zho-hans-100mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1705199492732417,
  "eval_steps": 3000,
  "global_step": 6000,
  "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
    }
  ],
  "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": 2.7034114664448e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}