6.377481174468994,
      "epoch": 1.831918837186616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004697571733954588,
      "loss": 6.1491,
      "mean_token_accuracy": 0.14308176189661026,
      "num_tokens": 31372295.0,
      "step": 9390
    },
    {
      "entropy": 6.310229778289795,
      "epoch": 1.832894351770559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004697219434983809,
      "loss": 6.1553,
      "mean_token_accuracy": 0.14781548380851744,
      "num_tokens": 31388915.0,
      "step": 9395
    },
    {
      "entropy": 6.414380264282227,
      "epoch": 1.8338698663545019,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046968669457376865,
      "loss": 6.3431,
      "mean_token_accuracy": 0.13076071813702583,
      "num_tokens": 31404934.0,
      "step": 9400
    },
    {
      "entropy": 6.387458944320679,
      "epoch": 1.834845380938445,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046965142662506646,
      "loss": 6.2679,
      "mean_token_accuracy": 0.1353147253394127,
      "num_tokens": 31421481.0,
      "step": 9405
    },
    {
      "entropy": 6.419200038909912,
      "epoch": 1.835820895522388,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004696161396557206,
      "loss": 6.1878,
      "mean_token_accuracy": 0.1447974279522896,
      "num_tokens": 31437155.0,
      "step": 9410
    },
    {
      "entropy": 6.357773923873902,
      "epoch": 1.8367964101063312,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046958083366917925,
      "loss": 6.1672,
      "mean_token_accuracy": 0.13997371196746827,
      "num_tokens": 31453303.0,
      "step": 9415
    },
    {
      "entropy": 6.366457986831665,
      "epoch": 1.8377719246902742,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046954550866889236,
      "loss": 6.2996,
      "mean_token_accuracy": 0.1401326209306717,
      "num_tokens": 31468818.0,
      "step": 9420
    },
    {
      "entropy": 6.389225673675537,
      "epoch": 1.8387474392742171,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695101646583116,
      "loss": 6.2843,
      "mean_token_accuracy": 0.12857346013188362,
      "num_tokens": 31485967.0,
      "step": 9425
    },
    {
      "entropy": 6.44317045211792,
      "epoch": 1.83972295385816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000469474801640891,
      "loss": 6.2044,
      "mean_token_accuracy": 0.1341864600777626,
      "num_tokens": 31501938.0,
      "step": 9430
    },
    {
      "entropy": 6.38956241607666,
      "epoch": 1.840698468442103,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046943941962008574,
      "loss": 6.2386,
      "mean_token_accuracy": 0.13364326059818268,
      "num_tokens": 31517793.0,
      "step": 9435
    },
    {
      "entropy": 6.429997873306275,
      "epoch": 1.8416739830260462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694040185993534,
      "loss": 6.3004,
      "mean_token_accuracy": 0.13642890453338624,
      "num_tokens": 31534186.0,
      "step": 9440
    },
    {
      "entropy": 6.4221148014068605,
      "epoch": 1.8426494976099894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004693685985821532,
      "loss": 6.2637,
      "mean_token_accuracy": 0.13464615270495414,
      "num_tokens": 31551460.0,
      "step": 9445
    },
    {
      "entropy": 6.238397026062012,
      "epoch": 1.8436250121939324,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046933315957194637,
      "loss": 6.0303,
      "mean_token_accuracy": 0.15888315811753273,
      "num_tokens": 31569403.0,
      "step": 9450
    },
    {
      "entropy": 6.366998529434204,
      "epoch": 1.8446005267778753,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004692977015721958,
      "loss": 6.2265,
      "mean_token_accuracy": 0.13784189671278,
      "num_tokens": 31586007.0,
      "step": 9455
    },
    {
      "entropy": 6.497064781188965,
      "epoch": 1.8455760413618183,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004692622245863663,
      "loss": 6.3274,
      "mean_token_accuracy": 0.12788506746292114,
      "num_tokens": 31603421.0,
      "step": 9460
    },
    {
      "entropy": 6.311890649795532,
      "epoch": 1.8465515559457613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046922672861792456,
      "loss": 6.2591,
      "mean_token_accuracy": 0.13616713061928748,
      "num_tokens": 31620173.0,
      "step": 9465
    },
    {
      "entropy": 6.39614462852478,
      "epoch": 1.8475270705297044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046919121367033917,
      "loss": 6.2004,
      "mean_token_accuracy": 0.1403808556497097,
      "num_tokens": 31636820.0,
      "step": 9470
    },
    {
      "entropy": 6.377713298797607,
      "epoch": 1.8485025851136474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004691556797470805,
      "loss": 6.2798,
      "mean_token_accuracy": 0.13367884904146193,
      "num_tokens": 31655099.0,
      "step": 9475
    },
    {
      "entropy": 6.42005181312561,
      "epoch": 1.8494780996975906,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046912012685162074,
      "loss": 6.2832,
      "mean_token_accuracy": 0.13551371693611144,
      "num_tokens": 31671875.0,
      "step": 9480
    },
    {
      "entropy": 6.348361539840698,
      "epoch": 1.8504536142815335,
      "grad_norm": 1.0,
      "learning_rate": 0.0004690845549874341,
      "loss": 6.2136,
      "mean_token_accuracy": 0.13749729320406914,
      "num_tokens": 31688542.0,
      "step": 9485
    },
    {
      "entropy": 6.33844575881958,
      "epoch": 1.8514291288654765,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004690489641579964,
      "loss": 6.1874,
      "mean_token_accuracy": 0.13903053998947143,
      "num_tokens": 31705139.0,
      "step": 9490
    },
    {
      "entropy": 6.452338695526123,
      "epoch": 1.8524046434494195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046901335436678554,
      "loss": 6.353,
      "mean_token_accuracy": 0.1285382829606533,
      "num_tokens": 31723038.0,
      "step": 9495
    },
    {
      "entropy": 6.390421915054321,
      "epoch": 1.8533801580333626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689777256172812,
      "loss": 6.2136,
      "mean_token_accuracy": 0.1382942281663418,
      "num_tokens": 31738920.0,
      "step": 9500
    },
    {
      "entropy": 6.280410575866699,
      "epoch": 1.8543556726173056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046894207791296484,
      "loss": 6.1697,
      "mean_token_accuracy": 0.1416476286947727,
      "num_tokens": 31755093.0,
      "step": 9505
    },
    {
      "entropy": 6.295342445373535,
      "epoch": 1.8553311872012488,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046890641125731987,
      "loss": 6.1535,
      "mean_token_accuracy": 0.14636207073926927,
      "num_tokens": 31775364.0,
      "step": 9510
    },
    {
      "entropy": 6.430967187881469,
      "epoch": 1.8563067017851917,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046887072565383154,
      "loss": 6.2942,
      "mean_token_accuracy": 0.13192014172673225,
      "num_tokens": 31790462.0,
      "step": 9515
    },
    {
      "entropy": 6.457190656661988,
      "epoch": 1.8572822163691347,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004688350211059868,
      "loss": 6.3105,
      "mean_token_accuracy": 0.12919219881296157,
      "num_tokens": 31806813.0,
      "step": 9520
    },
    {
      "entropy": 6.4903340339660645,
      "epoch": 1.8582577309530777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004687992976172746,
      "loss": 6.2999,
      "mean_token_accuracy": 0.13215499892830848,
      "num_tokens": 31823592.0,
      "step": 9525
    },
    {
      "entropy": 6.315087795257568,
      "epoch": 1.8592332455370206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046876355519118586,
      "loss": 6.1251,
      "mean_token_accuracy": 0.14368800520896913,
      "num_tokens": 31840110.0,
      "step": 9530
    },
    {
      "entropy": 6.380385398864746,
      "epoch": 1.8602087601209638,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468727793831213,
      "loss": 6.2003,
      "mean_token_accuracy": 0.14167005196213722,
      "num_tokens": 31858338.0,
      "step": 9535
    },
    {
      "entropy": 6.434043121337891,
      "epoch": 1.861184274704907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046869201354085056,
      "loss": 6.2824,
      "mean_token_accuracy": 0.13134352639317512,
      "num_tokens": 31874793.0,
      "step": 9540
    },
    {
      "entropy": 6.36174464225769,
      "epoch": 1.86215978928885,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046865621432359497,
      "loss": 6.2634,
      "mean_token_accuracy": 0.13526437804102898,
      "num_tokens": 31891339.0,
      "step": 9545
    },
    {
      "entropy": 6.474738359451294,
      "epoch": 1.863135303872793,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046862039618294425,
      "loss": 6.3607,
      "mean_token_accuracy": 0.128442994505167,
      "num_tokens": 31908407.0,
      "step": 9550
    },
    {
      "entropy": 6.42780122756958,
      "epoch": 1.8641108184567359,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004685845591223985,
      "loss": 6.2304,
      "mean_token_accuracy": 0.1355710119009018,
      "num_tokens": 31923653.0,
      "step": 9555
    },
    {
      "entropy": 6.254728555679321,
      "epoch": 1.8650863330406788,
      "grad_norm": 1.0,
      "learning_rate": 0.00046854870314545966,
      "loss": 6.0881,
      "mean_token_accuracy": 0.1496527001261711,
      "num_tokens": 31940531.0,
      "step": 9560
    },
    {
      "entropy": 6.253374242782593,
      "epoch": 1.866061847624622,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004685128282556314,
      "loss": 6.1073,
      "mean_token_accuracy": 0.1473422683775425,
      "num_tokens": 31958730.0,
      "step": 9565
    },
    {
      "entropy": 6.398071813583374,
      "epoch": 1.867037362208565,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004684769344564191,
      "loss": 6.2775,
      "mean_token_accuracy": 0.13654083535075187,
      "num_tokens": 31976200.0,
      "step": 9570
    },
    {
      "entropy": 6.5001633644104,
      "epoch": 1.8680128767925082,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046844102175133033,
      "loss": 6.3764,
      "mean_token_accuracy": 0.12947179600596428,
      "num_tokens": 31992791.0,
      "step": 9575
    },
    {
      "entropy": 6.362196969985962,
      "epoch": 1.8689883913764511,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046840509014387443,
      "loss": 6.1877,
      "mean_token_accuracy": 0.137497179210186,
      "num_tokens": 32008936.0,
      "step": 9580
    },
    {
      "entropy": 6.391362190246582,
      "epoch": 1.869963905960394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004683691396375625,
      "loss": 6.2438,
      "mean_token_accuracy": 0.13349375277757644,
      "num_tokens": 32024651.0,
      "step": 9585
    },
    {
      "entropy": 6.435646629333496,
      "epoch": 1.870939420544337,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004683331702359073,
      "loss": 6.293,
      "mean_token_accuracy": 0.13425374105572702,
      "num_tokens": 32040879.0,
      "step": 9590
    },
    {
      "entropy": 6.404850482940674,
      "epoch": 1.8719149351282802,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046829718194242375,
      "loss": 6.2647,
      "mean_token_accuracy": 0.13727355524897575,
      "num_tokens": 32058624.0,
      "step": 9595
    },
    {
      "entropy": 6.312521696090698,
      "epoch": 1.8728904497122232,
      "grad_norm": 1.0,
      "learning_rate": 0.0004682611747606286,
      "loss": 6.1423,
      "mean_token_accuracy": 0.14502613097429276,
      "num_tokens": 32074594.0,
      "step": 9600
    },
    {
      "entropy": 6.387927150726318,
      "epoch": 1.8738659642961664,
      "grad_norm": 1.0,
      "learning_rate": 0.0004682251486940402,
      "loss": 6.2567,
      "mean_token_accuracy": 0.13194873332977294,
      "num_tokens": 32090431.0,
      "step": 9605
    },
    {
      "entropy": 6.320067262649536,
      "epoch": 1.8748414788801093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046818910374617895,
      "loss": 6.2024,
      "mean_token_accuracy": 0.13491860628128052,
      "num_tokens": 32108353.0,
      "step": 9610
    },
    {
      "entropy": 6.399367380142212,
      "epoch": 1.8758169934640523,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046815303992056703,
      "loss": 6.247,
      "mean_token_accuracy": 0.13238520920276642,
      "num_tokens": 32125747.0,
      "step": 9615
    },
    {
      "entropy": 6.327607774734497,
      "epoch": 1.8767925080479952,
      "grad_norm": 1.0,
      "learning_rate": 0.0004681169572207284,
      "loss": 6.1587,
      "mean_token_accuracy": 0.1378806471824646,
      "num_tokens": 32141578.0,
      "step": 9620
    },
    {
      "entropy": 6.408616733551026,
      "epoch": 1.8777680226319382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046808085565018905,
      "loss": 6.3102,
      "mean_token_accuracy": 0.13556066602468492,
      "num_tokens": 32158939.0,
      "step": 9625
    },
    {
      "entropy": 6.410214853286743,
      "epoch": 1.8787435372158814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004680447352124766,
      "loss": 6.2561,
      "mean_token_accuracy": 0.13738429099321364,
      "num_tokens": 32175355.0,
      "step": 9630
    },
    {
      "entropy": 6.4065690517425535,
      "epoch": 1.8797190517998246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046800859591112064,
      "loss": 6.283,
      "mean_token_accuracy": 0.13723159953951836,
      "num_tokens": 32190827.0,
      "step": 9635
    },
    {
      "entropy": 6.478393459320069,
      "epoch": 1.8806945663837675,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046797243774965265,
      "loss": 6.2997,
      "mean_token_accuracy": 0.13692883253097535,
      "num_tokens": 32207911.0,
      "step": 9640
    },
    {
      "entropy": 6.394640493392944,
      "epoch": 1.8816700809677105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004679362607316057,
      "loss": 6.2419,
      "mean_token_accuracy": 0.14067244306206703,
      "num_tokens": 32225200.0,
      "step": 9645
    },
    {
      "entropy": 6.326661157608032,
      "epoch": 1.8826455955516535,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004679000648605151,
      "loss": 6.2155,
      "mean_token_accuracy": 0.14661810919642448,
      "num_tokens": 32241933.0,
      "step": 9650
    },
    {
      "entropy": 6.398635816574097,
      "epoch": 1.8836211101355964,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678638501399176,
      "loss": 6.2375,
      "mean_token_accuracy": 0.13663963377475738,
      "num_tokens": 32257856.0,
      "step": 9655
    },
    {
      "entropy": 6.402212333679199,
      "epoch": 1.8845966247195396,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046782761657335203,
      "loss": 6.1804,
      "mean_token_accuracy": 0.13570712357759476,
      "num_tokens": 32273973.0,
      "step": 9660
    },
    {
      "entropy": 6.381634759902954,
      "epoch": 1.8855721393034826,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004677913641643591,
      "loss": 6.2881,
      "mean_token_accuracy": 0.13265275359153747,
      "num_tokens": 32289731.0,
      "step": 9665
    },
    {
      "entropy": 6.427105951309204,
      "epoch": 1.8865476538874257,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004677550929164812,
      "loss": 6.2455,
      "mean_token_accuracy": 0.1366285651922226,
      "num_tokens": 32305839.0,
      "step": 9670
    },
    {
      "entropy": 6.351392316818237,
      "epoch": 1.8875231684713687,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046771880283326254,
      "loss": 6.1965,
      "mean_token_accuracy": 0.14740199521183966,
      "num_tokens": 32322393.0,
      "step": 9675
    },
    {
      "entropy": 6.283886384963989,
      "epoch": 1.8884986830553117,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004676824939182493,
      "loss": 6.1225,
      "mean_token_accuracy": 0.1389647200703621,
      "num_tokens": 32340535.0,
      "step": 9680
    },
    {
      "entropy": 6.411776304244995,
      "epoch": 1.8894741976392546,
      "grad_norm": 1.0,
      "learning_rate": 0.00046764616617498957,
      "loss": 6.2684,
      "mean_token_accuracy": 0.1376809783279896,
      "num_tokens": 32357104.0,
      "step": 9685
    },
    {
      "entropy": 6.330094528198242,
      "epoch": 1.8904497122231976,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467609819607033,
      "loss": 6.2566,
      "mean_token_accuracy": 0.13713766038417816,
      "num_tokens": 32373969.0,
      "step": 9690
    },
    {
      "entropy": 6.433128118515015,
      "epoch": 1.8914252268071408,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004675734542179313,
      "loss": 6.2695,
      "mean_token_accuracy": 0.13298409432172775,
      "num_tokens": 32391021.0,
      "step": 9695
    },
    {
      "entropy": 6.283124160766602,
      "epoch": 1.892400741391084,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004675370700112382,
      "loss": 6.1751,
      "mean_token_accuracy": 0.1394822731614113,
      "num_tokens": 32407415.0,
      "step": 9700
    },
    {
      "entropy": 6.4129070281982425,
      "epoch": 1.893376255975027,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046750066699050865,
      "loss": 6.1766,
      "mean_token_accuracy": 0.14455540105700493,
      "num_tokens": 32424525.0,
      "step": 9705
    },
    {
      "entropy": 6.411465263366699,
      "epoch": 1.8943517705589699,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046746424515930004,
      "loss": 6.2711,
      "mean_token_accuracy": 0.13164868503808974,
      "num_tokens": 32440975.0,
      "step": 9710
    },
    {
      "entropy": 6.1548644542694095,
      "epoch": 1.8953272851429128,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004674278045211714,
      "loss": 5.9686,
      "mean_token_accuracy": 0.16147546246647834,
      "num_tokens": 32458021.0,
      "step": 9715
    },
    {
      "entropy": 6.359719085693359,
      "epoch": 1.8963027997268558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046739134507968347,
      "loss": 6.2834,
      "mean_token_accuracy": 0.13299766108393668,
      "num_tokens": 32475354.0,
      "step": 9720
    },
    {
      "entropy": 6.3961522579193115,
      "epoch": 1.897278314310799,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046735486683839903,
      "loss": 6.2569,
      "mean_token_accuracy": 0.13953315168619157,
      "num_tokens": 32491686.0,
      "step": 9725
    },
    {
      "entropy": 6.424519824981689,
      "epoch": 1.898253828894742,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046731836980088256,
      "loss": 6.2214,
      "mean_token_accuracy": 0.14066344425082206,
      "num_tokens": 32508785.0,
      "step": 9730
    },
    {
      "entropy": 6.290301752090454,
      "epoch": 1.899229343478685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046728185397070047,
      "loss": 6.1288,
      "mean_token_accuracy": 0.14431832805275918,
      "num_tokens": 32525558.0,
      "step": 9735
    },
    {
      "entropy": 6.408667755126953,
      "epoch": 1.900204858062628,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046724531935142077,
      "loss": 6.3096,
      "mean_token_accuracy": 0.1289863795042038,
      "num_tokens": 32541743.0,
      "step": 9740
    },
    {
      "entropy": 6.407957267761231,
      "epoch": 1.901180372646571,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046720876594661374,
      "loss": 6.2409,
      "mean_token_accuracy": 0.13563432022929192,
      "num_tokens": 32558492.0,
      "step": 9745
    },
    {
      "entropy": 6.417195415496826,
      "epoch": 1.902155887230514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004671721937598512,
      "loss": 6.1872,
      "mean_token_accuracy": 0.1437838762998581,
      "num_tokens": 32574782.0,
      "step": 9750
    },
    {
      "entropy": 6.263223695755005,
      "epoch": 1.9031314018144572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004671356027947067,
      "loss": 6.2469,
      "mean_token_accuracy": 0.13862508833408355,
      "num_tokens": 32591630.0,
      "step": 9755
    },
    {
      "entropy": 6.272764062881469,
      "epoch": 1.9041069163984001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004670989930547559,
      "loss": 6.1517,
      "mean_token_accuracy": 0.142232496291399,
      "num_tokens": 32607388.0,
      "step": 9760
    },
    {
      "entropy": 6.484280729293824,
      "epoch": 1.9050824309823433,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004670623645435762,
      "loss": 6.2967,
      "mean_token_accuracy": 0.13402714729309081,
      "num_tokens": 32624598.0,
      "step": 9765
    },
    {
      "entropy": 6.369623756408691,
      "epoch": 1.9060579455662863,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004670257172647467,
      "loss": 6.2628,
      "mean_token_accuracy": 0.13176015093922616,
      "num_tokens": 32641646.0,
      "step": 9770
    },
    {
      "entropy": 6.417615556716919,
      "epoch": 1.9070334601502292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046698905122184855,
      "loss": 6.3104,
      "mean_token_accuracy": 0.12836852222681044,
      "num_tokens": 32659760.0,
      "step": 9775
    },
    {
      "entropy": 6.282630777359008,
      "epoch": 1.9080089747341722,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004669523664184646,
      "loss": 6.1226,
      "mean_token_accuracy": 0.15455398410558702,
      "num_tokens": 32675820.0,
      "step": 9780
    },
    {
      "entropy": 6.3346906185150145,
      "epoch": 1.9089844893181152,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004669156628581794,
      "loss": 6.1942,
      "mean_token_accuracy": 0.14405579790472983,
      "num_tokens": 32694675.0,
      "step": 9785
    },
    {
      "entropy": 6.438905096054077,
      "epoch": 1.9099600039020583,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004668789405445797,
      "loss": 6.2737,
      "mean_token_accuracy": 0.13169640228152274,
      "num_tokens": 32710265.0,
      "step": 9790
    },
    {
      "entropy": 6.40919680595398,
      "epoch": 1.9109355184860015,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046684219948125397,
      "loss": 6.2464,
      "mean_token_accuracy": 0.1377132274210453,
      "num_tokens": 32726709.0,
      "step": 9795
    },
    {
      "entropy": 6.358468532562256,
      "epoch": 1.9119110330699445,
      "grad_norm": 1.03125,
      "learning_rate": 0.000466805439671792,
      "loss": 6.2309,
      "mean_token_accuracy": 0.14157585576176643,
      "num_tokens": 32742043.0,
      "step": 9800
    },
    {
      "entropy": 6.299545669555664,
      "epoch": 1.9128865476538874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046676866111978614,
      "loss": 6.1612,
      "mean_token_accuracy": 0.14495604187250138,
      "num_tokens": 32758932.0,
      "step": 9805
    },
    {
      "entropy": 6.4308281421661375,
      "epoch": 1.9138620622378304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046673186382883026,
      "loss": 6.2337,
      "mean_token_accuracy": 0.13529407531023024,
      "num_tokens": 32775012.0,
      "step": 9810
    },
    {
      "entropy": 6.317173671722412,
      "epoch": 1.9148375768217734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046669504780251985,
      "loss": 6.2148,
      "mean_token_accuracy": 0.13794417008757592,
      "num_tokens": 32790312.0,
      "step": 9815
    },
    {
      "entropy": 6.330607557296753,
      "epoch": 1.9158130914057165,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046665821304445274,
      "loss": 6.2668,
      "mean_token_accuracy": 0.1403467111289501,
      "num_tokens": 32807926.0,
      "step": 9820
    },
    {
      "entropy": 6.448745346069336,
      "epoch": 1.9167886059896595,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046662135955822804,
      "loss": 6.2486,
      "mean_token_accuracy": 0.1364908441901207,
      "num_tokens": 32825129.0,
      "step": 9825
    },
    {
      "entropy": 6.353630638122558,
      "epoch": 1.9177641205736027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046658448734744715,
      "loss": 6.2171,
      "mean_token_accuracy": 0.13952165320515633,
      "num_tokens": 32841688.0,
      "step": 9830
    },
    {
      "entropy": 6.379477262496948,
      "epoch": 1.9187396351575456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004665475964157128,
      "loss": 6.2523,
      "mean_token_accuracy": 0.14466929882764817,
      "num_tokens": 32860004.0,
      "step": 9835
    },
    {
      "entropy": 6.375952291488647,
      "epoch": 1.9197151497414886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004665106867666301,
      "loss": 6.1703,
      "mean_token_accuracy": 0.14359717071056366,
      "num_tokens": 32877633.0,
      "step": 9840
    },
    {
      "entropy": 6.385291481018067,
      "epoch": 1.9206906643254316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004664737584038057,
      "loss": 6.2137,
      "mean_token_accuracy": 0.13904341533780099,
      "num_tokens": 32892826.0,
      "step": 9845
    },
    {
      "entropy": 6.31505012512207,
      "epoch": 1.9216661789093747,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046643681133084797,
      "loss": 6.1909,
      "mean_token_accuracy": 0.14110593721270562,
      "num_tokens": 32909832.0,
      "step": 9850
    },
    {
      "entropy": 6.351682138442993,
      "epoch": 1.9226416934933177,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004663998455513673,
      "loss": 6.1983,
      "mean_token_accuracy": 0.13740734681487082,
      "num_tokens": 32926687.0,
      "step": 9855
    },
    {
      "entropy": 6.281480884552002,
      "epoch": 1.9236172080772609,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046636286106897595,
      "loss": 6.1075,
      "mean_token_accuracy": 0.15055232346057892,
      "num_tokens": 32944155.0,
      "step": 9860
    },
    {
      "entropy": 6.3911130905151365,
      "epoch": 1.9245927226612038,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004663258578872879,
      "loss": 6.2283,
      "mean_token_accuracy": 0.13994019329547883,
      "num_tokens": 32961013.0,
      "step": 9865
    },
    {
      "entropy": 6.418384790420532,
      "epoch": 1.9255682372451468,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004662888360099189,
      "loss": 6.309,
      "mean_token_accuracy": 0.13048052936792373,
      "num_tokens": 32977358.0,
      "step": 9870
    },
    {
      "entropy": 6.3143316268920895,
      "epoch": 1.9265437518290898,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046625179544048653,
      "loss": 6.1153,
      "mean_token_accuracy": 0.14734323918819428,
      "num_tokens": 32994668.0,
      "step": 9875
    },
    {
      "entropy": 6.460422563552856,
      "epoch": 1.9275192664130327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004662147361826103,
      "loss": 6.3218,
      "mean_token_accuracy": 0.1293775372207165,
      "num_tokens": 33009949.0,
      "step": 9880
    },
    {
      "entropy": 6.248913240432739,
      "epoch": 1.928494780996976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004661776582399117,
      "loss": 6.056,
      "mean_token_accuracy": 0.14552169740200044,
      "num_tokens": 33026122.0,
      "step": 9885
    },
    {
      "entropy": 6.417372703552246,
      "epoch": 1.929470295580919,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046614056161601366,
      "loss": 6.2586,
      "mean_token_accuracy": 0.1397307276725769,
      "num_tokens": 33043406.0,
      "step": 9890
    },
    {
      "entropy": 6.308789587020874,
      "epoch": 1.930445810164862,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046610344631454114,
      "loss": 6.1477,
      "mean_token_accuracy": 0.14535077959299086,
      "num_tokens": 33058788.0,
      "step": 9895
    },
    {
      "entropy": 6.36555118560791,
      "epoch": 1.931421324748805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000466066312339121,
      "loss": 6.1974,
      "mean_token_accuracy": 0.13645484745502473,
      "num_tokens": 33076111.0,
      "step": 9900
    },
    {
      "entropy": 6.385289573669434,
      "epoch": 1.932396839332748,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046602915969338175,
      "loss": 6.2183,
      "mean_token_accuracy": 0.13597076013684273,
      "num_tokens": 33093531.0,
      "step": 9905
    },
    {
      "entropy": 6.3503624439239506,
      "epoch": 1.933372353916691,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004659919883809538,
      "loss": 6.1227,
      "mean_token_accuracy": 0.14721708819270135,
      "num_tokens": 33110621.0,
      "step": 9910
    },
    {
      "entropy": 6.190105199813843,
      "epoch": 1.9343478685006341,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004659547984054695,
      "loss": 6.0224,
      "mean_token_accuracy": 0.15250845849514008,
      "num_tokens": 33126933.0,
      "step": 9915
    },
    {
      "entropy": 6.365618562698364,
      "epoch": 1.935323383084577,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046591758977056294,
      "loss": 6.2885,
      "mean_token_accuracy": 0.13069705590605735,
      "num_tokens": 33143947.0,
      "step": 9920
    },
    {
      "entropy": 6.466252326965332,
      "epoch": 1.9362988976685203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004658803624798699,
      "loss": 6.2927,
      "mean_token_accuracy": 0.13383173644542695,
      "num_tokens": 33160375.0,
      "step": 9925
    },
    {
      "entropy": 6.413086986541748,
      "epoch": 1.9372744122524632,
      "grad_norm": 1.125,
      "learning_rate": 0.00046584311653702815,
      "loss": 6.2869,
      "mean_token_accuracy": 0.13437009155750274,
      "num_tokens": 33177649.0,
      "step": 9930
    },
    {
      "entropy": 6.422484445571899,
      "epoch": 1.9382499268364062,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004658058519456772,
      "loss": 6.2938,
      "mean_token_accuracy": 0.13376794308423995,
      "num_tokens": 33194938.0,
      "step": 9935
    },
    {
      "entropy": 6.364679956436158,
      "epoch": 1.9392254414203491,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657685687094585,
      "loss": 6.179,
      "mean_token_accuracy": 0.13833021596074105,
      "num_tokens": 33210049.0,
      "step": 9940
    },
    {
      "entropy": 6.415758419036865,
      "epoch": 1.940200956004292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046573126683201513,
      "loss": 6.2493,
      "mean_token_accuracy": 0.13633688911795616,
      "num_tokens": 33227589.0,
      "step": 9945
    },
    {
      "entropy": 6.402521324157715,
      "epoch": 1.9411764705882353,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046569394631699215,
      "loss": 6.2033,
      "mean_token_accuracy": 0.13320761919021606,
      "num_tokens": 33245812.0,
      "step": 9950
    },
    {
      "entropy": 6.369533920288086,
      "epoch": 1.9421519851721785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004656566071680364,
      "loss": 6.2485,
      "mean_token_accuracy": 0.13407781571149827,
      "num_tokens": 33261792.0,
      "step": 9955
    },
    {
      "entropy": 6.330699825286866,
      "epoch": 1.9431274997561214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004656192493887965,
      "loss": 6.2248,
      "mean_token_accuracy": 0.1443083442747593,
      "num_tokens": 33279083.0,
      "step": 9960
    },
    {
      "entropy": 6.222974061965942,
      "epoch": 1.9441030143400644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004655818729829229,
      "loss": 6.0157,
      "mean_token_accuracy": 0.1511776216328144,
      "num_tokens": 33296391.0,
      "step": 9965
    },
    {
      "entropy": 6.344937562942505,
      "epoch": 1.9450785289240073,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004655444779540679,
      "loss": 6.1284,
      "mean_token_accuracy": 0.1412220150232315,
      "num_tokens": 33313168.0,
      "step": 9970
    },
    {
      "entropy": 6.359477519989014,
      "epoch": 1.9460540435079503,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046550706430588574,
      "loss": 6.2679,
      "mean_token_accuracy": 0.13444903045892714,
      "num_tokens": 33330156.0,
      "step": 9975
    },
    {
      "entropy": 6.318238401412964,
      "epoch": 1.9470295580918935,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046546963204203214,
      "loss": 6.1486,
      "mean_token_accuracy": 0.14393093138933183,
      "num_tokens": 33347496.0,
      "step": 9980
    },
    {
      "entropy": 6.3290793895721436,
      "epoch": 1.9480050726758367,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004654321811661649,
      "loss": 6.2105,
      "mean_token_accuracy": 0.13718379959464072,
      "num_tokens": 33364521.0,
      "step": 9985
    },
    {
      "entropy": 6.346898174285888,
      "epoch": 1.9489805872597796,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004653947116819438,
      "loss": 6.2015,
      "mean_token_accuracy": 0.14146358519792557,
      "num_tokens": 33381902.0,
      "step": 9990
    },
    {
      "entropy": 6.391664600372314,
      "epoch": 1.9499561018437226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653572235930299,
      "loss": 6.2941,
      "mean_token_accuracy": 0.13201361522078514,
      "num_tokens": 33397654.0,
      "step": 9995
    },
    {
      "entropy": 6.416857957839966,
      "epoch": 1.9509316164276655,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004653197169030866,
      "loss": 6.263,
      "mean_token_accuracy": 0.1347549192607403,
      "num_tokens": 33415440.0,
      "step": 10000
    },
    {
      "entropy": 6.315218257904053,
      "epoch": 1.9519071310116085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046528219161577893,
      "loss": 6.1124,
      "mean_token_accuracy": 0.143174596875906,
      "num_tokens": 33431488.0,
      "step": 10005
    },
    {
      "entropy": 6.297128391265869,
      "epoch": 1.9528826455955517,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046524464773477355,
      "loss": 6.1052,
      "mean_token_accuracy": 0.14074433892965316,
      "num_tokens": 33447750.0,
      "step": 10010
    },
    {
      "entropy": 6.282841205596924,
      "epoch": 1.9538581601794947,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00046520708526373935,
      "loss": 6.1572,
      "mean_token_accuracy": 0.14235602393746377,
      "num_tokens": 33464420.0,
      "step": 10015
    },
    {
      "entropy": 6.291812133789063,
      "epoch": 1.9548336747634378,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004651695042063467,
      "loss": 6.0753,
      "mean_token_accuracy": 0.14952076449990273,
      "num_tokens": 33482114.0,
      "step": 10020
    },
    {
      "entropy": 6.463514280319214,
      "epoch": 1.9558091893473808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004651319045662678,
      "loss": 6.2677,
      "mean_token_accuracy": 0.13815312087535858,
      "num_tokens": 33500239.0,
      "step": 10025
    },
    {
      "entropy": 6.397865343093872,
      "epoch": 1.9567847039313238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004650942863471768,
      "loss": 6.2753,
      "mean_token_accuracy": 0.13365463837981223,
      "num_tokens": 33516832.0,
      "step": 10030
    },
    {
      "entropy": 6.336740398406983,
      "epoch": 1.9577602185152667,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046505664955274974,
      "loss": 6.2237,
      "mean_token_accuracy": 0.14432193487882614,
      "num_tokens": 33532899.0,
      "step": 10035
    },
    {
      "entropy": 6.44476432800293,
      "epoch": 1.9587357330992097,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004650189941866643,
      "loss": 6.2689,
      "mean_token_accuracy": 0.13195819333195685,
      "num_tokens": 33548704.0,
      "step": 10040
    },
    {
      "entropy": 6.325990056991577,
      "epoch": 1.9597112476831529,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004649813202525999,
      "loss": 6.1781,
      "mean_token_accuracy": 0.14325996860861778,
      "num_tokens": 33566179.0,
      "step": 10045
    },
    {
      "entropy": 6.430149841308594,
      "epoch": 1.960686762267096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649436277542379,
      "loss": 6.3188,
      "mean_token_accuracy": 0.1313358724117279,
      "num_tokens": 33582504.0,
      "step": 10050
    },
    {
      "entropy": 6.403002166748047,
      "epoch": 1.961662276851039,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046490591669526174,
      "loss": 6.2067,
      "mean_token_accuracy": 0.1337461695075035,
      "num_tokens": 33597931.0,
      "step": 10055
    },
    {
      "entropy": 6.300563621520996,
      "epoch": 1.962637791434982,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004648681870793561,
      "loss": 6.1722,
      "mean_token_accuracy": 0.14268806278705598,
      "num_tokens": 33615027.0,
      "step": 10060
    },
    {
      "entropy": 6.402237224578857,
      "epoch": 1.963613306018925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000464830438910208,
      "loss": 6.2607,
      "mean_token_accuracy": 0.13468398451805114,
      "num_tokens": 33632484.0,
      "step": 10065
    },
    {
      "entropy": 6.436352920532227,
      "epoch": 1.9645888206028679,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046479267219150597,
      "loss": 6.3135,
      "mean_token_accuracy": 0.13108855485916138,
      "num_tokens": 33648889.0,
      "step": 10070
    },
    {
      "entropy": 6.378850269317627,
      "epoch": 1.965564335186811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004647548869269404,
      "loss": 6.2226,
      "mean_token_accuracy": 0.14269087389111518,
      "num_tokens": 33666105.0,
      "step": 10075
    },
    {
      "entropy": 6.401835823059082,
      "epoch": 1.966539849770754,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004647170831202036,
      "loss": 6.2695,
      "mean_token_accuracy": 0.13535130470991136,
      "num_tokens": 33683175.0,
      "step": 10080
    },
    {
      "entropy": 6.529552030563354,
      "epoch": 1.9675153643546972,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004646792607749896,
      "loss": 6.2611,
      "mean_token_accuracy": 0.12866187170147897,
      "num_tokens": 33698202.0,
      "step": 10085
    },
    {
      "entropy": 6.3124425411224365,
      "epoch": 1.9684908789386402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046464141989499433,
      "loss": 6.146,
      "mean_token_accuracy": 0.14231472089886665,
      "num_tokens": 33714573.0,
      "step": 10090
    },
    {
      "entropy": 6.306052732467651,
      "epoch": 1.9694663935225831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004646035604839153,
      "loss": 6.2053,
      "mean_token_accuracy": 0.13531112670898438,
      "num_tokens": 33731031.0,
      "step": 10095
    },
    {
      "entropy": 6.451034307479858,
      "epoch": 1.970441908106526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004645656825454522,
      "loss": 6.312,
      "mean_token_accuracy": 0.1342950388789177,
      "num_tokens": 33748228.0,
      "step": 10100
    },
    {
      "entropy": 6.333255290985107,
      "epoch": 1.9714174226904693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004645277860833062,
      "loss": 6.1637,
      "mean_token_accuracy": 0.1331145904958248,
      "num_tokens": 33765361.0,
      "step": 10105
    },
    {
      "entropy": 6.398478555679321,
      "epoch": 1.9723929372744122,
      "grad_norm": 0.875,
      "learning_rate": 0.00046448987110118034,
      "loss": 6.1772,
      "mean_token_accuracy": 0.14224255532026292,
      "num_tokens": 33781800.0,
      "step": 10110
    },
    {
      "entropy": 6.428394508361817,
      "epoch": 1.9733684518583554,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046445193760277976,
      "loss": 6.3102,
      "mean_token_accuracy": 0.13713693395256996,
      "num_tokens": 33797280.0,
      "step": 10115
    },
    {
      "entropy": 6.3418967723846436,
      "epoch": 1.9743439664422984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046441398559181096,
      "loss": 6.1549,
      "mean_token_accuracy": 0.13758799210190772,
      "num_tokens": 33814408.0,
      "step": 10120
    },
    {
      "entropy": 6.400722455978394,
      "epoch": 1.9753194810262413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046437601507198263,
      "loss": 6.2964,
      "mean_token_accuracy": 0.14066311568021775,
      "num_tokens": 33831499.0,
      "step": 10125
    },
    {
      "entropy": 6.473283338546753,
      "epoch": 1.9762949956101843,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046433802604700506,
      "loss": 6.3416,
      "mean_token_accuracy": 0.12951863706111907,
      "num_tokens": 33848456.0,
      "step": 10130
    },
    {
      "entropy": 6.413709735870361,
      "epoch": 1.9772705101941273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004643000185205903,
      "loss": 6.3556,
      "mean_token_accuracy": 0.12665458172559738,
      "num_tokens": 33863824.0,
      "step": 10135
    },
    {
      "entropy": 6.41361517906189,
      "epoch": 1.9782460247780704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046426199249645253,
      "loss": 6.2638,
      "mean_token_accuracy": 0.12781478092074394,
      "num_tokens": 33879382.0,
      "step": 10140
    },
    {
      "entropy": 6.331200504302979,
      "epoch": 1.9792215393620136,
      "grad_norm": 1.0,
      "learning_rate": 0.0004642239479783073,
      "loss": 6.1989,
      "mean_token_accuracy": 0.1379562310874462,
      "num_tokens": 33895930.0,
      "step": 10145
    },
    {
      "entropy": 6.1597757816314695,
      "epoch": 1.9801970539459566,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046418588496987233,
      "loss": 6.017,
      "mean_token_accuracy": 0.16122057139873505,
      "num_tokens": 33912067.0,
      "step": 10150
    },
    {
      "entropy": 6.446319055557251,
      "epoch": 1.9811725685298995,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000464147803474867,
      "loss": 6.2966,
      "mean_token_accuracy": 0.12850443720817567,
      "num_tokens": 33927481.0,
      "step": 10155
    },
    {
      "entropy": 6.389091491699219,
      "epoch": 1.9821480831138425,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004641097034970123,
      "loss": 6.1647,
      "mean_token_accuracy": 0.1418360337615013,
      "num_tokens": 33945373.0,
      "step": 10160
    },
    {
      "entropy": 6.383556079864502,
      "epoch": 1.9831235976977855,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046407158504003143,
      "loss": 6.2543,
      "mean_token_accuracy": 0.13663430139422417,
      "num_tokens": 33962441.0,
      "step": 10165
    },
    {
      "entropy": 6.2562073230743405,
      "epoch": 1.9840991122817286,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046403344810764933,
      "loss": 6.1274,
      "mean_token_accuracy": 0.14631522372364997,
      "num_tokens": 33979597.0,
      "step": 10170
    },
    {
      "entropy": 6.447936201095581,
      "epoch": 1.9850746268656716,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046399529270359214,
      "loss": 6.2614,
      "mean_token_accuracy": 0.1408032774925232,
      "num_tokens": 33996079.0,
      "step": 10175
    },
    {
      "entropy": 6.210469055175781,
      "epoch": 1.9860501414496148,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046395711883158864,
      "loss": 6.1112,
      "mean_token_accuracy": 0.14614792317152023,
      "num_tokens": 34014106.0,
      "step": 10180
    },
    {
      "entropy": 6.379308366775513,
      "epoch": 1.9870256560335577,
      "grad_norm": 1.03125,
      "learning_rate": 0.000463918926495369,
      "loss": 6.2238,
      "mean_token_accuracy": 0.1340431921184063,
      "num_tokens": 34030815.0,
      "step": 10185
    },
    {
      "entropy": 6.381000137329101,
      "epoch": 1.9880011706175007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004638807156986651,
      "loss": 6.1796,
      "mean_token_accuracy": 0.14317063093185425,
      "num_tokens": 34047453.0,
      "step": 10190
    },
    {
      "entropy": 6.358499431610108,
      "epoch": 1.9889766852014437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004638424864452109,
      "loss": 6.1648,
      "mean_token_accuracy": 0.13990716338157655,
      "num_tokens": 34065442.0,
      "step": 10195
    },
    {
      "entropy": 6.310580253601074,
      "epoch": 1.9899521997853868,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004638042387387419,
      "loss": 6.1459,
      "mean_token_accuracy": 0.14379386305809022,
      "num_tokens": 34082073.0,
      "step": 10200
    },
    {
      "entropy": 6.458771991729736,
      "epoch": 1.9909277143693298,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004637659725829957,
      "loss": 6.2713,
      "mean_token_accuracy": 0.13161378875374793,
      "num_tokens": 34097542.0,
      "step": 10205
    },
    {
      "entropy": 6.2479596614837645,
      "epoch": 1.991903228953273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004637276879817114,
      "loss": 6.1144,
      "mean_token_accuracy": 0.14662338569760322,
      "num_tokens": 34114961.0,
      "step": 10210
    },
    {
      "entropy": 6.323702812194824,
      "epoch": 1.992878743537216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046368938493863,
      "loss": 6.145,
      "mean_token_accuracy": 0.14644043445587157,
      "num_tokens": 34131774.0,
      "step": 10215
    },
    {
      "entropy": 6.4197783946990965,
      "epoch": 1.993854258121159,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046365106345749445,
      "loss": 6.2382,
      "mean_token_accuracy": 0.13264575526118277,
      "num_tokens": 34147865.0,
      "step": 10220
    },
    {
      "entropy": 6.285823917388916,
      "epoch": 1.9948297727051019,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004636127235420493,
      "loss": 6.1773,
      "mean_token_accuracy": 0.14502398371696473,
      "num_tokens": 34164029.0,
      "step": 10225
    },
    {
      "entropy": 6.211844635009766,
      "epoch": 1.9958052872890448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046357436519604107,
      "loss": 5.9879,
      "mean_token_accuracy": 0.1631839692592621,
      "num_tokens": 34181026.0,
      "step": 10230
    },
    {
      "entropy": 6.3903114795684814,
      "epoch": 1.996780801872988,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463535988423218,
      "loss": 6.1927,
      "mean_token_accuracy": 0.14325509667396547,
      "num_tokens": 34197129.0,
      "step": 10235
    },
    {
      "entropy": 6.238297319412231,
      "epoch": 1.9977563164569312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004634975932273301,
      "loss": 6.1554,
      "mean_token_accuracy": 0.14841860458254813,
      "num_tokens": 34214277.0,
      "step": 10240
    },
    {
      "entropy": 6.3491109848022464,
      "epoch": 1.9987318310408742,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046345917961212915,
      "loss": 6.1984,
      "mean_token_accuracy": 0.1424255460500717,
      "num_tokens": 34231491.0,
      "step": 10245
    },
    {
      "entropy": 6.4141514778137205,
      "epoch": 1.9997073456248171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046342074758136886,
      "loss": 6.206,
      "mean_token_accuracy": 0.13676658719778062,
      "num_tokens": 34248000.0,
      "step": 10250
    },
    {
      "entropy": 6.31264861424764,
      "epoch": 2.0005853087503658,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004633822971388047,
      "loss": 6.131,
      "mean_token_accuracy": 0.1551901267634498,
      "num_tokens": 34262606.0,
      "step": 10255
    },
    {
      "entropy": 6.231221342086792,
      "epoch": 2.0015608233343087,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046334382828819385,
      "loss": 5.9704,
      "mean_token_accuracy": 0.15866035372018814,
      "num_tokens": 34280759.0,
      "step": 10260
    },
    {
      "entropy": 6.334379434585571,
      "epoch": 2.0025363379182517,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004633053410332953,
      "loss": 6.122,
      "mean_token_accuracy": 0.14013804793357848,
      "num_tokens": 34296420.0,
      "step": 10265
    },
    {
      "entropy": 6.423362350463867,
      "epoch": 2.003511852502195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046326683537787006,
      "loss": 6.1857,
      "mean_token_accuracy": 0.1354950875043869,
      "num_tokens": 34313171.0,
      "step": 10270
    },
    {
      "entropy": 6.372868442535401,
      "epoch": 2.004487367086138,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046322831132568067,
      "loss": 6.0958,
      "mean_token_accuracy": 0.1399066776037216,
      "num_tokens": 34330885.0,
      "step": 10275
    },
    {
      "entropy": 6.347308015823364,
      "epoch": 2.005462881670081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004631897688804914,
      "loss": 6.0437,
      "mean_token_accuracy": 0.13959520533680916,
      "num_tokens": 34346661.0,
      "step": 10280
    },
    {
      "entropy": 6.346377325057984,
      "epoch": 2.006438396254024,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004631512080460688,
      "loss": 6.1378,
      "mean_token_accuracy": 0.13864003121852875,
      "num_tokens": 34362480.0,
      "step": 10285
    },
    {
      "entropy": 6.25256462097168,
      "epoch": 2.007413910837967,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046311262882618064,
      "loss": 5.9994,
      "mean_token_accuracy": 0.14761052429676055,
      "num_tokens": 34379719.0,
      "step": 10290
    },
    {
      "entropy": 6.380419063568115,
      "epoch": 2.00838942542191,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004630740312245969,
      "loss": 6.1396,
      "mean_token_accuracy": 0.1373915985226631,
      "num_tokens": 34396088.0,
      "step": 10295
    },
    {
      "entropy": 6.319455242156982,
      "epoch": 2.0093649400058533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046303541524508906,
      "loss": 6.076,
      "mean_token_accuracy": 0.141716568171978,
      "num_tokens": 34413658.0,
      "step": 10300
    },
    {
      "entropy": 6.4406133651733395,
      "epoch": 2.0103404545897963,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004629967808914307,
      "loss": 6.2218,
      "mean_token_accuracy": 0.13505981042981147,
      "num_tokens": 34430799.0,
      "step": 10305
    },
    {
      "entropy": 6.4476865291595455,
      "epoch": 2.0113159691737392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004629581281673969,
      "loss": 6.1817,
      "mean_token_accuracy": 0.14121901020407676,
      "num_tokens": 34447177.0,
      "step": 10310
    },
    {
      "entropy": 6.400524759292603,
      "epoch": 2.012291483757682,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046291945707676473,
      "loss": 6.0926,
      "mean_token_accuracy": 0.1380524791777134,
      "num_tokens": 34463175.0,
      "step": 10315
    },
    {
      "entropy": 6.233956909179687,
      "epoch": 2.013266998341625,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046288076762331294,
      "loss": 6.0283,
      "mean_token_accuracy": 0.14747570157051088,
      "num_tokens": 34482233.0,
      "step": 10320
    },
    {
      "entropy": 6.398956680297852,
      "epoch": 2.014242512925568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004628420598108222,
      "loss": 6.1882,
      "mean_token_accuracy": 0.13884783759713173,
      "num_tokens": 34498780.0,
      "step": 10325
    },
    {
      "entropy": 6.293699836730957,
      "epoch": 2.015218027509511,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046280333364307483,
      "loss": 5.9686,
      "mean_token_accuracy": 0.14757298827171325,
      "num_tokens": 34516723.0,
      "step": 10330
    },
    {
      "entropy": 6.317723608016967,
      "epoch": 2.0161935420934545,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000462764589123855,
      "loss": 6.1121,
      "mean_token_accuracy": 0.13512558862566948,
      "num_tokens": 34534291.0,
      "step": 10335
    },
    {
      "entropy": 6.2723602771759035,
      "epoch": 2.0171690566773974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004627258262569487,
      "loss": 5.9633,
      "mean_token_accuracy": 0.14921482652425766,
      "num_tokens": 34551349.0,
      "step": 10340
    },
    {
      "entropy": 6.249519395828247,
      "epoch": 2.0181445712613404,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046268704504614383,
      "loss": 5.9441,
      "mean_token_accuracy": 0.15334743037819862,
      "num_tokens": 34569014.0,
      "step": 10345
    },
    {
      "entropy": 6.4147961139678955,
      "epoch": 2.0191200858452834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004626482454952298,
      "loss": 6.2003,
      "mean_token_accuracy": 0.135830120742321,
      "num_tokens": 34586855.0,
      "step": 10350
    },
    {
      "entropy": 6.354845476150513,
      "epoch": 2.0200956004292263,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000462609427607998,
      "loss": 6.0256,
      "mean_token_accuracy": 0.15259339213371276,
      "num_tokens": 34602809.0,
      "step": 10355
    },
    {
      "entropy": 6.322290849685669,
      "epoch": 2.0210711150131693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004625705913882416,
      "loss": 6.1138,
      "mean_token_accuracy": 0.14134227856993675,
      "num_tokens": 34618371.0,
      "step": 10360
    },
    {
      "entropy": 6.22515115737915,
      "epoch": 2.0220466295971127,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046253173683975557,
      "loss": 6.0269,
      "mean_token_accuracy": 0.13785478994250297,
      "num_tokens": 34635118.0,
      "step": 10365
    },
    {
      "entropy": 6.36070556640625,
      "epoch": 2.0230221441810556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004624928639663365,
      "loss": 6.0792,
      "mean_token_accuracy": 0.14499531537294388,
      "num_tokens": 34651827.0,
      "step": 10370
    },
    {
      "entropy": 6.349448251724243,
      "epoch": 2.0239976587649986,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004624539727717831,
      "loss": 6.1134,
      "mean_token_accuracy": 0.13573916554450988,
      "num_tokens": 34668698.0,
      "step": 10375
    },
    {
      "entropy": 6.425476312637329,
      "epoch": 2.0249731733489416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046241506325989546,
      "loss": 6.1366,
      "mean_token_accuracy": 0.14368387460708618,
      "num_tokens": 34686257.0,
      "step": 10380
    },
    {
      "entropy": 6.353345203399658,
      "epoch": 2.0259486879328845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004623761354344759,
      "loss": 6.0903,
      "mean_token_accuracy": 0.1392493113875389,
      "num_tokens": 34702088.0,
      "step": 10385
    },
    {
      "entropy": 6.299949789047242,
      "epoch": 2.0269242025168275,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004623371892993281,
      "loss": 6.0852,
      "mean_token_accuracy": 0.149676176905632,
      "num_tokens": 34718940.0,
      "step": 10390
    },
    {
      "entropy": 6.323379611968994,
      "epoch": 2.0278997171007704,
      "grad_norm": 1.015625,
      "learning_rate": 0.000462298224858258,
      "loss": 6.1141,
      "mean_token_accuracy": 0.14253847524523736,
      "num_tokens": 34735228.0,
      "step": 10395
    },
    {
      "entropy": 6.432223224639893,
      "epoch": 2.028875231684714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046225924211507275,
      "loss": 6.1736,
      "mean_token_accuracy": 0.13810604810714722,
      "num_tokens": 34751741.0,
      "step": 10400
    },
    {
      "entropy": 6.300885105133057,
      "epoch": 2.029850746268657,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004622202410735818,
      "loss": 6.0328,
      "mean_token_accuracy": 0.14386026263237,
      "num_tokens": 34767152.0,
      "step": 10405
    },
    {
      "entropy": 6.317934656143189,
      "epoch": 2.0308262608525998,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046218122173759626,
      "loss": 6.1044,
      "mean_token_accuracy": 0.1387723922729492,
      "num_tokens": 34784419.0,
      "step": 10410
    },
    {
      "entropy": 6.252539491653442,
      "epoch": 2.0318017754365427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046214218411092867,
      "loss": 5.9978,
      "mean_token_accuracy": 0.15149002522230148,
      "num_tokens": 34799877.0,
      "step": 10415
    },
    {
      "entropy": 6.383495759963989,
      "epoch": 2.0327772900204857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046210312819739406,
      "loss": 6.1652,
      "mean_token_accuracy": 0.135135205835104,
      "num_tokens": 34816332.0,
      "step": 10420
    },
    {
      "entropy": 6.431286191940307,
      "epoch": 2.0337528046044286,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004620640540008084,
      "loss": 6.1524,
      "mean_token_accuracy": 0.1369701586663723,
      "num_tokens": 34833364.0,
      "step": 10425
    },
    {
      "entropy": 6.209193897247315,
      "epoch": 2.034728319188372,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004620249615249902,
      "loss": 5.8911,
      "mean_token_accuracy": 0.16134502738714218,
      "num_tokens": 34851419.0,
      "step": 10430
    },
    {
      "entropy": 6.318286991119384,
      "epoch": 2.035703833772315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004619858507737594,
      "loss": 6.1315,
      "mean_token_accuracy": 0.14211411625146866,
      "num_tokens": 34869237.0,
      "step": 10435
    },
    {
      "entropy": 6.27009768486023,
      "epoch": 2.036679348356258,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004619467217509375,
      "loss": 6.0297,
      "mean_token_accuracy": 0.15027124136686326,
      "num_tokens": 34888254.0,
      "step": 10440
    },
    {
      "entropy": 6.3556770324707035,
      "epoch": 2.037654862940201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046190757446034844,
      "loss": 6.1397,
      "mean_token_accuracy": 0.14206055030226708,
      "num_tokens": 34905257.0,
      "step": 10445
    },
    {
      "entropy": 6.332166481018066,
      "epoch": 2.038630377524144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046186840890581724,
      "loss": 6.047,
      "mean_token_accuracy": 0.14275565296411513,
      "num_tokens": 34922358.0,
      "step": 10450
    },
    {
      "entropy": 6.369397926330566,
      "epoch": 2.039605892108087,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046182922509117115,
      "loss": 6.0514,
      "mean_token_accuracy": 0.14699846357107163,
      "num_tokens": 34939746.0,
      "step": 10455
    },
    {
      "entropy": 6.262325859069824,
      "epoch": 2.0405814066920303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004617900230202391,
      "loss": 6.0091,
      "mean_token_accuracy": 0.1448981337249279,
      "num_tokens": 34957188.0,
      "step": 10460
    },
    {
      "entropy": 6.268700790405274,
      "epoch": 2.041556921275973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046175080269685167,
      "loss": 6.0735,
      "mean_token_accuracy": 0.14958121851086617,
      "num_tokens": 34974859.0,
      "step": 10465
    },
    {
      "entropy": 6.2782923698425295,
      "epoch": 2.042532435859916,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046171156412484136,
      "loss": 6.0111,
      "mean_token_accuracy": 0.15425544530153273,
      "num_tokens": 34990806.0,
      "step": 10470
    },
    {
      "entropy": 6.292570495605469,
      "epoch": 2.043507950443859,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046167230730804256,
      "loss": 6.0697,
      "mean_token_accuracy": 0.14735029712319375,
      "num_tokens": 35006644.0,
      "step": 10475
    },
    {
      "entropy": 6.364605188369751,
      "epoch": 2.044483465027802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004616330322502911,
      "loss": 6.1347,
      "mean_token_accuracy": 0.1410667836666107,
      "num_tokens": 35023818.0,
      "step": 10480
    },
    {
      "entropy": 6.384490013122559,
      "epoch": 2.045458979611745,
      "grad_norm": 1.0,
      "learning_rate": 0.00046159373895542506,
      "loss": 6.132,
      "mean_token_accuracy": 0.1391760766506195,
      "num_tokens": 35039447.0,
      "step": 10485
    },
    {
      "entropy": 6.308459663391114,
      "epoch": 2.046434494195688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004615544274272837,
      "loss": 6.085,
      "mean_token_accuracy": 0.14373720288276673,
      "num_tokens": 35054831.0,
      "step": 10490
    },
    {
      "entropy": 6.349425554275513,
      "epoch": 2.0474100087796314,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004615150976697087,
      "loss": 6.0547,
      "mean_token_accuracy": 0.14313050508499145,
      "num_tokens": 35071266.0,
      "step": 10495
    },
    {
      "entropy": 6.397823619842529,
      "epoch": 2.0483855233635744,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000461475749686543,
      "loss": 6.1356,
      "mean_token_accuracy": 0.14042172133922576,
      "num_tokens": 35087696.0,
      "step": 10500
    },
    {
      "entropy": 6.2558544158935545,
      "epoch": 2.0493610379475173,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004614363834816318,
      "loss": 6.0212,
      "mean_token_accuracy": 0.14533352851867676,
      "num_tokens": 35104914.0,
      "step": 10505
    },
    {
      "entropy": 6.322768640518189,
      "epoch": 2.0503365525314603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004613969990588216,
      "loss": 6.1212,
      "mean_token_accuracy": 0.14560258537530898,
      "num_tokens": 35121777.0,
      "step": 10510
    },
    {
      "entropy": 6.263347578048706,
      "epoch": 2.0513120671154033,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046135759642196094,
      "loss": 6.0793,
      "mean_token_accuracy": 0.15307567566633223,
      "num_tokens": 35141293.0,
      "step": 10515
    },
    {
      "entropy": 6.432005405426025,
      "epoch": 2.052287581699346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046131817557490016,
      "loss": 6.1787,
      "mean_token_accuracy": 0.13740066066384315,
      "num_tokens": 35157596.0,
      "step": 10520
    },
    {
      "entropy": 6.27400393486023,
      "epoch": 2.0532630962832896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004612787365214914,
      "loss": 6.047,
      "mean_token_accuracy": 0.14720142185688018,
      "num_tokens": 35173060.0,
      "step": 10525
    },
    {
      "entropy": 6.345219898223877,
      "epoch": 2.0542386108672326,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046123927926558836,
      "loss": 6.056,
      "mean_token_accuracy": 0.13511715456843376,
      "num_tokens": 35189846.0,
      "step": 10530
    },
    {
      "entropy": 6.328106641769409,
      "epoch": 2.0552141254511755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004611998038110467,
      "loss": 6.13,
      "mean_token_accuracy": 0.14349254667758943,
      "num_tokens": 35206355.0,
      "step": 10535
    },
    {
      "entropy": 6.228447484970093,
      "epoch": 2.0561896400351185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004611603101617239,
      "loss": 5.9829,
      "mean_token_accuracy": 0.14944793432950973,
      "num_tokens": 35223535.0,
      "step": 10540
    },
    {
      "entropy": 6.2420190334320065,
      "epoch": 2.0571651546190615,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046112079832147894,
      "loss": 5.9086,
      "mean_token_accuracy": 0.15675691664218902,
      "num_tokens": 35241283.0,
      "step": 10545
    },
    {
      "entropy": 6.296031761169433,
      "epoch": 2.0581406692030044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000461081268294173,
      "loss": 6.0796,
      "mean_token_accuracy": 0.1385519489645958,
      "num_tokens": 35257398.0,
      "step": 10550
    },
    {
      "entropy": 6.2507524490356445,
      "epoch": 2.059116183786948,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004610417200836688,
      "loss": 5.9708,
      "mean_token_accuracy": 0.15392484813928603,
      "num_tokens": 35275121.0,
      "step": 10555
    },
    {
      "entropy": 6.2712461948394775,
      "epoch": 2.060091698370891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004610021536938307,
      "loss": 6.0339,
      "mean_token_accuracy": 0.13858166113495826,
      "num_tokens": 35290169.0,
      "step": 10560
    },
    {
      "entropy": 6.259926748275757,
      "epoch": 2.0610672129548337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046096256912852505,
      "loss": 6.0099,
      "mean_token_accuracy": 0.14764256849884988,
      "num_tokens": 35305824.0,
      "step": 10565
    },
    {
      "entropy": 6.388303899765015,
      "epoch": 2.0620427275387767,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004609229663916199,
      "loss": 6.088,
      "mean_token_accuracy": 0.13928432762622833,
      "num_tokens": 35322182.0,
      "step": 10570
    },
    {
      "entropy": 6.336553621292114,
      "epoch": 2.0630182421227197,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046088334548698525,
      "loss": 6.1883,
      "mean_token_accuracy": 0.14324963390827178,
      "num_tokens": 35339046.0,
      "step": 10575
    },
    {
      "entropy": 6.311831283569336,
      "epoch": 2.0639937567066626,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004608437064184925,
      "loss": 6.0821,
      "mean_token_accuracy": 0.14232448637485504,
      "num_tokens": 35355448.0,
      "step": 10580
    },
    {
      "entropy": 6.307275819778442,
      "epoch": 2.0649692712906056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046080404919001506,
      "loss": 5.9768,
      "mean_token_accuracy": 0.1490872673690319,
      "num_tokens": 35372738.0,
      "step": 10585
    },
    {
      "entropy": 6.279605960845947,
      "epoch": 2.065944785874549,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046076437380542824,
      "loss": 6.045,
      "mean_token_accuracy": 0.14438512921333313,
      "num_tokens": 35389344.0,
      "step": 10590
    },
    {
      "entropy": 6.213056516647339,
      "epoch": 2.066920300458492,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004607246802686089,
      "loss": 5.9458,
      "mean_token_accuracy": 0.14828964620828627,
      "num_tokens": 35405897.0,
      "step": 10595
    },
    {
      "entropy": 6.2875138282775875,
      "epoch": 2.067895815042435,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004606849685834356,
      "loss": 6.0597,
      "mean_token_accuracy": 0.1478186346590519,
      "num_tokens": 35423084.0,
      "step": 10600
    },
    {
      "entropy": 6.2869476795196535,
      "epoch": 2.068871329626378,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004606452387537891,
      "loss": 6.0148,
      "mean_token_accuracy": 0.15416245386004448,
      "num_tokens": 35440411.0,
      "step": 10605
    },
    {
      "entropy": 6.374316263198852,
      "epoch": 2.069846844210321,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004606054907835514,
      "loss": 6.151,
      "mean_token_accuracy": 0.1400749869644642,
      "num_tokens": 35456519.0,
      "step": 10610
    },
    {
      "entropy": 6.355369424819946,
      "epoch": 2.070822358794264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004605657246766068,
      "loss": 6.0672,
      "mean_token_accuracy": 0.14235804304480554,
      "num_tokens": 35473271.0,
      "step": 10615
    },
    {
      "entropy": 6.341520118713379,
      "epoch": 2.071797873378207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046052594043684086,
      "loss": 6.1449,
      "mean_token_accuracy": 0.141668139398098,
      "num_tokens": 35491360.0,
      "step": 10620
    },
    {
      "entropy": 6.397606611251831,
      "epoch": 2.07277338796215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046048613806814125,
      "loss": 6.1207,
      "mean_token_accuracy": 0.1400507003068924,
      "num_tokens": 35508481.0,
      "step": 10625
    },
    {
      "entropy": 6.369201564788819,
      "epoch": 2.073748902546093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004604463175743974,
      "loss": 6.1148,
      "mean_token_accuracy": 0.1439947634935379,
      "num_tokens": 35524895.0,
      "step": 10630
    },
    {
      "entropy": 6.300136661529541,
      "epoch": 2.074724417130036,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046040647895950016,
      "loss": 6.1214,
      "mean_token_accuracy": 0.13854775205254555,
      "num_tokens": 35541644.0,
      "step": 10635
    },
    {
      "entropy": 6.334607124328613,
      "epoch": 2.075699931713979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046036662222734276,
      "loss": 6.0685,
      "mean_token_accuracy": 0.13893636614084243,
      "num_tokens": 35557211.0,
      "step": 10640
    },
    {
      "entropy": 6.332695341110229,
      "epoch": 2.076675446297922,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004603267473818196,
      "loss": 6.0709,
      "mean_token_accuracy": 0.14095504209399223,
      "num_tokens": 35574366.0,
      "step": 10645
    },
    {
      "entropy": 6.226788902282715,
      "epoch": 2.077650960881865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004602868544268272,
      "loss": 5.9448,
      "mean_token_accuracy": 0.15750849917531012,
      "num_tokens": 35592172.0,
      "step": 10650
    },
    {
      "entropy": 6.327335929870605,
      "epoch": 2.0786264754658084,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004602469433662639,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1493236839771271,
      "num_tokens": 35608225.0,
      "step": 10655
    },
    {
      "entropy": 6.310957717895508,
      "epoch": 2.0796019900497513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046020701420402936,
      "loss": 6.0826,
      "mean_token_accuracy": 0.13885179534554482,
      "num_tokens": 35624474.0,
      "step": 10660
    },
    {
      "entropy": 6.384568119049073,
      "epoch": 2.0805775046336943,
      "grad_norm": 1.0,
      "learning_rate": 0.00046016706694402556,
      "loss": 6.1283,
      "mean_token_accuracy": 0.13876380026340485,
      "num_tokens": 35639800.0,
      "step": 10665
    },
    {
      "entropy": 6.234939813613892,
      "epoch": 2.0815530192176372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004601271015901559,
      "loss": 6.0056,
      "mean_token_accuracy": 0.14909723699092864,
      "num_tokens": 35656830.0,
      "step": 10670
    },
    {
      "entropy": 6.302708768844605,
      "epoch": 2.08252853380158,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004600871181463257,
      "loss": 6.0789,
      "mean_token_accuracy": 0.14402605444192887,
      "num_tokens": 35673394.0,
      "step": 10675
    },
    {
      "entropy": 6.3660026550292965,
      "epoch": 2.083504048385523,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004600471166164419,
      "loss": 6.0443,
      "mean_token_accuracy": 0.14694800078868867,
      "num_tokens": 35690135.0,
      "step": 10680
    },
    {
      "entropy": 6.363548421859742,
      "epoch": 2.0844795629694666,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004600070970044135,
      "loss": 6.1948,
      "mean_token_accuracy": 0.13991641104221345,
      "num_tokens": 35706817.0,
      "step": 10685
    },
    {
      "entropy": 6.291651248931885,
      "epoch": 2.0854550775534095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045996705931415096,
      "loss": 6.0616,
      "mean_token_accuracy": 0.14560936838388444,
      "num_tokens": 35724929.0,
      "step": 10690
    },
    {
      "entropy": 6.200566816329956,
      "epoch": 2.0864305921373525,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004599270035495665,
      "loss": 5.935,
      "mean_token_accuracy": 0.1588302344083786,
      "num_tokens": 35741892.0,
      "step": 10695
    },
    {
      "entropy": 6.2709449291229244,
      "epoch": 2.0874061067212955,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045988692971457443,
      "loss": 5.9735,
      "mean_token_accuracy": 0.14722650945186616,
      "num_tokens": 35758486.0,
      "step": 10700
    },
    {
      "entropy": 6.298744487762451,
      "epoch": 2.0883816213052384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004598468378130906,
      "loss": 6.0738,
      "mean_token_accuracy": 0.14366669803857804,
      "num_tokens": 35774118.0,
      "step": 10705
    },
    {
      "entropy": 6.231470203399658,
      "epoch": 2.0893571358891814,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045980672784903245,
      "loss": 6.0,
      "mean_token_accuracy": 0.14757902547717094,
      "num_tokens": 35790442.0,
      "step": 10710
    },
    {
      "entropy": 6.275332403182984,
      "epoch": 2.0903326504731248,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045976659982631954,
      "loss": 6.0364,
      "mean_token_accuracy": 0.14612374007701873,
      "num_tokens": 35805498.0,
      "step": 10715
    },
    {
      "entropy": 6.288800621032715,
      "epoch": 2.0913081650570677,
      "grad_norm": 1.171875,
      "learning_rate": 0.000459726453748873,
      "loss": 6.0266,
      "mean_token_accuracy": 0.15030321031808852,
      "num_tokens": 35821477.0,
      "step": 10720
    },
    {
      "entropy": 6.287510776519776,
      "epoch": 2.0922836796410107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045968628962061585,
      "loss": 6.0583,
      "mean_token_accuracy": 0.1467307075858116,
      "num_tokens": 35838462.0,
      "step": 10725
    },
    {
      "entropy": 6.328286790847779,
      "epoch": 2.0932591942249537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004596461074454727,
      "loss": 6.0866,
      "mean_token_accuracy": 0.13928924649953842,
      "num_tokens": 35854814.0,
      "step": 10730
    },
    {
      "entropy": 6.343556976318359,
      "epoch": 2.0942347088088966,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045960590722737,
      "loss": 6.1303,
      "mean_token_accuracy": 0.13566041588783265,
      "num_tokens": 35870117.0,
      "step": 10735
    },
    {
      "entropy": 6.421367359161377,
      "epoch": 2.0952102233928396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000459565688970236,
      "loss": 6.1115,
      "mean_token_accuracy": 0.1419231429696083,
      "num_tokens": 35884900.0,
      "step": 10740
    },
    {
      "entropy": 6.317118787765503,
      "epoch": 2.096185737976783,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004595254526780007,
      "loss": 6.0583,
      "mean_token_accuracy": 0.1504852794110775,
      "num_tokens": 35901280.0,
      "step": 10745
    },
    {
      "entropy": 6.288026285171509,
      "epoch": 2.097161252560726,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045948519835459575,
      "loss": 6.0831,
      "mean_token_accuracy": 0.14773936420679093,
      "num_tokens": 35918015.0,
      "step": 10750
    },
    {
      "entropy": 6.355995750427246,
      "epoch": 2.098136767144669,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045944492600395476,
      "loss": 6.1393,
      "mean_token_accuracy": 0.14304338693618773,
      "num_tokens": 35935367.0,
      "step": 10755
    },
    {
      "entropy": 6.328325033187866,
      "epoch": 2.099112281728612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000459404635630013,
      "loss": 6.0982,
      "mean_token_accuracy": 0.14105041176080704,
      "num_tokens": 35951697.0,
      "step": 10760
    },
    {
      "entropy": 6.3235735416412355,
      "epoch": 2.100087796312555,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004593643272367075,
      "loss": 6.0555,
      "mean_token_accuracy": 0.14655024111270903,
      "num_tokens": 35968321.0,
      "step": 10765
    },
    {
      "entropy": 6.366960716247559,
      "epoch": 2.101063310896498,
      "grad_norm": 1.046875,
      "learning_rate": 0.000459324000827977,
      "loss": 6.0524,
      "mean_token_accuracy": 0.14268654584884644,
      "num_tokens": 35984192.0,
      "step": 10770
    },
    {
      "entropy": 6.320577001571655,
      "epoch": 2.1020388254804407,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004592836564077621,
      "loss": 6.1081,
      "mean_token_accuracy": 0.13892083317041398,
      "num_tokens": 36001208.0,
      "step": 10775
    },
    {
      "entropy": 6.345907545089721,
      "epoch": 2.103014340064384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000459243293980005,
      "loss": 6.0713,
      "mean_token_accuracy": 0.1401558220386505,
      "num_tokens": 36018253.0,
      "step": 10780
    },
    {
      "entropy": 6.218587160110474,
      "epoch": 2.103989854648327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045920291354865,
      "loss": 5.9679,
      "mean_token_accuracy": 0.14950090795755386,
      "num_tokens": 36035041.0,
      "step": 10785
    },
    {
      "entropy": 6.3904133319854735,
      "epoch": 2.10496536923227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045916251511764265,
      "loss": 6.1677,
      "mean_token_accuracy": 0.13784012347459793,
      "num_tokens": 36051497.0,
      "step": 10790
    },
    {
      "entropy": 6.32241415977478,
      "epoch": 2.105940883816213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004591220986909309,
      "loss": 6.0645,
      "mean_token_accuracy": 0.13994555920362473,
      "num_tokens": 36068117.0,
      "step": 10795
    },
    {
      "entropy": 6.283169603347778,
      "epoch": 2.106916398400156,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045908166427246377,
      "loss": 6.0688,
      "mean_token_accuracy": 0.14199872314929962,
      "num_tokens": 36084489.0,
      "step": 10800
    },
    {
      "entropy": 6.229799890518189,
      "epoch": 2.107891912984099,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004590412118661925,
      "loss": 5.9281,
      "mean_token_accuracy": 0.1469028562307358,
      "num_tokens": 36102633.0,
      "step": 10805
    },
    {
      "entropy": 6.190728092193604,
      "epoch": 2.1088674275680424,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004590007414760698,
      "loss": 5.9629,
      "mean_token_accuracy": 0.15746331140398978,
      "num_tokens": 36121581.0,
      "step": 10810
    },
    {
      "entropy": 6.234311151504516,
      "epoch": 2.1098429421519853,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045896025310605067,
      "loss": 6.0345,
      "mean_token_accuracy": 0.14943696409463883,
      "num_tokens": 36138266.0,
      "step": 10815
    },
    {
      "entropy": 6.188042306900025,
      "epoch": 2.1108184567359283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004589197467600911,
      "loss": 5.9741,
      "mean_token_accuracy": 0.15357098430395127,
      "num_tokens": 36154365.0,
      "step": 10820
    },
    {
      "entropy": 6.254004049301147,
      "epoch": 2.1117939713198712,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004588792224421494,
      "loss": 6.0337,
      "mean_token_accuracy": 0.15057696104049684,
      "num_tokens": 36173469.0,
      "step": 10825
    },
    {
      "entropy": 6.333424758911133,
      "epoch": 2.112769485903814,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045883868015618554,
      "loss": 6.1019,
      "mean_token_accuracy": 0.14176615253090857,
      "num_tokens": 36190032.0,
      "step": 10830
    },
    {
      "entropy": 6.283120012283325,
      "epoch": 2.113745000487757,
      "grad_norm": 0.9375,
      "learning_rate": 0.000458798119906161,
      "loss": 6.0782,
      "mean_token_accuracy": 0.14284923076629638,
      "num_tokens": 36207681.0,
      "step": 10835
    },
    {
      "entropy": 6.320326042175293,
      "epoch": 2.1147205150717,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004587575416960391,
      "loss": 6.1074,
      "mean_token_accuracy": 0.14424423575401307,
      "num_tokens": 36225031.0,
      "step": 10840
    },
    {
      "entropy": 6.3230860233306885,
      "epoch": 2.1156960296556435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045871694552978534,
      "loss": 5.9724,
      "mean_token_accuracy": 0.15131415501236917,
      "num_tokens": 36241579.0,
      "step": 10845
    },
    {
      "entropy": 6.210815191268921,
      "epoch": 2.1166715442395865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045867633141136643,
      "loss": 6.0158,
      "mean_token_accuracy": 0.15266728699207305,
      "num_tokens": 36257978.0,
      "step": 10850
    },
    {
      "entropy": 6.3533852100372314,
      "epoch": 2.1176470588235294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045863569934475104,
      "loss": 6.1143,
      "mean_token_accuracy": 0.13661568164825438,
      "num_tokens": 36274068.0,
      "step": 10855
    },
    {
      "entropy": 6.097590732574463,
      "epoch": 2.1186225734074724,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045859504933390947,
      "loss": 5.8661,
      "mean_token_accuracy": 0.16164280474185944,
      "num_tokens": 36291033.0,
      "step": 10860
    },
    {
      "entropy": 6.271762800216675,
      "epoch": 2.1195980879914154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004585543813828141,
      "loss": 6.0596,
      "mean_token_accuracy": 0.14740503281354905,
      "num_tokens": 36307430.0,
      "step": 10865
    },
    {
      "entropy": 6.45457878112793,
      "epoch": 2.1205736025753583,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004585136954954389,
      "loss": 6.2054,
      "mean_token_accuracy": 0.13325666189193724,
      "num_tokens": 36324257.0,
      "step": 10870
    },
    {
      "entropy": 6.3684813499450685,
      "epoch": 2.1215491171593017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004584729916757593,
      "loss": 6.1244,
      "mean_token_accuracy": 0.1387636363506317,
      "num_tokens": 36339864.0,
      "step": 10875
    },
    {
      "entropy": 6.234913444519043,
      "epoch": 2.1225246317432447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045843226992775294,
      "loss": 6.028,
      "mean_token_accuracy": 0.14726742208003998,
      "num_tokens": 36356066.0,
      "step": 10880
    },
    {
      "entropy": 6.264840507507325,
      "epoch": 2.1235001463271876,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045839153025539883,
      "loss": 6.0137,
      "mean_token_accuracy": 0.15147039592266082,
      "num_tokens": 36374420.0,
      "step": 10885
    },
    {
      "entropy": 6.333682060241699,
      "epoch": 2.1244756609111306,
      "grad_norm": 0.875,
      "learning_rate": 0.00045835077266267803,
      "loss": 6.149,
      "mean_token_accuracy": 0.1363453172147274,
      "num_tokens": 36390884.0,
      "step": 10890
    },
    {
      "entropy": 6.350889444351196,
      "epoch": 2.1254511754950736,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045830999715357326,
      "loss": 6.0612,
      "mean_token_accuracy": 0.1402958333492279,
      "num_tokens": 36407730.0,
      "step": 10895
    },
    {
      "entropy": 6.256021165847779,
      "epoch": 2.1264266900790165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045826920373206884,
      "loss": 6.0922,
      "mean_token_accuracy": 0.13767525553703308,
      "num_tokens": 36423470.0,
      "step": 10900
    },
    {
      "entropy": 6.305959558486938,
      "epoch": 2.1274022046629595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000458228392402151,
      "loss": 6.0701,
      "mean_token_accuracy": 0.14378181472420692,
      "num_tokens": 36439478.0,
      "step": 10905
    },
    {
      "entropy": 6.352621984481812,
      "epoch": 2.128377719246903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004581875631678077,
      "loss": 6.1379,
      "mean_token_accuracy": 0.13772692009806634,
      "num_tokens": 36455732.0,
      "step": 10910
    },
    {
      "entropy": 6.319253253936767,
      "epoch": 2.129353233830846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004581467160330287,
      "loss": 6.013,
      "mean_token_accuracy": 0.14840034693479537,
      "num_tokens": 36471220.0,
      "step": 10915
    },
    {
      "entropy": 6.1324005126953125,
      "epoch": 2.130328748414789,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004581058510018053,
      "loss": 5.835,
      "mean_token_accuracy": 0.15908453464508057,
      "num_tokens": 36488408.0,
      "step": 10920
    },
    {
      "entropy": 6.2738440990447994,
      "epoch": 2.1313042629987318,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004580649680781307,
      "loss": 5.9578,
      "mean_token_accuracy": 0.15111517757177353,
      "num_tokens": 36504785.0,
      "step": 10925
    },
    {
      "entropy": 6.339068841934204,
      "epoch": 2.1322797775826747,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045802406726599997,
      "loss": 6.1808,
      "mean_token_accuracy": 0.1336050473153591,
      "num_tokens": 36521880.0,
      "step": 10930
    },
    {
      "entropy": 6.250499486923218,
      "epoch": 2.1332552921666177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045798314856940956,
      "loss": 5.9692,
      "mean_token_accuracy": 0.14363300129771234,
      "num_tokens": 36537075.0,
      "step": 10935
    },
    {
      "entropy": 6.263360500335693,
      "epoch": 2.134230806750561,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045794221199235827,
      "loss": 6.0179,
      "mean_token_accuracy": 0.15232848525047302,
      "num_tokens": 36554244.0,
      "step": 10940
    },
    {
      "entropy": 6.15951886177063,
      "epoch": 2.135206321334504,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045790125753884575,
      "loss": 5.91,
      "mean_token_accuracy": 0.16019859686493873,
      "num_tokens": 36571900.0,
      "step": 10945
    },
    {
      "entropy": 6.246325588226318,
      "epoch": 2.136181835918447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004578602852128744,
      "loss": 5.9774,
      "mean_token_accuracy": 0.14971997141838073,
      "num_tokens": 36588420.0,
      "step": 10950
    },
    {
      "entropy": 6.321615886688233,
      "epoch": 2.13715735050239,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045781929501844765,
      "loss": 6.1605,
      "mean_token_accuracy": 0.1463972419500351,
      "num_tokens": 36605371.0,
      "step": 10955
    },
    {
      "entropy": 6.1381847858428955,
      "epoch": 2.138132865086333,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045777828695957095,
      "loss": 5.9241,
      "mean_token_accuracy": 0.16474867165088652,
      "num_tokens": 36620665.0,
      "step": 10960
    },
    {
      "entropy": 6.208252096176148,
      "epoch": 2.139108379670276,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045773726104025153,
      "loss": 6.0187,
      "mean_token_accuracy": 0.14822878390550615,
      "num_tokens": 36637067.0,
      "step": 10965
    },
    {
      "entropy": 6.3570576190948485,
      "epoch": 2.1400838942542193,
      "grad_norm": 1.0,
      "learning_rate": 0.0004576962172644982,
      "loss": 6.0675,
      "mean_token_accuracy": 0.1408511221408844,
      "num_tokens": 36652952.0,
      "step": 10970
    },
    {
      "entropy": 6.30081353187561,
      "epoch": 2.1410594088381623,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004576551556363217,
      "loss": 6.0255,
      "mean_token_accuracy": 0.1462226152420044,
      "num_tokens": 36668275.0,
      "step": 10975
    },
    {
      "entropy": 6.256220102310181,
      "epoch": 2.142034923422105,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045761407615973446,
      "loss": 6.0885,
      "mean_token_accuracy": 0.1478171706199646,
      "num_tokens": 36684718.0,
      "step": 10980
    },
    {
      "entropy": 6.289742422103882,
      "epoch": 2.143010438006048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575729788387504,
      "loss": 6.0292,
      "mean_token_accuracy": 0.14095955938100815,
      "num_tokens": 36700259.0,
      "step": 10985
    },
    {
      "entropy": 6.283935260772705,
      "epoch": 2.143985952589991,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045753186367738564,
      "loss": 6.0679,
      "mean_token_accuracy": 0.15009886175394058,
      "num_tokens": 36716949.0,
      "step": 10990
    },
    {
      "entropy": 6.382436084747314,
      "epoch": 2.144961467173934,
      "grad_norm": 1.0,
      "learning_rate": 0.0004574907306796577,
      "loss": 6.1049,
      "mean_token_accuracy": 0.14010793045163156,
      "num_tokens": 36732927.0,
      "step": 10995
    },
    {
      "entropy": 6.276382303237915,
      "epoch": 2.1459369817578775,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004574495798495859,
      "loss": 6.1033,
      "mean_token_accuracy": 0.14093181416392325,
      "num_tokens": 36748841.0,
      "step": 11000
    },
    {
      "entropy": 6.2993786334991455,
      "epoch": 2.1469124963418205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045740841119119147,
      "loss": 6.1096,
      "mean_token_accuracy": 0.14778874516487123,
      "num_tokens": 36766048.0,
      "step": 11005
    },
    {
      "entropy": 6.222366571426392,
      "epoch": 2.1478880109257634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004573672247084972,
      "loss": 6.035,
      "mean_token_accuracy": 0.1539136826992035,
      "num_tokens": 36782471.0,
      "step": 11010
    },
    {
      "entropy": 6.2309516906738285,
      "epoch": 2.1488635255097064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004573260204055277,
      "loss": 5.9358,
      "mean_token_accuracy": 0.16051677614450455,
      "num_tokens": 36797789.0,
      "step": 11015
    },
    {
      "entropy": 6.159729623794556,
      "epoch": 2.1498390400936493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004572847982863094,
      "loss": 5.9722,
      "mean_token_accuracy": 0.15461728870868682,
      "num_tokens": 36814466.0,
      "step": 11020
    },
    {
      "entropy": 6.167754030227661,
      "epoch": 2.1508145546775923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004572435583548702,
      "loss": 5.9326,
      "mean_token_accuracy": 0.1572544127702713,
      "num_tokens": 36831579.0,
      "step": 11025
    },
    {
      "entropy": 6.397600841522217,
      "epoch": 2.1517900692615353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045720230061524,
      "loss": 6.1645,
      "mean_token_accuracy": 0.1374441161751747,
      "num_tokens": 36847260.0,
      "step": 11030
    },
    {
      "entropy": 6.3738878726959225,
      "epoch": 2.1527655838454787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571610250714504,
      "loss": 6.1653,
      "mean_token_accuracy": 0.13669291883707047,
      "num_tokens": 36865380.0,
      "step": 11035
    },
    {
      "entropy": 6.346575307846069,
      "epoch": 2.1537410984294216,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004571197317275346,
      "loss": 6.1078,
      "mean_token_accuracy": 0.14707719683647155,
      "num_tokens": 36883410.0,
      "step": 11040
    },
    {
      "entropy": 6.310169076919555,
      "epoch": 2.1547166130133646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004570784205875278,
      "loss": 6.1428,
      "mean_token_accuracy": 0.13469286113977433,
      "num_tokens": 36900270.0,
      "step": 11045
    },
    {
      "entropy": 6.232250356674195,
      "epoch": 2.1556921275973076,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004570370916554665,
      "loss": 6.0149,
      "mean_token_accuracy": 0.1487743616104126,
      "num_tokens": 36916076.0,
      "step": 11050
    },
    {
      "entropy": 6.361471176147461,
      "epoch": 2.1566676421812505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045699574493538966,
      "loss": 6.1272,
      "mean_token_accuracy": 0.1434021033346653,
      "num_tokens": 36931966.0,
      "step": 11055
    },
    {
      "entropy": 6.296859169006348,
      "epoch": 2.1576431567651935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045695438043133707,
      "loss": 6.0553,
      "mean_token_accuracy": 0.1482545480132103,
      "num_tokens": 36946615.0,
      "step": 11060
    },
    {
      "entropy": 6.269865989685059,
      "epoch": 2.158618671349137,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000456912998147351,
      "loss": 6.0094,
      "mean_token_accuracy": 0.14847919195890427,
      "num_tokens": 36964830.0,
      "step": 11065
    },
    {
      "entropy": 6.346645736694336,
      "epoch": 2.15959418593308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004568715980874751,
      "loss": 6.1185,
      "mean_token_accuracy": 0.13933421671390533,
      "num_tokens": 36982248.0,
      "step": 11070
    },
    {
      "entropy": 6.2892554759979244,
      "epoch": 2.160569700517023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004568301802557548,
      "loss": 6.0287,
      "mean_token_accuracy": 0.15040020495653153,
      "num_tokens": 36998543.0,
      "step": 11075
    },
    {
      "entropy": 6.292067146301269,
      "epoch": 2.1615452151009658,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004567887446562373,
      "loss": 6.0383,
      "mean_token_accuracy": 0.14460055977106095,
      "num_tokens": 37014385.0,
      "step": 11080
    },
    {
      "entropy": 6.329105949401855,
      "epoch": 2.1625207296849087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045674729129297165,
      "loss": 6.0962,
      "mean_token_accuracy": 0.1394900806248188,
      "num_tokens": 37030562.0,
      "step": 11085
    },
    {
      "entropy": 6.282892799377441,
      "epoch": 2.1634962442688517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045670582017000847,
      "loss": 6.0262,
      "mean_token_accuracy": 0.14748286381363868,
      "num_tokens": 37045618.0,
      "step": 11090
    },
    {
      "entropy": 6.276543521881104,
      "epoch": 2.1644717588527946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004566643312914001,
      "loss": 6.079,
      "mean_token_accuracy": 0.14771506637334825,
      "num_tokens": 37061356.0,
      "step": 11095
    },
    {
      "entropy": 6.253771686553955,
      "epoch": 2.165447273436738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045662282466120077,
      "loss": 5.992,
      "mean_token_accuracy": 0.15310063585639,
      "num_tokens": 37078250.0,
      "step": 11100
    },
    {
      "entropy": 6.353211069107056,
      "epoch": 2.166422788020681,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004565813002834663,
      "loss": 6.1233,
      "mean_token_accuracy": 0.13640418276190758,
      "num_tokens": 37093774.0,
      "step": 11105
    },
    {
      "entropy": 6.3375030040740965,
      "epoch": 2.167398302604624,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045653975816225433,
      "loss": 6.1271,
      "mean_token_accuracy": 0.13971150144934655,
      "num_tokens": 37109720.0,
      "step": 11110
    },
    {
      "entropy": 6.099762773513794,
      "epoch": 2.168373817188567,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004564981983016242,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1638236165046692,
      "num_tokens": 37126872.0,
      "step": 11115
    },
    {
      "entropy": 6.294187402725219,
      "epoch": 2.16934933177251,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004564566207056371,
      "loss": 6.0631,
      "mean_token_accuracy": 0.1452108584344387,
      "num_tokens": 37144227.0,
      "step": 11120
    },
    {
      "entropy": 6.387281179428101,
      "epoch": 2.170324846356453,
      "grad_norm": 1.0,
      "learning_rate": 0.00045641502537835567,
      "loss": 6.1586,
      "mean_token_accuracy": 0.13880186825990676,
      "num_tokens": 37161419.0,
      "step": 11125
    },
    {
      "entropy": 6.308630847930909,
      "epoch": 2.1713003609403962,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045637341232384454,
      "loss": 6.0479,
      "mean_token_accuracy": 0.14609692096710206,
      "num_tokens": 37177070.0,
      "step": 11130
    },
    {
      "entropy": 6.239846181869507,
      "epoch": 2.172275875524339,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045633178154617,
      "loss": 5.9381,
      "mean_token_accuracy": 0.1506247952580452,
      "num_tokens": 37193910.0,
      "step": 11135
    },
    {
      "entropy": 6.25131778717041,
      "epoch": 2.173251390108282,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045629013304940004,
      "loss": 6.1298,
      "mean_token_accuracy": 0.13882508724927903,
      "num_tokens": 37210636.0,
      "step": 11140
    },
    {
      "entropy": 6.327555322647095,
      "epoch": 2.174226904692225,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004562484668376044,
      "loss": 6.0598,
      "mean_token_accuracy": 0.1471230834722519,
      "num_tokens": 37229194.0,
      "step": 11145
    },
    {
      "entropy": 6.390720367431641,
      "epoch": 2.175202419276168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045620678291485454,
      "loss": 6.1384,
      "mean_token_accuracy": 0.14192619398236275,
      "num_tokens": 37245409.0,
      "step": 11150
    },
    {
      "entropy": 6.319381046295166,
      "epoch": 2.176177933860111,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045616508128522374,
      "loss": 6.1532,
      "mean_token_accuracy": 0.13709580823779105,
      "num_tokens": 37260892.0,
      "step": 11155
    },
    {
      "entropy": 6.28315052986145,
      "epoch": 2.177153448444054,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045612336195278687,
      "loss": 5.9454,
      "mean_token_accuracy": 0.15261341109871865,
      "num_tokens": 37277549.0,
      "step": 11160
    },
    {
      "entropy": 6.2526084899902346,
      "epoch": 2.1781289630279974,
      "grad_norm": 1.0,
      "learning_rate": 0.0004560816249216206,
      "loss": 5.9734,
      "mean_token_accuracy": 0.15466092228889466,
      "num_tokens": 37295169.0,
      "step": 11165
    },
    {
      "entropy": 6.208287572860717,
      "epoch": 2.1791044776119404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045603987019580326,
      "loss": 5.9765,
      "mean_token_accuracy": 0.14528486058115958,
      "num_tokens": 37311521.0,
      "step": 11170
    },
    {
      "entropy": 6.27390456199646,
      "epoch": 2.1800799921958833,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004559980977794152,
      "loss": 6.0885,
      "mean_token_accuracy": 0.14239205569028854,
      "num_tokens": 37328294.0,
      "step": 11175
    },
    {
      "entropy": 6.2034379005432125,
      "epoch": 2.1810555067798263,
      "grad_norm": 1.171875,
      "learning_rate": 0.000455956307676538,
      "loss": 5.9597,
      "mean_token_accuracy": 0.15514098331332207,
      "num_tokens": 37344818.0,
      "step": 11180
    },
    {
      "entropy": 6.426271438598633,
      "epoch": 2.1820310213637693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045591449989125526,
      "loss": 6.203,
      "mean_token_accuracy": 0.13491289392113687,
      "num_tokens": 37361863.0,
      "step": 11185
    },
    {
      "entropy": 6.327885866165161,
      "epoch": 2.183006535947712,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004558726744276526,
      "loss": 6.1083,
      "mean_token_accuracy": 0.14299367219209672,
      "num_tokens": 37377577.0,
      "step": 11190
    },
    {
      "entropy": 6.215516090393066,
      "epoch": 2.1839820505316556,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045583083128981666,
      "loss": 6.0252,
      "mean_token_accuracy": 0.15314200147986412,
      "num_tokens": 37397204.0,
      "step": 11195
    },
    {
      "entropy": 6.334845876693725,
      "epoch": 2.1849575651155986,
      "grad_norm": 1.0,
      "learning_rate": 0.0004557889704818365,
      "loss": 6.1058,
      "mean_token_accuracy": 0.1395121321082115,
      "num_tokens": 37414053.0,
      "step": 11200
    },
    {
      "entropy": 6.321520233154297,
      "epoch": 2.1859330796995415,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045574709200780236,
      "loss": 6.1244,
      "mean_token_accuracy": 0.14247528612613677,
      "num_tokens": 37431389.0,
      "step": 11205
    },
    {
      "entropy": 6.355411577224731,
      "epoch": 2.1869085942834845,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045570519587180666,
      "loss": 6.0588,
      "mean_token_accuracy": 0.14684050232172013,
      "num_tokens": 37448330.0,
      "step": 11210
    },
    {
      "entropy": 6.358974647521973,
      "epoch": 2.1878841088674275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004556632820779433,
      "loss": 6.2003,
      "mean_token_accuracy": 0.1348964162170887,
      "num_tokens": 37465149.0,
      "step": 11215
    },
    {
      "entropy": 6.332659912109375,
      "epoch": 2.1888596234513704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045562135063030785,
      "loss": 6.0902,
      "mean_token_accuracy": 0.1450905680656433,
      "num_tokens": 37481252.0,
      "step": 11220
    },
    {
      "entropy": 6.283927774429321,
      "epoch": 2.189835138035314,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045557940153299783,
      "loss": 6.0776,
      "mean_token_accuracy": 0.14235224053263665,
      "num_tokens": 37498021.0,
      "step": 11225
    },
    {
      "entropy": 6.371999788284302,
      "epoch": 2.190810652619257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004555374347901122,
      "loss": 6.0965,
      "mean_token_accuracy": 0.151792211830616,
      "num_tokens": 37514715.0,
      "step": 11230
    },
    {
      "entropy": 6.303356885910034,
      "epoch": 2.1917861672031997,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004554954504057519,
      "loss": 6.1487,
      "mean_token_accuracy": 0.1356409214437008,
      "num_tokens": 37531069.0,
      "step": 11235
    },
    {
      "entropy": 6.205270671844483,
      "epoch": 2.1927616817871427,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004554534483840195,
      "loss": 5.9847,
      "mean_token_accuracy": 0.1535641521215439,
      "num_tokens": 37548434.0,
      "step": 11240
    },
    {
      "entropy": 6.15211329460144,
      "epoch": 2.1937371963710857,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004554114287290194,
      "loss": 5.881,
      "mean_token_accuracy": 0.1538038395345211,
      "num_tokens": 37564233.0,
      "step": 11245
    },
    {
      "entropy": 6.405444717407226,
      "epoch": 2.1947127109550286,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045536939144485733,
      "loss": 6.1377,
      "mean_token_accuracy": 0.13965703845024108,
      "num_tokens": 37579356.0,
      "step": 11250
    },
    {
      "entropy": 6.252398920059204,
      "epoch": 2.195688225538972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004553273365356412,
      "loss": 6.0781,
      "mean_token_accuracy": 0.13738964945077897,
      "num_tokens": 37595458.0,
      "step": 11255
    },
    {
      "entropy": 6.3251337051391605,
      "epoch": 2.196663740122915,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045528526400548044,
      "loss": 6.1123,
      "mean_token_accuracy": 0.14555007293820382,
      "num_tokens": 37613421.0,
      "step": 11260
    },
    {
      "entropy": 6.127777338027954,
      "epoch": 2.197639254706858,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045524317385848623,
      "loss": 5.8098,
      "mean_token_accuracy": 0.16558764427900313,
      "num_tokens": 37630610.0,
      "step": 11265
    },
    {
      "entropy": 6.20623140335083,
      "epoch": 2.198614769290801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045520106609877144,
      "loss": 5.9722,
      "mean_token_accuracy": 0.14371296763420105,
      "num_tokens": 37646520.0,
      "step": 11270
    },
    {
      "entropy": 6.200391149520874,
      "epoch": 2.199590283874744,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004551589407304508,
      "loss": 6.0195,
      "mean_token_accuracy": 0.14713364019989966,
      "num_tokens": 37662874.0,
      "step": 11275
    },
    {
      "entropy": 6.19166145324707,
      "epoch": 2.200565798458687,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045511679775764053,
      "loss": 5.9688,
      "mean_token_accuracy": 0.16136980056762695,
      "num_tokens": 37679745.0,
      "step": 11280
    },
    {
      "entropy": 6.360001468658448,
      "epoch": 2.20154131304263,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045507463718445876,
      "loss": 6.1023,
      "mean_token_accuracy": 0.14563947021961213,
      "num_tokens": 37697681.0,
      "step": 11285
    },
    {
      "entropy": 6.363406038284301,
      "epoch": 2.202516827626573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045503245901502517,
      "loss": 6.1717,
      "mean_token_accuracy": 0.13791734725236893,
      "num_tokens": 37714333.0,
      "step": 11290
    },
    {
      "entropy": 6.354908752441406,
      "epoch": 2.203492342210516,
      "grad_norm": 1.0,
      "learning_rate": 0.0004549902632534614,
      "loss": 6.1091,
      "mean_token_accuracy": 0.14059773236513137,
      "num_tokens": 37730439.0,
      "step": 11295
    },
    {
      "entropy": 6.232324314117432,
      "epoch": 2.204467856794459,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045494804990389053,
      "loss": 5.9701,
      "mean_token_accuracy": 0.14659513533115387,
      "num_tokens": 37746836.0,
      "step": 11300
    },
    {
      "entropy": 6.265915298461914,
      "epoch": 2.205443371378402,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045490581897043766,
      "loss": 6.0197,
      "mean_token_accuracy": 0.14496007412672043,
      "num_tokens": 37763755.0,
      "step": 11305
    },
    {
      "entropy": 6.3625816822052,
      "epoch": 2.206418885962345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004548635704572293,
      "loss": 6.1416,
      "mean_token_accuracy": 0.13809728547930716,
      "num_tokens": 37780065.0,
      "step": 11310
    },
    {
      "entropy": 6.316467523574829,
      "epoch": 2.207394400546288,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004548213043683939,
      "loss": 6.0978,
      "mean_token_accuracy": 0.1409959688782692,
      "num_tokens": 37796630.0,
      "step": 11315
    },
    {
      "entropy": 6.3641551494598385,
      "epoch": 2.2083699151302314,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045477902070806156,
      "loss": 6.1266,
      "mean_token_accuracy": 0.13916173875331878,
      "num_tokens": 37812257.0,
      "step": 11320
    },
    {
      "entropy": 6.278663396835327,
      "epoch": 2.2093454297141744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000454736719480364,
      "loss": 6.0313,
      "mean_token_accuracy": 0.1448860928416252,
      "num_tokens": 37829702.0,
      "step": 11325
    },
    {
      "entropy": 6.288736820220947,
      "epoch": 2.2103209442981173,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004546944006894348,
      "loss": 6.1155,
      "mean_token_accuracy": 0.13996743112802507,
      "num_tokens": 37846335.0,
      "step": 11330
    },
    {
      "entropy": 6.298521947860718,
      "epoch": 2.2112964588820603,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004546520643394093,
      "loss": 5.9915,
      "mean_token_accuracy": 0.14569065645337104,
      "num_tokens": 37863341.0,
      "step": 11335
    },
    {
      "entropy": 6.236401510238648,
      "epoch": 2.2122719734660032,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004546097104344242,
      "loss": 6.0591,
      "mean_token_accuracy": 0.15536320731043815,
      "num_tokens": 37879277.0,
      "step": 11340
    },
    {
      "entropy": 6.251049947738648,
      "epoch": 2.213247488049946,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004545673389786185,
      "loss": 6.0137,
      "mean_token_accuracy": 0.14558098763227462,
      "num_tokens": 37895719.0,
      "step": 11345
    },
    {
      "entropy": 6.212751960754394,
      "epoch": 2.214223002633889,
      "grad_norm": 1.0,
      "learning_rate": 0.00045452494997613236,
      "loss": 5.961,
      "mean_token_accuracy": 0.1488008677959442,
      "num_tokens": 37912799.0,
      "step": 11350
    },
    {
      "entropy": 6.31485595703125,
      "epoch": 2.2151985172178326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000454482543431108,
      "loss": 6.0587,
      "mean_token_accuracy": 0.1408642202615738,
      "num_tokens": 37930875.0,
      "step": 11355
    },
    {
      "entropy": 6.175679922103882,
      "epoch": 2.2161740318017755,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004544401193476891,
      "loss": 6.0241,
      "mean_token_accuracy": 0.15754064843058585,
      "num_tokens": 37948545.0,
      "step": 11360
    },
    {
      "entropy": 6.307265758514404,
      "epoch": 2.2171495463857185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543976777300213,
      "loss": 5.9782,
      "mean_token_accuracy": 0.15520357191562653,
      "num_tokens": 37965967.0,
      "step": 11365
    },
    {
      "entropy": 6.0906885147094725,
      "epoch": 2.2181250609696614,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045435521858225184,
      "loss": 5.8681,
      "mean_token_accuracy": 0.16699595004320145,
      "num_tokens": 37984285.0,
      "step": 11370
    },
    {
      "entropy": 6.380165767669678,
      "epoch": 2.2191005755536044,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543127419085296,
      "loss": 6.1446,
      "mean_token_accuracy": 0.13776771873235702,
      "num_tokens": 38001022.0,
      "step": 11375
    },
    {
      "entropy": 6.268902730941773,
      "epoch": 2.2200760901375474,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00045427024771300527,
      "loss": 6.0766,
      "mean_token_accuracy": 0.1495864786207676,
      "num_tokens": 38017724.0,
      "step": 11380
    },
    {
      "entropy": 6.275279903411866,
      "epoch": 2.2210516047214908,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045422773599983137,
      "loss": 6.0951,
      "mean_token_accuracy": 0.14188117906451225,
      "num_tokens": 38033873.0,
      "step": 11385
    },
    {
      "entropy": 6.316410255432129,
      "epoch": 2.2220271193054337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045418520677316187,
      "loss": 6.0344,
      "mean_token_accuracy": 0.1489210844039917,
      "num_tokens": 38049705.0,
      "step": 11390
    },
    {
      "entropy": 6.291846036911011,
      "epoch": 2.2230026338893767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004541426600371526,
      "loss": 6.0874,
      "mean_token_accuracy": 0.14294905662536622,
      "num_tokens": 38066756.0,
      "step": 11395
    },
    {
      "entropy": 6.408981609344482,
      "epoch": 2.2239781484733196,
      "grad_norm": 1.125,
      "learning_rate": 0.000454100095795961,
      "loss": 6.1834,
      "mean_token_accuracy": 0.13673478364944458,
      "num_tokens": 38083103.0,
      "step": 11400
    },
    {
      "entropy": 6.2666895389556885,
      "epoch": 2.2249536630572626,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004540575140537464,
      "loss": 5.9956,
      "mean_token_accuracy": 0.1507670447230339,
      "num_tokens": 38100149.0,
      "step": 11405
    },
    {
      "entropy": 6.232724857330322,
      "epoch": 2.2259291776412056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045401491481466974,
      "loss": 5.996,
      "mean_token_accuracy": 0.15326189398765563,
      "num_tokens": 38118295.0,
      "step": 11410
    },
    {
      "entropy": 6.30423150062561,
      "epoch": 2.226904692225149,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004539722980828935,
      "loss": 6.0496,
      "mean_token_accuracy": 0.1431081138551235,
      "num_tokens": 38134861.0,
      "step": 11415
    },
    {
      "entropy": 6.326748943328857,
      "epoch": 2.227880206809092,
      "grad_norm": 1.0,
      "learning_rate": 0.00045392966386258225,
      "loss": 6.0938,
      "mean_token_accuracy": 0.14336548298597335,
      "num_tokens": 38151070.0,
      "step": 11420
    },
    {
      "entropy": 6.252478790283203,
      "epoch": 2.228855721393035,
      "grad_norm": 0.953125,
      "learning_rate": 0.000453887012157902,
      "loss": 6.0319,
      "mean_token_accuracy": 0.14819028228521347,
      "num_tokens": 38168939.0,
      "step": 11425
    },
    {
      "entropy": 6.25355076789856,
      "epoch": 2.229831235976978,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045384434297302044,
      "loss": 6.0793,
      "mean_token_accuracy": 0.1443414606153965,
      "num_tokens": 38185779.0,
      "step": 11430
    },
    {
      "entropy": 6.360746145248413,
      "epoch": 2.230806750560921,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045380165631210713,
      "loss": 6.115,
      "mean_token_accuracy": 0.1432499572634697,
      "num_tokens": 38203415.0,
      "step": 11435
    },
    {
      "entropy": 6.360417747497559,
      "epoch": 2.2317822651448638,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004537589521793332,
      "loss": 6.0633,
      "mean_token_accuracy": 0.13918056041002275,
      "num_tokens": 38219953.0,
      "step": 11440
    },
    {
      "entropy": 6.33971939086914,
      "epoch": 2.232757779728807,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004537162305788716,
      "loss": 6.154,
      "mean_token_accuracy": 0.13568906337022782,
      "num_tokens": 38236258.0,
      "step": 11445
    },
    {
      "entropy": 6.387169647216797,
      "epoch": 2.23373329431275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045367349151489685,
      "loss": 6.1474,
      "mean_token_accuracy": 0.13952315524220466,
      "num_tokens": 38252878.0,
      "step": 11450
    },
    {
      "entropy": 6.21766471862793,
      "epoch": 2.234708808896693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004536307349915853,
      "loss": 5.9625,
      "mean_token_accuracy": 0.15926968902349473,
      "num_tokens": 38270215.0,
      "step": 11455
    },
    {
      "entropy": 6.166777467727661,
      "epoch": 2.235684323480636,
      "grad_norm": 0.96875,
      "learning_rate": 0.000453587961013115,
      "loss": 6.0026,
      "mean_token_accuracy": 0.14995893090963364,
      "num_tokens": 38287970.0,
      "step": 11460
    },
    {
      "entropy": 6.228988218307495,
      "epoch": 2.236659838064579,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045354516958366565,
      "loss": 5.9876,
      "mean_token_accuracy": 0.15570079535245895,
      "num_tokens": 38305345.0,
      "step": 11465
    },
    {
      "entropy": 6.239770889282227,
      "epoch": 2.237635352648522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004535023607074187,
      "loss": 6.0065,
      "mean_token_accuracy": 0.15097958520054816,
      "num_tokens": 38322410.0,
      "step": 11470
    },
    {
      "entropy": 6.257023000717163,
      "epoch": 2.238610867232465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045345953438855725,
      "loss": 6.0288,
      "mean_token_accuracy": 0.14757018610835076,
      "num_tokens": 38339708.0,
      "step": 11475
    },
    {
      "entropy": 6.35485053062439,
      "epoch": 2.2395863818164083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004534166906312661,
      "loss": 6.0973,
      "mean_token_accuracy": 0.14308699667453767,
      "num_tokens": 38356776.0,
      "step": 11480
    },
    {
      "entropy": 6.286906337738037,
      "epoch": 2.2405618964003513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004533738294397319,
      "loss": 6.0292,
      "mean_token_accuracy": 0.14716243967413903,
      "num_tokens": 38372916.0,
      "step": 11485
    },
    {
      "entropy": 6.210446453094482,
      "epoch": 2.2415374109842943,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045333095081814275,
      "loss": 6.0449,
      "mean_token_accuracy": 0.14971258789300917,
      "num_tokens": 38389401.0,
      "step": 11490
    },
    {
      "entropy": 6.320865058898926,
      "epoch": 2.2425129255682372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045328805477068873,
      "loss": 6.0729,
      "mean_token_accuracy": 0.14771794229745866,
      "num_tokens": 38404985.0,
      "step": 11495
    },
    {
      "entropy": 6.296080827713013,
      "epoch": 2.24348844015218,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004532451413015614,
      "loss": 6.0681,
      "mean_token_accuracy": 0.14591746032238007,
      "num_tokens": 38420748.0,
      "step": 11500
    },
    {
      "entropy": 6.35866756439209,
      "epoch": 2.244463954736123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004532022104149541,
      "loss": 6.125,
      "mean_token_accuracy": 0.14224493354558945,
      "num_tokens": 38437043.0,
      "step": 11505
    },
    {
      "entropy": 6.3126648426055905,
      "epoch": 2.2454394693200666,
      "grad_norm": 0.984375,
      "learning_rate": 0.000453159262115062,
      "loss": 6.1374,
      "mean_token_accuracy": 0.14544578045606613,
      "num_tokens": 38453469.0,
      "step": 11510
    },
    {
      "entropy": 6.411281776428223,
      "epoch": 2.2464149839040095,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004531162964060817,
      "loss": 6.1427,
      "mean_token_accuracy": 0.1385834999382496,
      "num_tokens": 38469162.0,
      "step": 11515
    },
    {
      "entropy": 6.383779811859131,
      "epoch": 2.2473904984879525,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004530733132922118,
      "loss": 6.1773,
      "mean_token_accuracy": 0.13614313974976539,
      "num_tokens": 38485425.0,
      "step": 11520
    },
    {
      "entropy": 6.068434476852417,
      "epoch": 2.2483660130718954,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045303031277765235,
      "loss": 5.8017,
      "mean_token_accuracy": 0.17115475237369537,
      "num_tokens": 38503437.0,
      "step": 11525
    },
    {
      "entropy": 6.2870275497436525,
      "epoch": 2.2493415276558384,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004529872948666052,
      "loss": 6.0458,
      "mean_token_accuracy": 0.14632475152611732,
      "num_tokens": 38520206.0,
      "step": 11530
    },
    {
      "entropy": 6.349869966506958,
      "epoch": 2.2503170422397814,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004529442595632741,
      "loss": 6.1183,
      "mean_token_accuracy": 0.14304069727659224,
      "num_tokens": 38536717.0,
      "step": 11535
    },
    {
      "entropy": 6.2541828632354735,
      "epoch": 2.2512925568237243,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045290120687186406,
      "loss": 6.0683,
      "mean_token_accuracy": 0.14719306975603103,
      "num_tokens": 38553930.0,
      "step": 11540
    },
    {
      "entropy": 6.224720287322998,
      "epoch": 2.2522680714076677,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004528581367965821,
      "loss": 5.9286,
      "mean_token_accuracy": 0.1593418002128601,
      "num_tokens": 38569798.0,
      "step": 11545
    },
    {
      "entropy": 6.21540641784668,
      "epoch": 2.2532435859916107,
      "grad_norm": 1.125,
      "learning_rate": 0.000452815049341637,
      "loss": 5.9697,
      "mean_token_accuracy": 0.147123284637928,
      "num_tokens": 38587878.0,
      "step": 11550
    },
    {
      "entropy": 6.23087067604065,
      "epoch": 2.2542191005755536,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000452771944511239,
      "loss": 5.9263,
      "mean_token_accuracy": 0.14912570416927337,
      "num_tokens": 38604601.0,
      "step": 11555
    },
    {
      "entropy": 6.242258024215698,
      "epoch": 2.2551946151594966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004527288223096001,
      "loss": 5.975,
      "mean_token_accuracy": 0.1471011683344841,
      "num_tokens": 38621086.0,
      "step": 11560
    },
    {
      "entropy": 6.132077646255493,
      "epoch": 2.2561701297434396,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004526856827409342,
      "loss": 5.852,
      "mean_token_accuracy": 0.16281613409519197,
      "num_tokens": 38637078.0,
      "step": 11565
    },
    {
      "entropy": 6.181963682174683,
      "epoch": 2.2571456443273825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045264252580945664,
      "loss": 5.9635,
      "mean_token_accuracy": 0.15805716961622238,
      "num_tokens": 38653957.0,
      "step": 11570
    },
    {
      "entropy": 6.286596775054932,
      "epoch": 2.258121158911326,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045259935151938454,
      "loss": 6.0964,
      "mean_token_accuracy": 0.14351747408509255,
      "num_tokens": 38671985.0,
      "step": 11575
    },
    {
      "entropy": 6.3811592102050785,
      "epoch": 2.259096673495269,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004525561598749368,
      "loss": 6.1485,
      "mean_token_accuracy": 0.13995025604963302,
      "num_tokens": 38688412.0,
      "step": 11580
    },
    {
      "entropy": 6.303156137466431,
      "epoch": 2.260072188079212,
      "grad_norm": 0.9375,
      "learning_rate": 0.000452512950880334,
      "loss": 6.04,
      "mean_token_accuracy": 0.14304190874099731,
      "num_tokens": 38705442.0,
      "step": 11585
    },
    {
      "entropy": 6.3157226085662845,
      "epoch": 2.261047702663155,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004524697245397983,
      "loss": 6.145,
      "mean_token_accuracy": 0.1426028326153755,
      "num_tokens": 38722094.0,
      "step": 11590
    },
    {
      "entropy": 6.292375802993774,
      "epoch": 2.2620232172470978,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004524264808575536,
      "loss": 6.039,
      "mean_token_accuracy": 0.15411513298749924,
      "num_tokens": 38738287.0,
      "step": 11595
    },
    {
      "entropy": 6.311616563796997,
      "epoch": 2.2629987318310407,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045238321983782556,
      "loss": 6.0927,
      "mean_token_accuracy": 0.1429364114999771,
      "num_tokens": 38754374.0,
      "step": 11600
    },
    {
      "entropy": 6.315369558334351,
      "epoch": 2.2639742464149837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045233994148484154,
      "loss": 6.0257,
      "mean_token_accuracy": 0.14552948176860808,
      "num_tokens": 38770397.0,
      "step": 11605
    },
    {
      "entropy": 6.276001834869385,
      "epoch": 2.264949760998927,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045229664580283036,
      "loss": 6.0933,
      "mean_token_accuracy": 0.14299559071660042,
      "num_tokens": 38785657.0,
      "step": 11610
    },
    {
      "entropy": 6.357353258132934,
      "epoch": 2.26592527558287,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004522533327960229,
      "loss": 6.0898,
      "mean_token_accuracy": 0.13894077837467195,
      "num_tokens": 38801644.0,
      "step": 11615
    },
    {
      "entropy": 6.329026794433593,
      "epoch": 2.266900790166813,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045221000246865157,
      "loss": 6.1161,
      "mean_token_accuracy": 0.13947243839502335,
      "num_tokens": 38818406.0,
      "step": 11620
    },
    {
      "entropy": 6.195217084884644,
      "epoch": 2.267876304750756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004521666548249503,
      "loss": 6.0067,
      "mean_token_accuracy": 0.15231175422668458,
      "num_tokens": 38834805.0,
      "step": 11625
    },
    {
      "entropy": 6.430713176727295,
      "epoch": 2.268851819334699,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045212328986915496,
      "loss": 6.1714,
      "mean_token_accuracy": 0.14298241287469865,
      "num_tokens": 38850571.0,
      "step": 11630
    },
    {
      "entropy": 6.269105386734009,
      "epoch": 2.2698273339186423,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045207990760550305,
      "loss": 5.9862,
      "mean_token_accuracy": 0.15101485922932625,
      "num_tokens": 38867476.0,
      "step": 11635
    },
    {
      "entropy": 6.25333890914917,
      "epoch": 2.2708028485025853,
      "grad_norm": 1.0,
      "learning_rate": 0.0004520365080382337,
      "loss": 6.0903,
      "mean_token_accuracy": 0.14641242772340773,
      "num_tokens": 38882979.0,
      "step": 11640
    },
    {
      "entropy": 6.316976928710938,
      "epoch": 2.2717783630865283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045199309117158773,
      "loss": 5.9745,
      "mean_token_accuracy": 0.15034198388457298,
      "num_tokens": 38899585.0,
      "step": 11645
    },
    {
      "entropy": 6.289170026779175,
      "epoch": 2.272753877670471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004519496570098077,
      "loss": 6.1007,
      "mean_token_accuracy": 0.13961717411875724,
      "num_tokens": 38915707.0,
      "step": 11650
    },
    {
      "entropy": 6.216695117950439,
      "epoch": 2.273729392254414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004519062055571379,
      "loss": 6.0237,
      "mean_token_accuracy": 0.14519270583987237,
      "num_tokens": 38931608.0,
      "step": 11655
    },
    {
      "entropy": 6.29925651550293,
      "epoch": 2.274704906838357,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004518627368178241,
      "loss": 6.0364,
      "mean_token_accuracy": 0.1466996192932129,
      "num_tokens": 38948675.0,
      "step": 11660
    },
    {
      "entropy": 6.265527248382568,
      "epoch": 2.2756804214223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004518192507961141,
      "loss": 6.0273,
      "mean_token_accuracy": 0.15223909616470338,
      "num_tokens": 38964568.0,
      "step": 11665
    },
    {
      "entropy": 6.292103099822998,
      "epoch": 2.276655936006243,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045177574749625713,
      "loss": 6.0566,
      "mean_token_accuracy": 0.14907958209514618,
      "num_tokens": 38981192.0,
      "step": 11670
    },
    {
      "entropy": 6.310835266113282,
      "epoch": 2.2776314505901865,
      "grad_norm": 1.0,
      "learning_rate": 0.00045173222692250405,
      "loss": 6.0022,
      "mean_token_accuracy": 0.1461956188082695,
      "num_tokens": 38997160.0,
      "step": 11675
    },
    {
      "entropy": 6.275320816040039,
      "epoch": 2.2786069651741294,
      "grad_norm": 1.0,
      "learning_rate": 0.00045168868907910763,
      "loss": 6.0786,
      "mean_token_accuracy": 0.14483651667833328,
      "num_tokens": 39013834.0,
      "step": 11680
    },
    {
      "entropy": 6.343396997451782,
      "epoch": 2.2795824797580724,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004516451339703224,
      "loss": 6.0889,
      "mean_token_accuracy": 0.1403576634824276,
      "num_tokens": 39029935.0,
      "step": 11685
    },
    {
      "entropy": 6.326666069030762,
      "epoch": 2.2805579943420153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045160156160040403,
      "loss": 6.0927,
      "mean_token_accuracy": 0.14191546589136123,
      "num_tokens": 39048022.0,
      "step": 11690
    },
    {
      "entropy": 6.369234466552735,
      "epoch": 2.2815335089259583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004515579719736107,
      "loss": 6.1514,
      "mean_token_accuracy": 0.13678107485175134,
      "num_tokens": 39064036.0,
      "step": 11695
    },
    {
      "entropy": 6.237329006195068,
      "epoch": 2.2825090235099017,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004515143650942015,
      "loss": 5.9845,
      "mean_token_accuracy": 0.15949834734201432,
      "num_tokens": 39081607.0,
      "step": 11700
    },
    {
      "entropy": 6.289684247970581,
      "epoch": 2.2834845380938447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045147074096643766,
      "loss": 5.9716,
      "mean_token_accuracy": 0.15260938107967376,
      "num_tokens": 39098101.0,
      "step": 11705
    },
    {
      "entropy": 6.262957954406739,
      "epoch": 2.2844600526777876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045142709959458194,
      "loss": 6.0745,
      "mean_token_accuracy": 0.14840612784028054,
      "num_tokens": 39115381.0,
      "step": 11710
    },
    {
      "entropy": 6.3340685844421385,
      "epoch": 2.2854355672617306,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045138344098289893,
      "loss": 6.1082,
      "mean_token_accuracy": 0.1399632915854454,
      "num_tokens": 39131601.0,
      "step": 11715
    },
    {
      "entropy": 6.260099220275879,
      "epoch": 2.2864110818456735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045133976513565463,
      "loss": 5.9732,
      "mean_token_accuracy": 0.15308795124292374,
      "num_tokens": 39146551.0,
      "step": 11720
    },
    {
      "entropy": 6.263168239593506,
      "epoch": 2.2873865964296165,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045129607205711696,
      "loss": 5.9617,
      "mean_token_accuracy": 0.1479753226041794,
      "num_tokens": 39163519.0,
      "step": 11725
    },
    {
      "entropy": 6.22260217666626,
      "epoch": 2.2883621110135595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004512523617515555,
      "loss": 6.0369,
      "mean_token_accuracy": 0.1460757814347744,
      "num_tokens": 39179428.0,
      "step": 11730
    },
    {
      "entropy": 6.293737888336182,
      "epoch": 2.289337625597503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045120863422324145,
      "loss": 6.052,
      "mean_token_accuracy": 0.1497077889740467,
      "num_tokens": 39196384.0,
      "step": 11735
    },
    {
      "entropy": 6.252483177185058,
      "epoch": 2.290313140181446,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004511648894764476,
      "loss": 6.0327,
      "mean_token_accuracy": 0.14560268744826316,
      "num_tokens": 39213375.0,
      "step": 11740
    },
    {
      "entropy": 6.321463251113892,
      "epoch": 2.291288654765389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004511211275154487,
      "loss": 6.045,
      "mean_token_accuracy": 0.1452957883477211,
      "num_tokens": 39229697.0,
      "step": 11745
    },
    {
      "entropy": 6.237533664703369,
      "epoch": 2.2922641693493317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045107734834452085,
      "loss": 6.0108,
      "mean_token_accuracy": 0.14895022362470628,
      "num_tokens": 39245763.0,
      "step": 11750
    },
    {
      "entropy": 6.3335052013397215,
      "epoch": 2.2932396839332747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004510335519679421,
      "loss": 6.0819,
      "mean_token_accuracy": 0.13789229840040207,
      "num_tokens": 39262298.0,
      "step": 11755
    },
    {
      "entropy": 6.272576713562012,
      "epoch": 2.2942151985172177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004509897383899921,
      "loss": 5.9721,
      "mean_token_accuracy": 0.14985468685626985,
      "num_tokens": 39278532.0,
      "step": 11760
    },
    {
      "entropy": 6.275338220596313,
      "epoch": 2.295190713101161,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045094590761495207,
      "loss": 6.122,
      "mean_token_accuracy": 0.14044289216399192,
      "num_tokens": 39294542.0,
      "step": 11765
    },
    {
      "entropy": 6.285881185531617,
      "epoch": 2.296166227685104,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045090205964710495,
      "loss": 6.0504,
      "mean_token_accuracy": 0.14118274599313735,
      "num_tokens": 39311318.0,
      "step": 11770
    },
    {
      "entropy": 6.296628999710083,
      "epoch": 2.297141742269047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508581944907356,
      "loss": 6.0463,
      "mean_token_accuracy": 0.1493461847305298,
      "num_tokens": 39328633.0,
      "step": 11775
    },
    {
      "entropy": 6.348930883407593,
      "epoch": 2.29811725685299,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004508143121501302,
      "loss": 6.1211,
      "mean_token_accuracy": 0.14254983216524125,
      "num_tokens": 39345063.0,
      "step": 11780
    },
    {
      "entropy": 6.330511283874512,
      "epoch": 2.299092771436933,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045077041262957685,
      "loss": 6.0924,
      "mean_token_accuracy": 0.14612421467900277,
      "num_tokens": 39360163.0,
      "step": 11785
    },
    {
      "entropy": 6.192141389846801,
      "epoch": 2.300068286020876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004507264959333652,
      "loss": 5.9958,
      "mean_token_accuracy": 0.1549705021083355,
      "num_tokens": 39376888.0,
      "step": 11790
    },
    {
      "entropy": 6.3352285861969,
      "epoch": 2.301043800604819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045068256206578665,
      "loss": 6.0714,
      "mean_token_accuracy": 0.14495841935276985,
      "num_tokens": 39394544.0,
      "step": 11795
    },
    {
      "entropy": 6.254873275756836,
      "epoch": 2.3020193151887622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045063861103113433,
      "loss": 5.9762,
      "mean_token_accuracy": 0.14909483045339583,
      "num_tokens": 39410118.0,
      "step": 11800
    },
    {
      "entropy": 6.301969528198242,
      "epoch": 2.302994829772705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045059464283370295,
      "loss": 6.1213,
      "mean_token_accuracy": 0.13822447806596755,
      "num_tokens": 39427307.0,
      "step": 11805
    },
    {
      "entropy": 6.320710372924805,
      "epoch": 2.303970344356648,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004505506574777888,
      "loss": 6.1112,
      "mean_token_accuracy": 0.13844417557120323,
      "num_tokens": 39443400.0,
      "step": 11810
    },
    {
      "entropy": 6.231529617309571,
      "epoch": 2.304945858940591,
      "grad_norm": 1.0,
      "learning_rate": 0.00045050665496769015,
      "loss": 6.0547,
      "mean_token_accuracy": 0.14154181331396104,
      "num_tokens": 39459392.0,
      "step": 11815
    },
    {
      "entropy": 6.328775835037232,
      "epoch": 2.305921373524534,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004504626353077068,
      "loss": 6.125,
      "mean_token_accuracy": 0.13726758435368538,
      "num_tokens": 39475795.0,
      "step": 11820
    },
    {
      "entropy": 6.252225542068482,
      "epoch": 2.306896888108477,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004504185985021399,
      "loss": 6.0288,
      "mean_token_accuracy": 0.15273562222719192,
      "num_tokens": 39492883.0,
      "step": 11825
    },
    {
      "entropy": 6.290733289718628,
      "epoch": 2.3078724026924204,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045037454455529286,
      "loss": 6.0698,
      "mean_token_accuracy": 0.14614976942539215,
      "num_tokens": 39510834.0,
      "step": 11830
    },
    {
      "entropy": 6.322775363922119,
      "epoch": 2.3088479172763634,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045033047347147034,
      "loss": 6.1401,
      "mean_token_accuracy": 0.13964953124523163,
      "num_tokens": 39527052.0,
      "step": 11835
    },
    {
      "entropy": 6.278546237945557,
      "epoch": 2.3098234318603064,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045028638525497884,
      "loss": 6.0153,
      "mean_token_accuracy": 0.1487372487783432,
      "num_tokens": 39543694.0,
      "step": 11840
    },
    {
      "entropy": 6.227156019210815,
      "epoch": 2.3107989464442493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004502422799101265,
      "loss": 6.0127,
      "mean_token_accuracy": 0.15150878429412842,
      "num_tokens": 39561223.0,
      "step": 11845
    },
    {
      "entropy": 6.327408790588379,
      "epoch": 2.3117744610281923,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045019815744122316,
      "loss": 6.0615,
      "mean_token_accuracy": 0.14629149436950684,
      "num_tokens": 39577513.0,
      "step": 11850
    },
    {
      "entropy": 6.238372945785523,
      "epoch": 2.3127499756121352,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045015401785258034,
      "loss": 5.9986,
      "mean_token_accuracy": 0.14844683706760406,
      "num_tokens": 39595172.0,
      "step": 11855
    },
    {
      "entropy": 6.267140626907349,
      "epoch": 2.313725490196078,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045010986114851103,
      "loss": 6.049,
      "mean_token_accuracy": 0.15747468322515487,
      "num_tokens": 39611285.0,
      "step": 11860
    },
    {
      "entropy": 6.371673679351806,
      "epoch": 2.3147010047800216,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045006568733333026,
      "loss": 6.1507,
      "mean_token_accuracy": 0.13574086204171182,
      "num_tokens": 39628545.0,
      "step": 11865
    },
    {
      "entropy": 6.315887069702148,
      "epoch": 2.3156765193639646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004500214964113545,
      "loss": 6.0622,
      "mean_token_accuracy": 0.14393752068281174,
      "num_tokens": 39645868.0,
      "step": 11870
    },
    {
      "entropy": 6.28657374382019,
      "epoch": 2.3166520339479075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004499772883869018,
      "loss": 6.0712,
      "mean_token_accuracy": 0.14565057903528214,
      "num_tokens": 39662781.0,
      "step": 11875
    },
    {
      "entropy": 6.265350866317749,
      "epoch": 2.3176275485318505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004499330632642922,
      "loss": 6.0556,
      "mean_token_accuracy": 0.14959292039275168,
      "num_tokens": 39679675.0,
      "step": 11880
    },
    {
      "entropy": 6.306356763839721,
      "epoch": 2.3186030631157935,
      "grad_norm": 1.0,
      "learning_rate": 0.00044988882104784713,
      "loss": 6.1156,
      "mean_token_accuracy": 0.1402053602039814,
      "num_tokens": 39697602.0,
      "step": 11885
    },
    {
      "entropy": 6.306491708755493,
      "epoch": 2.319578577699737,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044984456174188974,
      "loss": 6.0759,
      "mean_token_accuracy": 0.13751645907759666,
      "num_tokens": 39713951.0,
      "step": 11890
    },
    {
      "entropy": 6.2470588207244875,
      "epoch": 2.32055409228368,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00044980028535074497,
      "loss": 6.0639,
      "mean_token_accuracy": 0.15039174556732177,
      "num_tokens": 39730648.0,
      "step": 11895
    },
    {
      "entropy": 6.240250110626221,
      "epoch": 2.3215296068676228,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044975599187873933,
      "loss": 5.9808,
      "mean_token_accuracy": 0.15490271747112275,
      "num_tokens": 39748228.0,
      "step": 11900
    },
    {
      "entropy": 6.25016508102417,
      "epoch": 2.3225051214515657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044971168133020096,
      "loss": 6.0034,
      "mean_token_accuracy": 0.1482518881559372,
      "num_tokens": 39765248.0,
      "step": 11905
    },
    {
      "entropy": 6.2859302997589115,
      "epoch": 2.3234806360355087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044966735370945973,
      "loss": 6.0426,
      "mean_token_accuracy": 0.13885195404291154,
      "num_tokens": 39781799.0,
      "step": 11910
    },
    {
      "entropy": 6.195252561569214,
      "epoch": 2.3244561506194517,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004496230090208473,
      "loss": 5.9742,
      "mean_token_accuracy": 0.1576461173593998,
      "num_tokens": 39798952.0,
      "step": 11915
    },
    {
      "entropy": 6.231085395812988,
      "epoch": 2.3254316652033946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004495786472686967,
      "loss": 6.0395,
      "mean_token_accuracy": 0.15358738079667092,
      "num_tokens": 39814923.0,
      "step": 11920
    },
    {
      "entropy": 6.292228746414184,
      "epoch": 2.3264071797873376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004495342684573431,
      "loss": 6.056,
      "mean_token_accuracy": 0.14196276813745498,
      "num_tokens": 39831629.0,
      "step": 11925
    },
    {
      "entropy": 6.252839088439941,
      "epoch": 2.327382694371281,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044948987259112264,
      "loss": 6.0036,
      "mean_token_accuracy": 0.14481675922870635,
      "num_tokens": 39848103.0,
      "step": 11930
    },
    {
      "entropy": 6.262899875640869,
      "epoch": 2.328358208955224,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044944545967437385,
      "loss": 6.081,
      "mean_token_accuracy": 0.14459359273314476,
      "num_tokens": 39864969.0,
      "step": 11935
    },
    {
      "entropy": 6.3197437763214115,
      "epoch": 2.329333723539167,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044940102971143637,
      "loss": 6.0867,
      "mean_token_accuracy": 0.1472649022936821,
      "num_tokens": 39880636.0,
      "step": 11940
    },
    {
      "entropy": 6.244090127944946,
      "epoch": 2.33030923812311,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004493565827066519,
      "loss": 6.0425,
      "mean_token_accuracy": 0.1429576061666012,
      "num_tokens": 39898712.0,
      "step": 11945
    },
    {
      "entropy": 6.337475156784057,
      "epoch": 2.331284752707053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044931211866436354,
      "loss": 6.0807,
      "mean_token_accuracy": 0.14553862139582635,
      "num_tokens": 39914601.0,
      "step": 11950
    },
    {
      "entropy": 6.2209460735321045,
      "epoch": 2.3322602672909962,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004492676375889163,
      "loss": 6.0246,
      "mean_token_accuracy": 0.14761527329683305,
      "num_tokens": 39931564.0,
      "step": 11955
    },
    {
      "entropy": 6.251804304122925,
      "epoch": 2.333235781874939,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004492231394846565,
      "loss": 6.0336,
      "mean_token_accuracy": 0.14935895800590515,
      "num_tokens": 39949191.0,
      "step": 11960
    },
    {
      "entropy": 6.301523017883301,
      "epoch": 2.334211296458882,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004491786243559325,
      "loss": 6.0098,
      "mean_token_accuracy": 0.14402705952525138,
      "num_tokens": 39965560.0,
      "step": 11965
    },
    {
      "entropy": 6.1381035327911375,
      "epoch": 2.335186811042825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004491340922070942,
      "loss": 5.8552,
      "mean_token_accuracy": 0.1614916056394577,
      "num_tokens": 39982113.0,
      "step": 11970
    },
    {
      "entropy": 6.1707995414733885,
      "epoch": 2.336162325626768,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044908954304249287,
      "loss": 6.0452,
      "mean_token_accuracy": 0.1479676604270935,
      "num_tokens": 39998449.0,
      "step": 11975
    },
    {
      "entropy": 6.243093252182007,
      "epoch": 2.337137840210711,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044904497686648197,
      "loss": 6.0605,
      "mean_token_accuracy": 0.14558803141117097,
      "num_tokens": 40015419.0,
      "step": 11980
    },
    {
      "entropy": 6.349440574645996,
      "epoch": 2.338113354794654,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044900039368341623,
      "loss": 6.0721,
      "mean_token_accuracy": 0.14763350337743758,
      "num_tokens": 40030698.0,
      "step": 11985
    },
    {
      "entropy": 6.190534257888794,
      "epoch": 2.3390888693785974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044895579349765215,
      "loss": 5.9706,
      "mean_token_accuracy": 0.1513451397418976,
      "num_tokens": 40047385.0,
      "step": 11990
    },
    {
      "entropy": 6.323347187042236,
      "epoch": 2.3400643839625404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004489111763135479,
      "loss": 6.0721,
      "mean_token_accuracy": 0.14205363318324088,
      "num_tokens": 40062709.0,
      "step": 11995
    },
    {
      "entropy": 6.400413370132446,
      "epoch": 2.3410398985464833,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044886654213546333,
      "loss": 6.1256,
      "mean_token_accuracy": 0.13968910202383994,
      "num_tokens": 40079777.0,
      "step": 12000
    },
    {
      "epoch": 2.3410398985464833,
      "eval_entropy": 6.167364163727675,
      "eval_loss": 6.2119975090026855,
      "eval_mean_token_accuracy": 0.14463540688839963,
      "eval_num_tokens": 40079777.0,
      "eval_runtime": 15.2603,
      "eval_samples_per_second": 1086.284,
      "eval_steps_per_second": 135.843,
      "step": 12000
    },
    {
      "entropy": 6.351669549942017,
      "epoch": 2.3420154131304263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044882189096775993,
      "loss": 6.1247,
      "mean_token_accuracy": 0.1413797453045845,
      "num_tokens": 40096778.0,
      "step": 12005
    },
    {
      "entropy": 6.243264675140381,
      "epoch": 2.3429909277143692,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044877722281480093,
      "loss": 6.0427,
      "mean_token_accuracy": 0.1519486427307129,
      "num_tokens": 40112105.0,
      "step": 12010
    },
    {
      "entropy": 6.097096633911133,
      "epoch": 2.343966442298312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000448732537680951,
      "loss": 5.872,
      "mean_token_accuracy": 0.1553135022521019,
      "num_tokens": 40128611.0,
      "step": 12015
    },
    {
      "entropy": 6.37248420715332,
      "epoch": 2.3449419568822556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004486878355705767,
      "loss": 6.1585,
      "mean_token_accuracy": 0.1356949180364609,
      "num_tokens": 40145530.0,
      "step": 12020
    },
    {
      "entropy": 6.338003635406494,
      "epoch": 2.3459174714661986,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044864311648804613,
      "loss": 6.0995,
      "mean_token_accuracy": 0.1402410164475441,
      "num_tokens": 40162175.0,
      "step": 12025
    },
    {
      "entropy": 6.317586421966553,
      "epoch": 2.3468929860501415,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004485983804377291,
      "loss": 6.1214,
      "mean_token_accuracy": 0.14386224150657653,
      "num_tokens": 40178119.0,
      "step": 12030
    },
    {
      "entropy": 6.2666877746582035,
      "epoch": 2.3478685006340845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044855362742399706,
      "loss": 6.0146,
      "mean_token_accuracy": 0.15478023439645766,
      "num_tokens": 40193465.0,
      "step": 12035
    },
    {
      "entropy": 6.170851325988769,
      "epoch": 2.3488440152180274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485088574512231,
      "loss": 5.9973,
      "mean_token_accuracy": 0.15413923263549806,
      "num_tokens": 40209673.0,
      "step": 12040
    },
    {
      "entropy": 6.255893993377685,
      "epoch": 2.3498195298019704,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044846407052378194,
      "loss": 6.0911,
      "mean_token_accuracy": 0.1483833223581314,
      "num_tokens": 40225174.0,
      "step": 12045
    },
    {
      "entropy": 6.358680725097656,
      "epoch": 2.3507950443859134,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044841926664605016,
      "loss": 6.0698,
      "mean_token_accuracy": 0.15399102717638016,
      "num_tokens": 40242211.0,
      "step": 12050
    },
    {
      "entropy": 6.323353719711304,
      "epoch": 2.3517705589698568,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004483744458224056,
      "loss": 6.1721,
      "mean_token_accuracy": 0.13725527077913285,
      "num_tokens": 40258371.0,
      "step": 12055
    },
    {
      "entropy": 6.259028720855713,
      "epoch": 2.3527460735537997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044832960805722824,
      "loss": 6.0929,
      "mean_token_accuracy": 0.1468133270740509,
      "num_tokens": 40275574.0,
      "step": 12060
    },
    {
      "entropy": 6.321178388595581,
      "epoch": 2.3537215881377427,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004482847533548992,
      "loss": 6.0133,
      "mean_token_accuracy": 0.14970429837703705,
      "num_tokens": 40292575.0,
      "step": 12065
    },
    {
      "entropy": 6.228995943069458,
      "epoch": 2.3546971027216856,
      "grad_norm": 1.0,
      "learning_rate": 0.00044823988171980173,
      "loss": 6.1204,
      "mean_token_accuracy": 0.14398185908794403,
      "num_tokens": 40310001.0,
      "step": 12070
    },
    {
      "entropy": 6.254848432540894,
      "epoch": 2.3556726173056286,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044819499315632043,
      "loss": 5.9743,
      "mean_token_accuracy": 0.14872586727142334,
      "num_tokens": 40327014.0,
      "step": 12075
    },
    {
      "entropy": 6.327195501327514,
      "epoch": 2.3566481318895716,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044815008766884167,
      "loss": 6.0722,
      "mean_token_accuracy": 0.14778528958559037,
      "num_tokens": 40344628.0,
      "step": 12080
    },
    {
      "entropy": 6.367212963104248,
      "epoch": 2.357623646473515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004481051652617535,
      "loss": 6.1774,
      "mean_token_accuracy": 0.13799878135323523,
      "num_tokens": 40361336.0,
      "step": 12085
    },
    {
      "entropy": 6.323347520828247,
      "epoch": 2.358599161057458,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480602259394455,
      "loss": 6.0179,
      "mean_token_accuracy": 0.15025805085897445,
      "num_tokens": 40378378.0,
      "step": 12090
    },
    {
      "entropy": 6.212512683868408,
      "epoch": 2.359574675641401,
      "grad_norm": 1.0,
      "learning_rate": 0.00044801526970630906,
      "loss": 5.9765,
      "mean_token_accuracy": 0.15030775219202042,
      "num_tokens": 40395580.0,
      "step": 12095
    },
    {
      "entropy": 6.2114739418029785,
      "epoch": 2.360550190225344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004479702965667371,
      "loss": 6.032,
      "mean_token_accuracy": 0.1468564085662365,
      "num_tokens": 40412395.0,
      "step": 12100
    },
    {
      "entropy": 6.290848016738892,
      "epoch": 2.361525704809287,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004479253065251242,
      "loss": 6.0579,
      "mean_token_accuracy": 0.14904629290103913,
      "num_tokens": 40428736.0,
      "step": 12105
    },
    {
      "entropy": 6.324514961242675,
      "epoch": 2.3625012193932298,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004478802995858667,
      "loss": 6.059,
      "mean_token_accuracy": 0.1444017395377159,
      "num_tokens": 40445890.0,
      "step": 12110
    },
    {
      "entropy": 6.2056632995605465,
      "epoch": 2.3634767339771727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004478352757533624,
      "loss": 5.9658,
      "mean_token_accuracy": 0.1553890511393547,
      "num_tokens": 40462268.0,
      "step": 12115
    },
    {
      "entropy": 6.299913930892944,
      "epoch": 2.364452248561116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044779023503201105,
      "loss": 6.0646,
      "mean_token_accuracy": 0.14748004600405692,
      "num_tokens": 40480111.0,
      "step": 12120
    },
    {
      "entropy": 6.278724002838135,
      "epoch": 2.365427763145059,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044774517742621374,
      "loss": 5.9434,
      "mean_token_accuracy": 0.14776492714881898,
      "num_tokens": 40496868.0,
      "step": 12125
    },
    {
      "entropy": 6.212432622909546,
      "epoch": 2.366403277729002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044770010294037333,
      "loss": 5.9797,
      "mean_token_accuracy": 0.15029366239905356,
      "num_tokens": 40512533.0,
      "step": 12130
    },
    {
      "entropy": 6.330637073516845,
      "epoch": 2.367378792312945,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044765501157889446,
      "loss": 6.099,
      "mean_token_accuracy": 0.13845668658614158,
      "num_tokens": 40530843.0,
      "step": 12135
    },
    {
      "entropy": 6.216795301437378,
      "epoch": 2.368354306896888,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044760990334618314,
      "loss": 5.9165,
      "mean_token_accuracy": 0.15343313217163085,
      "num_tokens": 40546513.0,
      "step": 12140
    },
    {
      "entropy": 6.340396547317505,
      "epoch": 2.3693298214808314,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044756477824664726,
      "loss": 6.1213,
      "mean_token_accuracy": 0.13628450706601142,
      "num_tokens": 40563763.0,
      "step": 12145
    },
    {
      "entropy": 6.219386911392212,
      "epoch": 2.3703053360647743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004475196362846963,
      "loss": 5.9586,
      "mean_token_accuracy": 0.1519125998020172,
      "num_tokens": 40580558.0,
      "step": 12150
    },
    {
      "entropy": 6.289233064651489,
      "epoch": 2.3712808506487173,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004474744774647413,
      "loss": 6.0459,
      "mean_token_accuracy": 0.14011761397123337,
      "num_tokens": 40598053.0,
      "step": 12155
    },
    {
      "entropy": 6.288386440277099,
      "epoch": 2.3722563652326603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044742930179119517,
      "loss": 6.0213,
      "mean_token_accuracy": 0.1468390479683876,
      "num_tokens": 40614435.0,
      "step": 12160
    },
    {
      "entropy": 6.30715537071228,
      "epoch": 2.373231879816603,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004473841092684722,
      "loss": 6.026,
      "mean_token_accuracy": 0.14793270379304885,
      "num_tokens": 40630070.0,
      "step": 12165
    },
    {
      "entropy": 6.240496301651001,
      "epoch": 2.374207394400546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044733889990098834,
      "loss": 6.0672,
      "mean_token_accuracy": 0.14481963366270065,
      "num_tokens": 40647777.0,
      "step": 12170
    },
    {
      "entropy": 6.336735582351684,
      "epoch": 2.375182908984489,
      "grad_norm": 1.0,
      "learning_rate": 0.0004472936736931615,
      "loss": 6.1358,
      "mean_token_accuracy": 0.13659907430410384,
      "num_tokens": 40665035.0,
      "step": 12175
    },
    {
      "entropy": 6.29972882270813,
      "epoch": 2.376158423568432,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044724843064941096,
      "loss": 6.0687,
      "mean_token_accuracy": 0.1408476047217846,
      "num_tokens": 40681507.0,
      "step": 12180
    },
    {
      "entropy": 6.2992511749267575,
      "epoch": 2.3771339381523755,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004472031707741577,
      "loss": 6.013,
      "mean_token_accuracy": 0.1465413212776184,
      "num_tokens": 40698613.0,
      "step": 12185
    },
    {
      "entropy": 6.19135890007019,
      "epoch": 2.3781094527363185,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004471578940718242,
      "loss": 5.8959,
      "mean_token_accuracy": 0.15895339399576186,
      "num_tokens": 40716127.0,
      "step": 12190
    },
    {
      "entropy": 6.126419019699097,
      "epoch": 2.3790849673202614,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044711260054683497,
      "loss": 5.8893,
      "mean_token_accuracy": 0.16706997901201248,
      "num_tokens": 40734847.0,
      "step": 12195
    },
    {
      "entropy": 6.324120426177979,
      "epoch": 2.3800604819042044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004470672902036158,
      "loss": 6.1074,
      "mean_token_accuracy": 0.14216447100043297,
      "num_tokens": 40752266.0,
      "step": 12200
    },
    {
      "entropy": 6.320329475402832,
      "epoch": 2.3810359964881473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470219630465943,
      "loss": 6.1475,
      "mean_token_accuracy": 0.13293766751885414,
      "num_tokens": 40769011.0,
      "step": 12205
    },
    {
      "entropy": 6.296940279006958,
      "epoch": 2.3820115110720907,
      "grad_norm": 1.0,
      "learning_rate": 0.00044697661908019967,
      "loss": 6.0237,
      "mean_token_accuracy": 0.1429208092391491,
      "num_tokens": 40785260.0,
      "step": 12210
    },
    {
      "entropy": 6.226141166687012,
      "epoch": 2.3829870256560337,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044693125830886276,
      "loss": 5.9811,
      "mean_token_accuracy": 0.15579407811164855,
      "num_tokens": 40802651.0,
      "step": 12215
    },
    {
      "entropy": 6.197859764099121,
      "epoch": 2.3839625402399767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004468858807370161,
      "loss": 5.9752,
      "mean_token_accuracy": 0.15280209928750993,
      "num_tokens": 40820290.0,
      "step": 12220
    },
    {
      "entropy": 6.19496865272522,
      "epoch": 2.3849380548239196,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044684048636909376,
      "loss": 5.9066,
      "mean_token_accuracy": 0.16068874597549437,
      "num_tokens": 40838030.0,
      "step": 12225
    },
    {
      "entropy": 6.29974627494812,
      "epoch": 2.3859135694078626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004467950752095316,
      "loss": 6.0286,
      "mean_token_accuracy": 0.15251473858952522,
      "num_tokens": 40854979.0,
      "step": 12230
    },
    {
      "entropy": 6.3008216381072994,
      "epoch": 2.3868890839918055,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044674964726276685,
      "loss": 6.1432,
      "mean_token_accuracy": 0.13841716796159745,
      "num_tokens": 40872474.0,
      "step": 12235
    },
    {
      "entropy": 6.324249649047852,
      "epoch": 2.3878645985757485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044670420253323874,
      "loss": 6.0958,
      "mean_token_accuracy": 0.13978761956095695,
      "num_tokens": 40888116.0,
      "step": 12240
    },
    {
      "entropy": 6.265179395675659,
      "epoch": 2.388840113159692,
      "grad_norm": 0.875,
      "learning_rate": 0.000446658741025388,
      "loss": 5.9938,
      "mean_token_accuracy": 0.14352766871452333,
      "num_tokens": 40904385.0,
      "step": 12245
    },
    {
      "entropy": 6.222695827484131,
      "epoch": 2.389815627743635,
      "grad_norm": 1.125,
      "learning_rate": 0.0004466132627436568,
      "loss": 5.9926,
      "mean_token_accuracy": 0.14651983380317687,
      "num_tokens": 40923117.0,
      "step": 12250
    },
    {
      "entropy": 6.308610725402832,
      "epoch": 2.390791142327578,
      "grad_norm": 1.0,
      "learning_rate": 0.00044656776769248926,
      "loss": 6.0559,
      "mean_token_accuracy": 0.14286171197891234,
      "num_tokens": 40940029.0,
      "step": 12255
    },
    {
      "entropy": 6.303072929382324,
      "epoch": 2.391766656911521,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004465222558763309,
      "loss": 6.0656,
      "mean_token_accuracy": 0.14188651591539383,
      "num_tokens": 40955629.0,
      "step": 12260
    },
    {
      "entropy": 6.366601276397705,
      "epoch": 2.3927421714954638,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000446476727299629,
      "loss": 6.2076,
      "mean_token_accuracy": 0.1378733828663826,
      "num_tokens": 40974665.0,
      "step": 12265
    },
    {
      "entropy": 6.262490749359131,
      "epoch": 2.3937176860794067,
      "grad_norm": 0.859375,
      "learning_rate": 0.00044643118196683245,
      "loss": 6.0554,
      "mean_token_accuracy": 0.15380342677235603,
      "num_tokens": 40992528.0,
      "step": 12270
    },
    {
      "entropy": 6.33275990486145,
      "epoch": 2.39469320066335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004463856198823919,
      "loss": 6.0452,
      "mean_token_accuracy": 0.14657412245869636,
      "num_tokens": 41009315.0,
      "step": 12275
    },
    {
      "entropy": 6.16519513130188,
      "epoch": 2.395668715247293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004463400410507593,
      "loss": 5.9303,
      "mean_token_accuracy": 0.15237510651350022,
      "num_tokens": 41027718.0,
      "step": 12280
    },
    {
      "entropy": 6.194326877593994,
      "epoch": 2.396644229831236,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044629444547638865,
      "loss": 5.9332,
      "mean_token_accuracy": 0.16109570115804672,
      "num_tokens": 41045268.0,
      "step": 12285
    },
    {
      "entropy": 6.36592698097229,
      "epoch": 2.397619744415179,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004462488331637352,
      "loss": 6.1343,
      "mean_token_accuracy": 0.1416682317852974,
      "num_tokens": 41062522.0,
      "step": 12290
    },
    {
      "entropy": 6.227939653396606,
      "epoch": 2.398595258999122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004462032041172562,
      "loss": 6.0779,
      "mean_token_accuracy": 0.13851461410522461,
      "num_tokens": 41079131.0,
      "step": 12295
    },
    {
      "entropy": 6.2889628410339355,
      "epoch": 2.399570773583065,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004461575583414102,
      "loss": 5.9757,
      "mean_token_accuracy": 0.151631635427475,
      "num_tokens": 41095138.0,
      "step": 12300
    },
    {
      "entropy": 6.3088830471038815,
      "epoch": 2.400546288167008,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044611189584065764,
      "loss": 6.1072,
      "mean_token_accuracy": 0.14574891924858094,
      "num_tokens": 41111360.0,
      "step": 12305
    },
    {
      "entropy": 6.301737451553345,
      "epoch": 2.4015218027509513,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044606621661946044,
      "loss": 6.0894,
      "mean_token_accuracy": 0.1471741557121277,
      "num_tokens": 41126017.0,
      "step": 12310
    },
    {
      "entropy": 6.280316877365112,
      "epoch": 2.4024973173348942,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044602052068228227,
      "loss": 6.0644,
      "mean_token_accuracy": 0.14316802099347115,
      "num_tokens": 41142418.0,
      "step": 12315
    },
    {
      "entropy": 6.295060777664185,
      "epoch": 2.403472831918837,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044597480803358836,
      "loss": 6.0097,
      "mean_token_accuracy": 0.14888952076435089,
      "num_tokens": 41158597.0,
      "step": 12320
    },
    {
      "entropy": 6.131647109985352,
      "epoch": 2.40444834650278,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044592907867784556,
      "loss": 5.8862,
      "mean_token_accuracy": 0.15795677080750464,
      "num_tokens": 41175692.0,
      "step": 12325
    },
    {
      "entropy": 6.304438161849975,
      "epoch": 2.405423861086723,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004458833326195224,
      "loss": 6.0097,
      "mean_token_accuracy": 0.14514154121279715,
      "num_tokens": 41192322.0,
      "step": 12330
    },
    {
      "entropy": 6.171199655532837,
      "epoch": 2.406399375670666,
      "grad_norm": 1.015625,
      "learning_rate": 0.000445837569863089,
      "loss": 5.9056,
      "mean_token_accuracy": 0.15975725501775742,
      "num_tokens": 41209978.0,
      "step": 12335
    },
    {
      "entropy": 6.228154420852661,
      "epoch": 2.4073748902546095,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044579179041301717,
      "loss": 6.0366,
      "mean_token_accuracy": 0.14759132266044617,
      "num_tokens": 41228053.0,
      "step": 12340
    },
    {
      "entropy": 6.212140083312988,
      "epoch": 2.4083504048385524,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044574599427378033,
      "loss": 6.0106,
      "mean_token_accuracy": 0.15278043746948242,
      "num_tokens": 41244871.0,
      "step": 12345
    },
    {
      "entropy": 6.287372064590454,
      "epoch": 2.4093259194224954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044570018144985334,
      "loss": 6.0761,
      "mean_token_accuracy": 0.1410802960395813,
      "num_tokens": 41262622.0,
      "step": 12350
    },
    {
      "entropy": 6.322216415405274,
      "epoch": 2.4103014340064384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004456543519457132,
      "loss": 6.1299,
      "mean_token_accuracy": 0.13906718641519547,
      "num_tokens": 41280045.0,
      "step": 12355
    },
    {
      "entropy": 6.243094444274902,
      "epoch": 2.4112769485903813,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044560850576583785,
      "loss": 5.9952,
      "mean_token_accuracy": 0.15035391598939896,
      "num_tokens": 41297223.0,
      "step": 12360
    },
    {
      "entropy": 6.2458092212677006,
      "epoch": 2.4122524631743243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044556264291470746,
      "loss": 5.9509,
      "mean_token_accuracy": 0.15526669695973397,
      "num_tokens": 41313558.0,
      "step": 12365
    },
    {
      "entropy": 6.308018684387207,
      "epoch": 2.4132279777582673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004455167633968035,
      "loss": 6.1447,
      "mean_token_accuracy": 0.14148341193795205,
      "num_tokens": 41329259.0,
      "step": 12370
    },
    {
      "entropy": 6.161374759674072,
      "epoch": 2.4142034923422107,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044547086721660913,
      "loss": 5.8958,
      "mean_token_accuracy": 0.1587733283638954,
      "num_tokens": 41344725.0,
      "step": 12375
    },
    {
      "entropy": 6.314972829818726,
      "epoch": 2.4151790069261536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004454249543786091,
      "loss": 6.0131,
      "mean_token_accuracy": 0.14932730942964553,
      "num_tokens": 41360512.0,
      "step": 12380
    },
    {
      "entropy": 6.191746473312378,
      "epoch": 2.4161545215100966,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044537902488729,
      "loss": 5.943,
      "mean_token_accuracy": 0.15969090163707733,
      "num_tokens": 41377915.0,
      "step": 12385
    },
    {
      "entropy": 6.324790287017822,
      "epoch": 2.4171300360940395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004453330787471399,
      "loss": 6.0201,
      "mean_token_accuracy": 0.14535043016076088,
      "num_tokens": 41394515.0,
      "step": 12390
    },
    {
      "entropy": 6.305230617523193,
      "epoch": 2.4181055506779825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004452871159626484,
      "loss": 6.0022,
      "mean_token_accuracy": 0.13994933143258095,
      "num_tokens": 41410725.0,
      "step": 12395
    },
    {
      "entropy": 6.251168537139892,
      "epoch": 2.419081065261926,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044524113653830673,
      "loss": 6.0363,
      "mean_token_accuracy": 0.1451465368270874,
      "num_tokens": 41428139.0,
      "step": 12400
    },
    {
      "entropy": 6.117955732345581,
      "epoch": 2.420056579845869,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004451951404786081,
      "loss": 5.882,
      "mean_token_accuracy": 0.1630956768989563,
      "num_tokens": 41444210.0,
      "step": 12405
    },
    {
      "entropy": 6.171045112609863,
      "epoch": 2.421032094429812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044514912778804684,
      "loss": 5.9564,
      "mean_token_accuracy": 0.1536673679947853,
      "num_tokens": 41461107.0,
      "step": 12410
    },
    {
      "entropy": 6.178147792816162,
      "epoch": 2.422007609013755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044510309847111925,
      "loss": 5.9035,
      "mean_token_accuracy": 0.16359627097845078,
      "num_tokens": 41476731.0,
      "step": 12415
    },
    {
      "entropy": 6.283373689651489,
      "epoch": 2.4229831235976977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044505705253232317,
      "loss": 6.0597,
      "mean_token_accuracy": 0.14441127628087996,
      "num_tokens": 41494467.0,
      "step": 12420
    },
    {
      "entropy": 6.184131240844726,
      "epoch": 2.4239586381816407,
      "grad_norm": 0.875,
      "learning_rate": 0.00044501098997615793,
      "loss": 5.9341,
      "mean_token_accuracy": 0.15990680158138276,
      "num_tokens": 41512094.0,
      "step": 12425
    },
    {
      "entropy": 6.242037582397461,
      "epoch": 2.4249341527655837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044496491080712477,
      "loss": 6.0283,
      "mean_token_accuracy": 0.14863826781511308,
      "num_tokens": 41529058.0,
      "step": 12430
    },
    {
      "entropy": 6.315050029754639,
      "epoch": 2.425909667349527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004449188150297262,
      "loss": 6.156,
      "mean_token_accuracy": 0.1422819674015045,
      "num_tokens": 41547051.0,
      "step": 12435
    },
    {
      "entropy": 6.455913352966308,
      "epoch": 2.42688518193347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044487270264846666,
      "loss": 6.1948,
      "mean_token_accuracy": 0.13306871950626373,
      "num_tokens": 41563847.0,
      "step": 12440
    },
    {
      "entropy": 6.23896369934082,
      "epoch": 2.427860696517413,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044482657366785205,
      "loss": 5.9953,
      "mean_token_accuracy": 0.14889064878225328,
      "num_tokens": 41581090.0,
      "step": 12445
    },
    {
      "entropy": 6.303030014038086,
      "epoch": 2.428836211101356,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044478042809239,
      "loss": 6.0423,
      "mean_token_accuracy": 0.15043525397777557,
      "num_tokens": 41598177.0,
      "step": 12450
    },
    {
      "entropy": 6.13528618812561,
      "epoch": 2.429811725685299,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044473426592658963,
      "loss": 5.7972,
      "mean_token_accuracy": 0.17047664225101472,
      "num_tokens": 41615389.0,
      "step": 12455
    },
    {
      "entropy": 6.311246252059936,
      "epoch": 2.430787240269242,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044468808717496163,
      "loss": 6.0669,
      "mean_token_accuracy": 0.14635440707206726,
      "num_tokens": 41632639.0,
      "step": 12460
    },
    {
      "entropy": 6.224834060668945,
      "epoch": 2.4317627548531853,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004446418918420186,
      "loss": 6.004,
      "mean_token_accuracy": 0.15361220836639405,
      "num_tokens": 41648523.0,
      "step": 12465
    },
    {
      "entropy": 6.2893939971923825,
      "epoch": 2.4327382694371282,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445956799322745,
      "loss": 6.0317,
      "mean_token_accuracy": 0.14374398961663246,
      "num_tokens": 41665999.0,
      "step": 12470
    },
    {
      "entropy": 6.2507270812988285,
      "epoch": 2.433713784021071,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044454945145024505,
      "loss": 5.9897,
      "mean_token_accuracy": 0.1512048102915287,
      "num_tokens": 41685737.0,
      "step": 12475
    },
    {
      "entropy": 6.342110633850098,
      "epoch": 2.434689298605014,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004445032064004474,
      "loss": 6.1181,
      "mean_token_accuracy": 0.14085302054882048,
      "num_tokens": 41703511.0,
      "step": 12480
    },
    {
      "entropy": 6.265038347244262,
      "epoch": 2.435664813188957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044445694478740063,
      "loss": 6.065,
      "mean_token_accuracy": 0.1464575447142124,
      "num_tokens": 41720775.0,
      "step": 12485
    },
    {
      "entropy": 6.232242774963379,
      "epoch": 2.4366403277729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004444106666156251,
      "loss": 5.9858,
      "mean_token_accuracy": 0.1564353808760643,
      "num_tokens": 41738836.0,
      "step": 12490
    },
    {
      "entropy": 6.255876064300537,
      "epoch": 2.437615842356843,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044436437188964304,
      "loss": 6.0882,
      "mean_token_accuracy": 0.14421744421124458,
      "num_tokens": 41755140.0,
      "step": 12495
    },
    {
      "entropy": 6.312948846817017,
      "epoch": 2.4385913569407864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044431806061397816,
      "loss": 6.1051,
      "mean_token_accuracy": 0.1385059781372547,
      "num_tokens": 41770794.0,
      "step": 12500
    },
    {
      "entropy": 6.3483359813690186,
      "epoch": 2.4395668715247294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044427173279315584,
      "loss": 6.0681,
      "mean_token_accuracy": 0.14584231972694398,
      "num_tokens": 41787464.0,
      "step": 12505
    },
    {
      "entropy": 6.310025453567505,
      "epoch": 2.4405423861086724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004442253884317032,
      "loss": 6.0753,
      "mean_token_accuracy": 0.14489177241921425,
      "num_tokens": 41804851.0,
      "step": 12510
    },
    {
      "entropy": 6.238664531707764,
      "epoch": 2.4415179006926153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004441790275341487,
      "loss": 6.0053,
      "mean_token_accuracy": 0.15377910882234574,
      "num_tokens": 41821959.0,
      "step": 12515
    },
    {
      "entropy": 6.2031578540802,
      "epoch": 2.4424934152765583,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004441326501050225,
      "loss": 5.9462,
      "mean_token_accuracy": 0.16224450021982192,
      "num_tokens": 41837581.0,
      "step": 12520
    },
    {
      "entropy": 6.206959009170532,
      "epoch": 2.4434689298605012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044408625614885663,
      "loss": 5.9471,
      "mean_token_accuracy": 0.15562666952610016,
      "num_tokens": 41854497.0,
      "step": 12525
    },
    {
      "entropy": 6.321823835372925,
      "epoch": 2.4444444444444446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004440398456701844,
      "loss": 6.1144,
      "mean_token_accuracy": 0.13645765110850333,
      "num_tokens": 41870968.0,
      "step": 12530
    },
    {
      "entropy": 6.265920734405517,
      "epoch": 2.4454199590283876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044399341867354093,
      "loss": 5.941,
      "mean_token_accuracy": 0.14915903136134148,
      "num_tokens": 41886734.0,
      "step": 12535
    },
    {
      "entropy": 6.23798680305481,
      "epoch": 2.4463954736123306,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044394697516346286,
      "loss": 6.0452,
      "mean_token_accuracy": 0.149243126809597,
      "num_tokens": 41903728.0,
      "step": 12540
    },
    {
      "entropy": 6.296313190460205,
      "epoch": 2.4473709881962735,
      "grad_norm": 1.0,
      "learning_rate": 0.00044390051514448854,
      "loss": 6.0676,
      "mean_token_accuracy": 0.13919911533594131,
      "num_tokens": 41919967.0,
      "step": 12545
    },
    {
      "entropy": 6.128534317016602,
      "epoch": 2.4483465027802165,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00044385403862115785,
      "loss": 5.834,
      "mean_token_accuracy": 0.16033875793218613,
      "num_tokens": 41937330.0,
      "step": 12550
    },
    {
      "entropy": 6.2394966125488285,
      "epoch": 2.4493220173641594,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044380754559801235,
      "loss": 6.0001,
      "mean_token_accuracy": 0.14488652125000953,
      "num_tokens": 41953261.0,
      "step": 12555
    },
    {
      "entropy": 6.334920597076416,
      "epoch": 2.4502975319481024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044376103607959513,
      "loss": 6.102,
      "mean_token_accuracy": 0.1358017437160015,
      "num_tokens": 41970468.0,
      "step": 12560
    },
    {
      "entropy": 6.2459827899932865,
      "epoch": 2.451273046532046,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044371451007045083,
      "loss": 6.0188,
      "mean_token_accuracy": 0.15032144635915756,
      "num_tokens": 41986235.0,
      "step": 12565
    },
    {
      "entropy": 6.237640380859375,
      "epoch": 2.4522485611159888,
      "grad_norm": 0.96875,
      "learning_rate": 0.000443667967575126,
      "loss": 5.9478,
      "mean_token_accuracy": 0.14223495423793792,
      "num_tokens": 42004539.0,
      "step": 12570
    },
    {
      "entropy": 6.290277051925659,
      "epoch": 2.4532240756999317,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436214085981686,
      "loss": 5.9772,
      "mean_token_accuracy": 0.1542574167251587,
      "num_tokens": 42022937.0,
      "step": 12575
    },
    {
      "entropy": 6.292674875259399,
      "epoch": 2.4541995902838747,
      "grad_norm": 1.09375,
      "learning_rate": 0.000443574833144128,
      "loss": 5.9741,
      "mean_token_accuracy": 0.14716312885284424,
      "num_tokens": 42039260.0,
      "step": 12580
    },
    {
      "entropy": 6.221227312088013,
      "epoch": 2.4551751048678176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004435282412175556,
      "loss": 6.0119,
      "mean_token_accuracy": 0.14429089799523354,
      "num_tokens": 42055894.0,
      "step": 12585
    },
    {
      "entropy": 6.256859397888183,
      "epoch": 2.456150619451761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044348163282300403,
      "loss": 6.0665,
      "mean_token_accuracy": 0.14724738150835037,
      "num_tokens": 42070748.0,
      "step": 12590
    },
    {
      "entropy": 6.324374866485596,
      "epoch": 2.457126134035704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044343500796502783,
      "loss": 6.0669,
      "mean_token_accuracy": 0.14583705812692643,
      "num_tokens": 42088016.0,
      "step": 12595
    },
    {
      "entropy": 6.320900917053223,
      "epoch": 2.458101648619647,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004433883666481829,
      "loss": 5.9623,
      "mean_token_accuracy": 0.15193734765052797,
      "num_tokens": 42105422.0,
      "step": 12600
    },
    {
      "entropy": 6.185057640075684,
      "epoch": 2.45907716320359,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000443341708877027,
      "loss": 6.0093,
      "mean_token_accuracy": 0.14901165962219237,
      "num_tokens": 42120534.0,
      "step": 12605
    },
    {
      "entropy": 6.241154718399048,
      "epoch": 2.460052677787533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004432950346561193,
      "loss": 5.9759,
      "mean_token_accuracy": 0.1556865930557251,
      "num_tokens": 42137413.0,
      "step": 12610
    },
    {
      "entropy": 6.2396448135375975,
      "epoch": 2.461028192371476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004432483439900206,
      "loss": 6.0387,
      "mean_token_accuracy": 0.14696442186832429,
      "num_tokens": 42153006.0,
      "step": 12615
    },
    {
      "entropy": 6.3366106986999515,
      "epoch": 2.462003706955419,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004432016368832934,
      "loss": 6.1507,
      "mean_token_accuracy": 0.1363580882549286,
      "num_tokens": 42170966.0,
      "step": 12620
    },
    {
      "entropy": 6.355908918380737,
      "epoch": 2.4629792215393618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044315491334050166,
      "loss": 6.0222,
      "mean_token_accuracy": 0.1411207564175129,
      "num_tokens": 42187349.0,
      "step": 12625
    },
    {
      "entropy": 6.2226966381072994,
      "epoch": 2.463954736123305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004431081733662112,
      "loss": 5.9866,
      "mean_token_accuracy": 0.1570926383137703,
      "num_tokens": 42204486.0,
      "step": 12630
    },
    {
      "entropy": 6.293699073791504,
      "epoch": 2.464930250707248,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044306141696498914,
      "loss": 6.0204,
      "mean_token_accuracy": 0.14643407016992568,
      "num_tokens": 42220316.0,
      "step": 12635
    },
    {
      "entropy": 6.1139482975006105,
      "epoch": 2.465905765291191,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044301464414140437,
      "loss": 5.9141,
      "mean_token_accuracy": 0.15361291468143462,
      "num_tokens": 42237734.0,
      "step": 12640
    },
    {
      "entropy": 6.357467794418335,
      "epoch": 2.466881279875134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004429678549000274,
      "loss": 6.1303,
      "mean_token_accuracy": 0.13426149412989616,
      "num_tokens": 42253458.0,
      "step": 12645
    },
    {
      "entropy": 6.144019269943238,
      "epoch": 2.467856794459077,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004429210492454303,
      "loss": 5.9263,
      "mean_token_accuracy": 0.1529343381524086,
      "num_tokens": 42270031.0,
      "step": 12650
    },
    {
      "entropy": 6.3559300899505615,
      "epoch": 2.4688323090430204,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044287422718218675,
      "loss": 6.0438,
      "mean_token_accuracy": 0.1498144283890724,
      "num_tokens": 42286227.0,
      "step": 12655
    },
    {
      "entropy": 6.280980730056763,
      "epoch": 2.4698078236269634,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044282738871487205,
      "loss": 6.0305,
      "mean_token_accuracy": 0.14498544931411744,
      "num_tokens": 42302361.0,
      "step": 12660
    },
    {
      "entropy": 6.269526481628418,
      "epoch": 2.4707833382109063,
      "grad_norm": 1.09375,
      "learning_rate": 0.000442780533848063,
      "loss": 6.0866,
      "mean_token_accuracy": 0.1418351337313652,
      "num_tokens": 42319462.0,
      "step": 12665
    },
    {
      "entropy": 6.280678129196167,
      "epoch": 2.4717588527948493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044273366258633823,
      "loss": 6.0222,
      "mean_token_accuracy": 0.14900000989437104,
      "num_tokens": 42335565.0,
      "step": 12670
    },
    {
      "entropy": 6.318631505966186,
      "epoch": 2.4727343673787923,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004426867749342778,
      "loss": 6.1181,
      "mean_token_accuracy": 0.13672236427664758,
      "num_tokens": 42352335.0,
      "step": 12675
    },
    {
      "entropy": 6.272025775909424,
      "epoch": 2.4737098819627352,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004426398708964633,
      "loss": 6.0146,
      "mean_token_accuracy": 0.15469063445925713,
      "num_tokens": 42368320.0,
      "step": 12680
    },
    {
      "entropy": 6.215971279144287,
      "epoch": 2.474685396546678,
      "grad_norm": 1.125,
      "learning_rate": 0.00044259295047747814,
      "loss": 5.9571,
      "mean_token_accuracy": 0.1511790484189987,
      "num_tokens": 42384224.0,
      "step": 12685
    },
    {
      "entropy": 6.247292470932007,
      "epoch": 2.4756609111306216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004425460136819071,
      "loss": 5.9834,
      "mean_token_accuracy": 0.15176981836557388,
      "num_tokens": 42399973.0,
      "step": 12690
    },
    {
      "entropy": 6.219344282150269,
      "epoch": 2.4766364257145645,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004424990605143369,
      "loss": 6.0336,
      "mean_token_accuracy": 0.14653353840112687,
      "num_tokens": 42415355.0,
      "step": 12695
    },
    {
      "entropy": 6.258107328414917,
      "epoch": 2.4776119402985075,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044245209097935537,
      "loss": 6.0609,
      "mean_token_accuracy": 0.14572559297084808,
      "num_tokens": 42431176.0,
      "step": 12700
    },
    {
      "entropy": 6.272924900054932,
      "epoch": 2.4785874548824505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004424051050815523,
      "loss": 6.0003,
      "mean_token_accuracy": 0.150200654566288,
      "num_tokens": 42446935.0,
      "step": 12705
    },
    {
      "entropy": 6.222360706329345,
      "epoch": 2.4795629694663934,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044235810282551905,
      "loss": 5.9254,
      "mean_token_accuracy": 0.16007743030786514,
      "num_tokens": 42463473.0,
      "step": 12710
    },
    {
      "entropy": 6.289364385604858,
      "epoch": 2.4805384840503364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004423110842158484,
      "loss": 5.9945,
      "mean_token_accuracy": 0.15365808606147766,
      "num_tokens": 42479460.0,
      "step": 12715
    },
    {
      "entropy": 6.2380951881408695,
      "epoch": 2.48151399863428,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000442264049257135,
      "loss": 6.0095,
      "mean_token_accuracy": 0.15078665614128112,
      "num_tokens": 42495213.0,
      "step": 12720
    },
    {
      "entropy": 6.321897506713867,
      "epoch": 2.4824895132182228,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004422169979539748,
      "loss": 6.0808,
      "mean_token_accuracy": 0.1432742953300476,
      "num_tokens": 42513104.0,
      "step": 12725
    },
    {
      "entropy": 6.279369306564331,
      "epoch": 2.4834650278021657,
      "grad_norm": 0.875,
      "learning_rate": 0.00044216993031096554,
      "loss": 5.9888,
      "mean_token_accuracy": 0.14911687523126602,
      "num_tokens": 42529774.0,
      "step": 12730
    },
    {
      "entropy": 6.239939832687378,
      "epoch": 2.4844405423861087,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004421228463327064,
      "loss": 6.0231,
      "mean_token_accuracy": 0.14864950329065324,
      "num_tokens": 42545570.0,
      "step": 12735
    },
    {
      "entropy": 6.276117944717408,
      "epoch": 2.4854160569700516,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044207574602379856,
      "loss": 6.0257,
      "mean_token_accuracy": 0.1482274278998375,
      "num_tokens": 42563165.0,
      "step": 12740
    },
    {
      "entropy": 6.2531046867370605,
      "epoch": 2.4863915715539946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004420286293888441,
      "loss": 6.0635,
      "mean_token_accuracy": 0.14765256717801095,
      "num_tokens": 42578870.0,
      "step": 12745
    },
    {
      "entropy": 6.209896993637085,
      "epoch": 2.4873670861379376,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004419814964324474,
      "loss": 5.9704,
      "mean_token_accuracy": 0.15499818101525306,
      "num_tokens": 42596247.0,
      "step": 12750
    },
    {
      "entropy": 6.355484962463379,
      "epoch": 2.488342600721881,
      "grad_norm": 1.03125,
      "learning_rate": 0.000441934347159214,
      "loss": 6.177,
      "mean_token_accuracy": 0.13772156313061715,
      "num_tokens": 42613196.0,
      "step": 12755
    },
    {
      "entropy": 6.299760198593139,
      "epoch": 2.489318115305824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000441887181573751,
      "loss": 6.0059,
      "mean_token_accuracy": 0.14364723563194276,
      "num_tokens": 42628242.0,
      "step": 12760
    },
    {
      "entropy": 6.285452079772949,
      "epoch": 2.490293629889767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004418399996806676,
      "loss": 6.0857,
      "mean_token_accuracy": 0.14097429513931276,
      "num_tokens": 42645687.0,
      "step": 12765
    },
    {
      "entropy": 6.299478483200073,
      "epoch": 2.49126914447371,
      "grad_norm": 1.0,
      "learning_rate": 0.00044179280148457395,
      "loss": 6.0812,
      "mean_token_accuracy": 0.13850736767053604,
      "num_tokens": 42661998.0,
      "step": 12770
    },
    {
      "entropy": 6.341281986236572,
      "epoch": 2.492244659057653,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044174558699008225,
      "loss": 6.1081,
      "mean_token_accuracy": 0.1444399692118168,
      "num_tokens": 42678276.0,
      "step": 12775
    },
    {
      "entropy": 6.33143196105957,
      "epoch": 2.4932201736415958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044169835620180603,
      "loss": 6.105,
      "mean_token_accuracy": 0.14379451274871827,
      "num_tokens": 42697094.0,
      "step": 12780
    },
    {
      "entropy": 6.322100830078125,
      "epoch": 2.494195688225539,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004416511091243606,
      "loss": 6.0642,
      "mean_token_accuracy": 0.1400900438427925,
      "num_tokens": 42713188.0,
      "step": 12785
    },
    {
      "entropy": 6.325028896331787,
      "epoch": 2.495171202809482,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004416038457623628,
      "loss": 6.0473,
      "mean_token_accuracy": 0.14424021542072296,
      "num_tokens": 42729210.0,
      "step": 12790
    },
    {
      "entropy": 6.230472803115845,
      "epoch": 2.496146717393425,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004415565661204309,
      "loss": 6.0223,
      "mean_token_accuracy": 0.15032730773091316,
      "num_tokens": 42746543.0,
      "step": 12795
    },
    {
      "entropy": 6.2818550109863285,
      "epoch": 2.497122231977368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000441509270203185,
      "loss": 6.0185,
      "mean_token_accuracy": 0.1463448852300644,
      "num_tokens": 42762265.0,
      "step": 12800
    },
    {
      "entropy": 6.293279409408569,
      "epoch": 2.498097746561311,
      "grad_norm": 0.828125,
      "learning_rate": 0.00044146195801524665,
      "loss": 6.049,
      "mean_token_accuracy": 0.15364544242620468,
      "num_tokens": 42779555.0,
      "step": 12805
    },
    {
      "entropy": 6.350290727615357,
      "epoch": 2.499073261145254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000441414629561239,
      "loss": 6.1085,
      "mean_token_accuracy": 0.13808467462658883,
      "num_tokens": 42797211.0,
      "step": 12810
    },
    {
      "entropy": 6.342711925506592,
      "epoch": 2.500048775729197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004413672848457869,
      "loss": 6.0774,
      "mean_token_accuracy": 0.140689480304718,
      "num_tokens": 42813827.0,
      "step": 12815
    },
    {
      "entropy": 6.305530595779419,
      "epoch": 2.5010242903131403,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004413199238735168,
      "loss": 6.049,
      "mean_token_accuracy": 0.14124192595481871,
      "num_tokens": 42829458.0,
      "step": 12820
    },
    {
      "entropy": 6.337877082824707,
      "epoch": 2.5019998048970833,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044127254664905636,
      "loss": 6.0853,
      "mean_token_accuracy": 0.14856272116303443,
      "num_tokens": 42844651.0,
      "step": 12825
    },
    {
      "entropy": 6.199969959259033,
      "epoch": 2.5029753194810263,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004412251531770353,
      "loss": 5.9256,
      "mean_token_accuracy": 0.1548518344759941,
      "num_tokens": 42860425.0,
      "step": 12830
    },
    {
      "entropy": 6.2976305961608885,
      "epoch": 2.503950834064969,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004411777434620847,
      "loss": 6.0836,
      "mean_token_accuracy": 0.1367059551179409,
      "num_tokens": 42877734.0,
      "step": 12835
    },
    {
      "entropy": 6.286861753463745,
      "epoch": 2.504926348648912,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004411303175088374,
      "loss": 6.0252,
      "mean_token_accuracy": 0.14148588553071023,
      "num_tokens": 42895143.0,
      "step": 12840
    },
    {
      "entropy": 6.289626693725586,
      "epoch": 2.5059018632328556,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004410828753219274,
      "loss": 6.0095,
      "mean_token_accuracy": 0.1414526730775833,
      "num_tokens": 42910777.0,
      "step": 12845
    },
    {
      "entropy": 6.3519463539123535,
      "epoch": 2.5068773778167985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004410354169059908,
      "loss": 6.0966,
      "mean_token_accuracy": 0.1383506529033184,
      "num_tokens": 42927594.0,
      "step": 12850
    },
    {
      "entropy": 6.246324491500855,
      "epoch": 2.5078528924007415,
      "grad_norm": 1.03125,
      "learning_rate": 0.000440987942265665,
      "loss": 6.0267,
      "mean_token_accuracy": 0.14618308693170548,
      "num_tokens": 42943896.0,
      "step": 12855
    },
    {
      "entropy": 6.175165748596191,
      "epoch": 2.5088284069846845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044094045140558914,
      "loss": 5.885,
      "mean_token_accuracy": 0.15295365154743196,
      "num_tokens": 42960762.0,
      "step": 12860
    },
    {
      "entropy": 6.214346027374267,
      "epoch": 2.5098039215686274,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004408929443304037,
      "loss": 5.8868,
      "mean_token_accuracy": 0.15817152410745622,
      "num_tokens": 42976994.0,
      "step": 12865
    },
    {
      "entropy": 6.21792688369751,
      "epoch": 2.5107794361525704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044084542104475113,
      "loss": 5.9983,
      "mean_token_accuracy": 0.14893289804458618,
      "num_tokens": 42994827.0,
      "step": 12870
    },
    {
      "entropy": 6.14800763130188,
      "epoch": 2.5117549507365133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044079788155327503,
      "loss": 5.8622,
      "mean_token_accuracy": 0.16573034524917601,
      "num_tokens": 43011318.0,
      "step": 12875
    },
    {
      "entropy": 6.27344536781311,
      "epoch": 2.5127304653204563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044075032586062083,
      "loss": 6.0661,
      "mean_token_accuracy": 0.1430560566484928,
      "num_tokens": 43028956.0,
      "step": 12880
    },
    {
      "entropy": 6.2259192943573,
      "epoch": 2.5137059799043997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044070275397143553,
      "loss": 5.9176,
      "mean_token_accuracy": 0.146781887114048,
      "num_tokens": 43045178.0,
      "step": 12885
    },
    {
      "entropy": 6.286348485946656,
      "epoch": 2.5146814944883427,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044065516589036777,
      "loss": 6.0701,
      "mean_token_accuracy": 0.14030610769987106,
      "num_tokens": 43061887.0,
      "step": 12890
    },
    {
      "entropy": 6.305762529373169,
      "epoch": 2.5156570090722856,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044060756162206757,
      "loss": 6.0471,
      "mean_token_accuracy": 0.14790304601192475,
      "num_tokens": 43078754.0,
      "step": 12895
    },
    {
      "entropy": 6.176844263076783,
      "epoch": 2.5166325236562286,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004405599411711866,
      "loss": 5.9543,
      "mean_token_accuracy": 0.1523220047354698,
      "num_tokens": 43096599.0,
      "step": 12900
    },
    {
      "entropy": 6.29280047416687,
      "epoch": 2.5176080382401715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004405123045423783,
      "loss": 6.013,
      "mean_token_accuracy": 0.14294768273830413,
      "num_tokens": 43111503.0,
      "step": 12905
    },
    {
      "entropy": 6.262187528610229,
      "epoch": 2.518583552824115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004404646517402975,
      "loss": 6.0241,
      "mean_token_accuracy": 0.1478027284145355,
      "num_tokens": 43129052.0,
      "step": 12910
    },
    {
      "entropy": 6.153257942199707,
      "epoch": 2.519559067408058,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004404169827696006,
      "loss": 5.9171,
      "mean_token_accuracy": 0.15552832186222076,
      "num_tokens": 43147456.0,
      "step": 12915
    },
    {
      "entropy": 6.234153413772583,
      "epoch": 2.520534581992001,
      "grad_norm": 1.0,
      "learning_rate": 0.0004403692976349458,
      "loss": 5.9803,
      "mean_token_accuracy": 0.1472913533449173,
      "num_tokens": 43163535.0,
      "step": 12920
    },
    {
      "entropy": 6.301495885848999,
      "epoch": 2.521510096575944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004403215963409926,
      "loss": 6.0596,
      "mean_token_accuracy": 0.14190984070301055,
      "num_tokens": 43181376.0,
      "step": 12925
    },
    {
      "entropy": 6.311179780960083,
      "epoch": 2.522485611159887,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044027387889240225,
      "loss": 6.0256,
      "mean_token_accuracy": 0.1501276671886444,
      "num_tokens": 43196993.0,
      "step": 12930
    },
    {
      "entropy": 6.25733437538147,
      "epoch": 2.5234611257438297,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004402261452938374,
      "loss": 6.0257,
      "mean_token_accuracy": 0.15047376304864885,
      "num_tokens": 43214355.0,
      "step": 12935
    },
    {
      "entropy": 6.303263187408447,
      "epoch": 2.5244366403277727,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004401783955499626,
      "loss": 6.1292,
      "mean_token_accuracy": 0.13927378058433532,
      "num_tokens": 43231031.0,
      "step": 12940
    },
    {
      "entropy": 6.272302103042603,
      "epoch": 2.5254121549117157,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044013062966544365,
      "loss": 5.9756,
      "mean_token_accuracy": 0.14924041628837587,
      "num_tokens": 43248603.0,
      "step": 12945
    },
    {
      "entropy": 6.350983810424805,
      "epoch": 2.526387669495659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004400828476449481,
      "loss": 6.0646,
      "mean_token_accuracy": 0.14613796770572662,
      "num_tokens": 43264053.0,
      "step": 12950
    },
    {
      "entropy": 6.289100265502929,
      "epoch": 2.527363184079602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044003504949314507,
      "loss": 6.0622,
      "mean_token_accuracy": 0.14560191631317138,
      "num_tokens": 43282739.0,
      "step": 12955
    },
    {
      "entropy": 6.160247659683227,
      "epoch": 2.528338698663545,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004399872352147051,
      "loss": 5.8968,
      "mean_token_accuracy": 0.16026347130537033,
      "num_tokens": 43298885.0,
      "step": 12960
    },
    {
      "entropy": 6.3531403064727785,
      "epoch": 2.529314213247488,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004399394048143007,
      "loss": 6.0685,
      "mean_token_accuracy": 0.14189369976520538,
      "num_tokens": 43314960.0,
      "step": 12965
    },
    {
      "entropy": 6.130657768249511,
      "epoch": 2.5302897278314314,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004398915582966054,
      "loss": 5.9447,
      "mean_token_accuracy": 0.15090589001774787,
      "num_tokens": 43331637.0,
      "step": 12970
    },
    {
      "entropy": 6.225008773803711,
      "epoch": 2.5312652424153743,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004398436956662948,
      "loss": 6.0264,
      "mean_token_accuracy": 0.1433355450630188,
      "num_tokens": 43349534.0,
      "step": 12975
    },
    {
      "entropy": 6.39310941696167,
      "epoch": 2.5322407569993173,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043979581692804565,
      "loss": 6.0333,
      "mean_token_accuracy": 0.1443559154868126,
      "num_tokens": 43366385.0,
      "step": 12980
    },
    {
      "entropy": 6.240903425216675,
      "epoch": 2.5332162715832602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043974792208653667,
      "loss": 5.9786,
      "mean_token_accuracy": 0.14629695788025857,
      "num_tokens": 43383204.0,
      "step": 12985
    },
    {
      "entropy": 6.123369026184082,
      "epoch": 2.534191786167203,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397000111464479,
      "loss": 5.8496,
      "mean_token_accuracy": 0.16136842370033264,
      "num_tokens": 43401509.0,
      "step": 12990
    },
    {
      "entropy": 6.1934168338775635,
      "epoch": 2.535167300751146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004396520841124611,
      "loss": 5.9106,
      "mean_token_accuracy": 0.15552762001752854,
      "num_tokens": 43418958.0,
      "step": 12995
    },
    {
      "entropy": 6.342613172531128,
      "epoch": 2.536142815335089,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004396041409892594,
      "loss": 6.1018,
      "mean_token_accuracy": 0.14021877348423004,
      "num_tokens": 43436032.0,
      "step": 13000
    },
    {
      "entropy": 6.330953359603882,
      "epoch": 2.537118329919032,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043955618178152773,
      "loss": 6.1243,
      "mean_token_accuracy": 0.13799187391996384,
      "num_tokens": 43453433.0,
      "step": 13005
    },
    {
      "entropy": 6.315371513366699,
      "epoch": 2.5380938445029755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043950820649395244,
      "loss": 6.064,
      "mean_token_accuracy": 0.1403285689651966,
      "num_tokens": 43469450.0,
      "step": 13010
    },
    {
      "entropy": 6.311705589294434,
      "epoch": 2.5390693590869184,
      "grad_norm": 1.0,
      "learning_rate": 0.0004394602151312216,
      "loss": 6.0513,
      "mean_token_accuracy": 0.14972377866506575,
      "num_tokens": 43486633.0,
      "step": 13015
    },
    {
      "entropy": 6.3151122570037845,
      "epoch": 2.5400448736708614,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043941220769802466,
      "loss": 6.0406,
      "mean_token_accuracy": 0.14675991609692574,
      "num_tokens": 43503056.0,
      "step": 13020
    },
    {
      "entropy": 6.179532051086426,
      "epoch": 2.5410203882548044,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043936418419905274,
      "loss": 5.9856,
      "mean_token_accuracy": 0.14892366230487825,
      "num_tokens": 43519691.0,
      "step": 13025
    },
    {
      "entropy": 6.299685621261597,
      "epoch": 2.5419959028387473,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004393161446389985,
      "loss": 6.0286,
      "mean_token_accuracy": 0.14791033864021302,
      "num_tokens": 43537152.0,
      "step": 13030
    },
    {
      "entropy": 6.336560535430908,
      "epoch": 2.5429714174226907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004392680890225563,
      "loss": 6.0428,
      "mean_token_accuracy": 0.14362604767084122,
      "num_tokens": 43554352.0,
      "step": 13035
    },
    {
      "entropy": 6.245951890945435,
      "epoch": 2.5439469320066337,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004392200173544219,
      "loss": 5.9926,
      "mean_token_accuracy": 0.14565864205360413,
      "num_tokens": 43570244.0,
      "step": 13040
    },
    {
      "entropy": 6.242842483520508,
      "epoch": 2.5449224465905766,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043917192963929266,
      "loss": 6.0007,
      "mean_token_accuracy": 0.1482628434896469,
      "num_tokens": 43585701.0,
      "step": 13045
    },
    {
      "entropy": 6.277584505081177,
      "epoch": 2.5458979611745196,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004391238258818676,
      "loss": 6.03,
      "mean_token_accuracy": 0.14479517340660095,
      "num_tokens": 43600966.0,
      "step": 13050
    },
    {
      "entropy": 6.1683002471923825,
      "epoch": 2.5468734757584626,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004390757060868473,
      "loss": 5.933,
      "mean_token_accuracy": 0.15264015048742294,
      "num_tokens": 43617318.0,
      "step": 13055
    },
    {
      "entropy": 6.318524360656738,
      "epoch": 2.5478489903424055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043902757025893364,
      "loss": 6.0551,
      "mean_token_accuracy": 0.1430969625711441,
      "num_tokens": 43633124.0,
      "step": 13060
    },
    {
      "entropy": 6.263995361328125,
      "epoch": 2.5488245049263485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043897941840283043,
      "loss": 6.0333,
      "mean_token_accuracy": 0.14895378649234772,
      "num_tokens": 43649281.0,
      "step": 13065
    },
    {
      "entropy": 6.303466176986694,
      "epoch": 2.5498000195102914,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043893125052324293,
      "loss": 6.0519,
      "mean_token_accuracy": 0.14803403243422508,
      "num_tokens": 43665572.0,
      "step": 13070
    },
    {
      "entropy": 6.320852661132813,
      "epoch": 2.550775534094235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004388830666248779,
      "loss": 6.1531,
      "mean_token_accuracy": 0.14422462806105613,
      "num_tokens": 43682447.0,
      "step": 13075
    },
    {
      "entropy": 6.314708089828491,
      "epoch": 2.551751048678178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004388348667124437,
      "loss": 6.0401,
      "mean_token_accuracy": 0.14436660408973695,
      "num_tokens": 43698528.0,
      "step": 13080
    },
    {
      "entropy": 6.237517023086548,
      "epoch": 2.5527265632621208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043878665079065025,
      "loss": 5.9776,
      "mean_token_accuracy": 0.14747550264000892,
      "num_tokens": 43715403.0,
      "step": 13085
    },
    {
      "entropy": 6.2601158142089846,
      "epoch": 2.5537020778460637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043873841886420893,
      "loss": 5.992,
      "mean_token_accuracy": 0.14936922639608383,
      "num_tokens": 43732029.0,
      "step": 13090
    },
    {
      "entropy": 6.252572631835937,
      "epoch": 2.5546775924300067,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043869017093783305,
      "loss": 6.0412,
      "mean_token_accuracy": 0.15265828669071196,
      "num_tokens": 43749951.0,
      "step": 13095
    },
    {
      "entropy": 6.377341270446777,
      "epoch": 2.55565310701395,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004386419070162369,
      "loss": 6.1103,
      "mean_token_accuracy": 0.1485545367002487,
      "num_tokens": 43766169.0,
      "step": 13100
    },
    {
      "entropy": 6.280043220520019,
      "epoch": 2.556628621597893,
      "grad_norm": 1.125,
      "learning_rate": 0.00043859362710413694,
      "loss": 6.0345,
      "mean_token_accuracy": 0.14880056381225587,
      "num_tokens": 43782751.0,
      "step": 13105
    },
    {
      "entropy": 6.2175291061401365,
      "epoch": 2.557604136181836,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004385453312062508,
      "loss": 5.9328,
      "mean_token_accuracy": 0.149529629945755,
      "num_tokens": 43799113.0,
      "step": 13110
    },
    {
      "entropy": 6.266657114028931,
      "epoch": 2.558579650765779,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004384970193272977,
      "loss": 5.9976,
      "mean_token_accuracy": 0.14766507148742675,
      "num_tokens": 43816926.0,
      "step": 13115
    },
    {
      "entropy": 6.263850212097168,
      "epoch": 2.559555165349722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043844869147199867,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1527656689286232,
      "num_tokens": 43834557.0,
      "step": 13120
    },
    {
      "entropy": 6.264004325866699,
      "epoch": 2.560530679933665,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000438400347645076,
      "loss": 6.0124,
      "mean_token_accuracy": 0.14782193750143052,
      "num_tokens": 43850527.0,
      "step": 13125
    },
    {
      "entropy": 6.2223780155181885,
      "epoch": 2.561506194517608,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043835198785125367,
      "loss": 5.9931,
      "mean_token_accuracy": 0.14672531187534332,
      "num_tokens": 43867889.0,
      "step": 13130
    },
    {
      "entropy": 6.312329053878784,
      "epoch": 2.562481709101551,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004383036120952574,
      "loss": 6.0906,
      "mean_token_accuracy": 0.1384254775941372,
      "num_tokens": 43885002.0,
      "step": 13135
    },
    {
      "entropy": 6.28605465888977,
      "epoch": 2.5634572236854942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004382552203818142,
      "loss": 6.0348,
      "mean_token_accuracy": 0.15201013237237931,
      "num_tokens": 43902161.0,
      "step": 13140
    },
    {
      "entropy": 6.285544490814209,
      "epoch": 2.564432738269437,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043820681271565265,
      "loss": 5.9457,
      "mean_token_accuracy": 0.15588801279664039,
      "num_tokens": 43917580.0,
      "step": 13145
    },
    {
      "entropy": 6.238213634490966,
      "epoch": 2.56540825285338,
      "grad_norm": 1.0625,
      "learning_rate": 0.000438158389101503,
      "loss": 6.0152,
      "mean_token_accuracy": 0.14443636387586595,
      "num_tokens": 43933709.0,
      "step": 13150
    },
    {
      "entropy": 6.253530025482178,
      "epoch": 2.566383767437323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004381099495440971,
      "loss": 5.9814,
      "mean_token_accuracy": 0.1472393497824669,
      "num_tokens": 43950264.0,
      "step": 13155
    },
    {
      "entropy": 6.1751336574554445,
      "epoch": 2.567359282021266,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004380614940481683,
      "loss": 5.8902,
      "mean_token_accuracy": 0.15786474868655204,
      "num_tokens": 43966722.0,
      "step": 13160
    },
    {
      "entropy": 6.217467451095581,
      "epoch": 2.5683347966052095,
      "grad_norm": 1.375,
      "learning_rate": 0.0004380130226184514,
      "loss": 5.9488,
      "mean_token_accuracy": 0.15517865121364594,
      "num_tokens": 43984780.0,
      "step": 13165
    },
    {
      "entropy": 6.251792287826538,
      "epoch": 2.5693103111891524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043796453525968304,
      "loss": 6.0002,
      "mean_token_accuracy": 0.14712864607572557,
      "num_tokens": 44000954.0,
      "step": 13170
    },
    {
      "entropy": 6.349868679046631,
      "epoch": 2.5702858257730954,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043791603197660095,
      "loss": 6.0658,
      "mean_token_accuracy": 0.14637986868619918,
      "num_tokens": 44017214.0,
      "step": 13175
    },
    {
      "entropy": 6.276723957061767,
      "epoch": 2.5712613403570383,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000437867512773945,
      "loss": 6.056,
      "mean_token_accuracy": 0.14488063752651215,
      "num_tokens": 44032793.0,
      "step": 13180
    },
    {
      "entropy": 6.244515657424927,
      "epoch": 2.5722368549409813,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043781897765645615,
      "loss": 5.9941,
      "mean_token_accuracy": 0.1539919674396515,
      "num_tokens": 44047831.0,
      "step": 13185
    },
    {
      "entropy": 6.296598768234253,
      "epoch": 2.5732123695249243,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437770426628877,
      "loss": 6.115,
      "mean_token_accuracy": 0.14259875789284707,
      "num_tokens": 44063861.0,
      "step": 13190
    },
    {
      "entropy": 6.3213887214660645,
      "epoch": 2.5741878841088672,
      "grad_norm": 0.984375,
      "learning_rate": 0.000437721859695952,
      "loss": 6.1007,
      "mean_token_accuracy": 0.1439078316092491,
      "num_tokens": 44080740.0,
      "step": 13195
    },
    {
      "entropy": 6.318906021118164,
      "epoch": 2.57516339869281,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004376732768624267,
      "loss": 5.9944,
      "mean_token_accuracy": 0.14360572174191474,
      "num_tokens": 44096581.0,
      "step": 13200
    },
    {
      "entropy": 6.232607412338257,
      "epoch": 2.5761389132767536,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004376246781330487,
      "loss": 5.9369,
      "mean_token_accuracy": 0.15373810827732087,
      "num_tokens": 44113556.0,
      "step": 13205
    },
    {
      "entropy": 6.185152339935303,
      "epoch": 2.5771144278606966,
      "grad_norm": 1.0,
      "learning_rate": 0.00043757606351256664,
      "loss": 5.9579,
      "mean_token_accuracy": 0.15744076818227767,
      "num_tokens": 44131440.0,
      "step": 13210
    },
    {
      "entropy": 6.201213598251343,
      "epoch": 2.5780899424446395,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043752743300573115,
      "loss": 6.007,
      "mean_token_accuracy": 0.1500581532716751,
      "num_tokens": 44148982.0,
      "step": 13215
    },
    {
      "entropy": 6.349516487121582,
      "epoch": 2.5790654570285825,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004374787866172942,
      "loss": 6.0092,
      "mean_token_accuracy": 0.14747979640960693,
      "num_tokens": 44165540.0,
      "step": 13220
    },
    {
      "entropy": 6.232906150817871,
      "epoch": 2.580040971612526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004374301243520092,
      "loss": 6.0395,
      "mean_token_accuracy": 0.1452867016196251,
      "num_tokens": 44182038.0,
      "step": 13225
    },
    {
      "entropy": 6.284265041351318,
      "epoch": 2.581016486196469,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043738144621463136,
      "loss": 6.118,
      "mean_token_accuracy": 0.1428047701716423,
      "num_tokens": 44198838.0,
      "step": 13230
    },
    {
      "entropy": 6.247124528884887,
      "epoch": 2.581992000780412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004373327522099174,
      "loss": 5.9228,
      "mean_token_accuracy": 0.15846910923719407,
      "num_tokens": 44215422.0,
      "step": 13235
    },
    {
      "entropy": 6.285521078109741,
      "epoch": 2.5829675153643548,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043728404234262547,
      "loss": 6.1131,
      "mean_token_accuracy": 0.14250001013278962,
      "num_tokens": 44232019.0,
      "step": 13240
    },
    {
      "entropy": 6.399219274520874,
      "epoch": 2.5839430299482977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004372353166175153,
      "loss": 6.0445,
      "mean_token_accuracy": 0.14396784007549285,
      "num_tokens": 44246675.0,
      "step": 13245
    },
    {
      "entropy": 6.2968706607818605,
      "epoch": 2.5849185445322407,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004371865750393482,
      "loss": 6.0923,
      "mean_token_accuracy": 0.14123225957155228,
      "num_tokens": 44262531.0,
      "step": 13250
    },
    {
      "entropy": 6.302400779724121,
      "epoch": 2.5858940591161836,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043713781761288705,
      "loss": 6.0078,
      "mean_token_accuracy": 0.14932631850242614,
      "num_tokens": 44279440.0,
      "step": 13255
    },
    {
      "entropy": 6.274488067626953,
      "epoch": 2.5868695737001266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004370890443428962,
      "loss": 6.0616,
      "mean_token_accuracy": 0.14439291805028914,
      "num_tokens": 44296384.0,
      "step": 13260
    },
    {
      "entropy": 6.166977643966675,
      "epoch": 2.58784508828407,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043704025523414163,
      "loss": 5.8784,
      "mean_token_accuracy": 0.1586136683821678,
      "num_tokens": 44313318.0,
      "step": 13265
    },
    {
      "entropy": 6.258075141906739,
      "epoch": 2.588820602868013,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004369914502913909,
      "loss": 5.9271,
      "mean_token_accuracy": 0.15079473853111267,
      "num_tokens": 44329752.0,
      "step": 13270
    },
    {
      "entropy": 6.264841938018799,
      "epoch": 2.589796117451956,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004369426295194129,
      "loss": 5.965,
      "mean_token_accuracy": 0.15016015619039536,
      "num_tokens": 44347318.0,
      "step": 13275
    },
    {
      "entropy": 6.311624240875244,
      "epoch": 2.590771632035899,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004368937929229783,
      "loss": 6.0346,
      "mean_token_accuracy": 0.1495633453130722,
      "num_tokens": 44362595.0,
      "step": 13280
    },
    {
      "entropy": 6.226622724533081,
      "epoch": 2.591747146619842,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043684494050685933,
      "loss": 6.0702,
      "mean_token_accuracy": 0.1387689858675003,
      "num_tokens": 44378611.0,
      "step": 13285
    },
    {
      "entropy": 6.296039962768555,
      "epoch": 2.5927226612037853,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004367960722758296,
      "loss": 6.0446,
      "mean_token_accuracy": 0.14530161619186402,
      "num_tokens": 44394427.0,
      "step": 13290
    },
    {
      "entropy": 6.271791315078735,
      "epoch": 2.593698175787728,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043674718823466437,
      "loss": 6.0174,
      "mean_token_accuracy": 0.1480724923312664,
      "num_tokens": 44411952.0,
      "step": 13295
    },
    {
      "entropy": 6.273071575164795,
      "epoch": 2.594673690371671,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043669828838814037,
      "loss": 5.996,
      "mean_token_accuracy": 0.14918098598718643,
      "num_tokens": 44428434.0,
      "step": 13300
    },
    {
      "entropy": 6.234520816802979,
      "epoch": 2.595649204955614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004366493727410359,
      "loss": 5.9614,
      "mean_token_accuracy": 0.15560984909534453,
      "num_tokens": 44447072.0,
      "step": 13305
    },
    {
      "entropy": 6.297246313095092,
      "epoch": 2.596624719539557,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043660044129813084,
      "loss": 6.0314,
      "mean_token_accuracy": 0.14357317686080934,
      "num_tokens": 44464135.0,
      "step": 13310
    },
    {
      "entropy": 6.25408706665039,
      "epoch": 2.5976002341235,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043655149406420657,
      "loss": 6.0842,
      "mean_token_accuracy": 0.14739783257246017,
      "num_tokens": 44479366.0,
      "step": 13315
    },
    {
      "entropy": 6.342901086807251,
      "epoch": 2.598575748707443,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004365025310440462,
      "loss": 6.0565,
      "mean_token_accuracy": 0.14116014465689658,
      "num_tokens": 44496957.0,
      "step": 13320
    },
    {
      "entropy": 6.372913360595703,
      "epoch": 2.599551263291386,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000436453552242434,
      "loss": 6.0747,
      "mean_token_accuracy": 0.1409519851207733,
      "num_tokens": 44514300.0,
      "step": 13325
    },
    {
      "entropy": 6.253871822357178,
      "epoch": 2.6005267778753294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043640455766415616,
      "loss": 6.0296,
      "mean_token_accuracy": 0.15072071701288223,
      "num_tokens": 44530354.0,
      "step": 13330
    },
    {
      "entropy": 6.311959266662598,
      "epoch": 2.6015022924592723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043635554731400014,
      "loss": 6.0322,
      "mean_token_accuracy": 0.1484083652496338,
      "num_tokens": 44547175.0,
      "step": 13335
    },
    {
      "entropy": 6.217301893234253,
      "epoch": 2.6024778070432153,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043630652119675515,
      "loss": 6.0337,
      "mean_token_accuracy": 0.14598288238048554,
      "num_tokens": 44564461.0,
      "step": 13340
    },
    {
      "entropy": 6.274942111968994,
      "epoch": 2.6034533216271583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043625747931721185,
      "loss": 6.0244,
      "mean_token_accuracy": 0.15020784884691238,
      "num_tokens": 44583031.0,
      "step": 13345
    },
    {
      "entropy": 6.263120222091675,
      "epoch": 2.604428836211101,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004362084216801624,
      "loss": 6.0246,
      "mean_token_accuracy": 0.14371549040079118,
      "num_tokens": 44600408.0,
      "step": 13350
    },
    {
      "entropy": 6.253127527236939,
      "epoch": 2.6054043507950446,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043615934829040057,
      "loss": 5.9719,
      "mean_token_accuracy": 0.15611988604068755,
      "num_tokens": 44618361.0,
      "step": 13355
    },
    {
      "entropy": 6.167724704742431,
      "epoch": 2.6063798653789876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004361102591527215,
      "loss": 5.796,
      "mean_token_accuracy": 0.164225035905838,
      "num_tokens": 44634246.0,
      "step": 13360
    },
    {
      "entropy": 6.105594444274902,
      "epoch": 2.6073553799629305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004360611542719222,
      "loss": 5.9107,
      "mean_token_accuracy": 0.15414296686649323,
      "num_tokens": 44650087.0,
      "step": 13365
    },
    {
      "entropy": 6.240294075012207,
      "epoch": 2.6083308945468735,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004360120336528009,
      "loss": 6.0194,
      "mean_token_accuracy": 0.15050989389419556,
      "num_tokens": 44666540.0,
      "step": 13370
    },
    {
      "entropy": 6.25012617111206,
      "epoch": 2.6093064091308165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043596289730015754,
      "loss": 6.0055,
      "mean_token_accuracy": 0.15165585651993752,
      "num_tokens": 44683856.0,
      "step": 13375
    },
    {
      "entropy": 6.262922859191894,
      "epoch": 2.6102819237147594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043591374521879353,
      "loss": 6.0006,
      "mean_token_accuracy": 0.14310139194130897,
      "num_tokens": 44700297.0,
      "step": 13380
    },
    {
      "entropy": 6.226145505905151,
      "epoch": 2.6112574382987024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004358645774135118,
      "loss": 5.9255,
      "mean_token_accuracy": 0.1524067297577858,
      "num_tokens": 44716878.0,
      "step": 13385
    },
    {
      "entropy": 6.296266412734985,
      "epoch": 2.6122329528826453,
      "grad_norm": 0.953125,
      "learning_rate": 0.000435815393889117,
      "loss": 6.0393,
      "mean_token_accuracy": 0.14700075685977937,
      "num_tokens": 44734656.0,
      "step": 13390
    },
    {
      "entropy": 6.281830930709839,
      "epoch": 2.6132084674665887,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043576619465041496,
      "loss": 6.0113,
      "mean_token_accuracy": 0.14244814068078995,
      "num_tokens": 44750857.0,
      "step": 13395
    },
    {
      "entropy": 6.286918020248413,
      "epoch": 2.6141839820505317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004357169797022134,
      "loss": 6.0249,
      "mean_token_accuracy": 0.14374616891145706,
      "num_tokens": 44767641.0,
      "step": 13400
    },
    {
      "entropy": 6.040944814682007,
      "epoch": 2.6151594966344747,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004356677490493214,
      "loss": 5.7197,
      "mean_token_accuracy": 0.17327279672026635,
      "num_tokens": 44785003.0,
      "step": 13405
    },
    {
      "entropy": 6.32127366065979,
      "epoch": 2.6161350112184176,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004356185026965496,
      "loss": 6.1147,
      "mean_token_accuracy": 0.14223199263215064,
      "num_tokens": 44800308.0,
      "step": 13410
    },
    {
      "entropy": 6.355884170532226,
      "epoch": 2.6171105258023606,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043556924064871023,
      "loss": 6.0558,
      "mean_token_accuracy": 0.14032940119504927,
      "num_tokens": 44817064.0,
      "step": 13415
    },
    {
      "entropy": 6.154467487335205,
      "epoch": 2.618086040386304,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043551996291061684,
      "loss": 5.8837,
      "mean_token_accuracy": 0.15721844881772995,
      "num_tokens": 44834863.0,
      "step": 13420
    },
    {
      "entropy": 6.146806621551514,
      "epoch": 2.619061554970247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004354706694870848,
      "loss": 5.9308,
      "mean_token_accuracy": 0.1544480301439762,
      "num_tokens": 44852453.0,
      "step": 13425
    },
    {
      "entropy": 6.302256393432617,
      "epoch": 2.62003706955419,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043542136038293087,
      "loss": 6.022,
      "mean_token_accuracy": 0.14277254194021224,
      "num_tokens": 44868976.0,
      "step": 13430
    },
    {
      "entropy": 6.2614258289337155,
      "epoch": 2.621012584138133,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004353720356029733,
      "loss": 5.9822,
      "mean_token_accuracy": 0.14690974354743958,
      "num_tokens": 44886490.0,
      "step": 13435
    },
    {
      "entropy": 6.346377897262573,
      "epoch": 2.621988098722076,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004353226951520321,
      "loss": 6.0743,
      "mean_token_accuracy": 0.14865155667066574,
      "num_tokens": 44902572.0,
      "step": 13440
    },
    {
      "entropy": 6.251208400726318,
      "epoch": 2.622963613306019,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043527333903492846,
      "loss": 6.0185,
      "mean_token_accuracy": 0.14650631099939346,
      "num_tokens": 44918181.0,
      "step": 13445
    },
    {
      "entropy": 6.307183647155762,
      "epoch": 2.6239391278899618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004352239672564854,
      "loss": 6.1359,
      "mean_token_accuracy": 0.14087493270635604,
      "num_tokens": 44934809.0,
      "step": 13450
    },
    {
      "entropy": 6.3155840873718265,
      "epoch": 2.6249146424739047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043517457982152725,
      "loss": 6.0673,
      "mean_token_accuracy": 0.14259479716420173,
      "num_tokens": 44951626.0,
      "step": 13455
    },
    {
      "entropy": 6.303248262405395,
      "epoch": 2.625890157057848,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004351251767348801,
      "loss": 6.0941,
      "mean_token_accuracy": 0.14108134806156158,
      "num_tokens": 44968211.0,
      "step": 13460
    },
    {
      "entropy": 6.240374422073364,
      "epoch": 2.626865671641791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043507575800137135,
      "loss": 5.902,
      "mean_token_accuracy": 0.1569766767323017,
      "num_tokens": 44984605.0,
      "step": 13465
    },
    {
      "entropy": 6.306239938735962,
      "epoch": 2.627841186225734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043502632362583005,
      "loss": 6.0414,
      "mean_token_accuracy": 0.1469213619828224,
      "num_tokens": 45000803.0,
      "step": 13470
    },
    {
      "entropy": 6.123032474517823,
      "epoch": 2.628816700809677,
      "grad_norm": 0.875,
      "learning_rate": 0.00043497687361308677,
      "loss": 5.8754,
      "mean_token_accuracy": 0.16591592282056808,
      "num_tokens": 45018362.0,
      "step": 13475
    },
    {
      "entropy": 6.180677032470703,
      "epoch": 2.6297922153936204,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043492740796797347,
      "loss": 5.8482,
      "mean_token_accuracy": 0.15900825709104538,
      "num_tokens": 45033160.0,
      "step": 13480
    },
    {
      "entropy": 6.32230920791626,
      "epoch": 2.6307677299775634,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004348779266953239,
      "loss": 6.0294,
      "mean_token_accuracy": 0.14965516030788423,
      "num_tokens": 45049631.0,
      "step": 13485
    },
    {
      "entropy": 6.1701515197753904,
      "epoch": 2.6317432445615063,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043482842979997323,
      "loss": 5.9396,
      "mean_token_accuracy": 0.15648388117551804,
      "num_tokens": 45067083.0,
      "step": 13490
    },
    {
      "entropy": 6.225676202774048,
      "epoch": 2.6327187591454493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004347789172867579,
      "loss": 6.0396,
      "mean_token_accuracy": 0.14824800491333007,
      "num_tokens": 45083576.0,
      "step": 13495
    },
    {
      "entropy": 6.245168542861938,
      "epoch": 2.6336942737293922,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043472938916051633,
      "loss": 5.9557,
      "mean_token_accuracy": 0.15691422075033187,
      "num_tokens": 45100738.0,
      "step": 13500
    },
    {
      "entropy": 6.232299137115478,
      "epoch": 2.634669788313335,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004346798454260882,
      "loss": 5.9933,
      "mean_token_accuracy": 0.14183720722794532,
      "num_tokens": 45117254.0,
      "step": 13505
    },
    {
      "entropy": 6.106680822372437,
      "epoch": 2.635645302897278,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004346302860883146,
      "loss": 5.7995,
      "mean_token_accuracy": 0.1742892012000084,
      "num_tokens": 45135160.0,
      "step": 13510
    },
    {
      "entropy": 6.240459299087524,
      "epoch": 2.636620817481221,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043458071115203836,
      "loss": 5.9496,
      "mean_token_accuracy": 0.15071164220571517,
      "num_tokens": 45152025.0,
      "step": 13515
    },
    {
      "entropy": 6.121939945220947,
      "epoch": 2.6375963320651645,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043453112062210377,
      "loss": 5.8793,
      "mean_token_accuracy": 0.1583749920129776,
      "num_tokens": 45169452.0,
      "step": 13520
    },
    {
      "entropy": 6.209090042114258,
      "epoch": 2.6385718466491075,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043448151450335666,
      "loss": 5.901,
      "mean_token_accuracy": 0.15899208933115005,
      "num_tokens": 45186401.0,
      "step": 13525
    },
    {
      "entropy": 6.2525135517120365,
      "epoch": 2.6395473612330504,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043443189280064444,
      "loss": 5.9698,
      "mean_token_accuracy": 0.15062050968408586,
      "num_tokens": 45201713.0,
      "step": 13530
    },
    {
      "entropy": 6.287310647964477,
      "epoch": 2.6405228758169934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043438225551881577,
      "loss": 6.0557,
      "mean_token_accuracy": 0.14876216351985933,
      "num_tokens": 45216727.0,
      "step": 13535
    },
    {
      "entropy": 6.239027833938598,
      "epoch": 2.6414983904009364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043433260266272116,
      "loss": 5.9662,
      "mean_token_accuracy": 0.14643826186656952,
      "num_tokens": 45231958.0,
      "step": 13540
    },
    {
      "entropy": 6.297397899627685,
      "epoch": 2.6424739049848798,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004342829342372125,
      "loss": 6.039,
      "mean_token_accuracy": 0.1433882988989353,
      "num_tokens": 45248620.0,
      "step": 13545
    },
    {
      "entropy": 6.201178503036499,
      "epoch": 2.6434494195688227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004342332502471432,
      "loss": 5.9051,
      "mean_token_accuracy": 0.157735738158226,
      "num_tokens": 45264065.0,
      "step": 13550
    },
    {
      "entropy": 6.285269069671631,
      "epoch": 2.6444249341527657,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043418355069736834,
      "loss": 6.0556,
      "mean_token_accuracy": 0.145534910261631,
      "num_tokens": 45279818.0,
      "step": 13555
    },
    {
      "entropy": 6.2139300346374515,
      "epoch": 2.6454004487367087,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043413383559274404,
      "loss": 5.8944,
      "mean_token_accuracy": 0.15798521488904954,
      "num_tokens": 45296282.0,
      "step": 13560
    },
    {
      "entropy": 6.204136705398559,
      "epoch": 2.6463759633206516,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004340841049381286,
      "loss": 5.9293,
      "mean_token_accuracy": 0.15141840949654578,
      "num_tokens": 45314345.0,
      "step": 13565
    },
    {
      "entropy": 6.220501518249511,
      "epoch": 2.6473514779045946,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004340343587383814,
      "loss": 5.9588,
      "mean_token_accuracy": 0.15503242760896682,
      "num_tokens": 45329975.0,
      "step": 13570
    },
    {
      "entropy": 6.243371629714966,
      "epoch": 2.6483269924885375,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004339845969983634,
      "loss": 5.9563,
      "mean_token_accuracy": 0.1514472171664238,
      "num_tokens": 45347163.0,
      "step": 13575
    },
    {
      "entropy": 6.235499095916748,
      "epoch": 2.6493025070724805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004339348197229373,
      "loss": 5.9496,
      "mean_token_accuracy": 0.1549084335565567,
      "num_tokens": 45362966.0,
      "step": 13580
    },
    {
      "entropy": 6.239350318908691,
      "epoch": 2.650278021656424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004338850269169671,
      "loss": 6.0245,
      "mean_token_accuracy": 0.14357788115739822,
      "num_tokens": 45379628.0,
      "step": 13585
    },
    {
      "entropy": 6.234705781936645,
      "epoch": 2.651253536240367,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004338352185853183,
      "loss": 5.9526,
      "mean_token_accuracy": 0.1505420334637165,
      "num_tokens": 45397088.0,
      "step": 13590
    },
    {
      "entropy": 6.247823429107666,
      "epoch": 2.65222905082431,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000433785394732858,
      "loss": 5.9569,
      "mean_token_accuracy": 0.15196189880371094,
      "num_tokens": 45414134.0,
      "step": 13595
    },
    {
      "entropy": 6.2247514724731445,
      "epoch": 2.653204565408253,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043373555536445487,
      "loss": 5.9346,
      "mean_token_accuracy": 0.15608946681022645,
      "num_tokens": 45431398.0,
      "step": 13600
    },
    {
      "entropy": 6.282065582275391,
      "epoch": 2.6541800799921957,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043368570048497895,
      "loss": 6.0523,
      "mean_token_accuracy": 0.14521847516298295,
      "num_tokens": 45448867.0,
      "step": 13605
    },
    {
      "entropy": 6.29674916267395,
      "epoch": 2.655155594576139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004336358300993021,
      "loss": 5.99,
      "mean_token_accuracy": 0.14829104840755464,
      "num_tokens": 45466388.0,
      "step": 13610
    },
    {
      "entropy": 6.225131320953369,
      "epoch": 2.656131109160082,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043358594421229717,
      "loss": 5.9668,
      "mean_token_accuracy": 0.15071047842502594,
      "num_tokens": 45483711.0,
      "step": 13615
    },
    {
      "entropy": 6.192389059066772,
      "epoch": 2.657106623744025,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000433536042828839,
      "loss": 5.9352,
      "mean_token_accuracy": 0.15134310126304626,
      "num_tokens": 45502228.0,
      "step": 13620
    },
    {
      "entropy": 6.216763496398926,
      "epoch": 2.658082138327968,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004334861259538038,
      "loss": 5.8878,
      "mean_token_accuracy": 0.1546794906258583,
      "num_tokens": 45519278.0,
      "step": 13625
    },
    {
      "entropy": 6.295430707931518,
      "epoch": 2.659057652911911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004334361935920693,
      "loss": 6.0667,
      "mean_token_accuracy": 0.14337946474552155,
      "num_tokens": 45535287.0,
      "step": 13630
    },
    {
      "entropy": 6.3030554294586185,
      "epoch": 2.660033167495854,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043338624574851454,
      "loss": 6.0193,
      "mean_token_accuracy": 0.14972767606377602,
      "num_tokens": 45551744.0,
      "step": 13635
    },
    {
      "entropy": 6.327000665664673,
      "epoch": 2.661008682079797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004333362824280204,
      "loss": 6.0897,
      "mean_token_accuracy": 0.1427089422941208,
      "num_tokens": 45569089.0,
      "step": 13640
    },
    {
      "entropy": 6.280104160308838,
      "epoch": 2.66198419666374,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043328630363546894,
      "loss": 6.0087,
      "mean_token_accuracy": 0.15162042602896691,
      "num_tokens": 45586750.0,
      "step": 13645
    },
    {
      "entropy": 6.305783081054687,
      "epoch": 2.6629597112476833,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004332363093757442,
      "loss": 5.9965,
      "mean_token_accuracy": 0.15209872722625734,
      "num_tokens": 45604103.0,
      "step": 13650
    },
    {
      "entropy": 6.211026239395141,
      "epoch": 2.6639352258316262,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004331862996537312,
      "loss": 6.0099,
      "mean_token_accuracy": 0.14638373106718064,
      "num_tokens": 45620588.0,
      "step": 13655
    },
    {
      "entropy": 6.308323001861572,
      "epoch": 2.664910740415569,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004331362744743168,
      "loss": 6.0353,
      "mean_token_accuracy": 0.1515328750014305,
      "num_tokens": 45638289.0,
      "step": 13660
    },
    {
      "entropy": 6.213793706893921,
      "epoch": 2.665886254999512,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043308623384238925,
      "loss": 5.8878,
      "mean_token_accuracy": 0.15954763889312745,
      "num_tokens": 45655904.0,
      "step": 13665
    },
    {
      "entropy": 6.226664209365845,
      "epoch": 2.666861769583455,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00043303617776283834,
      "loss": 6.0008,
      "mean_token_accuracy": 0.14912529587745665,
      "num_tokens": 45673314.0,
      "step": 13670
    },
    {
      "entropy": 6.277876281738282,
      "epoch": 2.6678372841673985,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004329861062405554,
      "loss": 5.9594,
      "mean_token_accuracy": 0.1576439768075943,
      "num_tokens": 45689082.0,
      "step": 13675
    },
    {
      "entropy": 6.3268028736114506,
      "epoch": 2.6688127987513415,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004329360192804332,
      "loss": 6.1076,
      "mean_token_accuracy": 0.14167659729719162,
      "num_tokens": 45705979.0,
      "step": 13680
    },
    {
      "entropy": 6.382341718673706,
      "epoch": 2.6697883133352844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328859168873661,
      "loss": 6.1637,
      "mean_token_accuracy": 0.13593902662396432,
      "num_tokens": 45722941.0,
      "step": 13685
    },
    {
      "entropy": 6.27644944190979,
      "epoch": 2.6707638279192274,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043283579906625003,
      "loss": 6.0,
      "mean_token_accuracy": 0.15136749744415284,
      "num_tokens": 45739625.0,
      "step": 13690
    },
    {
      "entropy": 6.269883346557617,
      "epoch": 2.6717393425031704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004327856658219821,
      "loss": 6.0415,
      "mean_token_accuracy": 0.1469683066010475,
      "num_tokens": 45756008.0,
      "step": 13695
    },
    {
      "entropy": 6.257565689086914,
      "epoch": 2.6727148570871133,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00043273551715946126,
      "loss": 6.0364,
      "mean_token_accuracy": 0.14647665470838547,
      "num_tokens": 45773043.0,
      "step": 13700
    },
    {
      "entropy": 6.15273699760437,
      "epoch": 2.6736903716710563,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004326853530835879,
      "loss": 5.8811,
      "mean_token_accuracy": 0.15811533331871033,
      "num_tokens": 45790987.0,
      "step": 13705
    },
    {
      "entropy": 6.306484794616699,
      "epoch": 2.6746658862549992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004326351735992638,
      "loss": 6.0511,
      "mean_token_accuracy": 0.14380398988723755,
      "num_tokens": 45807867.0,
      "step": 13710
    },
    {
      "entropy": 6.3193480491638185,
      "epoch": 2.6756414008389426,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004325849787113924,
      "loss": 6.0702,
      "mean_token_accuracy": 0.14513107091188432,
      "num_tokens": 45824806.0,
      "step": 13715
    },
    {
      "entropy": 6.241525650024414,
      "epoch": 2.6766169154228856,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004325347684248785,
      "loss": 5.9495,
      "mean_token_accuracy": 0.1500966265797615,
      "num_tokens": 45841013.0,
      "step": 13720
    },
    {
      "entropy": 6.339110898971557,
      "epoch": 2.6775924300068286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004324845427446285,
      "loss": 6.0373,
      "mean_token_accuracy": 0.1449911817908287,
      "num_tokens": 45857618.0,
      "step": 13725
    },
    {
      "entropy": 6.253673791885376,
      "epoch": 2.6785679445907715,
      "grad_norm": 0.90625,
      "learning_rate": 0.00043243430167555025,
      "loss": 5.9815,
      "mean_token_accuracy": 0.1479606434702873,
      "num_tokens": 45875807.0,
      "step": 13730
    },
    {
      "entropy": 6.306834602355957,
      "epoch": 2.679543459174715,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004323840452225532,
      "loss": 6.0261,
      "mean_token_accuracy": 0.1433855965733528,
      "num_tokens": 45892580.0,
      "step": 13735
    },
    {
      "entropy": 6.2318745136260985,
      "epoch": 2.680518973758658,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004323337733905481,
      "loss": 6.0457,
      "mean_token_accuracy": 0.1472772479057312,
      "num_tokens": 45908717.0,
      "step": 13740
    },
    {
      "entropy": 6.204717111587525,
      "epoch": 2.681494488342601,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043228348618444746,
      "loss": 5.9113,
      "mean_token_accuracy": 0.15959796011447908,
      "num_tokens": 45925200.0,
      "step": 13745
    },
    {
      "entropy": 6.2878752708435055,
      "epoch": 2.682470002926544,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004322331836091651,
      "loss": 6.0221,
      "mean_token_accuracy": 0.1450769543647766,
      "num_tokens": 45942384.0,
      "step": 13750
    },
    {
      "entropy": 6.302674341201782,
      "epoch": 2.6834455175104868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004321828656696165,
      "loss": 6.048,
      "mean_token_accuracy": 0.14828323870897292,
      "num_tokens": 45959040.0,
      "step": 13755
    },
    {
      "entropy": 6.28315749168396,
      "epoch": 2.6844210320944297,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043213253237071837,
      "loss": 6.0121,
      "mean_token_accuracy": 0.14851996004581453,
      "num_tokens": 45975559.0,
      "step": 13760
    },
    {
      "entropy": 6.271198320388794,
      "epoch": 2.6853965466783727,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043208218371738926,
      "loss": 6.0714,
      "mean_token_accuracy": 0.1424812540411949,
      "num_tokens": 45992587.0,
      "step": 13765
    },
    {
      "entropy": 6.331692361831665,
      "epoch": 2.6863720612623156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043203181971454904,
      "loss": 6.0663,
      "mean_token_accuracy": 0.14337299540638923,
      "num_tokens": 46008612.0,
      "step": 13770
    },
    {
      "entropy": 6.326591157913208,
      "epoch": 2.687347575846259,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043198144036711903,
      "loss": 6.1296,
      "mean_token_accuracy": 0.1458910509943962,
      "num_tokens": 46025730.0,
      "step": 13775
    },
    {
      "entropy": 6.170970392227173,
      "epoch": 2.688323090430202,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043193104568002214,
      "loss": 5.9282,
      "mean_token_accuracy": 0.16123348325490952,
      "num_tokens": 46041389.0,
      "step": 13780
    },
    {
      "entropy": 6.371288871765136,
      "epoch": 2.689298605014145,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043188063565818277,
      "loss": 6.0486,
      "mean_token_accuracy": 0.1496119663119316,
      "num_tokens": 46056805.0,
      "step": 13785
    },
    {
      "entropy": 6.200106859207153,
      "epoch": 2.690274119598088,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004318302103065269,
      "loss": 5.9323,
      "mean_token_accuracy": 0.15840858817100525,
      "num_tokens": 46073219.0,
      "step": 13790
    },
    {
      "entropy": 6.303269004821777,
      "epoch": 2.691249634182031,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004317797696299818,
      "loss": 6.0617,
      "mean_token_accuracy": 0.1478768303990364,
      "num_tokens": 46089009.0,
      "step": 13795
    },
    {
      "entropy": 6.323823022842407,
      "epoch": 2.6922251487659743,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004317293136334763,
      "loss": 6.0457,
      "mean_token_accuracy": 0.14508724436163903,
      "num_tokens": 46104394.0,
      "step": 13800
    },
    {
      "entropy": 6.309295415878296,
      "epoch": 2.6932006633499173,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043167884232194094,
      "loss": 6.0567,
      "mean_token_accuracy": 0.14321453794836997,
      "num_tokens": 46121183.0,
      "step": 13805
    },
    {
      "entropy": 6.1802160263061525,
      "epoch": 2.69417617793386,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004316283557003075,
      "loss": 5.8697,
      "mean_token_accuracy": 0.16711682677268982,
      "num_tokens": 46138594.0,
      "step": 13810
    },
    {
      "entropy": 6.242872285842895,
      "epoch": 2.695151692517803,
      "grad_norm": 1.0,
      "learning_rate": 0.0004315778537735094,
      "loss": 6.0233,
      "mean_token_accuracy": 0.15076958239078522,
      "num_tokens": 46155601.0,
      "step": 13815
    },
    {
      "entropy": 6.282159090042114,
      "epoch": 2.696127207101746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043152733654648146,
      "loss": 5.9351,
      "mean_token_accuracy": 0.1542927145957947,
      "num_tokens": 46172282.0,
      "step": 13820
    },
    {
      "entropy": 6.252718734741211,
      "epoch": 2.697102721685689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043147680402416,
      "loss": 6.0182,
      "mean_token_accuracy": 0.14810230284929277,
      "num_tokens": 46188609.0,
      "step": 13825
    },
    {
      "entropy": 6.139023399353027,
      "epoch": 2.698078236269632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000431426256211483,
      "loss": 5.8468,
      "mean_token_accuracy": 0.166483074426651,
      "num_tokens": 46204624.0,
      "step": 13830
    },
    {
      "entropy": 6.186560106277466,
      "epoch": 2.699053750853575,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043137569311338977,
      "loss": 5.8645,
      "mean_token_accuracy": 0.16286583691835405,
      "num_tokens": 46221516.0,
      "step": 13835
    },
    {
      "entropy": 6.225598573684692,
      "epoch": 2.7000292654375184,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004313251147348211,
      "loss": 5.9792,
      "mean_token_accuracy": 0.1499457150697708,
      "num_tokens": 46238983.0,
      "step": 13840
    },
    {
      "entropy": 6.3537074565887455,
      "epoch": 2.7010047800214614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043127452108071947,
      "loss": 6.0257,
      "mean_token_accuracy": 0.1432358831167221,
      "num_tokens": 46254158.0,
      "step": 13845
    },
    {
      "entropy": 6.333734703063965,
      "epoch": 2.7019802946054043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004312239121560285,
      "loss": 6.0871,
      "mean_token_accuracy": 0.1407615974545479,
      "num_tokens": 46270566.0,
      "step": 13850
    },
    {
      "entropy": 6.301610803604126,
      "epoch": 2.7029558091893473,
      "grad_norm": 0.859375,
      "learning_rate": 0.00043117328796569374,
      "loss": 6.019,
      "mean_token_accuracy": 0.14550378918647766,
      "num_tokens": 46288136.0,
      "step": 13855
    },
    {
      "entropy": 6.154617357254028,
      "epoch": 2.7039313237732903,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043112264851466177,
      "loss": 5.8814,
      "mean_token_accuracy": 0.1554368667304516,
      "num_tokens": 46305418.0,
      "step": 13860
    },
    {
      "entropy": 6.2809265613555905,
      "epoch": 2.7049068383572337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004310719938078811,
      "loss": 6.0022,
      "mean_token_accuracy": 0.14634008556604386,
      "num_tokens": 46321897.0,
      "step": 13865
    },
    {
      "entropy": 6.342752742767334,
      "epoch": 2.7058823529411766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004310213238503014,
      "loss": 6.1222,
      "mean_token_accuracy": 0.1457255356013775,
      "num_tokens": 46338936.0,
      "step": 13870
    },
    {
      "entropy": 6.282695198059082,
      "epoch": 2.7068578675251196,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004309706386468741,
      "loss": 5.9433,
      "mean_token_accuracy": 0.1463232882320881,
      "num_tokens": 46354453.0,
      "step": 13875
    },
    {
      "entropy": 6.213294982910156,
      "epoch": 2.7078333821090625,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043091993820255177,
      "loss": 5.9738,
      "mean_token_accuracy": 0.14886244535446166,
      "num_tokens": 46369441.0,
      "step": 13880
    },
    {
      "entropy": 6.225582265853882,
      "epoch": 2.7088088966930055,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00043086922252228886,
      "loss": 5.9132,
      "mean_token_accuracy": 0.15595605671405793,
      "num_tokens": 46386273.0,
      "step": 13885
    },
    {
      "entropy": 6.1641417980194095,
      "epoch": 2.7097844112769485,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043081849161104096,
      "loss": 5.9451,
      "mean_token_accuracy": 0.15429385453462602,
      "num_tokens": 46403274.0,
      "step": 13890
    },
    {
      "entropy": 6.1988990783691404,
      "epoch": 2.7107599258608914,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043076774547376554,
      "loss": 5.996,
      "mean_token_accuracy": 0.15462515652179717,
      "num_tokens": 46422601.0,
      "step": 13895
    },
    {
      "entropy": 6.291223573684692,
      "epoch": 2.7117354404448344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004307169841154211,
      "loss": 6.0218,
      "mean_token_accuracy": 0.14766983091831207,
      "num_tokens": 46437861.0,
      "step": 13900
    },
    {
      "entropy": 6.349146127700806,
      "epoch": 2.712710955028778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043066620754096793,
      "loss": 6.0593,
      "mean_token_accuracy": 0.14336006119847297,
      "num_tokens": 46454558.0,
      "step": 13905
    },
    {
      "entropy": 6.170024108886719,
      "epoch": 2.7136864696127208,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004306154157553679,
      "loss": 5.8464,
      "mean_token_accuracy": 0.16552362889051436,
      "num_tokens": 46470975.0,
      "step": 13910
    },
    {
      "entropy": 6.195564270019531,
      "epoch": 2.7146619841966637,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043056460876358396,
      "loss": 5.9587,
      "mean_token_accuracy": 0.1563336506485939,
      "num_tokens": 46487791.0,
      "step": 13915
    },
    {
      "entropy": 6.164254426956177,
      "epoch": 2.7156374987806067,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004305137865705809,
      "loss": 5.8507,
      "mean_token_accuracy": 0.16007353812456132,
      "num_tokens": 46504668.0,
      "step": 13920
    },
    {
      "entropy": 6.231702709197998,
      "epoch": 2.7166130133645496,
      "grad_norm": 1.0,
      "learning_rate": 0.00043046294918132496,
      "loss": 5.9578,
      "mean_token_accuracy": 0.14800213873386384,
      "num_tokens": 46519994.0,
      "step": 13925
    },
    {
      "entropy": 6.122684717178345,
      "epoch": 2.717588527948493,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004304120966007835,
      "loss": 5.905,
      "mean_token_accuracy": 0.1591530904173851,
      "num_tokens": 46537407.0,
      "step": 13930
    },
    {
      "entropy": 6.268473482131958,
      "epoch": 2.718564042532436,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000430361228833926,
      "loss": 5.9767,
      "mean_token_accuracy": 0.1560317173600197,
      "num_tokens": 46555518.0,
      "step": 13935
    },
    {
      "entropy": 6.282375764846802,
      "epoch": 2.719539557116379,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043031034588572294,
      "loss": 6.018,
      "mean_token_accuracy": 0.14537838995456695,
      "num_tokens": 46573404.0,
      "step": 13940
    },
    {
      "entropy": 6.162353610992431,
      "epoch": 2.720515071700322,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004302594477611464,
      "loss": 5.8512,
      "mean_token_accuracy": 0.1593487098813057,
      "num_tokens": 46588634.0,
      "step": 13945
    },
    {
      "entropy": 6.187953472137451,
      "epoch": 2.721490586284265,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043020853446517005,
      "loss": 5.9431,
      "mean_token_accuracy": 0.15483950078487396,
      "num_tokens": 46606105.0,
      "step": 13950
    },
    {
      "entropy": 6.224151849746704,
      "epoch": 2.722466100868208,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043015760600276877,
      "loss": 5.8555,
      "mean_token_accuracy": 0.1628679007291794,
      "num_tokens": 46623586.0,
      "step": 13955
    },
    {
      "entropy": 6.186040210723877,
      "epoch": 2.723441615452151,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00043010666237891927,
      "loss": 5.913,
      "mean_token_accuracy": 0.15840320885181428,
      "num_tokens": 46640198.0,
      "step": 13960
    },
    {
      "entropy": 6.19617075920105,
      "epoch": 2.724417130036094,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004300557035985995,
      "loss": 6.0276,
      "mean_token_accuracy": 0.14880677610635756,
      "num_tokens": 46656892.0,
      "step": 13965
    },
    {
      "entropy": 6.296930122375488,
      "epoch": 2.725392644620037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043000472966678906,
      "loss": 5.9996,
      "mean_token_accuracy": 0.15561263859272004,
      "num_tokens": 46673796.0,
      "step": 13970
    },
    {
      "entropy": 6.158070421218872,
      "epoch": 2.72636815920398,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004299537405884689,
      "loss": 5.8629,
      "mean_token_accuracy": 0.16362271159887315,
      "num_tokens": 46691399.0,
      "step": 13975
    },
    {
      "entropy": 6.147817468643188,
      "epoch": 2.727343673787923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042990273636862137,
      "loss": 5.854,
      "mean_token_accuracy": 0.15416584461927413,
      "num_tokens": 46708738.0,
      "step": 13980
    },
    {
      "entropy": 6.293470001220703,
      "epoch": 2.728319188371866,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004298517170122307,
      "loss": 6.015,
      "mean_token_accuracy": 0.1513020932674408,
      "num_tokens": 46725481.0,
      "step": 13985
    },
    {
      "entropy": 6.2503642559051515,
      "epoch": 2.7292947029558094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042980068252428203,
      "loss": 5.9533,
      "mean_token_accuracy": 0.15050876140594482,
      "num_tokens": 46741758.0,
      "step": 13990
    },
    {
      "entropy": 6.288397693634034,
      "epoch": 2.7302702175397524,
      "grad_norm": 1.0,
      "learning_rate": 0.00042974963290976235,
      "loss": 6.0456,
      "mean_token_accuracy": 0.14406616985797882,
      "num_tokens": 46758448.0,
      "step": 13995
    },
    {
      "entropy": 6.213140678405762,
      "epoch": 2.7312457321236954,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004296985681736601,
      "loss": 5.938,
      "mean_token_accuracy": 0.14957604855298995,
      "num_tokens": 46775261.0,
      "step": 14000
    },
    {
      "entropy": 6.204946708679199,
      "epoch": 2.7322212467076383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042964748832096517,
      "loss": 5.8947,
      "mean_token_accuracy": 0.15721533745527266,
      "num_tokens": 46792484.0,
      "step": 14005
    },
    {
      "entropy": 6.288388252258301,
      "epoch": 2.7331967612915813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004295963933566688,
      "loss": 6.0658,
      "mean_token_accuracy": 0.13781430721282958,
      "num_tokens": 46810078.0,
      "step": 14010
    },
    {
      "entropy": 6.228634357452393,
      "epoch": 2.7341722758755242,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004295452832857639,
      "loss": 5.9197,
      "mean_token_accuracy": 0.15069997310638428,
      "num_tokens": 46826524.0,
      "step": 14015
    },
    {
      "entropy": 6.217752885818482,
      "epoch": 2.735147790459467,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004294941581132447,
      "loss": 5.9193,
      "mean_token_accuracy": 0.1554874062538147,
      "num_tokens": 46842939.0,
      "step": 14020
    },
    {
      "entropy": 6.230658626556396,
      "epoch": 2.73612330504341,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042944301784410705,
      "loss": 6.0826,
      "mean_token_accuracy": 0.14324986040592194,
      "num_tokens": 46858060.0,
      "step": 14025
    },
    {
      "entropy": 6.230405235290528,
      "epoch": 2.7370988196273536,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004293918624833481,
      "loss": 5.9302,
      "mean_token_accuracy": 0.15402874499559402,
      "num_tokens": 46873434.0,
      "step": 14030
    },
    {
      "entropy": 6.293866968154907,
      "epoch": 2.7380743342112965,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004293406920359665,
      "loss": 6.0515,
      "mean_token_accuracy": 0.15064638033509253,
      "num_tokens": 46890689.0,
      "step": 14035
    },
    {
      "entropy": 6.367683076858521,
      "epoch": 2.7390498487952395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042928950650696274,
      "loss": 6.0458,
      "mean_token_accuracy": 0.14491432011127472,
      "num_tokens": 46906384.0,
      "step": 14040
    },
    {
      "entropy": 6.225796747207641,
      "epoch": 2.7400253633791825,
      "grad_norm": 1.0,
      "learning_rate": 0.00042923830590133825,
      "loss": 6.0424,
      "mean_token_accuracy": 0.14755972176790239,
      "num_tokens": 46922154.0,
      "step": 14045
    },
    {
      "entropy": 6.350451135635376,
      "epoch": 2.7410008779631254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042918709022409624,
      "loss": 6.0689,
      "mean_token_accuracy": 0.140258264541626,
      "num_tokens": 46939213.0,
      "step": 14050
    },
    {
      "entropy": 6.258037137985229,
      "epoch": 2.741976392547069,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004291358594802412,
      "loss": 5.9813,
      "mean_token_accuracy": 0.15142300128936767,
      "num_tokens": 46956207.0,
      "step": 14055
    },
    {
      "entropy": 6.291506433486939,
      "epoch": 2.742951907131012,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004290846136747795,
      "loss": 6.0371,
      "mean_token_accuracy": 0.15100251734256745,
      "num_tokens": 46972228.0,
      "step": 14060
    },
    {
      "entropy": 6.352012634277344,
      "epoch": 2.7439274217149547,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004290333528127184,
      "loss": 6.0511,
      "mean_token_accuracy": 0.1506814554333687,
      "num_tokens": 46988398.0,
      "step": 14065
    },
    {
      "entropy": 6.216901397705078,
      "epoch": 2.7449029362988977,
      "grad_norm": 0.921875,
      "learning_rate": 0.000428982076899067,
      "loss": 5.9163,
      "mean_token_accuracy": 0.1591764897108078,
      "num_tokens": 47006178.0,
      "step": 14070
    },
    {
      "entropy": 6.236685562133789,
      "epoch": 2.7458784508828407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004289307859388358,
      "loss": 6.0777,
      "mean_token_accuracy": 0.14751118719577788,
      "num_tokens": 47022703.0,
      "step": 14075
    },
    {
      "entropy": 6.26066780090332,
      "epoch": 2.7468539654667836,
      "grad_norm": 1.046875,
      "learning_rate": 0.000428879479937037,
      "loss": 6.0588,
      "mean_token_accuracy": 0.14729162007570268,
      "num_tokens": 47038164.0,
      "step": 14080
    },
    {
      "entropy": 6.21960940361023,
      "epoch": 2.7478294800507266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004288281588986838,
      "loss": 5.9756,
      "mean_token_accuracy": 0.14948850125074387,
      "num_tokens": 47055080.0,
      "step": 14085
    },
    {
      "entropy": 6.374417495727539,
      "epoch": 2.7488049946346695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004287768228287912,
      "loss": 6.0202,
      "mean_token_accuracy": 0.14467756897211076,
      "num_tokens": 47071121.0,
      "step": 14090
    },
    {
      "entropy": 6.043356037139892,
      "epoch": 2.749780509218613,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004287254717323754,
      "loss": 5.69,
      "mean_token_accuracy": 0.18274304047226905,
      "num_tokens": 47088303.0,
      "step": 14095
    },
    {
      "entropy": 6.306983613967896,
      "epoch": 2.750756023802556,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042867410561445445,
      "loss": 6.0879,
      "mean_token_accuracy": 0.14324308335781097,
      "num_tokens": 47104637.0,
      "step": 14100
    },
    {
      "entropy": 6.3412103176116945,
      "epoch": 2.751731538386499,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004286227244800477,
      "loss": 6.0642,
      "mean_token_accuracy": 0.14634467512369156,
      "num_tokens": 47121234.0,
      "step": 14105
    },
    {
      "entropy": 6.095263910293579,
      "epoch": 2.752707052970442,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042857132833417575,
      "loss": 5.8872,
      "mean_token_accuracy": 0.16767469942569732,
      "num_tokens": 47139264.0,
      "step": 14110
    },
    {
      "entropy": 6.180713510513305,
      "epoch": 2.753682567554385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004285199171818609,
      "loss": 5.8255,
      "mean_token_accuracy": 0.15847878977656366,
      "num_tokens": 47155546.0,
      "step": 14115
    },
    {
      "entropy": 6.2306554317474365,
      "epoch": 2.754658082138328,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042846849102812695,
      "loss": 5.9449,
      "mean_token_accuracy": 0.15078910440206528,
      "num_tokens": 47171468.0,
      "step": 14120
    },
    {
      "entropy": 6.115953874588013,
      "epoch": 2.755633596722271,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004284170498779991,
      "loss": 5.8181,
      "mean_token_accuracy": 0.16409843862056733,
      "num_tokens": 47187937.0,
      "step": 14125
    },
    {
      "entropy": 6.244295930862426,
      "epoch": 2.756609111306214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004283655937365038,
      "loss": 5.9548,
      "mean_token_accuracy": 0.15630403980612756,
      "num_tokens": 47204306.0,
      "step": 14130
    },
    {
      "entropy": 6.204940128326416,
      "epoch": 2.757584625890157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004283141226086693,
      "loss": 5.9187,
      "mean_token_accuracy": 0.15545208901166915,
      "num_tokens": 47222040.0,
      "step": 14135
    },
    {
      "entropy": 6.154622888565063,
      "epoch": 2.7585601404741,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004282626364995252,
      "loss": 5.8528,
      "mean_token_accuracy": 0.15775507539510727,
      "num_tokens": 47237889.0,
      "step": 14140
    },
    {
      "entropy": 6.147297716140747,
      "epoch": 2.759535655058043,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042821113541410247,
      "loss": 5.9462,
      "mean_token_accuracy": 0.1511112555861473,
      "num_tokens": 47253752.0,
      "step": 14145
    },
    {
      "entropy": 6.242524337768555,
      "epoch": 2.760511169641986,
      "grad_norm": 1.0,
      "learning_rate": 0.00042815961935743367,
      "loss": 5.9539,
      "mean_token_accuracy": 0.1511687383055687,
      "num_tokens": 47271469.0,
      "step": 14150
    },
    {
      "entropy": 6.248548364639282,
      "epoch": 2.761486684225929,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042810808833455274,
      "loss": 6.0105,
      "mean_token_accuracy": 0.1551550105214119,
      "num_tokens": 47288746.0,
      "step": 14155
    },
    {
      "entropy": 6.254455184936523,
      "epoch": 2.7624621988098723,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00042805654235049514,
      "loss": 5.9348,
      "mean_token_accuracy": 0.15668729692697525,
      "num_tokens": 47308101.0,
      "step": 14160
    },
    {
      "entropy": 6.280220603942871,
      "epoch": 2.7634377133938153,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004280049814102976,
      "loss": 6.0194,
      "mean_token_accuracy": 0.15045560598373414,
      "num_tokens": 47324441.0,
      "step": 14165
    },
    {
      "entropy": 6.294740056991577,
      "epoch": 2.7644132279777582,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004279534055189987,
      "loss": 6.0276,
      "mean_token_accuracy": 0.15112961307168007,
      "num_tokens": 47341664.0,
      "step": 14170
    },
    {
      "entropy": 6.254258012771606,
      "epoch": 2.765388742561701,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042790181468163813,
      "loss": 6.0514,
      "mean_token_accuracy": 0.14850770831108093,
      "num_tokens": 47358424.0,
      "step": 14175
    },
    {
      "entropy": 6.176885032653809,
      "epoch": 2.7663642571456446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004278502089032572,
      "loss": 5.8983,
      "mean_token_accuracy": 0.15907283127307892,
      "num_tokens": 47375211.0,
      "step": 14180
    },
    {
      "entropy": 6.166579675674439,
      "epoch": 2.7673397717295876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004277985881888986,
      "loss": 5.8625,
      "mean_token_accuracy": 0.1588609531521797,
      "num_tokens": 47393947.0,
      "step": 14185
    },
    {
      "entropy": 6.345116662979126,
      "epoch": 2.7683152863135305,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004277469525436065,
      "loss": 6.0231,
      "mean_token_accuracy": 0.14439835920929908,
      "num_tokens": 47410519.0,
      "step": 14190
    },
    {
      "entropy": 6.245634412765503,
      "epoch": 2.7692908008974735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042769530197242674,
      "loss": 5.9827,
      "mean_token_accuracy": 0.15034462213516236,
      "num_tokens": 47426672.0,
      "step": 14195
    },
    {
      "entropy": 6.354568004608154,
      "epoch": 2.7702663154814164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004276436364804061,
      "loss": 6.1026,
      "mean_token_accuracy": 0.14300211220979692,
      "num_tokens": 47443545.0,
      "step": 14200
    },
    {
      "entropy": 6.306677055358887,
      "epoch": 2.7712418300653594,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042759195607259344,
      "loss": 6.026,
      "mean_token_accuracy": 0.1447511188685894,
      "num_tokens": 47459543.0,
      "step": 14205
    },
    {
      "entropy": 6.263290786743164,
      "epoch": 2.7722173446493024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042754026075403866,
      "loss": 6.0116,
      "mean_token_accuracy": 0.14618819504976271,
      "num_tokens": 47476240.0,
      "step": 14210
    },
    {
      "entropy": 6.137014675140381,
      "epoch": 2.7731928592332453,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042748855052979326,
      "loss": 5.8241,
      "mean_token_accuracy": 0.17489274740219116,
      "num_tokens": 47493855.0,
      "step": 14215
    },
    {
      "entropy": 6.152178621292114,
      "epoch": 2.7741683738171887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042743682540491017,
      "loss": 5.9114,
      "mean_token_accuracy": 0.1546615719795227,
      "num_tokens": 47510175.0,
      "step": 14220
    },
    {
      "entropy": 6.163287734985351,
      "epoch": 2.7751438884011317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042738508538444383,
      "loss": 5.871,
      "mean_token_accuracy": 0.1566540263593197,
      "num_tokens": 47528346.0,
      "step": 14225
    },
    {
      "entropy": 6.325794410705567,
      "epoch": 2.7761194029850746,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042733333047345005,
      "loss": 5.9726,
      "mean_token_accuracy": 0.15593766719102858,
      "num_tokens": 47545069.0,
      "step": 14230
    },
    {
      "entropy": 6.123036670684814,
      "epoch": 2.7770949175690176,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042728156067698607,
      "loss": 5.7991,
      "mean_token_accuracy": 0.1701396957039833,
      "num_tokens": 47562255.0,
      "step": 14235
    },
    {
      "entropy": 6.267027902603149,
      "epoch": 2.7780704321529606,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004272297760001108,
      "loss": 5.968,
      "mean_token_accuracy": 0.15189952105283738,
      "num_tokens": 47577940.0,
      "step": 14240
    },
    {
      "entropy": 6.306848382949829,
      "epoch": 2.779045946736904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004271779764478844,
      "loss": 6.004,
      "mean_token_accuracy": 0.1468572273850441,
      "num_tokens": 47593728.0,
      "step": 14245
    },
    {
      "entropy": 6.2137610912323,
      "epoch": 2.780021461320847,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042712616202536844,
      "loss": 6.0228,
      "mean_token_accuracy": 0.15945882573723794,
      "num_tokens": 47610367.0,
      "step": 14250
    },
    {
      "entropy": 6.3204528331756595,
      "epoch": 2.78099697590479,
      "grad_norm": 1.0,
      "learning_rate": 0.00042707433273762613,
      "loss": 6.0957,
      "mean_token_accuracy": 0.14594335705041886,
      "num_tokens": 47628690.0,
      "step": 14255
    },
    {
      "entropy": 6.280133485794067,
      "epoch": 2.781972490488733,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042702248858972206,
      "loss": 6.0135,
      "mean_token_accuracy": 0.1520581178367138,
      "num_tokens": 47643787.0,
      "step": 14260
    },
    {
      "entropy": 6.140622091293335,
      "epoch": 2.782948005072676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042697062958672215,
      "loss": 5.8853,
      "mean_token_accuracy": 0.16206456869840621,
      "num_tokens": 47660163.0,
      "step": 14265
    },
    {
      "entropy": 6.328909683227539,
      "epoch": 2.7839235196566188,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004269187557336941,
      "loss": 6.0537,
      "mean_token_accuracy": 0.14676631838083268,
      "num_tokens": 47675959.0,
      "step": 14270
    },
    {
      "entropy": 6.383520793914795,
      "epoch": 2.7848990342405617,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00042686686703570655,
      "loss": 6.1173,
      "mean_token_accuracy": 0.1407186672091484,
      "num_tokens": 47693318.0,
      "step": 14275
    },
    {
      "entropy": 6.314073276519776,
      "epoch": 2.7858745488245047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004268149634978301,
      "loss": 6.0581,
      "mean_token_accuracy": 0.14372048974037172,
      "num_tokens": 47709130.0,
      "step": 14280
    },
    {
      "entropy": 6.306469058990478,
      "epoch": 2.786850063408448,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004267630451251364,
      "loss": 6.0846,
      "mean_token_accuracy": 0.1405602902173996,
      "num_tokens": 47725015.0,
      "step": 14285
    },
    {
      "entropy": 6.3256713390350345,
      "epoch": 2.787825577992391,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004267111119226989,
      "loss": 6.0498,
      "mean_token_accuracy": 0.14702777862548827,
      "num_tokens": 47741322.0,
      "step": 14290
    },
    {
      "entropy": 6.257889461517334,
      "epoch": 2.788801092576334,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042665916389559216,
      "loss": 5.9759,
      "mean_token_accuracy": 0.15046907514333724,
      "num_tokens": 47759252.0,
      "step": 14295
    },
    {
      "entropy": 6.295649290084839,
      "epoch": 2.789776607160277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042660720104889256,
      "loss": 6.0085,
      "mean_token_accuracy": 0.1497651979327202,
      "num_tokens": 47775258.0,
      "step": 14300
    },
    {
      "entropy": 6.1861003875732425,
      "epoch": 2.79075212174422,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042655522338767756,
      "loss": 5.9276,
      "mean_token_accuracy": 0.1530047044157982,
      "num_tokens": 47791072.0,
      "step": 14305
    },
    {
      "entropy": 6.319885778427124,
      "epoch": 2.7917276363281633,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004265032309170263,
      "loss": 6.0611,
      "mean_token_accuracy": 0.14946260154247284,
      "num_tokens": 47807946.0,
      "step": 14310
    },
    {
      "entropy": 6.323369693756104,
      "epoch": 2.7927031509121063,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004264512236420193,
      "loss": 6.0532,
      "mean_token_accuracy": 0.1429503493010998,
      "num_tokens": 47825295.0,
      "step": 14315
    },
    {
      "entropy": 6.356843137741089,
      "epoch": 2.7936786654960493,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004263992015677384,
      "loss": 6.078,
      "mean_token_accuracy": 0.14610934555530547,
      "num_tokens": 47841302.0,
      "step": 14320
    },
    {
      "entropy": 6.315498924255371,
      "epoch": 2.7946541800799922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042634716469926723,
      "loss": 5.9872,
      "mean_token_accuracy": 0.1482444941997528,
      "num_tokens": 47856871.0,
      "step": 14325
    },
    {
      "entropy": 6.208625316619873,
      "epoch": 2.795629694663935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004262951130416905,
      "loss": 5.9811,
      "mean_token_accuracy": 0.1525020867586136,
      "num_tokens": 47872842.0,
      "step": 14330
    },
    {
      "entropy": 6.256116724014282,
      "epoch": 2.796605209247878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004262430466000946,
      "loss": 6.0119,
      "mean_token_accuracy": 0.15051274523139,
      "num_tokens": 47889093.0,
      "step": 14335
    },
    {
      "entropy": 6.378304529190063,
      "epoch": 2.797580723831821,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042619096537956715,
      "loss": 6.0694,
      "mean_token_accuracy": 0.14408605843782424,
      "num_tokens": 47904953.0,
      "step": 14340
    },
    {
      "entropy": 6.2262977123260494,
      "epoch": 2.798556238415764,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004261388693851974,
      "loss": 5.9506,
      "mean_token_accuracy": 0.1459345906972885,
      "num_tokens": 47921591.0,
      "step": 14345
    },
    {
      "entropy": 6.285395097732544,
      "epoch": 2.7995317529997075,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004260867586220761,
      "loss": 6.0268,
      "mean_token_accuracy": 0.14418798834085464,
      "num_tokens": 47938463.0,
      "step": 14350
    },
    {
      "entropy": 6.132035589218139,
      "epoch": 2.8005072675836504,
      "grad_norm": 0.84375,
      "learning_rate": 0.00042603463309529515,
      "loss": 5.822,
      "mean_token_accuracy": 0.1685475617647171,
      "num_tokens": 47956247.0,
      "step": 14355
    },
    {
      "entropy": 6.274279260635376,
      "epoch": 2.8014827821675934,
      "grad_norm": 1.0,
      "learning_rate": 0.00042598249280994817,
      "loss": 5.9957,
      "mean_token_accuracy": 0.15175690948963166,
      "num_tokens": 47973305.0,
      "step": 14360
    },
    {
      "entropy": 6.293311214447021,
      "epoch": 2.8024582967515363,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042593033777113017,
      "loss": 5.9331,
      "mean_token_accuracy": 0.16335161179304122,
      "num_tokens": 47989674.0,
      "step": 14365
    },
    {
      "entropy": 6.2166626930236815,
      "epoch": 2.8034338113354793,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004258781679839374,
      "loss": 5.9295,
      "mean_token_accuracy": 0.1521068975329399,
      "num_tokens": 48004814.0,
      "step": 14370
    },
    {
      "entropy": 6.262455987930298,
      "epoch": 2.8044093259194227,
      "grad_norm": 1.0,
      "learning_rate": 0.00042582598345346787,
      "loss": 5.99,
      "mean_token_accuracy": 0.15162694007158278,
      "num_tokens": 48021174.0,
      "step": 14375
    },
    {
      "entropy": 6.283260631561279,
      "epoch": 2.8053848405033657,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004257737841848207,
      "loss": 6.0551,
      "mean_token_accuracy": 0.14559864103794098,
      "num_tokens": 48038179.0,
      "step": 14380
    },
    {
      "entropy": 6.323927736282348,
      "epoch": 2.8063603550873086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042572157018309675,
      "loss": 6.015,
      "mean_token_accuracy": 0.14929332882165908,
      "num_tokens": 48054933.0,
      "step": 14385
    },
    {
      "entropy": 6.196407842636108,
      "epoch": 2.8073358696712516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042566934145339816,
      "loss": 5.9503,
      "mean_token_accuracy": 0.15831852108240127,
      "num_tokens": 48071908.0,
      "step": 14390
    },
    {
      "entropy": 6.2944793701171875,
      "epoch": 2.8083113842551946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042561709800082847,
      "loss": 6.0183,
      "mean_token_accuracy": 0.1447554811835289,
      "num_tokens": 48088385.0,
      "step": 14395
    },
    {
      "entropy": 6.1927001953125,
      "epoch": 2.8092868988391375,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004255648398304928,
      "loss": 5.8924,
      "mean_token_accuracy": 0.16171323284506797,
      "num_tokens": 48105331.0,
      "step": 14400
    },
    {
      "entropy": 6.029235696792602,
      "epoch": 2.8102624134230805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042551256694749764,
      "loss": 5.7583,
      "mean_token_accuracy": 0.17291318327188493,
      "num_tokens": 48123092.0,
      "step": 14405
    },
    {
      "entropy": 6.238957166671753,
      "epoch": 2.8112379280070234,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004254602793569509,
      "loss": 5.9422,
      "mean_token_accuracy": 0.15361684411764145,
      "num_tokens": 48140096.0,
      "step": 14410
    },
    {
      "entropy": 6.315847444534302,
      "epoch": 2.812213442590967,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004254079770639618,
      "loss": 6.0649,
      "mean_token_accuracy": 0.14537217244505882,
      "num_tokens": 48156427.0,
      "step": 14415
    },
    {
      "entropy": 6.385224676132202,
      "epoch": 2.81318895717491,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004253556600736413,
      "loss": 6.0874,
      "mean_token_accuracy": 0.14568108841776847,
      "num_tokens": 48172867.0,
      "step": 14420
    },
    {
      "entropy": 6.24599061012268,
      "epoch": 2.8141644717588528,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042530332839110156,
      "loss": 5.9861,
      "mean_token_accuracy": 0.15245443135499953,
      "num_tokens": 48189106.0,
      "step": 14425
    },
    {
      "entropy": 6.266282463073731,
      "epoch": 2.8151399863427957,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004252509820214563,
      "loss": 5.9868,
      "mean_token_accuracy": 0.14558991491794587,
      "num_tokens": 48205400.0,
      "step": 14430
    },
    {
      "entropy": 6.283788204193115,
      "epoch": 2.816115500926739,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042519862096982054,
      "loss": 6.0312,
      "mean_token_accuracy": 0.1507223889231682,
      "num_tokens": 48221937.0,
      "step": 14435
    },
    {
      "entropy": 6.344908237457275,
      "epoch": 2.817091015510682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004251462452413109,
      "loss": 6.0237,
      "mean_token_accuracy": 0.14841895997524263,
      "num_tokens": 48238060.0,
      "step": 14440
    },
    {
      "entropy": 6.142515707015991,
      "epoch": 2.818066530094625,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042509385484104526,
      "loss": 5.8917,
      "mean_token_accuracy": 0.156676784157753,
      "num_tokens": 48255813.0,
      "step": 14445
    },
    {
      "entropy": 6.209969806671142,
      "epoch": 2.819042044678568,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042504144977414303,
      "loss": 5.9246,
      "mean_token_accuracy": 0.15752413868904114,
      "num_tokens": 48272989.0,
      "step": 14450
    },
    {
      "entropy": 6.30368185043335,
      "epoch": 2.820017559262511,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042498903004572515,
      "loss": 5.9962,
      "mean_token_accuracy": 0.15020231753587723,
      "num_tokens": 48289415.0,
      "step": 14455
    },
    {
      "entropy": 6.180605030059814,
      "epoch": 2.820993073846454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004249365956609139,
      "loss": 5.9156,
      "mean_token_accuracy": 0.15705854594707488,
      "num_tokens": 48306740.0,
      "step": 14460
    },
    {
      "entropy": 6.207580709457398,
      "epoch": 2.821968588430397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004248841466248327,
      "loss": 5.9374,
      "mean_token_accuracy": 0.1504146859049797,
      "num_tokens": 48322934.0,
      "step": 14465
    },
    {
      "entropy": 6.231025838851929,
      "epoch": 2.82294410301434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042483168294260706,
      "loss": 5.9667,
      "mean_token_accuracy": 0.15620044022798538,
      "num_tokens": 48338835.0,
      "step": 14470
    },
    {
      "entropy": 6.147571659088134,
      "epoch": 2.8239196175982832,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004247792046193633,
      "loss": 5.8654,
      "mean_token_accuracy": 0.1550082191824913,
      "num_tokens": 48355996.0,
      "step": 14475
    },
    {
      "entropy": 6.379227066040039,
      "epoch": 2.824895132182226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042472671166022955,
      "loss": 6.1357,
      "mean_token_accuracy": 0.13951852023601533,
      "num_tokens": 48372775.0,
      "step": 14480
    },
    {
      "entropy": 6.307221364974976,
      "epoch": 2.825870646766169,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004246742040703351,
      "loss": 5.9763,
      "mean_token_accuracy": 0.14925282150506974,
      "num_tokens": 48389018.0,
      "step": 14485
    },
    {
      "entropy": 6.227575206756592,
      "epoch": 2.826846161350112,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004246216818548109,
      "loss": 5.9733,
      "mean_token_accuracy": 0.1552024506032467,
      "num_tokens": 48404298.0,
      "step": 14490
    },
    {
      "entropy": 6.253101015090943,
      "epoch": 2.827821675934055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042456914501878926,
      "loss": 6.0052,
      "mean_token_accuracy": 0.14760172963142396,
      "num_tokens": 48421477.0,
      "step": 14495
    },
    {
      "entropy": 6.292385578155518,
      "epoch": 2.8287971905179985,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004245165935674038,
      "loss": 5.9764,
      "mean_token_accuracy": 0.1549450233578682,
      "num_tokens": 48439357.0,
      "step": 14500
    },
    {
      "entropy": 6.212532567977905,
      "epoch": 2.8297727051019415,
      "grad_norm": 1.0,
      "learning_rate": 0.00042446402750578974,
      "loss": 5.9033,
      "mean_token_accuracy": 0.15282618552446364,
      "num_tokens": 48454109.0,
      "step": 14505
    },
    {
      "entropy": 6.234435081481934,
      "epoch": 2.8307482196858844,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042441144683908356,
      "loss": 5.9845,
      "mean_token_accuracy": 0.14737728759646415,
      "num_tokens": 48470274.0,
      "step": 14510
    },
    {
      "entropy": 6.228418207168579,
      "epoch": 2.8317237342698274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042435885157242326,
      "loss": 5.9219,
      "mean_token_accuracy": 0.1531671941280365,
      "num_tokens": 48485190.0,
      "step": 14515
    },
    {
      "entropy": 6.276603364944458,
      "epoch": 2.8326992488537703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004243062417109484,
      "loss": 6.0114,
      "mean_token_accuracy": 0.14706870168447495,
      "num_tokens": 48501757.0,
      "step": 14520
    },
    {
      "entropy": 6.265215110778809,
      "epoch": 2.8336747634377133,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042425361725979973,
      "loss": 5.9945,
      "mean_token_accuracy": 0.1478548526763916,
      "num_tokens": 48518539.0,
      "step": 14525
    },
    {
      "entropy": 6.3145349502563475,
      "epoch": 2.8346502780216563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042420097822411945,
      "loss": 5.9814,
      "mean_token_accuracy": 0.14563742876052857,
      "num_tokens": 48535575.0,
      "step": 14530
    },
    {
      "entropy": 6.234805727005005,
      "epoch": 2.835625792605599,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042414832460905145,
      "loss": 5.9784,
      "mean_token_accuracy": 0.15164869725704194,
      "num_tokens": 48552569.0,
      "step": 14535
    },
    {
      "entropy": 6.271487045288086,
      "epoch": 2.8366013071895426,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004240956564197407,
      "loss": 6.0888,
      "mean_token_accuracy": 0.14863712340593338,
      "num_tokens": 48571151.0,
      "step": 14540
    },
    {
      "entropy": 6.238283491134643,
      "epoch": 2.8375768217734856,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042404297366133376,
      "loss": 5.8881,
      "mean_token_accuracy": 0.15609751343727113,
      "num_tokens": 48588228.0,
      "step": 14545
    },
    {
      "entropy": 6.156391334533692,
      "epoch": 2.8385523363574285,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004239902763389786,
      "loss": 5.9525,
      "mean_token_accuracy": 0.14819619208574294,
      "num_tokens": 48604936.0,
      "step": 14550
    },
    {
      "entropy": 6.298422050476074,
      "epoch": 2.8395278509413715,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042393756445782466,
      "loss": 5.9633,
      "mean_token_accuracy": 0.15301454961299896,
      "num_tokens": 48622064.0,
      "step": 14555
    },
    {
      "entropy": 6.197825479507446,
      "epoch": 2.8405033655253145,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042388483802302285,
      "loss": 5.9028,
      "mean_token_accuracy": 0.15383419692516326,
      "num_tokens": 48638161.0,
      "step": 14560
    },
    {
      "entropy": 6.229183340072632,
      "epoch": 2.841478880109258,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042383209703972523,
      "loss": 5.9386,
      "mean_token_accuracy": 0.1569916769862175,
      "num_tokens": 48654521.0,
      "step": 14565
    },
    {
      "entropy": 6.25848240852356,
      "epoch": 2.842454394693201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042377934151308555,
      "loss": 5.9939,
      "mean_token_accuracy": 0.15142636895179748,
      "num_tokens": 48671083.0,
      "step": 14570
    },
    {
      "entropy": 6.373251008987427,
      "epoch": 2.843429909277144,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042372657144825886,
      "loss": 6.0426,
      "mean_token_accuracy": 0.1487225130200386,
      "num_tokens": 48687917.0,
      "step": 14575
    },
    {
      "entropy": 6.248176574707031,
      "epoch": 2.8444054238610867,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004236737868504017,
      "loss": 6.0214,
      "mean_token_accuracy": 0.1515406385064125,
      "num_tokens": 48703626.0,
      "step": 14580
    },
    {
      "entropy": 6.286380910873413,
      "epoch": 2.8453809384450297,
      "grad_norm": 1.046875,
      "learning_rate": 0.000423620987724672,
      "loss": 6.045,
      "mean_token_accuracy": 0.14219587594270705,
      "num_tokens": 48720020.0,
      "step": 14585
    },
    {
      "entropy": 6.3597959041595455,
      "epoch": 2.8463564530289727,
      "grad_norm": 0.890625,
      "learning_rate": 0.000423568174076229,
      "loss": 6.0723,
      "mean_token_accuracy": 0.14285866692662239,
      "num_tokens": 48737185.0,
      "step": 14590
    },
    {
      "entropy": 6.208530378341675,
      "epoch": 2.8473319676129156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042351534591023363,
      "loss": 5.8906,
      "mean_token_accuracy": 0.15731287747621536,
      "num_tokens": 48754106.0,
      "step": 14595
    },
    {
      "entropy": 6.196780776977539,
      "epoch": 2.8483074821968586,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004234625032318479,
      "loss": 5.9256,
      "mean_token_accuracy": 0.15754527300596238,
      "num_tokens": 48770847.0,
      "step": 14600
    },
    {
      "entropy": 6.189958000183106,
      "epoch": 2.849282996780802,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00042340964604623547,
      "loss": 5.897,
      "mean_token_accuracy": 0.15768026411533356,
      "num_tokens": 48787891.0,
      "step": 14605
    },
    {
      "entropy": 6.27116060256958,
      "epoch": 2.850258511364745,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004233567743585613,
      "loss": 6.022,
      "mean_token_accuracy": 0.14967958331108094,
      "num_tokens": 48804418.0,
      "step": 14610
    },
    {
      "entropy": 6.257380962371826,
      "epoch": 2.851234025948688,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004233038881739921,
      "loss": 5.9213,
      "mean_token_accuracy": 0.1603957325220108,
      "num_tokens": 48821521.0,
      "step": 14615
    },
    {
      "entropy": 6.2533739566802975,
      "epoch": 2.852209540532631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042325098749769534,
      "loss": 5.9755,
      "mean_token_accuracy": 0.14536822140216826,
      "num_tokens": 48838325.0,
      "step": 14620
    },
    {
      "entropy": 6.289679479598999,
      "epoch": 2.853185055116574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004231980723348406,
      "loss": 5.9494,
      "mean_token_accuracy": 0.15203264653682708,
      "num_tokens": 48853857.0,
      "step": 14625
    },
    {
      "entropy": 6.210837650299072,
      "epoch": 2.8541605697005172,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042314514269059826,
      "loss": 5.9602,
      "mean_token_accuracy": 0.15454787909984588,
      "num_tokens": 48870782.0,
      "step": 14630
    },
    {
      "entropy": 6.1820801258087155,
      "epoch": 2.85513608428446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004230921985701407,
      "loss": 5.9103,
      "mean_token_accuracy": 0.15477660968899726,
      "num_tokens": 48888068.0,
      "step": 14635
    },
    {
      "entropy": 6.186975717544556,
      "epoch": 2.856111598868403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004230392399786413,
      "loss": 5.7817,
      "mean_token_accuracy": 0.16512124389410018,
      "num_tokens": 48904174.0,
      "step": 14640
    },
    {
      "entropy": 6.284388256072998,
      "epoch": 2.857087113452346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004229862669212749,
      "loss": 5.9832,
      "mean_token_accuracy": 0.14674447625875472,
      "num_tokens": 48921224.0,
      "step": 14645
    },
    {
      "entropy": 6.165310621261597,
      "epoch": 2.858062628036289,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042293327940321797,
      "loss": 5.8661,
      "mean_token_accuracy": 0.15821752697229385,
      "num_tokens": 48937626.0,
      "step": 14650
    },
    {
      "entropy": 6.206501531600952,
      "epoch": 2.859038142620232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042288027742964825,
      "loss": 5.8825,
      "mean_token_accuracy": 0.15783081203699112,
      "num_tokens": 48955148.0,
      "step": 14655
    },
    {
      "entropy": 6.2908933639526365,
      "epoch": 2.860013657204175,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042282726100574487,
      "loss": 5.9854,
      "mean_token_accuracy": 0.14825415313243867,
      "num_tokens": 48972619.0,
      "step": 14660
    },
    {
      "entropy": 6.2766107559204105,
      "epoch": 2.860989171788118,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042277423013668835,
      "loss": 6.0002,
      "mean_token_accuracy": 0.14608921557664872,
      "num_tokens": 48989909.0,
      "step": 14665
    },
    {
      "entropy": 6.215120077133179,
      "epoch": 2.8619646863720614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004227211848276608,
      "loss": 5.9226,
      "mean_token_accuracy": 0.1551208883523941,
      "num_tokens": 49005312.0,
      "step": 14670
    },
    {
      "entropy": 6.2488508224487305,
      "epoch": 2.8629402009560043,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042266812508384553,
      "loss": 5.9483,
      "mean_token_accuracy": 0.15670492947101594,
      "num_tokens": 49021002.0,
      "step": 14675
    },
    {
      "entropy": 6.217446660995483,
      "epoch": 2.8639157155399473,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004226150509104274,
      "loss": 5.976,
      "mean_token_accuracy": 0.15447665452957154,
      "num_tokens": 49037813.0,
      "step": 14680
    },
    {
      "entropy": 6.16425256729126,
      "epoch": 2.8648912301238902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004225619623125926,
      "loss": 5.8798,
      "mean_token_accuracy": 0.15499567836523057,
      "num_tokens": 49054477.0,
      "step": 14685
    },
    {
      "entropy": 6.306948137283325,
      "epoch": 2.8658667447078336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042250885929552876,
      "loss": 6.0181,
      "mean_token_accuracy": 0.14739200174808503,
      "num_tokens": 49070432.0,
      "step": 14690
    },
    {
      "entropy": 6.282090711593628,
      "epoch": 2.8668422592917766,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004224557418644249,
      "loss": 5.9954,
      "mean_token_accuracy": 0.14879863411188127,
      "num_tokens": 49087984.0,
      "step": 14695
    },
    {
      "entropy": 6.332457113265991,
      "epoch": 2.8678177738757196,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004224026100244715,
      "loss": 6.0663,
      "mean_token_accuracy": 0.14297546073794365,
      "num_tokens": 49104899.0,
      "step": 14700
    },
    {
      "entropy": 6.253933668136597,
      "epoch": 2.8687932884596625,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042234946378086043,
      "loss": 5.9919,
      "mean_token_accuracy": 0.14902078360319138,
      "num_tokens": 49121306.0,
      "step": 14705
    },
    {
      "entropy": 6.241885137557984,
      "epoch": 2.8697688030436055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222963031387848,
      "loss": 5.9959,
      "mean_token_accuracy": 0.14571028649806977,
      "num_tokens": 49136793.0,
      "step": 14710
    },
    {
      "entropy": 6.294543123245239,
      "epoch": 2.8707443176275484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004222431281034395,
      "loss": 5.9868,
      "mean_token_accuracy": 0.15176960229873657,
      "num_tokens": 49154241.0,
      "step": 14715
    },
    {
      "entropy": 6.212490892410278,
      "epoch": 2.8717198322114914,
      "grad_norm": 1.0,
      "learning_rate": 0.00042218993868002055,
      "loss": 5.9966,
      "mean_token_accuracy": 0.15239147394895552,
      "num_tokens": 49170978.0,
      "step": 14720
    },
    {
      "entropy": 6.262880516052246,
      "epoch": 2.8726953467954344,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004221367348737253,
      "loss": 6.0372,
      "mean_token_accuracy": 0.1467800185084343,
      "num_tokens": 49187576.0,
      "step": 14725
    },
    {
      "entropy": 6.302109956741333,
      "epoch": 2.8736708613793778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042208351668975274,
      "loss": 6.009,
      "mean_token_accuracy": 0.1502528429031372,
      "num_tokens": 49206404.0,
      "step": 14730
    },
    {
      "entropy": 6.29430661201477,
      "epoch": 2.8746463759633207,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004220302841333032,
      "loss": 6.0291,
      "mean_token_accuracy": 0.15240800604224206,
      "num_tokens": 49222365.0,
      "step": 14735
    },
    {
      "entropy": 6.204982662200928,
      "epoch": 2.8756218905472637,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004219770372095783,
      "loss": 5.9038,
      "mean_token_accuracy": 0.1589774399995804,
      "num_tokens": 49240471.0,
      "step": 14740
    },
    {
      "entropy": 6.179146146774292,
      "epoch": 2.8765974051312067,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004219237759237811,
      "loss": 5.9612,
      "mean_token_accuracy": 0.15662383213639258,
      "num_tokens": 49255718.0,
      "step": 14745
    },
    {
      "entropy": 6.220907974243164,
      "epoch": 2.8775729197151496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004218705002811162,
      "loss": 5.9697,
      "mean_token_accuracy": 0.153243812918663,
      "num_tokens": 49272746.0,
      "step": 14750
    },
    {
      "entropy": 6.344614171981812,
      "epoch": 2.878548434299093,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004218172102867895,
      "loss": 6.0003,
      "mean_token_accuracy": 0.1490359365940094,
      "num_tokens": 49290478.0,
      "step": 14755
    },
    {
      "entropy": 6.21785159111023,
      "epoch": 2.879523948883036,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004217639059460082,
      "loss": 5.9107,
      "mean_token_accuracy": 0.15352722704410554,
      "num_tokens": 49308105.0,
      "step": 14760
    },
    {
      "entropy": 6.231669855117798,
      "epoch": 2.880499463466979,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042171058726398115,
      "loss": 5.8973,
      "mean_token_accuracy": 0.15445386320352555,
      "num_tokens": 49323519.0,
      "step": 14765
    },
    {
      "entropy": 6.293832874298095,
      "epoch": 2.881474978050922,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004216572542459184,
      "loss": 5.9608,
      "mean_token_accuracy": 0.1489347353577614,
      "num_tokens": 49340162.0,
      "step": 14770
    },
    {
      "entropy": 6.281808090209961,
      "epoch": 2.882450492634865,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042160390689703137,
      "loss": 6.0087,
      "mean_token_accuracy": 0.1462470054626465,
      "num_tokens": 49356831.0,
      "step": 14775
    },
    {
      "entropy": 6.276211309432983,
      "epoch": 2.883426007218808,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042155054522253317,
      "loss": 5.9722,
      "mean_token_accuracy": 0.14988208562135696,
      "num_tokens": 49374368.0,
      "step": 14780
    },
    {
      "entropy": 6.241279363632202,
      "epoch": 2.884401521802751,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000421497169227638,
      "loss": 5.9411,
      "mean_token_accuracy": 0.155579574406147,
      "num_tokens": 49390124.0,
      "step": 14785
    },
    {
      "entropy": 6.258439826965332,
      "epoch": 2.8853770363866937,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042144377891756155,
      "loss": 6.0188,
      "mean_token_accuracy": 0.14231052845716477,
      "num_tokens": 49406826.0,
      "step": 14790
    },
    {
      "entropy": 6.186502885818482,
      "epoch": 2.886352550970637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042139037429752093,
      "loss": 5.86,
      "mean_token_accuracy": 0.16209660172462464,
      "num_tokens": 49422539.0,
      "step": 14795
    },
    {
      "entropy": 6.333443021774292,
      "epoch": 2.88732806555458,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004213369553727347,
      "loss": 6.112,
      "mean_token_accuracy": 0.14058787971735,
      "num_tokens": 49438958.0,
      "step": 14800
    },
    {
      "entropy": 6.317953586578369,
      "epoch": 2.888303580138523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042128352214842275,
      "loss": 6.0079,
      "mean_token_accuracy": 0.14946997165679932,
      "num_tokens": 49454044.0,
      "step": 14805
    },
    {
      "entropy": 6.170296382904053,
      "epoch": 2.889279094722466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004212300746298064,
      "loss": 5.8565,
      "mean_token_accuracy": 0.16735574156045913,
      "num_tokens": 49472198.0,
      "step": 14810
    },
    {
      "entropy": 6.182820415496826,
      "epoch": 2.890254609306409,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042117661282210827,
      "loss": 5.9721,
      "mean_token_accuracy": 0.15472621470689774,
      "num_tokens": 49489198.0,
      "step": 14815
    },
    {
      "entropy": 6.232686519622803,
      "epoch": 2.8912301238903524,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042112313673055263,
      "loss": 5.9028,
      "mean_token_accuracy": 0.15921423733234405,
      "num_tokens": 49504730.0,
      "step": 14820
    },
    {
      "entropy": 6.136650276184082,
      "epoch": 2.8922056384742953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004210696463603647,
      "loss": 5.8754,
      "mean_token_accuracy": 0.16636044383049012,
      "num_tokens": 49521348.0,
      "step": 14825
    },
    {
      "entropy": 6.202272224426269,
      "epoch": 2.8931811530582383,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042101614171677164,
      "loss": 5.8732,
      "mean_token_accuracy": 0.1510866716504097,
      "num_tokens": 49537618.0,
      "step": 14830
    },
    {
      "entropy": 6.2911577224731445,
      "epoch": 2.8941566676421813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004209626228050017,
      "loss": 6.0213,
      "mean_token_accuracy": 0.15113298147916793,
      "num_tokens": 49555898.0,
      "step": 14835
    },
    {
      "entropy": 6.0038713932037355,
      "epoch": 2.8951321822261242,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042090908963028433,
      "loss": 5.6612,
      "mean_token_accuracy": 0.1822880044579506,
      "num_tokens": 49573310.0,
      "step": 14840
    },
    {
      "entropy": 6.266212463378906,
      "epoch": 2.896107696810067,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042085554219785083,
      "loss": 5.9713,
      "mean_token_accuracy": 0.15955851078033448,
      "num_tokens": 49588931.0,
      "step": 14845
    },
    {
      "entropy": 6.2216216087341305,
      "epoch": 2.89708321139401,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042080198051293354,
      "loss": 5.9645,
      "mean_token_accuracy": 0.15080207735300064,
      "num_tokens": 49605331.0,
      "step": 14850
    },
    {
      "entropy": 6.282920980453492,
      "epoch": 2.898058725977953,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042074840458076644,
      "loss": 6.0335,
      "mean_token_accuracy": 0.14948488622903824,
      "num_tokens": 49621917.0,
      "step": 14855
    },
    {
      "entropy": 6.270652914047242,
      "epoch": 2.8990342405618965,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004206948144065846,
      "loss": 5.9262,
      "mean_token_accuracy": 0.1537351831793785,
      "num_tokens": 49638809.0,
      "step": 14860
    },
    {
      "entropy": 6.2626447677612305,
      "epoch": 2.9000097551458395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004206412099956247,
      "loss": 6.0066,
      "mean_token_accuracy": 0.15117238759994506,
      "num_tokens": 49654363.0,
      "step": 14865
    },
    {
      "entropy": 6.241929578781128,
      "epoch": 2.9009852697297824,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042058759135312504,
      "loss": 5.9095,
      "mean_token_accuracy": 0.15801897495985032,
      "num_tokens": 49671894.0,
      "step": 14870
    },
    {
      "entropy": 6.2151957035064695,
      "epoch": 2.9019607843137254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004205339584843247,
      "loss": 6.0102,
      "mean_token_accuracy": 0.1524176076054573,
      "num_tokens": 49689690.0,
      "step": 14875
    },
    {
      "entropy": 6.266592979431152,
      "epoch": 2.9029362988976684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004204803113944646,
      "loss": 5.9985,
      "mean_token_accuracy": 0.15099334940314293,
      "num_tokens": 49706308.0,
      "step": 14880
    },
    {
      "entropy": 6.222029495239258,
      "epoch": 2.9039118134816118,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004204266500887871,
      "loss": 5.8227,
      "mean_token_accuracy": 0.16221504807472228,
      "num_tokens": 49723531.0,
      "step": 14885
    },
    {
      "entropy": 6.011346387863159,
      "epoch": 2.9048873280655547,
      "grad_norm": 1.0,
      "learning_rate": 0.00042037297457253554,
      "loss": 5.754,
      "mean_token_accuracy": 0.17068906426429747,
      "num_tokens": 49740644.0,
      "step": 14890
    },
    {
      "entropy": 6.2566005229949955,
      "epoch": 2.9058628426494977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004203192848509551,
      "loss": 6.0726,
      "mean_token_accuracy": 0.13970333859324455,
      "num_tokens": 49758789.0,
      "step": 14895
    },
    {
      "entropy": 6.348239755630493,
      "epoch": 2.9068383572334406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042026558092929205,
      "loss": 6.0117,
      "mean_token_accuracy": 0.14818378090858458,
      "num_tokens": 49775516.0,
      "step": 14900
    },
    {
      "entropy": 6.261579942703247,
      "epoch": 2.9078138718173836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042021186281279417,
      "loss": 5.9576,
      "mean_token_accuracy": 0.14732573479413985,
      "num_tokens": 49790834.0,
      "step": 14905
    },
    {
      "entropy": 6.261695480346679,
      "epoch": 2.9087893864013266,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004201581305067106,
      "loss": 5.9658,
      "mean_token_accuracy": 0.14777673929929733,
      "num_tokens": 49808939.0,
      "step": 14910
    },
    {
      "entropy": 6.335543012619018,
      "epoch": 2.9097649009852695,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004201043840162919,
      "loss": 6.0408,
      "mean_token_accuracy": 0.14706171303987503,
      "num_tokens": 49825512.0,
      "step": 14915
    },
    {
      "entropy": 6.21590404510498,
      "epoch": 2.910740415569213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004200506233467899,
      "loss": 5.8964,
      "mean_token_accuracy": 0.15719028264284135,
      "num_tokens": 49841239.0,
      "step": 14920
    },
    {
      "entropy": 6.2738770008087155,
      "epoch": 2.911715930153156,
      "grad_norm": 1.0,
      "learning_rate": 0.000419996848503458,
      "loss": 5.9819,
      "mean_token_accuracy": 0.14869838953018188,
      "num_tokens": 49858352.0,
      "step": 14925
    },
    {
      "entropy": 6.289169645309448,
      "epoch": 2.912691444737099,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041994305949155083,
      "loss": 5.9965,
      "mean_token_accuracy": 0.15124712586402894,
      "num_tokens": 49874862.0,
      "step": 14930
    },
    {
      "entropy": 6.246123552322388,
      "epoch": 2.913666959321042,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004198892563163244,
      "loss": 5.9339,
      "mean_token_accuracy": 0.15423132926225663,
      "num_tokens": 49891575.0,
      "step": 14935
    },
    {
      "entropy": 6.264307451248169,
      "epoch": 2.9146424739049848,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041983543898303636,
      "loss": 5.9585,
      "mean_token_accuracy": 0.1498850017786026,
      "num_tokens": 49907607.0,
      "step": 14940
    },
    {
      "entropy": 6.292089176177979,
      "epoch": 2.915617988488928,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004197816074969453,
      "loss": 6.0852,
      "mean_token_accuracy": 0.14615197628736495,
      "num_tokens": 49924525.0,
      "step": 14945
    },
    {
      "entropy": 6.246337795257569,
      "epoch": 2.916593503072871,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041972776186331164,
      "loss": 6.0287,
      "mean_token_accuracy": 0.14452096521854402,
      "num_tokens": 49942456.0,
      "step": 14950
    },
    {
      "entropy": 6.337501382827758,
      "epoch": 2.917569017656814,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041967390208739693,
      "loss": 5.9851,
      "mean_token_accuracy": 0.14316869378089905,
      "num_tokens": 49961481.0,
      "step": 14955
    },
    {
      "entropy": 6.3040375232696535,
      "epoch": 2.918544532240757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041962002817446406,
      "loss": 6.0153,
      "mean_token_accuracy": 0.15149709284305574,
      "num_tokens": 49978397.0,
      "step": 14960
    },
    {
      "entropy": 6.192004013061523,
      "epoch": 2.9195200468247,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041956614012977753,
      "loss": 5.9153,
      "mean_token_accuracy": 0.16235812455415727,
      "num_tokens": 49995334.0,
      "step": 14965
    },
    {
      "entropy": 6.245798873901367,
      "epoch": 2.920495561408643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000419512237958603,
      "loss": 5.9673,
      "mean_token_accuracy": 0.15782852917909623,
      "num_tokens": 50012342.0,
      "step": 14970
    },
    {
      "entropy": 6.296828269958496,
      "epoch": 2.921471075992586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041945832166620766,
      "loss": 6.0245,
      "mean_token_accuracy": 0.1440238043665886,
      "num_tokens": 50028679.0,
      "step": 14975
    },
    {
      "entropy": 6.256477975845337,
      "epoch": 2.922446590576529,
      "grad_norm": 1.0,
      "learning_rate": 0.00041940439125785994,
      "loss": 5.9497,
      "mean_token_accuracy": 0.14638334587216378,
      "num_tokens": 50044280.0,
      "step": 14980
    },
    {
      "entropy": 6.323039245605469,
      "epoch": 2.9234221051604723,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004193504467388298,
      "loss": 5.986,
      "mean_token_accuracy": 0.14644398242235185,
      "num_tokens": 50061069.0,
      "step": 14985
    },
    {
      "entropy": 6.198437929153442,
      "epoch": 2.9243976197444153,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041929648811438847,
      "loss": 5.9315,
      "mean_token_accuracy": 0.15375867933034898,
      "num_tokens": 50077552.0,
      "step": 14990
    },
    {
      "entropy": 6.2147050380706785,
      "epoch": 2.925373134328358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004192425153898087,
      "loss": 5.9573,
      "mean_token_accuracy": 0.1528772935271263,
      "num_tokens": 50095296.0,
      "step": 14995
    },
    {
      "entropy": 6.3130711078643795,
      "epoch": 2.926348648912301,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041918852857036426,
      "loss": 5.9636,
      "mean_token_accuracy": 0.15090934038162232,
      "num_tokens": 50112472.0,
      "step": 15000
    },
    {
      "epoch": 2.926348648912301,
      "eval_entropy": 6.143111495079191,
      "eval_loss": 6.079898834228516,
      "eval_mean_token_accuracy": 0.1527553586998224,
      "eval_num_tokens": 50112472.0,
      "eval_runtime": 15.2978,
      "eval_samples_per_second": 1083.619,
      "eval_steps_per_second": 135.51,
      "step": 15000
    },
    {
      "entropy": 6.303101062774658,
      "epoch": 2.927324163496244,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004191345276613308,
      "loss": 6.041,
      "mean_token_accuracy": 0.1426026441156864,
      "num_tokens": 50130152.0,
      "step": 15005
    },
    {
      "entropy": 6.286333703994751,
      "epoch": 2.9282996780801875,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041908051266798495,
      "loss": 6.053,
      "mean_token_accuracy": 0.14602067396044732,
      "num_tokens": 50146572.0,
      "step": 15010
    },
    {
      "entropy": 6.210633230209351,
      "epoch": 2.9292751926641305,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000419026483595605,
      "loss": 5.9134,
      "mean_token_accuracy": 0.1608864888548851,
      "num_tokens": 50163501.0,
      "step": 15015
    },
    {
      "entropy": 6.240325546264648,
      "epoch": 2.9302507072480735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004189724404494703,
      "loss": 5.9859,
      "mean_token_accuracy": 0.14985650032758713,
      "num_tokens": 50181417.0,
      "step": 15020
    },
    {
      "entropy": 6.332682943344116,
      "epoch": 2.9312262218320164,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041891838323486187,
      "loss": 6.1148,
      "mean_token_accuracy": 0.14455197006464005,
      "num_tokens": 50198071.0,
      "step": 15025
    },
    {
      "entropy": 6.309230804443359,
      "epoch": 2.9322017364159594,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000418864311957062,
      "loss": 6.0179,
      "mean_token_accuracy": 0.15279347449541092,
      "num_tokens": 50213577.0,
      "step": 15030
    },
    {
      "entropy": 6.189103364944458,
      "epoch": 2.9331772509999023,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041881022662135426,
      "loss": 5.9852,
      "mean_token_accuracy": 0.15366155058145523,
      "num_tokens": 50231023.0,
      "step": 15035
    },
    {
      "entropy": 6.188130044937134,
      "epoch": 2.9341527655838453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041875612723302376,
      "loss": 5.8093,
      "mean_token_accuracy": 0.15542302131652833,
      "num_tokens": 50247435.0,
      "step": 15040
    },
    {
      "entropy": 6.273208379745483,
      "epoch": 2.9351282801677883,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041870201379735683,
      "loss": 6.0148,
      "mean_token_accuracy": 0.1499328337609768,
      "num_tokens": 50264937.0,
      "step": 15045
    },
    {
      "entropy": 6.207729196548462,
      "epoch": 2.9361037947517317,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004186478863196413,
      "loss": 5.8919,
      "mean_token_accuracy": 0.1554173484444618,
      "num_tokens": 50280653.0,
      "step": 15050
    },
    {
      "entropy": 6.360844087600708,
      "epoch": 2.9370793093356746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041859374480516625,
      "loss": 6.0822,
      "mean_token_accuracy": 0.14427170306444168,
      "num_tokens": 50297327.0,
      "step": 15055
    },
    {
      "entropy": 6.216488599777222,
      "epoch": 2.9380548239196176,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004185395892592223,
      "loss": 5.9489,
      "mean_token_accuracy": 0.1486012101173401,
      "num_tokens": 50314457.0,
      "step": 15060
    },
    {
      "entropy": 6.181719446182251,
      "epoch": 2.9390303385035605,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004184854196871012,
      "loss": 5.8601,
      "mean_token_accuracy": 0.1658857673406601,
      "num_tokens": 50330249.0,
      "step": 15065
    },
    {
      "entropy": 6.244798040390014,
      "epoch": 2.9400058530875035,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041843123609409635,
      "loss": 5.93,
      "mean_token_accuracy": 0.15899043381214142,
      "num_tokens": 50347494.0,
      "step": 15070
    },
    {
      "entropy": 6.284605836868286,
      "epoch": 2.940981367671447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004183770384855022,
      "loss": 5.962,
      "mean_token_accuracy": 0.15181366056203843,
      "num_tokens": 50364530.0,
      "step": 15075
    },
    {
      "entropy": 6.299973726272583,
      "epoch": 2.94195688225539,
      "grad_norm": 0.953125,
      "learning_rate": 0.000418322826866615,
      "loss": 5.9617,
      "mean_token_accuracy": 0.15193284898996354,
      "num_tokens": 50382689.0,
      "step": 15080
    },
    {
      "entropy": 6.242220497131347,
      "epoch": 2.942932396839333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004182686012427319,
      "loss": 5.988,
      "mean_token_accuracy": 0.15313626527786256,
      "num_tokens": 50400091.0,
      "step": 15085
    },
    {
      "entropy": 6.232745027542114,
      "epoch": 2.943907911423276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041821436161915165,
      "loss": 5.9159,
      "mean_token_accuracy": 0.15395527780056,
      "num_tokens": 50417441.0,
      "step": 15090
    },
    {
      "entropy": 6.235224056243896,
      "epoch": 2.9448834260072188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004181601080011745,
      "loss": 5.9148,
      "mean_token_accuracy": 0.16130338460206986,
      "num_tokens": 50434539.0,
      "step": 15095
    },
    {
      "entropy": 6.171309137344361,
      "epoch": 2.9458589405911617,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041810584039410174,
      "loss": 5.8568,
      "mean_token_accuracy": 0.1585104450583458,
      "num_tokens": 50450074.0,
      "step": 15100
    },
    {
      "entropy": 6.179873704910278,
      "epoch": 2.9468344551751047,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041805155880323644,
      "loss": 5.9443,
      "mean_token_accuracy": 0.1564841613173485,
      "num_tokens": 50467987.0,
      "step": 15105
    },
    {
      "entropy": 6.15431227684021,
      "epoch": 2.9478099697590476,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004179972632338825,
      "loss": 5.8351,
      "mean_token_accuracy": 0.16147438287734986,
      "num_tokens": 50486168.0,
      "step": 15110
    },
    {
      "entropy": 6.26186900138855,
      "epoch": 2.948785484342991,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004179429536913457,
      "loss": 5.9427,
      "mean_token_accuracy": 0.1499002307653427,
      "num_tokens": 50502556.0,
      "step": 15115
    },
    {
      "entropy": 6.191634654998779,
      "epoch": 2.949760998926934,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041788863018093293,
      "loss": 5.9025,
      "mean_token_accuracy": 0.16017318218946458,
      "num_tokens": 50518347.0,
      "step": 15120
    },
    {
      "entropy": 6.1695386409759525,
      "epoch": 2.950736513510877,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004178342927079525,
      "loss": 5.8681,
      "mean_token_accuracy": 0.16408715546131133,
      "num_tokens": 50533847.0,
      "step": 15125
    },
    {
      "entropy": 6.314264965057373,
      "epoch": 2.95171202809482,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000417779941277714,
      "loss": 5.9814,
      "mean_token_accuracy": 0.14714889824390412,
      "num_tokens": 50549410.0,
      "step": 15130
    },
    {
      "entropy": 6.230092763900757,
      "epoch": 2.9526875426787633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041772557589552856,
      "loss": 5.9008,
      "mean_token_accuracy": 0.149276664853096,
      "num_tokens": 50566554.0,
      "step": 15135
    },
    {
      "entropy": 6.14343957901001,
      "epoch": 2.9536630572627063,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00041767119656670854,
      "loss": 5.9304,
      "mean_token_accuracy": 0.16260657608509063,
      "num_tokens": 50584688.0,
      "step": 15140
    },
    {
      "entropy": 6.2516645908355715,
      "epoch": 2.9546385718466492,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041761680329656766,
      "loss": 5.9446,
      "mean_token_accuracy": 0.14965897053480148,
      "num_tokens": 50601085.0,
      "step": 15145
    },
    {
      "entropy": 6.225920486450195,
      "epoch": 2.955614086430592,
      "grad_norm": 1.0,
      "learning_rate": 0.00041756239609042105,
      "loss": 6.0043,
      "mean_token_accuracy": 0.1469335988163948,
      "num_tokens": 50619030.0,
      "step": 15150
    },
    {
      "entropy": 6.32417402267456,
      "epoch": 2.956589601014535,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004175079749535852,
      "loss": 5.986,
      "mean_token_accuracy": 0.14966932833194732,
      "num_tokens": 50634511.0,
      "step": 15155
    },
    {
      "entropy": 6.293870830535889,
      "epoch": 2.957565115598478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000417453539891378,
      "loss": 5.9739,
      "mean_token_accuracy": 0.15078936964273454,
      "num_tokens": 50651824.0,
      "step": 15160
    },
    {
      "entropy": 6.152067708969116,
      "epoch": 2.958540630182421,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004173990909091186,
      "loss": 5.9169,
      "mean_token_accuracy": 0.16830268651247024,
      "num_tokens": 50668951.0,
      "step": 15165
    },
    {
      "entropy": 6.195869016647339,
      "epoch": 2.959516144766364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041734462801212743,
      "loss": 5.9033,
      "mean_token_accuracy": 0.1515965148806572,
      "num_tokens": 50684864.0,
      "step": 15170
    },
    {
      "entropy": 6.261364126205445,
      "epoch": 2.9604916593503074,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041729015120572655,
      "loss": 6.0053,
      "mean_token_accuracy": 0.15173157751560212,
      "num_tokens": 50701112.0,
      "step": 15175
    },
    {
      "entropy": 6.323415231704712,
      "epoch": 2.9614671739342504,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004172356604952393,
      "loss": 6.0101,
      "mean_token_accuracy": 0.15478409081697464,
      "num_tokens": 50718752.0,
      "step": 15180
    },
    {
      "entropy": 6.2544944286346436,
      "epoch": 2.9624426885181934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004171811558859902,
      "loss": 6.003,
      "mean_token_accuracy": 0.14452821463346482,
      "num_tokens": 50736808.0,
      "step": 15185
    },
    {
      "entropy": 6.234484767913818,
      "epoch": 2.9634182031021363,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041712663738330535,
      "loss": 5.9812,
      "mean_token_accuracy": 0.1522173836827278,
      "num_tokens": 50752389.0,
      "step": 15190
    },
    {
      "entropy": 6.239856004714966,
      "epoch": 2.9643937176860793,
      "grad_norm": 1.0,
      "learning_rate": 0.000417072104992512,
      "loss": 5.9675,
      "mean_token_accuracy": 0.15388938039541245,
      "num_tokens": 50768586.0,
      "step": 15195
    },
    {
      "entropy": 6.277125883102417,
      "epoch": 2.9653692322700227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041701755871893895,
      "loss": 5.9892,
      "mean_token_accuracy": 0.15313664674758912,
      "num_tokens": 50784716.0,
      "step": 15200
    },
    {
      "entropy": 6.326681470870971,
      "epoch": 2.9663447468539657,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004169629985679162,
      "loss": 6.0283,
      "mean_token_accuracy": 0.14183567240834236,
      "num_tokens": 50801684.0,
      "step": 15205
    },
    {
      "entropy": 6.179004144668579,
      "epoch": 2.9673202614379086,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00041690842454477516,
      "loss": 5.903,
      "mean_token_accuracy": 0.16154223382472993,
      "num_tokens": 50817823.0,
      "step": 15210
    },
    {
      "entropy": 6.365161609649658,
      "epoch": 2.9682957760218516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041685383665484866,
      "loss": 6.1116,
      "mean_token_accuracy": 0.14083350747823714,
      "num_tokens": 50834578.0,
      "step": 15215
    },
    {
      "entropy": 6.210379028320313,
      "epoch": 2.9692712906057945,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041679923490347084,
      "loss": 5.8564,
      "mean_token_accuracy": 0.16429907381534575,
      "num_tokens": 50851596.0,
      "step": 15220
    },
    {
      "entropy": 6.0610044479370115,
      "epoch": 2.9702468051897375,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004167446192959771,
      "loss": 5.8319,
      "mean_token_accuracy": 0.16751340925693511,
      "num_tokens": 50868827.0,
      "step": 15225
    },
    {
      "entropy": 6.300719690322876,
      "epoch": 2.9712223197736805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041668998983770446,
      "loss": 6.0291,
      "mean_token_accuracy": 0.14799614399671554,
      "num_tokens": 50884979.0,
      "step": 15230
    },
    {
      "entropy": 6.2920175075531,
      "epoch": 2.9721978343576234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041663534653399095,
      "loss": 5.9757,
      "mean_token_accuracy": 0.1474774017930031,
      "num_tokens": 50900700.0,
      "step": 15235
    },
    {
      "entropy": 6.2237293243408205,
      "epoch": 2.973173348941567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041658068939017613,
      "loss": 5.9642,
      "mean_token_accuracy": 0.14832511097192763,
      "num_tokens": 50917508.0,
      "step": 15240
    },
    {
      "entropy": 6.2732964038848875,
      "epoch": 2.97414886352551,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000416526018411601,
      "loss": 5.9733,
      "mean_token_accuracy": 0.15311516225337982,
      "num_tokens": 50934127.0,
      "step": 15245
    },
    {
      "entropy": 6.363986206054688,
      "epoch": 2.9751243781094527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041647133360360765,
      "loss": 6.0711,
      "mean_token_accuracy": 0.14645902961492538,
      "num_tokens": 50951144.0,
      "step": 15250
    },
    {
      "entropy": 6.277458047866821,
      "epoch": 2.9760998926933957,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041641663497153994,
      "loss": 5.8973,
      "mean_token_accuracy": 0.15397393852472305,
      "num_tokens": 50967506.0,
      "step": 15255
    },
    {
      "entropy": 6.225163316726684,
      "epoch": 2.9770754072773387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004163619225207426,
      "loss": 6.0009,
      "mean_token_accuracy": 0.1511413723230362,
      "num_tokens": 50983720.0,
      "step": 15260
    },
    {
      "entropy": 6.373676013946533,
      "epoch": 2.978050921861282,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004163071962565621,
      "loss": 6.0629,
      "mean_token_accuracy": 0.14627072289586068,
      "num_tokens": 51000613.0,
      "step": 15265
    },
    {
      "entropy": 6.167000675201416,
      "epoch": 2.979026436445225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004162524561843459,
      "loss": 5.7763,
      "mean_token_accuracy": 0.16843343079090117,
      "num_tokens": 51017345.0,
      "step": 15270
    },
    {
      "entropy": 6.1769506454467775,
      "epoch": 2.980001951029168,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041619770230944317,
      "loss": 5.9303,
      "mean_token_accuracy": 0.15417159646749495,
      "num_tokens": 51034117.0,
      "step": 15275
    },
    {
      "entropy": 6.303287601470947,
      "epoch": 2.980977465613111,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041614293463720415,
      "loss": 5.9532,
      "mean_token_accuracy": 0.15297964811325074,
      "num_tokens": 51049958.0,
      "step": 15280
    },
    {
      "entropy": 6.1950836181640625,
      "epoch": 2.981952980197054,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00041608815317298065,
      "loss": 5.9094,
      "mean_token_accuracy": 0.15152375996112824,
      "num_tokens": 51067721.0,
      "step": 15285
    },
    {
      "entropy": 6.173806428909302,
      "epoch": 2.982928494780997,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004160333579221257,
      "loss": 5.8955,
      "mean_token_accuracy": 0.15730760246515274,
      "num_tokens": 51083078.0,
      "step": 15290
    },
    {
      "entropy": 6.2199108600616455,
      "epoch": 2.98390400936494,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041597854888999363,
      "loss": 5.9105,
      "mean_token_accuracy": 0.15903644412755966,
      "num_tokens": 51099510.0,
      "step": 15295
    },
    {
      "entropy": 6.186666870117188,
      "epoch": 2.984879523948883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041592372608194024,
      "loss": 5.9267,
      "mean_token_accuracy": 0.1564860463142395,
      "num_tokens": 51115855.0,
      "step": 15300
    },
    {
      "entropy": 6.1129974842071535,
      "epoch": 2.985855038532826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041586888950332257,
      "loss": 5.7207,
      "mean_token_accuracy": 0.17064978182315826,
      "num_tokens": 51131912.0,
      "step": 15305
    },
    {
      "entropy": 6.270316886901855,
      "epoch": 2.986830553116769,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041581403915949913,
      "loss": 6.0717,
      "mean_token_accuracy": 0.1428835064172745,
      "num_tokens": 51149268.0,
      "step": 15310
    },
    {
      "entropy": 6.251433658599853,
      "epoch": 2.987806067700712,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004157591750558296,
      "loss": 5.9392,
      "mean_token_accuracy": 0.14893487840890884,
      "num_tokens": 51166809.0,
      "step": 15315
    },
    {
      "entropy": 6.278151178359986,
      "epoch": 2.988781582284655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041570429719767514,
      "loss": 5.9726,
      "mean_token_accuracy": 0.14885919392108918,
      "num_tokens": 51184155.0,
      "step": 15320
    },
    {
      "entropy": 6.162713050842285,
      "epoch": 2.989757096868598,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041564940559039827,
      "loss": 5.85,
      "mean_token_accuracy": 0.1595679357647896,
      "num_tokens": 51200158.0,
      "step": 15325
    },
    {
      "entropy": 6.2317163944244385,
      "epoch": 2.9907326114525414,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041559450023936277,
      "loss": 5.9387,
      "mean_token_accuracy": 0.15413075983524321,
      "num_tokens": 51216921.0,
      "step": 15330
    },
    {
      "entropy": 6.082292795181274,
      "epoch": 2.9917081260364844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004155395811499338,
      "loss": 5.7807,
      "mean_token_accuracy": 0.17233011797070502,
      "num_tokens": 51234587.0,
      "step": 15335
    },
    {
      "entropy": 6.256633377075195,
      "epoch": 2.9926836406204274,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004154846483274779,
      "loss": 5.8963,
      "mean_token_accuracy": 0.15769733861088753,
      "num_tokens": 51249379.0,
      "step": 15340
    },
    {
      "entropy": 6.132475900650024,
      "epoch": 2.9936591552043703,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004154297017773627,
      "loss": 5.7574,
      "mean_token_accuracy": 0.16926292330026627,
      "num_tokens": 51267449.0,
      "step": 15345
    },
    {
      "entropy": 6.254632663726807,
      "epoch": 2.9946346697883133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004153747415049577,
      "loss": 5.9785,
      "mean_token_accuracy": 0.15180247128009797,
      "num_tokens": 51284976.0,
      "step": 15350
    },
    {
      "entropy": 6.269231843948364,
      "epoch": 2.9956101843722562,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004153197675156333,
      "loss": 6.018,
      "mean_token_accuracy": 0.1522519752383232,
      "num_tokens": 51301534.0,
      "step": 15355
    },
    {
      "entropy": 6.169466304779053,
      "epoch": 2.996585698956199,
      "grad_norm": 1.25,
      "learning_rate": 0.0004152647798147612,
      "loss": 5.9335,
      "mean_token_accuracy": 0.15745105147361754,
      "num_tokens": 51318932.0,
      "step": 15360
    },
    {
      "entropy": 6.191586637496949,
      "epoch": 2.997561213540142,
      "grad_norm": 0.875,
      "learning_rate": 0.0004152097784077148,
      "loss": 5.8717,
      "mean_token_accuracy": 0.15502889305353165,
      "num_tokens": 51336724.0,
      "step": 15365
    },
    {
      "entropy": 6.237853002548218,
      "epoch": 2.9985367281240856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004151547632998685,
      "loss": 5.9936,
      "mean_token_accuracy": 0.15845445394515992,
      "num_tokens": 51353443.0,
      "step": 15370
    },
    {
      "entropy": 6.323828554153442,
      "epoch": 2.9995122427080285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004150997344965983,
      "loss": 5.9886,
      "mean_token_accuracy": 0.15453446954488753,
      "num_tokens": 51371068.0,
      "step": 15375
    },
    {
      "entropy": 6.203205373552111,
      "epoch": 3.000390205833577,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004150446920032814,
      "loss": 5.9681,
      "mean_token_accuracy": 0.14381685356299082,
      "num_tokens": 51385621.0,
      "step": 15380
    },
    {
      "entropy": 6.305203628540039,
      "epoch": 3.00136572041752,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004149896358252962,
      "loss": 5.9371,
      "mean_token_accuracy": 0.15141319781541823,
      "num_tokens": 51401776.0,
      "step": 15385
    },
    {
      "entropy": 6.182918548583984,
      "epoch": 3.002341235001463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004149345659680228,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16675786375999452,
      "num_tokens": 51418966.0,
      "step": 15390
    },
    {
      "entropy": 6.234851121902466,
      "epoch": 3.0033167495854065,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041487948243684237,
      "loss": 5.8626,
      "mean_token_accuracy": 0.15540676414966584,
      "num_tokens": 51436227.0,
      "step": 15395
    },
    {
      "entropy": 6.202904748916626,
      "epoch": 3.0042922641693495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041482438523713744,
      "loss": 5.8579,
      "mean_token_accuracy": 0.15769557207822799,
      "num_tokens": 51452772.0,
      "step": 15400
    },
    {
      "entropy": 6.332231950759888,
      "epoch": 3.0052677787532924,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004147692743742919,
      "loss": 5.8965,
      "mean_token_accuracy": 0.15307178795337678,
      "num_tokens": 51470050.0,
      "step": 15405
    },
    {
      "entropy": 6.249373054504394,
      "epoch": 3.0062432933372354,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004147141498536911,
      "loss": 5.8709,
      "mean_token_accuracy": 0.15163935124874114,
      "num_tokens": 51487037.0,
      "step": 15410
    },
    {
      "entropy": 6.056521940231323,
      "epoch": 3.0072188079211783,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041465901168072144,
      "loss": 5.6656,
      "mean_token_accuracy": 0.17384843081235885,
      "num_tokens": 51505385.0,
      "step": 15415
    },
    {
      "entropy": 6.109113597869873,
      "epoch": 3.0081943225051213,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000414603859860771,
      "loss": 5.6703,
      "mean_token_accuracy": 0.17725851088762284,
      "num_tokens": 51521664.0,
      "step": 15420
    },
    {
      "entropy": 6.254049062728882,
      "epoch": 3.0091698370890643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004145486943992288,
      "loss": 5.9341,
      "mean_token_accuracy": 0.15328772887587547,
      "num_tokens": 51537765.0,
      "step": 15425
    },
    {
      "entropy": 6.2281961917877195,
      "epoch": 3.0101453516730077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041449351530148566,
      "loss": 5.8702,
      "mean_token_accuracy": 0.15621715784072876,
      "num_tokens": 51553698.0,
      "step": 15430
    },
    {
      "entropy": 6.211122703552246,
      "epoch": 3.0111208662569506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004144383225729333,
      "loss": 5.7949,
      "mean_token_accuracy": 0.16805047690868377,
      "num_tokens": 51570234.0,
      "step": 15435
    },
    {
      "entropy": 6.184336090087891,
      "epoch": 3.0120963808408936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041438311621896505,
      "loss": 5.8012,
      "mean_token_accuracy": 0.1658303588628769,
      "num_tokens": 51586692.0,
      "step": 15440
    },
    {
      "entropy": 6.275726747512818,
      "epoch": 3.0130718954248366,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041432789624497546,
      "loss": 5.8792,
      "mean_token_accuracy": 0.1553502470254898,
      "num_tokens": 51603521.0,
      "step": 15445
    },
    {
      "entropy": 6.213590908050537,
      "epoch": 3.0140474100087795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041427266265636043,
      "loss": 5.7945,
      "mean_token_accuracy": 0.16276625245809556,
      "num_tokens": 51618055.0,
      "step": 15450
    },
    {
      "entropy": 6.2823827266693115,
      "epoch": 3.0150229245927225,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004142174154585172,
      "loss": 5.8585,
      "mean_token_accuracy": 0.15878407806158065,
      "num_tokens": 51633309.0,
      "step": 15455
    },
    {
      "entropy": 6.159121084213257,
      "epoch": 3.015998439176666,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004141621546568443,
      "loss": 5.779,
      "mean_token_accuracy": 0.164939583837986,
      "num_tokens": 51650327.0,
      "step": 15460
    },
    {
      "entropy": 6.227036190032959,
      "epoch": 3.016973953760609,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041410688025674154,
      "loss": 5.85,
      "mean_token_accuracy": 0.1567177176475525,
      "num_tokens": 51666216.0,
      "step": 15465
    },
    {
      "entropy": 6.282097101211548,
      "epoch": 3.017949468344552,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004140515922636103,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1569633290171623,
      "num_tokens": 51682361.0,
      "step": 15470
    },
    {
      "entropy": 6.32126750946045,
      "epoch": 3.0189249829284948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041399629068285303,
      "loss": 5.8919,
      "mean_token_accuracy": 0.14931862950325012,
      "num_tokens": 51698826.0,
      "step": 15475
    },
    {
      "entropy": 6.221215152740479,
      "epoch": 3.0199004975124377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041394097551987366,
      "loss": 5.8576,
      "mean_token_accuracy": 0.15243782550096513,
      "num_tokens": 51714838.0,
      "step": 15480
    },
    {
      "entropy": 6.254670095443726,
      "epoch": 3.0208760120963807,
      "grad_norm": 1.0,
      "learning_rate": 0.00041388564678007726,
      "loss": 5.8347,
      "mean_token_accuracy": 0.1517318457365036,
      "num_tokens": 51731185.0,
      "step": 15485
    },
    {
      "entropy": 6.3758830547332765,
      "epoch": 3.021851526680324,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041383030446887054,
      "loss": 6.0336,
      "mean_token_accuracy": 0.15067343264818192,
      "num_tokens": 51749007.0,
      "step": 15490
    },
    {
      "entropy": 6.203380298614502,
      "epoch": 3.022827041264267,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041377494859166133,
      "loss": 5.7838,
      "mean_token_accuracy": 0.1633370876312256,
      "num_tokens": 51767088.0,
      "step": 15495
    },
    {
      "entropy": 6.213956594467163,
      "epoch": 3.02380255584821,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041371957915385866,
      "loss": 5.7719,
      "mean_token_accuracy": 0.16114173829555511,
      "num_tokens": 51783361.0,
      "step": 15500
    },
    {
      "entropy": 6.206213903427124,
      "epoch": 3.024778070432153,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041366419616087314,
      "loss": 5.8073,
      "mean_token_accuracy": 0.15744010210037232,
      "num_tokens": 51799555.0,
      "step": 15505
    },
    {
      "entropy": 6.217703056335449,
      "epoch": 3.025753585016096,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004136087996181166,
      "loss": 5.7717,
      "mean_token_accuracy": 0.16473789215087892,
      "num_tokens": 51815973.0,
      "step": 15510
    },
    {
      "entropy": 6.23214316368103,
      "epoch": 3.026729099600039,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004135533895310021,
      "loss": 5.7975,
      "mean_token_accuracy": 0.1580188602209091,
      "num_tokens": 51832337.0,
      "step": 15515
    },
    {
      "entropy": 6.182824230194091,
      "epoch": 3.027704614183982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004134979659049443,
      "loss": 5.7987,
      "mean_token_accuracy": 0.15779075771570206,
      "num_tokens": 51848465.0,
      "step": 15520
    },
    {
      "entropy": 6.224671649932861,
      "epoch": 3.0286801287679253,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041344252874535883,
      "loss": 5.8562,
      "mean_token_accuracy": 0.1570225715637207,
      "num_tokens": 51865477.0,
      "step": 15525
    },
    {
      "entropy": 6.274149990081787,
      "epoch": 3.029655643351868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004133870780576628,
      "loss": 5.9094,
      "mean_token_accuracy": 0.1578225463628769,
      "num_tokens": 51881444.0,
      "step": 15530
    },
    {
      "entropy": 6.290549325942993,
      "epoch": 3.030631157935811,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004133316138472748,
      "loss": 5.9374,
      "mean_token_accuracy": 0.15501331984996797,
      "num_tokens": 51897564.0,
      "step": 15535
    },
    {
      "entropy": 6.165888357162475,
      "epoch": 3.031606672519754,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004132761361196145,
      "loss": 5.8211,
      "mean_token_accuracy": 0.15926152169704438,
      "num_tokens": 51916964.0,
      "step": 15540
    },
    {
      "entropy": 6.25823483467102,
      "epoch": 3.032582187103697,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004132206448801031,
      "loss": 5.8299,
      "mean_token_accuracy": 0.16031641662120819,
      "num_tokens": 51934530.0,
      "step": 15545
    },
    {
      "entropy": 6.179529809951783,
      "epoch": 3.03355770168764,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041316514013416274,
      "loss": 5.7617,
      "mean_token_accuracy": 0.1647062286734581,
      "num_tokens": 51951355.0,
      "step": 15550
    },
    {
      "entropy": 6.207970905303955,
      "epoch": 3.0345332162715835,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041310962188721744,
      "loss": 5.8069,
      "mean_token_accuracy": 0.1645549088716507,
      "num_tokens": 51969406.0,
      "step": 15555
    },
    {
      "entropy": 6.260521507263183,
      "epoch": 3.0355087308555264,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041305409014469205,
      "loss": 5.8563,
      "mean_token_accuracy": 0.15705254822969436,
      "num_tokens": 51986093.0,
      "step": 15560
    },
    {
      "entropy": 6.249173736572265,
      "epoch": 3.0364842454394694,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000412998544912013,
      "loss": 5.8911,
      "mean_token_accuracy": 0.14906765818595885,
      "num_tokens": 52004360.0,
      "step": 15565
    },
    {
      "entropy": 6.2143724918365475,
      "epoch": 3.0374597600234123,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041294298619460804,
      "loss": 5.8191,
      "mean_token_accuracy": 0.15840015262365342,
      "num_tokens": 52018991.0,
      "step": 15570
    },
    {
      "entropy": 6.306128263473511,
      "epoch": 3.0384352746073553,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000412887413997906,
      "loss": 5.8855,
      "mean_token_accuracy": 0.15039998888969422,
      "num_tokens": 52035724.0,
      "step": 15575
    },
    {
      "entropy": 6.175100946426392,
      "epoch": 3.0394107891912983,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004128318283273374,
      "loss": 5.7668,
      "mean_token_accuracy": 0.1662082254886627,
      "num_tokens": 52052369.0,
      "step": 15580
    },
    {
      "entropy": 6.284959888458252,
      "epoch": 3.040386303775241,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041277622918833374,
      "loss": 5.9545,
      "mean_token_accuracy": 0.1538613647222519,
      "num_tokens": 52069161.0,
      "step": 15585
    },
    {
      "entropy": 6.1939386367797855,
      "epoch": 3.0413618183591846,
      "grad_norm": 0.96875,
      "learning_rate": 0.000412720616586328,
      "loss": 5.7509,
      "mean_token_accuracy": 0.1739802524447441,
      "num_tokens": 52084980.0,
      "step": 15590
    },
    {
      "entropy": 6.145208835601807,
      "epoch": 3.0423373329431276,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004126649905267545,
      "loss": 5.7073,
      "mean_token_accuracy": 0.16504894345998763,
      "num_tokens": 52103566.0,
      "step": 15595
    },
    {
      "entropy": 6.238322305679321,
      "epoch": 3.0433128475270705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041260935101504877,
      "loss": 5.9191,
      "mean_token_accuracy": 0.15181829333305358,
      "num_tokens": 52119535.0,
      "step": 15600
    },
    {
      "entropy": 6.294637966156006,
      "epoch": 3.0442883621110135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041255369805664766,
      "loss": 5.9084,
      "mean_token_accuracy": 0.1512921541929245,
      "num_tokens": 52135222.0,
      "step": 15605
    },
    {
      "entropy": 6.265599346160888,
      "epoch": 3.0452638766949565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041249803165698944,
      "loss": 5.9069,
      "mean_token_accuracy": 0.15809064358472824,
      "num_tokens": 52151242.0,
      "step": 15610
    },
    {
      "entropy": 6.162208127975464,
      "epoch": 3.0462393912788994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004124423518215136,
      "loss": 5.7952,
      "mean_token_accuracy": 0.1592549592256546,
      "num_tokens": 52168366.0,
      "step": 15615
    },
    {
      "entropy": 6.109570026397705,
      "epoch": 3.047214905862843,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00041238665855566106,
      "loss": 5.7356,
      "mean_token_accuracy": 0.1675531804561615,
      "num_tokens": 52184327.0,
      "step": 15620
    },
    {
      "entropy": 6.271342706680298,
      "epoch": 3.048190420446786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041233095186487385,
      "loss": 5.8871,
      "mean_token_accuracy": 0.15730815529823303,
      "num_tokens": 52200974.0,
      "step": 15625
    },
    {
      "entropy": 6.242988777160645,
      "epoch": 3.0491659350307287,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004122752317545955,
      "loss": 5.831,
      "mean_token_accuracy": 0.16103765815496446,
      "num_tokens": 52217245.0,
      "step": 15630
    },
    {
      "entropy": 6.155996656417846,
      "epoch": 3.0501414496146717,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004122194982302707,
      "loss": 5.7709,
      "mean_token_accuracy": 0.16508322954177856,
      "num_tokens": 52234924.0,
      "step": 15635
    },
    {
      "entropy": 6.26767053604126,
      "epoch": 3.0511169641986147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041216375129734563,
      "loss": 5.9275,
      "mean_token_accuracy": 0.15586531460285186,
      "num_tokens": 52252375.0,
      "step": 15640
    },
    {
      "entropy": 6.275266551971436,
      "epoch": 3.0520924787825576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004121079909612677,
      "loss": 5.8701,
      "mean_token_accuracy": 0.16325312107801437,
      "num_tokens": 52269751.0,
      "step": 15645
    },
    {
      "entropy": 6.052535152435302,
      "epoch": 3.053067993366501,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004120522172274855,
      "loss": 5.6403,
      "mean_token_accuracy": 0.1681852363049984,
      "num_tokens": 52286967.0,
      "step": 15650
    },
    {
      "entropy": 6.221571969985962,
      "epoch": 3.054043507950444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041199643010144906,
      "loss": 5.8312,
      "mean_token_accuracy": 0.15694714188575745,
      "num_tokens": 52303712.0,
      "step": 15655
    },
    {
      "entropy": 6.15099515914917,
      "epoch": 3.055019022534387,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004119406295886098,
      "loss": 5.6974,
      "mean_token_accuracy": 0.17305730581283568,
      "num_tokens": 52320706.0,
      "step": 15660
    },
    {
      "entropy": 6.200251913070678,
      "epoch": 3.05599453711833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041188481569442027,
      "loss": 5.8169,
      "mean_token_accuracy": 0.1621382623910904,
      "num_tokens": 52335681.0,
      "step": 15665
    },
    {
      "entropy": 6.177006578445434,
      "epoch": 3.056970051702273,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004118289884243344,
      "loss": 5.7953,
      "mean_token_accuracy": 0.1591392919421196,
      "num_tokens": 52354088.0,
      "step": 15670
    },
    {
      "entropy": 6.224957466125488,
      "epoch": 3.057945566286216,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004117731477838074,
      "loss": 5.919,
      "mean_token_accuracy": 0.14982060492038726,
      "num_tokens": 52372086.0,
      "step": 15675
    },
    {
      "entropy": 6.247678184509278,
      "epoch": 3.058921080870159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000411717293778296,
      "loss": 5.8925,
      "mean_token_accuracy": 0.1512695387005806,
      "num_tokens": 52389460.0,
      "step": 15680
    },
    {
      "entropy": 6.209524297714234,
      "epoch": 3.059896595454102,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004116614264132578,
      "loss": 5.7759,
      "mean_token_accuracy": 0.16067336797714232,
      "num_tokens": 52404589.0,
      "step": 15685
    },
    {
      "entropy": 6.090528392791748,
      "epoch": 3.060872110038045,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004116055456941521,
      "loss": 5.7112,
      "mean_token_accuracy": 0.17178263813257216,
      "num_tokens": 52423225.0,
      "step": 15690
    },
    {
      "entropy": 6.208738183975219,
      "epoch": 3.061847624621988,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041154965162643937,
      "loss": 5.8336,
      "mean_token_accuracy": 0.15590507984161378,
      "num_tokens": 52438289.0,
      "step": 15695
    },
    {
      "entropy": 6.180991363525391,
      "epoch": 3.062823139205931,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004114937442155814,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16165985763072968,
      "num_tokens": 52456418.0,
      "step": 15700
    },
    {
      "entropy": 6.273389148712158,
      "epoch": 3.063798653789874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041143782346704115,
      "loss": 5.8804,
      "mean_token_accuracy": 0.15453929379582404,
      "num_tokens": 52472327.0,
      "step": 15705
    },
    {
      "entropy": 6.2602705478668215,
      "epoch": 3.064774168373817,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041138188938628315,
      "loss": 5.9311,
      "mean_token_accuracy": 0.15122750401496887,
      "num_tokens": 52488530.0,
      "step": 15710
    },
    {
      "entropy": 6.2434405326843265,
      "epoch": 3.0657496829577604,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004113259419787728,
      "loss": 5.9057,
      "mean_token_accuracy": 0.15725244879722594,
      "num_tokens": 52505576.0,
      "step": 15715
    },
    {
      "entropy": 6.272645473480225,
      "epoch": 3.0667251975417034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041126998124997745,
      "loss": 5.8,
      "mean_token_accuracy": 0.16223448663949966,
      "num_tokens": 52521984.0,
      "step": 15720
    },
    {
      "entropy": 6.20276141166687,
      "epoch": 3.0677007121256463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004112140072053652,
      "loss": 5.8161,
      "mean_token_accuracy": 0.16602576225996019,
      "num_tokens": 52539139.0,
      "step": 15725
    },
    {
      "entropy": 6.156534910202026,
      "epoch": 3.0686762267095893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004111580198504055,
      "loss": 5.8056,
      "mean_token_accuracy": 0.16154175624251366,
      "num_tokens": 52555652.0,
      "step": 15730
    },
    {
      "entropy": 6.230987787246704,
      "epoch": 3.0696517412935322,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041110201919056954,
      "loss": 5.8536,
      "mean_token_accuracy": 0.1631019115447998,
      "num_tokens": 52572013.0,
      "step": 15735
    },
    {
      "entropy": 6.254735946655273,
      "epoch": 3.070627255877475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041104600523132924,
      "loss": 5.9047,
      "mean_token_accuracy": 0.15291812121868134,
      "num_tokens": 52587915.0,
      "step": 15740
    },
    {
      "entropy": 6.153646659851074,
      "epoch": 3.0716027704614186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004109899779781583,
      "loss": 5.7662,
      "mean_token_accuracy": 0.1669243335723877,
      "num_tokens": 52605406.0,
      "step": 15745
    },
    {
      "entropy": 6.2264299392700195,
      "epoch": 3.0725782850453616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004109339374365312,
      "loss": 5.7393,
      "mean_token_accuracy": 0.16284470856189728,
      "num_tokens": 52623246.0,
      "step": 15750
    },
    {
      "entropy": 6.129952621459961,
      "epoch": 3.0735537996293045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004108778836119244,
      "loss": 5.7535,
      "mean_token_accuracy": 0.16437994986772536,
      "num_tokens": 52640528.0,
      "step": 15755
    },
    {
      "entropy": 6.176532030105591,
      "epoch": 3.0745293142132475,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000410821816509815,
      "loss": 5.7544,
      "mean_token_accuracy": 0.16719566881656647,
      "num_tokens": 52657579.0,
      "step": 15760
    },
    {
      "entropy": 6.259911346435547,
      "epoch": 3.0755048287971904,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041076573613568176,
      "loss": 5.8935,
      "mean_token_accuracy": 0.16165215969085694,
      "num_tokens": 52673250.0,
      "step": 15765
    },
    {
      "entropy": 6.125877332687378,
      "epoch": 3.0764803433811334,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004107096424950047,
      "loss": 5.7692,
      "mean_token_accuracy": 0.16789505779743194,
      "num_tokens": 52689727.0,
      "step": 15770
    },
    {
      "entropy": 6.173990535736084,
      "epoch": 3.0774558579650764,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004106535355932651,
      "loss": 5.7916,
      "mean_token_accuracy": 0.16765566170215607,
      "num_tokens": 52706724.0,
      "step": 15775
    },
    {
      "entropy": 6.173914051055908,
      "epoch": 3.0784313725490198,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041059741543594537,
      "loss": 5.8013,
      "mean_token_accuracy": 0.159796904027462,
      "num_tokens": 52723540.0,
      "step": 15780
    },
    {
      "entropy": 6.140379571914673,
      "epoch": 3.0794068871329627,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041054128202852954,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16304728239774705,
      "num_tokens": 52740276.0,
      "step": 15785
    },
    {
      "entropy": 6.2918860912323,
      "epoch": 3.0803824017169057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004104851353765027,
      "loss": 5.8751,
      "mean_token_accuracy": 0.16240522116422654,
      "num_tokens": 52756952.0,
      "step": 15790
    },
    {
      "entropy": 6.238530254364013,
      "epoch": 3.0813579163008487,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004104289754853513,
      "loss": 5.8639,
      "mean_token_accuracy": 0.15881982147693635,
      "num_tokens": 52773088.0,
      "step": 15795
    },
    {
      "entropy": 6.027067613601685,
      "epoch": 3.0823334308847916,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004103728023605631,
      "loss": 5.6652,
      "mean_token_accuracy": 0.17601378262043,
      "num_tokens": 52790918.0,
      "step": 15800
    },
    {
      "entropy": 6.258475637435913,
      "epoch": 3.0833089454687346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041031661600762705,
      "loss": 5.8791,
      "mean_token_accuracy": 0.1608082264661789,
      "num_tokens": 52807000.0,
      "step": 15805
    },
    {
      "entropy": 6.192395210266113,
      "epoch": 3.084284460052678,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041026041643203373,
      "loss": 5.8217,
      "mean_token_accuracy": 0.1593296080827713,
      "num_tokens": 52822994.0,
      "step": 15810
    },
    {
      "entropy": 6.21892728805542,
      "epoch": 3.085259974636621,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004102042036392744,
      "loss": 5.821,
      "mean_token_accuracy": 0.15342558771371842,
      "num_tokens": 52838268.0,
      "step": 15815
    },
    {
      "entropy": 6.313898801803589,
      "epoch": 3.086235489220564,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004101479776348423,
      "loss": 5.9345,
      "mean_token_accuracy": 0.15287290662527084,
      "num_tokens": 52855683.0,
      "step": 15820
    },
    {
      "entropy": 6.240259838104248,
      "epoch": 3.087211003804507,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004100917384242315,
      "loss": 5.8608,
      "mean_token_accuracy": 0.16069925129413604,
      "num_tokens": 52872141.0,
      "step": 15825
    },
    {
      "entropy": 6.285158109664917,
      "epoch": 3.08818651838845,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041003548601293747,
      "loss": 5.8433,
      "mean_token_accuracy": 0.15858289003372192,
      "num_tokens": 52888919.0,
      "step": 15830
    },
    {
      "entropy": 6.204867601394653,
      "epoch": 3.089162032972393,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040997922040645696,
      "loss": 5.8129,
      "mean_token_accuracy": 0.15653839707374573,
      "num_tokens": 52905903.0,
      "step": 15835
    },
    {
      "entropy": 6.174923515319824,
      "epoch": 3.0901375475563357,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004099229416102881,
      "loss": 5.8364,
      "mean_token_accuracy": 0.1601391389966011,
      "num_tokens": 52924224.0,
      "step": 15840
    },
    {
      "entropy": 6.23770318031311,
      "epoch": 3.091113062140279,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004098666496299304,
      "loss": 5.8741,
      "mean_token_accuracy": 0.15787013471126557,
      "num_tokens": 52940510.0,
      "step": 15845
    },
    {
      "entropy": 6.067630577087402,
      "epoch": 3.092088576724222,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00040981034447088425,
      "loss": 5.6767,
      "mean_token_accuracy": 0.1787683293223381,
      "num_tokens": 52958458.0,
      "step": 15850
    },
    {
      "entropy": 6.270606660842896,
      "epoch": 3.093064091308165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004097540261386518,
      "loss": 5.88,
      "mean_token_accuracy": 0.15550704300403595,
      "num_tokens": 52973709.0,
      "step": 15855
    },
    {
      "entropy": 6.239150524139404,
      "epoch": 3.094039605892108,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040969769463873617,
      "loss": 5.8341,
      "mean_token_accuracy": 0.15689529329538346,
      "num_tokens": 52989203.0,
      "step": 15860
    },
    {
      "entropy": 6.153495836257934,
      "epoch": 3.095015120476051,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040964134997664183,
      "loss": 5.7746,
      "mean_token_accuracy": 0.15787402987480165,
      "num_tokens": 53005833.0,
      "step": 15865
    },
    {
      "entropy": 6.294810485839844,
      "epoch": 3.095990635059994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040958499215787463,
      "loss": 5.9504,
      "mean_token_accuracy": 0.15326353311538696,
      "num_tokens": 53022616.0,
      "step": 15870
    },
    {
      "entropy": 6.192452907562256,
      "epoch": 3.0969661496439373,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004095286211879418,
      "loss": 5.776,
      "mean_token_accuracy": 0.1687918022274971,
      "num_tokens": 53040473.0,
      "step": 15875
    },
    {
      "entropy": 6.168641805648804,
      "epoch": 3.0979416642278803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040947223707235147,
      "loss": 5.7191,
      "mean_token_accuracy": 0.17535959035158158,
      "num_tokens": 53057142.0,
      "step": 15880
    },
    {
      "entropy": 6.083554077148437,
      "epoch": 3.0989171788118233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004094158398166134,
      "loss": 5.713,
      "mean_token_accuracy": 0.16701506376266478,
      "num_tokens": 53073909.0,
      "step": 15885
    },
    {
      "entropy": 6.2647223472595215,
      "epoch": 3.0998926933957662,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004093594294262386,
      "loss": 5.8882,
      "mean_token_accuracy": 0.149155193567276,
      "num_tokens": 53091327.0,
      "step": 15890
    },
    {
      "entropy": 6.231508684158325,
      "epoch": 3.100868207979709,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040930300590673915,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1521793246269226,
      "num_tokens": 53107589.0,
      "step": 15895
    },
    {
      "entropy": 6.116573524475098,
      "epoch": 3.101843722563652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040924656926362863,
      "loss": 5.7248,
      "mean_token_accuracy": 0.17394813150167465,
      "num_tokens": 53125165.0,
      "step": 15900
    },
    {
      "entropy": 6.249022626876831,
      "epoch": 3.1028192371475956,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004091901195024218,
      "loss": 5.8162,
      "mean_token_accuracy": 0.16056041717529296,
      "num_tokens": 53141704.0,
      "step": 15905
    },
    {
      "entropy": 6.259309005737305,
      "epoch": 3.1037947517315385,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004091336566286348,
      "loss": 5.8375,
      "mean_token_accuracy": 0.15840044021606445,
      "num_tokens": 53158220.0,
      "step": 15910
    },
    {
      "entropy": 6.204025840759277,
      "epoch": 3.1047702663154815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004090771806477849,
      "loss": 5.8738,
      "mean_token_accuracy": 0.1604410484433174,
      "num_tokens": 53173947.0,
      "step": 15915
    },
    {
      "entropy": 6.197992038726807,
      "epoch": 3.1057457808994244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004090206915653908,
      "loss": 5.8387,
      "mean_token_accuracy": 0.1658250406384468,
      "num_tokens": 53190885.0,
      "step": 15920
    },
    {
      "entropy": 6.2658271312713625,
      "epoch": 3.1067212954833674,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040896418938697246,
      "loss": 5.9209,
      "mean_token_accuracy": 0.15142602622509002,
      "num_tokens": 53209555.0,
      "step": 15925
    },
    {
      "entropy": 6.317733907699585,
      "epoch": 3.1076968100673104,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040890767411805084,
      "loss": 5.8931,
      "mean_token_accuracy": 0.1528881847858429,
      "num_tokens": 53225794.0,
      "step": 15930
    },
    {
      "entropy": 6.261526012420655,
      "epoch": 3.1086723246512538,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004088511457641487,
      "loss": 5.8641,
      "mean_token_accuracy": 0.15458890348672866,
      "num_tokens": 53243070.0,
      "step": 15935
    },
    {
      "entropy": 6.138437700271607,
      "epoch": 3.1096478392351967,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040879460433078954,
      "loss": 5.7681,
      "mean_token_accuracy": 0.16324483901262282,
      "num_tokens": 53260734.0,
      "step": 15940
    },
    {
      "entropy": 6.293890905380249,
      "epoch": 3.1106233538191397,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004087380498234985,
      "loss": 5.9392,
      "mean_token_accuracy": 0.14930154979228974,
      "num_tokens": 53278462.0,
      "step": 15945
    },
    {
      "entropy": 6.282012128829956,
      "epoch": 3.1115988684030826,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004086814822478019,
      "loss": 5.8093,
      "mean_token_accuracy": 0.16268643289804458,
      "num_tokens": 53294010.0,
      "step": 15950
    },
    {
      "entropy": 6.239280843734742,
      "epoch": 3.1125743829870256,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040862490160922734,
      "loss": 5.9085,
      "mean_token_accuracy": 0.15065233334898948,
      "num_tokens": 53310546.0,
      "step": 15955
    },
    {
      "entropy": 6.156354570388794,
      "epoch": 3.1135498975709686,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004085683079133036,
      "loss": 5.7718,
      "mean_token_accuracy": 0.16968608200550078,
      "num_tokens": 53328646.0,
      "step": 15960
    },
    {
      "entropy": 5.998792409896851,
      "epoch": 3.1145254121549115,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040851170116556085,
      "loss": 5.5598,
      "mean_token_accuracy": 0.1896073579788208,
      "num_tokens": 53346116.0,
      "step": 15965
    },
    {
      "entropy": 6.183960962295532,
      "epoch": 3.115500926738855,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004084550813715306,
      "loss": 5.804,
      "mean_token_accuracy": 0.15908256769180298,
      "num_tokens": 53363067.0,
      "step": 15970
    },
    {
      "entropy": 6.244590139389038,
      "epoch": 3.116476441322798,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040839844853674537,
      "loss": 5.8285,
      "mean_token_accuracy": 0.1583217903971672,
      "num_tokens": 53378992.0,
      "step": 15975
    },
    {
      "entropy": 6.164586782455444,
      "epoch": 3.117451955906741,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040834180266673917,
      "loss": 5.8117,
      "mean_token_accuracy": 0.1599603071808815,
      "num_tokens": 53395761.0,
      "step": 15980
    },
    {
      "entropy": 6.235380363464356,
      "epoch": 3.118427470490684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040828514376704724,
      "loss": 5.9218,
      "mean_token_accuracy": 0.1485462263226509,
      "num_tokens": 53412717.0,
      "step": 15985
    },
    {
      "entropy": 6.248510360717773,
      "epoch": 3.1194029850746268,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040822847184320617,
      "loss": 5.8148,
      "mean_token_accuracy": 0.16175041049718858,
      "num_tokens": 53429525.0,
      "step": 15990
    },
    {
      "entropy": 6.232184982299804,
      "epoch": 3.1203784996585697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040817178690075363,
      "loss": 5.8622,
      "mean_token_accuracy": 0.15731247365474701,
      "num_tokens": 53447195.0,
      "step": 15995
    },
    {
      "entropy": 6.2001848220825195,
      "epoch": 3.121354014242513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004081150889452287,
      "loss": 5.8011,
      "mean_token_accuracy": 0.15992796123027803,
      "num_tokens": 53463671.0,
      "step": 16000
    },
    {
      "entropy": 6.220301532745362,
      "epoch": 3.122329528826456,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004080583779821717,
      "loss": 5.8288,
      "mean_token_accuracy": 0.15825044512748718,
      "num_tokens": 53480721.0,
      "step": 16005
    },
    {
      "entropy": 6.177873706817627,
      "epoch": 3.123305043410399,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040800165401712427,
      "loss": 5.8173,
      "mean_token_accuracy": 0.1588008299469948,
      "num_tokens": 53498292.0,
      "step": 16010
    },
    {
      "entropy": 6.215986061096191,
      "epoch": 3.124280557994342,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040794491705562917,
      "loss": 5.8004,
      "mean_token_accuracy": 0.1672125518321991,
      "num_tokens": 53515747.0,
      "step": 16015
    },
    {
      "entropy": 6.107724714279175,
      "epoch": 3.125256072578285,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040788816710323063,
      "loss": 5.7129,
      "mean_token_accuracy": 0.17033711075782776,
      "num_tokens": 53532895.0,
      "step": 16020
    },
    {
      "entropy": 6.114950037002563,
      "epoch": 3.126231587162228,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004078314041654741,
      "loss": 5.7444,
      "mean_token_accuracy": 0.16796301156282425,
      "num_tokens": 53549241.0,
      "step": 16025
    },
    {
      "entropy": 6.271958589553833,
      "epoch": 3.127207101746171,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004077746282479061,
      "loss": 5.8635,
      "mean_token_accuracy": 0.15805317163467408,
      "num_tokens": 53565882.0,
      "step": 16030
    },
    {
      "entropy": 6.263635063171387,
      "epoch": 3.1281826163301143,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004077178393560746,
      "loss": 5.8998,
      "mean_token_accuracy": 0.1539486214518547,
      "num_tokens": 53582770.0,
      "step": 16035
    },
    {
      "entropy": 6.178380346298217,
      "epoch": 3.1291581309140573,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004076610374955289,
      "loss": 5.7941,
      "mean_token_accuracy": 0.16475779563188553,
      "num_tokens": 53600649.0,
      "step": 16040
    },
    {
      "entropy": 6.219222497940064,
      "epoch": 3.130133645498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004076042226718195,
      "loss": 5.8277,
      "mean_token_accuracy": 0.16607412397861482,
      "num_tokens": 53616568.0,
      "step": 16045
    },
    {
      "entropy": 6.278246784210205,
      "epoch": 3.131109160081943,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000407547394890498,
      "loss": 5.8767,
      "mean_token_accuracy": 0.15641167610883713,
      "num_tokens": 53634717.0,
      "step": 16050
    },
    {
      "entropy": 6.20270414352417,
      "epoch": 3.132084674665886,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040749055415711743,
      "loss": 5.8003,
      "mean_token_accuracy": 0.1572565421462059,
      "num_tokens": 53650877.0,
      "step": 16055
    },
    {
      "entropy": 6.21031813621521,
      "epoch": 3.133060189249829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004074337004772322,
      "loss": 5.8005,
      "mean_token_accuracy": 0.16208639442920686,
      "num_tokens": 53667906.0,
      "step": 16060
    },
    {
      "entropy": 6.123543787002563,
      "epoch": 3.1340357038337725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040737683385639776,
      "loss": 5.7649,
      "mean_token_accuracy": 0.1619967043399811,
      "num_tokens": 53682942.0,
      "step": 16065
    },
    {
      "entropy": 6.187923622131348,
      "epoch": 3.1350112184177155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040731995430017085,
      "loss": 5.86,
      "mean_token_accuracy": 0.15874305963516236,
      "num_tokens": 53699906.0,
      "step": 16070
    },
    {
      "entropy": 6.273453950881958,
      "epoch": 3.1359867330016584,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004072630618141096,
      "loss": 5.8641,
      "mean_token_accuracy": 0.1550448164343834,
      "num_tokens": 53716278.0,
      "step": 16075
    },
    {
      "entropy": 6.141275024414062,
      "epoch": 3.1369622475856014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040720615640377347,
      "loss": 5.7174,
      "mean_token_accuracy": 0.1739761859178543,
      "num_tokens": 53733317.0,
      "step": 16080
    },
    {
      "entropy": 6.212957429885864,
      "epoch": 3.1379377621695443,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040714923807472274,
      "loss": 5.8145,
      "mean_token_accuracy": 0.15944063514471055,
      "num_tokens": 53750477.0,
      "step": 16085
    },
    {
      "entropy": 6.258506107330322,
      "epoch": 3.1389132767534873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004070923068325196,
      "loss": 5.8712,
      "mean_token_accuracy": 0.15852162092924119,
      "num_tokens": 53765755.0,
      "step": 16090
    },
    {
      "entropy": 6.1709802627563475,
      "epoch": 3.1398887913374303,
      "grad_norm": 1.0,
      "learning_rate": 0.00040703536268272693,
      "loss": 5.8304,
      "mean_token_accuracy": 0.16711232513189317,
      "num_tokens": 53782173.0,
      "step": 16095
    },
    {
      "entropy": 6.096932363510132,
      "epoch": 3.1408643059213737,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004069784056309093,
      "loss": 5.7086,
      "mean_token_accuracy": 0.17271096408367156,
      "num_tokens": 53799407.0,
      "step": 16100
    },
    {
      "entropy": 6.153721714019776,
      "epoch": 3.1418398205053166,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004069214356826322,
      "loss": 5.7538,
      "mean_token_accuracy": 0.1713261991739273,
      "num_tokens": 53816363.0,
      "step": 16105
    },
    {
      "entropy": 6.126596212387085,
      "epoch": 3.1428153350892596,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040686445284346254,
      "loss": 5.7219,
      "mean_token_accuracy": 0.16990626603364944,
      "num_tokens": 53833022.0,
      "step": 16110
    },
    {
      "entropy": 6.187149333953857,
      "epoch": 3.1437908496732025,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004068074571189685,
      "loss": 5.8011,
      "mean_token_accuracy": 0.17070655673742294,
      "num_tokens": 53849924.0,
      "step": 16115
    },
    {
      "entropy": 6.225136137008667,
      "epoch": 3.1447663642571455,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040675044851471966,
      "loss": 5.8607,
      "mean_token_accuracy": 0.16492980867624282,
      "num_tokens": 53866291.0,
      "step": 16120
    },
    {
      "entropy": 6.2570525169372555,
      "epoch": 3.1457418788410885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040669342703628645,
      "loss": 5.9164,
      "mean_token_accuracy": 0.15732453912496566,
      "num_tokens": 53882510.0,
      "step": 16125
    },
    {
      "entropy": 6.2320818424224855,
      "epoch": 3.146717393425032,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004066363926892409,
      "loss": 5.8071,
      "mean_token_accuracy": 0.1584446206688881,
      "num_tokens": 53898833.0,
      "step": 16130
    },
    {
      "entropy": 6.276405906677246,
      "epoch": 3.147692908008975,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040657934547915625,
      "loss": 5.8845,
      "mean_token_accuracy": 0.15611676275730133,
      "num_tokens": 53916945.0,
      "step": 16135
    },
    {
      "entropy": 6.266270399093628,
      "epoch": 3.148668422592918,
      "grad_norm": 0.984375,
      "learning_rate": 0.000406522285411607,
      "loss": 5.8976,
      "mean_token_accuracy": 0.15739496797323227,
      "num_tokens": 53933853.0,
      "step": 16140
    },
    {
      "entropy": 6.251980447769165,
      "epoch": 3.1496439371768608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004064652124921686,
      "loss": 5.8498,
      "mean_token_accuracy": 0.15368182361125945,
      "num_tokens": 53950098.0,
      "step": 16145
    },
    {
      "entropy": 6.205639410018921,
      "epoch": 3.1506194517608037,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040640812672641834,
      "loss": 5.7767,
      "mean_token_accuracy": 0.16516460329294205,
      "num_tokens": 53966070.0,
      "step": 16150
    },
    {
      "entropy": 6.148417663574219,
      "epoch": 3.1515949663447467,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040635102811993417,
      "loss": 5.726,
      "mean_token_accuracy": 0.1680666372179985,
      "num_tokens": 53981715.0,
      "step": 16155
    },
    {
      "entropy": 6.152086448669434,
      "epoch": 3.15257048092869,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004062939166782958,
      "loss": 5.7332,
      "mean_token_accuracy": 0.16802968084812164,
      "num_tokens": 53999179.0,
      "step": 16160
    },
    {
      "entropy": 5.966910171508789,
      "epoch": 3.153545995512633,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004062367924070837,
      "loss": 5.5916,
      "mean_token_accuracy": 0.17749583274126052,
      "num_tokens": 54016024.0,
      "step": 16165
    },
    {
      "entropy": 6.2389974117279055,
      "epoch": 3.154521510096576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040617965531188007,
      "loss": 5.8811,
      "mean_token_accuracy": 0.1594734385609627,
      "num_tokens": 54033378.0,
      "step": 16170
    },
    {
      "entropy": 6.268914270401001,
      "epoch": 3.155497024680519,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004061225053982681,
      "loss": 5.842,
      "mean_token_accuracy": 0.1592211753129959,
      "num_tokens": 54049432.0,
      "step": 16175
    },
    {
      "entropy": 6.1743567943572994,
      "epoch": 3.156472539264462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004060653426718321,
      "loss": 5.7402,
      "mean_token_accuracy": 0.16705299764871598,
      "num_tokens": 54064688.0,
      "step": 16180
    },
    {
      "entropy": 6.067214632034302,
      "epoch": 3.157448053848405,
      "grad_norm": 0.9375,
      "learning_rate": 0.000406008167138158,
      "loss": 5.7267,
      "mean_token_accuracy": 0.17143253982067108,
      "num_tokens": 54080195.0,
      "step": 16185
    },
    {
      "entropy": 6.263535022735596,
      "epoch": 3.1584235684323483,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040595097880283276,
      "loss": 5.8862,
      "mean_token_accuracy": 0.15880610942840576,
      "num_tokens": 54096138.0,
      "step": 16190
    },
    {
      "entropy": 6.179228448867798,
      "epoch": 3.1593990830162912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040589377767144455,
      "loss": 5.7967,
      "mean_token_accuracy": 0.17139531821012496,
      "num_tokens": 54112411.0,
      "step": 16195
    },
    {
      "entropy": 6.193606328964234,
      "epoch": 3.160374597600234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040583656374958285,
      "loss": 5.82,
      "mean_token_accuracy": 0.1598365604877472,
      "num_tokens": 54128207.0,
      "step": 16200
    },
    {
      "entropy": 6.23356065750122,
      "epoch": 3.161350112184177,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040577933704283856,
      "loss": 5.8906,
      "mean_token_accuracy": 0.1547669678926468,
      "num_tokens": 54144997.0,
      "step": 16205
    },
    {
      "entropy": 6.1544195175170895,
      "epoch": 3.16232562676812,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004057220975568035,
      "loss": 5.8274,
      "mean_token_accuracy": 0.16332604587078095,
      "num_tokens": 54162733.0,
      "step": 16210
    },
    {
      "entropy": 6.223440361022949,
      "epoch": 3.163301141352063,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004056648452970709,
      "loss": 5.8778,
      "mean_token_accuracy": 0.14905786663293838,
      "num_tokens": 54180712.0,
      "step": 16215
    },
    {
      "entropy": 6.19349365234375,
      "epoch": 3.164276655936006,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004056075802692354,
      "loss": 5.7478,
      "mean_token_accuracy": 0.16409339904785156,
      "num_tokens": 54196797.0,
      "step": 16220
    },
    {
      "entropy": 6.149651384353637,
      "epoch": 3.1652521705199494,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004055503024788927,
      "loss": 5.7895,
      "mean_token_accuracy": 0.16169643700122832,
      "num_tokens": 54214911.0,
      "step": 16225
    },
    {
      "entropy": 6.262200927734375,
      "epoch": 3.1662276851038924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004054930119316397,
      "loss": 5.9209,
      "mean_token_accuracy": 0.1559016674757004,
      "num_tokens": 54230049.0,
      "step": 16230
    },
    {
      "entropy": 6.161082458496094,
      "epoch": 3.1672031996878354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004054357086330745,
      "loss": 5.8425,
      "mean_token_accuracy": 0.16396384239196776,
      "num_tokens": 54247490.0,
      "step": 16235
    },
    {
      "entropy": 6.304794645309448,
      "epoch": 3.1681787142717783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040537839258879694,
      "loss": 5.8867,
      "mean_token_accuracy": 0.15680582970380783,
      "num_tokens": 54263731.0,
      "step": 16240
    },
    {
      "entropy": 6.255164670944214,
      "epoch": 3.1691542288557213,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004053210638044075,
      "loss": 5.8867,
      "mean_token_accuracy": 0.1556261494755745,
      "num_tokens": 54279820.0,
      "step": 16245
    },
    {
      "entropy": 6.292652702331543,
      "epoch": 3.1701297434396642,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040526372228550804,
      "loss": 5.8816,
      "mean_token_accuracy": 0.1568897545337677,
      "num_tokens": 54296728.0,
      "step": 16250
    },
    {
      "entropy": 6.145977878570557,
      "epoch": 3.1711052580236077,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040520636803770206,
      "loss": 5.6999,
      "mean_token_accuracy": 0.17211344391107558,
      "num_tokens": 54314526.0,
      "step": 16255
    },
    {
      "entropy": 6.234192562103272,
      "epoch": 3.1720807726075506,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004051490010665938,
      "loss": 5.8231,
      "mean_token_accuracy": 0.15466077029705047,
      "num_tokens": 54330750.0,
      "step": 16260
    },
    {
      "entropy": 6.175535583496094,
      "epoch": 3.1730562871914936,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000405091621377789,
      "loss": 5.7793,
      "mean_token_accuracy": 0.16825142353773118,
      "num_tokens": 54349380.0,
      "step": 16265
    },
    {
      "entropy": 6.2836731433868405,
      "epoch": 3.1740318017754365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004050342289768946,
      "loss": 5.9005,
      "mean_token_accuracy": 0.15241361707448958,
      "num_tokens": 54366034.0,
      "step": 16270
    },
    {
      "entropy": 6.214749717712403,
      "epoch": 3.1750073163593795,
      "grad_norm": 1.0,
      "learning_rate": 0.00040497682386951886,
      "loss": 5.8662,
      "mean_token_accuracy": 0.15514660030603408,
      "num_tokens": 54382293.0,
      "step": 16275
    },
    {
      "entropy": 6.227333021163941,
      "epoch": 3.1759828309433225,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040491940606127115,
      "loss": 5.8919,
      "mean_token_accuracy": 0.16120505183935166,
      "num_tokens": 54399183.0,
      "step": 16280
    },
    {
      "entropy": 6.2955262660980225,
      "epoch": 3.1769583455272654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040486197555776197,
      "loss": 5.9233,
      "mean_token_accuracy": 0.15021163523197173,
      "num_tokens": 54414611.0,
      "step": 16285
    },
    {
      "entropy": 6.227910947799683,
      "epoch": 3.177933860111209,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040480453236460355,
      "loss": 5.7637,
      "mean_token_accuracy": 0.16292756646871567,
      "num_tokens": 54431025.0,
      "step": 16290
    },
    {
      "entropy": 6.174452257156372,
      "epoch": 3.178909374695152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040474707648740875,
      "loss": 5.7904,
      "mean_token_accuracy": 0.162223719060421,
      "num_tokens": 54447604.0,
      "step": 16295
    },
    {
      "entropy": 6.189231824874878,
      "epoch": 3.1798848892790947,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004046896079317921,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16105469018220903,
      "num_tokens": 54463519.0,
      "step": 16300
    },
    {
      "entropy": 6.169848966598511,
      "epoch": 3.1808604038630377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040463212670336924,
      "loss": 5.7542,
      "mean_token_accuracy": 0.16388167440891266,
      "num_tokens": 54479303.0,
      "step": 16305
    },
    {
      "entropy": 5.983514261245728,
      "epoch": 3.1818359184469807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040457463280775687,
      "loss": 5.5605,
      "mean_token_accuracy": 0.18276263773441315,
      "num_tokens": 54495462.0,
      "step": 16310
    },
    {
      "entropy": 6.202028751373291,
      "epoch": 3.1828114330309236,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004045171262505732,
      "loss": 5.873,
      "mean_token_accuracy": 0.15258457213640214,
      "num_tokens": 54512849.0,
      "step": 16315
    },
    {
      "entropy": 6.215277576446534,
      "epoch": 3.183786947614867,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004044596070374376,
      "loss": 5.8576,
      "mean_token_accuracy": 0.16044628322124482,
      "num_tokens": 54529678.0,
      "step": 16320
    },
    {
      "entropy": 6.279133939743042,
      "epoch": 3.18476246219881,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004044020751739706,
      "loss": 5.8607,
      "mean_token_accuracy": 0.15442074090242386,
      "num_tokens": 54546317.0,
      "step": 16325
    },
    {
      "entropy": 6.2088275909423825,
      "epoch": 3.185737976782753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004043445306657941,
      "loss": 5.8437,
      "mean_token_accuracy": 0.15946442037820815,
      "num_tokens": 54562983.0,
      "step": 16330
    },
    {
      "entropy": 6.3201103687286375,
      "epoch": 3.186713491366696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040428697351853094,
      "loss": 5.964,
      "mean_token_accuracy": 0.15317586809396744,
      "num_tokens": 54578708.0,
      "step": 16335
    },
    {
      "entropy": 6.096599006652832,
      "epoch": 3.187689005950639,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040422940373780554,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1733556717634201,
      "num_tokens": 54594555.0,
      "step": 16340
    },
    {
      "entropy": 6.241952753067016,
      "epoch": 3.188664520534582,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004041718213292434,
      "loss": 5.855,
      "mean_token_accuracy": 0.1577516958117485,
      "num_tokens": 54611430.0,
      "step": 16345
    },
    {
      "entropy": 6.191230630874633,
      "epoch": 3.1896400351185252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040411422629847124,
      "loss": 5.7735,
      "mean_token_accuracy": 0.16073525324463844,
      "num_tokens": 54629467.0,
      "step": 16350
    },
    {
      "entropy": 6.260449361801148,
      "epoch": 3.190615549702468,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000404056618651117,
      "loss": 5.8878,
      "mean_token_accuracy": 0.15365138053894042,
      "num_tokens": 54644732.0,
      "step": 16355
    },
    {
      "entropy": 6.183058261871338,
      "epoch": 3.191591064286411,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040399899839281013,
      "loss": 5.8065,
      "mean_token_accuracy": 0.16587045639753342,
      "num_tokens": 54661639.0,
      "step": 16360
    },
    {
      "entropy": 6.184934616088867,
      "epoch": 3.192566578870354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004039413655291807,
      "loss": 5.7727,
      "mean_token_accuracy": 0.16484688073396683,
      "num_tokens": 54678153.0,
      "step": 16365
    },
    {
      "entropy": 6.210353708267212,
      "epoch": 3.193542093454297,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004038837200658608,
      "loss": 5.7969,
      "mean_token_accuracy": 0.16377971023321153,
      "num_tokens": 54694215.0,
      "step": 16370
    },
    {
      "entropy": 6.162817859649659,
      "epoch": 3.19451760803824,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040382606200848285,
      "loss": 5.809,
      "mean_token_accuracy": 0.16345303803682326,
      "num_tokens": 54710637.0,
      "step": 16375
    },
    {
      "entropy": 6.12972583770752,
      "epoch": 3.1954931226221834,
      "grad_norm": 1.0,
      "learning_rate": 0.00040376839136268145,
      "loss": 5.7748,
      "mean_token_accuracy": 0.15955910235643386,
      "num_tokens": 54726638.0,
      "step": 16380
    },
    {
      "entropy": 6.192163801193237,
      "epoch": 3.1964686372061264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004037107081340918,
      "loss": 5.7777,
      "mean_token_accuracy": 0.160294371843338,
      "num_tokens": 54744581.0,
      "step": 16385
    },
    {
      "entropy": 6.1786141872406,
      "epoch": 3.1974441517900694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004036530123283504,
      "loss": 5.7746,
      "mean_token_accuracy": 0.16840081438422203,
      "num_tokens": 54761561.0,
      "step": 16390
    },
    {
      "entropy": 6.208243989944458,
      "epoch": 3.1984196663740123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004035953039510952,
      "loss": 5.8344,
      "mean_token_accuracy": 0.162686026096344,
      "num_tokens": 54778457.0,
      "step": 16395
    },
    {
      "entropy": 6.212801313400268,
      "epoch": 3.1993951809579553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040353758300796524,
      "loss": 5.7825,
      "mean_token_accuracy": 0.16219274401664735,
      "num_tokens": 54794992.0,
      "step": 16400
    },
    {
      "entropy": 6.233416175842285,
      "epoch": 3.2003706955418982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040347984950460077,
      "loss": 5.8842,
      "mean_token_accuracy": 0.15783315002918244,
      "num_tokens": 54812263.0,
      "step": 16405
    },
    {
      "entropy": 6.217382383346558,
      "epoch": 3.201346210125841,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040342210344664325,
      "loss": 5.9169,
      "mean_token_accuracy": 0.15325677394866943,
      "num_tokens": 54828059.0,
      "step": 16410
    },
    {
      "entropy": 6.306053543090821,
      "epoch": 3.2023217247097846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004033643448397356,
      "loss": 5.9686,
      "mean_token_accuracy": 0.15171040147542952,
      "num_tokens": 54844615.0,
      "step": 16415
    },
    {
      "entropy": 6.2280679702758786,
      "epoch": 3.2032972392937276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004033065736895215,
      "loss": 5.7955,
      "mean_token_accuracy": 0.15872325152158737,
      "num_tokens": 54861760.0,
      "step": 16420
    },
    {
      "entropy": 6.334616184234619,
      "epoch": 3.2042727538776705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040324879000164644,
      "loss": 5.9232,
      "mean_token_accuracy": 0.15518457889556886,
      "num_tokens": 54878426.0,
      "step": 16425
    },
    {
      "entropy": 6.213060140609741,
      "epoch": 3.2052482684616135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031909937817567,
      "loss": 5.8474,
      "mean_token_accuracy": 0.1592486619949341,
      "num_tokens": 54895173.0,
      "step": 16430
    },
    {
      "entropy": 6.307892179489135,
      "epoch": 3.2062237830455564,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040313318503549996,
      "loss": 5.9374,
      "mean_token_accuracy": 0.14698827415704727,
      "num_tokens": 54911054.0,
      "step": 16435
    },
    {
      "entropy": 6.272445964813232,
      "epoch": 3.2071992976294994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040307536376852504,
      "loss": 5.8561,
      "mean_token_accuracy": 0.1599495604634285,
      "num_tokens": 54926250.0,
      "step": 16440
    },
    {
      "entropy": 6.181629943847656,
      "epoch": 3.208174812213443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040301752998648193,
      "loss": 5.8127,
      "mean_token_accuracy": 0.17114293575286865,
      "num_tokens": 54941810.0,
      "step": 16445
    },
    {
      "entropy": 6.2239542484283445,
      "epoch": 3.2091503267973858,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004029596836950222,
      "loss": 5.9078,
      "mean_token_accuracy": 0.1590113967657089,
      "num_tokens": 54957628.0,
      "step": 16450
    },
    {
      "entropy": 6.19017014503479,
      "epoch": 3.2101258413813287,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040290182489979815,
      "loss": 5.8104,
      "mean_token_accuracy": 0.15814623534679412,
      "num_tokens": 54974349.0,
      "step": 16455
    },
    {
      "entropy": 6.144648742675781,
      "epoch": 3.2111013559652717,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040284395360646366,
      "loss": 5.7403,
      "mean_token_accuracy": 0.1696278214454651,
      "num_tokens": 54991936.0,
      "step": 16460
    },
    {
      "entropy": 6.294503688812256,
      "epoch": 3.2120768705492146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040278606982067363,
      "loss": 5.9496,
      "mean_token_accuracy": 0.15386193245649338,
      "num_tokens": 55009444.0,
      "step": 16465
    },
    {
      "entropy": 6.265997886657715,
      "epoch": 3.2130523851331576,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040272817354808425,
      "loss": 5.8437,
      "mean_token_accuracy": 0.15976627320051193,
      "num_tokens": 55025183.0,
      "step": 16470
    },
    {
      "entropy": 6.140140199661255,
      "epoch": 3.2140278997171006,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004026702647943531,
      "loss": 5.7404,
      "mean_token_accuracy": 0.16760076582431793,
      "num_tokens": 55040996.0,
      "step": 16475
    },
    {
      "entropy": 6.097470998764038,
      "epoch": 3.215003414301044,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040261234356513863,
      "loss": 5.7479,
      "mean_token_accuracy": 0.16538316011428833,
      "num_tokens": 55056979.0,
      "step": 16480
    },
    {
      "entropy": 6.097005033493042,
      "epoch": 3.215978928884987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004025544098661008,
      "loss": 5.6906,
      "mean_token_accuracy": 0.17214101552963257,
      "num_tokens": 55075276.0,
      "step": 16485
    },
    {
      "entropy": 6.146110963821411,
      "epoch": 3.21695444346893,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040249646370290064,
      "loss": 5.7022,
      "mean_token_accuracy": 0.17502323240041734,
      "num_tokens": 55091564.0,
      "step": 16490
    },
    {
      "entropy": 6.124863243103027,
      "epoch": 3.217929958052873,
      "grad_norm": 1.0,
      "learning_rate": 0.0004024385050812004,
      "loss": 5.7831,
      "mean_token_accuracy": 0.1691673532128334,
      "num_tokens": 55109801.0,
      "step": 16495
    },
    {
      "entropy": 6.166194343566895,
      "epoch": 3.218905472636816,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040238053400666374,
      "loss": 5.7956,
      "mean_token_accuracy": 0.16304438859224318,
      "num_tokens": 55127625.0,
      "step": 16500
    },
    {
      "entropy": 6.192845106124878,
      "epoch": 3.2198809872207588,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004023225504849552,
      "loss": 5.8496,
      "mean_token_accuracy": 0.16441356241703034,
      "num_tokens": 55146895.0,
      "step": 16505
    },
    {
      "entropy": 6.239700746536255,
      "epoch": 3.220856501804702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040226455452174086,
      "loss": 5.8578,
      "mean_token_accuracy": 0.15608281046152114,
      "num_tokens": 55163540.0,
      "step": 16510
    },
    {
      "entropy": 6.23892650604248,
      "epoch": 3.221832016388645,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004022065461226879,
      "loss": 5.8394,
      "mean_token_accuracy": 0.15737112760543823,
      "num_tokens": 55180938.0,
      "step": 16515
    },
    {
      "entropy": 6.099393081665039,
      "epoch": 3.222807530972588,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040214852529346445,
      "loss": 5.7127,
      "mean_token_accuracy": 0.17836003750562668,
      "num_tokens": 55197661.0,
      "step": 16520
    },
    {
      "entropy": 6.145405673980713,
      "epoch": 3.223783045556531,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040209049203974035,
      "loss": 5.7385,
      "mean_token_accuracy": 0.16118369102478028,
      "num_tokens": 55214013.0,
      "step": 16525
    },
    {
      "entropy": 6.10662841796875,
      "epoch": 3.224758560140474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004020324463671864,
      "loss": 5.6829,
      "mean_token_accuracy": 0.18397630155086517,
      "num_tokens": 55231443.0,
      "step": 16530
    },
    {
      "entropy": 6.164712142944336,
      "epoch": 3.225734074724417,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004019743882814745,
      "loss": 5.7243,
      "mean_token_accuracy": 0.1731916144490242,
      "num_tokens": 55249367.0,
      "step": 16535
    },
    {
      "entropy": 6.304480409622192,
      "epoch": 3.22670958930836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004019163177882779,
      "loss": 5.9665,
      "mean_token_accuracy": 0.1521339848637581,
      "num_tokens": 55266440.0,
      "step": 16540
    },
    {
      "entropy": 6.242803335189819,
      "epoch": 3.2276851038923033,
      "grad_norm": 1.015625,
      "learning_rate": 0.000401858234893271,
      "loss": 5.8384,
      "mean_token_accuracy": 0.1613006189465523,
      "num_tokens": 55282708.0,
      "step": 16545
    },
    {
      "entropy": 6.035299491882324,
      "epoch": 3.2286606184762463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004018001396021296,
      "loss": 5.6476,
      "mean_token_accuracy": 0.18091916292905807,
      "num_tokens": 55299604.0,
      "step": 16550
    },
    {
      "entropy": 6.191159200668335,
      "epoch": 3.2296361330601893,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040174203192053047,
      "loss": 5.7821,
      "mean_token_accuracy": 0.1591540515422821,
      "num_tokens": 55316865.0,
      "step": 16555
    },
    {
      "entropy": 6.187320423126221,
      "epoch": 3.2306116476441322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004016839118541517,
      "loss": 5.6899,
      "mean_token_accuracy": 0.171175779402256,
      "num_tokens": 55333855.0,
      "step": 16560
    },
    {
      "entropy": 6.192883491516113,
      "epoch": 3.231587162228075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004016257794086726,
      "loss": 5.812,
      "mean_token_accuracy": 0.16298629194498063,
      "num_tokens": 55350793.0,
      "step": 16565
    },
    {
      "entropy": 6.232262039184571,
      "epoch": 3.232562676812018,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040156763458977366,
      "loss": 5.8168,
      "mean_token_accuracy": 0.16072232127189637,
      "num_tokens": 55367367.0,
      "step": 16570
    },
    {
      "entropy": 6.2683452606201175,
      "epoch": 3.2335381913959615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004015094774031365,
      "loss": 5.9386,
      "mean_token_accuracy": 0.14972513169050217,
      "num_tokens": 55385931.0,
      "step": 16575
    },
    {
      "entropy": 6.2345093250274655,
      "epoch": 3.2345137059799045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004014513078544441,
      "loss": 5.9054,
      "mean_token_accuracy": 0.15762123763561248,
      "num_tokens": 55402540.0,
      "step": 16580
    },
    {
      "entropy": 6.100069427490235,
      "epoch": 3.2354892205638475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040139312594938053,
      "loss": 5.6856,
      "mean_token_accuracy": 0.1731901526451111,
      "num_tokens": 55420237.0,
      "step": 16585
    },
    {
      "entropy": 6.163342094421386,
      "epoch": 3.2364647351477904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040133493169363125,
      "loss": 5.7781,
      "mean_token_accuracy": 0.1635628491640091,
      "num_tokens": 55437274.0,
      "step": 16590
    },
    {
      "entropy": 6.325007772445678,
      "epoch": 3.2374402497317334,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040127672509288265,
      "loss": 5.9384,
      "mean_token_accuracy": 0.15247176885604857,
      "num_tokens": 55455129.0,
      "step": 16595
    },
    {
      "entropy": 6.1517250537872314,
      "epoch": 3.2384157643156763,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004012185061528226,
      "loss": 5.8179,
      "mean_token_accuracy": 0.16713568717241287,
      "num_tokens": 55471962.0,
      "step": 16600
    },
    {
      "entropy": 6.202768039703369,
      "epoch": 3.2393912788996198,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004011602748791399,
      "loss": 5.7911,
      "mean_token_accuracy": 0.1664888307452202,
      "num_tokens": 55490566.0,
      "step": 16605
    },
    {
      "entropy": 6.111893367767334,
      "epoch": 3.2403667934835627,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040110203127752486,
      "loss": 5.7705,
      "mean_token_accuracy": 0.16321417987346648,
      "num_tokens": 55507578.0,
      "step": 16610
    },
    {
      "entropy": 6.22696361541748,
      "epoch": 3.2413423080675057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004010437753536687,
      "loss": 5.8359,
      "mean_token_accuracy": 0.16170321255922318,
      "num_tokens": 55523097.0,
      "step": 16615
    },
    {
      "entropy": 6.217107582092285,
      "epoch": 3.2423178226514486,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004009855071132641,
      "loss": 5.8101,
      "mean_token_accuracy": 0.15812478959560394,
      "num_tokens": 55541304.0,
      "step": 16620
    },
    {
      "entropy": 6.278928232192993,
      "epoch": 3.2432933372353916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004009272265620047,
      "loss": 5.9014,
      "mean_token_accuracy": 0.14708152264356614,
      "num_tokens": 55557767.0,
      "step": 16625
    },
    {
      "entropy": 6.292943716049194,
      "epoch": 3.2442688518193346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040086893370558553,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1531719446182251,
      "num_tokens": 55573172.0,
      "step": 16630
    },
    {
      "entropy": 6.271873998641968,
      "epoch": 3.245244366403278,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040081062854970277,
      "loss": 5.8585,
      "mean_token_accuracy": 0.15640549510717391,
      "num_tokens": 55589921.0,
      "step": 16635
    },
    {
      "entropy": 6.240350818634033,
      "epoch": 3.246219880987221,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040075231110005385,
      "loss": 5.9316,
      "mean_token_accuracy": 0.14952096119523048,
      "num_tokens": 55605647.0,
      "step": 16640
    },
    {
      "entropy": 6.144240379333496,
      "epoch": 3.247195395571164,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040069398136233724,
      "loss": 5.7221,
      "mean_token_accuracy": 0.16929909586906433,
      "num_tokens": 55623308.0,
      "step": 16645
    },
    {
      "entropy": 6.165787124633789,
      "epoch": 3.248170910155107,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004006356393422528,
      "loss": 5.7075,
      "mean_token_accuracy": 0.17533146291971208,
      "num_tokens": 55640245.0,
      "step": 16650
    },
    {
      "entropy": 6.182398414611816,
      "epoch": 3.24914642473905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040057728504550136,
      "loss": 5.8214,
      "mean_token_accuracy": 0.16282541304826736,
      "num_tokens": 55657842.0,
      "step": 16655
    },
    {
      "entropy": 6.2330152034759525,
      "epoch": 3.2501219393229928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004005189184777852,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16191678941249849,
      "num_tokens": 55672580.0,
      "step": 16660
    },
    {
      "entropy": 6.220091915130615,
      "epoch": 3.2510974539069357,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004004605396448078,
      "loss": 5.8102,
      "mean_token_accuracy": 0.16015965044498442,
      "num_tokens": 55688806.0,
      "step": 16665
    },
    {
      "entropy": 6.226820945739746,
      "epoch": 3.252072968490879,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040040214855227346,
      "loss": 5.8239,
      "mean_token_accuracy": 0.16656205654144288,
      "num_tokens": 55704989.0,
      "step": 16670
    },
    {
      "entropy": 6.070832538604736,
      "epoch": 3.253048483074822,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004003437452058882,
      "loss": 5.7167,
      "mean_token_accuracy": 0.17278805673122405,
      "num_tokens": 55722448.0,
      "step": 16675
    },
    {
      "entropy": 6.273216438293457,
      "epoch": 3.254023997658765,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004002853296113589,
      "loss": 5.9044,
      "mean_token_accuracy": 0.15791408866643905,
      "num_tokens": 55738862.0,
      "step": 16680
    },
    {
      "entropy": 6.119503545761108,
      "epoch": 3.254999512242708,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004002269017743937,
      "loss": 5.6911,
      "mean_token_accuracy": 0.17048951238393784,
      "num_tokens": 55754593.0,
      "step": 16685
    },
    {
      "entropy": 6.196415901184082,
      "epoch": 3.255975026826651,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040016846170070193,
      "loss": 5.8548,
      "mean_token_accuracy": 0.1565488636493683,
      "num_tokens": 55772039.0,
      "step": 16690
    },
    {
      "entropy": 6.26135401725769,
      "epoch": 3.256950541410594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004001100093959943,
      "loss": 5.8547,
      "mean_token_accuracy": 0.16053869128227233,
      "num_tokens": 55788977.0,
      "step": 16695
    },
    {
      "entropy": 6.225171279907227,
      "epoch": 3.2579260559945373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004000515448659823,
      "loss": 5.87,
      "mean_token_accuracy": 0.15695779472589494,
      "num_tokens": 55805997.0,
      "step": 16700
    },
    {
      "entropy": 6.155216836929322,
      "epoch": 3.2589015705784803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039999306811637915,
      "loss": 5.8054,
      "mean_token_accuracy": 0.1627048671245575,
      "num_tokens": 55822108.0,
      "step": 16705
    },
    {
      "entropy": 6.289952421188355,
      "epoch": 3.2598770851624232,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003999345791528988,
      "loss": 5.9333,
      "mean_token_accuracy": 0.15459804832935334,
      "num_tokens": 55838840.0,
      "step": 16710
    },
    {
      "entropy": 6.237365293502807,
      "epoch": 3.260852599746366,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039987607798125667,
      "loss": 5.8504,
      "mean_token_accuracy": 0.1593086987733841,
      "num_tokens": 55855023.0,
      "step": 16715
    },
    {
      "entropy": 6.210836791992188,
      "epoch": 3.261828114330309,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003998175646071693,
      "loss": 5.8407,
      "mean_token_accuracy": 0.16193434447050095,
      "num_tokens": 55870994.0,
      "step": 16720
    },
    {
      "entropy": 6.177848672866821,
      "epoch": 3.262803628914252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039975903903635435,
      "loss": 5.7916,
      "mean_token_accuracy": 0.16959375888109207,
      "num_tokens": 55887816.0,
      "step": 16725
    },
    {
      "entropy": 6.131797456741333,
      "epoch": 3.263779143498195,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003997005012745307,
      "loss": 5.7369,
      "mean_token_accuracy": 0.16626511737704278,
      "num_tokens": 55905075.0,
      "step": 16730
    },
    {
      "entropy": 6.252405500411987,
      "epoch": 3.2647546580821385,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039964195132741854,
      "loss": 5.8617,
      "mean_token_accuracy": 0.15385382920503615,
      "num_tokens": 55923240.0,
      "step": 16735
    },
    {
      "entropy": 6.188016891479492,
      "epoch": 3.2657301726660815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039958338920073914,
      "loss": 5.8176,
      "mean_token_accuracy": 0.16084687858819963,
      "num_tokens": 55938854.0,
      "step": 16740
    },
    {
      "entropy": 6.192647171020508,
      "epoch": 3.2667056872500244,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039952481490021497,
      "loss": 5.8593,
      "mean_token_accuracy": 0.15907461196184158,
      "num_tokens": 55956748.0,
      "step": 16745
    },
    {
      "entropy": 6.247330999374389,
      "epoch": 3.2676812018339674,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039946622843156973,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16195733547210694,
      "num_tokens": 55971773.0,
      "step": 16750
    },
    {
      "entropy": 6.114775800704956,
      "epoch": 3.2686567164179103,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003994076298005282,
      "loss": 5.7231,
      "mean_token_accuracy": 0.17137501388788223,
      "num_tokens": 55990160.0,
      "step": 16755
    },
    {
      "entropy": 6.215562725067139,
      "epoch": 3.2696322310018533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003993490190128165,
      "loss": 5.789,
      "mean_token_accuracy": 0.16184793561697006,
      "num_tokens": 56007343.0,
      "step": 16760
    },
    {
      "entropy": 6.123782348632813,
      "epoch": 3.2706077455857967,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003992903960741619,
      "loss": 5.7338,
      "mean_token_accuracy": 0.17111219763755797,
      "num_tokens": 56022882.0,
      "step": 16765
    },
    {
      "entropy": 6.173250484466553,
      "epoch": 3.2715832601697397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003992317609902927,
      "loss": 5.807,
      "mean_token_accuracy": 0.1592779502272606,
      "num_tokens": 56039020.0,
      "step": 16770
    },
    {
      "entropy": 6.228856182098388,
      "epoch": 3.2725587747536826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003991731137669386,
      "loss": 5.7931,
      "mean_token_accuracy": 0.16047764122486113,
      "num_tokens": 56054990.0,
      "step": 16775
    },
    {
      "entropy": 6.12404146194458,
      "epoch": 3.2735342893376256,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003991144544098304,
      "loss": 5.7469,
      "mean_token_accuracy": 0.16184705197811128,
      "num_tokens": 56071107.0,
      "step": 16780
    },
    {
      "entropy": 6.2727855205535885,
      "epoch": 3.2745098039215685,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003990557829247,
      "loss": 5.8881,
      "mean_token_accuracy": 0.15721462517976761,
      "num_tokens": 56086275.0,
      "step": 16785
    },
    {
      "entropy": 6.086425018310547,
      "epoch": 3.2754853185055115,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003989970993172808,
      "loss": 5.7189,
      "mean_token_accuracy": 0.17568438798189162,
      "num_tokens": 56104010.0,
      "step": 16790
    },
    {
      "entropy": 6.185404348373413,
      "epoch": 3.2764608330894545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003989384035933068,
      "loss": 5.8074,
      "mean_token_accuracy": 0.16384632289409637,
      "num_tokens": 56121145.0,
      "step": 16795
    },
    {
      "entropy": 6.0659221649169925,
      "epoch": 3.277436347673398,
      "grad_norm": 1.0,
      "learning_rate": 0.0003988796957585139,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17920274138450623,
      "num_tokens": 56138927.0,
      "step": 16800
    },
    {
      "entropy": 6.183497428894043,
      "epoch": 3.278411862257341,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003988209758186386,
      "loss": 5.819,
      "mean_token_accuracy": 0.16573209762573243,
      "num_tokens": 56157550.0,
      "step": 16805
    },
    {
      "entropy": 6.1791003227233885,
      "epoch": 3.279387376841284,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003987622437794188,
      "loss": 5.8109,
      "mean_token_accuracy": 0.1596543475985527,
      "num_tokens": 56174011.0,
      "step": 16810
    },
    {
      "entropy": 6.154550170898437,
      "epoch": 3.2803628914252267,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003987034996465936,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1608641341328621,
      "num_tokens": 56190630.0,
      "step": 16815
    },
    {
      "entropy": 6.292879915237426,
      "epoch": 3.2813384060091697,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039864474342590333,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1649249479174614,
      "num_tokens": 56206397.0,
      "step": 16820
    },
    {
      "entropy": 6.058073663711548,
      "epoch": 3.282313920593113,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039858597512308945,
      "loss": 5.613,
      "mean_token_accuracy": 0.17877697795629502,
      "num_tokens": 56223112.0,
      "step": 16825
    },
    {
      "entropy": 6.254243087768555,
      "epoch": 3.283289435177056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039852719474389454,
      "loss": 5.8785,
      "mean_token_accuracy": 0.15745313167572023,
      "num_tokens": 56238520.0,
      "step": 16830
    },
    {
      "entropy": 6.204789400100708,
      "epoch": 3.284264949760999,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003984684022940624,
      "loss": 5.8421,
      "mean_token_accuracy": 0.16015105545520783,
      "num_tokens": 56256255.0,
      "step": 16835
    },
    {
      "entropy": 6.138209581375122,
      "epoch": 3.285240464344942,
      "grad_norm": 1.015625,
      "learning_rate": 0.000398409597779338,
      "loss": 5.7857,
      "mean_token_accuracy": 0.15960948020219803,
      "num_tokens": 56271725.0,
      "step": 16840
    },
    {
      "entropy": 6.178225135803222,
      "epoch": 3.286215978928885,
      "grad_norm": 1.0,
      "learning_rate": 0.00039835078120546756,
      "loss": 5.7953,
      "mean_token_accuracy": 0.1662612348794937,
      "num_tokens": 56287421.0,
      "step": 16845
    },
    {
      "entropy": 6.236384916305542,
      "epoch": 3.287191493512828,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003982919525781985,
      "loss": 5.9059,
      "mean_token_accuracy": 0.1530366614460945,
      "num_tokens": 56304557.0,
      "step": 16850
    },
    {
      "entropy": 6.219915390014648,
      "epoch": 3.288167008096771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039823311190327913,
      "loss": 5.9257,
      "mean_token_accuracy": 0.16314796060323716,
      "num_tokens": 56323026.0,
      "step": 16855
    },
    {
      "entropy": 6.243680095672607,
      "epoch": 3.289142522680714,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003981742591864594,
      "loss": 5.8267,
      "mean_token_accuracy": 0.15656643956899643,
      "num_tokens": 56340231.0,
      "step": 16860
    },
    {
      "entropy": 6.149862432479859,
      "epoch": 3.2901180372646572,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039811539443349,
      "loss": 5.8247,
      "mean_token_accuracy": 0.16227519810199736,
      "num_tokens": 56357987.0,
      "step": 16865
    },
    {
      "entropy": 6.324792814254761,
      "epoch": 3.2910935518486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000398056517650123,
      "loss": 5.9642,
      "mean_token_accuracy": 0.15354969650506972,
      "num_tokens": 56374854.0,
      "step": 16870
    },
    {
      "entropy": 6.145672702789307,
      "epoch": 3.292069066432543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003979976288421117,
      "loss": 5.744,
      "mean_token_accuracy": 0.17193729877471925,
      "num_tokens": 56391620.0,
      "step": 16875
    },
    {
      "entropy": 6.138344144821167,
      "epoch": 3.293044581016486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003979387280152106,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16336254477500917,
      "num_tokens": 56407892.0,
      "step": 16880
    },
    {
      "entropy": 6.26220736503601,
      "epoch": 3.294020095600429,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039787981517517505,
      "loss": 5.8612,
      "mean_token_accuracy": 0.15254429429769517,
      "num_tokens": 56424005.0,
      "step": 16885
    },
    {
      "entropy": 6.228219795227051,
      "epoch": 3.2949956101843725,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003978208903277619,
      "loss": 5.8075,
      "mean_token_accuracy": 0.15827288925647737,
      "num_tokens": 56439936.0,
      "step": 16890
    },
    {
      "entropy": 6.188315916061401,
      "epoch": 3.2959711247683154,
      "grad_norm": 1.0,
      "learning_rate": 0.0003977619534787292,
      "loss": 5.7876,
      "mean_token_accuracy": 0.15779519230127334,
      "num_tokens": 56456332.0,
      "step": 16895
    },
    {
      "entropy": 6.239431667327881,
      "epoch": 3.2969466393522584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039770300463383575,
      "loss": 5.8215,
      "mean_token_accuracy": 0.159769144654274,
      "num_tokens": 56472294.0,
      "step": 16900
    },
    {
      "entropy": 6.172833013534546,
      "epoch": 3.2979221539362014,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039764404379884216,
      "loss": 5.7533,
      "mean_token_accuracy": 0.16469765603542327,
      "num_tokens": 56489414.0,
      "step": 16905
    },
    {
      "entropy": 6.241561794281006,
      "epoch": 3.2988976685201443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039758507097950965,
      "loss": 5.9116,
      "mean_token_accuracy": 0.1548190027475357,
      "num_tokens": 56506834.0,
      "step": 16910
    },
    {
      "entropy": 6.094025707244873,
      "epoch": 3.2998731831040873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003975260861816009,
      "loss": 5.7365,
      "mean_token_accuracy": 0.1703884020447731,
      "num_tokens": 56523958.0,
      "step": 16915
    },
    {
      "entropy": 6.2822949409484865,
      "epoch": 3.3008486976880302,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003974670894108798,
      "loss": 5.8308,
      "mean_token_accuracy": 0.16249848455190657,
      "num_tokens": 56539889.0,
      "step": 16920
    },
    {
      "entropy": 6.243182277679443,
      "epoch": 3.3018242122719736,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039740808067311107,
      "loss": 5.9009,
      "mean_token_accuracy": 0.1590711370110512,
      "num_tokens": 56556879.0,
      "step": 16925
    },
    {
      "entropy": 6.28345799446106,
      "epoch": 3.3027997268559166,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000397349059974061,
      "loss": 5.9535,
      "mean_token_accuracy": 0.1580958142876625,
      "num_tokens": 56573077.0,
      "step": 16930
    },
    {
      "entropy": 6.275404930114746,
      "epoch": 3.3037752414398596,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039729002731949694,
      "loss": 5.8525,
      "mean_token_accuracy": 0.16064162999391557,
      "num_tokens": 56589183.0,
      "step": 16935
    },
    {
      "entropy": 6.1312634468078615,
      "epoch": 3.3047507560238025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039723098271518716,
      "loss": 5.7607,
      "mean_token_accuracy": 0.16959540098905562,
      "num_tokens": 56607060.0,
      "step": 16940
    },
    {
      "entropy": 6.229372453689575,
      "epoch": 3.3057262706077455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003971719261669014,
      "loss": 5.8708,
      "mean_token_accuracy": 0.15982984751462936,
      "num_tokens": 56623503.0,
      "step": 16945
    },
    {
      "entropy": 6.133183288574219,
      "epoch": 3.3067017851916884,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039711285768041045,
      "loss": 5.7446,
      "mean_token_accuracy": 0.16775506138801574,
      "num_tokens": 56639442.0,
      "step": 16950
    },
    {
      "entropy": 6.176585292816162,
      "epoch": 3.307677299775632,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003970537772614863,
      "loss": 5.8285,
      "mean_token_accuracy": 0.15809365063905717,
      "num_tokens": 56658451.0,
      "step": 16955
    },
    {
      "entropy": 6.248908567428589,
      "epoch": 3.308652814359575,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039699468491590206,
      "loss": 5.8433,
      "mean_token_accuracy": 0.15834278911352156,
      "num_tokens": 56674826.0,
      "step": 16960
    },
    {
      "entropy": 6.158338785171509,
      "epoch": 3.3096283289435178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003969355806494319,
      "loss": 5.7044,
      "mean_token_accuracy": 0.16989672780036927,
      "num_tokens": 56690617.0,
      "step": 16965
    },
    {
      "entropy": 6.159194135665894,
      "epoch": 3.3106038435274607,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003968764644678515,
      "loss": 5.7684,
      "mean_token_accuracy": 0.1611621454358101,
      "num_tokens": 56707169.0,
      "step": 16970
    },
    {
      "entropy": 6.162480592727661,
      "epoch": 3.3115793581114037,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003968173363769373,
      "loss": 5.775,
      "mean_token_accuracy": 0.16317256391048432,
      "num_tokens": 56723448.0,
      "step": 16975
    },
    {
      "entropy": 6.186424779891968,
      "epoch": 3.3125548726953467,
      "grad_norm": 1.0,
      "learning_rate": 0.0003967581963824672,
      "loss": 5.7643,
      "mean_token_accuracy": 0.16840381920337677,
      "num_tokens": 56740535.0,
      "step": 16980
    },
    {
      "entropy": 6.105634021759033,
      "epoch": 3.3135303872792896,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003966990444902201,
      "loss": 5.8229,
      "mean_token_accuracy": 0.1657811388373375,
      "num_tokens": 56758375.0,
      "step": 16985
    },
    {
      "entropy": 6.048989152908325,
      "epoch": 3.314505901863233,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039663988070597604,
      "loss": 5.5766,
      "mean_token_accuracy": 0.17813056111335754,
      "num_tokens": 56776090.0,
      "step": 16990
    },
    {
      "entropy": 6.346331310272217,
      "epoch": 3.315481416447176,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039658070503551646,
      "loss": 6.024,
      "mean_token_accuracy": 0.14637826085090638,
      "num_tokens": 56792627.0,
      "step": 16995
    },
    {
      "entropy": 6.318522787094116,
      "epoch": 3.316456931031119,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039652151748462367,
      "loss": 5.8943,
      "mean_token_accuracy": 0.1503227949142456,
      "num_tokens": 56809025.0,
      "step": 17000
    },
    {
      "entropy": 6.050733518600464,
      "epoch": 3.317432445615062,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003964623180590813,
      "loss": 5.6702,
      "mean_token_accuracy": 0.16749214828014375,
      "num_tokens": 56826043.0,
      "step": 17005
    },
    {
      "entropy": 6.346860694885254,
      "epoch": 3.318407960199005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003964031067646741,
      "loss": 5.9246,
      "mean_token_accuracy": 0.1542916014790535,
      "num_tokens": 56841974.0,
      "step": 17010
    },
    {
      "entropy": 6.237962436676026,
      "epoch": 3.319383474782948,
      "grad_norm": 1.125,
      "learning_rate": 0.00039634388360718797,
      "loss": 5.9087,
      "mean_token_accuracy": 0.15067324489355088,
      "num_tokens": 56859404.0,
      "step": 17015
    },
    {
      "entropy": 6.216332960128784,
      "epoch": 3.3203589893668912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003962846485924101,
      "loss": 5.8025,
      "mean_token_accuracy": 0.1643927127122879,
      "num_tokens": 56875727.0,
      "step": 17020
    },
    {
      "entropy": 6.19213719367981,
      "epoch": 3.321334503950834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003962254017261286,
      "loss": 5.7664,
      "mean_token_accuracy": 0.16317408233880998,
      "num_tokens": 56891212.0,
      "step": 17025
    },
    {
      "entropy": 6.132361650466919,
      "epoch": 3.322310018534777,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039616614301413286,
      "loss": 5.7215,
      "mean_token_accuracy": 0.17384137362241744,
      "num_tokens": 56908005.0,
      "step": 17030
    },
    {
      "entropy": 6.237157535552979,
      "epoch": 3.32328553311872,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003961068724622135,
      "loss": 5.7579,
      "mean_token_accuracy": 0.1687219724059105,
      "num_tokens": 56924993.0,
      "step": 17035
    },
    {
      "entropy": 6.200999355316162,
      "epoch": 3.324261047702663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003960475900761622,
      "loss": 5.8816,
      "mean_token_accuracy": 0.15814142525196076,
      "num_tokens": 56941354.0,
      "step": 17040
    },
    {
      "entropy": 6.1370948314666744,
      "epoch": 3.325236562286606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039598829586177177,
      "loss": 5.7198,
      "mean_token_accuracy": 0.1686364695429802,
      "num_tokens": 56957908.0,
      "step": 17045
    },
    {
      "entropy": 6.24730167388916,
      "epoch": 3.326212076870549,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003959289898248363,
      "loss": 5.8568,
      "mean_token_accuracy": 0.15801601409912108,
      "num_tokens": 56973342.0,
      "step": 17050
    },
    {
      "entropy": 6.2092118740081785,
      "epoch": 3.3271875914544924,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039586967197115095,
      "loss": 5.8028,
      "mean_token_accuracy": 0.16271702200174332,
      "num_tokens": 56990688.0,
      "step": 17055
    },
    {
      "entropy": 6.087743330001831,
      "epoch": 3.3281631060384353,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003958103423065121,
      "loss": 5.8101,
      "mean_token_accuracy": 0.1685144066810608,
      "num_tokens": 57007516.0,
      "step": 17060
    },
    {
      "entropy": 6.1041816711425785,
      "epoch": 3.3291386206223783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003957510008367171,
      "loss": 5.7265,
      "mean_token_accuracy": 0.16595582365989686,
      "num_tokens": 57023771.0,
      "step": 17065
    },
    {
      "entropy": 6.254126405715942,
      "epoch": 3.3301141352063213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003956916475675647,
      "loss": 5.8361,
      "mean_token_accuracy": 0.15764425396919252,
      "num_tokens": 57040018.0,
      "step": 17070
    },
    {
      "entropy": 6.090141534805298,
      "epoch": 3.3310896497902642,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003956322825048546,
      "loss": 5.7466,
      "mean_token_accuracy": 0.1720593363046646,
      "num_tokens": 57055385.0,
      "step": 17075
    },
    {
      "entropy": 6.268667697906494,
      "epoch": 3.3320651643742076,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039557290565438787,
      "loss": 5.8353,
      "mean_token_accuracy": 0.15989254117012025,
      "num_tokens": 57072172.0,
      "step": 17080
    },
    {
      "entropy": 6.191109275817871,
      "epoch": 3.3330406789581506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003955135170219665,
      "loss": 5.8303,
      "mean_token_accuracy": 0.16267281174659728,
      "num_tokens": 57088944.0,
      "step": 17085
    },
    {
      "entropy": 6.264589643478393,
      "epoch": 3.3340161935420936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003954541166133937,
      "loss": 5.8534,
      "mean_token_accuracy": 0.15574304908514022,
      "num_tokens": 57106284.0,
      "step": 17090
    },
    {
      "entropy": 6.183551025390625,
      "epoch": 3.3349917081260365,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000395394704434474,
      "loss": 5.7442,
      "mean_token_accuracy": 0.17194441109895706,
      "num_tokens": 57123507.0,
      "step": 17095
    },
    {
      "entropy": 6.170405721664428,
      "epoch": 3.3359672227099795,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003953352804910129,
      "loss": 5.7973,
      "mean_token_accuracy": 0.16517878025770188,
      "num_tokens": 57140944.0,
      "step": 17100
    },
    {
      "entropy": 6.241258716583252,
      "epoch": 3.3369427372939224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000395275844788817,
      "loss": 5.8398,
      "mean_token_accuracy": 0.16453058421611785,
      "num_tokens": 57156582.0,
      "step": 17105
    },
    {
      "entropy": 6.193296623229981,
      "epoch": 3.3379182518778654,
      "grad_norm": 1.0,
      "learning_rate": 0.0003952163973336943,
      "loss": 5.8194,
      "mean_token_accuracy": 0.1608797565102577,
      "num_tokens": 57172230.0,
      "step": 17110
    },
    {
      "entropy": 6.263594627380371,
      "epoch": 3.3388937664618084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003951569381314537,
      "loss": 5.8811,
      "mean_token_accuracy": 0.1557580664753914,
      "num_tokens": 57189176.0,
      "step": 17115
    },
    {
      "entropy": 6.164998960494995,
      "epoch": 3.3398692810457518,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003950974671879053,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1770977035164833,
      "num_tokens": 57205802.0,
      "step": 17120
    },
    {
      "entropy": 6.1226606369018555,
      "epoch": 3.3408447956296947,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003950379845088605,
      "loss": 5.6866,
      "mean_token_accuracy": 0.165744286775589,
      "num_tokens": 57222224.0,
      "step": 17125
    },
    {
      "entropy": 6.119365215301514,
      "epoch": 3.3418203102136377,
      "grad_norm": 1.0,
      "learning_rate": 0.0003949784901001315,
      "loss": 5.7648,
      "mean_token_accuracy": 0.16292551159858704,
      "num_tokens": 57236920.0,
      "step": 17130
    },
    {
      "entropy": 6.148439025878906,
      "epoch": 3.3427958247975806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039491898396753225,
      "loss": 5.7679,
      "mean_token_accuracy": 0.16937180012464523,
      "num_tokens": 57254076.0,
      "step": 17135
    },
    {
      "entropy": 6.058057689666748,
      "epoch": 3.3437713393815236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039485946611687725,
      "loss": 5.6485,
      "mean_token_accuracy": 0.1775268331170082,
      "num_tokens": 57269932.0,
      "step": 17140
    },
    {
      "entropy": 6.2618796825408936,
      "epoch": 3.344746853965467,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039479993655398236,
      "loss": 5.9295,
      "mean_token_accuracy": 0.1499936029314995,
      "num_tokens": 57286571.0,
      "step": 17145
    },
    {
      "entropy": 6.230014038085938,
      "epoch": 3.34572236854941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003947403952846647,
      "loss": 5.7982,
      "mean_token_accuracy": 0.17145951539278032,
      "num_tokens": 57304076.0,
      "step": 17150
    },
    {
      "entropy": 6.211432409286499,
      "epoch": 3.346697883133353,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003946808423147423,
      "loss": 5.8294,
      "mean_token_accuracy": 0.16094666421413423,
      "num_tokens": 57322032.0,
      "step": 17155
    },
    {
      "entropy": 6.236704874038696,
      "epoch": 3.347673397717296,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003946212776500346,
      "loss": 5.914,
      "mean_token_accuracy": 0.16148820221424104,
      "num_tokens": 57338101.0,
      "step": 17160
    },
    {
      "entropy": 6.313989591598511,
      "epoch": 3.348648912301239,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000394561701296362,
      "loss": 5.9253,
      "mean_token_accuracy": 0.15732854902744292,
      "num_tokens": 57355964.0,
      "step": 17165
    },
    {
      "entropy": 6.210937452316284,
      "epoch": 3.349624426885182,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000394502113259546,
      "loss": 5.8401,
      "mean_token_accuracy": 0.15907954126596452,
      "num_tokens": 57372645.0,
      "step": 17170
    },
    {
      "entropy": 6.118713855743408,
      "epoch": 3.3505999414691248,
      "grad_norm": 1.0,
      "learning_rate": 0.0003944425135454095,
      "loss": 5.6749,
      "mean_token_accuracy": 0.17475084215402603,
      "num_tokens": 57388415.0,
      "step": 17175
    },
    {
      "entropy": 6.224962043762207,
      "epoch": 3.351575456053068,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003943829021597762,
      "loss": 5.8491,
      "mean_token_accuracy": 0.1588088721036911,
      "num_tokens": 57404027.0,
      "step": 17180
    },
    {
      "entropy": 6.189910745620727,
      "epoch": 3.352550970637011,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003943232791084712,
      "loss": 5.8656,
      "mean_token_accuracy": 0.1635876327753067,
      "num_tokens": 57421229.0,
      "step": 17185
    },
    {
      "entropy": 6.235906171798706,
      "epoch": 3.353526485220954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003942636443973207,
      "loss": 5.8407,
      "mean_token_accuracy": 0.15879758894443513,
      "num_tokens": 57436151.0,
      "step": 17190
    },
    {
      "entropy": 6.18425669670105,
      "epoch": 3.354501999804897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003942039980321519,
      "loss": 5.7642,
      "mean_token_accuracy": 0.17047473043203354,
      "num_tokens": 57452832.0,
      "step": 17195
    },
    {
      "entropy": 6.0099029541015625,
      "epoch": 3.35547751438884,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039414434001879326,
      "loss": 5.634,
      "mean_token_accuracy": 0.18321929574012757,
      "num_tokens": 57470926.0,
      "step": 17200
    },
    {
      "entropy": 6.167320632934571,
      "epoch": 3.356453028972783,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039408467036307435,
      "loss": 5.7971,
      "mean_token_accuracy": 0.166639444231987,
      "num_tokens": 57489120.0,
      "step": 17205
    },
    {
      "entropy": 6.253600120544434,
      "epoch": 3.3574285435567264,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039402498907082587,
      "loss": 5.8238,
      "mean_token_accuracy": 0.16470373719930648,
      "num_tokens": 57505553.0,
      "step": 17210
    },
    {
      "entropy": 6.165506505966187,
      "epoch": 3.3584040581406693,
      "grad_norm": 1.0,
      "learning_rate": 0.00039396529614787963,
      "loss": 5.7869,
      "mean_token_accuracy": 0.16203253865242004,
      "num_tokens": 57521827.0,
      "step": 17215
    },
    {
      "entropy": 6.151413774490356,
      "epoch": 3.3593795727246123,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039390559160006874,
      "loss": 5.7988,
      "mean_token_accuracy": 0.16629012078046798,
      "num_tokens": 57537221.0,
      "step": 17220
    },
    {
      "entropy": 6.147352504730224,
      "epoch": 3.3603550873085553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003938458754332272,
      "loss": 5.7499,
      "mean_token_accuracy": 0.17478955090045928,
      "num_tokens": 57555468.0,
      "step": 17225
    },
    {
      "entropy": 6.1846216201782225,
      "epoch": 3.361330601892498,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003937861476531902,
      "loss": 5.787,
      "mean_token_accuracy": 0.1651604637503624,
      "num_tokens": 57573616.0,
      "step": 17230
    },
    {
      "entropy": 6.24920687675476,
      "epoch": 3.362306116476441,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039372640826579433,
      "loss": 5.8129,
      "mean_token_accuracy": 0.16480155438184738,
      "num_tokens": 57591393.0,
      "step": 17235
    },
    {
      "entropy": 6.2184549331665036,
      "epoch": 3.363281631060384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039366665727687695,
      "loss": 5.849,
      "mean_token_accuracy": 0.1566319927573204,
      "num_tokens": 57607149.0,
      "step": 17240
    },
    {
      "entropy": 6.248511695861817,
      "epoch": 3.3642571456443275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003936068946922767,
      "loss": 5.862,
      "mean_token_accuracy": 0.1569177657365799,
      "num_tokens": 57623344.0,
      "step": 17245
    },
    {
      "entropy": 6.1625152111053465,
      "epoch": 3.3652326602282705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039354712051783346,
      "loss": 5.8031,
      "mean_token_accuracy": 0.16246427595615387,
      "num_tokens": 57640190.0,
      "step": 17250
    },
    {
      "entropy": 6.1889348983764645,
      "epoch": 3.3662081748122135,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003934873347593882,
      "loss": 5.8002,
      "mean_token_accuracy": 0.1591624066233635,
      "num_tokens": 57657205.0,
      "step": 17255
    },
    {
      "entropy": 6.218704509735107,
      "epoch": 3.3671836893961564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003934275374227828,
      "loss": 5.8355,
      "mean_token_accuracy": 0.16099630370736123,
      "num_tokens": 57673241.0,
      "step": 17260
    },
    {
      "entropy": 6.029776954650879,
      "epoch": 3.3681592039800994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003933677285138605,
      "loss": 5.7017,
      "mean_token_accuracy": 0.18283040523529054,
      "num_tokens": 57691316.0,
      "step": 17265
    },
    {
      "entropy": 6.230903577804566,
      "epoch": 3.3691347185640423,
      "grad_norm": 1.125,
      "learning_rate": 0.00039330790803846573,
      "loss": 5.8509,
      "mean_token_accuracy": 0.1573049396276474,
      "num_tokens": 57708216.0,
      "step": 17270
    },
    {
      "entropy": 6.233055973052979,
      "epoch": 3.3701102331479857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003932480760024437,
      "loss": 5.8197,
      "mean_token_accuracy": 0.15629468113183975,
      "num_tokens": 57725567.0,
      "step": 17275
    },
    {
      "entropy": 6.176669645309448,
      "epoch": 3.3710857477319287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039318823241164125,
      "loss": 5.7524,
      "mean_token_accuracy": 0.17170178443193435,
      "num_tokens": 57741984.0,
      "step": 17280
    },
    {
      "entropy": 6.244108724594116,
      "epoch": 3.3720612623158717,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003931283772719059,
      "loss": 5.8837,
      "mean_token_accuracy": 0.1573283314704895,
      "num_tokens": 57758532.0,
      "step": 17285
    },
    {
      "entropy": 6.159806632995606,
      "epoch": 3.3730367768998146,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003930685105890866,
      "loss": 5.834,
      "mean_token_accuracy": 0.1640821576118469,
      "num_tokens": 57776602.0,
      "step": 17290
    },
    {
      "entropy": 6.092983627319336,
      "epoch": 3.3740122914837576,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00039300863236903314,
      "loss": 5.6982,
      "mean_token_accuracy": 0.16620202511548995,
      "num_tokens": 57795269.0,
      "step": 17295
    },
    {
      "entropy": 6.103784561157227,
      "epoch": 3.3749878060677005,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003929487426175968,
      "loss": 5.7179,
      "mean_token_accuracy": 0.17325290888547898,
      "num_tokens": 57813301.0,
      "step": 17300
    },
    {
      "entropy": 6.179039525985718,
      "epoch": 3.3759633206516435,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039288884134062974,
      "loss": 5.8417,
      "mean_token_accuracy": 0.16849416196346284,
      "num_tokens": 57830314.0,
      "step": 17305
    },
    {
      "entropy": 6.211490201950073,
      "epoch": 3.376938835235587,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003928289285439852,
      "loss": 5.8114,
      "mean_token_accuracy": 0.16158153414726256,
      "num_tokens": 57847309.0,
      "step": 17310
    },
    {
      "entropy": 6.260879468917847,
      "epoch": 3.37791434981953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003927690042335178,
      "loss": 5.8775,
      "mean_token_accuracy": 0.15851065516471863,
      "num_tokens": 57863050.0,
      "step": 17315
    },
    {
      "entropy": 6.220606899261474,
      "epoch": 3.378889864403473,
      "grad_norm": 0.921875,
      "learning_rate": 0.000392709068415083,
      "loss": 5.8849,
      "mean_token_accuracy": 0.15837692618370056,
      "num_tokens": 57880469.0,
      "step": 17320
    },
    {
      "entropy": 6.247674512863159,
      "epoch": 3.379865378987416,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039264912109453746,
      "loss": 5.8796,
      "mean_token_accuracy": 0.15866979956626892,
      "num_tokens": 57898090.0,
      "step": 17325
    },
    {
      "entropy": 6.218291521072388,
      "epoch": 3.3808408935713588,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003925891622777392,
      "loss": 5.7798,
      "mean_token_accuracy": 0.16376422196626664,
      "num_tokens": 57914343.0,
      "step": 17330
    },
    {
      "entropy": 6.276689052581787,
      "epoch": 3.381816408155302,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003925291919705472,
      "loss": 5.8704,
      "mean_token_accuracy": 0.15594725906848908,
      "num_tokens": 57930471.0,
      "step": 17335
    },
    {
      "entropy": 5.998125791549683,
      "epoch": 3.382791922739245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039246921017882123,
      "loss": 5.6187,
      "mean_token_accuracy": 0.17995610237121581,
      "num_tokens": 57947766.0,
      "step": 17340
    },
    {
      "entropy": 6.30804409980774,
      "epoch": 3.383767437323188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003924092169084228,
      "loss": 5.9482,
      "mean_token_accuracy": 0.15212342143058777,
      "num_tokens": 57964205.0,
      "step": 17345
    },
    {
      "entropy": 6.224425649642944,
      "epoch": 3.384742951907131,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039234921216521417,
      "loss": 5.8886,
      "mean_token_accuracy": 0.15928155183792114,
      "num_tokens": 57982307.0,
      "step": 17350
    },
    {
      "entropy": 6.191834020614624,
      "epoch": 3.385718466491074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039228919595505875,
      "loss": 5.786,
      "mean_token_accuracy": 0.1654949501156807,
      "num_tokens": 57999302.0,
      "step": 17355
    },
    {
      "entropy": 6.190549898147583,
      "epoch": 3.386693981075017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003922291682838211,
      "loss": 5.7767,
      "mean_token_accuracy": 0.16740838438272476,
      "num_tokens": 58016578.0,
      "step": 17360
    },
    {
      "entropy": 6.182544612884522,
      "epoch": 3.38766949565896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000392169129157367,
      "loss": 5.7771,
      "mean_token_accuracy": 0.16047187447547911,
      "num_tokens": 58033770.0,
      "step": 17365
    },
    {
      "entropy": 6.105577898025513,
      "epoch": 3.3886450102429033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003921090785815632,
      "loss": 5.6997,
      "mean_token_accuracy": 0.16595925986766816,
      "num_tokens": 58050075.0,
      "step": 17370
    },
    {
      "entropy": 6.245540761947632,
      "epoch": 3.3896205248268463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003920490165622775,
      "loss": 5.8786,
      "mean_token_accuracy": 0.16037721931934357,
      "num_tokens": 58067020.0,
      "step": 17375
    },
    {
      "entropy": 6.215511131286621,
      "epoch": 3.3905960394107892,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039198894310537923,
      "loss": 5.8009,
      "mean_token_accuracy": 0.1629573941230774,
      "num_tokens": 58081922.0,
      "step": 17380
    },
    {
      "entropy": 6.202944564819336,
      "epoch": 3.391571553994732,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003919288582167383,
      "loss": 5.8049,
      "mean_token_accuracy": 0.16351406127214432,
      "num_tokens": 58097929.0,
      "step": 17385
    },
    {
      "entropy": 6.320071268081665,
      "epoch": 3.392547068578675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039186876190222607,
      "loss": 5.8884,
      "mean_token_accuracy": 0.15634384304285048,
      "num_tokens": 58115299.0,
      "step": 17390
    },
    {
      "entropy": 6.201424503326416,
      "epoch": 3.393522583162618,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039180865416771506,
      "loss": 5.7664,
      "mean_token_accuracy": 0.16695133000612258,
      "num_tokens": 58131333.0,
      "step": 17395
    },
    {
      "entropy": 6.097601795196534,
      "epoch": 3.3944980977465615,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003917485350190786,
      "loss": 5.7527,
      "mean_token_accuracy": 0.16887993067502977,
      "num_tokens": 58147556.0,
      "step": 17400
    },
    {
      "entropy": 6.224594497680664,
      "epoch": 3.3954736123305045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039168840446219145,
      "loss": 5.8724,
      "mean_token_accuracy": 0.15910903066396714,
      "num_tokens": 58163731.0,
      "step": 17405
    },
    {
      "entropy": 6.274510526657105,
      "epoch": 3.3964491269144474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039162826250292926,
      "loss": 5.8659,
      "mean_token_accuracy": 0.15857531279325485,
      "num_tokens": 58180842.0,
      "step": 17410
    },
    {
      "entropy": 6.164380216598511,
      "epoch": 3.3974246414983904,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039156810914716897,
      "loss": 5.8112,
      "mean_token_accuracy": 0.1632729724049568,
      "num_tokens": 58198064.0,
      "step": 17415
    },
    {
      "entropy": 6.2249531745910645,
      "epoch": 3.3984001560823334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039150794440078845,
      "loss": 5.8007,
      "mean_token_accuracy": 0.1617382809519768,
      "num_tokens": 58214233.0,
      "step": 17420
    },
    {
      "entropy": 6.175877857208252,
      "epoch": 3.3993756706662763,
      "grad_norm": 1.0,
      "learning_rate": 0.0003914477682696668,
      "loss": 5.7734,
      "mean_token_accuracy": 0.1663191944360733,
      "num_tokens": 58231940.0,
      "step": 17425
    },
    {
      "entropy": 6.238563823699951,
      "epoch": 3.4003511852502193,
      "grad_norm": 1.0,
      "learning_rate": 0.0003913875807596844,
      "loss": 5.8164,
      "mean_token_accuracy": 0.16034474074840546,
      "num_tokens": 58248076.0,
      "step": 17430
    },
    {
      "entropy": 6.214708757400513,
      "epoch": 3.4013266998341627,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003913273818767223,
      "loss": 5.8564,
      "mean_token_accuracy": 0.15924559235572816,
      "num_tokens": 58264548.0,
      "step": 17435
    },
    {
      "entropy": 6.180126047134399,
      "epoch": 3.4023022144181057,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039126717162666307,
      "loss": 5.7897,
      "mean_token_accuracy": 0.1640253856778145,
      "num_tokens": 58280177.0,
      "step": 17440
    },
    {
      "entropy": 6.217389249801636,
      "epoch": 3.4032777290020486,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039120695001539026,
      "loss": 5.8077,
      "mean_token_accuracy": 0.16347962021827697,
      "num_tokens": 58297083.0,
      "step": 17445
    },
    {
      "entropy": 6.248660326004028,
      "epoch": 3.4042532435859916,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039114671704878844,
      "loss": 5.8557,
      "mean_token_accuracy": 0.16372911930084227,
      "num_tokens": 58313098.0,
      "step": 17450
    },
    {
      "entropy": 6.148215055465698,
      "epoch": 3.4052287581699345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003910864727327434,
      "loss": 5.7493,
      "mean_token_accuracy": 0.17352089136838914,
      "num_tokens": 58329618.0,
      "step": 17455
    },
    {
      "entropy": 6.21085033416748,
      "epoch": 3.4062042727538775,
      "grad_norm": 1.0,
      "learning_rate": 0.00039102621707314195,
      "loss": 5.738,
      "mean_token_accuracy": 0.16883869767189025,
      "num_tokens": 58347133.0,
      "step": 17460
    },
    {
      "entropy": 6.1128833293914795,
      "epoch": 3.407179787337821,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003909659500758721,
      "loss": 5.7695,
      "mean_token_accuracy": 0.16443999111652374,
      "num_tokens": 58363550.0,
      "step": 17465
    },
    {
      "entropy": 6.176812267303466,
      "epoch": 3.408155301921764,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000390905671746823,
      "loss": 5.7116,
      "mean_token_accuracy": 0.17116509079933168,
      "num_tokens": 58379437.0,
      "step": 17470
    },
    {
      "entropy": 6.21966028213501,
      "epoch": 3.409130816505707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003908453820918846,
      "loss": 5.8795,
      "mean_token_accuracy": 0.1597173884510994,
      "num_tokens": 58396459.0,
      "step": 17475
    },
    {
      "entropy": 6.159523820877075,
      "epoch": 3.41010633108965,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039078508111694844,
      "loss": 5.7589,
      "mean_token_accuracy": 0.1704690396785736,
      "num_tokens": 58411926.0,
      "step": 17480
    },
    {
      "entropy": 6.1946361541748045,
      "epoch": 3.4110818456735927,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003907247688279068,
      "loss": 5.8542,
      "mean_token_accuracy": 0.1603941038250923,
      "num_tokens": 58428024.0,
      "step": 17485
    },
    {
      "entropy": 6.163990592956543,
      "epoch": 3.4120573602575357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003906644452306532,
      "loss": 5.7995,
      "mean_token_accuracy": 0.168952077627182,
      "num_tokens": 58446112.0,
      "step": 17490
    },
    {
      "entropy": 6.20738959312439,
      "epoch": 3.4130328748414787,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039060411033108223,
      "loss": 5.8044,
      "mean_token_accuracy": 0.16794513612985612,
      "num_tokens": 58462903.0,
      "step": 17495
    },
    {
      "entropy": 6.172256231307983,
      "epoch": 3.414008389425422,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003905437641350897,
      "loss": 5.7454,
      "mean_token_accuracy": 0.16748764663934707,
      "num_tokens": 58478257.0,
      "step": 17500
    },
    {
      "entropy": 6.252355337142944,
      "epoch": 3.414983904009365,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039048340664857225,
      "loss": 5.7827,
      "mean_token_accuracy": 0.17153455018997193,
      "num_tokens": 58493742.0,
      "step": 17505
    },
    {
      "entropy": 5.928999042510986,
      "epoch": 3.415959418593308,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003904230378774279,
      "loss": 5.5661,
      "mean_token_accuracy": 0.18449084758758544,
      "num_tokens": 58512808.0,
      "step": 17510
    },
    {
      "entropy": 6.255723714828491,
      "epoch": 3.416934933177251,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003903626578275557,
      "loss": 5.8408,
      "mean_token_accuracy": 0.165750914812088,
      "num_tokens": 58530001.0,
      "step": 17515
    },
    {
      "entropy": 6.169854736328125,
      "epoch": 3.417910447761194,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039030226650485576,
      "loss": 5.7549,
      "mean_token_accuracy": 0.1715557634830475,
      "num_tokens": 58546710.0,
      "step": 17520
    },
    {
      "entropy": 6.096439599990845,
      "epoch": 3.4188859623451373,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003902418639152293,
      "loss": 5.7392,
      "mean_token_accuracy": 0.17267160713672638,
      "num_tokens": 58563931.0,
      "step": 17525
    },
    {
      "entropy": 6.237244081497193,
      "epoch": 3.4198614769290803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003901814500645785,
      "loss": 5.8299,
      "mean_token_accuracy": 0.1613943710923195,
      "num_tokens": 58579228.0,
      "step": 17530
    },
    {
      "entropy": 6.231747150421143,
      "epoch": 3.4208369915130232,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000390121024958807,
      "loss": 5.86,
      "mean_token_accuracy": 0.1534218043088913,
      "num_tokens": 58595177.0,
      "step": 17535
    },
    {
      "entropy": 6.090013694763184,
      "epoch": 3.421812506096966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003900605886038192,
      "loss": 5.6902,
      "mean_token_accuracy": 0.17683519423007965,
      "num_tokens": 58612465.0,
      "step": 17540
    },
    {
      "entropy": 6.149916696548462,
      "epoch": 3.422788020680909,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039000014100552084,
      "loss": 5.7583,
      "mean_token_accuracy": 0.174143947660923,
      "num_tokens": 58628696.0,
      "step": 17545
    },
    {
      "entropy": 6.191145563125611,
      "epoch": 3.423763535264852,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038993968216981847,
      "loss": 5.7937,
      "mean_token_accuracy": 0.16245743185281752,
      "num_tokens": 58645972.0,
      "step": 17550
    },
    {
      "entropy": 6.199252653121948,
      "epoch": 3.424739049848795,
      "grad_norm": 1.0,
      "learning_rate": 0.00038987921210262007,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1640791490674019,
      "num_tokens": 58662453.0,
      "step": 17555
    },
    {
      "entropy": 5.983352899551392,
      "epoch": 3.425714564432738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003898187308098345,
      "loss": 5.5489,
      "mean_token_accuracy": 0.19246771484613417,
      "num_tokens": 58680046.0,
      "step": 17560
    },
    {
      "entropy": 6.283593225479126,
      "epoch": 3.4266900790166814,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003897582382973717,
      "loss": 5.8955,
      "mean_token_accuracy": 0.15345927476882934,
      "num_tokens": 58696851.0,
      "step": 17565
    },
    {
      "entropy": 6.169867372512817,
      "epoch": 3.4276655936006244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038969773457114295,
      "loss": 5.792,
      "mean_token_accuracy": 0.16575236171483992,
      "num_tokens": 58713095.0,
      "step": 17570
    },
    {
      "entropy": 6.081604909896851,
      "epoch": 3.4286411081845674,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038963721963706035,
      "loss": 5.6946,
      "mean_token_accuracy": 0.17334526479244233,
      "num_tokens": 58729997.0,
      "step": 17575
    },
    {
      "entropy": 6.171984004974365,
      "epoch": 3.4296166227685103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003895766935010372,
      "loss": 5.8296,
      "mean_token_accuracy": 0.1678292602300644,
      "num_tokens": 58749063.0,
      "step": 17580
    },
    {
      "entropy": 6.29621205329895,
      "epoch": 3.4305921373524533,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038951615616898797,
      "loss": 5.8545,
      "mean_token_accuracy": 0.15899531841278075,
      "num_tokens": 58765158.0,
      "step": 17585
    },
    {
      "entropy": 6.1590064525604244,
      "epoch": 3.4315676519363967,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038945560764682807,
      "loss": 5.79,
      "mean_token_accuracy": 0.16476368010044098,
      "num_tokens": 58781523.0,
      "step": 17590
    },
    {
      "entropy": 6.176547336578369,
      "epoch": 3.4325431665203396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038939504794047414,
      "loss": 5.713,
      "mean_token_accuracy": 0.16856082826852797,
      "num_tokens": 58798436.0,
      "step": 17595
    },
    {
      "entropy": 6.247156476974487,
      "epoch": 3.4335186811042826,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003893344770558438,
      "loss": 5.8907,
      "mean_token_accuracy": 0.1612651228904724,
      "num_tokens": 58814839.0,
      "step": 17600
    },
    {
      "entropy": 6.247379446029663,
      "epoch": 3.4344941956882256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038927389499885583,
      "loss": 5.8816,
      "mean_token_accuracy": 0.15997202694416046,
      "num_tokens": 58831752.0,
      "step": 17605
    },
    {
      "entropy": 6.195089721679688,
      "epoch": 3.4354697102721685,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003892133017754302,
      "loss": 5.8042,
      "mean_token_accuracy": 0.17042580246925354,
      "num_tokens": 58848975.0,
      "step": 17610
    },
    {
      "entropy": 6.239693975448608,
      "epoch": 3.4364452248561115,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038915269739148774,
      "loss": 5.8534,
      "mean_token_accuracy": 0.15912223905324935,
      "num_tokens": 58864413.0,
      "step": 17615
    },
    {
      "entropy": 6.084351634979248,
      "epoch": 3.4374207394400544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003890920818529506,
      "loss": 5.6304,
      "mean_token_accuracy": 0.17553916722536086,
      "num_tokens": 58881151.0,
      "step": 17620
    },
    {
      "entropy": 6.20175576210022,
      "epoch": 3.438396254023998,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003890314551657417,
      "loss": 5.7455,
      "mean_token_accuracy": 0.17051013559103012,
      "num_tokens": 58899699.0,
      "step": 17625
    },
    {
      "entropy": 6.173539829254151,
      "epoch": 3.439371768607941,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003889708173357856,
      "loss": 5.7422,
      "mean_token_accuracy": 0.16583328396081926,
      "num_tokens": 58915193.0,
      "step": 17630
    },
    {
      "entropy": 6.0965910911560055,
      "epoch": 3.4403472831918838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003889101683690073,
      "loss": 5.6829,
      "mean_token_accuracy": 0.17717451900243758,
      "num_tokens": 58931385.0,
      "step": 17635
    },
    {
      "entropy": 6.228127813339233,
      "epoch": 3.4413227977758267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038884950827133336,
      "loss": 5.8621,
      "mean_token_accuracy": 0.16035036891698837,
      "num_tokens": 58946498.0,
      "step": 17640
    },
    {
      "entropy": 6.237435150146484,
      "epoch": 3.4422983123597697,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038878883704869126,
      "loss": 5.8212,
      "mean_token_accuracy": 0.16369762122631074,
      "num_tokens": 58964885.0,
      "step": 17645
    },
    {
      "entropy": 6.204268121719361,
      "epoch": 3.4432738269437126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003887281547070095,
      "loss": 5.7905,
      "mean_token_accuracy": 0.16579457223415375,
      "num_tokens": 58981866.0,
      "step": 17650
    },
    {
      "entropy": 6.118002462387085,
      "epoch": 3.444249341527656,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003886674612522179,
      "loss": 5.7671,
      "mean_token_accuracy": 0.17365244776010513,
      "num_tokens": 58999688.0,
      "step": 17655
    },
    {
      "entropy": 6.164096164703369,
      "epoch": 3.445224856111599,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038860675669024703,
      "loss": 5.7211,
      "mean_token_accuracy": 0.17746117115020751,
      "num_tokens": 59016385.0,
      "step": 17660
    },
    {
      "entropy": 6.21614179611206,
      "epoch": 3.446200370695542,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003885460410270288,
      "loss": 5.8401,
      "mean_token_accuracy": 0.1587155357003212,
      "num_tokens": 59031825.0,
      "step": 17665
    },
    {
      "entropy": 6.24473090171814,
      "epoch": 3.447175885279485,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003884853142684961,
      "loss": 5.8757,
      "mean_token_accuracy": 0.1566457986831665,
      "num_tokens": 59048310.0,
      "step": 17670
    },
    {
      "entropy": 6.137642908096313,
      "epoch": 3.448151399863428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038842457642058304,
      "loss": 5.7234,
      "mean_token_accuracy": 0.17018169462680816,
      "num_tokens": 59066440.0,
      "step": 17675
    },
    {
      "entropy": 6.120011615753174,
      "epoch": 3.449126914447371,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038836382748922457,
      "loss": 5.7296,
      "mean_token_accuracy": 0.17363833785057067,
      "num_tokens": 59083904.0,
      "step": 17680
    },
    {
      "entropy": 6.106198167800903,
      "epoch": 3.450102429031314,
      "grad_norm": 1.0,
      "learning_rate": 0.00038830306748035686,
      "loss": 5.7052,
      "mean_token_accuracy": 0.1731279507279396,
      "num_tokens": 59099991.0,
      "step": 17685
    },
    {
      "entropy": 6.195065450668335,
      "epoch": 3.451077943615257,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003882422963999173,
      "loss": 5.8227,
      "mean_token_accuracy": 0.16623908430337905,
      "num_tokens": 59116770.0,
      "step": 17690
    },
    {
      "entropy": 6.289421701431275,
      "epoch": 3.4520534581992,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003881815142538442,
      "loss": 5.8463,
      "mean_token_accuracy": 0.16629332304000854,
      "num_tokens": 59134722.0,
      "step": 17695
    },
    {
      "entropy": 6.182572698593139,
      "epoch": 3.453028972783143,
      "grad_norm": 1.0,
      "learning_rate": 0.00038812072104807676,
      "loss": 5.7765,
      "mean_token_accuracy": 0.16728079468011856,
      "num_tokens": 59151345.0,
      "step": 17700
    },
    {
      "entropy": 6.1536908626556395,
      "epoch": 3.454004487367086,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003880599167885557,
      "loss": 5.7319,
      "mean_token_accuracy": 0.17319904267787933,
      "num_tokens": 59166492.0,
      "step": 17705
    },
    {
      "entropy": 6.246331310272216,
      "epoch": 3.454980001951029,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038799910148122255,
      "loss": 5.8738,
      "mean_token_accuracy": 0.16505365520715715,
      "num_tokens": 59183570.0,
      "step": 17710
    },
    {
      "entropy": 6.2877795696258545,
      "epoch": 3.455955516534972,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038793827513201987,
      "loss": 5.9303,
      "mean_token_accuracy": 0.15277693271636963,
      "num_tokens": 59200175.0,
      "step": 17715
    },
    {
      "entropy": 6.246874618530273,
      "epoch": 3.4569310311189154,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003878774377468915,
      "loss": 5.8393,
      "mean_token_accuracy": 0.16410023719072342,
      "num_tokens": 59216275.0,
      "step": 17720
    },
    {
      "entropy": 6.138171100616455,
      "epoch": 3.4579065457028584,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038781658933178224,
      "loss": 5.7032,
      "mean_token_accuracy": 0.17320085018873216,
      "num_tokens": 59233341.0,
      "step": 17725
    },
    {
      "entropy": 6.084492158889771,
      "epoch": 3.4588820602868013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003877557298926379,
      "loss": 5.7218,
      "mean_token_accuracy": 0.17377087473869324,
      "num_tokens": 59248631.0,
      "step": 17730
    },
    {
      "entropy": 6.2406312942504885,
      "epoch": 3.4598575748707443,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003876948594354056,
      "loss": 5.8572,
      "mean_token_accuracy": 0.16592158377170563,
      "num_tokens": 59264645.0,
      "step": 17735
    },
    {
      "entropy": 6.256051731109619,
      "epoch": 3.4608330894546873,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003876339779660332,
      "loss": 5.7846,
      "mean_token_accuracy": 0.1601938858628273,
      "num_tokens": 59279830.0,
      "step": 17740
    },
    {
      "entropy": 6.1917469024658205,
      "epoch": 3.46180860403863,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038757308549047,
      "loss": 5.8714,
      "mean_token_accuracy": 0.1622008591890335,
      "num_tokens": 59297958.0,
      "step": 17745
    },
    {
      "entropy": 6.218881034851075,
      "epoch": 3.462784118622573,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038751218201466605,
      "loss": 5.7853,
      "mean_token_accuracy": 0.16518965512514114,
      "num_tokens": 59314119.0,
      "step": 17750
    },
    {
      "entropy": 6.200621175765991,
      "epoch": 3.4637596332065166,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003874512675445726,
      "loss": 5.8261,
      "mean_token_accuracy": 0.1666320502758026,
      "num_tokens": 59330391.0,
      "step": 17755
    },
    {
      "entropy": 6.240032434463501,
      "epoch": 3.4647351477904595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003873903420861422,
      "loss": 5.845,
      "mean_token_accuracy": 0.17020692229270934,
      "num_tokens": 59346449.0,
      "step": 17760
    },
    {
      "entropy": 6.2273268699646,
      "epoch": 3.4657106623744025,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038732940564532804,
      "loss": 5.9033,
      "mean_token_accuracy": 0.15746424943208695,
      "num_tokens": 59363077.0,
      "step": 17765
    },
    {
      "entropy": 6.1968484878540036,
      "epoch": 3.4666861769583455,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003872684582280847,
      "loss": 5.7167,
      "mean_token_accuracy": 0.1739965483546257,
      "num_tokens": 59381542.0,
      "step": 17770
    },
    {
      "entropy": 6.16641206741333,
      "epoch": 3.4676616915422884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003872074998403678,
      "loss": 5.7617,
      "mean_token_accuracy": 0.1688603088259697,
      "num_tokens": 59398045.0,
      "step": 17775
    },
    {
      "entropy": 6.208156156539917,
      "epoch": 3.468637206126232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038714653048813394,
      "loss": 5.9397,
      "mean_token_accuracy": 0.15496391355991362,
      "num_tokens": 59414958.0,
      "step": 17780
    },
    {
      "entropy": 6.306074762344361,
      "epoch": 3.469612720710175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038708555017734076,
      "loss": 5.8886,
      "mean_token_accuracy": 0.160356704890728,
      "num_tokens": 59431585.0,
      "step": 17785
    },
    {
      "entropy": 6.238780689239502,
      "epoch": 3.4705882352941178,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038702455891394716,
      "loss": 5.7379,
      "mean_token_accuracy": 0.1625151127576828,
      "num_tokens": 59447724.0,
      "step": 17790
    },
    {
      "entropy": 6.047742128372192,
      "epoch": 3.4715637498780607,
      "grad_norm": 1.0,
      "learning_rate": 0.00038696355670391294,
      "loss": 5.6413,
      "mean_token_accuracy": 0.18405098766088485,
      "num_tokens": 59465116.0,
      "step": 17795
    },
    {
      "entropy": 6.228205347061158,
      "epoch": 3.4725392644620037,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038690254355319893,
      "loss": 5.8661,
      "mean_token_accuracy": 0.15875468403100967,
      "num_tokens": 59481541.0,
      "step": 17800
    },
    {
      "entropy": 6.230705404281617,
      "epoch": 3.4735147790459466,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038684151946776727,
      "loss": 5.8081,
      "mean_token_accuracy": 0.16494135409593583,
      "num_tokens": 59498300.0,
      "step": 17805
    },
    {
      "entropy": 6.217336368560791,
      "epoch": 3.4744902936298896,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003867804844535809,
      "loss": 5.8376,
      "mean_token_accuracy": 0.16019289046525956,
      "num_tokens": 59514844.0,
      "step": 17810
    },
    {
      "entropy": 6.163929224014282,
      "epoch": 3.4754658082138326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003867194385166039,
      "loss": 5.7648,
      "mean_token_accuracy": 0.16796428263187407,
      "num_tokens": 59531820.0,
      "step": 17815
    },
    {
      "entropy": 6.250615882873535,
      "epoch": 3.476441322797776,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038665838166280163,
      "loss": 5.7937,
      "mean_token_accuracy": 0.16144197285175324,
      "num_tokens": 59549744.0,
      "step": 17820
    },
    {
      "entropy": 6.294581413269043,
      "epoch": 3.477416837381719,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003865973138981403,
      "loss": 5.8925,
      "mean_token_accuracy": 0.15827423483133315,
      "num_tokens": 59567598.0,
      "step": 17825
    },
    {
      "entropy": 6.197109222412109,
      "epoch": 3.478392351965662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038653623522858714,
      "loss": 5.8124,
      "mean_token_accuracy": 0.16211321353912353,
      "num_tokens": 59583332.0,
      "step": 17830
    },
    {
      "entropy": 6.220383405685425,
      "epoch": 3.479367866549605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003864751456601106,
      "loss": 5.8284,
      "mean_token_accuracy": 0.16350522190332412,
      "num_tokens": 59600056.0,
      "step": 17835
    },
    {
      "entropy": 6.2544337749481205,
      "epoch": 3.480343381133548,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038641404519868016,
      "loss": 5.8719,
      "mean_token_accuracy": 0.16131356805562974,
      "num_tokens": 59616065.0,
      "step": 17840
    },
    {
      "entropy": 6.181321525573731,
      "epoch": 3.481318895717491,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003863529338502663,
      "loss": 5.7929,
      "mean_token_accuracy": 0.16911821514368058,
      "num_tokens": 59633533.0,
      "step": 17845
    },
    {
      "entropy": 6.2149206638336185,
      "epoch": 3.482294410301434,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038629181162084055,
      "loss": 5.7566,
      "mean_token_accuracy": 0.17265586256980897,
      "num_tokens": 59649011.0,
      "step": 17850
    },
    {
      "entropy": 6.206947898864746,
      "epoch": 3.483269924885377,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038623067851637573,
      "loss": 5.7878,
      "mean_token_accuracy": 0.15875309109687805,
      "num_tokens": 59664638.0,
      "step": 17855
    },
    {
      "entropy": 6.310023832321167,
      "epoch": 3.48424543946932,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003861695345428453,
      "loss": 5.9282,
      "mean_token_accuracy": 0.1521035373210907,
      "num_tokens": 59680173.0,
      "step": 17860
    },
    {
      "entropy": 6.292608451843262,
      "epoch": 3.485220954053263,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038610837970622433,
      "loss": 5.921,
      "mean_token_accuracy": 0.15696401596069337,
      "num_tokens": 59696156.0,
      "step": 17865
    },
    {
      "entropy": 6.293128919601441,
      "epoch": 3.486196468637206,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003860472140124884,
      "loss": 5.8547,
      "mean_token_accuracy": 0.16548086404800416,
      "num_tokens": 59712449.0,
      "step": 17870
    },
    {
      "entropy": 6.080825424194336,
      "epoch": 3.487171983221149,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038598603746761446,
      "loss": 5.7401,
      "mean_token_accuracy": 0.17039635181427001,
      "num_tokens": 59730896.0,
      "step": 17875
    },
    {
      "entropy": 6.189550447463989,
      "epoch": 3.4881474978050924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003859248500775805,
      "loss": 5.7578,
      "mean_token_accuracy": 0.16403668224811555,
      "num_tokens": 59747040.0,
      "step": 17880
    },
    {
      "entropy": 6.3021697998046875,
      "epoch": 3.4891230123890353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038586365184836556,
      "loss": 5.86,
      "mean_token_accuracy": 0.15835288763046265,
      "num_tokens": 59763931.0,
      "step": 17885
    },
    {
      "entropy": 6.178811359405517,
      "epoch": 3.4900985269729783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038580244278594976,
      "loss": 5.7462,
      "mean_token_accuracy": 0.16282061636447906,
      "num_tokens": 59779581.0,
      "step": 17890
    },
    {
      "entropy": 6.248637342453003,
      "epoch": 3.4910740415569212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003857412228963141,
      "loss": 5.858,
      "mean_token_accuracy": 0.16202323138713837,
      "num_tokens": 59796388.0,
      "step": 17895
    },
    {
      "entropy": 6.128023242950439,
      "epoch": 3.492049556140864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003856799921854408,
      "loss": 5.7532,
      "mean_token_accuracy": 0.17261091321706773,
      "num_tokens": 59814398.0,
      "step": 17900
    },
    {
      "entropy": 6.154485034942627,
      "epoch": 3.493025070724807,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003856187506593132,
      "loss": 5.81,
      "mean_token_accuracy": 0.1664289876818657,
      "num_tokens": 59831772.0,
      "step": 17905
    },
    {
      "entropy": 5.910578775405884,
      "epoch": 3.4940005853087506,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003855574983239155,
      "loss": 5.4808,
      "mean_token_accuracy": 0.20043767243623734,
      "num_tokens": 59849621.0,
      "step": 17910
    },
    {
      "entropy": 6.205603456497192,
      "epoch": 3.4949760998926935,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003854962351852331,
      "loss": 5.7672,
      "mean_token_accuracy": 0.16468181014060973,
      "num_tokens": 59867231.0,
      "step": 17915
    },
    {
      "entropy": 6.175866413116455,
      "epoch": 3.4959516144766365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038543496124925243,
      "loss": 5.7764,
      "mean_token_accuracy": 0.1679822400212288,
      "num_tokens": 59884040.0,
      "step": 17920
    },
    {
      "entropy": 6.177528285980225,
      "epoch": 3.4969271290605795,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003853736765219609,
      "loss": 5.6955,
      "mean_token_accuracy": 0.17502517402172088,
      "num_tokens": 59900593.0,
      "step": 17925
    },
    {
      "entropy": 6.299563455581665,
      "epoch": 3.4979026436445224,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038531238100934713,
      "loss": 5.8763,
      "mean_token_accuracy": 0.15755697786808015,
      "num_tokens": 59916389.0,
      "step": 17930
    },
    {
      "entropy": 6.143807744979858,
      "epoch": 3.4988781582284654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038525107471740064,
      "loss": 5.815,
      "mean_token_accuracy": 0.1638144537806511,
      "num_tokens": 59933536.0,
      "step": 17935
    },
    {
      "entropy": 6.133168697357178,
      "epoch": 3.4998536728124083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003851897576521121,
      "loss": 5.7185,
      "mean_token_accuracy": 0.17354726195335388,
      "num_tokens": 59949756.0,
      "step": 17940
    },
    {
      "entropy": 6.238433504104615,
      "epoch": 3.5008291873963517,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038512842981947297,
      "loss": 5.7629,
      "mean_token_accuracy": 0.16270435154438018,
      "num_tokens": 59965791.0,
      "step": 17945
    },
    {
      "entropy": 6.11470513343811,
      "epoch": 3.5018047019802947,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038506709122547645,
      "loss": 5.6903,
      "mean_token_accuracy": 0.18155694156885147,
      "num_tokens": 59981471.0,
      "step": 17950
    },
    {
      "entropy": 6.224485921859741,
      "epoch": 3.5027802165642377,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038500574187611594,
      "loss": 5.7639,
      "mean_token_accuracy": 0.16050916016101838,
      "num_tokens": 59996896.0,
      "step": 17955
    },
    {
      "entropy": 6.187601804733276,
      "epoch": 3.5037557311481806,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003849443817773863,
      "loss": 5.8312,
      "mean_token_accuracy": 0.1640125721693039,
      "num_tokens": 60013110.0,
      "step": 17960
    },
    {
      "entropy": 6.179777145385742,
      "epoch": 3.5047312457321236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003848830109352837,
      "loss": 5.766,
      "mean_token_accuracy": 0.168218170106411,
      "num_tokens": 60029763.0,
      "step": 17965
    },
    {
      "entropy": 6.1938494682312015,
      "epoch": 3.505706760316067,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003848216293558048,
      "loss": 5.7535,
      "mean_token_accuracy": 0.16942343562841417,
      "num_tokens": 60046078.0,
      "step": 17970
    },
    {
      "entropy": 6.264446687698364,
      "epoch": 3.50668227490001,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038476023704494774,
      "loss": 5.885,
      "mean_token_accuracy": 0.15428589135408402,
      "num_tokens": 60062469.0,
      "step": 17975
    },
    {
      "entropy": 6.078772830963135,
      "epoch": 3.507657789483953,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003846988340087114,
      "loss": 5.6687,
      "mean_token_accuracy": 0.17712630927562714,
      "num_tokens": 60078183.0,
      "step": 17980
    },
    {
      "entropy": 6.0640161514282225,
      "epoch": 3.508633304067896,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003846374202530961,
      "loss": 5.6679,
      "mean_token_accuracy": 0.1763103649020195,
      "num_tokens": 60094811.0,
      "step": 17985
    },
    {
      "entropy": 6.22886061668396,
      "epoch": 3.509608818651839,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038457599578410286,
      "loss": 5.7719,
      "mean_token_accuracy": 0.16446950137615204,
      "num_tokens": 60112478.0,
      "step": 17990
    },
    {
      "entropy": 6.200900363922119,
      "epoch": 3.510584333235782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038451456060773376,
      "loss": 5.7832,
      "mean_token_accuracy": 0.16458000093698502,
      "num_tokens": 60128486.0,
      "step": 17995
    },
    {
      "entropy": 6.160733222961426,
      "epoch": 3.5115598478197247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003844531147299921,
      "loss": 5.7591,
      "mean_token_accuracy": 0.16959148198366164,
      "num_tokens": 60144303.0,
      "step": 18000
    },
    {
      "epoch": 3.5115598478197247,
      "eval_entropy": 5.987648142056941,
      "eval_loss": 5.96008825302124,
      "eval_mean_token_accuracy": 0.1644608318065309,
      "eval_num_tokens": 60144303.0,
      "eval_runtime": 15.2211,
      "eval_samples_per_second": 1089.079,
      "eval_steps_per_second": 136.192,
      "step": 18000
    },
    {
      "entropy": 6.020863580703735,
      "epoch": 3.5125353624036677,
      "grad_norm": 0.84375,
      "learning_rate": 0.00038439165815688233,
      "loss": 5.6735,
      "mean_token_accuracy": 0.17593095302581788,
      "num_tokens": 60161750.0,
      "step": 18005
    },
    {
      "entropy": 6.240034914016723,
      "epoch": 3.513510876987611,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003843301908944094,
      "loss": 5.83,
      "mean_token_accuracy": 0.16284202486276628,
      "num_tokens": 60178144.0,
      "step": 18010
    },
    {
      "entropy": 6.0533167839050295,
      "epoch": 3.514486391571554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038426871294858,
      "loss": 5.6062,
      "mean_token_accuracy": 0.18057896494865416,
      "num_tokens": 60196152.0,
      "step": 18015
    },
    {
      "entropy": 6.135756969451904,
      "epoch": 3.515461906155497,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038420722432540137,
      "loss": 5.7218,
      "mean_token_accuracy": 0.16804761588573455,
      "num_tokens": 60212049.0,
      "step": 18020
    },
    {
      "entropy": 6.154497671127319,
      "epoch": 3.51643742073944,
      "grad_norm": 0.859375,
      "learning_rate": 0.000384145725030882,
      "loss": 5.7684,
      "mean_token_accuracy": 0.1672029748558998,
      "num_tokens": 60230028.0,
      "step": 18025
    },
    {
      "entropy": 6.275807619094849,
      "epoch": 3.517412935323383,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038408421507103144,
      "loss": 5.9036,
      "mean_token_accuracy": 0.15860126912593842,
      "num_tokens": 60247265.0,
      "step": 18030
    },
    {
      "entropy": 6.171633243560791,
      "epoch": 3.5183884499073264,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003840226944518601,
      "loss": 5.7329,
      "mean_token_accuracy": 0.16708520501852037,
      "num_tokens": 60263685.0,
      "step": 18035
    },
    {
      "entropy": 6.209921360015869,
      "epoch": 3.5193639644912693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003839611631793796,
      "loss": 5.8551,
      "mean_token_accuracy": 0.16212470829486847,
      "num_tokens": 60279574.0,
      "step": 18040
    },
    {
      "entropy": 6.187422370910644,
      "epoch": 3.5203394790752123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003838996212596027,
      "loss": 5.836,
      "mean_token_accuracy": 0.1651304304599762,
      "num_tokens": 60296302.0,
      "step": 18045
    },
    {
      "entropy": 6.219190692901611,
      "epoch": 3.5213149936591552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003838380686985428,
      "loss": 5.8448,
      "mean_token_accuracy": 0.16849314868450166,
      "num_tokens": 60313595.0,
      "step": 18050
    },
    {
      "entropy": 6.069160652160645,
      "epoch": 3.522290508243098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038377650550221475,
      "loss": 5.6583,
      "mean_token_accuracy": 0.17737076804041862,
      "num_tokens": 60331286.0,
      "step": 18055
    },
    {
      "entropy": 6.188697481155396,
      "epoch": 3.523266022827041,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038371493167663434,
      "loss": 5.787,
      "mean_token_accuracy": 0.16704670786857606,
      "num_tokens": 60347704.0,
      "step": 18060
    },
    {
      "entropy": 6.212107706069946,
      "epoch": 3.524241537410984,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003836533472278183,
      "loss": 5.8377,
      "mean_token_accuracy": 0.16251338571310042,
      "num_tokens": 60364702.0,
      "step": 18065
    },
    {
      "entropy": 6.205227661132812,
      "epoch": 3.525217051994927,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038359175216178427,
      "loss": 5.7934,
      "mean_token_accuracy": 0.15955734103918076,
      "num_tokens": 60381052.0,
      "step": 18070
    },
    {
      "entropy": 6.182648754119873,
      "epoch": 3.5261925665788705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038353014648455136,
      "loss": 5.8382,
      "mean_token_accuracy": 0.16116272062063217,
      "num_tokens": 60396772.0,
      "step": 18075
    },
    {
      "entropy": 6.1336136817932125,
      "epoch": 3.5271680811628134,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003834685302021393,
      "loss": 5.7303,
      "mean_token_accuracy": 0.16955242156982422,
      "num_tokens": 60412951.0,
      "step": 18080
    },
    {
      "entropy": 6.1568011283874515,
      "epoch": 3.5281435957467564,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038340690332056904,
      "loss": 5.7389,
      "mean_token_accuracy": 0.1700511872768402,
      "num_tokens": 60429560.0,
      "step": 18085
    },
    {
      "entropy": 6.259678697586059,
      "epoch": 3.5291191103306994,
      "grad_norm": 1.0,
      "learning_rate": 0.0003833452658458626,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16623667478561402,
      "num_tokens": 60444962.0,
      "step": 18090
    },
    {
      "entropy": 6.095765447616577,
      "epoch": 3.5300946249146423,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038328361778404294,
      "loss": 5.7042,
      "mean_token_accuracy": 0.1719483755528927,
      "num_tokens": 60462796.0,
      "step": 18095
    },
    {
      "entropy": 6.216716289520264,
      "epoch": 3.5310701394985857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003832219591411341,
      "loss": 5.7849,
      "mean_token_accuracy": 0.16315912157297136,
      "num_tokens": 60478639.0,
      "step": 18100
    },
    {
      "entropy": 6.056381559371948,
      "epoch": 3.5320456540825287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038316028992316106,
      "loss": 5.6393,
      "mean_token_accuracy": 0.1835736647248268,
      "num_tokens": 60493114.0,
      "step": 18105
    },
    {
      "entropy": 6.187199974060059,
      "epoch": 3.5330211686664716,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038309861013614995,
      "loss": 5.7388,
      "mean_token_accuracy": 0.17634321004152298,
      "num_tokens": 60507726.0,
      "step": 18110
    },
    {
      "entropy": 6.242354154586792,
      "epoch": 3.5339966832504146,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038303691978612797,
      "loss": 5.8828,
      "mean_token_accuracy": 0.15838723182678222,
      "num_tokens": 60525835.0,
      "step": 18115
    },
    {
      "entropy": 6.1323404788970945,
      "epoch": 3.5349721978343576,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003829752188791233,
      "loss": 5.7454,
      "mean_token_accuracy": 0.17011684477329253,
      "num_tokens": 60543333.0,
      "step": 18120
    },
    {
      "entropy": 6.103044700622559,
      "epoch": 3.5359477124183005,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000382913507421165,
      "loss": 5.6996,
      "mean_token_accuracy": 0.1681746944785118,
      "num_tokens": 60559369.0,
      "step": 18125
    },
    {
      "entropy": 5.9992677688598635,
      "epoch": 3.5369232270022435,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003828517854182834,
      "loss": 5.5482,
      "mean_token_accuracy": 0.19012025892734527,
      "num_tokens": 60574575.0,
      "step": 18130
    },
    {
      "entropy": 6.131134414672852,
      "epoch": 3.5378987415861864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003827900528765097,
      "loss": 5.7371,
      "mean_token_accuracy": 0.16964758783578873,
      "num_tokens": 60591185.0,
      "step": 18135
    },
    {
      "entropy": 6.133888292312622,
      "epoch": 3.53887425617013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003827283098018762,
      "loss": 5.7683,
      "mean_token_accuracy": 0.16518858224153518,
      "num_tokens": 60607942.0,
      "step": 18140
    },
    {
      "entropy": 6.184584760665894,
      "epoch": 3.539849770754073,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003826665562004162,
      "loss": 5.771,
      "mean_token_accuracy": 0.16982644200325012,
      "num_tokens": 60625127.0,
      "step": 18145
    },
    {
      "entropy": 6.288888692855835,
      "epoch": 3.5408252853380158,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003826047920781641,
      "loss": 5.8659,
      "mean_token_accuracy": 0.15970002859830856,
      "num_tokens": 60641980.0,
      "step": 18150
    },
    {
      "entropy": 6.117878866195679,
      "epoch": 3.5418007999219587,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003825430174411552,
      "loss": 5.7499,
      "mean_token_accuracy": 0.16974127143621445,
      "num_tokens": 60659171.0,
      "step": 18155
    },
    {
      "entropy": 6.210864305496216,
      "epoch": 3.542776314505902,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003824812322954259,
      "loss": 5.8427,
      "mean_token_accuracy": 0.1656330019235611,
      "num_tokens": 60675320.0,
      "step": 18160
    },
    {
      "entropy": 6.272024488449096,
      "epoch": 3.543751829089845,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038241943664701374,
      "loss": 5.9566,
      "mean_token_accuracy": 0.15754733830690384,
      "num_tokens": 60692876.0,
      "step": 18165
    },
    {
      "entropy": 6.254979753494263,
      "epoch": 3.544727343673788,
      "grad_norm": 0.984375,
      "learning_rate": 0.000382357630501957,
      "loss": 5.869,
      "mean_token_accuracy": 0.1598404750227928,
      "num_tokens": 60708442.0,
      "step": 18170
    },
    {
      "entropy": 6.169257116317749,
      "epoch": 3.545702858257731,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003822958138662953,
      "loss": 5.7079,
      "mean_token_accuracy": 0.1678064540028572,
      "num_tokens": 60726103.0,
      "step": 18175
    },
    {
      "entropy": 6.138762664794922,
      "epoch": 3.546678372841674,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000382233986746069,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1748331680893898,
      "num_tokens": 60741990.0,
      "step": 18180
    },
    {
      "entropy": 6.20259199142456,
      "epoch": 3.547653887425617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003821721491473198,
      "loss": 5.8101,
      "mean_token_accuracy": 0.1683477982878685,
      "num_tokens": 60758877.0,
      "step": 18185
    },
    {
      "entropy": 6.195466327667236,
      "epoch": 3.54862940200956,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038211030107609,
      "loss": 5.773,
      "mean_token_accuracy": 0.16771250888705252,
      "num_tokens": 60774762.0,
      "step": 18190
    },
    {
      "entropy": 6.169298696517944,
      "epoch": 3.549604916593503,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038204844253842345,
      "loss": 5.8167,
      "mean_token_accuracy": 0.16711457520723344,
      "num_tokens": 60791196.0,
      "step": 18195
    },
    {
      "entropy": 6.1648579120635985,
      "epoch": 3.5505804311774463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038198657354036457,
      "loss": 5.8035,
      "mean_token_accuracy": 0.16542958766222,
      "num_tokens": 60807688.0,
      "step": 18200
    },
    {
      "entropy": 6.256539630889892,
      "epoch": 3.551555945761389,
      "grad_norm": 1.0,
      "learning_rate": 0.00038192469408795905,
      "loss": 5.809,
      "mean_token_accuracy": 0.16498955935239792,
      "num_tokens": 60823713.0,
      "step": 18205
    },
    {
      "entropy": 6.196027231216431,
      "epoch": 3.552531460345332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038186280418725353,
      "loss": 5.8489,
      "mean_token_accuracy": 0.1647099018096924,
      "num_tokens": 60840433.0,
      "step": 18210
    },
    {
      "entropy": 6.218382453918457,
      "epoch": 3.553506974929275,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038180090384429564,
      "loss": 5.8813,
      "mean_token_accuracy": 0.16241230368614196,
      "num_tokens": 60854953.0,
      "step": 18215
    },
    {
      "entropy": 6.249703788757325,
      "epoch": 3.554482489513218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038173899306513404,
      "loss": 5.8945,
      "mean_token_accuracy": 0.159484626352787,
      "num_tokens": 60871973.0,
      "step": 18220
    },
    {
      "entropy": 6.228696298599243,
      "epoch": 3.5554580040971615,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003816770718558185,
      "loss": 5.8025,
      "mean_token_accuracy": 0.16178472489118575,
      "num_tokens": 60888135.0,
      "step": 18225
    },
    {
      "entropy": 6.031142711639404,
      "epoch": 3.5564335186811045,
      "grad_norm": 0.875,
      "learning_rate": 0.00038161514022239974,
      "loss": 5.6177,
      "mean_token_accuracy": 0.1823108375072479,
      "num_tokens": 60905831.0,
      "step": 18230
    },
    {
      "entropy": 6.189695024490357,
      "epoch": 3.5574090332650474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003815531981709294,
      "loss": 5.7895,
      "mean_token_accuracy": 0.15632833838462828,
      "num_tokens": 60923670.0,
      "step": 18235
    },
    {
      "entropy": 6.145584678649902,
      "epoch": 3.5583845478489904,
      "grad_norm": 1.0,
      "learning_rate": 0.00038149124570746037,
      "loss": 5.7579,
      "mean_token_accuracy": 0.1658823534846306,
      "num_tokens": 60939007.0,
      "step": 18240
    },
    {
      "entropy": 6.190569067001343,
      "epoch": 3.5593600624329333,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003814292828380463,
      "loss": 5.7504,
      "mean_token_accuracy": 0.1620073437690735,
      "num_tokens": 60956421.0,
      "step": 18245
    },
    {
      "entropy": 5.999208974838257,
      "epoch": 3.5603355770168763,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000381367309568742,
      "loss": 5.5121,
      "mean_token_accuracy": 0.1915200561285019,
      "num_tokens": 60975047.0,
      "step": 18250
    },
    {
      "entropy": 6.139942169189453,
      "epoch": 3.5613110916008193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003813053259056033,
      "loss": 5.7055,
      "mean_token_accuracy": 0.17430460304021836,
      "num_tokens": 60991482.0,
      "step": 18255
    },
    {
      "entropy": 6.095325326919555,
      "epoch": 3.5622866061847622,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038124333185468714,
      "loss": 5.6853,
      "mean_token_accuracy": 0.1752055913209915,
      "num_tokens": 61009162.0,
      "step": 18260
    },
    {
      "entropy": 6.181165552139282,
      "epoch": 3.5632621207687056,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003811813274220511,
      "loss": 5.7604,
      "mean_token_accuracy": 0.1674210399389267,
      "num_tokens": 61026654.0,
      "step": 18265
    },
    {
      "entropy": 6.164965200424194,
      "epoch": 3.5642376353526486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038111931261375424,
      "loss": 5.7754,
      "mean_token_accuracy": 0.16887067556381224,
      "num_tokens": 61042808.0,
      "step": 18270
    },
    {
      "entropy": 6.23197455406189,
      "epoch": 3.5652131499365916,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038105728743585644,
      "loss": 5.8387,
      "mean_token_accuracy": 0.15244799107313156,
      "num_tokens": 61059349.0,
      "step": 18275
    },
    {
      "entropy": 6.240943908691406,
      "epoch": 3.5661886645205345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003809952518944185,
      "loss": 5.8756,
      "mean_token_accuracy": 0.16527860313653947,
      "num_tokens": 61077830.0,
      "step": 18280
    },
    {
      "entropy": 6.147238731384277,
      "epoch": 3.5671641791044775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003809332059955022,
      "loss": 5.7459,
      "mean_token_accuracy": 0.1754624977707863,
      "num_tokens": 61094812.0,
      "step": 18285
    },
    {
      "entropy": 6.134187126159668,
      "epoch": 3.568139693688421,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003808711497451706,
      "loss": 5.7484,
      "mean_token_accuracy": 0.17744943201541902,
      "num_tokens": 61113756.0,
      "step": 18290
    },
    {
      "entropy": 6.1909407615661625,
      "epoch": 3.569115208272364,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038080908314948764,
      "loss": 5.7634,
      "mean_token_accuracy": 0.17293113768100737,
      "num_tokens": 61128395.0,
      "step": 18295
    },
    {
      "entropy": 6.05263237953186,
      "epoch": 3.570090722856307,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003807470062145181,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17588766366243364,
      "num_tokens": 61144823.0,
      "step": 18300
    },
    {
      "entropy": 6.193591690063476,
      "epoch": 3.5710662374402498,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038068491894632805,
      "loss": 5.8165,
      "mean_token_accuracy": 0.1638689324259758,
      "num_tokens": 61161693.0,
      "step": 18305
    },
    {
      "entropy": 6.246031284332275,
      "epoch": 3.5720417520241927,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003806228213509844,
      "loss": 5.9098,
      "mean_token_accuracy": 0.15628963708877563,
      "num_tokens": 61178807.0,
      "step": 18310
    },
    {
      "entropy": 6.093202447891235,
      "epoch": 3.5730172666081357,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038056071343455514,
      "loss": 5.6248,
      "mean_token_accuracy": 0.17875364422798157,
      "num_tokens": 61195154.0,
      "step": 18315
    },
    {
      "entropy": 6.197946834564209,
      "epoch": 3.5739927811920786,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003804985952031091,
      "loss": 5.8494,
      "mean_token_accuracy": 0.16149336248636245,
      "num_tokens": 61211825.0,
      "step": 18320
    },
    {
      "entropy": 6.252932453155518,
      "epoch": 3.5749682957760216,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003804364666627164,
      "loss": 5.8795,
      "mean_token_accuracy": 0.16075246334075927,
      "num_tokens": 61228548.0,
      "step": 18325
    },
    {
      "entropy": 6.289742565155029,
      "epoch": 3.575943810359965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038037432781944797,
      "loss": 5.8745,
      "mean_token_accuracy": 0.15801834613084792,
      "num_tokens": 61245490.0,
      "step": 18330
    },
    {
      "entropy": 6.124907684326172,
      "epoch": 3.576919324943908,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038031217867937566,
      "loss": 5.6591,
      "mean_token_accuracy": 0.17501956224441528,
      "num_tokens": 61261604.0,
      "step": 18335
    },
    {
      "entropy": 6.148367595672608,
      "epoch": 3.577894839527851,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038025001924857273,
      "loss": 5.7435,
      "mean_token_accuracy": 0.17596617639064788,
      "num_tokens": 61279100.0,
      "step": 18340
    },
    {
      "entropy": 6.205999088287354,
      "epoch": 3.578870354111794,
      "grad_norm": 1.015625,
      "learning_rate": 0.000380187849533113,
      "loss": 5.7557,
      "mean_token_accuracy": 0.1677320420742035,
      "num_tokens": 61295184.0,
      "step": 18345
    },
    {
      "entropy": 6.1625205993652346,
      "epoch": 3.579845868695737,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038012566953907147,
      "loss": 5.7651,
      "mean_token_accuracy": 0.17378849387168885,
      "num_tokens": 61313070.0,
      "step": 18350
    },
    {
      "entropy": 6.24502215385437,
      "epoch": 3.5808213832796802,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003800634792725242,
      "loss": 5.8499,
      "mean_token_accuracy": 0.1648797184228897,
      "num_tokens": 61330199.0,
      "step": 18355
    },
    {
      "entropy": 6.292352819442749,
      "epoch": 3.581796897863623,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038000127873954824,
      "loss": 5.8913,
      "mean_token_accuracy": 0.15497296452522277,
      "num_tokens": 61345525.0,
      "step": 18360
    },
    {
      "entropy": 6.257273530960083,
      "epoch": 3.582772412447566,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037993906794622144,
      "loss": 5.8791,
      "mean_token_accuracy": 0.16260384172201156,
      "num_tokens": 61363064.0,
      "step": 18365
    },
    {
      "entropy": 6.2132251262664795,
      "epoch": 3.583747927031509,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037987684689862314,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1631417378783226,
      "num_tokens": 61381092.0,
      "step": 18370
    },
    {
      "entropy": 6.176194524765014,
      "epoch": 3.584723441615452,
      "grad_norm": 1.0,
      "learning_rate": 0.000379814615602833,
      "loss": 5.7757,
      "mean_token_accuracy": 0.16854993253946304,
      "num_tokens": 61397505.0,
      "step": 18375
    },
    {
      "entropy": 6.184675979614258,
      "epoch": 3.585698956199395,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037975237406493215,
      "loss": 5.7258,
      "mean_token_accuracy": 0.16846155375242233,
      "num_tokens": 61414986.0,
      "step": 18380
    },
    {
      "entropy": 5.988455057144165,
      "epoch": 3.586674470783338,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037969012229100276,
      "loss": 5.5448,
      "mean_token_accuracy": 0.19250905215740205,
      "num_tokens": 61431878.0,
      "step": 18385
    },
    {
      "entropy": 6.222172784805298,
      "epoch": 3.587649985367281,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037962786028712767,
      "loss": 5.8003,
      "mean_token_accuracy": 0.16030510067939757,
      "num_tokens": 61447872.0,
      "step": 18390
    },
    {
      "entropy": 6.1583373069763185,
      "epoch": 3.5886254999512244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037956558805939095,
      "loss": 5.739,
      "mean_token_accuracy": 0.17175170928239822,
      "num_tokens": 61465005.0,
      "step": 18395
    },
    {
      "entropy": 6.090008783340454,
      "epoch": 3.5896010145351673,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003795033056138777,
      "loss": 5.7105,
      "mean_token_accuracy": 0.16649671047925949,
      "num_tokens": 61481482.0,
      "step": 18400
    },
    {
      "entropy": 6.2067742347717285,
      "epoch": 3.5905765291191103,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003794410129566738,
      "loss": 5.8122,
      "mean_token_accuracy": 0.16084552109241484,
      "num_tokens": 61496611.0,
      "step": 18405
    },
    {
      "entropy": 6.187848377227783,
      "epoch": 3.5915520437030533,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003793787100938665,
      "loss": 5.7353,
      "mean_token_accuracy": 0.17416218519210816,
      "num_tokens": 61515309.0,
      "step": 18410
    },
    {
      "entropy": 6.126193618774414,
      "epoch": 3.5925275582869967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003793163970315435,
      "loss": 5.7188,
      "mean_token_accuracy": 0.17330708652734755,
      "num_tokens": 61531375.0,
      "step": 18415
    },
    {
      "entropy": 6.277548599243164,
      "epoch": 3.5935030728709396,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037925407377579405,
      "loss": 5.8299,
      "mean_token_accuracy": 0.16164597868919373,
      "num_tokens": 61548626.0,
      "step": 18420
    },
    {
      "entropy": 6.138858890533447,
      "epoch": 3.5944785874548826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003791917403327081,
      "loss": 5.7453,
      "mean_token_accuracy": 0.172882279753685,
      "num_tokens": 61566514.0,
      "step": 18425
    },
    {
      "entropy": 6.1951087474822994,
      "epoch": 3.5954541020388255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003791293967083766,
      "loss": 5.7961,
      "mean_token_accuracy": 0.16031832247972488,
      "num_tokens": 61582558.0,
      "step": 18430
    },
    {
      "entropy": 6.2448395729064945,
      "epoch": 3.5964296166227685,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003790670429088915,
      "loss": 5.8569,
      "mean_token_accuracy": 0.16610682606697083,
      "num_tokens": 61599045.0,
      "step": 18435
    },
    {
      "entropy": 6.05029525756836,
      "epoch": 3.5974051312067115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037900467894034606,
      "loss": 5.6501,
      "mean_token_accuracy": 0.17141792327165603,
      "num_tokens": 61613661.0,
      "step": 18440
    },
    {
      "entropy": 6.175707244873047,
      "epoch": 3.5983806457906544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003789423048088339,
      "loss": 5.7349,
      "mean_token_accuracy": 0.1690747395157814,
      "num_tokens": 61628579.0,
      "step": 18445
    },
    {
      "entropy": 6.0825389385223385,
      "epoch": 3.5993561603745974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003788799205204502,
      "loss": 5.6694,
      "mean_token_accuracy": 0.17572338730096818,
      "num_tokens": 61644334.0,
      "step": 18450
    },
    {
      "entropy": 6.233411455154419,
      "epoch": 3.600331674958541,
      "grad_norm": 1.046875,
      "learning_rate": 0.000378817526081291,
      "loss": 5.7854,
      "mean_token_accuracy": 0.1580802544951439,
      "num_tokens": 61661279.0,
      "step": 18455
    },
    {
      "entropy": 6.215374517440796,
      "epoch": 3.6013071895424837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003787551214974531,
      "loss": 5.7473,
      "mean_token_accuracy": 0.16390259861946105,
      "num_tokens": 61677745.0,
      "step": 18460
    },
    {
      "entropy": 6.182989454269409,
      "epoch": 3.6022827041264267,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003786927067750345,
      "loss": 5.7989,
      "mean_token_accuracy": 0.17163249105215073,
      "num_tokens": 61693875.0,
      "step": 18465
    },
    {
      "entropy": 6.2072227001190186,
      "epoch": 3.6032582187103697,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00037863028192013424,
      "loss": 5.8551,
      "mean_token_accuracy": 0.16420358568429946,
      "num_tokens": 61711047.0,
      "step": 18470
    },
    {
      "entropy": 5.998494100570679,
      "epoch": 3.6042337332943126,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003785678469388521,
      "loss": 5.5776,
      "mean_token_accuracy": 0.1853575512766838,
      "num_tokens": 61728260.0,
      "step": 18475
    },
    {
      "entropy": 6.2506004810333256,
      "epoch": 3.605209247878256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003785054018372891,
      "loss": 5.7994,
      "mean_token_accuracy": 0.16152718663215637,
      "num_tokens": 61744056.0,
      "step": 18480
    },
    {
      "entropy": 6.05032753944397,
      "epoch": 3.606184762462199,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003784429466215471,
      "loss": 5.6158,
      "mean_token_accuracy": 0.18810852766036987,
      "num_tokens": 61760692.0,
      "step": 18485
    },
    {
      "entropy": 6.012847518920898,
      "epoch": 3.607160277046142,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037838048129772904,
      "loss": 5.587,
      "mean_token_accuracy": 0.19589415639638902,
      "num_tokens": 61777537.0,
      "step": 18490
    },
    {
      "entropy": 6.192381095886231,
      "epoch": 3.608135791630085,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037831800587193883,
      "loss": 5.7793,
      "mean_token_accuracy": 0.16944967061281205,
      "num_tokens": 61792535.0,
      "step": 18495
    },
    {
      "entropy": 6.13362832069397,
      "epoch": 3.609111306214028,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037825552035028134,
      "loss": 5.7894,
      "mean_token_accuracy": 0.16626467257738115,
      "num_tokens": 61808979.0,
      "step": 18500
    },
    {
      "entropy": 6.27005820274353,
      "epoch": 3.610086820797971,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037819302473886236,
      "loss": 5.779,
      "mean_token_accuracy": 0.16801147013902665,
      "num_tokens": 61824524.0,
      "step": 18505
    },
    {
      "entropy": 6.172961902618408,
      "epoch": 3.611062335381914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003781305190437888,
      "loss": 5.7363,
      "mean_token_accuracy": 0.16825105547904967,
      "num_tokens": 61840196.0,
      "step": 18510
    },
    {
      "entropy": 6.252478313446045,
      "epoch": 3.6120378499658568,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003780680032711685,
      "loss": 5.9265,
      "mean_token_accuracy": 0.15643012821674346,
      "num_tokens": 61856708.0,
      "step": 18515
    },
    {
      "entropy": 6.201608324050904,
      "epoch": 3.6130133645498,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003780054774271102,
      "loss": 5.7678,
      "mean_token_accuracy": 0.16237518638372422,
      "num_tokens": 61873405.0,
      "step": 18520
    },
    {
      "entropy": 6.23112645149231,
      "epoch": 3.613988879133743,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037794294151772385,
      "loss": 5.8293,
      "mean_token_accuracy": 0.16696261167526244,
      "num_tokens": 61891066.0,
      "step": 18525
    },
    {
      "entropy": 6.245869922637939,
      "epoch": 3.614964393717686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037788039554912013,
      "loss": 5.8782,
      "mean_token_accuracy": 0.16304274201393126,
      "num_tokens": 61906858.0,
      "step": 18530
    },
    {
      "entropy": 6.206060838699341,
      "epoch": 3.615939908301629,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003778178395274108,
      "loss": 5.832,
      "mean_token_accuracy": 0.16026730388402938,
      "num_tokens": 61922871.0,
      "step": 18535
    },
    {
      "entropy": 6.1581337451934814,
      "epoch": 3.616915422885572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003777552734587087,
      "loss": 5.785,
      "mean_token_accuracy": 0.1699697867035866,
      "num_tokens": 61938041.0,
      "step": 18540
    },
    {
      "entropy": 6.215324735641479,
      "epoch": 3.6178909374695154,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037769269734912745,
      "loss": 5.8665,
      "mean_token_accuracy": 0.15949883311986923,
      "num_tokens": 61954027.0,
      "step": 18545
    },
    {
      "entropy": 6.185980939865113,
      "epoch": 3.6188664520534584,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003776301112047818,
      "loss": 5.7124,
      "mean_token_accuracy": 0.16956378966569902,
      "num_tokens": 61968666.0,
      "step": 18550
    },
    {
      "entropy": 6.161270427703857,
      "epoch": 3.6198419666374013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037756751503178755,
      "loss": 5.795,
      "mean_token_accuracy": 0.16490214318037033,
      "num_tokens": 61985320.0,
      "step": 18555
    },
    {
      "entropy": 6.128264856338501,
      "epoch": 3.6208174812213443,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003775049088362612,
      "loss": 5.7127,
      "mean_token_accuracy": 0.1727745994925499,
      "num_tokens": 62001905.0,
      "step": 18560
    },
    {
      "entropy": 6.196975135803223,
      "epoch": 3.6217929958052872,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003774422926243206,
      "loss": 5.756,
      "mean_token_accuracy": 0.17088237702846526,
      "num_tokens": 62017280.0,
      "step": 18565
    },
    {
      "entropy": 6.116398715972901,
      "epoch": 3.62276851038923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037737966640208413,
      "loss": 5.657,
      "mean_token_accuracy": 0.17125205397605897,
      "num_tokens": 62034919.0,
      "step": 18570
    },
    {
      "entropy": 6.1519773483276365,
      "epoch": 3.623744024973173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003773170301756716,
      "loss": 5.7461,
      "mean_token_accuracy": 0.16807809770107268,
      "num_tokens": 62051575.0,
      "step": 18575
    },
    {
      "entropy": 6.142057752609253,
      "epoch": 3.624719539557116,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037725438395120356,
      "loss": 5.7849,
      "mean_token_accuracy": 0.16925605535507202,
      "num_tokens": 62067684.0,
      "step": 18580
    },
    {
      "entropy": 6.172507858276367,
      "epoch": 3.6256950541410595,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003771917277348016,
      "loss": 5.7885,
      "mean_token_accuracy": 0.16813289672136306,
      "num_tokens": 62085030.0,
      "step": 18585
    },
    {
      "entropy": 6.267319822311402,
      "epoch": 3.6266705687250025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037712906153258813,
      "loss": 5.9026,
      "mean_token_accuracy": 0.16296535730361938,
      "num_tokens": 62101662.0,
      "step": 18590
    },
    {
      "entropy": 6.076683759689331,
      "epoch": 3.6276460833089454,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037706638535068686,
      "loss": 5.6886,
      "mean_token_accuracy": 0.17094110548496247,
      "num_tokens": 62118879.0,
      "step": 18595
    },
    {
      "entropy": 6.18033652305603,
      "epoch": 3.6286215978928884,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003770036991952221,
      "loss": 5.7819,
      "mean_token_accuracy": 0.17086105644702912,
      "num_tokens": 62134641.0,
      "step": 18600
    },
    {
      "entropy": 6.254193925857544,
      "epoch": 3.6295971124768314,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037694100307231947,
      "loss": 5.8747,
      "mean_token_accuracy": 0.16002056300640105,
      "num_tokens": 62151616.0,
      "step": 18605
    },
    {
      "entropy": 6.208031892776489,
      "epoch": 3.6305726270607748,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037687829698810526,
      "loss": 5.8134,
      "mean_token_accuracy": 0.170817331969738,
      "num_tokens": 62169307.0,
      "step": 18610
    },
    {
      "entropy": 6.333951044082641,
      "epoch": 3.6315481416447177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003768155809487071,
      "loss": 5.9049,
      "mean_token_accuracy": 0.16121673434972764,
      "num_tokens": 62185665.0,
      "step": 18615
    },
    {
      "entropy": 6.167591094970703,
      "epoch": 3.6325236562286607,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037675285496025315,
      "loss": 5.8097,
      "mean_token_accuracy": 0.16250770390033722,
      "num_tokens": 62199901.0,
      "step": 18620
    },
    {
      "entropy": 6.2418632984161375,
      "epoch": 3.6334991708126037,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003766901190288729,
      "loss": 5.8743,
      "mean_token_accuracy": 0.16013457328081132,
      "num_tokens": 62216813.0,
      "step": 18625
    },
    {
      "entropy": 6.128393697738647,
      "epoch": 3.6344746853965466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003766273731606966,
      "loss": 5.6759,
      "mean_token_accuracy": 0.1770881488919258,
      "num_tokens": 62232578.0,
      "step": 18630
    },
    {
      "entropy": 6.176776981353759,
      "epoch": 3.6354501999804896,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003765646173618557,
      "loss": 5.8042,
      "mean_token_accuracy": 0.16275352984666824,
      "num_tokens": 62248652.0,
      "step": 18635
    },
    {
      "entropy": 6.127478551864624,
      "epoch": 3.6364257145644325,
      "grad_norm": 0.875,
      "learning_rate": 0.0003765018516384823,
      "loss": 5.6771,
      "mean_token_accuracy": 0.1751556098461151,
      "num_tokens": 62265170.0,
      "step": 18640
    },
    {
      "entropy": 6.053696966171264,
      "epoch": 3.6374012291483755,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037643907599670974,
      "loss": 5.6109,
      "mean_token_accuracy": 0.1886995777487755,
      "num_tokens": 62282809.0,
      "step": 18645
    },
    {
      "entropy": 6.249782276153565,
      "epoch": 3.638376743732319,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003763762904426722,
      "loss": 5.8873,
      "mean_token_accuracy": 0.1555943548679352,
      "num_tokens": 62301036.0,
      "step": 18650
    },
    {
      "entropy": 6.187583303451538,
      "epoch": 3.639352258316262,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003763134949825048,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16908892542123793,
      "num_tokens": 62317434.0,
      "step": 18655
    },
    {
      "entropy": 6.165703773498535,
      "epoch": 3.640327772900205,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003762506896223439,
      "loss": 5.7486,
      "mean_token_accuracy": 0.16442972868680955,
      "num_tokens": 62333158.0,
      "step": 18660
    },
    {
      "entropy": 6.1969263553619385,
      "epoch": 3.6413032874841478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003761878743683264,
      "loss": 5.8203,
      "mean_token_accuracy": 0.16449378281831742,
      "num_tokens": 62349075.0,
      "step": 18665
    },
    {
      "entropy": 6.139987421035767,
      "epoch": 3.642278802068091,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003761250492265904,
      "loss": 5.6588,
      "mean_token_accuracy": 0.176103076338768,
      "num_tokens": 62366322.0,
      "step": 18670
    },
    {
      "entropy": 6.230803346633911,
      "epoch": 3.643254316652034,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037606221420327505,
      "loss": 5.8088,
      "mean_token_accuracy": 0.16503652781248093,
      "num_tokens": 62382140.0,
      "step": 18675
    },
    {
      "entropy": 6.125056791305542,
      "epoch": 3.644229831235977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003759993693045204,
      "loss": 5.7935,
      "mean_token_accuracy": 0.16821264028549193,
      "num_tokens": 62399255.0,
      "step": 18680
    },
    {
      "entropy": 6.148307943344117,
      "epoch": 3.64520534581992,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003759365145364673,
      "loss": 5.6977,
      "mean_token_accuracy": 0.17661461383104324,
      "num_tokens": 62415451.0,
      "step": 18685
    },
    {
      "entropy": 6.121060371398926,
      "epoch": 3.646180860403863,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003758736499052577,
      "loss": 5.6975,
      "mean_token_accuracy": 0.17350801527500154,
      "num_tokens": 62432486.0,
      "step": 18690
    },
    {
      "entropy": 6.170613861083984,
      "epoch": 3.647156374987806,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003758107754170346,
      "loss": 5.7613,
      "mean_token_accuracy": 0.1751369282603264,
      "num_tokens": 62450117.0,
      "step": 18695
    },
    {
      "entropy": 6.143352127075195,
      "epoch": 3.648131889571749,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037574789107794175,
      "loss": 5.7397,
      "mean_token_accuracy": 0.16379004567861558,
      "num_tokens": 62465628.0,
      "step": 18700
    },
    {
      "entropy": 6.1681372165679935,
      "epoch": 3.649107404155692,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003756849968941241,
      "loss": 5.7377,
      "mean_token_accuracy": 0.1781106173992157,
      "num_tokens": 62483386.0,
      "step": 18705
    },
    {
      "entropy": 6.076944732666016,
      "epoch": 3.6500829187396353,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003756220928717274,
      "loss": 5.6775,
      "mean_token_accuracy": 0.17739930897951126,
      "num_tokens": 62499618.0,
      "step": 18710
    },
    {
      "entropy": 6.225219249725342,
      "epoch": 3.6510584333235783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003755591790168984,
      "loss": 5.7867,
      "mean_token_accuracy": 0.16239728033542633,
      "num_tokens": 62516226.0,
      "step": 18715
    },
    {
      "entropy": 6.231039428710938,
      "epoch": 3.6520339479075212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003754962553357849,
      "loss": 5.797,
      "mean_token_accuracy": 0.16661376655101776,
      "num_tokens": 62533360.0,
      "step": 18720
    },
    {
      "entropy": 6.187308645248413,
      "epoch": 3.653009462491464,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003754333218345354,
      "loss": 5.7841,
      "mean_token_accuracy": 0.17276209592819214,
      "num_tokens": 62549175.0,
      "step": 18725
    },
    {
      "entropy": 6.1404181003570555,
      "epoch": 3.653984977075407,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037537037851929965,
      "loss": 5.7566,
      "mean_token_accuracy": 0.17168797701597213,
      "num_tokens": 62566750.0,
      "step": 18730
    },
    {
      "entropy": 6.192042255401612,
      "epoch": 3.6549604916593506,
      "grad_norm": 1.0,
      "learning_rate": 0.00037530742539622835,
      "loss": 5.7404,
      "mean_token_accuracy": 0.17043737173080445,
      "num_tokens": 62582692.0,
      "step": 18735
    },
    {
      "entropy": 6.1526100635528564,
      "epoch": 3.6559360062432935,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00037524446247147285,
      "loss": 5.7367,
      "mean_token_accuracy": 0.17706303298473358,
      "num_tokens": 62599500.0,
      "step": 18740
    },
    {
      "entropy": 6.221291542053223,
      "epoch": 3.6569115208272365,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00037518148975118577,
      "loss": 5.8117,
      "mean_token_accuracy": 0.16868925243616104,
      "num_tokens": 62617528.0,
      "step": 18745
    },
    {
      "entropy": 6.1551628589630125,
      "epoch": 3.6578870354111794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003751185072415206,
      "loss": 5.6878,
      "mean_token_accuracy": 0.17879130095243453,
      "num_tokens": 62634062.0,
      "step": 18750
    },
    {
      "entropy": 6.127584505081177,
      "epoch": 3.6588625499951224,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037505551494863165,
      "loss": 5.7483,
      "mean_token_accuracy": 0.16839566230773925,
      "num_tokens": 62650684.0,
      "step": 18755
    },
    {
      "entropy": 6.174635505676269,
      "epoch": 3.6598380645790654,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003749925128786744,
      "loss": 5.7844,
      "mean_token_accuracy": 0.15988298803567885,
      "num_tokens": 62666967.0,
      "step": 18760
    },
    {
      "entropy": 6.24542007446289,
      "epoch": 3.6608135791630083,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037492950103780524,
      "loss": 5.8086,
      "mean_token_accuracy": 0.16849263608455659,
      "num_tokens": 62684128.0,
      "step": 18765
    },
    {
      "entropy": 6.1499903202056885,
      "epoch": 3.6617890937469513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003748664794321813,
      "loss": 5.7782,
      "mean_token_accuracy": 0.16495711654424666,
      "num_tokens": 62700218.0,
      "step": 18770
    },
    {
      "entropy": 6.171836090087891,
      "epoch": 3.6627646083308947,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037480344806796104,
      "loss": 5.7158,
      "mean_token_accuracy": 0.16704132705926894,
      "num_tokens": 62716326.0,
      "step": 18775
    },
    {
      "entropy": 6.280378961563111,
      "epoch": 3.6637401229148376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037474040695130353,
      "loss": 5.8297,
      "mean_token_accuracy": 0.16352569460868835,
      "num_tokens": 62732534.0,
      "step": 18780
    },
    {
      "entropy": 6.218710088729859,
      "epoch": 3.6647156374987806,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003746773560883688,
      "loss": 5.8026,
      "mean_token_accuracy": 0.17222462594509125,
      "num_tokens": 62749018.0,
      "step": 18785
    },
    {
      "entropy": 6.174106550216675,
      "epoch": 3.6656911520827236,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003746142954853182,
      "loss": 5.7409,
      "mean_token_accuracy": 0.1707424119114876,
      "num_tokens": 62765122.0,
      "step": 18790
    },
    {
      "entropy": 6.16850700378418,
      "epoch": 3.6666666666666665,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00037455122514831366,
      "loss": 5.7094,
      "mean_token_accuracy": 0.17385227978229523,
      "num_tokens": 62781598.0,
      "step": 18795
    },
    {
      "entropy": 6.084231233596801,
      "epoch": 3.66764218125061,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037448814508351814,
      "loss": 5.6764,
      "mean_token_accuracy": 0.17079636305570603,
      "num_tokens": 62799060.0,
      "step": 18800
    },
    {
      "entropy": 6.189211225509643,
      "epoch": 3.668617695834553,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037442505529709574,
      "loss": 5.751,
      "mean_token_accuracy": 0.17387328147888184,
      "num_tokens": 62815265.0,
      "step": 18805
    },
    {
      "entropy": 6.2556602478027346,
      "epoch": 3.669593210418496,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037436195579521124,
      "loss": 5.8739,
      "mean_token_accuracy": 0.16206503212451934,
      "num_tokens": 62832644.0,
      "step": 18810
    },
    {
      "entropy": 6.10641655921936,
      "epoch": 3.670568725002439,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037429884658403065,
      "loss": 5.6976,
      "mean_token_accuracy": 0.175459848344326,
      "num_tokens": 62850522.0,
      "step": 18815
    },
    {
      "entropy": 5.9086048126220705,
      "epoch": 3.6715442395863818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037423572766972054,
      "loss": 5.4717,
      "mean_token_accuracy": 0.2049630329012871,
      "num_tokens": 62868627.0,
      "step": 18820
    },
    {
      "entropy": 6.222802972793579,
      "epoch": 3.6725197541703247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003741725990584489,
      "loss": 5.8359,
      "mean_token_accuracy": 0.1652204692363739,
      "num_tokens": 62884635.0,
      "step": 18825
    },
    {
      "entropy": 6.28203501701355,
      "epoch": 3.6734952687542677,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003741094607563843,
      "loss": 5.9136,
      "mean_token_accuracy": 0.1547347292304039,
      "num_tokens": 62901345.0,
      "step": 18830
    },
    {
      "entropy": 6.1746461391448975,
      "epoch": 3.6744707833382106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037404631276969643,
      "loss": 5.7712,
      "mean_token_accuracy": 0.1761167123913765,
      "num_tokens": 62917594.0,
      "step": 18835
    },
    {
      "entropy": 6.1129309177398685,
      "epoch": 3.675446297922154,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037398315510455587,
      "loss": 5.7183,
      "mean_token_accuracy": 0.17252165973186492,
      "num_tokens": 62934705.0,
      "step": 18840
    },
    {
      "entropy": 6.190115976333618,
      "epoch": 3.676421812506097,
      "grad_norm": 1.0,
      "learning_rate": 0.00037391998776713426,
      "loss": 5.752,
      "mean_token_accuracy": 0.16814497709274293,
      "num_tokens": 62950720.0,
      "step": 18845
    },
    {
      "entropy": 6.243361949920654,
      "epoch": 3.67739732709004,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003738568107636039,
      "loss": 5.838,
      "mean_token_accuracy": 0.1682617872953415,
      "num_tokens": 62965808.0,
      "step": 18850
    },
    {
      "entropy": 6.179583215713501,
      "epoch": 3.678372841673983,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003737936241001384,
      "loss": 5.7657,
      "mean_token_accuracy": 0.1705608084797859,
      "num_tokens": 62982506.0,
      "step": 18855
    },
    {
      "entropy": 6.2146484375,
      "epoch": 3.6793483562579263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000373730427782912,
      "loss": 5.8106,
      "mean_token_accuracy": 0.1607588455080986,
      "num_tokens": 62997975.0,
      "step": 18860
    },
    {
      "entropy": 6.191304397583008,
      "epoch": 3.6803238708418693,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037366722181810013,
      "loss": 5.7442,
      "mean_token_accuracy": 0.164962475001812,
      "num_tokens": 63014838.0,
      "step": 18865
    },
    {
      "entropy": 6.273276615142822,
      "epoch": 3.6812993854258123,
      "grad_norm": 0.953125,
      "learning_rate": 0.000373604006211879,
      "loss": 5.831,
      "mean_token_accuracy": 0.15738908648490907,
      "num_tokens": 63031545.0,
      "step": 18870
    },
    {
      "entropy": 6.072709321975708,
      "epoch": 3.682274900009755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037354078097042576,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18736693263053894,
      "num_tokens": 63048834.0,
      "step": 18875
    },
    {
      "entropy": 6.220409917831421,
      "epoch": 3.683250414593698,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037347754609991864,
      "loss": 5.8089,
      "mean_token_accuracy": 0.16335874944925308,
      "num_tokens": 63064043.0,
      "step": 18880
    },
    {
      "entropy": 6.126245737075806,
      "epoch": 3.684225929177641,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037341430160653665,
      "loss": 5.7305,
      "mean_token_accuracy": 0.17356715351343155,
      "num_tokens": 63079478.0,
      "step": 18885
    },
    {
      "entropy": 6.227221441268921,
      "epoch": 3.685201443761584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037335104749645995,
      "loss": 5.8607,
      "mean_token_accuracy": 0.16213216185569762,
      "num_tokens": 63097301.0,
      "step": 18890
    },
    {
      "entropy": 6.1437256813049315,
      "epoch": 3.686176958345527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003732877837758694,
      "loss": 5.7102,
      "mean_token_accuracy": 0.17161976397037507,
      "num_tokens": 63114243.0,
      "step": 18895
    },
    {
      "entropy": 6.141433906555176,
      "epoch": 3.6871524729294705,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003732245104509468,
      "loss": 5.7492,
      "mean_token_accuracy": 0.16636846363544464,
      "num_tokens": 63130810.0,
      "step": 18900
    },
    {
      "entropy": 6.201108407974243,
      "epoch": 3.6881279875134134,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00037316122752787523,
      "loss": 5.7801,
      "mean_token_accuracy": 0.16464801132678986,
      "num_tokens": 63147189.0,
      "step": 18905
    },
    {
      "entropy": 6.232687902450562,
      "epoch": 3.6891035020973564,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003730979350128384,
      "loss": 5.8268,
      "mean_token_accuracy": 0.16504455953836442,
      "num_tokens": 63164027.0,
      "step": 18910
    },
    {
      "entropy": 6.202819061279297,
      "epoch": 3.6900790166812993,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003730346329120208,
      "loss": 5.7994,
      "mean_token_accuracy": 0.17158216387033462,
      "num_tokens": 63182244.0,
      "step": 18915
    },
    {
      "entropy": 6.202239274978638,
      "epoch": 3.6910545312652423,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037297132123160847,
      "loss": 5.8055,
      "mean_token_accuracy": 0.16321628391742707,
      "num_tokens": 63198764.0,
      "step": 18920
    },
    {
      "entropy": 6.259067106246948,
      "epoch": 3.6920300458491857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037290799997778776,
      "loss": 5.7827,
      "mean_token_accuracy": 0.16606993228197098,
      "num_tokens": 63215316.0,
      "step": 18925
    },
    {
      "entropy": 6.192262601852417,
      "epoch": 3.6930055604331287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037284466915674623,
      "loss": 5.7487,
      "mean_token_accuracy": 0.17440325915813445,
      "num_tokens": 63232241.0,
      "step": 18930
    },
    {
      "entropy": 6.144844388961792,
      "epoch": 3.6939810750170716,
      "grad_norm": 1.0,
      "learning_rate": 0.0003727813287746723,
      "loss": 5.6681,
      "mean_token_accuracy": 0.1676945149898529,
      "num_tokens": 63249077.0,
      "step": 18935
    },
    {
      "entropy": 6.0738372802734375,
      "epoch": 3.6949565896010146,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003727179788377555,
      "loss": 5.6604,
      "mean_token_accuracy": 0.1818731904029846,
      "num_tokens": 63266421.0,
      "step": 18940
    },
    {
      "entropy": 6.137603807449341,
      "epoch": 3.6959321041849575,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000372654619352186,
      "loss": 5.7506,
      "mean_token_accuracy": 0.17311102002859116,
      "num_tokens": 63281776.0,
      "step": 18945
    },
    {
      "entropy": 6.1220532894134525,
      "epoch": 3.6969076187689005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003725912503241552,
      "loss": 5.725,
      "mean_token_accuracy": 0.17396869957447053,
      "num_tokens": 63298200.0,
      "step": 18950
    },
    {
      "entropy": 6.067586660385132,
      "epoch": 3.6978831333528435,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037252787175985527,
      "loss": 5.5702,
      "mean_token_accuracy": 0.18371629416942598,
      "num_tokens": 63314905.0,
      "step": 18955
    },
    {
      "entropy": 6.102956008911133,
      "epoch": 3.6988586479367864,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003724644836654792,
      "loss": 5.767,
      "mean_token_accuracy": 0.17202410250902175,
      "num_tokens": 63331523.0,
      "step": 18960
    },
    {
      "entropy": 6.237825679779053,
      "epoch": 3.69983416252073,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003724010860472213,
      "loss": 5.7963,
      "mean_token_accuracy": 0.166073215007782,
      "num_tokens": 63348850.0,
      "step": 18965
    },
    {
      "entropy": 6.157182693481445,
      "epoch": 3.700809677104673,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003723376789112763,
      "loss": 5.7106,
      "mean_token_accuracy": 0.1721932053565979,
      "num_tokens": 63366918.0,
      "step": 18970
    },
    {
      "entropy": 6.2039073467254635,
      "epoch": 3.7017851916886158,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003722742622638403,
      "loss": 5.807,
      "mean_token_accuracy": 0.16313412487506868,
      "num_tokens": 63385398.0,
      "step": 18975
    },
    {
      "entropy": 6.254578161239624,
      "epoch": 3.7027607062725587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037221083611111,
      "loss": 5.8099,
      "mean_token_accuracy": 0.16658218801021576,
      "num_tokens": 63402075.0,
      "step": 18980
    },
    {
      "entropy": 6.169415521621704,
      "epoch": 3.7037362208565017,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037214740045928333,
      "loss": 5.6733,
      "mean_token_accuracy": 0.1790804535150528,
      "num_tokens": 63418833.0,
      "step": 18985
    },
    {
      "entropy": 6.192022037506104,
      "epoch": 3.704711735440445,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003720839553145589,
      "loss": 5.8204,
      "mean_token_accuracy": 0.16644377410411834,
      "num_tokens": 63434112.0,
      "step": 18990
    },
    {
      "entropy": 6.0996263980865475,
      "epoch": 3.705687250024388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003720205006831364,
      "loss": 5.6353,
      "mean_token_accuracy": 0.18476065546274184,
      "num_tokens": 63450737.0,
      "step": 18995
    },
    {
      "entropy": 6.207696533203125,
      "epoch": 3.706662764608331,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003719570365712164,
      "loss": 5.7937,
      "mean_token_accuracy": 0.1680603489279747,
      "num_tokens": 63466553.0,
      "step": 19000
    },
    {
      "entropy": 6.077584505081177,
      "epoch": 3.707638279192274,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003718935629850003,
      "loss": 5.6772,
      "mean_token_accuracy": 0.18403969407081605,
      "num_tokens": 63482179.0,
      "step": 19005
    },
    {
      "entropy": 6.1179423332214355,
      "epoch": 3.708613793776217,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0003718300799306905,
      "loss": 5.6649,
      "mean_token_accuracy": 0.17540279626846314,
      "num_tokens": 63499435.0,
      "step": 19010
    },
    {
      "entropy": 6.0985342502594,
      "epoch": 3.70958930836016,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037176658741449036,
      "loss": 5.7116,
      "mean_token_accuracy": 0.18008563965559005,
      "num_tokens": 63518351.0,
      "step": 19015
    },
    {
      "entropy": 6.135393714904785,
      "epoch": 3.710564822944103,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00037170308544260436,
      "loss": 5.6402,
      "mean_token_accuracy": 0.18153681755065917,
      "num_tokens": 63534133.0,
      "step": 19020
    },
    {
      "entropy": 6.20279712677002,
      "epoch": 3.711540337528046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003716395740212374,
      "loss": 5.8042,
      "mean_token_accuracy": 0.16535217314958572,
      "num_tokens": 63551076.0,
      "step": 19025
    },
    {
      "entropy": 6.237430763244629,
      "epoch": 3.712515852111989,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037157605315659564,
      "loss": 5.8692,
      "mean_token_accuracy": 0.15615181475877762,
      "num_tokens": 63567380.0,
      "step": 19030
    },
    {
      "entropy": 6.243799543380737,
      "epoch": 3.713491366695932,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003715125228548863,
      "loss": 5.8346,
      "mean_token_accuracy": 0.164350263774395,
      "num_tokens": 63583318.0,
      "step": 19035
    },
    {
      "entropy": 6.232751941680908,
      "epoch": 3.714466881279875,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037144898312231714,
      "loss": 5.8077,
      "mean_token_accuracy": 0.16383724808692932,
      "num_tokens": 63601667.0,
      "step": 19040
    },
    {
      "entropy": 6.2863123416900635,
      "epoch": 3.715442395863818,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003713854339650971,
      "loss": 5.9132,
      "mean_token_accuracy": 0.15848598331212999,
      "num_tokens": 63620335.0,
      "step": 19045
    },
    {
      "entropy": 6.231946849822998,
      "epoch": 3.716417910447761,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037132187538943594,
      "loss": 5.7998,
      "mean_token_accuracy": 0.17460187077522277,
      "num_tokens": 63637337.0,
      "step": 19050
    },
    {
      "entropy": 6.03164119720459,
      "epoch": 3.7173934250317044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003712583074015445,
      "loss": 5.5971,
      "mean_token_accuracy": 0.18645251393318177,
      "num_tokens": 63654358.0,
      "step": 19055
    },
    {
      "entropy": 6.271990013122559,
      "epoch": 3.7183689396156474,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037119473000763424,
      "loss": 5.8993,
      "mean_token_accuracy": 0.15949896574020386,
      "num_tokens": 63670448.0,
      "step": 19060
    },
    {
      "entropy": 6.179245042800903,
      "epoch": 3.7193444541995904,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003711311432139179,
      "loss": 5.6591,
      "mean_token_accuracy": 0.1823075920343399,
      "num_tokens": 63687912.0,
      "step": 19065
    },
    {
      "entropy": 6.177372074127197,
      "epoch": 3.7203199687835333,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037106754702660884,
      "loss": 5.8349,
      "mean_token_accuracy": 0.16445671319961547,
      "num_tokens": 63703949.0,
      "step": 19070
    },
    {
      "entropy": 6.130731201171875,
      "epoch": 3.7212954833674763,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003710039414519215,
      "loss": 5.7198,
      "mean_token_accuracy": 0.17165960520505905,
      "num_tokens": 63722182.0,
      "step": 19075
    },
    {
      "entropy": 6.193016576766968,
      "epoch": 3.7222709979514192,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000370940326496071,
      "loss": 5.731,
      "mean_token_accuracy": 0.17041251361370086,
      "num_tokens": 63738055.0,
      "step": 19080
    },
    {
      "entropy": 6.272898578643799,
      "epoch": 3.723246512535362,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037087670216527386,
      "loss": 5.8612,
      "mean_token_accuracy": 0.16127733439207076,
      "num_tokens": 63755570.0,
      "step": 19085
    },
    {
      "entropy": 6.183527898788452,
      "epoch": 3.724222027119305,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003708130684657471,
      "loss": 5.7915,
      "mean_token_accuracy": 0.17279002964496612,
      "num_tokens": 63771032.0,
      "step": 19090
    },
    {
      "entropy": 6.190408372879029,
      "epoch": 3.7251975417032486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037074942540370876,
      "loss": 5.7378,
      "mean_token_accuracy": 0.16861989796161653,
      "num_tokens": 63788402.0,
      "step": 19095
    },
    {
      "entropy": 6.190306711196899,
      "epoch": 3.7261730562871915,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003706857729853778,
      "loss": 5.7443,
      "mean_token_accuracy": 0.17203402668237686,
      "num_tokens": 63805609.0,
      "step": 19100
    },
    {
      "entropy": 6.182787227630615,
      "epoch": 3.7271485708711345,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003706221112169742,
      "loss": 5.7902,
      "mean_token_accuracy": 0.17226737588644028,
      "num_tokens": 63821146.0,
      "step": 19105
    },
    {
      "entropy": 6.117856645584107,
      "epoch": 3.7281240854550775,
      "grad_norm": 1.0,
      "learning_rate": 0.0003705584401047186,
      "loss": 5.6916,
      "mean_token_accuracy": 0.17361207455396652,
      "num_tokens": 63838948.0,
      "step": 19110
    },
    {
      "entropy": 6.167585277557373,
      "epoch": 3.729099600039021,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003704947596548329,
      "loss": 5.7568,
      "mean_token_accuracy": 0.1700035274028778,
      "num_tokens": 63854887.0,
      "step": 19115
    },
    {
      "entropy": 6.133130788803101,
      "epoch": 3.730075114622964,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003704310698735395,
      "loss": 5.7033,
      "mean_token_accuracy": 0.17647046893835067,
      "num_tokens": 63873903.0,
      "step": 19120
    },
    {
      "entropy": 6.180143880844116,
      "epoch": 3.7310506292069068,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003703673707670622,
      "loss": 5.767,
      "mean_token_accuracy": 0.16708868741989136,
      "num_tokens": 63890540.0,
      "step": 19125
    },
    {
      "entropy": 6.140535974502564,
      "epoch": 3.7320261437908497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037030366234162533,
      "loss": 5.6999,
      "mean_token_accuracy": 0.17434243708848954,
      "num_tokens": 63905842.0,
      "step": 19130
    },
    {
      "entropy": 6.163537645339966,
      "epoch": 3.7330016583747927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003702399446034542,
      "loss": 5.649,
      "mean_token_accuracy": 0.17997652888298035,
      "num_tokens": 63921505.0,
      "step": 19135
    },
    {
      "entropy": 6.123333549499511,
      "epoch": 3.7339771729587357,
      "grad_norm": 1.0,
      "learning_rate": 0.0003701762175587752,
      "loss": 5.7294,
      "mean_token_accuracy": 0.16886010617017747,
      "num_tokens": 63936375.0,
      "step": 19140
    },
    {
      "entropy": 6.180980634689331,
      "epoch": 3.7349526875426786,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003701124812138153,
      "loss": 5.8728,
      "mean_token_accuracy": 0.15658996999263763,
      "num_tokens": 63953129.0,
      "step": 19145
    },
    {
      "entropy": 6.35427508354187,
      "epoch": 3.7359282021266216,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003700487355748028,
      "loss": 6.0069,
      "mean_token_accuracy": 0.15302463620901108,
      "num_tokens": 63969806.0,
      "step": 19150
    },
    {
      "entropy": 6.25797176361084,
      "epoch": 3.736903716710565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036998498064796665,
      "loss": 5.8449,
      "mean_token_accuracy": 0.1629552349448204,
      "num_tokens": 63985536.0,
      "step": 19155
    },
    {
      "entropy": 6.18504490852356,
      "epoch": 3.737879231294508,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036992121643953665,
      "loss": 5.8277,
      "mean_token_accuracy": 0.16229649037122726,
      "num_tokens": 64002346.0,
      "step": 19160
    },
    {
      "entropy": 6.2867261409759525,
      "epoch": 3.738854745878451,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003698574429557437,
      "loss": 5.848,
      "mean_token_accuracy": 0.1614130973815918,
      "num_tokens": 64019430.0,
      "step": 19165
    },
    {
      "entropy": 6.034282207489014,
      "epoch": 3.739830260462394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036979366020281954,
      "loss": 5.5293,
      "mean_token_accuracy": 0.19006356298923494,
      "num_tokens": 64036802.0,
      "step": 19170
    },
    {
      "entropy": 6.15432448387146,
      "epoch": 3.740805775046337,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003697298681869967,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16870122849941255,
      "num_tokens": 64053818.0,
      "step": 19175
    },
    {
      "entropy": 6.158634757995605,
      "epoch": 3.7417812896302802,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003696660669145088,
      "loss": 5.7132,
      "mean_token_accuracy": 0.17451692074537278,
      "num_tokens": 64070424.0,
      "step": 19180
    },
    {
      "entropy": 6.207339239120484,
      "epoch": 3.742756804214223,
      "grad_norm": 1.0,
      "learning_rate": 0.0003696022563915903,
      "loss": 5.7957,
      "mean_token_accuracy": 0.16740633994340898,
      "num_tokens": 64089103.0,
      "step": 19185
    },
    {
      "entropy": 6.135363531112671,
      "epoch": 3.743732318798166,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036953843662447636,
      "loss": 5.6687,
      "mean_token_accuracy": 0.1800054132938385,
      "num_tokens": 64105743.0,
      "step": 19190
    },
    {
      "entropy": 6.1244744777679445,
      "epoch": 3.744707833382109,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003694746076194034,
      "loss": 5.7468,
      "mean_token_accuracy": 0.1731436789035797,
      "num_tokens": 64124535.0,
      "step": 19195
    },
    {
      "entropy": 6.242048597335815,
      "epoch": 3.745683347966052,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036941076938260854,
      "loss": 5.83,
      "mean_token_accuracy": 0.17342693954706193,
      "num_tokens": 64142430.0,
      "step": 19200
    },
    {
      "entropy": 6.139105224609375,
      "epoch": 3.746658862549995,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003693469219203297,
      "loss": 5.7175,
      "mean_token_accuracy": 0.17142631709575654,
      "num_tokens": 64159592.0,
      "step": 19205
    },
    {
      "entropy": 6.1197922706604,
      "epoch": 3.747634377133938,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036928306523880593,
      "loss": 5.647,
      "mean_token_accuracy": 0.17079166918992997,
      "num_tokens": 64176023.0,
      "step": 19210
    },
    {
      "entropy": 6.081421136856079,
      "epoch": 3.748609891717881,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036921919934427704,
      "loss": 5.6095,
      "mean_token_accuracy": 0.1856187254190445,
      "num_tokens": 64192372.0,
      "step": 19215
    },
    {
      "entropy": 6.224906969070434,
      "epoch": 3.7495854063018244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036915532424298383,
      "loss": 5.7911,
      "mean_token_accuracy": 0.1649242103099823,
      "num_tokens": 64210420.0,
      "step": 19220
    },
    {
      "entropy": 6.212627410888672,
      "epoch": 3.7505609208857673,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003690914399411678,
      "loss": 5.8156,
      "mean_token_accuracy": 0.16752398014068604,
      "num_tokens": 64228281.0,
      "step": 19225
    },
    {
      "entropy": 6.075934648513794,
      "epoch": 3.7515364354697103,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036902754644507177,
      "loss": 5.6419,
      "mean_token_accuracy": 0.18270799815654754,
      "num_tokens": 64244002.0,
      "step": 19230
    },
    {
      "entropy": 6.1755448341369625,
      "epoch": 3.7525119500536532,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036896364376093896,
      "loss": 5.7623,
      "mean_token_accuracy": 0.17099791169166564,
      "num_tokens": 64259849.0,
      "step": 19235
    },
    {
      "entropy": 6.198581743240356,
      "epoch": 3.753487464637596,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036889973189501373,
      "loss": 5.7251,
      "mean_token_accuracy": 0.17427580654621125,
      "num_tokens": 64278760.0,
      "step": 19240
    },
    {
      "entropy": 6.1257518291473385,
      "epoch": 3.7544629792215396,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003688358108535414,
      "loss": 5.672,
      "mean_token_accuracy": 0.17886927723884583,
      "num_tokens": 64296633.0,
      "step": 19245
    },
    {
      "entropy": 6.1979461193084715,
      "epoch": 3.7554384938054826,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036877188064276807,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1709323450922966,
      "num_tokens": 64313278.0,
      "step": 19250
    },
    {
      "entropy": 6.063502454757691,
      "epoch": 3.7564140083894255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003687079412689408,
      "loss": 5.6406,
      "mean_token_accuracy": 0.18523443043231963,
      "num_tokens": 64331521.0,
      "step": 19255
    },
    {
      "entropy": 6.0526903629302975,
      "epoch": 3.7573895229733685,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036864399273830745,
      "loss": 5.5657,
      "mean_token_accuracy": 0.18310745805501938,
      "num_tokens": 64347633.0,
      "step": 19260
    },
    {
      "entropy": 6.245188283920288,
      "epoch": 3.7583650375573114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036858003505711685,
      "loss": 5.8705,
      "mean_token_accuracy": 0.1576077789068222,
      "num_tokens": 64363955.0,
      "step": 19265
    },
    {
      "entropy": 6.1186792850494385,
      "epoch": 3.7593405521412544,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003685160682316188,
      "loss": 5.6764,
      "mean_token_accuracy": 0.17720528095960617,
      "num_tokens": 64382109.0,
      "step": 19270
    },
    {
      "entropy": 6.224259519577027,
      "epoch": 3.7603160667251974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036845209226806386,
      "loss": 5.8316,
      "mean_token_accuracy": 0.16274044662714005,
      "num_tokens": 64398912.0,
      "step": 19275
    },
    {
      "entropy": 6.1868620872497555,
      "epoch": 3.7612915813091403,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003683881071727035,
      "loss": 5.715,
      "mean_token_accuracy": 0.17318140864372253,
      "num_tokens": 64415809.0,
      "step": 19280
    },
    {
      "entropy": 6.0231041431427,
      "epoch": 3.7622670958930837,
      "grad_norm": 0.828125,
      "learning_rate": 0.0003683241129517902,
      "loss": 5.525,
      "mean_token_accuracy": 0.19030694365501405,
      "num_tokens": 64431900.0,
      "step": 19285
    },
    {
      "entropy": 6.17198805809021,
      "epoch": 3.7632426104770267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036826010961157717,
      "loss": 5.7625,
      "mean_token_accuracy": 0.1698137938976288,
      "num_tokens": 64447491.0,
      "step": 19290
    },
    {
      "entropy": 6.177663993835449,
      "epoch": 3.7642181250609696,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036819609715831866,
      "loss": 5.7905,
      "mean_token_accuracy": 0.16731888353824614,
      "num_tokens": 64465121.0,
      "step": 19295
    },
    {
      "entropy": 6.123001194000244,
      "epoch": 3.7651936396449126,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036813207559826963,
      "loss": 5.7035,
      "mean_token_accuracy": 0.17303216457366943,
      "num_tokens": 64481044.0,
      "step": 19300
    },
    {
      "entropy": 6.052958679199219,
      "epoch": 3.7661691542288556,
      "grad_norm": 1.0,
      "learning_rate": 0.0003680680449376861,
      "loss": 5.5954,
      "mean_token_accuracy": 0.17716532349586486,
      "num_tokens": 64498228.0,
      "step": 19305
    },
    {
      "entropy": 6.114635610580445,
      "epoch": 3.767144668812799,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036800400518282494,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1753512889146805,
      "num_tokens": 64514815.0,
      "step": 19310
    },
    {
      "entropy": 6.168979454040527,
      "epoch": 3.768120183396742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036793995633994383,
      "loss": 5.7849,
      "mean_token_accuracy": 0.17642999589443206,
      "num_tokens": 64533453.0,
      "step": 19315
    },
    {
      "entropy": 6.243974447250366,
      "epoch": 3.769095697980685,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003678758984153015,
      "loss": 5.8662,
      "mean_token_accuracy": 0.16315729171037674,
      "num_tokens": 64551118.0,
      "step": 19320
    },
    {
      "entropy": 6.174286890029907,
      "epoch": 3.770071212564628,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003678118314151573,
      "loss": 5.7934,
      "mean_token_accuracy": 0.17143791168928146,
      "num_tokens": 64567472.0,
      "step": 19325
    },
    {
      "entropy": 6.203152799606324,
      "epoch": 3.771046727148571,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036774775534577174,
      "loss": 5.8093,
      "mean_token_accuracy": 0.16886450797319413,
      "num_tokens": 64583526.0,
      "step": 19330
    },
    {
      "entropy": 6.204670095443726,
      "epoch": 3.7720222417325138,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036768367021340615,
      "loss": 5.8177,
      "mean_token_accuracy": 0.16477226167917253,
      "num_tokens": 64600647.0,
      "step": 19335
    },
    {
      "entropy": 6.071776294708252,
      "epoch": 3.7729977563164567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003676195760243226,
      "loss": 5.6031,
      "mean_token_accuracy": 0.18077300637960433,
      "num_tokens": 64616963.0,
      "step": 19340
    },
    {
      "entropy": 6.140085077285766,
      "epoch": 3.7739732709003997,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003675554727847842,
      "loss": 5.7211,
      "mean_token_accuracy": 0.17769014835357666,
      "num_tokens": 64634078.0,
      "step": 19345
    },
    {
      "entropy": 6.218139266967773,
      "epoch": 3.774948785484343,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036749136050105485,
      "loss": 5.7834,
      "mean_token_accuracy": 0.164497110247612,
      "num_tokens": 64650056.0,
      "step": 19350
    },
    {
      "entropy": 6.237177419662475,
      "epoch": 3.775924300068286,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003674272391793994,
      "loss": 5.8349,
      "mean_token_accuracy": 0.16330410093069075,
      "num_tokens": 64667667.0,
      "step": 19355
    },
    {
      "entropy": 5.965578699111939,
      "epoch": 3.776899814652229,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0003673631088260836,
      "loss": 5.5457,
      "mean_token_accuracy": 0.1991736561059952,
      "num_tokens": 64685618.0,
      "step": 19360
    },
    {
      "entropy": 6.164388036727905,
      "epoch": 3.777875329236172,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036729896944737395,
      "loss": 5.778,
      "mean_token_accuracy": 0.1695023760199547,
      "num_tokens": 64702551.0,
      "step": 19365
    },
    {
      "entropy": 6.22850489616394,
      "epoch": 3.7788508438201154,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003672348210495379,
      "loss": 5.8932,
      "mean_token_accuracy": 0.16205960810184478,
      "num_tokens": 64719648.0,
      "step": 19370
    },
    {
      "entropy": 6.2424756526947025,
      "epoch": 3.7798263584040583,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000367170663638844,
      "loss": 5.8426,
      "mean_token_accuracy": 0.16450008153915405,
      "num_tokens": 64736047.0,
      "step": 19375
    },
    {
      "entropy": 6.089214897155761,
      "epoch": 3.7808018729880013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003671064972215613,
      "loss": 5.7429,
      "mean_token_accuracy": 0.17135773450136185,
      "num_tokens": 64751832.0,
      "step": 19380
    },
    {
      "entropy": 6.187060832977295,
      "epoch": 3.7817773875719443,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036704232180396004,
      "loss": 5.7979,
      "mean_token_accuracy": 0.16131854355335234,
      "num_tokens": 64768744.0,
      "step": 19385
    },
    {
      "entropy": 6.125353860855102,
      "epoch": 3.782752902155887,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000366978137392311,
      "loss": 5.6681,
      "mean_token_accuracy": 0.17659587264060975,
      "num_tokens": 64785133.0,
      "step": 19390
    },
    {
      "entropy": 6.108879518508911,
      "epoch": 3.78372841673983,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036691394399288626,
      "loss": 5.724,
      "mean_token_accuracy": 0.178168186545372,
      "num_tokens": 64801965.0,
      "step": 19395
    },
    {
      "entropy": 6.242418241500855,
      "epoch": 3.784703931323773,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036684974161195846,
      "loss": 5.8415,
      "mean_token_accuracy": 0.16508519053459167,
      "num_tokens": 64819821.0,
      "step": 19400
    },
    {
      "entropy": 6.175282192230225,
      "epoch": 3.785679445907716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036678553025580124,
      "loss": 5.7078,
      "mean_token_accuracy": 0.166381274163723,
      "num_tokens": 64836115.0,
      "step": 19405
    },
    {
      "entropy": 6.239202451705933,
      "epoch": 3.7866549604916595,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036672130993068927,
      "loss": 5.9186,
      "mean_token_accuracy": 0.15369196534156798,
      "num_tokens": 64853225.0,
      "step": 19410
    },
    {
      "entropy": 6.246558904647827,
      "epoch": 3.7876304750756025,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036665708064289774,
      "loss": 5.8171,
      "mean_token_accuracy": 0.17237705886363983,
      "num_tokens": 64870400.0,
      "step": 19415
    },
    {
      "entropy": 6.085214900970459,
      "epoch": 3.7886059896595454,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003665928423987029,
      "loss": 5.5985,
      "mean_token_accuracy": 0.18455255627632142,
      "num_tokens": 64887567.0,
      "step": 19420
    },
    {
      "entropy": 5.9836132526397705,
      "epoch": 3.7895815042434884,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003665285952043819,
      "loss": 5.5543,
      "mean_token_accuracy": 0.1820947289466858,
      "num_tokens": 64903700.0,
      "step": 19425
    },
    {
      "entropy": 6.18402738571167,
      "epoch": 3.7905570188274313,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003664643390662129,
      "loss": 5.7852,
      "mean_token_accuracy": 0.16822271049022675,
      "num_tokens": 64921386.0,
      "step": 19430
    },
    {
      "entropy": 6.147191953659058,
      "epoch": 3.7915325334113748,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036640007399047457,
      "loss": 5.728,
      "mean_token_accuracy": 0.1750026375055313,
      "num_tokens": 64938180.0,
      "step": 19435
    },
    {
      "entropy": 6.237868595123291,
      "epoch": 3.7925080479953177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036633579998344686,
      "loss": 5.8457,
      "mean_token_accuracy": 0.16194757968187332,
      "num_tokens": 64954489.0,
      "step": 19440
    },
    {
      "entropy": 6.133492422103882,
      "epoch": 3.7934835625792607,
      "grad_norm": 0.875,
      "learning_rate": 0.0003662715170514103,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1773966759443283,
      "num_tokens": 64971492.0,
      "step": 19445
    },
    {
      "entropy": 6.243435573577881,
      "epoch": 3.7944590771632036,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003662072252006464,
      "loss": 5.8123,
      "mean_token_accuracy": 0.1615425243973732,
      "num_tokens": 64988322.0,
      "step": 19450
    },
    {
      "entropy": 6.2443701267242435,
      "epoch": 3.7954345917471466,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003661429244374374,
      "loss": 5.898,
      "mean_token_accuracy": 0.16262697726488112,
      "num_tokens": 65004973.0,
      "step": 19455
    },
    {
      "entropy": 6.205704307556152,
      "epoch": 3.7964101063310896,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003660786147680668,
      "loss": 5.821,
      "mean_token_accuracy": 0.16743358820676804,
      "num_tokens": 65021532.0,
      "step": 19460
    },
    {
      "entropy": 6.146353340148925,
      "epoch": 3.7973856209150325,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036601429619881847,
      "loss": 5.6513,
      "mean_token_accuracy": 0.17818712145090104,
      "num_tokens": 65037896.0,
      "step": 19465
    },
    {
      "entropy": 6.212851858139038,
      "epoch": 3.7983611354989755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003659499687359776,
      "loss": 5.7678,
      "mean_token_accuracy": 0.16556842923164367,
      "num_tokens": 65054000.0,
      "step": 19470
    },
    {
      "entropy": 6.185162734985352,
      "epoch": 3.799336650082919,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003658856323858299,
      "loss": 5.855,
      "mean_token_accuracy": 0.16444188505411148,
      "num_tokens": 65070818.0,
      "step": 19475
    },
    {
      "entropy": 6.153944969177246,
      "epoch": 3.800312164666862,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036582128715466214,
      "loss": 5.7085,
      "mean_token_accuracy": 0.17902918457984923,
      "num_tokens": 65087129.0,
      "step": 19480
    },
    {
      "entropy": 6.247226238250732,
      "epoch": 3.801287679250805,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003657569330487618,
      "loss": 5.9109,
      "mean_token_accuracy": 0.16112104952335357,
      "num_tokens": 65103541.0,
      "step": 19485
    },
    {
      "entropy": 6.204540205001831,
      "epoch": 3.8022631938347478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036569257007441745,
      "loss": 5.7652,
      "mean_token_accuracy": 0.16653676480054855,
      "num_tokens": 65118924.0,
      "step": 19490
    },
    {
      "entropy": 6.144047784805298,
      "epoch": 3.8032387084186907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003656281982379185,
      "loss": 5.7231,
      "mean_token_accuracy": 0.17948416471481324,
      "num_tokens": 65134690.0,
      "step": 19495
    },
    {
      "entropy": 6.147839832305908,
      "epoch": 3.804214223002634,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036556381754555497,
      "loss": 5.7041,
      "mean_token_accuracy": 0.17178822308778763,
      "num_tokens": 65150495.0,
      "step": 19500
    },
    {
      "entropy": 6.218462324142456,
      "epoch": 3.805189737586577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036549942800361793,
      "loss": 5.7847,
      "mean_token_accuracy": 0.1649606019258499,
      "num_tokens": 65166148.0,
      "step": 19505
    },
    {
      "entropy": 6.230331087112427,
      "epoch": 3.80616525217052,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036543502961839937,
      "loss": 5.798,
      "mean_token_accuracy": 0.1626672312617302,
      "num_tokens": 65182136.0,
      "step": 19510
    },
    {
      "entropy": 6.242058801651001,
      "epoch": 3.807140766754463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036537062239619194,
      "loss": 5.8001,
      "mean_token_accuracy": 0.16737321615219117,
      "num_tokens": 65198884.0,
      "step": 19515
    },
    {
      "entropy": 6.145125532150269,
      "epoch": 3.808116281338406,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036530620634328944,
      "loss": 5.7826,
      "mean_token_accuracy": 0.16620512455701827,
      "num_tokens": 65217274.0,
      "step": 19520
    },
    {
      "entropy": 6.137088060379028,
      "epoch": 3.809091795922349,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003652417814659864,
      "loss": 5.751,
      "mean_token_accuracy": 0.17126494646072388,
      "num_tokens": 65234052.0,
      "step": 19525
    },
    {
      "entropy": 6.252723169326782,
      "epoch": 3.810067310506292,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036517734777057796,
      "loss": 5.7635,
      "mean_token_accuracy": 0.16443238854408265,
      "num_tokens": 65251103.0,
      "step": 19530
    },
    {
      "entropy": 6.196247959136963,
      "epoch": 3.811042825090235,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003651129052633605,
      "loss": 5.7619,
      "mean_token_accuracy": 0.17321498394012452,
      "num_tokens": 65266735.0,
      "step": 19535
    },
    {
      "entropy": 6.184046602249145,
      "epoch": 3.8120183396741782,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003650484539506312,
      "loss": 5.6929,
      "mean_token_accuracy": 0.17756633162498475,
      "num_tokens": 65283990.0,
      "step": 19540
    },
    {
      "entropy": 6.096623754501342,
      "epoch": 3.812993854258121,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036498399383868787,
      "loss": 5.721,
      "mean_token_accuracy": 0.17746609598398208,
      "num_tokens": 65301394.0,
      "step": 19545
    },
    {
      "entropy": 6.1919537544250485,
      "epoch": 3.813969368842064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003649195249338294,
      "loss": 5.7602,
      "mean_token_accuracy": 0.16875679641962052,
      "num_tokens": 65317809.0,
      "step": 19550
    },
    {
      "entropy": 6.104783248901367,
      "epoch": 3.814944883426007,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036485504724235535,
      "loss": 5.6594,
      "mean_token_accuracy": 0.17699486017227173,
      "num_tokens": 65335129.0,
      "step": 19555
    },
    {
      "entropy": 6.067898607254028,
      "epoch": 3.81592039800995,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003647905607705664,
      "loss": 5.6018,
      "mean_token_accuracy": 0.18944627344608306,
      "num_tokens": 65352057.0,
      "step": 19560
    },
    {
      "entropy": 6.234423494338989,
      "epoch": 3.8168959125938935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036472606552476384,
      "loss": 5.8123,
      "mean_token_accuracy": 0.16957614719867706,
      "num_tokens": 65369386.0,
      "step": 19565
    },
    {
      "entropy": 6.143637561798096,
      "epoch": 3.8178714271778365,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036466156151125,
      "loss": 5.7081,
      "mean_token_accuracy": 0.17160283625125886,
      "num_tokens": 65385775.0,
      "step": 19570
    },
    {
      "entropy": 6.180300617218018,
      "epoch": 3.8188469417617794,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036459704873632794,
      "loss": 5.8313,
      "mean_token_accuracy": 0.1601707637310028,
      "num_tokens": 65401795.0,
      "step": 19575
    },
    {
      "entropy": 6.2466246604919435,
      "epoch": 3.8198224563457224,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003645325272063016,
      "loss": 5.8458,
      "mean_token_accuracy": 0.16211240291595458,
      "num_tokens": 65416772.0,
      "step": 19580
    },
    {
      "entropy": 5.968429183959961,
      "epoch": 3.8207979709296653,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036446799692747573,
      "loss": 5.4919,
      "mean_token_accuracy": 0.18807943612337114,
      "num_tokens": 65432485.0,
      "step": 19585
    },
    {
      "entropy": 6.176430749893188,
      "epoch": 3.8217734855136083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003644034579061561,
      "loss": 5.7296,
      "mean_token_accuracy": 0.17845242470502853,
      "num_tokens": 65449141.0,
      "step": 19590
    },
    {
      "entropy": 6.059508895874023,
      "epoch": 3.8227490000975513,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003643389101486492,
      "loss": 5.5662,
      "mean_token_accuracy": 0.19405182898044587,
      "num_tokens": 65467753.0,
      "step": 19595
    },
    {
      "entropy": 6.175834083557129,
      "epoch": 3.823724514681494,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003642743536612624,
      "loss": 5.7472,
      "mean_token_accuracy": 0.17134518623352052,
      "num_tokens": 65485577.0,
      "step": 19600
    },
    {
      "entropy": 6.172492265701294,
      "epoch": 3.8247000292654376,
      "grad_norm": 0.984375,
      "learning_rate": 0.000364209788450304,
      "loss": 5.7534,
      "mean_token_accuracy": 0.17260310649871827,
      "num_tokens": 65501824.0,
      "step": 19605
    },
    {
      "entropy": 6.112289237976074,
      "epoch": 3.8256755438493806,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000364145214522083,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1796448841691017,
      "num_tokens": 65518710.0,
      "step": 19610
    },
    {
      "entropy": 6.073272657394409,
      "epoch": 3.8266510584333235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036408063188290925,
      "loss": 5.6041,
      "mean_token_accuracy": 0.1870590016245842,
      "num_tokens": 65534914.0,
      "step": 19615
    },
    {
      "entropy": 5.979992818832398,
      "epoch": 3.8276265730172665,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003640160405390938,
      "loss": 5.617,
      "mean_token_accuracy": 0.18748721927404405,
      "num_tokens": 65550990.0,
      "step": 19620
    },
    {
      "entropy": 6.1619902610778805,
      "epoch": 3.82860208760121,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000363951440496948,
      "loss": 5.7775,
      "mean_token_accuracy": 0.17063190937042236,
      "num_tokens": 65569255.0,
      "step": 19625
    },
    {
      "entropy": 6.2124286651611325,
      "epoch": 3.829577602185153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003638868317627845,
      "loss": 5.7764,
      "mean_token_accuracy": 0.16955794245004654,
      "num_tokens": 65585926.0,
      "step": 19630
    },
    {
      "entropy": 6.125852775573731,
      "epoch": 3.830553116769096,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00036382221434291665,
      "loss": 5.6985,
      "mean_token_accuracy": 0.17486192435026168,
      "num_tokens": 65602530.0,
      "step": 19635
    },
    {
      "entropy": 6.241074466705323,
      "epoch": 3.831528631353039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003637575882436585,
      "loss": 5.8449,
      "mean_token_accuracy": 0.15833071768283843,
      "num_tokens": 65619566.0,
      "step": 19640
    },
    {
      "entropy": 6.13237361907959,
      "epoch": 3.8325041459369817,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036369295347132526,
      "loss": 5.6876,
      "mean_token_accuracy": 0.1860792577266693,
      "num_tokens": 65636631.0,
      "step": 19645
    },
    {
      "entropy": 6.161198234558105,
      "epoch": 3.8334796605209247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036362831003223264,
      "loss": 5.8099,
      "mean_token_accuracy": 0.16040909737348558,
      "num_tokens": 65652536.0,
      "step": 19650
    },
    {
      "entropy": 6.1604022026062015,
      "epoch": 3.8344551751048677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003635636579326975,
      "loss": 5.706,
      "mean_token_accuracy": 0.17544066905975342,
      "num_tokens": 65667934.0,
      "step": 19655
    },
    {
      "entropy": 6.048477125167847,
      "epoch": 3.8354306896888106,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036349899717903734,
      "loss": 5.5336,
      "mean_token_accuracy": 0.1918069079518318,
      "num_tokens": 65683508.0,
      "step": 19660
    },
    {
      "entropy": 6.223748970031738,
      "epoch": 3.836406204272754,
      "grad_norm": 1.0,
      "learning_rate": 0.00036343432777757054,
      "loss": 5.717,
      "mean_token_accuracy": 0.17113741040229796,
      "num_tokens": 65698198.0,
      "step": 19665
    },
    {
      "entropy": 6.066596221923828,
      "epoch": 3.837381718856697,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003633696497346164,
      "loss": 5.7153,
      "mean_token_accuracy": 0.17955327183008193,
      "num_tokens": 65715867.0,
      "step": 19670
    },
    {
      "entropy": 6.216776418685913,
      "epoch": 3.83835723344064,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003633049630564951,
      "loss": 5.7875,
      "mean_token_accuracy": 0.16393525004386902,
      "num_tokens": 65732466.0,
      "step": 19675
    },
    {
      "entropy": 6.176986837387085,
      "epoch": 3.839332748024583,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036324026774952754,
      "loss": 5.7407,
      "mean_token_accuracy": 0.177761447429657,
      "num_tokens": 65750454.0,
      "step": 19680
    },
    {
      "entropy": 6.223819160461426,
      "epoch": 3.840308262608526,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036317556382003545,
      "loss": 5.8726,
      "mean_token_accuracy": 0.16711514592170715,
      "num_tokens": 65767221.0,
      "step": 19685
    },
    {
      "entropy": 5.976307964324951,
      "epoch": 3.8412837771924693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003631108512743417,
      "loss": 5.4964,
      "mean_token_accuracy": 0.1945937156677246,
      "num_tokens": 65783894.0,
      "step": 19690
    },
    {
      "entropy": 6.114663600921631,
      "epoch": 3.8422592917764122,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00036304613011876934,
      "loss": 5.7075,
      "mean_token_accuracy": 0.1787665531039238,
      "num_tokens": 65799823.0,
      "step": 19695
    },
    {
      "entropy": 6.043512201309204,
      "epoch": 3.843234806360355,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036298140035964313,
      "loss": 5.6486,
      "mean_token_accuracy": 0.1807727947831154,
      "num_tokens": 65816285.0,
      "step": 19700
    },
    {
      "entropy": 6.168577241897583,
      "epoch": 3.844210320944298,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000362916662003288,
      "loss": 5.7764,
      "mean_token_accuracy": 0.174776291847229,
      "num_tokens": 65832092.0,
      "step": 19705
    },
    {
      "entropy": 6.156113958358764,
      "epoch": 3.845185835528241,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036285191505602994,
      "loss": 5.7887,
      "mean_token_accuracy": 0.16780510395765305,
      "num_tokens": 65849232.0,
      "step": 19710
    },
    {
      "entropy": 6.185300207138061,
      "epoch": 3.846161350112184,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036278715952419593,
      "loss": 5.6826,
      "mean_token_accuracy": 0.17975089251995086,
      "num_tokens": 65866723.0,
      "step": 19715
    },
    {
      "entropy": 6.215818548202515,
      "epoch": 3.847136864696127,
      "grad_norm": 1.0,
      "learning_rate": 0.0003627223954141136,
      "loss": 5.7567,
      "mean_token_accuracy": 0.17060885280370713,
      "num_tokens": 65883311.0,
      "step": 19720
    },
    {
      "entropy": 6.263116836547852,
      "epoch": 3.84811237928007,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003626576227321114,
      "loss": 5.8403,
      "mean_token_accuracy": 0.16553237289190292,
      "num_tokens": 65901087.0,
      "step": 19725
    },
    {
      "entropy": 6.1831389427185055,
      "epoch": 3.8490878938640134,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00036259284148451864,
      "loss": 5.68,
      "mean_token_accuracy": 0.176888869702816,
      "num_tokens": 65918404.0,
      "step": 19730
    },
    {
      "entropy": 6.15406231880188,
      "epoch": 3.8500634084479564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036252805167766566,
      "loss": 5.8089,
      "mean_token_accuracy": 0.17201164960861207,
      "num_tokens": 65934512.0,
      "step": 19735
    },
    {
      "entropy": 6.107724618911743,
      "epoch": 3.8510389230318993,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003624632533178835,
      "loss": 5.6884,
      "mean_token_accuracy": 0.17938963025808335,
      "num_tokens": 65951553.0,
      "step": 19740
    },
    {
      "entropy": 6.108796453475952,
      "epoch": 3.8520144376158423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003623984464115038,
      "loss": 5.666,
      "mean_token_accuracy": 0.17970816195011138,
      "num_tokens": 65968573.0,
      "step": 19745
    },
    {
      "entropy": 6.042445850372315,
      "epoch": 3.8529899521997852,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036233363096485946,
      "loss": 5.568,
      "mean_token_accuracy": 0.18913066685199736,
      "num_tokens": 65985270.0,
      "step": 19750
    },
    {
      "entropy": 6.1805590152740475,
      "epoch": 3.8539654667837286,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000362268806984284,
      "loss": 5.8006,
      "mean_token_accuracy": 0.1680007681250572,
      "num_tokens": 66002484.0,
      "step": 19755
    },
    {
      "entropy": 6.136266565322876,
      "epoch": 3.8549409813676716,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036220397447611174,
      "loss": 5.6869,
      "mean_token_accuracy": 0.17690560072660447,
      "num_tokens": 66018724.0,
      "step": 19760
    },
    {
      "entropy": 6.1818407535552975,
      "epoch": 3.8559164959516146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036213913344667795,
      "loss": 5.7287,
      "mean_token_accuracy": 0.17155335694551468,
      "num_tokens": 66035225.0,
      "step": 19765
    },
    {
      "entropy": 6.1932087421417235,
      "epoch": 3.8568920105355575,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003620742839023186,
      "loss": 5.8138,
      "mean_token_accuracy": 0.16670722812414168,
      "num_tokens": 66051766.0,
      "step": 19770
    },
    {
      "entropy": 6.242376041412354,
      "epoch": 3.8578675251195005,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036200942584937047,
      "loss": 5.8294,
      "mean_token_accuracy": 0.1644139900803566,
      "num_tokens": 66068360.0,
      "step": 19775
    },
    {
      "entropy": 6.2383284091949465,
      "epoch": 3.8588430397034434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036194455929417136,
      "loss": 5.7916,
      "mean_token_accuracy": 0.16790044009685517,
      "num_tokens": 66083224.0,
      "step": 19780
    },
    {
      "entropy": 6.221960878372192,
      "epoch": 3.8598185542873864,
      "grad_norm": 1.0,
      "learning_rate": 0.0003618796842430599,
      "loss": 5.7566,
      "mean_token_accuracy": 0.17501020282506943,
      "num_tokens": 66098752.0,
      "step": 19785
    },
    {
      "entropy": 6.210503816604614,
      "epoch": 3.8607940688713294,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036181480070237524,
      "loss": 5.8367,
      "mean_token_accuracy": 0.1630653351545334,
      "num_tokens": 66115556.0,
      "step": 19790
    },
    {
      "entropy": 6.217724323272705,
      "epoch": 3.8617695834552728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003617499086784577,
      "loss": 5.7251,
      "mean_token_accuracy": 0.17165531516075133,
      "num_tokens": 66133195.0,
      "step": 19795
    },
    {
      "entropy": 6.1783136367797855,
      "epoch": 3.8627450980392157,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036168500817764833,
      "loss": 5.8253,
      "mean_token_accuracy": 0.16386525928974152,
      "num_tokens": 66150331.0,
      "step": 19800
    },
    {
      "entropy": 6.176911497116089,
      "epoch": 3.8637206126231587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003616200992062888,
      "loss": 5.7833,
      "mean_token_accuracy": 0.16914025098085403,
      "num_tokens": 66168948.0,
      "step": 19805
    },
    {
      "entropy": 6.1450823783874515,
      "epoch": 3.8646961272071017,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000361555181770722,
      "loss": 5.6691,
      "mean_token_accuracy": 0.17635659426450728,
      "num_tokens": 66185837.0,
      "step": 19810
    },
    {
      "entropy": 5.953151226043701,
      "epoch": 3.8656716417910446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003614902558772912,
      "loss": 5.5868,
      "mean_token_accuracy": 0.18919734209775924,
      "num_tokens": 66203872.0,
      "step": 19815
    },
    {
      "entropy": 6.183252000808716,
      "epoch": 3.866647156374988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036142532153234097,
      "loss": 5.7725,
      "mean_token_accuracy": 0.16480569690465927,
      "num_tokens": 66220017.0,
      "step": 19820
    },
    {
      "entropy": 6.162132978439331,
      "epoch": 3.867622670958931,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003613603787422163,
      "loss": 5.7136,
      "mean_token_accuracy": 0.17316519767045974,
      "num_tokens": 66237277.0,
      "step": 19825
    },
    {
      "entropy": 6.16609001159668,
      "epoch": 3.868598185542874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003612954275132632,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1754566937685013,
      "num_tokens": 66254856.0,
      "step": 19830
    },
    {
      "entropy": 6.256608247756958,
      "epoch": 3.869573700126817,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036123046785182855,
      "loss": 5.8441,
      "mean_token_accuracy": 0.15768484324216842,
      "num_tokens": 66271094.0,
      "step": 19835
    },
    {
      "entropy": 6.2579728126525875,
      "epoch": 3.87054921471076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003611654997642598,
      "loss": 5.8455,
      "mean_token_accuracy": 0.16408242285251617,
      "num_tokens": 66288600.0,
      "step": 19840
    },
    {
      "entropy": 6.191664934158325,
      "epoch": 3.871524729294703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003611005232569056,
      "loss": 5.6818,
      "mean_token_accuracy": 0.17102333456277846,
      "num_tokens": 66304395.0,
      "step": 19845
    },
    {
      "entropy": 6.141848707199097,
      "epoch": 3.8725002438786458,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036103553833611505,
      "loss": 5.7658,
      "mean_token_accuracy": 0.16786793917417525,
      "num_tokens": 66320129.0,
      "step": 19850
    },
    {
      "entropy": 6.269152688980102,
      "epoch": 3.873475758462589,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036097054500823847,
      "loss": 5.8186,
      "mean_token_accuracy": 0.16247932314872743,
      "num_tokens": 66337229.0,
      "step": 19855
    },
    {
      "entropy": 6.124242448806763,
      "epoch": 3.874451273046532,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036090554327962654,
      "loss": 5.6365,
      "mean_token_accuracy": 0.18473364263772965,
      "num_tokens": 66352842.0,
      "step": 19860
    },
    {
      "entropy": 6.126213216781617,
      "epoch": 3.875426787630475,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003608405331566311,
      "loss": 5.7388,
      "mean_token_accuracy": 0.16707332730293273,
      "num_tokens": 66369289.0,
      "step": 19865
    },
    {
      "entropy": 6.199384975433349,
      "epoch": 3.876402302214418,
      "grad_norm": 1.0,
      "learning_rate": 0.0003607755146456047,
      "loss": 5.8017,
      "mean_token_accuracy": 0.17692138999700546,
      "num_tokens": 66385584.0,
      "step": 19870
    },
    {
      "entropy": 6.105945873260498,
      "epoch": 3.877377816798361,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003607104877529007,
      "loss": 5.736,
      "mean_token_accuracy": 0.1762722209095955,
      "num_tokens": 66404163.0,
      "step": 19875
    },
    {
      "entropy": 6.1417535781860355,
      "epoch": 3.8783533313823044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003606454524848734,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1846093714237213,
      "num_tokens": 66420111.0,
      "step": 19880
    },
    {
      "entropy": 6.083583879470825,
      "epoch": 3.8793288459662474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003605804088478777,
      "loss": 5.6526,
      "mean_token_accuracy": 0.17951032370328904,
      "num_tokens": 66437040.0,
      "step": 19885
    },
    {
      "entropy": 6.092736148834229,
      "epoch": 3.8803043605501903,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003605153568482694,
      "loss": 5.6412,
      "mean_token_accuracy": 0.19087485522031783,
      "num_tokens": 66453985.0,
      "step": 19890
    },
    {
      "entropy": 6.161043787002564,
      "epoch": 3.8812798751341333,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003604502964924053,
      "loss": 5.7195,
      "mean_token_accuracy": 0.17693374454975128,
      "num_tokens": 66471703.0,
      "step": 19895
    },
    {
      "entropy": 6.233561658859253,
      "epoch": 3.8822553897180763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003603852277866427,
      "loss": 5.8085,
      "mean_token_accuracy": 0.16612978428602218,
      "num_tokens": 66488747.0,
      "step": 19900
    },
    {
      "entropy": 6.161090230941772,
      "epoch": 3.8832309043020192,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00036032015073734,
      "loss": 5.776,
      "mean_token_accuracy": 0.1757804423570633,
      "num_tokens": 66505858.0,
      "step": 19905
    },
    {
      "entropy": 6.198017024993897,
      "epoch": 3.884206418885962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003602550653508562,
      "loss": 5.7664,
      "mean_token_accuracy": 0.16455168128013611,
      "num_tokens": 66522157.0,
      "step": 19910
    },
    {
      "entropy": 6.188524341583252,
      "epoch": 3.885181933469905,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036018997163355136,
      "loss": 5.6707,
      "mean_token_accuracy": 0.1751059338450432,
      "num_tokens": 66537837.0,
      "step": 19915
    },
    {
      "entropy": 6.160026693344117,
      "epoch": 3.8861574480538486,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000360124869591786,
      "loss": 5.7932,
      "mean_token_accuracy": 0.1757144644856453,
      "num_tokens": 66554236.0,
      "step": 19920
    },
    {
      "entropy": 6.250861501693725,
      "epoch": 3.8871329626377915,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036005975923192183,
      "loss": 5.8298,
      "mean_token_accuracy": 0.16599425822496414,
      "num_tokens": 66569628.0,
      "step": 19925
    },
    {
      "entropy": 6.254523277282715,
      "epoch": 3.8881084772217345,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035999464056032124,
      "loss": 5.7842,
      "mean_token_accuracy": 0.17369529753923416,
      "num_tokens": 66586291.0,
      "step": 19930
    },
    {
      "entropy": 6.143277883529663,
      "epoch": 3.8890839918056774,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035992951358334717,
      "loss": 5.7058,
      "mean_token_accuracy": 0.17462162524461747,
      "num_tokens": 66602350.0,
      "step": 19935
    },
    {
      "entropy": 6.142317819595337,
      "epoch": 3.8900595063896204,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035986437830736373,
      "loss": 5.6904,
      "mean_token_accuracy": 0.17717300802469255,
      "num_tokens": 66619611.0,
      "step": 19940
    },
    {
      "entropy": 6.071225166320801,
      "epoch": 3.891035020973564,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003597992347387358,
      "loss": 5.6513,
      "mean_token_accuracy": 0.1823953405022621,
      "num_tokens": 66637133.0,
      "step": 19945
    },
    {
      "entropy": 6.025386953353882,
      "epoch": 3.8920105355575068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035973408288382877,
      "loss": 5.5758,
      "mean_token_accuracy": 0.18829991668462753,
      "num_tokens": 66655470.0,
      "step": 19950
    },
    {
      "entropy": 6.118005561828613,
      "epoch": 3.8929860501414497,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035966892274900925,
      "loss": 5.7132,
      "mean_token_accuracy": 0.17737381160259247,
      "num_tokens": 66672510.0,
      "step": 19955
    },
    {
      "entropy": 6.204590749740601,
      "epoch": 3.8939615647253927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035960375434064426,
      "loss": 5.7644,
      "mean_token_accuracy": 0.1687963753938675,
      "num_tokens": 66687829.0,
      "step": 19960
    },
    {
      "entropy": 6.1814676284790036,
      "epoch": 3.8949370793093356,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000359538577665102,
      "loss": 5.652,
      "mean_token_accuracy": 0.18026262372732163,
      "num_tokens": 66703287.0,
      "step": 19965
    },
    {
      "entropy": 6.188808917999268,
      "epoch": 3.8959125938932786,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003594733927287513,
      "loss": 5.7935,
      "mean_token_accuracy": 0.17090900242328644,
      "num_tokens": 66719573.0,
      "step": 19970
    },
    {
      "entropy": 6.256528329849243,
      "epoch": 3.8968881084772216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035940819953796154,
      "loss": 5.8261,
      "mean_token_accuracy": 0.16266619116067887,
      "num_tokens": 66735459.0,
      "step": 19975
    },
    {
      "entropy": 6.097744131088257,
      "epoch": 3.8978636230611645,
      "grad_norm": 0.859375,
      "learning_rate": 0.00035934299809910355,
      "loss": 5.6842,
      "mean_token_accuracy": 0.18049687147140503,
      "num_tokens": 66752522.0,
      "step": 19980
    },
    {
      "entropy": 6.210787391662597,
      "epoch": 3.898839137645108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003592777884185483,
      "loss": 5.8118,
      "mean_token_accuracy": 0.17204325646162033,
      "num_tokens": 66769380.0,
      "step": 19985
    },
    {
      "entropy": 6.062454748153686,
      "epoch": 3.899814652229051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003592125705026679,
      "loss": 5.5912,
      "mean_token_accuracy": 0.1884119138121605,
      "num_tokens": 66786974.0,
      "step": 19990
    },
    {
      "entropy": 6.184155035018921,
      "epoch": 3.900790166812994,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035914734435783536,
      "loss": 5.7161,
      "mean_token_accuracy": 0.17544183880090714,
      "num_tokens": 66804637.0,
      "step": 19995
    },
    {
      "entropy": 6.191705179214478,
      "epoch": 3.901765681396937,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003590821099904242,
      "loss": 5.7925,
      "mean_token_accuracy": 0.16870720237493514,
      "num_tokens": 66820473.0,
      "step": 20000
    },
    {
      "entropy": 6.111003112792969,
      "epoch": 3.9027411959808798,
      "grad_norm": 0.875,
      "learning_rate": 0.0003590168674068089,
      "loss": 5.7241,
      "mean_token_accuracy": 0.17460666596889496,
      "num_tokens": 66836629.0,
      "step": 20005
    },
    {
      "entropy": 6.204890727996826,
      "epoch": 3.903716710564823,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035895161661336474,
      "loss": 5.7746,
      "mean_token_accuracy": 0.1694620132446289,
      "num_tokens": 66853315.0,
      "step": 20010
    },
    {
      "entropy": 6.175518131256103,
      "epoch": 3.904692225148766,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035888635761646784,
      "loss": 5.7622,
      "mean_token_accuracy": 0.16846730709075927,
      "num_tokens": 66869862.0,
      "step": 20015
    },
    {
      "entropy": 6.058056020736695,
      "epoch": 3.905667739732709,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035882109042249507,
      "loss": 5.6979,
      "mean_token_accuracy": 0.17967069000005723,
      "num_tokens": 66888081.0,
      "step": 20020
    },
    {
      "entropy": 6.184330034255981,
      "epoch": 3.906643254316652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035875581503782406,
      "loss": 5.6296,
      "mean_token_accuracy": 0.17336546629667282,
      "num_tokens": 66903312.0,
      "step": 20025
    },
    {
      "entropy": 6.127900218963623,
      "epoch": 3.907618768900595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003586905314688333,
      "loss": 5.7261,
      "mean_token_accuracy": 0.1745365634560585,
      "num_tokens": 66920078.0,
      "step": 20030
    },
    {
      "entropy": 6.128669357299804,
      "epoch": 3.908594283484538,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00035862523972190216,
      "loss": 5.7363,
      "mean_token_accuracy": 0.17329704910516738,
      "num_tokens": 66938333.0,
      "step": 20035
    },
    {
      "entropy": 6.127836465835571,
      "epoch": 3.909569798068481,
      "grad_norm": 1.0,
      "learning_rate": 0.00035855993980341054,
      "loss": 5.6801,
      "mean_token_accuracy": 0.17912220656871797,
      "num_tokens": 66954857.0,
      "step": 20040
    },
    {
      "entropy": 6.193211364746094,
      "epoch": 3.910545312652424,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035849463171973945,
      "loss": 5.8286,
      "mean_token_accuracy": 0.17217591106891633,
      "num_tokens": 66971268.0,
      "step": 20045
    },
    {
      "entropy": 6.1122353076934814,
      "epoch": 3.9115208272363673,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003584293154772705,
      "loss": 5.6739,
      "mean_token_accuracy": 0.1725878745317459,
      "num_tokens": 66987666.0,
      "step": 20050
    },
    {
      "entropy": 6.236019897460937,
      "epoch": 3.9124963418203103,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003583639910823862,
      "loss": 5.7861,
      "mean_token_accuracy": 0.1710006669163704,
      "num_tokens": 67003250.0,
      "step": 20055
    },
    {
      "entropy": 6.145016431808472,
      "epoch": 3.913471856404253,
      "grad_norm": 1.0,
      "learning_rate": 0.0003582986585414699,
      "loss": 5.8063,
      "mean_token_accuracy": 0.17181132435798646,
      "num_tokens": 67020247.0,
      "step": 20060
    },
    {
      "entropy": 6.131876802444458,
      "epoch": 3.914447370988196,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00035823331786090547,
      "loss": 5.6617,
      "mean_token_accuracy": 0.1793724149465561,
      "num_tokens": 67038443.0,
      "step": 20065
    },
    {
      "entropy": 6.06286358833313,
      "epoch": 3.9154228855721396,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003581679690470779,
      "loss": 5.5836,
      "mean_token_accuracy": 0.1926560178399086,
      "num_tokens": 67056207.0,
      "step": 20070
    },
    {
      "entropy": 6.041671419143677,
      "epoch": 3.9163984001560825,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003581026121063728,
      "loss": 5.6478,
      "mean_token_accuracy": 0.1905737414956093,
      "num_tokens": 67073250.0,
      "step": 20075
    },
    {
      "entropy": 6.065449571609497,
      "epoch": 3.9173739147400255,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035803724704517665,
      "loss": 5.6385,
      "mean_token_accuracy": 0.18222409039735793,
      "num_tokens": 67091138.0,
      "step": 20080
    },
    {
      "entropy": 6.241698741912842,
      "epoch": 3.9183494293239685,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035797187386987664,
      "loss": 5.8062,
      "mean_token_accuracy": 0.16266168653964996,
      "num_tokens": 67108637.0,
      "step": 20085
    },
    {
      "entropy": 6.084622621536255,
      "epoch": 3.9193249439079114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003579064925868608,
      "loss": 5.6322,
      "mean_token_accuracy": 0.18721264302730561,
      "num_tokens": 67126984.0,
      "step": 20090
    },
    {
      "entropy": 6.245590114593506,
      "epoch": 3.9203004584918544,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035784110320251806,
      "loss": 5.8739,
      "mean_token_accuracy": 0.16404805034399034,
      "num_tokens": 67144165.0,
      "step": 20095
    },
    {
      "entropy": 6.131417512893677,
      "epoch": 3.9212759730757973,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035777570572323794,
      "loss": 5.6538,
      "mean_token_accuracy": 0.1807382419705391,
      "num_tokens": 67160798.0,
      "step": 20100
    },
    {
      "entropy": 6.144091653823852,
      "epoch": 3.9222514876597403,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035771030015541085,
      "loss": 5.6688,
      "mean_token_accuracy": 0.18014919012784958,
      "num_tokens": 67177104.0,
      "step": 20105
    },
    {
      "entropy": 6.221422433853149,
      "epoch": 3.9232270022436837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003576448865054281,
      "loss": 5.8047,
      "mean_token_accuracy": 0.17163495868444442,
      "num_tokens": 67192746.0,
      "step": 20110
    },
    {
      "entropy": 6.204389476776123,
      "epoch": 3.9242025168276267,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003575794647796815,
      "loss": 5.7581,
      "mean_token_accuracy": 0.16920051723718643,
      "num_tokens": 67208934.0,
      "step": 20115
    },
    {
      "entropy": 6.177174806594849,
      "epoch": 3.9251780314115696,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000357514034984564,
      "loss": 5.7742,
      "mean_token_accuracy": 0.1735398143529892,
      "num_tokens": 67224810.0,
      "step": 20120
    },
    {
      "entropy": 6.2173677444458,
      "epoch": 3.9261535459955126,
      "grad_norm": 1.046875,
      "learning_rate": 0.000357448597126469,
      "loss": 5.8059,
      "mean_token_accuracy": 0.16145186871290207,
      "num_tokens": 67240205.0,
      "step": 20125
    },
    {
      "entropy": 6.281117105484009,
      "epoch": 3.9271290605794555,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000357383151211791,
      "loss": 5.8122,
      "mean_token_accuracy": 0.1696818083524704,
      "num_tokens": 67256952.0,
      "step": 20130
    },
    {
      "entropy": 6.2306678771972654,
      "epoch": 3.928104575163399,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000357317697246925,
      "loss": 5.8166,
      "mean_token_accuracy": 0.16630115658044814,
      "num_tokens": 67272953.0,
      "step": 20135
    },
    {
      "entropy": 6.105654191970825,
      "epoch": 3.929080089747342,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003572522352382671,
      "loss": 5.6849,
      "mean_token_accuracy": 0.18077637553215026,
      "num_tokens": 67289024.0,
      "step": 20140
    },
    {
      "entropy": 6.147957992553711,
      "epoch": 3.930055604331285,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035718676519221396,
      "loss": 5.6853,
      "mean_token_accuracy": 0.17516709864139557,
      "num_tokens": 67306375.0,
      "step": 20145
    },
    {
      "entropy": 6.148039722442627,
      "epoch": 3.931031118915228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003571212871151629,
      "loss": 5.7033,
      "mean_token_accuracy": 0.174236723780632,
      "num_tokens": 67322565.0,
      "step": 20150
    },
    {
      "entropy": 6.124021863937378,
      "epoch": 3.932006633499171,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035705580101351247,
      "loss": 5.663,
      "mean_token_accuracy": 0.17815431356430053,
      "num_tokens": 67338901.0,
      "step": 20155
    },
    {
      "entropy": 6.211487102508545,
      "epoch": 3.9329821480831137,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003569903068936616,
      "loss": 5.8481,
      "mean_token_accuracy": 0.164604027569294,
      "num_tokens": 67354555.0,
      "step": 20160
    },
    {
      "entropy": 6.115857934951782,
      "epoch": 3.9339576626670567,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035692480476201007,
      "loss": 5.6842,
      "mean_token_accuracy": 0.18089483678340912,
      "num_tokens": 67371620.0,
      "step": 20165
    },
    {
      "entropy": 6.197987365722656,
      "epoch": 3.9349331772509997,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035685929462495866,
      "loss": 5.731,
      "mean_token_accuracy": 0.17315839231014252,
      "num_tokens": 67387336.0,
      "step": 20170
    },
    {
      "entropy": 6.205096912384033,
      "epoch": 3.935908691834943,
      "grad_norm": 1.0,
      "learning_rate": 0.00035679377648890876,
      "loss": 5.8008,
      "mean_token_accuracy": 0.16929546743631363,
      "num_tokens": 67403541.0,
      "step": 20175
    },
    {
      "entropy": 6.256868743896485,
      "epoch": 3.936884206418886,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035672825036026253,
      "loss": 5.8291,
      "mean_token_accuracy": 0.1722355529665947,
      "num_tokens": 67418777.0,
      "step": 20180
    },
    {
      "entropy": 6.185796022415161,
      "epoch": 3.937859721002829,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0003566627162454229,
      "loss": 5.7679,
      "mean_token_accuracy": 0.17150929868221282,
      "num_tokens": 67435423.0,
      "step": 20185
    },
    {
      "entropy": 6.173885440826416,
      "epoch": 3.938835235586772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003565971741507938,
      "loss": 5.7255,
      "mean_token_accuracy": 0.1689460888504982,
      "num_tokens": 67452189.0,
      "step": 20190
    },
    {
      "entropy": 6.184897804260254,
      "epoch": 3.939810750170715,
      "grad_norm": 1.0,
      "learning_rate": 0.00035653162408277953,
      "loss": 5.7528,
      "mean_token_accuracy": 0.1750512719154358,
      "num_tokens": 67470161.0,
      "step": 20195
    },
    {
      "entropy": 6.186540365219116,
      "epoch": 3.9407862647546583,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003564660660477856,
      "loss": 5.7462,
      "mean_token_accuracy": 0.17427713721990584,
      "num_tokens": 67487392.0,
      "step": 20200
    },
    {
      "entropy": 6.149524831771851,
      "epoch": 3.9417617793386013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003564005000522181,
      "loss": 5.7553,
      "mean_token_accuracy": 0.171650929749012,
      "num_tokens": 67505920.0,
      "step": 20205
    },
    {
      "entropy": 6.105236530303955,
      "epoch": 3.9427372939225442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003563349261024838,
      "loss": 5.6546,
      "mean_token_accuracy": 0.17529852390289308,
      "num_tokens": 67523040.0,
      "step": 20210
    },
    {
      "entropy": 6.205237197875976,
      "epoch": 3.943712808506487,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003562693442049904,
      "loss": 5.7379,
      "mean_token_accuracy": 0.1678093358874321,
      "num_tokens": 67538349.0,
      "step": 20215
    },
    {
      "entropy": 6.2240705490112305,
      "epoch": 3.94468832309043,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003562037543661464,
      "loss": 5.8219,
      "mean_token_accuracy": 0.16585012078285216,
      "num_tokens": 67554033.0,
      "step": 20220
    },
    {
      "entropy": 6.17977819442749,
      "epoch": 3.945663837674373,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035613815659236086,
      "loss": 5.7562,
      "mean_token_accuracy": 0.1735121339559555,
      "num_tokens": 67571427.0,
      "step": 20225
    },
    {
      "entropy": 6.16907811164856,
      "epoch": 3.946639352258316,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003560725508900438,
      "loss": 5.7102,
      "mean_token_accuracy": 0.17327420860528947,
      "num_tokens": 67588667.0,
      "step": 20230
    },
    {
      "entropy": 6.236766529083252,
      "epoch": 3.947614866842259,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003560069372656061,
      "loss": 5.7477,
      "mean_token_accuracy": 0.16837377697229386,
      "num_tokens": 67604465.0,
      "step": 20235
    },
    {
      "entropy": 6.122318267822266,
      "epoch": 3.9485903814262024,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003559413157254592,
      "loss": 5.6824,
      "mean_token_accuracy": 0.18036585301160812,
      "num_tokens": 67620469.0,
      "step": 20240
    },
    {
      "entropy": 6.137354040145874,
      "epoch": 3.9495658960101454,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003558756862760153,
      "loss": 5.7202,
      "mean_token_accuracy": 0.17955804765224456,
      "num_tokens": 67636880.0,
      "step": 20245
    },
    {
      "entropy": 6.164858722686768,
      "epoch": 3.9505414105940884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003558100489236877,
      "loss": 5.7274,
      "mean_token_accuracy": 0.1770495966076851,
      "num_tokens": 67653979.0,
      "step": 20250
    },
    {
      "entropy": 6.11784725189209,
      "epoch": 3.9515169251780313,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035574440367489013,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1866457238793373,
      "num_tokens": 67672751.0,
      "step": 20255
    },
    {
      "entropy": 6.149512577056885,
      "epoch": 3.9524924397619743,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003556787505360372,
      "loss": 5.6639,
      "mean_token_accuracy": 0.17832530289888382,
      "num_tokens": 67690291.0,
      "step": 20260
    },
    {
      "entropy": 6.171418809890747,
      "epoch": 3.9534679543459177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003556130895135442,
      "loss": 5.7714,
      "mean_token_accuracy": 0.17051052004098893,
      "num_tokens": 67708362.0,
      "step": 20265
    },
    {
      "entropy": 6.254381084442139,
      "epoch": 3.9544434689298607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003555474206138275,
      "loss": 5.8081,
      "mean_token_accuracy": 0.1691737338900566,
      "num_tokens": 67725624.0,
      "step": 20270
    },
    {
      "entropy": 6.099515390396118,
      "epoch": 3.9554189835138036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035548174384330395,
      "loss": 5.6325,
      "mean_token_accuracy": 0.18449808210134505,
      "num_tokens": 67742712.0,
      "step": 20275
    },
    {
      "entropy": 6.045726299285889,
      "epoch": 3.9563944980977466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035541605920839123,
      "loss": 5.5628,
      "mean_token_accuracy": 0.18602059185504913,
      "num_tokens": 67759162.0,
      "step": 20280
    },
    {
      "entropy": 6.151945972442627,
      "epoch": 3.9573700126816895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003553503667155078,
      "loss": 5.6604,
      "mean_token_accuracy": 0.17783304452896118,
      "num_tokens": 67774206.0,
      "step": 20285
    },
    {
      "entropy": 6.201687526702881,
      "epoch": 3.9583455272656325,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035528466637107294,
      "loss": 5.8313,
      "mean_token_accuracy": 0.16381054520606994,
      "num_tokens": 67790793.0,
      "step": 20290
    },
    {
      "entropy": 6.109421968460083,
      "epoch": 3.9593210418495755,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003552189581815065,
      "loss": 5.6414,
      "mean_token_accuracy": 0.17537594735622405,
      "num_tokens": 67809224.0,
      "step": 20295
    },
    {
      "entropy": 6.1532923698425295,
      "epoch": 3.9602965564335184,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035515324215322947,
      "loss": 5.7133,
      "mean_token_accuracy": 0.16980111449956894,
      "num_tokens": 67825976.0,
      "step": 20300
    },
    {
      "entropy": 6.156664848327637,
      "epoch": 3.961272071017462,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003550875182926633,
      "loss": 5.6792,
      "mean_token_accuracy": 0.17345758974552156,
      "num_tokens": 67843215.0,
      "step": 20305
    },
    {
      "entropy": 6.160940027236938,
      "epoch": 3.9622475856014048,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003550217866062303,
      "loss": 5.7585,
      "mean_token_accuracy": 0.17232984602451323,
      "num_tokens": 67861242.0,
      "step": 20310
    },
    {
      "entropy": 6.078514862060547,
      "epoch": 3.9632231001853477,
      "grad_norm": 0.875,
      "learning_rate": 0.0003549560471003535,
      "loss": 5.6843,
      "mean_token_accuracy": 0.1841524511575699,
      "num_tokens": 67877141.0,
      "step": 20315
    },
    {
      "entropy": 6.224091958999634,
      "epoch": 3.9641986147692907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035489029978145667,
      "loss": 5.7967,
      "mean_token_accuracy": 0.16613465547561646,
      "num_tokens": 67893322.0,
      "step": 20320
    },
    {
      "entropy": 6.119479942321777,
      "epoch": 3.965174129353234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003548245446559645,
      "loss": 5.6046,
      "mean_token_accuracy": 0.18102187365293504,
      "num_tokens": 67909498.0,
      "step": 20325
    },
    {
      "entropy": 6.164345979690552,
      "epoch": 3.966149643937177,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035475878173030234,
      "loss": 5.7082,
      "mean_token_accuracy": 0.17078330963850022,
      "num_tokens": 67926067.0,
      "step": 20330
    },
    {
      "entropy": 6.2148740768432615,
      "epoch": 3.96712515852112,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003546930110108963,
      "loss": 5.8046,
      "mean_token_accuracy": 0.16338644325733184,
      "num_tokens": 67941500.0,
      "step": 20335
    },
    {
      "entropy": 6.2225090026855465,
      "epoch": 3.968100673105063,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003546272325041733,
      "loss": 5.7569,
      "mean_token_accuracy": 0.16856616735458374,
      "num_tokens": 67958351.0,
      "step": 20340
    },
    {
      "entropy": 6.22070345878601,
      "epoch": 3.969076187689006,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003545614462165608,
      "loss": 5.757,
      "mean_token_accuracy": 0.17374424040317535,
      "num_tokens": 67974664.0,
      "step": 20345
    },
    {
      "entropy": 5.9679523468017575,
      "epoch": 3.970051702272949,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035449565215448756,
      "loss": 5.5326,
      "mean_token_accuracy": 0.18923618793487548,
      "num_tokens": 67992055.0,
      "step": 20350
    },
    {
      "entropy": 6.222035312652588,
      "epoch": 3.971027216856892,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035442985032438243,
      "loss": 5.7697,
      "mean_token_accuracy": 0.1671732559800148,
      "num_tokens": 68006973.0,
      "step": 20355
    },
    {
      "entropy": 6.125344848632812,
      "epoch": 3.972002731440835,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003543640407326753,
      "loss": 5.6427,
      "mean_token_accuracy": 0.18095045238733293,
      "num_tokens": 68022513.0,
      "step": 20360
    },
    {
      "entropy": 6.181301403045654,
      "epoch": 3.9729782460247782,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035429822338579696,
      "loss": 5.7034,
      "mean_token_accuracy": 0.1763250336050987,
      "num_tokens": 68039090.0,
      "step": 20365
    },
    {
      "entropy": 6.170670413970948,
      "epoch": 3.973953760608721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000354232398290179,
      "loss": 5.8013,
      "mean_token_accuracy": 0.16881915926933289,
      "num_tokens": 68055173.0,
      "step": 20370
    },
    {
      "entropy": 6.204822540283203,
      "epoch": 3.974929275192664,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003541665654522534,
      "loss": 5.7647,
      "mean_token_accuracy": 0.16795723140239716,
      "num_tokens": 68071900.0,
      "step": 20375
    },
    {
      "entropy": 6.215801239013672,
      "epoch": 3.975904789776607,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003541007248784532,
      "loss": 5.779,
      "mean_token_accuracy": 0.1691139817237854,
      "num_tokens": 68089206.0,
      "step": 20380
    },
    {
      "entropy": 6.101759338378907,
      "epoch": 3.97688030436055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003540348765752121,
      "loss": 5.631,
      "mean_token_accuracy": 0.1784880429506302,
      "num_tokens": 68107057.0,
      "step": 20385
    },
    {
      "entropy": 6.099558687210083,
      "epoch": 3.9778558189444935,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035396902054896453,
      "loss": 5.6439,
      "mean_token_accuracy": 0.18536680042743683,
      "num_tokens": 68123499.0,
      "step": 20390
    },
    {
      "entropy": 6.233685493469238,
      "epoch": 3.9788313335284364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035390315680614573,
      "loss": 5.781,
      "mean_token_accuracy": 0.1661030814051628,
      "num_tokens": 68140830.0,
      "step": 20395
    },
    {
      "entropy": 6.1660877704620365,
      "epoch": 3.9798068481123794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003538372853531916,
      "loss": 5.7721,
      "mean_token_accuracy": 0.17455434799194336,
      "num_tokens": 68156649.0,
      "step": 20400
    },
    {
      "entropy": 6.150478410720825,
      "epoch": 3.9807823626963224,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035377140619653904,
      "loss": 5.7681,
      "mean_token_accuracy": 0.17633209079504014,
      "num_tokens": 68175174.0,
      "step": 20405
    },
    {
      "entropy": 6.11451210975647,
      "epoch": 3.9817578772802653,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003537055193426253,
      "loss": 5.6184,
      "mean_token_accuracy": 0.17801520079374314,
      "num_tokens": 68191434.0,
      "step": 20410
    },
    {
      "entropy": 6.245803260803223,
      "epoch": 3.9827333918642083,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035363962479788877,
      "loss": 5.8169,
      "mean_token_accuracy": 0.1694601684808731,
      "num_tokens": 68208026.0,
      "step": 20415
    },
    {
      "entropy": 6.173400020599365,
      "epoch": 3.9837089064481512,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003535737225687684,
      "loss": 5.7077,
      "mean_token_accuracy": 0.17360395789146424,
      "num_tokens": 68224515.0,
      "step": 20420
    },
    {
      "entropy": 6.180664825439453,
      "epoch": 3.984684421032094,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035350781266170386,
      "loss": 5.7258,
      "mean_token_accuracy": 0.1713791996240616,
      "num_tokens": 68240162.0,
      "step": 20425
    },
    {
      "entropy": 6.25550856590271,
      "epoch": 3.9856599356160376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003534418950831357,
      "loss": 5.7623,
      "mean_token_accuracy": 0.17351239621639253,
      "num_tokens": 68255884.0,
      "step": 20430
    },
    {
      "entropy": 6.173336553573608,
      "epoch": 3.9866354501999806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003533759698395051,
      "loss": 5.77,
      "mean_token_accuracy": 0.17218856811523436,
      "num_tokens": 68272109.0,
      "step": 20435
    },
    {
      "entropy": 6.060657978057861,
      "epoch": 3.9876109647839235,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003533100369372541,
      "loss": 5.6102,
      "mean_token_accuracy": 0.18569861501455306,
      "num_tokens": 68289198.0,
      "step": 20440
    },
    {
      "entropy": 6.082625389099121,
      "epoch": 3.9885864793678665,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003532440963828254,
      "loss": 5.6532,
      "mean_token_accuracy": 0.17659110575914383,
      "num_tokens": 68306438.0,
      "step": 20445
    },
    {
      "entropy": 6.161987066268921,
      "epoch": 3.9895619939518094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035317814818266243,
      "loss": 5.752,
      "mean_token_accuracy": 0.17621419131755828,
      "num_tokens": 68323380.0,
      "step": 20450
    },
    {
      "entropy": 6.157684469223023,
      "epoch": 3.990537508535753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003531121923432095,
      "loss": 5.6598,
      "mean_token_accuracy": 0.18317321240901946,
      "num_tokens": 68339568.0,
      "step": 20455
    },
    {
      "entropy": 6.15892653465271,
      "epoch": 3.991513023119696,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003530462288709114,
      "loss": 5.7213,
      "mean_token_accuracy": 0.17723555862903595,
      "num_tokens": 68357375.0,
      "step": 20460
    },
    {
      "entropy": 5.981211948394775,
      "epoch": 3.9924885377036388,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035298025777221414,
      "loss": 5.5109,
      "mean_token_accuracy": 0.18969102054834366,
      "num_tokens": 68376231.0,
      "step": 20465
    },
    {
      "entropy": 5.994700288772583,
      "epoch": 3.9934640522875817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035291427905356386,
      "loss": 5.5449,
      "mean_token_accuracy": 0.18655870109796524,
      "num_tokens": 68391877.0,
      "step": 20470
    },
    {
      "entropy": 6.019798040390015,
      "epoch": 3.9944395668715247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000352848292721408,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1922982692718506,
      "num_tokens": 68409793.0,
      "step": 20475
    },
    {
      "entropy": 6.056041431427002,
      "epoch": 3.9954150814554676,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003527822987821945,
      "loss": 5.6287,
      "mean_token_accuracy": 0.17569854706525803,
      "num_tokens": 68427420.0,
      "step": 20480
    },
    {
      "entropy": 6.125441598892212,
      "epoch": 3.9963905960394106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000352716297242372,
      "loss": 5.5929,
      "mean_token_accuracy": 0.19064355343580247,
      "num_tokens": 68445907.0,
      "step": 20485
    },
    {
      "entropy": 6.245910215377807,
      "epoch": 3.9973661106233536,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035265028810838977,
      "loss": 5.8432,
      "mean_token_accuracy": 0.16442357748746872,
      "num_tokens": 68462305.0,
      "step": 20490
    },
    {
      "entropy": 6.134955549240113,
      "epoch": 3.998341625207297,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035258427138669824,
      "loss": 5.7509,
      "mean_token_accuracy": 0.1740370661020279,
      "num_tokens": 68478173.0,
      "step": 20495
    },
    {
      "entropy": 6.070293188095093,
      "epoch": 3.99931713979124,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035251824708374817,
      "loss": 5.6153,
      "mean_token_accuracy": 0.17728565633296967,
      "num_tokens": 68494480.0,
      "step": 20500
    },
    {
      "entropy": 6.108280075920953,
      "epoch": 4.000195102916789,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003524522152059914,
      "loss": 5.6113,
      "mean_token_accuracy": 0.17906073232491812,
      "num_tokens": 68507877.0,
      "step": 20505
    },
    {
      "entropy": 6.214677810668945,
      "epoch": 4.0011706175007316,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035238617575988,
      "loss": 5.6802,
      "mean_token_accuracy": 0.17341448068618776,
      "num_tokens": 68523954.0,
      "step": 20510
    },
    {
      "entropy": 6.014162063598633,
      "epoch": 4.0021461320846745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003523201287518675,
      "loss": 5.5177,
      "mean_token_accuracy": 0.1899395391345024,
      "num_tokens": 68542109.0,
      "step": 20515
    },
    {
      "entropy": 6.188377523422242,
      "epoch": 4.0031216466686175,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035225407418840757,
      "loss": 5.6782,
      "mean_token_accuracy": 0.17432186305522918,
      "num_tokens": 68558888.0,
      "step": 20520
    },
    {
      "entropy": 6.111111164093018,
      "epoch": 4.00409716125256,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035218801207595473,
      "loss": 5.6355,
      "mean_token_accuracy": 0.1817752867937088,
      "num_tokens": 68575727.0,
      "step": 20525
    },
    {
      "entropy": 6.185637044906616,
      "epoch": 4.005072675836503,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003521219424209645,
      "loss": 5.5972,
      "mean_token_accuracy": 0.18436148911714553,
      "num_tokens": 68593545.0,
      "step": 20530
    },
    {
      "entropy": 6.029224061965943,
      "epoch": 4.006048190420446,
      "grad_norm": 1.078125,
      "learning_rate": 0.000352055865229893,
      "loss": 5.554,
      "mean_token_accuracy": 0.19071259945631028,
      "num_tokens": 68611564.0,
      "step": 20535
    },
    {
      "entropy": 6.120980978012085,
      "epoch": 4.00702370500439,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035198978050919686,
      "loss": 5.6629,
      "mean_token_accuracy": 0.1691782593727112,
      "num_tokens": 68627079.0,
      "step": 20540
    },
    {
      "entropy": 6.183647871017456,
      "epoch": 4.007999219588333,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035192368826533386,
      "loss": 5.6854,
      "mean_token_accuracy": 0.17644384354352952,
      "num_tokens": 68643633.0,
      "step": 20545
    },
    {
      "entropy": 6.108878898620605,
      "epoch": 4.008974734172276,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035185758850476214,
      "loss": 5.539,
      "mean_token_accuracy": 0.19031617045402527,
      "num_tokens": 68659570.0,
      "step": 20550
    },
    {
      "entropy": 6.196605205535889,
      "epoch": 4.009950248756219,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035179148123394083,
      "loss": 5.6771,
      "mean_token_accuracy": 0.18198518604040145,
      "num_tokens": 68674724.0,
      "step": 20555
    },
    {
      "entropy": 6.2471349239349365,
      "epoch": 4.010925763340162,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003517253664593297,
      "loss": 5.7428,
      "mean_token_accuracy": 0.17193799912929536,
      "num_tokens": 68689610.0,
      "step": 20560
    },
    {
      "entropy": 6.1636412143707275,
      "epoch": 4.011901277924105,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035165924418738916,
      "loss": 5.6758,
      "mean_token_accuracy": 0.17951057255268096,
      "num_tokens": 68706647.0,
      "step": 20565
    },
    {
      "entropy": 6.089786529541016,
      "epoch": 4.012876792508048,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035159311442458047,
      "loss": 5.5826,
      "mean_token_accuracy": 0.18836510479450225,
      "num_tokens": 68722653.0,
      "step": 20570
    },
    {
      "entropy": 6.232775497436523,
      "epoch": 4.013852307091991,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035152697717736567,
      "loss": 5.6935,
      "mean_token_accuracy": 0.17708128243684768,
      "num_tokens": 68739036.0,
      "step": 20575
    },
    {
      "entropy": 6.0165845394134525,
      "epoch": 4.014827821675934,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035146083245220737,
      "loss": 5.461,
      "mean_token_accuracy": 0.1947704866528511,
      "num_tokens": 68754607.0,
      "step": 20580
    },
    {
      "entropy": 6.145127058029175,
      "epoch": 4.015803336259877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003513946802555691,
      "loss": 5.6709,
      "mean_token_accuracy": 0.17856974452733992,
      "num_tokens": 68771880.0,
      "step": 20585
    },
    {
      "entropy": 6.199996709823608,
      "epoch": 4.01677885084382,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035132852059391485,
      "loss": 5.6448,
      "mean_token_accuracy": 0.1784682884812355,
      "num_tokens": 68789321.0,
      "step": 20590
    },
    {
      "entropy": 6.132694435119629,
      "epoch": 4.017754365427763,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003512623534737097,
      "loss": 5.6032,
      "mean_token_accuracy": 0.18290592730045319,
      "num_tokens": 68805875.0,
      "step": 20595
    },
    {
      "entropy": 6.177619075775146,
      "epoch": 4.018729880011707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035119617890141914,
      "loss": 5.7017,
      "mean_token_accuracy": 0.17752931267023087,
      "num_tokens": 68821358.0,
      "step": 20600
    },
    {
      "entropy": 6.1756171703338625,
      "epoch": 4.01970539459565,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003511299968835096,
      "loss": 5.6708,
      "mean_token_accuracy": 0.1821786105632782,
      "num_tokens": 68837441.0,
      "step": 20605
    },
    {
      "entropy": 6.181574010848999,
      "epoch": 4.0206809091795925,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000351063807426448,
      "loss": 5.6755,
      "mean_token_accuracy": 0.177424256503582,
      "num_tokens": 68855122.0,
      "step": 20610
    },
    {
      "entropy": 6.143614339828491,
      "epoch": 4.0216564237635355,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035099761053670233,
      "loss": 5.6537,
      "mean_token_accuracy": 0.18061917275190353,
      "num_tokens": 68873005.0,
      "step": 20615
    },
    {
      "entropy": 6.17504849433899,
      "epoch": 4.0226319383474785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035093140622074093,
      "loss": 5.6273,
      "mean_token_accuracy": 0.17477470636367798,
      "num_tokens": 68887996.0,
      "step": 20620
    },
    {
      "entropy": 6.142250633239746,
      "epoch": 4.023607452931421,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003508651944850332,
      "loss": 5.6496,
      "mean_token_accuracy": 0.18222428411245345,
      "num_tokens": 68903696.0,
      "step": 20625
    },
    {
      "entropy": 6.1381628036499025,
      "epoch": 4.024582967515364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035079897533604905,
      "loss": 5.6537,
      "mean_token_accuracy": 0.17974578589200974,
      "num_tokens": 68920507.0,
      "step": 20630
    },
    {
      "entropy": 6.183375549316406,
      "epoch": 4.025558482099307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003507327487802591,
      "loss": 5.5871,
      "mean_token_accuracy": 0.18413725793361663,
      "num_tokens": 68937363.0,
      "step": 20635
    },
    {
      "entropy": 6.088832521438599,
      "epoch": 4.02653399668325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035066651482413495,
      "loss": 5.6021,
      "mean_token_accuracy": 0.18092174530029298,
      "num_tokens": 68954544.0,
      "step": 20640
    },
    {
      "entropy": 6.085489082336426,
      "epoch": 4.027509511267193,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003506002734741486,
      "loss": 5.5508,
      "mean_token_accuracy": 0.1907321810722351,
      "num_tokens": 68970991.0,
      "step": 20645
    },
    {
      "entropy": 6.098512029647827,
      "epoch": 4.028485025851136,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00035053402473677307,
      "loss": 5.5376,
      "mean_token_accuracy": 0.19015760570764542,
      "num_tokens": 68988438.0,
      "step": 20650
    },
    {
      "entropy": 6.111618947982788,
      "epoch": 4.029460540435079,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003504677686184817,
      "loss": 5.6101,
      "mean_token_accuracy": 0.18202681988477706,
      "num_tokens": 69004367.0,
      "step": 20655
    },
    {
      "entropy": 6.11430139541626,
      "epoch": 4.030436055019022,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0003504015051257491,
      "loss": 5.6028,
      "mean_token_accuracy": 0.18822069615125656,
      "num_tokens": 69021714.0,
      "step": 20660
    },
    {
      "entropy": 6.173597908020019,
      "epoch": 4.031411569602966,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0003503352342650501,
      "loss": 5.6685,
      "mean_token_accuracy": 0.17485352456569672,
      "num_tokens": 69037501.0,
      "step": 20665
    },
    {
      "entropy": 6.183787536621094,
      "epoch": 4.032387084186909,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003502689560428604,
      "loss": 5.6818,
      "mean_token_accuracy": 0.17252670973539352,
      "num_tokens": 69054521.0,
      "step": 20670
    },
    {
      "entropy": 6.064186143875122,
      "epoch": 4.033362598770852,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035020267046565666,
      "loss": 5.5264,
      "mean_token_accuracy": 0.18919183015823365,
      "num_tokens": 69073028.0,
      "step": 20675
    },
    {
      "entropy": 6.148317289352417,
      "epoch": 4.034338113354795,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000350136377539916,
      "loss": 5.5551,
      "mean_token_accuracy": 0.17950645089149475,
      "num_tokens": 69090670.0,
      "step": 20680
    },
    {
      "entropy": 6.125371599197388,
      "epoch": 4.035313627938738,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035007007727211635,
      "loss": 5.6488,
      "mean_token_accuracy": 0.17863638550043107,
      "num_tokens": 69106388.0,
      "step": 20685
    },
    {
      "entropy": 6.193325662612915,
      "epoch": 4.036289142522681,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035000376966873623,
      "loss": 5.6539,
      "mean_token_accuracy": 0.17704471200704575,
      "num_tokens": 69121681.0,
      "step": 20690
    },
    {
      "entropy": 6.134922027587891,
      "epoch": 4.037264657106624,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003499374547362551,
      "loss": 5.6733,
      "mean_token_accuracy": 0.1751808062195778,
      "num_tokens": 69138183.0,
      "step": 20695
    },
    {
      "entropy": 6.127878856658936,
      "epoch": 4.038240171690567,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034987113248115297,
      "loss": 5.6063,
      "mean_token_accuracy": 0.1786319226026535,
      "num_tokens": 69156739.0,
      "step": 20700
    },
    {
      "entropy": 6.2097679615020756,
      "epoch": 4.03921568627451,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034980480290991065,
      "loss": 5.6595,
      "mean_token_accuracy": 0.17434917092323304,
      "num_tokens": 69173208.0,
      "step": 20705
    },
    {
      "entropy": 6.212058877944946,
      "epoch": 4.040191200858453,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003497384660290095,
      "loss": 5.7055,
      "mean_token_accuracy": 0.17997449934482573,
      "num_tokens": 69189597.0,
      "step": 20710
    },
    {
      "entropy": 5.977896976470947,
      "epoch": 4.041166715442396,
      "grad_norm": 0.8125,
      "learning_rate": 0.0003496721218449319,
      "loss": 5.4464,
      "mean_token_accuracy": 0.19758476316928864,
      "num_tokens": 69207238.0,
      "step": 20715
    },
    {
      "entropy": 6.072599220275879,
      "epoch": 4.0421422300263385,
      "grad_norm": 0.90625,
      "learning_rate": 0.00034960577036416066,
      "loss": 5.5556,
      "mean_token_accuracy": 0.1923348382115364,
      "num_tokens": 69223760.0,
      "step": 20720
    },
    {
      "entropy": 6.205385732650757,
      "epoch": 4.0431177446102815,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00034953941159317946,
      "loss": 5.6588,
      "mean_token_accuracy": 0.17428401708602906,
      "num_tokens": 69242244.0,
      "step": 20725
    },
    {
      "entropy": 6.086724042892456,
      "epoch": 4.044093259194225,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003494730455384727,
      "loss": 5.5323,
      "mean_token_accuracy": 0.19740307927131653,
      "num_tokens": 69259637.0,
      "step": 20730
    },
    {
      "entropy": 6.0896224021911625,
      "epoch": 4.045068773778168,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034940667220652517,
      "loss": 5.5369,
      "mean_token_accuracy": 0.18830291479825972,
      "num_tokens": 69275721.0,
      "step": 20735
    },
    {
      "entropy": 6.156922769546509,
      "epoch": 4.046044288362111,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000349340291603823,
      "loss": 5.6306,
      "mean_token_accuracy": 0.18006037771701813,
      "num_tokens": 69292638.0,
      "step": 20740
    },
    {
      "entropy": 6.055719423294067,
      "epoch": 4.047019802946054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034927390373685243,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1948936477303505,
      "num_tokens": 69309070.0,
      "step": 20745
    },
    {
      "entropy": 6.083170652389526,
      "epoch": 4.047995317529997,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003492075086121006,
      "loss": 5.5615,
      "mean_token_accuracy": 0.17895137667655944,
      "num_tokens": 69326225.0,
      "step": 20750
    },
    {
      "entropy": 6.098231983184815,
      "epoch": 4.04897083211394,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003491411062360557,
      "loss": 5.5837,
      "mean_token_accuracy": 0.18715597540140153,
      "num_tokens": 69343360.0,
      "step": 20755
    },
    {
      "entropy": 6.1559075832366945,
      "epoch": 4.049946346697883,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034907469661520595,
      "loss": 5.7059,
      "mean_token_accuracy": 0.17363443821668625,
      "num_tokens": 69361143.0,
      "step": 20760
    },
    {
      "entropy": 6.143121862411499,
      "epoch": 4.050921861281826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034900827975604097,
      "loss": 5.6561,
      "mean_token_accuracy": 0.18056463450193405,
      "num_tokens": 69378039.0,
      "step": 20765
    },
    {
      "entropy": 6.096863842010498,
      "epoch": 4.051897375865769,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00034894185566505066,
      "loss": 5.5873,
      "mean_token_accuracy": 0.18796045929193497,
      "num_tokens": 69395383.0,
      "step": 20770
    },
    {
      "entropy": 6.197028255462646,
      "epoch": 4.052872890449712,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003488754243487257,
      "loss": 5.6358,
      "mean_token_accuracy": 0.18079105019569397,
      "num_tokens": 69412870.0,
      "step": 20775
    },
    {
      "entropy": 6.197768068313598,
      "epoch": 4.053848405033655,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003488089858135576,
      "loss": 5.6716,
      "mean_token_accuracy": 0.17654359489679336,
      "num_tokens": 69430510.0,
      "step": 20780
    },
    {
      "entropy": 6.108411550521851,
      "epoch": 4.054823919617598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034874254006603847,
      "loss": 5.6245,
      "mean_token_accuracy": 0.18141791075468064,
      "num_tokens": 69447381.0,
      "step": 20785
    },
    {
      "entropy": 6.227818489074707,
      "epoch": 4.055799434201541,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034867608711266107,
      "loss": 5.7312,
      "mean_token_accuracy": 0.17247421890497208,
      "num_tokens": 69464505.0,
      "step": 20790
    },
    {
      "entropy": 6.149707794189453,
      "epoch": 4.056774948785485,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034860962695991914,
      "loss": 5.5319,
      "mean_token_accuracy": 0.1858424261212349,
      "num_tokens": 69479696.0,
      "step": 20795
    },
    {
      "entropy": 5.998798656463623,
      "epoch": 4.057750463369428,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003485431596143068,
      "loss": 5.5607,
      "mean_token_accuracy": 0.18495958149433137,
      "num_tokens": 69495522.0,
      "step": 20800
    },
    {
      "entropy": 6.2026228427886965,
      "epoch": 4.058725977953371,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00034847668508231896,
      "loss": 5.7145,
      "mean_token_accuracy": 0.1747339889407158,
      "num_tokens": 69511123.0,
      "step": 20805
    },
    {
      "entropy": 6.191488027572632,
      "epoch": 4.059701492537314,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003484102033704513,
      "loss": 5.6283,
      "mean_token_accuracy": 0.18087586760520935,
      "num_tokens": 69527207.0,
      "step": 20810
    },
    {
      "entropy": 6.20435357093811,
      "epoch": 4.060677007121257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003483437144852003,
      "loss": 5.7448,
      "mean_token_accuracy": 0.17149455100297928,
      "num_tokens": 69544927.0,
      "step": 20815
    },
    {
      "entropy": 5.968436574935913,
      "epoch": 4.0616525217051995,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003482772184330629,
      "loss": 5.4149,
      "mean_token_accuracy": 0.20681223422288894,
      "num_tokens": 69561712.0,
      "step": 20820
    },
    {
      "entropy": 6.1538825035095215,
      "epoch": 4.0626280362891425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003482107152205368,
      "loss": 5.6248,
      "mean_token_accuracy": 0.1772807940840721,
      "num_tokens": 69578564.0,
      "step": 20825
    },
    {
      "entropy": 6.213495397567749,
      "epoch": 4.0636035508730854,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003481442048541206,
      "loss": 5.6593,
      "mean_token_accuracy": 0.1823177680373192,
      "num_tokens": 69594454.0,
      "step": 20830
    },
    {
      "entropy": 6.1029644966125485,
      "epoch": 4.064579065457028,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034807768734031343,
      "loss": 5.5415,
      "mean_token_accuracy": 0.1860174372792244,
      "num_tokens": 69612256.0,
      "step": 20835
    },
    {
      "entropy": 5.94959979057312,
      "epoch": 4.065554580040971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000348011162685615,
      "loss": 5.4577,
      "mean_token_accuracy": 0.20194588899612426,
      "num_tokens": 69630135.0,
      "step": 20840
    },
    {
      "entropy": 6.132473516464233,
      "epoch": 4.066530094624914,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034794463089652604,
      "loss": 5.6211,
      "mean_token_accuracy": 0.186427241563797,
      "num_tokens": 69647613.0,
      "step": 20845
    },
    {
      "entropy": 6.034402894973755,
      "epoch": 4.067505609208857,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003478780919795477,
      "loss": 5.5608,
      "mean_token_accuracy": 0.19428885132074356,
      "num_tokens": 69665884.0,
      "step": 20850
    },
    {
      "entropy": 5.994107627868653,
      "epoch": 4.068481123792801,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00034781154594118197,
      "loss": 5.5085,
      "mean_token_accuracy": 0.19382786750793457,
      "num_tokens": 69682688.0,
      "step": 20855
    },
    {
      "entropy": 6.2067827701568605,
      "epoch": 4.069456638376744,
      "grad_norm": 0.875,
      "learning_rate": 0.00034774499278793155,
      "loss": 5.7194,
      "mean_token_accuracy": 0.17667319625616074,
      "num_tokens": 69699990.0,
      "step": 20860
    },
    {
      "entropy": 6.0967950344085695,
      "epoch": 4.070432152960687,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003476784325262996,
      "loss": 5.544,
      "mean_token_accuracy": 0.18592432588338853,
      "num_tokens": 69718250.0,
      "step": 20865
    },
    {
      "entropy": 6.124383401870728,
      "epoch": 4.07140766754463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003476118651627903,
      "loss": 5.6382,
      "mean_token_accuracy": 0.18186887055635453,
      "num_tokens": 69734371.0,
      "step": 20870
    },
    {
      "entropy": 6.236646366119385,
      "epoch": 4.072383182128573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034754529070390826,
      "loss": 5.7137,
      "mean_token_accuracy": 0.17552950382232665,
      "num_tokens": 69750299.0,
      "step": 20875
    },
    {
      "entropy": 6.194391250610352,
      "epoch": 4.073358696712516,
      "grad_norm": 1.015625,
      "learning_rate": 0.000347478709156159,
      "loss": 5.7013,
      "mean_token_accuracy": 0.17276897728443147,
      "num_tokens": 69765956.0,
      "step": 20880
    },
    {
      "entropy": 6.056909847259521,
      "epoch": 4.074334211296459,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003474121205260486,
      "loss": 5.4969,
      "mean_token_accuracy": 0.18442903012037276,
      "num_tokens": 69782859.0,
      "step": 20885
    },
    {
      "entropy": 6.188111782073975,
      "epoch": 4.075309725880402,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003473455248200839,
      "loss": 5.6883,
      "mean_token_accuracy": 0.1753315806388855,
      "num_tokens": 69799378.0,
      "step": 20890
    },
    {
      "entropy": 6.07768702507019,
      "epoch": 4.076285240464345,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034727892204477237,
      "loss": 5.5625,
      "mean_token_accuracy": 0.19394800513982774,
      "num_tokens": 69816917.0,
      "step": 20895
    },
    {
      "entropy": 6.099189138412475,
      "epoch": 4.077260755048288,
      "grad_norm": 0.875,
      "learning_rate": 0.00034721231220662216,
      "loss": 5.6081,
      "mean_token_accuracy": 0.18858096301555632,
      "num_tokens": 69834095.0,
      "step": 20900
    },
    {
      "entropy": 6.199332141876221,
      "epoch": 4.078236269632231,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003471456953121421,
      "loss": 5.6742,
      "mean_token_accuracy": 0.1801038607954979,
      "num_tokens": 69850951.0,
      "step": 20905
    },
    {
      "entropy": 6.175704669952393,
      "epoch": 4.079211784216174,
      "grad_norm": 0.90625,
      "learning_rate": 0.00034707907136784195,
      "loss": 5.6264,
      "mean_token_accuracy": 0.1845935180783272,
      "num_tokens": 69867318.0,
      "step": 20910
    },
    {
      "entropy": 5.918591403961182,
      "epoch": 4.080187298800117,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003470124403802318,
      "loss": 5.3968,
      "mean_token_accuracy": 0.2084179624915123,
      "num_tokens": 69883415.0,
      "step": 20915
    },
    {
      "entropy": 6.192277574539185,
      "epoch": 4.0811628133840605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034694580235582255,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1762828126549721,
      "num_tokens": 69898731.0,
      "step": 20920
    },
    {
      "entropy": 6.176850700378418,
      "epoch": 4.0821383279680035,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003468791573011261,
      "loss": 5.6389,
      "mean_token_accuracy": 0.1857429713010788,
      "num_tokens": 69916255.0,
      "step": 20925
    },
    {
      "entropy": 6.185256099700927,
      "epoch": 4.083113842551946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034681250522265435,
      "loss": 5.7429,
      "mean_token_accuracy": 0.17218397408723832,
      "num_tokens": 69933773.0,
      "step": 20930
    },
    {
      "entropy": 6.146423053741455,
      "epoch": 4.084089357135889,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034674584612692067,
      "loss": 5.6321,
      "mean_token_accuracy": 0.18217306733131408,
      "num_tokens": 69949529.0,
      "step": 20935
    },
    {
      "entropy": 6.212532472610474,
      "epoch": 4.085064871719832,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003466791800204386,
      "loss": 5.6878,
      "mean_token_accuracy": 0.1832932412624359,
      "num_tokens": 69965262.0,
      "step": 20940
    },
    {
      "entropy": 6.233250379562378,
      "epoch": 4.086040386303775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003466125069097225,
      "loss": 5.7141,
      "mean_token_accuracy": 0.16856715828180313,
      "num_tokens": 69981185.0,
      "step": 20945
    },
    {
      "entropy": 6.065553379058838,
      "epoch": 4.087015900887718,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003465458268012875,
      "loss": 5.6219,
      "mean_token_accuracy": 0.18438326865434645,
      "num_tokens": 69997876.0,
      "step": 20950
    },
    {
      "entropy": 6.195099353790283,
      "epoch": 4.087991415471661,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034647913970164924,
      "loss": 5.6499,
      "mean_token_accuracy": 0.1763538807630539,
      "num_tokens": 70013375.0,
      "step": 20955
    },
    {
      "entropy": 6.19053168296814,
      "epoch": 4.088966930055604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003464124456173243,
      "loss": 5.683,
      "mean_token_accuracy": 0.17902383357286453,
      "num_tokens": 70029231.0,
      "step": 20960
    },
    {
      "entropy": 6.126196670532226,
      "epoch": 4.089942444639547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003463457445548296,
      "loss": 5.6139,
      "mean_token_accuracy": 0.18113370835781098,
      "num_tokens": 70045313.0,
      "step": 20965
    },
    {
      "entropy": 6.162135362625122,
      "epoch": 4.09091795922349,
      "grad_norm": 0.953125,
      "learning_rate": 0.000346279036520683,
      "loss": 5.6042,
      "mean_token_accuracy": 0.18258559852838516,
      "num_tokens": 70061599.0,
      "step": 20970
    },
    {
      "entropy": 6.120928525924683,
      "epoch": 4.091893473807433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000346212321521403,
      "loss": 5.6255,
      "mean_token_accuracy": 0.1860387831926346,
      "num_tokens": 70078573.0,
      "step": 20975
    },
    {
      "entropy": 6.0185634136199955,
      "epoch": 4.092868988391376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003461455995635088,
      "loss": 5.484,
      "mean_token_accuracy": 0.19744864851236343,
      "num_tokens": 70094845.0,
      "step": 20980
    },
    {
      "entropy": 6.205918407440185,
      "epoch": 4.09384450297532,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034607887065352024,
      "loss": 5.7068,
      "mean_token_accuracy": 0.17087829262018203,
      "num_tokens": 70112269.0,
      "step": 20985
    },
    {
      "entropy": 6.079029893875122,
      "epoch": 4.094820017559263,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034601213479795766,
      "loss": 5.5321,
      "mean_token_accuracy": 0.1929800420999527,
      "num_tokens": 70128827.0,
      "step": 20990
    },
    {
      "entropy": 6.118289995193481,
      "epoch": 4.095795532143206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003459453920033424,
      "loss": 5.6005,
      "mean_token_accuracy": 0.1822095513343811,
      "num_tokens": 70145203.0,
      "step": 20995
    },
    {
      "entropy": 6.008833742141723,
      "epoch": 4.096771046727149,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003458786422761963,
      "loss": 5.4664,
      "mean_token_accuracy": 0.19815705269575118,
      "num_tokens": 70164015.0,
      "step": 21000
    },
    {
      "epoch": 4.096771046727149,
      "eval_entropy": 5.9673730875062185,
      "eval_loss": 5.847629070281982,
      "eval_mean_token_accuracy": 0.17631852944699164,
      "eval_num_tokens": 70164015.0,
      "eval_runtime": 15.483,
      "eval_samples_per_second": 1070.661,
      "eval_steps_per_second": 133.889,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 51250,
  "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": 9.48552847488e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}