rus-cyrl-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-12000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0486760464913047,
  "eval_steps": 3000,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.896485424041748,
      "epoch": 0.0004369483527047103,
      "grad_norm": 19.125,
      "learning_rate": 2e-06,
      "loss": 14.1403,
      "mean_token_accuracy": 0.0004707288753706962,
      "num_tokens": 10478.0,
      "step": 5
    },
    {
      "entropy": 4.8918811798095705,
      "epoch": 0.0008738967054094206,
      "grad_norm": 19.125,
      "learning_rate": 4.5e-06,
      "loss": 14.1677,
      "mean_token_accuracy": 0.0009102159005124122,
      "num_tokens": 21152.0,
      "step": 10
    },
    {
      "entropy": 4.9260358810424805,
      "epoch": 0.0013108450581141308,
      "grad_norm": 21.25,
      "learning_rate": 7e-06,
      "loss": 14.0135,
      "mean_token_accuracy": 0.00043343613506294785,
      "num_tokens": 31246.0,
      "step": 15
    },
    {
      "entropy": 5.002536630630493,
      "epoch": 0.0017477934108188413,
      "grad_norm": 27.75,
      "learning_rate": 9.5e-06,
      "loss": 13.7803,
      "mean_token_accuracy": 0.00043386173201724886,
      "num_tokens": 41934.0,
      "step": 20
    },
    {
      "entropy": 5.2952552318573,
      "epoch": 0.0021847417635235517,
      "grad_norm": 33.5,
      "learning_rate": 1.2e-05,
      "loss": 13.2447,
      "mean_token_accuracy": 0.00043462926405481996,
      "num_tokens": 52830.0,
      "step": 25
    },
    {
      "entropy": 6.374371004104614,
      "epoch": 0.0026216901162282617,
      "grad_norm": 37.75,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 12.2103,
      "mean_token_accuracy": 0.0001860347867477685,
      "num_tokens": 62706.0,
      "step": 30
    },
    {
      "entropy": 9.60359296798706,
      "epoch": 0.003058638468932972,
      "grad_norm": 3.5625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.9691,
      "mean_token_accuracy": 0.0,
      "num_tokens": 72619.0,
      "step": 35
    },
    {
      "entropy": 10.6598690032959,
      "epoch": 0.0034955868216376825,
      "grad_norm": 3.984375,
      "learning_rate": 1.95e-05,
      "loss": 10.6826,
      "mean_token_accuracy": 0.0027851671096868814,
      "num_tokens": 82903.0,
      "step": 40
    },
    {
      "entropy": 10.72455415725708,
      "epoch": 0.003932535174342393,
      "grad_norm": 3.09375,
      "learning_rate": 2.2e-05,
      "loss": 10.5546,
      "mean_token_accuracy": 0.011984702153131365,
      "num_tokens": 94243.0,
      "step": 45
    },
    {
      "entropy": 10.698091506958008,
      "epoch": 0.004369483527047103,
      "grad_norm": 2.515625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.3469,
      "mean_token_accuracy": 0.017781304474920035,
      "num_tokens": 103618.0,
      "step": 50
    },
    {
      "entropy": 10.646620655059815,
      "epoch": 0.004806431879751813,
      "grad_norm": 2.09375,
      "learning_rate": 2.7e-05,
      "loss": 10.1899,
      "mean_token_accuracy": 0.01909907590597868,
      "num_tokens": 112180.0,
      "step": 55
    },
    {
      "entropy": 10.682601261138917,
      "epoch": 0.005243380232456523,
      "grad_norm": 2.296875,
      "learning_rate": 2.95e-05,
      "loss": 10.0898,
      "mean_token_accuracy": 0.01639490285888314,
      "num_tokens": 122462.0,
      "step": 60
    },
    {
      "entropy": 10.703626728057861,
      "epoch": 0.005680328585161234,
      "grad_norm": 1.890625,
      "learning_rate": 3.2e-05,
      "loss": 9.9926,
      "mean_token_accuracy": 0.029707580246031283,
      "num_tokens": 132741.0,
      "step": 65
    },
    {
      "entropy": 10.696713733673096,
      "epoch": 0.006117276937865944,
      "grad_norm": 2.0625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8566,
      "mean_token_accuracy": 0.05769682787358761,
      "num_tokens": 142973.0,
      "step": 70
    },
    {
      "entropy": 10.690417957305907,
      "epoch": 0.006554225290570655,
      "grad_norm": 1.796875,
      "learning_rate": 3.7e-05,
      "loss": 9.8804,
      "mean_token_accuracy": 0.05056782606989145,
      "num_tokens": 153199.0,
      "step": 75
    },
    {
      "entropy": 10.685645198822021,
      "epoch": 0.006991173643275365,
      "grad_norm": 1.8359375,
      "learning_rate": 3.95e-05,
      "loss": 9.768,
      "mean_token_accuracy": 0.05694139041006565,
      "num_tokens": 163079.0,
      "step": 80
    },
    {
      "entropy": 10.66970682144165,
      "epoch": 0.0074281219959800755,
      "grad_norm": 1.765625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.71,
      "mean_token_accuracy": 0.06373655460774899,
      "num_tokens": 173308.0,
      "step": 85
    },
    {
      "entropy": 10.657975387573241,
      "epoch": 0.007865070348684786,
      "grad_norm": 1.6640625,
      "learning_rate": 4.45e-05,
      "loss": 9.6122,
      "mean_token_accuracy": 0.0695813860744238,
      "num_tokens": 183856.0,
      "step": 90
    },
    {
      "entropy": 10.634114265441895,
      "epoch": 0.008302018701389496,
      "grad_norm": 1.796875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5337,
      "mean_token_accuracy": 0.060128339752554896,
      "num_tokens": 193368.0,
      "step": 95
    },
    {
      "entropy": 10.613899898529052,
      "epoch": 0.008738967054094207,
      "grad_norm": 1.75,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4681,
      "mean_token_accuracy": 0.06204190067946911,
      "num_tokens": 202686.0,
      "step": 100
    },
    {
      "entropy": 10.591835975646973,
      "epoch": 0.009175915406798917,
      "grad_norm": 1.7265625,
      "learning_rate": 5.2e-05,
      "loss": 9.3641,
      "mean_token_accuracy": 0.06793517135083675,
      "num_tokens": 213237.0,
      "step": 105
    },
    {
      "entropy": 10.56372594833374,
      "epoch": 0.009612863759503626,
      "grad_norm": 1.59375,
      "learning_rate": 5.45e-05,
      "loss": 9.3006,
      "mean_token_accuracy": 0.06861048303544522,
      "num_tokens": 222406.0,
      "step": 110
    },
    {
      "entropy": 10.523597049713135,
      "epoch": 0.010049812112208336,
      "grad_norm": 1.6640625,
      "learning_rate": 5.7e-05,
      "loss": 9.2349,
      "mean_token_accuracy": 0.06383556276559829,
      "num_tokens": 232139.0,
      "step": 115
    },
    {
      "entropy": 10.474488925933837,
      "epoch": 0.010486760464913047,
      "grad_norm": 1.40625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.1603,
      "mean_token_accuracy": 0.06192168295383453,
      "num_tokens": 242438.0,
      "step": 120
    },
    {
      "entropy": 10.428980922698974,
      "epoch": 0.010923708817617757,
      "grad_norm": 1.484375,
      "learning_rate": 6.2e-05,
      "loss": 9.0417,
      "mean_token_accuracy": 0.06691903881728649,
      "num_tokens": 253082.0,
      "step": 125
    },
    {
      "entropy": 10.344741821289062,
      "epoch": 0.011360657170322468,
      "grad_norm": 1.515625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.9769,
      "mean_token_accuracy": 0.07010326012969018,
      "num_tokens": 264067.0,
      "step": 130
    },
    {
      "entropy": 10.246838283538818,
      "epoch": 0.011797605523027178,
      "grad_norm": 1.4453125,
      "learning_rate": 6.7e-05,
      "loss": 8.8954,
      "mean_token_accuracy": 0.061506147310137746,
      "num_tokens": 274571.0,
      "step": 135
    },
    {
      "entropy": 10.181257820129394,
      "epoch": 0.012234553875731888,
      "grad_norm": 1.4921875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.7923,
      "mean_token_accuracy": 0.06518892385065556,
      "num_tokens": 284561.0,
      "step": 140
    },
    {
      "entropy": 10.101185703277588,
      "epoch": 0.012671502228436599,
      "grad_norm": 1.2421875,
      "learning_rate": 7.2e-05,
      "loss": 8.653,
      "mean_token_accuracy": 0.06580178663134575,
      "num_tokens": 295366.0,
      "step": 145
    },
    {
      "entropy": 10.015565013885498,
      "epoch": 0.01310845058114131,
      "grad_norm": 1.0625,
      "learning_rate": 7.45e-05,
      "loss": 8.6783,
      "mean_token_accuracy": 0.06620895192027092,
      "num_tokens": 305423.0,
      "step": 150
    },
    {
      "entropy": 9.899238300323486,
      "epoch": 0.01354539893384602,
      "grad_norm": 1.109375,
      "learning_rate": 7.7e-05,
      "loss": 8.5289,
      "mean_token_accuracy": 0.06539785154163838,
      "num_tokens": 315182.0,
      "step": 155
    },
    {
      "entropy": 9.770205783843995,
      "epoch": 0.01398234728655073,
      "grad_norm": 1.1171875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.4638,
      "mean_token_accuracy": 0.0697988323867321,
      "num_tokens": 324780.0,
      "step": 160
    },
    {
      "entropy": 9.595983505249023,
      "epoch": 0.01441929563925544,
      "grad_norm": 0.83984375,
      "learning_rate": 8.2e-05,
      "loss": 8.4388,
      "mean_token_accuracy": 0.06649183593690396,
      "num_tokens": 335104.0,
      "step": 165
    },
    {
      "entropy": 9.459788513183593,
      "epoch": 0.014856243991960151,
      "grad_norm": 0.9453125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.361,
      "mean_token_accuracy": 0.06445354036986828,
      "num_tokens": 345073.0,
      "step": 170
    },
    {
      "entropy": 9.294847583770752,
      "epoch": 0.015293192344664861,
      "grad_norm": 0.8984375,
      "learning_rate": 8.7e-05,
      "loss": 8.32,
      "mean_token_accuracy": 0.057173581421375276,
      "num_tokens": 354342.0,
      "step": 175
    },
    {
      "entropy": 9.100028991699219,
      "epoch": 0.015730140697369572,
      "grad_norm": 0.75390625,
      "learning_rate": 8.95e-05,
      "loss": 8.21,
      "mean_token_accuracy": 0.06767738796770573,
      "num_tokens": 364441.0,
      "step": 180
    },
    {
      "entropy": 8.956842231750489,
      "epoch": 0.01616708905007428,
      "grad_norm": 1.140625,
      "learning_rate": 9.2e-05,
      "loss": 8.2102,
      "mean_token_accuracy": 0.06345368176698685,
      "num_tokens": 373918.0,
      "step": 185
    },
    {
      "entropy": 8.87299690246582,
      "epoch": 0.016604037402778993,
      "grad_norm": 0.75390625,
      "learning_rate": 9.45e-05,
      "loss": 8.2134,
      "mean_token_accuracy": 0.061258262023329735,
      "num_tokens": 383273.0,
      "step": 190
    },
    {
      "entropy": 8.69699010848999,
      "epoch": 0.0170409857554837,
      "grad_norm": 0.8671875,
      "learning_rate": 9.7e-05,
      "loss": 8.0918,
      "mean_token_accuracy": 0.07180375047028065,
      "num_tokens": 393262.0,
      "step": 195
    },
    {
      "entropy": 8.610151290893555,
      "epoch": 0.017477934108188414,
      "grad_norm": 0.66796875,
      "learning_rate": 9.95e-05,
      "loss": 8.194,
      "mean_token_accuracy": 0.061346839740872385,
      "num_tokens": 402389.0,
      "step": 200
    },
    {
      "entropy": 8.630728244781494,
      "epoch": 0.017914882460893122,
      "grad_norm": 0.74609375,
      "learning_rate": 0.000102,
      "loss": 8.1296,
      "mean_token_accuracy": 0.06509425193071365,
      "num_tokens": 412997.0,
      "step": 205
    },
    {
      "entropy": 8.581557655334473,
      "epoch": 0.018351830813597834,
      "grad_norm": 0.6875,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.1323,
      "mean_token_accuracy": 0.06776191368699074,
      "num_tokens": 422222.0,
      "step": 210
    },
    {
      "entropy": 8.459476566314697,
      "epoch": 0.018788779166302543,
      "grad_norm": 0.66015625,
      "learning_rate": 0.000107,
      "loss": 8.0772,
      "mean_token_accuracy": 0.06964709646999837,
      "num_tokens": 432539.0,
      "step": 215
    },
    {
      "entropy": 8.583655071258544,
      "epoch": 0.019225727519007252,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0001095,
      "loss": 8.0763,
      "mean_token_accuracy": 0.07036846987903118,
      "num_tokens": 442311.0,
      "step": 220
    },
    {
      "entropy": 8.413945198059082,
      "epoch": 0.019662675871711964,
      "grad_norm": 0.70703125,
      "learning_rate": 0.000112,
      "loss": 8.0586,
      "mean_token_accuracy": 0.06845476403832436,
      "num_tokens": 452727.0,
      "step": 225
    },
    {
      "entropy": 8.500650024414062,
      "epoch": 0.020099624224416673,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001145,
      "loss": 8.0851,
      "mean_token_accuracy": 0.07086162865161896,
      "num_tokens": 462461.0,
      "step": 230
    },
    {
      "entropy": 8.524639797210693,
      "epoch": 0.020536572577121385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.0572,
      "mean_token_accuracy": 0.07282712981104851,
      "num_tokens": 472362.0,
      "step": 235
    },
    {
      "entropy": 8.431649208068848,
      "epoch": 0.020973520929826094,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.0476,
      "mean_token_accuracy": 0.07426296398043633,
      "num_tokens": 482718.0,
      "step": 240
    },
    {
      "entropy": 8.454423236846925,
      "epoch": 0.021410469282530806,
      "grad_norm": 0.703125,
      "learning_rate": 0.000122,
      "loss": 8.0537,
      "mean_token_accuracy": 0.07146380059421062,
      "num_tokens": 492947.0,
      "step": 245
    },
    {
      "entropy": 8.405343437194825,
      "epoch": 0.021847417635235514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0001245,
      "loss": 8.0819,
      "mean_token_accuracy": 0.06912912614643574,
      "num_tokens": 502957.0,
      "step": 250
    },
    {
      "entropy": 8.408589458465576,
      "epoch": 0.022284365987940227,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000127,
      "loss": 8.037,
      "mean_token_accuracy": 0.07069992832839489,
      "num_tokens": 513220.0,
      "step": 255
    },
    {
      "entropy": 8.416099166870117,
      "epoch": 0.022721314340644935,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0001295,
      "loss": 7.9798,
      "mean_token_accuracy": 0.07073883488774299,
      "num_tokens": 523217.0,
      "step": 260
    },
    {
      "entropy": 8.420017719268799,
      "epoch": 0.023158262693349647,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000132,
      "loss": 8.0086,
      "mean_token_accuracy": 0.06777435280382633,
      "num_tokens": 533789.0,
      "step": 265
    },
    {
      "entropy": 8.379833126068116,
      "epoch": 0.023595211046054356,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.9839,
      "mean_token_accuracy": 0.07827216312289238,
      "num_tokens": 544146.0,
      "step": 270
    },
    {
      "entropy": 8.337002563476563,
      "epoch": 0.02403215939875907,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.0224,
      "mean_token_accuracy": 0.06755217313766479,
      "num_tokens": 554215.0,
      "step": 275
    },
    {
      "entropy": 8.356577014923095,
      "epoch": 0.024469107751463777,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0001395,
      "loss": 7.9919,
      "mean_token_accuracy": 0.0684168841689825,
      "num_tokens": 564587.0,
      "step": 280
    },
    {
      "entropy": 8.399262714385987,
      "epoch": 0.024906056104168486,
      "grad_norm": 0.8125,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.0404,
      "mean_token_accuracy": 0.07138842865824699,
      "num_tokens": 574819.0,
      "step": 285
    },
    {
      "entropy": 8.372067642211913,
      "epoch": 0.025343004456873198,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001445,
      "loss": 7.9538,
      "mean_token_accuracy": 0.07292748279869557,
      "num_tokens": 585178.0,
      "step": 290
    },
    {
      "entropy": 8.315776348114014,
      "epoch": 0.025779952809577907,
      "grad_norm": 0.859375,
      "learning_rate": 0.000147,
      "loss": 7.9432,
      "mean_token_accuracy": 0.07378942035138607,
      "num_tokens": 595086.0,
      "step": 295
    },
    {
      "entropy": 8.330533981323242,
      "epoch": 0.02621690116228262,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0001495,
      "loss": 7.9452,
      "mean_token_accuracy": 0.07910944558680058,
      "num_tokens": 604942.0,
      "step": 300
    },
    {
      "entropy": 8.304435920715331,
      "epoch": 0.026653849514987327,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000152,
      "loss": 7.9291,
      "mean_token_accuracy": 0.07603786624968052,
      "num_tokens": 616082.0,
      "step": 305
    },
    {
      "entropy": 8.352412796020507,
      "epoch": 0.02709079786769204,
      "grad_norm": 0.921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.9481,
      "mean_token_accuracy": 0.07412574216723442,
      "num_tokens": 626375.0,
      "step": 310
    },
    {
      "entropy": 8.253109741210938,
      "epoch": 0.027527746220396748,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000157,
      "loss": 7.9133,
      "mean_token_accuracy": 0.07124261260032654,
      "num_tokens": 637563.0,
      "step": 315
    },
    {
      "entropy": 8.222493076324463,
      "epoch": 0.02796469457310146,
      "grad_norm": 0.859375,
      "learning_rate": 0.0001595,
      "loss": 7.9205,
      "mean_token_accuracy": 0.07289120145142078,
      "num_tokens": 647256.0,
      "step": 320
    },
    {
      "entropy": 8.32628002166748,
      "epoch": 0.02840164292580617,
      "grad_norm": 0.78125,
      "learning_rate": 0.000162,
      "loss": 7.9747,
      "mean_token_accuracy": 0.07336639389395713,
      "num_tokens": 657247.0,
      "step": 325
    },
    {
      "entropy": 8.336363315582275,
      "epoch": 0.02883859127851088,
      "grad_norm": 1.125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.8957,
      "mean_token_accuracy": 0.07113592401146888,
      "num_tokens": 667613.0,
      "step": 330
    },
    {
      "entropy": 8.263669490814209,
      "epoch": 0.02927553963121559,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.9176,
      "mean_token_accuracy": 0.07490575201809406,
      "num_tokens": 678048.0,
      "step": 335
    },
    {
      "entropy": 8.235548973083496,
      "epoch": 0.029712487983920302,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.9202,
      "mean_token_accuracy": 0.07333524562418461,
      "num_tokens": 688468.0,
      "step": 340
    },
    {
      "entropy": 8.21406421661377,
      "epoch": 0.03014943633662501,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.885,
      "mean_token_accuracy": 0.078083186596632,
      "num_tokens": 699809.0,
      "step": 345
    },
    {
      "entropy": 8.228328514099122,
      "epoch": 0.030586384689329723,
      "grad_norm": 0.78125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.8454,
      "mean_token_accuracy": 0.07926662042737007,
      "num_tokens": 709227.0,
      "step": 350
    },
    {
      "entropy": 8.25043773651123,
      "epoch": 0.03102333304203443,
      "grad_norm": 1.0,
      "learning_rate": 0.000177,
      "loss": 7.9612,
      "mean_token_accuracy": 0.07075673900544643,
      "num_tokens": 720178.0,
      "step": 355
    },
    {
      "entropy": 8.20638427734375,
      "epoch": 0.031460281394739144,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0001795,
      "loss": 7.8735,
      "mean_token_accuracy": 0.07839432805776596,
      "num_tokens": 730514.0,
      "step": 360
    },
    {
      "entropy": 8.303699684143066,
      "epoch": 0.03189722974744385,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000182,
      "loss": 7.9139,
      "mean_token_accuracy": 0.07378935068845749,
      "num_tokens": 741345.0,
      "step": 365
    },
    {
      "entropy": 8.170354652404786,
      "epoch": 0.03233417810014856,
      "grad_norm": 0.890625,
      "learning_rate": 0.0001845,
      "loss": 7.8564,
      "mean_token_accuracy": 0.07958053424954414,
      "num_tokens": 751163.0,
      "step": 370
    },
    {
      "entropy": 8.218745231628418,
      "epoch": 0.03277112645285327,
      "grad_norm": 1.21875,
      "learning_rate": 0.000187,
      "loss": 7.8577,
      "mean_token_accuracy": 0.07420928552746772,
      "num_tokens": 762090.0,
      "step": 375
    },
    {
      "entropy": 8.191785526275634,
      "epoch": 0.033208074805557986,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001895,
      "loss": 7.831,
      "mean_token_accuracy": 0.07720481157302857,
      "num_tokens": 772049.0,
      "step": 380
    },
    {
      "entropy": 8.150869464874267,
      "epoch": 0.03364502315826269,
      "grad_norm": 1.0625,
      "learning_rate": 0.000192,
      "loss": 7.7882,
      "mean_token_accuracy": 0.08157530575990676,
      "num_tokens": 781280.0,
      "step": 385
    },
    {
      "entropy": 8.15115146636963,
      "epoch": 0.0340819715109674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001945,
      "loss": 7.8737,
      "mean_token_accuracy": 0.07709059566259384,
      "num_tokens": 791642.0,
      "step": 390
    },
    {
      "entropy": 8.20585823059082,
      "epoch": 0.034518919863672115,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.8441,
      "mean_token_accuracy": 0.07621582746505737,
      "num_tokens": 801224.0,
      "step": 395
    },
    {
      "entropy": 8.199099254608154,
      "epoch": 0.03495586821637683,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.7916,
      "mean_token_accuracy": 0.07633770778775215,
      "num_tokens": 810727.0,
      "step": 400
    },
    {
      "entropy": 8.255587577819824,
      "epoch": 0.03539281656908153,
      "grad_norm": 1.046875,
      "learning_rate": 0.000202,
      "loss": 7.8705,
      "mean_token_accuracy": 0.07444445230066776,
      "num_tokens": 819943.0,
      "step": 405
    },
    {
      "entropy": 8.048962116241455,
      "epoch": 0.035829764921786245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.7977,
      "mean_token_accuracy": 0.07640674635767937,
      "num_tokens": 829474.0,
      "step": 410
    },
    {
      "entropy": 8.137162065505981,
      "epoch": 0.03626671327449096,
      "grad_norm": 1.015625,
      "learning_rate": 0.000207,
      "loss": 7.7691,
      "mean_token_accuracy": 0.0750108852982521,
      "num_tokens": 839288.0,
      "step": 415
    },
    {
      "entropy": 8.195742893218995,
      "epoch": 0.03670366162719567,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0002095,
      "loss": 7.793,
      "mean_token_accuracy": 0.07822757884860039,
      "num_tokens": 850064.0,
      "step": 420
    },
    {
      "entropy": 8.037298011779786,
      "epoch": 0.037140609979900374,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000212,
      "loss": 7.7651,
      "mean_token_accuracy": 0.08010210767388344,
      "num_tokens": 860732.0,
      "step": 425
    },
    {
      "entropy": 8.097493267059326,
      "epoch": 0.037577558332605086,
      "grad_norm": 0.875,
      "learning_rate": 0.0002145,
      "loss": 7.7724,
      "mean_token_accuracy": 0.0774189442396164,
      "num_tokens": 870076.0,
      "step": 430
    },
    {
      "entropy": 8.05912742614746,
      "epoch": 0.0380145066853098,
      "grad_norm": 0.90625,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.6706,
      "mean_token_accuracy": 0.08113809823989868,
      "num_tokens": 880201.0,
      "step": 435
    },
    {
      "entropy": 8.00949149131775,
      "epoch": 0.038451455038014504,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002195,
      "loss": 7.7294,
      "mean_token_accuracy": 0.08016021810472011,
      "num_tokens": 890344.0,
      "step": 440
    },
    {
      "entropy": 7.997669124603272,
      "epoch": 0.038888403390719216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000222,
      "loss": 7.7322,
      "mean_token_accuracy": 0.0784931242465973,
      "num_tokens": 900443.0,
      "step": 445
    },
    {
      "entropy": 8.134346008300781,
      "epoch": 0.03932535174342393,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0002245,
      "loss": 7.7529,
      "mean_token_accuracy": 0.07578973248600959,
      "num_tokens": 910284.0,
      "step": 450
    },
    {
      "entropy": 8.158567905426025,
      "epoch": 0.03976230009612864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.8343,
      "mean_token_accuracy": 0.07556049451231957,
      "num_tokens": 920986.0,
      "step": 455
    },
    {
      "entropy": 7.912718629837036,
      "epoch": 0.040199248448833345,
      "grad_norm": 0.875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.7393,
      "mean_token_accuracy": 0.08137748837471008,
      "num_tokens": 930481.0,
      "step": 460
    },
    {
      "entropy": 8.1414954662323,
      "epoch": 0.04063619680153806,
      "grad_norm": 0.875,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.764,
      "mean_token_accuracy": 0.08360744118690491,
      "num_tokens": 940554.0,
      "step": 465
    },
    {
      "entropy": 8.014440441131592,
      "epoch": 0.04107314515424277,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.6943,
      "mean_token_accuracy": 0.08052223920822144,
      "num_tokens": 950005.0,
      "step": 470
    },
    {
      "entropy": 8.032914018630981,
      "epoch": 0.04151009350694748,
      "grad_norm": 0.7265625,
      "learning_rate": 0.000237,
      "loss": 7.6964,
      "mean_token_accuracy": 0.0742336355149746,
      "num_tokens": 959414.0,
      "step": 475
    },
    {
      "entropy": 8.050236415863036,
      "epoch": 0.04194704185965219,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002395,
      "loss": 7.7366,
      "mean_token_accuracy": 0.07452398836612702,
      "num_tokens": 969352.0,
      "step": 480
    },
    {
      "entropy": 8.132762765884399,
      "epoch": 0.0423839902123569,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000242,
      "loss": 7.7524,
      "mean_token_accuracy": 0.07607904821634293,
      "num_tokens": 980294.0,
      "step": 485
    },
    {
      "entropy": 7.920047473907471,
      "epoch": 0.04282093856506161,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002445,
      "loss": 7.6372,
      "mean_token_accuracy": 0.0846704125404358,
      "num_tokens": 990453.0,
      "step": 490
    },
    {
      "entropy": 8.083022117614746,
      "epoch": 0.04325788691776632,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000247,
      "loss": 7.6619,
      "mean_token_accuracy": 0.0859624132514,
      "num_tokens": 1000509.0,
      "step": 495
    },
    {
      "entropy": 8.04277811050415,
      "epoch": 0.04369483527047103,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002495,
      "loss": 7.7908,
      "mean_token_accuracy": 0.07289138622581959,
      "num_tokens": 1011747.0,
      "step": 500
    },
    {
      "entropy": 8.003987789154053,
      "epoch": 0.04413178362317574,
      "grad_norm": 1.109375,
      "learning_rate": 0.000252,
      "loss": 7.7292,
      "mean_token_accuracy": 0.07683630660176277,
      "num_tokens": 1021511.0,
      "step": 505
    },
    {
      "entropy": 8.018040466308594,
      "epoch": 0.04456873197588045,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002545,
      "loss": 7.7397,
      "mean_token_accuracy": 0.07697788439691067,
      "num_tokens": 1032116.0,
      "step": 510
    },
    {
      "entropy": 7.920564413070679,
      "epoch": 0.04500568032858516,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000257,
      "loss": 7.6298,
      "mean_token_accuracy": 0.08404637426137924,
      "num_tokens": 1041356.0,
      "step": 515
    },
    {
      "entropy": 7.942330694198608,
      "epoch": 0.04544262868128987,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002595,
      "loss": 7.7028,
      "mean_token_accuracy": 0.07776187285780907,
      "num_tokens": 1051314.0,
      "step": 520
    },
    {
      "entropy": 8.040958166122437,
      "epoch": 0.04587957703399458,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000262,
      "loss": 7.6852,
      "mean_token_accuracy": 0.08220702111721039,
      "num_tokens": 1060857.0,
      "step": 525
    },
    {
      "entropy": 8.020647382736206,
      "epoch": 0.046316525386699295,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.6951,
      "mean_token_accuracy": 0.07845601513981819,
      "num_tokens": 1071744.0,
      "step": 530
    },
    {
      "entropy": 7.9700311660766605,
      "epoch": 0.046753473739404,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.6381,
      "mean_token_accuracy": 0.08291278034448624,
      "num_tokens": 1080777.0,
      "step": 535
    },
    {
      "entropy": 7.986575984954834,
      "epoch": 0.04719042209210871,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.6373,
      "mean_token_accuracy": 0.08483078628778458,
      "num_tokens": 1090951.0,
      "step": 540
    },
    {
      "entropy": 7.9460187435150145,
      "epoch": 0.047627370444813424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.6187,
      "mean_token_accuracy": 0.07862685099244118,
      "num_tokens": 1101667.0,
      "step": 545
    },
    {
      "entropy": 7.9729594707489015,
      "epoch": 0.04806431879751814,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002745,
      "loss": 7.7162,
      "mean_token_accuracy": 0.07553440034389496,
      "num_tokens": 1112652.0,
      "step": 550
    },
    {
      "entropy": 7.875039529800415,
      "epoch": 0.04850126715022284,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000277,
      "loss": 7.6821,
      "mean_token_accuracy": 0.08174617812037469,
      "num_tokens": 1122518.0,
      "step": 555
    },
    {
      "entropy": 8.032792949676514,
      "epoch": 0.048938215502927554,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002795,
      "loss": 7.6805,
      "mean_token_accuracy": 0.08198884315788746,
      "num_tokens": 1132867.0,
      "step": 560
    },
    {
      "entropy": 7.95278844833374,
      "epoch": 0.049375163855632266,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.7381,
      "mean_token_accuracy": 0.07585004493594169,
      "num_tokens": 1142524.0,
      "step": 565
    },
    {
      "entropy": 8.021573352813721,
      "epoch": 0.04981211220833697,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002845,
      "loss": 7.6283,
      "mean_token_accuracy": 0.08681625351309777,
      "num_tokens": 1152983.0,
      "step": 570
    },
    {
      "entropy": 7.975176620483398,
      "epoch": 0.050249060561041684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000287,
      "loss": 7.6872,
      "mean_token_accuracy": 0.08164099827408791,
      "num_tokens": 1164271.0,
      "step": 575
    },
    {
      "entropy": 7.9183494567871096,
      "epoch": 0.050686008913746396,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002895,
      "loss": 7.541,
      "mean_token_accuracy": 0.08640064373612404,
      "num_tokens": 1173567.0,
      "step": 580
    },
    {
      "entropy": 7.924293804168701,
      "epoch": 0.05112295726645111,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000292,
      "loss": 7.5932,
      "mean_token_accuracy": 0.08321595788002015,
      "num_tokens": 1183619.0,
      "step": 585
    },
    {
      "entropy": 7.873846387863159,
      "epoch": 0.05155990561915581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002945,
      "loss": 7.5453,
      "mean_token_accuracy": 0.0821442298591137,
      "num_tokens": 1193922.0,
      "step": 590
    },
    {
      "entropy": 7.94070348739624,
      "epoch": 0.051996853971860525,
      "grad_norm": 1.171875,
      "learning_rate": 0.000297,
      "loss": 7.6536,
      "mean_token_accuracy": 0.0783975936472416,
      "num_tokens": 1203665.0,
      "step": 595
    },
    {
      "entropy": 7.832107400894165,
      "epoch": 0.05243380232456524,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002995,
      "loss": 7.6423,
      "mean_token_accuracy": 0.0841694563627243,
      "num_tokens": 1214119.0,
      "step": 600
    },
    {
      "entropy": 7.960584402084351,
      "epoch": 0.05287075067726995,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000302,
      "loss": 7.6685,
      "mean_token_accuracy": 0.07720493376255036,
      "num_tokens": 1225275.0,
      "step": 605
    },
    {
      "entropy": 7.838187599182129,
      "epoch": 0.053307699029974655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003045,
      "loss": 7.5356,
      "mean_token_accuracy": 0.08203468546271324,
      "num_tokens": 1235199.0,
      "step": 610
    },
    {
      "entropy": 7.830732774734497,
      "epoch": 0.05374464738267937,
      "grad_norm": 0.84375,
      "learning_rate": 0.000307,
      "loss": 7.5585,
      "mean_token_accuracy": 0.08500108271837234,
      "num_tokens": 1245648.0,
      "step": 615
    },
    {
      "entropy": 7.852081918716431,
      "epoch": 0.05418159573538408,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003095,
      "loss": 7.4995,
      "mean_token_accuracy": 0.08836419135332108,
      "num_tokens": 1254793.0,
      "step": 620
    },
    {
      "entropy": 7.985955429077149,
      "epoch": 0.05461854408808879,
      "grad_norm": 0.96875,
      "learning_rate": 0.000312,
      "loss": 7.5573,
      "mean_token_accuracy": 0.0857199877500534,
      "num_tokens": 1264429.0,
      "step": 625
    },
    {
      "entropy": 7.8001152038574215,
      "epoch": 0.055055492440793496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003145,
      "loss": 7.5317,
      "mean_token_accuracy": 0.08314449116587638,
      "num_tokens": 1274449.0,
      "step": 630
    },
    {
      "entropy": 7.813184595108032,
      "epoch": 0.05549244079349821,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000317,
      "loss": 7.6222,
      "mean_token_accuracy": 0.07996979467570782,
      "num_tokens": 1284987.0,
      "step": 635
    },
    {
      "entropy": 7.950613355636596,
      "epoch": 0.05592938914620292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003195,
      "loss": 7.6343,
      "mean_token_accuracy": 0.07661416307091713,
      "num_tokens": 1295984.0,
      "step": 640
    },
    {
      "entropy": 7.814332628250122,
      "epoch": 0.056366337498907626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000322,
      "loss": 7.5214,
      "mean_token_accuracy": 0.08348236829042435,
      "num_tokens": 1305738.0,
      "step": 645
    },
    {
      "entropy": 7.804314756393433,
      "epoch": 0.05680328585161234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.4795,
      "mean_token_accuracy": 0.07995264530181885,
      "num_tokens": 1315691.0,
      "step": 650
    },
    {
      "entropy": 7.752275705337524,
      "epoch": 0.05724023420431705,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.4882,
      "mean_token_accuracy": 0.09246278628706932,
      "num_tokens": 1326195.0,
      "step": 655
    },
    {
      "entropy": 7.765911626815796,
      "epoch": 0.05767718255702176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.5164,
      "mean_token_accuracy": 0.08554120138287544,
      "num_tokens": 1336916.0,
      "step": 660
    },
    {
      "entropy": 7.800547170639038,
      "epoch": 0.05811413090972647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.5327,
      "mean_token_accuracy": 0.08561304733157157,
      "num_tokens": 1347386.0,
      "step": 665
    },
    {
      "entropy": 7.849845457077026,
      "epoch": 0.05855107926243118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.5301,
      "mean_token_accuracy": 0.08292968645691871,
      "num_tokens": 1357959.0,
      "step": 670
    },
    {
      "entropy": 7.800776195526123,
      "epoch": 0.05898802761513589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000337,
      "loss": 7.4917,
      "mean_token_accuracy": 0.08568408265709877,
      "num_tokens": 1368135.0,
      "step": 675
    },
    {
      "entropy": 7.8050871849060055,
      "epoch": 0.059424975967840604,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003395,
      "loss": 7.5337,
      "mean_token_accuracy": 0.08384500592947006,
      "num_tokens": 1378702.0,
      "step": 680
    },
    {
      "entropy": 7.720953702926636,
      "epoch": 0.05986192432054531,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000342,
      "loss": 7.3887,
      "mean_token_accuracy": 0.08918830305337906,
      "num_tokens": 1387902.0,
      "step": 685
    },
    {
      "entropy": 7.801330518722534,
      "epoch": 0.06029887267325002,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.4145,
      "mean_token_accuracy": 0.09191453829407692,
      "num_tokens": 1397288.0,
      "step": 690
    },
    {
      "entropy": 7.620101451873779,
      "epoch": 0.060735821025954734,
      "grad_norm": 0.96875,
      "learning_rate": 0.000347,
      "loss": 7.4673,
      "mean_token_accuracy": 0.08888877630233764,
      "num_tokens": 1407416.0,
      "step": 695
    },
    {
      "entropy": 7.842352247238159,
      "epoch": 0.061172769378659446,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003495,
      "loss": 7.5187,
      "mean_token_accuracy": 0.08510556891560554,
      "num_tokens": 1417733.0,
      "step": 700
    },
    {
      "entropy": 7.77461576461792,
      "epoch": 0.06160971773136415,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000352,
      "loss": 7.4219,
      "mean_token_accuracy": 0.0847130298614502,
      "num_tokens": 1426723.0,
      "step": 705
    },
    {
      "entropy": 7.718229103088379,
      "epoch": 0.06204666608406886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003545,
      "loss": 7.4886,
      "mean_token_accuracy": 0.08168335855007172,
      "num_tokens": 1437290.0,
      "step": 710
    },
    {
      "entropy": 7.7265349388122555,
      "epoch": 0.062483614436773575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000357,
      "loss": 7.4101,
      "mean_token_accuracy": 0.08986271098256111,
      "num_tokens": 1446756.0,
      "step": 715
    },
    {
      "entropy": 7.732719802856446,
      "epoch": 0.06292056278947829,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003595,
      "loss": 7.4766,
      "mean_token_accuracy": 0.08280604630708695,
      "num_tokens": 1457074.0,
      "step": 720
    },
    {
      "entropy": 7.706831645965576,
      "epoch": 0.06335751114218299,
      "grad_norm": 1.109375,
      "learning_rate": 0.000362,
      "loss": 7.5222,
      "mean_token_accuracy": 0.08136801272630692,
      "num_tokens": 1467785.0,
      "step": 725
    },
    {
      "entropy": 7.702896404266357,
      "epoch": 0.0637944594948877,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003645,
      "loss": 7.4124,
      "mean_token_accuracy": 0.08612973466515542,
      "num_tokens": 1477532.0,
      "step": 730
    },
    {
      "entropy": 7.657266235351562,
      "epoch": 0.06423140784759242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000367,
      "loss": 7.354,
      "mean_token_accuracy": 0.08939684703946113,
      "num_tokens": 1487102.0,
      "step": 735
    },
    {
      "entropy": 7.676237678527832,
      "epoch": 0.06466835620029712,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003695,
      "loss": 7.4367,
      "mean_token_accuracy": 0.08506104946136475,
      "num_tokens": 1497459.0,
      "step": 740
    },
    {
      "entropy": 7.6919153213500975,
      "epoch": 0.06510530455300184,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000372,
      "loss": 7.4116,
      "mean_token_accuracy": 0.08696235418319702,
      "num_tokens": 1508756.0,
      "step": 745
    },
    {
      "entropy": 7.655456924438477,
      "epoch": 0.06554225290570655,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0003745,
      "loss": 7.334,
      "mean_token_accuracy": 0.09340226426720619,
      "num_tokens": 1518735.0,
      "step": 750
    },
    {
      "entropy": 7.632836103439331,
      "epoch": 0.06597920125841125,
      "grad_norm": 1.09375,
      "learning_rate": 0.000377,
      "loss": 7.373,
      "mean_token_accuracy": 0.0900121010839939,
      "num_tokens": 1528632.0,
      "step": 755
    },
    {
      "entropy": 7.59655385017395,
      "epoch": 0.06641614961111597,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003795,
      "loss": 7.3537,
      "mean_token_accuracy": 0.09101043492555619,
      "num_tokens": 1538180.0,
      "step": 760
    },
    {
      "entropy": 7.675789928436279,
      "epoch": 0.06685309796382068,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000382,
      "loss": 7.3641,
      "mean_token_accuracy": 0.09065457805991173,
      "num_tokens": 1547788.0,
      "step": 765
    },
    {
      "entropy": 7.655756425857544,
      "epoch": 0.06729004631652538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003845,
      "loss": 7.3488,
      "mean_token_accuracy": 0.09222541823983192,
      "num_tokens": 1558008.0,
      "step": 770
    },
    {
      "entropy": 7.611498165130615,
      "epoch": 0.0677269946692301,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.3996,
      "mean_token_accuracy": 0.08548890501260757,
      "num_tokens": 1567728.0,
      "step": 775
    },
    {
      "entropy": 7.6110608100891115,
      "epoch": 0.0681639430219348,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.3381,
      "mean_token_accuracy": 0.08824502006173134,
      "num_tokens": 1578488.0,
      "step": 780
    },
    {
      "entropy": 7.646148300170898,
      "epoch": 0.06860089137463951,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.3251,
      "mean_token_accuracy": 0.09305397123098373,
      "num_tokens": 1588070.0,
      "step": 785
    },
    {
      "entropy": 7.577291822433471,
      "epoch": 0.06903783972734423,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.4018,
      "mean_token_accuracy": 0.08525708541274071,
      "num_tokens": 1598572.0,
      "step": 790
    },
    {
      "entropy": 7.725144910812378,
      "epoch": 0.06947478808004894,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.3944,
      "mean_token_accuracy": 0.0888714350759983,
      "num_tokens": 1607982.0,
      "step": 795
    },
    {
      "entropy": 7.575393867492676,
      "epoch": 0.06991173643275365,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003995,
      "loss": 7.4,
      "mean_token_accuracy": 0.09165657237172127,
      "num_tokens": 1618785.0,
      "step": 800
    },
    {
      "entropy": 7.640562582015991,
      "epoch": 0.07034868478545836,
      "grad_norm": 1.15625,
      "learning_rate": 0.000402,
      "loss": 7.3264,
      "mean_token_accuracy": 0.09215063080191613,
      "num_tokens": 1627988.0,
      "step": 805
    },
    {
      "entropy": 7.606148958206177,
      "epoch": 0.07078563313816306,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004045,
      "loss": 7.4088,
      "mean_token_accuracy": 0.08241811096668243,
      "num_tokens": 1638024.0,
      "step": 810
    },
    {
      "entropy": 7.635509967803955,
      "epoch": 0.07122258149086778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.3984,
      "mean_token_accuracy": 0.08276447355747223,
      "num_tokens": 1649215.0,
      "step": 815
    },
    {
      "entropy": 7.5351354598999025,
      "epoch": 0.07165952984357249,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004095,
      "loss": 7.2742,
      "mean_token_accuracy": 0.09341849088668823,
      "num_tokens": 1659198.0,
      "step": 820
    },
    {
      "entropy": 7.591096258163452,
      "epoch": 0.0720964781962772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412,
      "loss": 7.3213,
      "mean_token_accuracy": 0.08587326556444168,
      "num_tokens": 1668972.0,
      "step": 825
    },
    {
      "entropy": 7.505528831481934,
      "epoch": 0.07253342654898191,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004145,
      "loss": 7.3155,
      "mean_token_accuracy": 0.09277390204370022,
      "num_tokens": 1679807.0,
      "step": 830
    },
    {
      "entropy": 7.6023828983306885,
      "epoch": 0.07297037490168662,
      "grad_norm": 1.0625,
      "learning_rate": 0.000417,
      "loss": 7.3551,
      "mean_token_accuracy": 0.08633830323815346,
      "num_tokens": 1690567.0,
      "step": 835
    },
    {
      "entropy": 7.440132284164429,
      "epoch": 0.07340732325439134,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004195,
      "loss": 7.2531,
      "mean_token_accuracy": 0.09126588180661202,
      "num_tokens": 1700939.0,
      "step": 840
    },
    {
      "entropy": 7.51141676902771,
      "epoch": 0.07384427160709604,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000422,
      "loss": 7.2943,
      "mean_token_accuracy": 0.09781363755464553,
      "num_tokens": 1711181.0,
      "step": 845
    },
    {
      "entropy": 7.651186609268189,
      "epoch": 0.07428121995980075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004245,
      "loss": 7.3272,
      "mean_token_accuracy": 0.08526805341243744,
      "num_tokens": 1721392.0,
      "step": 850
    },
    {
      "entropy": 7.507527637481689,
      "epoch": 0.07471816831250547,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000427,
      "loss": 7.2344,
      "mean_token_accuracy": 0.09549994468688965,
      "num_tokens": 1731942.0,
      "step": 855
    },
    {
      "entropy": 7.518540573120117,
      "epoch": 0.07515511666521017,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004295,
      "loss": 7.3444,
      "mean_token_accuracy": 0.08721731156110764,
      "num_tokens": 1743409.0,
      "step": 860
    },
    {
      "entropy": 7.473425197601318,
      "epoch": 0.07559206501791488,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000432,
      "loss": 7.2332,
      "mean_token_accuracy": 0.09332804158329963,
      "num_tokens": 1753425.0,
      "step": 865
    },
    {
      "entropy": 7.484533405303955,
      "epoch": 0.0760290133706196,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004345,
      "loss": 7.263,
      "mean_token_accuracy": 0.09156408160924911,
      "num_tokens": 1762722.0,
      "step": 870
    },
    {
      "entropy": 7.63799090385437,
      "epoch": 0.0764659617233243,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000437,
      "loss": 7.3416,
      "mean_token_accuracy": 0.08278580531477928,
      "num_tokens": 1771895.0,
      "step": 875
    },
    {
      "entropy": 7.450203514099121,
      "epoch": 0.07690291007602901,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004395,
      "loss": 7.2387,
      "mean_token_accuracy": 0.09428637102246284,
      "num_tokens": 1782175.0,
      "step": 880
    },
    {
      "entropy": 7.548505687713623,
      "epoch": 0.07733985842873373,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000442,
      "loss": 7.3003,
      "mean_token_accuracy": 0.09130567386746406,
      "num_tokens": 1792409.0,
      "step": 885
    },
    {
      "entropy": 7.5660957336425785,
      "epoch": 0.07777680678143843,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004445,
      "loss": 7.342,
      "mean_token_accuracy": 0.08967537879943847,
      "num_tokens": 1802721.0,
      "step": 890
    },
    {
      "entropy": 7.403235387802124,
      "epoch": 0.07821375513414315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000447,
      "loss": 7.1835,
      "mean_token_accuracy": 0.09422190114855766,
      "num_tokens": 1812064.0,
      "step": 895
    },
    {
      "entropy": 7.539788246154785,
      "epoch": 0.07865070348684786,
      "grad_norm": 1.0,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.2802,
      "mean_token_accuracy": 0.0889531396329403,
      "num_tokens": 1822284.0,
      "step": 900
    },
    {
      "entropy": 7.418588542938233,
      "epoch": 0.07908765183955256,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.1734,
      "mean_token_accuracy": 0.10003943592309952,
      "num_tokens": 1832150.0,
      "step": 905
    },
    {
      "entropy": 7.502434968948364,
      "epoch": 0.07952460019225728,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.3111,
      "mean_token_accuracy": 0.08902120292186737,
      "num_tokens": 1842870.0,
      "step": 910
    },
    {
      "entropy": 7.440501403808594,
      "epoch": 0.07996154854496199,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.1386,
      "mean_token_accuracy": 0.10121746063232422,
      "num_tokens": 1852771.0,
      "step": 915
    },
    {
      "entropy": 7.499157047271728,
      "epoch": 0.08039849689766669,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.2641,
      "mean_token_accuracy": 0.08988009616732598,
      "num_tokens": 1863095.0,
      "step": 920
    },
    {
      "entropy": 7.418407726287842,
      "epoch": 0.08083544525037141,
      "grad_norm": 0.96875,
      "learning_rate": 0.000462,
      "loss": 7.2466,
      "mean_token_accuracy": 0.09539340287446976,
      "num_tokens": 1873060.0,
      "step": 925
    },
    {
      "entropy": 7.4301539897918705,
      "epoch": 0.08127239360307612,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004645,
      "loss": 7.1842,
      "mean_token_accuracy": 0.09798025488853454,
      "num_tokens": 1882574.0,
      "step": 930
    },
    {
      "entropy": 7.432036447525024,
      "epoch": 0.08170934195578082,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000467,
      "loss": 7.2438,
      "mean_token_accuracy": 0.09283970594406128,
      "num_tokens": 1892227.0,
      "step": 935
    },
    {
      "entropy": 7.447924375534058,
      "epoch": 0.08214629030848554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695,
      "loss": 7.2762,
      "mean_token_accuracy": 0.0921746976673603,
      "num_tokens": 1902868.0,
      "step": 940
    },
    {
      "entropy": 7.454946565628052,
      "epoch": 0.08258323866119024,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000472,
      "loss": 7.2095,
      "mean_token_accuracy": 0.091494170576334,
      "num_tokens": 1912781.0,
      "step": 945
    },
    {
      "entropy": 7.3900199890136715,
      "epoch": 0.08302018701389496,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004745,
      "loss": 7.2071,
      "mean_token_accuracy": 0.09347877204418183,
      "num_tokens": 1923638.0,
      "step": 950
    },
    {
      "entropy": 7.410589408874512,
      "epoch": 0.08345713536659967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000477,
      "loss": 7.1738,
      "mean_token_accuracy": 0.0997669093310833,
      "num_tokens": 1933498.0,
      "step": 955
    },
    {
      "entropy": 7.345382642745972,
      "epoch": 0.08389408371930437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004795,
      "loss": 7.1795,
      "mean_token_accuracy": 0.09802865833044053,
      "num_tokens": 1943798.0,
      "step": 960
    },
    {
      "entropy": 7.41419472694397,
      "epoch": 0.0843310320720091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 7.2289,
      "mean_token_accuracy": 0.08998453319072723,
      "num_tokens": 1954235.0,
      "step": 965
    },
    {
      "entropy": 7.420694160461426,
      "epoch": 0.0847679804247138,
      "grad_norm": 0.875,
      "learning_rate": 0.0004845,
      "loss": 7.1975,
      "mean_token_accuracy": 0.08924680277705192,
      "num_tokens": 1964421.0,
      "step": 970
    },
    {
      "entropy": 7.348307037353516,
      "epoch": 0.0852049287774185,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000487,
      "loss": 7.1912,
      "mean_token_accuracy": 0.09613144919276237,
      "num_tokens": 1975862.0,
      "step": 975
    },
    {
      "entropy": 7.43187837600708,
      "epoch": 0.08564187713012322,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004895,
      "loss": 7.1916,
      "mean_token_accuracy": 0.09355136156082153,
      "num_tokens": 1985758.0,
      "step": 980
    },
    {
      "entropy": 7.304799795150757,
      "epoch": 0.08607882548282793,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492,
      "loss": 7.1507,
      "mean_token_accuracy": 0.09453744739294052,
      "num_tokens": 1996904.0,
      "step": 985
    },
    {
      "entropy": 7.544142961502075,
      "epoch": 0.08651577383553263,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945,
      "loss": 7.2387,
      "mean_token_accuracy": 0.09486401230096816,
      "num_tokens": 2006114.0,
      "step": 990
    },
    {
      "entropy": 7.3271314144134525,
      "epoch": 0.08695272218823735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497,
      "loss": 7.187,
      "mean_token_accuracy": 0.09526639729738236,
      "num_tokens": 2016349.0,
      "step": 995
    },
    {
      "entropy": 7.326001262664795,
      "epoch": 0.08738967054094206,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995,
      "loss": 7.1294,
      "mean_token_accuracy": 0.09752104729413986,
      "num_tokens": 2026701.0,
      "step": 1000
    },
    {
      "entropy": 7.363681697845459,
      "epoch": 0.08782661889364678,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999999986190023,
      "loss": 7.1313,
      "mean_token_accuracy": 0.09394256100058555,
      "num_tokens": 2036950.0,
      "step": 1005
    },
    {
      "entropy": 7.389071273803711,
      "epoch": 0.08826356724635148,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999999930086991,
      "loss": 7.1271,
      "mean_token_accuracy": 0.09964141845703126,
      "num_tokens": 2047655.0,
      "step": 1010
    },
    {
      "entropy": 7.2618661403656,
      "epoch": 0.08870051559905619,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499999983082778,
      "loss": 7.0867,
      "mean_token_accuracy": 0.09885328710079193,
      "num_tokens": 2056800.0,
      "step": 1015
    },
    {
      "entropy": 7.379116010665894,
      "epoch": 0.0891374639517609,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999688412394,
      "loss": 7.1357,
      "mean_token_accuracy": 0.09616564437747002,
      "num_tokens": 2066862.0,
      "step": 1020
    },
    {
      "entropy": 7.303624439239502,
      "epoch": 0.08957441230446561,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999999502840836,
      "loss": 7.1615,
      "mean_token_accuracy": 0.09557585939764976,
      "num_tokens": 2076585.0,
      "step": 1025
    },
    {
      "entropy": 7.345246982574463,
      "epoch": 0.09001136065717032,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999999274113107,
      "loss": 7.2304,
      "mean_token_accuracy": 0.09007581695914268,
      "num_tokens": 2087768.0,
      "step": 1030
    },
    {
      "entropy": 7.375523328781128,
      "epoch": 0.09044830900987504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999999002229214,
      "loss": 7.1331,
      "mean_token_accuracy": 0.09556613713502884,
      "num_tokens": 2097027.0,
      "step": 1035
    },
    {
      "entropy": 7.343056964874267,
      "epoch": 0.09088525736257974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999998687189161,
      "loss": 7.0493,
      "mean_token_accuracy": 0.10475894957780837,
      "num_tokens": 2107055.0,
      "step": 1040
    },
    {
      "entropy": 7.232951641082764,
      "epoch": 0.09132220571528446,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999998328992954,
      "loss": 6.9891,
      "mean_token_accuracy": 0.09808351546525955,
      "num_tokens": 2116771.0,
      "step": 1045
    },
    {
      "entropy": 7.251108074188233,
      "epoch": 0.09175915406798917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049999979276406,
      "loss": 7.1275,
      "mean_token_accuracy": 0.09823357462882995,
      "num_tokens": 2126881.0,
      "step": 1050
    },
    {
      "entropy": 7.334365940093994,
      "epoch": 0.09219610242069387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999997483132108,
      "loss": 7.1095,
      "mean_token_accuracy": 0.09823260232806205,
      "num_tokens": 2136964.0,
      "step": 1055
    },
    {
      "entropy": 7.285873460769653,
      "epoch": 0.09263305077339859,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999996995467483,
      "loss": 7.1465,
      "mean_token_accuracy": 0.09619581401348114,
      "num_tokens": 2147263.0,
      "step": 1060
    },
    {
      "entropy": 7.269978046417236,
      "epoch": 0.0930699991261033,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999996464646738,
      "loss": 7.1421,
      "mean_token_accuracy": 0.09753268361091613,
      "num_tokens": 2157474.0,
      "step": 1065
    },
    {
      "entropy": 7.301066493988037,
      "epoch": 0.093506947478808,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999995890669882,
      "loss": 7.0887,
      "mean_token_accuracy": 0.09752897918224335,
      "num_tokens": 2167701.0,
      "step": 1070
    },
    {
      "entropy": 7.1919232368469235,
      "epoch": 0.09394389583151272,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999995273536925,
      "loss": 7.0485,
      "mean_token_accuracy": 0.10060491785407066,
      "num_tokens": 2178581.0,
      "step": 1075
    },
    {
      "entropy": 7.291404294967651,
      "epoch": 0.09438084418421742,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499999461324788,
      "loss": 7.0571,
      "mean_token_accuracy": 0.0973673515021801,
      "num_tokens": 2189529.0,
      "step": 1080
    },
    {
      "entropy": 7.293972587585449,
      "epoch": 0.09481779253692213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999993909802759,
      "loss": 7.1293,
      "mean_token_accuracy": 0.09355574771761895,
      "num_tokens": 2199805.0,
      "step": 1085
    },
    {
      "entropy": 7.219992399215698,
      "epoch": 0.09525474088962685,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999993163201577,
      "loss": 7.0882,
      "mean_token_accuracy": 0.09490575790405273,
      "num_tokens": 2210108.0,
      "step": 1090
    },
    {
      "entropy": 7.3503416061401365,
      "epoch": 0.09569168924233155,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999992373444347,
      "loss": 7.0748,
      "mean_token_accuracy": 0.09794154390692711,
      "num_tokens": 2220325.0,
      "step": 1095
    },
    {
      "entropy": 7.212787294387818,
      "epoch": 0.09612863759503627,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999991540531084,
      "loss": 7.078,
      "mean_token_accuracy": 0.09457284510135651,
      "num_tokens": 2230521.0,
      "step": 1100
    },
    {
      "entropy": 7.296592044830322,
      "epoch": 0.09656558594774098,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999990664461804,
      "loss": 7.1008,
      "mean_token_accuracy": 0.0982304073870182,
      "num_tokens": 2240372.0,
      "step": 1105
    },
    {
      "entropy": 7.249142646789551,
      "epoch": 0.09700253430044568,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999989745236524,
      "loss": 7.0654,
      "mean_token_accuracy": 0.10029410496354103,
      "num_tokens": 2250522.0,
      "step": 1110
    },
    {
      "entropy": 7.172446870803833,
      "epoch": 0.0974394826531504,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999988782855261,
      "loss": 7.0139,
      "mean_token_accuracy": 0.0998832106590271,
      "num_tokens": 2260142.0,
      "step": 1115
    },
    {
      "entropy": 7.262675666809082,
      "epoch": 0.09787643100585511,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999987777318035,
      "loss": 7.0449,
      "mean_token_accuracy": 0.0934498131275177,
      "num_tokens": 2270372.0,
      "step": 1120
    },
    {
      "entropy": 7.145792722702026,
      "epoch": 0.09831337935855981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999986728624865,
      "loss": 6.9603,
      "mean_token_accuracy": 0.1059371255338192,
      "num_tokens": 2280517.0,
      "step": 1125
    },
    {
      "entropy": 7.340591526031494,
      "epoch": 0.09875032771126453,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499998563677577,
      "loss": 7.0665,
      "mean_token_accuracy": 0.097840715944767,
      "num_tokens": 2290778.0,
      "step": 1130
    },
    {
      "entropy": 7.222999715805054,
      "epoch": 0.09918727606396924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999984501770771,
      "loss": 7.1046,
      "mean_token_accuracy": 0.09563738405704499,
      "num_tokens": 2301692.0,
      "step": 1135
    },
    {
      "entropy": 7.17807970046997,
      "epoch": 0.09962422441667394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999983323609891,
      "loss": 6.9737,
      "mean_token_accuracy": 0.10515563935041428,
      "num_tokens": 2312759.0,
      "step": 1140
    },
    {
      "entropy": 7.259163951873779,
      "epoch": 0.10006117276937866,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999982102293153,
      "loss": 6.9729,
      "mean_token_accuracy": 0.10644988417625427,
      "num_tokens": 2322016.0,
      "step": 1145
    },
    {
      "entropy": 7.230524587631225,
      "epoch": 0.10049812112208337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999980837820578,
      "loss": 6.9835,
      "mean_token_accuracy": 0.10811527222394943,
      "num_tokens": 2332476.0,
      "step": 1150
    },
    {
      "entropy": 7.183452081680298,
      "epoch": 0.10093506947478809,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999979530192192,
      "loss": 7.0065,
      "mean_token_accuracy": 0.0977212980389595,
      "num_tokens": 2342914.0,
      "step": 1155
    },
    {
      "entropy": 7.134286832809448,
      "epoch": 0.10137201782749279,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999978179408019,
      "loss": 7.0586,
      "mean_token_accuracy": 0.09377534911036492,
      "num_tokens": 2353381.0,
      "step": 1160
    },
    {
      "entropy": 7.30219407081604,
      "epoch": 0.1018089661801975,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999976785468085,
      "loss": 7.0465,
      "mean_token_accuracy": 0.10484937131404877,
      "num_tokens": 2364207.0,
      "step": 1165
    },
    {
      "entropy": 7.110436010360718,
      "epoch": 0.10224591453290222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999975348372418,
      "loss": 6.9636,
      "mean_token_accuracy": 0.10252279341220856,
      "num_tokens": 2374985.0,
      "step": 1170
    },
    {
      "entropy": 7.178512048721314,
      "epoch": 0.10268286288560692,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999973868121044,
      "loss": 7.0417,
      "mean_token_accuracy": 0.10154772102832794,
      "num_tokens": 2385684.0,
      "step": 1175
    },
    {
      "entropy": 7.308528852462769,
      "epoch": 0.10311981123831163,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999972344713994,
      "loss": 6.9786,
      "mean_token_accuracy": 0.10362603068351746,
      "num_tokens": 2395177.0,
      "step": 1180
    },
    {
      "entropy": 7.079153251647949,
      "epoch": 0.10355675959101635,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999970778151295,
      "loss": 6.9996,
      "mean_token_accuracy": 0.1028733141720295,
      "num_tokens": 2405154.0,
      "step": 1185
    },
    {
      "entropy": 7.193738412857056,
      "epoch": 0.10399370794372105,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999969168432976,
      "loss": 6.9717,
      "mean_token_accuracy": 0.10190057530999183,
      "num_tokens": 2415818.0,
      "step": 1190
    },
    {
      "entropy": 7.1862547397613525,
      "epoch": 0.10443065629642576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999967515559071,
      "loss": 6.948,
      "mean_token_accuracy": 0.10600498467683792,
      "num_tokens": 2425072.0,
      "step": 1195
    },
    {
      "entropy": 7.127433347702026,
      "epoch": 0.10486760464913047,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499996581952961,
      "loss": 7.0112,
      "mean_token_accuracy": 0.10471701920032501,
      "num_tokens": 2435055.0,
      "step": 1200
    },
    {
      "entropy": 7.183917045593262,
      "epoch": 0.10530455300183518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999964080344625,
      "loss": 6.9105,
      "mean_token_accuracy": 0.10828652903437615,
      "num_tokens": 2444434.0,
      "step": 1205
    },
    {
      "entropy": 7.123368644714356,
      "epoch": 0.1057415013545399,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999962298004151,
      "loss": 6.9342,
      "mean_token_accuracy": 0.10267825126647949,
      "num_tokens": 2453860.0,
      "step": 1210
    },
    {
      "entropy": 7.156991624832154,
      "epoch": 0.1061784497072446,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999960472508219,
      "loss": 7.0422,
      "mean_token_accuracy": 0.09686045348644257,
      "num_tokens": 2464235.0,
      "step": 1215
    },
    {
      "entropy": 7.170539712905883,
      "epoch": 0.10661539805994931,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999958603856869,
      "loss": 6.9567,
      "mean_token_accuracy": 0.1066193401813507,
      "num_tokens": 2474561.0,
      "step": 1220
    },
    {
      "entropy": 7.07088451385498,
      "epoch": 0.10705234641265403,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999956692050132,
      "loss": 6.9471,
      "mean_token_accuracy": 0.10514729768037796,
      "num_tokens": 2484574.0,
      "step": 1225
    },
    {
      "entropy": 7.207800436019897,
      "epoch": 0.10748929476535873,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999954737088048,
      "loss": 6.9074,
      "mean_token_accuracy": 0.11164658814668656,
      "num_tokens": 2494146.0,
      "step": 1230
    },
    {
      "entropy": 7.073509597778321,
      "epoch": 0.10792624311806344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999952738970652,
      "loss": 6.9986,
      "mean_token_accuracy": 0.10540383160114289,
      "num_tokens": 2504157.0,
      "step": 1235
    },
    {
      "entropy": 7.107709217071533,
      "epoch": 0.10836319147076816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999950697697983,
      "loss": 6.9445,
      "mean_token_accuracy": 0.10252791419625282,
      "num_tokens": 2512979.0,
      "step": 1240
    },
    {
      "entropy": 7.164508867263794,
      "epoch": 0.10880013982347286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999948613270081,
      "loss": 6.9823,
      "mean_token_accuracy": 0.09834418892860412,
      "num_tokens": 2523202.0,
      "step": 1245
    },
    {
      "entropy": 7.125106716156006,
      "epoch": 0.10923708817617758,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999946485686987,
      "loss": 6.9302,
      "mean_token_accuracy": 0.1016211062669754,
      "num_tokens": 2532790.0,
      "step": 1250
    },
    {
      "entropy": 7.114097547531128,
      "epoch": 0.10967403652888229,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499994431494874,
      "loss": 6.9191,
      "mean_token_accuracy": 0.10513179227709771,
      "num_tokens": 2542337.0,
      "step": 1255
    },
    {
      "entropy": 7.029775667190552,
      "epoch": 0.11011098488158699,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999942101055381,
      "loss": 6.8665,
      "mean_token_accuracy": 0.11162453517317772,
      "num_tokens": 2552762.0,
      "step": 1260
    },
    {
      "entropy": 7.129621982574463,
      "epoch": 0.11054793323429171,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999939844006955,
      "loss": 6.9116,
      "mean_token_accuracy": 0.10495460107922554,
      "num_tokens": 2563261.0,
      "step": 1265
    },
    {
      "entropy": 7.162447929382324,
      "epoch": 0.11098488158699642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999937543803503,
      "loss": 6.8825,
      "mean_token_accuracy": 0.10844132751226425,
      "num_tokens": 2573089.0,
      "step": 1270
    },
    {
      "entropy": 7.0640336036682125,
      "epoch": 0.11142182993970112,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499993520044507,
      "loss": 6.9735,
      "mean_token_accuracy": 0.10243528857827186,
      "num_tokens": 2584358.0,
      "step": 1275
    },
    {
      "entropy": 7.16273775100708,
      "epoch": 0.11185877829240584,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999932813931702,
      "loss": 6.9331,
      "mean_token_accuracy": 0.1072621412575245,
      "num_tokens": 2594981.0,
      "step": 1280
    },
    {
      "entropy": 7.042669677734375,
      "epoch": 0.11229572664511055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999930384263441,
      "loss": 6.938,
      "mean_token_accuracy": 0.10397336781024932,
      "num_tokens": 2603508.0,
      "step": 1285
    },
    {
      "entropy": 7.144971084594727,
      "epoch": 0.11273267499781525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499992791144034,
      "loss": 6.9346,
      "mean_token_accuracy": 0.10637803673744202,
      "num_tokens": 2613666.0,
      "step": 1290
    },
    {
      "entropy": 7.241996955871582,
      "epoch": 0.11316962335051997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999925395462439,
      "loss": 6.9627,
      "mean_token_accuracy": 0.09948229417204857,
      "num_tokens": 2624441.0,
      "step": 1295
    },
    {
      "entropy": 7.0352034091949465,
      "epoch": 0.11360657170322468,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999922836329793,
      "loss": 6.923,
      "mean_token_accuracy": 0.10289243534207344,
      "num_tokens": 2635261.0,
      "step": 1300
    },
    {
      "entropy": 7.011289453506469,
      "epoch": 0.1140435200559294,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999920234042446,
      "loss": 6.8854,
      "mean_token_accuracy": 0.11003832370042801,
      "num_tokens": 2645493.0,
      "step": 1305
    },
    {
      "entropy": 7.1208052158355715,
      "epoch": 0.1144804684086341,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999917588600451,
      "loss": 6.852,
      "mean_token_accuracy": 0.11447785347700119,
      "num_tokens": 2655430.0,
      "step": 1310
    },
    {
      "entropy": 7.098808240890503,
      "epoch": 0.1149174167613388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999914900003857,
      "loss": 6.9128,
      "mean_token_accuracy": 0.10278257057070732,
      "num_tokens": 2665675.0,
      "step": 1315
    },
    {
      "entropy": 7.085298490524292,
      "epoch": 0.11535436511404352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999912168252717,
      "loss": 6.8695,
      "mean_token_accuracy": 0.10640941336750984,
      "num_tokens": 2675962.0,
      "step": 1320
    },
    {
      "entropy": 7.03753924369812,
      "epoch": 0.11579131346674823,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999909393347083,
      "loss": 6.9073,
      "mean_token_accuracy": 0.10800022035837173,
      "num_tokens": 2687015.0,
      "step": 1325
    },
    {
      "entropy": 7.064095830917358,
      "epoch": 0.11622826181945294,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999906575287006,
      "loss": 6.9613,
      "mean_token_accuracy": 0.10142027139663697,
      "num_tokens": 2698065.0,
      "step": 1330
    },
    {
      "entropy": 7.07314624786377,
      "epoch": 0.11666521017215765,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999903714072544,
      "loss": 6.847,
      "mean_token_accuracy": 0.11568356677889824,
      "num_tokens": 2709046.0,
      "step": 1335
    },
    {
      "entropy": 7.1079401016235355,
      "epoch": 0.11710215852486236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999900809703749,
      "loss": 6.9191,
      "mean_token_accuracy": 0.10752326548099518,
      "num_tokens": 2719705.0,
      "step": 1340
    },
    {
      "entropy": 7.064958286285401,
      "epoch": 0.11753910687756706,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999897862180678,
      "loss": 6.8041,
      "mean_token_accuracy": 0.11670546159148217,
      "num_tokens": 2729556.0,
      "step": 1345
    },
    {
      "entropy": 6.963649272918701,
      "epoch": 0.11797605523027178,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999894871503386,
      "loss": 6.9142,
      "mean_token_accuracy": 0.10688972845673561,
      "num_tokens": 2741144.0,
      "step": 1350
    },
    {
      "entropy": 7.1297893047332765,
      "epoch": 0.11841300358297649,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999891837671932,
      "loss": 6.8313,
      "mean_token_accuracy": 0.11042980030179024,
      "num_tokens": 2751384.0,
      "step": 1355
    },
    {
      "entropy": 7.002114248275757,
      "epoch": 0.11884995193568121,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999888760686375,
      "loss": 6.8575,
      "mean_token_accuracy": 0.10566901117563247,
      "num_tokens": 2761044.0,
      "step": 1360
    },
    {
      "entropy": 6.988346195220947,
      "epoch": 0.11928690028838591,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499988564054677,
      "loss": 6.8086,
      "mean_token_accuracy": 0.11115672811865807,
      "num_tokens": 2770946.0,
      "step": 1365
    },
    {
      "entropy": 7.10092191696167,
      "epoch": 0.11972384864109062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999882477253182,
      "loss": 6.8658,
      "mean_token_accuracy": 0.10557007491588592,
      "num_tokens": 2781247.0,
      "step": 1370
    },
    {
      "entropy": 6.943843507766724,
      "epoch": 0.12016079699379534,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999879270805667,
      "loss": 6.7182,
      "mean_token_accuracy": 0.11588958874344826,
      "num_tokens": 2792078.0,
      "step": 1375
    },
    {
      "entropy": 7.026294994354248,
      "epoch": 0.12059774534650004,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999876021204291,
      "loss": 6.8013,
      "mean_token_accuracy": 0.10884896218776703,
      "num_tokens": 2802458.0,
      "step": 1380
    },
    {
      "entropy": 6.9799354553222654,
      "epoch": 0.12103469369920475,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999872728449112,
      "loss": 6.9463,
      "mean_token_accuracy": 0.10275301113724708,
      "num_tokens": 2813352.0,
      "step": 1385
    },
    {
      "entropy": 7.118425130844116,
      "epoch": 0.12147164205190947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999869392540195,
      "loss": 6.8378,
      "mean_token_accuracy": 0.10523320958018303,
      "num_tokens": 2822954.0,
      "step": 1390
    },
    {
      "entropy": 7.004513549804687,
      "epoch": 0.12190859040461417,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999866013477605,
      "loss": 6.8669,
      "mean_token_accuracy": 0.10665129348635674,
      "num_tokens": 2833375.0,
      "step": 1395
    },
    {
      "entropy": 7.019338464736938,
      "epoch": 0.12234553875731889,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999862591261407,
      "loss": 6.7998,
      "mean_token_accuracy": 0.10706322714686393,
      "num_tokens": 2843103.0,
      "step": 1400
    },
    {
      "entropy": 7.052399921417236,
      "epoch": 0.1227824871100236,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999859125891664,
      "loss": 6.8627,
      "mean_token_accuracy": 0.10739990398287773,
      "num_tokens": 2853801.0,
      "step": 1405
    },
    {
      "entropy": 7.02172212600708,
      "epoch": 0.1232194354627283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999855617368445,
      "loss": 6.8554,
      "mean_token_accuracy": 0.10320612639188767,
      "num_tokens": 2863851.0,
      "step": 1410
    },
    {
      "entropy": 6.942949008941651,
      "epoch": 0.12365638381543302,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999852065691816,
      "loss": 6.7966,
      "mean_token_accuracy": 0.11276845708489418,
      "num_tokens": 2875073.0,
      "step": 1415
    },
    {
      "entropy": 7.042190361022949,
      "epoch": 0.12409333216813773,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999848470861846,
      "loss": 6.8247,
      "mean_token_accuracy": 0.11047491133213043,
      "num_tokens": 2885676.0,
      "step": 1420
    },
    {
      "entropy": 7.036431312561035,
      "epoch": 0.12453028052084243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999844832878602,
      "loss": 6.835,
      "mean_token_accuracy": 0.11439931541681289,
      "num_tokens": 2896663.0,
      "step": 1425
    },
    {
      "entropy": 6.9360435009002686,
      "epoch": 0.12496722887354715,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999841151742156,
      "loss": 6.8149,
      "mean_token_accuracy": 0.1112043671309948,
      "num_tokens": 2907555.0,
      "step": 1430
    },
    {
      "entropy": 7.092131185531616,
      "epoch": 0.12540417722625186,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999837427452578,
      "loss": 6.8264,
      "mean_token_accuracy": 0.11113670319318772,
      "num_tokens": 2917173.0,
      "step": 1435
    },
    {
      "entropy": 6.94360384941101,
      "epoch": 0.12584112557895658,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499983366000994,
      "loss": 6.8074,
      "mean_token_accuracy": 0.10906619876623154,
      "num_tokens": 2928268.0,
      "step": 1440
    },
    {
      "entropy": 7.049355125427246,
      "epoch": 0.12627807393166127,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999829849414312,
      "loss": 6.7696,
      "mean_token_accuracy": 0.11551543027162552,
      "num_tokens": 2938203.0,
      "step": 1445
    },
    {
      "entropy": 6.926217317581177,
      "epoch": 0.12671502228436599,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499982599566577,
      "loss": 6.8056,
      "mean_token_accuracy": 0.11572478860616683,
      "num_tokens": 2948094.0,
      "step": 1450
    },
    {
      "entropy": 7.031111526489258,
      "epoch": 0.1271519706370707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999822098764385,
      "loss": 6.8032,
      "mean_token_accuracy": 0.10698436871170998,
      "num_tokens": 2957865.0,
      "step": 1455
    },
    {
      "entropy": 6.903406238555908,
      "epoch": 0.1275889189897754,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999818158710234,
      "loss": 6.7279,
      "mean_token_accuracy": 0.11594445258378983,
      "num_tokens": 2967880.0,
      "step": 1460
    },
    {
      "entropy": 6.977146100997925,
      "epoch": 0.12802586734248012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999814175503392,
      "loss": 6.7898,
      "mean_token_accuracy": 0.11242105588316917,
      "num_tokens": 2979537.0,
      "step": 1465
    },
    {
      "entropy": 6.962574195861817,
      "epoch": 0.12846281569518483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999810149143935,
      "loss": 6.7715,
      "mean_token_accuracy": 0.10995072573423385,
      "num_tokens": 2989700.0,
      "step": 1470
    },
    {
      "entropy": 6.879771089553833,
      "epoch": 0.12889976404788953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499980607963194,
      "loss": 6.7658,
      "mean_token_accuracy": 0.11039941981434823,
      "num_tokens": 2999141.0,
      "step": 1475
    },
    {
      "entropy": 6.992912673950196,
      "epoch": 0.12933671240059424,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999801966967486,
      "loss": 6.7351,
      "mean_token_accuracy": 0.11658515930175781,
      "num_tokens": 3008259.0,
      "step": 1480
    },
    {
      "entropy": 7.007751703262329,
      "epoch": 0.12977366075329896,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999797811150652,
      "loss": 6.771,
      "mean_token_accuracy": 0.10813285559415817,
      "num_tokens": 3018281.0,
      "step": 1485
    },
    {
      "entropy": 6.9875898361206055,
      "epoch": 0.13021060910600368,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999793612181516,
      "loss": 6.7811,
      "mean_token_accuracy": 0.11372531205415726,
      "num_tokens": 3029062.0,
      "step": 1490
    },
    {
      "entropy": 6.8786805152893065,
      "epoch": 0.13064755745870837,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999789370060161,
      "loss": 6.7488,
      "mean_token_accuracy": 0.11443367153406143,
      "num_tokens": 3038142.0,
      "step": 1495
    },
    {
      "entropy": 7.021120643615722,
      "epoch": 0.1310845058114131,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999785084786667,
      "loss": 6.845,
      "mean_token_accuracy": 0.1017142303287983,
      "num_tokens": 3049123.0,
      "step": 1500
    },
    {
      "entropy": 6.958976125717163,
      "epoch": 0.1315214541641178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999780756361114,
      "loss": 6.7553,
      "mean_token_accuracy": 0.10809942036867141,
      "num_tokens": 3059031.0,
      "step": 1505
    },
    {
      "entropy": 6.929660081863403,
      "epoch": 0.1319584025168225,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999776384783589,
      "loss": 6.787,
      "mean_token_accuracy": 0.11125006899237633,
      "num_tokens": 3068830.0,
      "step": 1510
    },
    {
      "entropy": 6.878285837173462,
      "epoch": 0.13239535086952722,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004999771970054174,
      "loss": 6.763,
      "mean_token_accuracy": 0.11348566561937332,
      "num_tokens": 3080124.0,
      "step": 1515
    },
    {
      "entropy": 7.005362796783447,
      "epoch": 0.13283229922223194,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999767512172953,
      "loss": 6.771,
      "mean_token_accuracy": 0.10827827528119087,
      "num_tokens": 3089607.0,
      "step": 1520
    },
    {
      "entropy": 6.95645546913147,
      "epoch": 0.13326924757493663,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999763011140013,
      "loss": 6.8462,
      "mean_token_accuracy": 0.10264672636985779,
      "num_tokens": 3099363.0,
      "step": 1525
    },
    {
      "entropy": 7.003346729278564,
      "epoch": 0.13370619592764135,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999758466955439,
      "loss": 6.7483,
      "mean_token_accuracy": 0.11257248520851135,
      "num_tokens": 3109552.0,
      "step": 1530
    },
    {
      "entropy": 6.890552520751953,
      "epoch": 0.13414314428034607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999753879619319,
      "loss": 6.7293,
      "mean_token_accuracy": 0.11494478285312652,
      "num_tokens": 3119536.0,
      "step": 1535
    },
    {
      "entropy": 6.977446985244751,
      "epoch": 0.13458009263305076,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004999749249131741,
      "loss": 6.7719,
      "mean_token_accuracy": 0.11656006649136544,
      "num_tokens": 3130552.0,
      "step": 1540
    },
    {
      "entropy": 6.943555307388306,
      "epoch": 0.13501704098575548,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999744575492793,
      "loss": 6.663,
      "mean_token_accuracy": 0.11371774896979332,
      "num_tokens": 3140723.0,
      "step": 1545
    },
    {
      "entropy": 6.942564630508423,
      "epoch": 0.1354539893384602,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999739858702567,
      "loss": 6.7176,
      "mean_token_accuracy": 0.12080485373735428,
      "num_tokens": 3151031.0,
      "step": 1550
    },
    {
      "entropy": 6.856326866149902,
      "epoch": 0.1358909376911649,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499973509876115,
      "loss": 6.7141,
      "mean_token_accuracy": 0.11423594728112221,
      "num_tokens": 3161373.0,
      "step": 1555
    },
    {
      "entropy": 6.9513795375823975,
      "epoch": 0.1363278860438696,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999730295668634,
      "loss": 6.7394,
      "mean_token_accuracy": 0.11483867838978767,
      "num_tokens": 3171094.0,
      "step": 1560
    },
    {
      "entropy": 6.879623651504517,
      "epoch": 0.13676483439657433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999725449425114,
      "loss": 6.6879,
      "mean_token_accuracy": 0.11481415182352066,
      "num_tokens": 3181646.0,
      "step": 1565
    },
    {
      "entropy": 6.871405601501465,
      "epoch": 0.13720178274927902,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499972056003068,
      "loss": 6.6551,
      "mean_token_accuracy": 0.11951215118169785,
      "num_tokens": 3190727.0,
      "step": 1570
    },
    {
      "entropy": 6.925275754928589,
      "epoch": 0.13763873110198374,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999715627485426,
      "loss": 6.761,
      "mean_token_accuracy": 0.10644741803407669,
      "num_tokens": 3200316.0,
      "step": 1575
    },
    {
      "entropy": 6.952586555480957,
      "epoch": 0.13807567945468846,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499971065178945,
      "loss": 6.7275,
      "mean_token_accuracy": 0.10815547704696656,
      "num_tokens": 3211830.0,
      "step": 1580
    },
    {
      "entropy": 6.832159805297851,
      "epoch": 0.13851262780739318,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999705632942843,
      "loss": 6.739,
      "mean_token_accuracy": 0.11103120297193528,
      "num_tokens": 3222711.0,
      "step": 1585
    },
    {
      "entropy": 6.977508449554444,
      "epoch": 0.13894957616009787,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999700570945704,
      "loss": 6.657,
      "mean_token_accuracy": 0.12132718712091446,
      "num_tokens": 3232810.0,
      "step": 1590
    },
    {
      "entropy": 6.748742961883545,
      "epoch": 0.1393865245128026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999695465798128,
      "loss": 6.722,
      "mean_token_accuracy": 0.11645966917276382,
      "num_tokens": 3243366.0,
      "step": 1595
    },
    {
      "entropy": 7.011563968658447,
      "epoch": 0.1398234728655073,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999690317500215,
      "loss": 6.7095,
      "mean_token_accuracy": 0.11738921105861663,
      "num_tokens": 3254000.0,
      "step": 1600
    },
    {
      "entropy": 6.944188117980957,
      "epoch": 0.140260421218212,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999685126052063,
      "loss": 6.7427,
      "mean_token_accuracy": 0.11331361308693885,
      "num_tokens": 3263601.0,
      "step": 1605
    },
    {
      "entropy": 6.870913124084472,
      "epoch": 0.14069736957091672,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999679891453772,
      "loss": 6.7751,
      "mean_token_accuracy": 0.10785396099090576,
      "num_tokens": 3273243.0,
      "step": 1610
    },
    {
      "entropy": 6.945408201217651,
      "epoch": 0.14113431792362144,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000499967461370544,
      "loss": 6.7204,
      "mean_token_accuracy": 0.11248369067907334,
      "num_tokens": 3284208.0,
      "step": 1615
    },
    {
      "entropy": 6.906504917144775,
      "epoch": 0.14157126627632613,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999669292807172,
      "loss": 6.6823,
      "mean_token_accuracy": 0.11680415868759156,
      "num_tokens": 3294216.0,
      "step": 1620
    },
    {
      "entropy": 6.915546798706055,
      "epoch": 0.14200821462903085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999663928759066,
      "loss": 6.7472,
      "mean_token_accuracy": 0.11192744225263596,
      "num_tokens": 3304576.0,
      "step": 1625
    },
    {
      "entropy": 6.897689294815064,
      "epoch": 0.14244516298173557,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999658521561229,
      "loss": 6.7145,
      "mean_token_accuracy": 0.10784238651394844,
      "num_tokens": 3314214.0,
      "step": 1630
    },
    {
      "entropy": 6.929960298538208,
      "epoch": 0.14288211133444026,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999653071213763,
      "loss": 6.7304,
      "mean_token_accuracy": 0.10703980401158333,
      "num_tokens": 3325527.0,
      "step": 1635
    },
    {
      "entropy": 6.783003330230713,
      "epoch": 0.14331905968714498,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999647577716771,
      "loss": 6.6364,
      "mean_token_accuracy": 0.11789060086011886,
      "num_tokens": 3334511.0,
      "step": 1640
    },
    {
      "entropy": 6.935398960113526,
      "epoch": 0.1437560080398497,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999642041070361,
      "loss": 6.6406,
      "mean_token_accuracy": 0.12125914841890335,
      "num_tokens": 3343922.0,
      "step": 1645
    },
    {
      "entropy": 6.825817012786866,
      "epoch": 0.1441929563925544,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999636461274637,
      "loss": 6.6609,
      "mean_token_accuracy": 0.11183852404356003,
      "num_tokens": 3353895.0,
      "step": 1650
    },
    {
      "entropy": 6.892046880722046,
      "epoch": 0.1446299047452591,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999630838329706,
      "loss": 6.712,
      "mean_token_accuracy": 0.11429799050092697,
      "num_tokens": 3364637.0,
      "step": 1655
    },
    {
      "entropy": 6.898443794250488,
      "epoch": 0.14506685309796383,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999625172235679,
      "loss": 6.6721,
      "mean_token_accuracy": 0.11531461402773857,
      "num_tokens": 3374308.0,
      "step": 1660
    },
    {
      "entropy": 6.860412073135376,
      "epoch": 0.14550380145066852,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499961946299266,
      "loss": 6.6627,
      "mean_token_accuracy": 0.11255936473608016,
      "num_tokens": 3383637.0,
      "step": 1665
    },
    {
      "entropy": 6.8619140625,
      "epoch": 0.14594074980337324,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999613710600763,
      "loss": 6.7309,
      "mean_token_accuracy": 0.10881676152348518,
      "num_tokens": 3394150.0,
      "step": 1670
    },
    {
      "entropy": 6.886914920806885,
      "epoch": 0.14637769815607796,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999607915060094,
      "loss": 6.6871,
      "mean_token_accuracy": 0.11321992725133896,
      "num_tokens": 3404445.0,
      "step": 1675
    },
    {
      "entropy": 6.786718940734863,
      "epoch": 0.14681464650878268,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004999602076370768,
      "loss": 6.6002,
      "mean_token_accuracy": 0.12198357656598091,
      "num_tokens": 3415183.0,
      "step": 1680
    },
    {
      "entropy": 6.8183753490448,
      "epoch": 0.14725159486148737,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999596194532895,
      "loss": 6.5695,
      "mean_token_accuracy": 0.1226851724088192,
      "num_tokens": 3424824.0,
      "step": 1685
    },
    {
      "entropy": 6.875542020797729,
      "epoch": 0.1476885432141921,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999590269546588,
      "loss": 6.6749,
      "mean_token_accuracy": 0.11421856135129929,
      "num_tokens": 3434789.0,
      "step": 1690
    },
    {
      "entropy": 6.737638378143311,
      "epoch": 0.1481254915668968,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999584301411961,
      "loss": 6.5652,
      "mean_token_accuracy": 0.1369031101465225,
      "num_tokens": 3445810.0,
      "step": 1695
    },
    {
      "entropy": 6.955615425109864,
      "epoch": 0.1485624399196015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999578290129129,
      "loss": 6.7108,
      "mean_token_accuracy": 0.11079858317971229,
      "num_tokens": 3455510.0,
      "step": 1700
    },
    {
      "entropy": 6.795374155044556,
      "epoch": 0.14899938827230622,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999572235698206,
      "loss": 6.6809,
      "mean_token_accuracy": 0.11354952603578568,
      "num_tokens": 3465785.0,
      "step": 1705
    },
    {
      "entropy": 6.860705327987671,
      "epoch": 0.14943633662501093,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999566138119308,
      "loss": 6.7346,
      "mean_token_accuracy": 0.10918433368206024,
      "num_tokens": 3477568.0,
      "step": 1710
    },
    {
      "entropy": 6.796298360824585,
      "epoch": 0.14987328497771563,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999559997392554,
      "loss": 6.5705,
      "mean_token_accuracy": 0.13033040314912797,
      "num_tokens": 3487703.0,
      "step": 1715
    },
    {
      "entropy": 6.803538990020752,
      "epoch": 0.15031023333042035,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999553813518061,
      "loss": 6.6467,
      "mean_token_accuracy": 0.11564805582165719,
      "num_tokens": 3497757.0,
      "step": 1720
    },
    {
      "entropy": 6.815519523620606,
      "epoch": 0.15074718168312506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999547586495946,
      "loss": 6.6123,
      "mean_token_accuracy": 0.11909747421741486,
      "num_tokens": 3507594.0,
      "step": 1725
    },
    {
      "entropy": 6.858670568466186,
      "epoch": 0.15118413003582976,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999541316326329,
      "loss": 6.6982,
      "mean_token_accuracy": 0.11421534195542335,
      "num_tokens": 3517889.0,
      "step": 1730
    },
    {
      "entropy": 6.789118671417237,
      "epoch": 0.15162107838853447,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999535003009332,
      "loss": 6.6842,
      "mean_token_accuracy": 0.11443898230791091,
      "num_tokens": 3528158.0,
      "step": 1735
    },
    {
      "entropy": 6.85327639579773,
      "epoch": 0.1520580267412392,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999528646545074,
      "loss": 6.6617,
      "mean_token_accuracy": 0.11684061735868453,
      "num_tokens": 3539099.0,
      "step": 1740
    },
    {
      "entropy": 6.868143224716187,
      "epoch": 0.15249497509394389,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999522246933678,
      "loss": 6.6744,
      "mean_token_accuracy": 0.11591163650155067,
      "num_tokens": 3549819.0,
      "step": 1745
    },
    {
      "entropy": 6.924664878845215,
      "epoch": 0.1529319234466486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999515804175268,
      "loss": 6.6708,
      "mean_token_accuracy": 0.1160687893629074,
      "num_tokens": 3560022.0,
      "step": 1750
    },
    {
      "entropy": 6.796440887451172,
      "epoch": 0.15336887179935332,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999509318269964,
      "loss": 6.6229,
      "mean_token_accuracy": 0.11483302041888237,
      "num_tokens": 3569681.0,
      "step": 1755
    },
    {
      "entropy": 6.931779527664185,
      "epoch": 0.15380582015205801,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999502789217894,
      "loss": 6.7187,
      "mean_token_accuracy": 0.11464278921484947,
      "num_tokens": 3580731.0,
      "step": 1760
    },
    {
      "entropy": 6.735879182815552,
      "epoch": 0.15424276850476273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999496217019181,
      "loss": 6.6771,
      "mean_token_accuracy": 0.12017816901206971,
      "num_tokens": 3590482.0,
      "step": 1765
    },
    {
      "entropy": 6.846047735214233,
      "epoch": 0.15467971685746745,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999489601673952,
      "loss": 6.6588,
      "mean_token_accuracy": 0.11184544190764427,
      "num_tokens": 3600988.0,
      "step": 1770
    },
    {
      "entropy": 6.966546869277954,
      "epoch": 0.15511666521017214,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999482943182334,
      "loss": 6.7207,
      "mean_token_accuracy": 0.11034572273492813,
      "num_tokens": 3611226.0,
      "step": 1775
    },
    {
      "entropy": 6.7466967582702635,
      "epoch": 0.15555361356287686,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999476241544454,
      "loss": 6.6406,
      "mean_token_accuracy": 0.11693034246563912,
      "num_tokens": 3621324.0,
      "step": 1780
    },
    {
      "entropy": 6.859796237945557,
      "epoch": 0.15599056191558158,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999469496760441,
      "loss": 6.5745,
      "mean_token_accuracy": 0.11757055521011353,
      "num_tokens": 3631650.0,
      "step": 1785
    },
    {
      "entropy": 6.793846988677979,
      "epoch": 0.1564275102682863,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999462708830425,
      "loss": 6.5964,
      "mean_token_accuracy": 0.13014772459864615,
      "num_tokens": 3641173.0,
      "step": 1790
    },
    {
      "entropy": 6.7543669700622555,
      "epoch": 0.156864458620991,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999455877754535,
      "loss": 6.6099,
      "mean_token_accuracy": 0.11783762127161027,
      "num_tokens": 3651846.0,
      "step": 1795
    },
    {
      "entropy": 6.839160680770874,
      "epoch": 0.1573014069736957,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999449003532903,
      "loss": 6.6181,
      "mean_token_accuracy": 0.11612390056252479,
      "num_tokens": 3661174.0,
      "step": 1800
    },
    {
      "entropy": 6.767947387695313,
      "epoch": 0.15773835532640043,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999442086165661,
      "loss": 6.5785,
      "mean_token_accuracy": 0.11563654839992524,
      "num_tokens": 3671042.0,
      "step": 1805
    },
    {
      "entropy": 6.836207246780395,
      "epoch": 0.15817530367910512,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499943512565294,
      "loss": 6.5404,
      "mean_token_accuracy": 0.11653607487678527,
      "num_tokens": 3681206.0,
      "step": 1810
    },
    {
      "entropy": 6.777206325531006,
      "epoch": 0.15861225203180984,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999428121994874,
      "loss": 6.6427,
      "mean_token_accuracy": 0.11898047327995301,
      "num_tokens": 3691748.0,
      "step": 1815
    },
    {
      "entropy": 6.702393054962158,
      "epoch": 0.15904920038451456,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999421075191599,
      "loss": 6.5483,
      "mean_token_accuracy": 0.12115112468600273,
      "num_tokens": 3701730.0,
      "step": 1820
    },
    {
      "entropy": 6.887295866012574,
      "epoch": 0.15948614873721925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999413985243249,
      "loss": 6.6747,
      "mean_token_accuracy": 0.1129259705543518,
      "num_tokens": 3712525.0,
      "step": 1825
    },
    {
      "entropy": 6.863398599624634,
      "epoch": 0.15992309708992397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499940685214996,
      "loss": 6.6086,
      "mean_token_accuracy": 0.11575344502925873,
      "num_tokens": 3722016.0,
      "step": 1830
    },
    {
      "entropy": 6.761053848266601,
      "epoch": 0.1603600454426287,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999399675911869,
      "loss": 6.5318,
      "mean_token_accuracy": 0.12198757007718086,
      "num_tokens": 3731299.0,
      "step": 1835
    },
    {
      "entropy": 6.743390274047852,
      "epoch": 0.16079699379533338,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999392456529113,
      "loss": 6.5806,
      "mean_token_accuracy": 0.12458991408348083,
      "num_tokens": 3740922.0,
      "step": 1840
    },
    {
      "entropy": 6.758446550369262,
      "epoch": 0.1612339421480381,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999385194001831,
      "loss": 6.5937,
      "mean_token_accuracy": 0.11810147613286973,
      "num_tokens": 3750339.0,
      "step": 1845
    },
    {
      "entropy": 6.787387943267822,
      "epoch": 0.16167089050074282,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999377888330164,
      "loss": 6.5467,
      "mean_token_accuracy": 0.12232641279697418,
      "num_tokens": 3761379.0,
      "step": 1850
    },
    {
      "entropy": 6.749542856216431,
      "epoch": 0.1621078388534475,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999370539514249,
      "loss": 6.6294,
      "mean_token_accuracy": 0.11457357108592987,
      "num_tokens": 3771649.0,
      "step": 1855
    },
    {
      "entropy": 6.757665014266967,
      "epoch": 0.16254478720615223,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999363147554229,
      "loss": 6.5497,
      "mean_token_accuracy": 0.12577750012278557,
      "num_tokens": 3782393.0,
      "step": 1860
    },
    {
      "entropy": 6.820600986480713,
      "epoch": 0.16298173555885695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999355712450244,
      "loss": 6.5986,
      "mean_token_accuracy": 0.12362995967268944,
      "num_tokens": 3792849.0,
      "step": 1865
    },
    {
      "entropy": 6.747550010681152,
      "epoch": 0.16341868391156164,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999348234202438,
      "loss": 6.5515,
      "mean_token_accuracy": 0.12130374014377594,
      "num_tokens": 3803848.0,
      "step": 1870
    },
    {
      "entropy": 6.627656269073486,
      "epoch": 0.16385563226426636,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999340712810956,
      "loss": 6.4677,
      "mean_token_accuracy": 0.13533140793442727,
      "num_tokens": 3814704.0,
      "step": 1875
    },
    {
      "entropy": 6.747098636627197,
      "epoch": 0.16429258061697108,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999333148275939,
      "loss": 6.6244,
      "mean_token_accuracy": 0.11492595672607422,
      "num_tokens": 3824955.0,
      "step": 1880
    },
    {
      "entropy": 6.810025024414062,
      "epoch": 0.1647295289696758,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999325540597535,
      "loss": 6.5992,
      "mean_token_accuracy": 0.12268827185034752,
      "num_tokens": 3836540.0,
      "step": 1885
    },
    {
      "entropy": 6.680448246002197,
      "epoch": 0.1651664773223805,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999317889775887,
      "loss": 6.4789,
      "mean_token_accuracy": 0.1307223454117775,
      "num_tokens": 3846472.0,
      "step": 1890
    },
    {
      "entropy": 6.797214794158935,
      "epoch": 0.1656034256750852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999310195811143,
      "loss": 6.6013,
      "mean_token_accuracy": 0.11527118012309075,
      "num_tokens": 3857187.0,
      "step": 1895
    },
    {
      "entropy": 6.796212434768677,
      "epoch": 0.16604037402778993,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999302458703451,
      "loss": 6.5513,
      "mean_token_accuracy": 0.12751033753156663,
      "num_tokens": 3866259.0,
      "step": 1900
    },
    {
      "entropy": 6.679151391983032,
      "epoch": 0.16647732238049462,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499929467845296,
      "loss": 6.6161,
      "mean_token_accuracy": 0.11295140087604523,
      "num_tokens": 3876869.0,
      "step": 1905
    },
    {
      "entropy": 6.8329041481018065,
      "epoch": 0.16691427073319934,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999286855059818,
      "loss": 6.59,
      "mean_token_accuracy": 0.12256261631846428,
      "num_tokens": 3888278.0,
      "step": 1910
    },
    {
      "entropy": 6.806984758377075,
      "epoch": 0.16735121908590406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999278988524177,
      "loss": 6.6406,
      "mean_token_accuracy": 0.11997235789895058,
      "num_tokens": 3898648.0,
      "step": 1915
    },
    {
      "entropy": 6.776583909988403,
      "epoch": 0.16778816743860875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999271078846184,
      "loss": 6.573,
      "mean_token_accuracy": 0.11964118704199792,
      "num_tokens": 3908282.0,
      "step": 1920
    },
    {
      "entropy": 6.679054212570191,
      "epoch": 0.16822511579131347,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999263126025994,
      "loss": 6.5397,
      "mean_token_accuracy": 0.12275912463665009,
      "num_tokens": 3918333.0,
      "step": 1925
    },
    {
      "entropy": 6.806920671463013,
      "epoch": 0.1686620641440182,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999255130063759,
      "loss": 6.5461,
      "mean_token_accuracy": 0.12174615561962128,
      "num_tokens": 3927403.0,
      "step": 1930
    },
    {
      "entropy": 6.764342546463013,
      "epoch": 0.16909901249672288,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999247090959633,
      "loss": 6.5812,
      "mean_token_accuracy": 0.12118082121014595,
      "num_tokens": 3937352.0,
      "step": 1935
    },
    {
      "entropy": 6.679003667831421,
      "epoch": 0.1695359608494276,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999239008713767,
      "loss": 6.4088,
      "mean_token_accuracy": 0.12937364503741264,
      "num_tokens": 3947164.0,
      "step": 1940
    },
    {
      "entropy": 6.770146179199219,
      "epoch": 0.16997290920213232,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999230883326319,
      "loss": 6.5737,
      "mean_token_accuracy": 0.1233607918024063,
      "num_tokens": 3956887.0,
      "step": 1945
    },
    {
      "entropy": 6.735566186904907,
      "epoch": 0.170409857554837,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999222714797445,
      "loss": 6.56,
      "mean_token_accuracy": 0.11969926655292511,
      "num_tokens": 3966542.0,
      "step": 1950
    },
    {
      "entropy": 6.785258102416992,
      "epoch": 0.17084680590754173,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049992145031273,
      "loss": 6.5065,
      "mean_token_accuracy": 0.1332083135843277,
      "num_tokens": 3976127.0,
      "step": 1955
    },
    {
      "entropy": 6.6167535305023195,
      "epoch": 0.17128375426024645,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999206248316043,
      "loss": 6.4685,
      "mean_token_accuracy": 0.12284221649169921,
      "num_tokens": 3987371.0,
      "step": 1960
    },
    {
      "entropy": 6.667543745040893,
      "epoch": 0.17172070261295114,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499919795036383,
      "loss": 6.4859,
      "mean_token_accuracy": 0.1272851638495922,
      "num_tokens": 3997924.0,
      "step": 1965
    },
    {
      "entropy": 6.740949630737305,
      "epoch": 0.17215765096565586,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999189609270823,
      "loss": 6.5751,
      "mean_token_accuracy": 0.12322761192917824,
      "num_tokens": 4009150.0,
      "step": 1970
    },
    {
      "entropy": 6.729563045501709,
      "epoch": 0.17259459931836058,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499918122503718,
      "loss": 6.5836,
      "mean_token_accuracy": 0.12200360670685768,
      "num_tokens": 4018899.0,
      "step": 1975
    },
    {
      "entropy": 6.730152893066406,
      "epoch": 0.17303154767106527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999172797663064,
      "loss": 6.5475,
      "mean_token_accuracy": 0.12035959288477897,
      "num_tokens": 4028652.0,
      "step": 1980
    },
    {
      "entropy": 6.713821458816528,
      "epoch": 0.17346849602376999,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999164327148634,
      "loss": 6.5693,
      "mean_token_accuracy": 0.12123406901955605,
      "num_tokens": 4039819.0,
      "step": 1985
    },
    {
      "entropy": 6.741577911376953,
      "epoch": 0.1739054443764747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999155813494053,
      "loss": 6.4755,
      "mean_token_accuracy": 0.1275830537080765,
      "num_tokens": 4050029.0,
      "step": 1990
    },
    {
      "entropy": 6.6802998065948485,
      "epoch": 0.17434239272917942,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999147256699487,
      "loss": 6.5211,
      "mean_token_accuracy": 0.12550380676984788,
      "num_tokens": 4060497.0,
      "step": 1995
    },
    {
      "entropy": 6.783202648162842,
      "epoch": 0.17477934108188412,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999138656765096,
      "loss": 6.4997,
      "mean_token_accuracy": 0.12641414254903793,
      "num_tokens": 4071477.0,
      "step": 2000
    },
    {
      "entropy": 6.732838535308838,
      "epoch": 0.17521628943458883,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999130013691049,
      "loss": 6.584,
      "mean_token_accuracy": 0.11925829127430916,
      "num_tokens": 4081869.0,
      "step": 2005
    },
    {
      "entropy": 6.760075616836548,
      "epoch": 0.17565323778729355,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999121327477509,
      "loss": 6.5618,
      "mean_token_accuracy": 0.1242613822221756,
      "num_tokens": 4092797.0,
      "step": 2010
    },
    {
      "entropy": 6.757962799072265,
      "epoch": 0.17609018613999824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999112598124642,
      "loss": 6.5178,
      "mean_token_accuracy": 0.12427477762103081,
      "num_tokens": 4103688.0,
      "step": 2015
    },
    {
      "entropy": 6.619799852371216,
      "epoch": 0.17652713449270296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999103825632618,
      "loss": 6.4814,
      "mean_token_accuracy": 0.12500133290886878,
      "num_tokens": 4114037.0,
      "step": 2020
    },
    {
      "entropy": 6.713038921356201,
      "epoch": 0.17696408284540768,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999095010001605,
      "loss": 6.4755,
      "mean_token_accuracy": 0.11943754628300667,
      "num_tokens": 4124164.0,
      "step": 2025
    },
    {
      "entropy": 6.727624893188477,
      "epoch": 0.17740103119811237,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499908615123177,
      "loss": 6.5808,
      "mean_token_accuracy": 0.1142855279147625,
      "num_tokens": 4134225.0,
      "step": 2030
    },
    {
      "entropy": 6.74792366027832,
      "epoch": 0.1778379795508171,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999077249323284,
      "loss": 6.6281,
      "mean_token_accuracy": 0.12126999646425247,
      "num_tokens": 4145962.0,
      "step": 2035
    },
    {
      "entropy": 6.74753704071045,
      "epoch": 0.1782749279035218,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999068304276319,
      "loss": 6.5234,
      "mean_token_accuracy": 0.12188527584075928,
      "num_tokens": 4156572.0,
      "step": 2040
    },
    {
      "entropy": 6.6771139144897464,
      "epoch": 0.1787118762562265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999059316091045,
      "loss": 6.5804,
      "mean_token_accuracy": 0.11935446932911872,
      "num_tokens": 4166960.0,
      "step": 2045
    },
    {
      "entropy": 6.767083215713501,
      "epoch": 0.17914882460893122,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999050284767634,
      "loss": 6.5431,
      "mean_token_accuracy": 0.12220514565706253,
      "num_tokens": 4177833.0,
      "step": 2050
    },
    {
      "entropy": 6.569900465011597,
      "epoch": 0.17958577296163594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999041210306261,
      "loss": 6.4581,
      "mean_token_accuracy": 0.12861283570528032,
      "num_tokens": 4187858.0,
      "step": 2055
    },
    {
      "entropy": 6.775664949417115,
      "epoch": 0.18002272131434063,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999032092707098,
      "loss": 6.5951,
      "mean_token_accuracy": 0.11695825308561325,
      "num_tokens": 4198207.0,
      "step": 2060
    },
    {
      "entropy": 6.746979188919068,
      "epoch": 0.18045966966704535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999022931970323,
      "loss": 6.4872,
      "mean_token_accuracy": 0.12345504090189933,
      "num_tokens": 4207594.0,
      "step": 2065
    },
    {
      "entropy": 6.626251983642578,
      "epoch": 0.18089661801975007,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999013728096107,
      "loss": 6.4911,
      "mean_token_accuracy": 0.12829106748104097,
      "num_tokens": 4217334.0,
      "step": 2070
    },
    {
      "entropy": 6.721229362487793,
      "epoch": 0.18133356637245476,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499900448108463,
      "loss": 6.4728,
      "mean_token_accuracy": 0.1287317655980587,
      "num_tokens": 4226407.0,
      "step": 2075
    },
    {
      "entropy": 6.631599426269531,
      "epoch": 0.18177051472515948,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499899519093607,
      "loss": 6.4147,
      "mean_token_accuracy": 0.1277851089835167,
      "num_tokens": 4237818.0,
      "step": 2080
    },
    {
      "entropy": 6.699046468734741,
      "epoch": 0.1822074630778642,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998985857650603,
      "loss": 6.5392,
      "mean_token_accuracy": 0.11802871823310852,
      "num_tokens": 4247783.0,
      "step": 2085
    },
    {
      "entropy": 6.62796630859375,
      "epoch": 0.18264441143056892,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998976481228409,
      "loss": 6.4955,
      "mean_token_accuracy": 0.12168094217777252,
      "num_tokens": 4258049.0,
      "step": 2090
    },
    {
      "entropy": 6.694241905212403,
      "epoch": 0.1830813597832736,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998967061669668,
      "loss": 6.4824,
      "mean_token_accuracy": 0.12094468101859093,
      "num_tokens": 4269357.0,
      "step": 2095
    },
    {
      "entropy": 6.694693088531494,
      "epoch": 0.18351830813597833,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998957598974559,
      "loss": 6.4142,
      "mean_token_accuracy": 0.13116534650325776,
      "num_tokens": 4279576.0,
      "step": 2100
    },
    {
      "entropy": 6.660217428207398,
      "epoch": 0.18395525648868305,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998948093143266,
      "loss": 6.4518,
      "mean_token_accuracy": 0.13070998936891556,
      "num_tokens": 4289761.0,
      "step": 2105
    },
    {
      "entropy": 6.606316137313843,
      "epoch": 0.18439220484138774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499893854417597,
      "loss": 6.3993,
      "mean_token_accuracy": 0.13042399287223816,
      "num_tokens": 4299231.0,
      "step": 2110
    },
    {
      "entropy": 6.696398591995239,
      "epoch": 0.18482915319409246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998928952072854,
      "loss": 6.5671,
      "mean_token_accuracy": 0.12065399885177612,
      "num_tokens": 4308687.0,
      "step": 2115
    },
    {
      "entropy": 6.724217700958252,
      "epoch": 0.18526610154679718,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998919316834102,
      "loss": 6.5186,
      "mean_token_accuracy": 0.1235981173813343,
      "num_tokens": 4318183.0,
      "step": 2120
    },
    {
      "entropy": 6.707517290115357,
      "epoch": 0.18570304989950187,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049989096384599,
      "loss": 6.544,
      "mean_token_accuracy": 0.11985958814620971,
      "num_tokens": 4328482.0,
      "step": 2125
    },
    {
      "entropy": 6.697282981872559,
      "epoch": 0.1861399982522066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998899916950431,
      "loss": 6.4329,
      "mean_token_accuracy": 0.1220651850104332,
      "num_tokens": 4339313.0,
      "step": 2130
    },
    {
      "entropy": 6.647680282592773,
      "epoch": 0.1865769466049113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998890152305885,
      "loss": 6.5417,
      "mean_token_accuracy": 0.12274084091186524,
      "num_tokens": 4350299.0,
      "step": 2135
    },
    {
      "entropy": 6.823352575302124,
      "epoch": 0.187013894957616,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998880344526447,
      "loss": 6.4685,
      "mean_token_accuracy": 0.12892222180962562,
      "num_tokens": 4361424.0,
      "step": 2140
    },
    {
      "entropy": 6.548090219497681,
      "epoch": 0.18745084331032072,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998870493612306,
      "loss": 6.4544,
      "mean_token_accuracy": 0.12490695714950562,
      "num_tokens": 4371480.0,
      "step": 2145
    },
    {
      "entropy": 6.629548788070679,
      "epoch": 0.18788779166302544,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998860599563649,
      "loss": 6.4753,
      "mean_token_accuracy": 0.12593501806259155,
      "num_tokens": 4381951.0,
      "step": 2150
    },
    {
      "entropy": 6.677478265762329,
      "epoch": 0.18832474001573013,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998850662380668,
      "loss": 6.5327,
      "mean_token_accuracy": 0.11975304633378983,
      "num_tokens": 4392290.0,
      "step": 2155
    },
    {
      "entropy": 6.600126171112061,
      "epoch": 0.18876168836843485,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998840682063553,
      "loss": 6.281,
      "mean_token_accuracy": 0.1401298873126507,
      "num_tokens": 4403350.0,
      "step": 2160
    },
    {
      "entropy": 6.602562713623047,
      "epoch": 0.18919863672113957,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998830658612496,
      "loss": 6.4852,
      "mean_token_accuracy": 0.12300979048013687,
      "num_tokens": 4413171.0,
      "step": 2165
    },
    {
      "entropy": 6.6313416957855225,
      "epoch": 0.18963558507384426,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998820592027688,
      "loss": 6.4376,
      "mean_token_accuracy": 0.12936783209443092,
      "num_tokens": 4424371.0,
      "step": 2170
    },
    {
      "entropy": 6.684165811538696,
      "epoch": 0.19007253342654898,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998810482309322,
      "loss": 6.4014,
      "mean_token_accuracy": 0.12163952067494392,
      "num_tokens": 4434040.0,
      "step": 2175
    },
    {
      "entropy": 6.727186775207519,
      "epoch": 0.1905094817792537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998800329457595,
      "loss": 6.5153,
      "mean_token_accuracy": 0.11811432018876075,
      "num_tokens": 4444663.0,
      "step": 2180
    },
    {
      "entropy": 6.65099720954895,
      "epoch": 0.1909464301319584,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998790133472697,
      "loss": 6.4614,
      "mean_token_accuracy": 0.12238967642188073,
      "num_tokens": 4454733.0,
      "step": 2185
    },
    {
      "entropy": 6.613533592224121,
      "epoch": 0.1913833784846631,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998779894354827,
      "loss": 6.4482,
      "mean_token_accuracy": 0.12639879956841468,
      "num_tokens": 4464286.0,
      "step": 2190
    },
    {
      "entropy": 6.706346750259399,
      "epoch": 0.19182032683736783,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499876961210418,
      "loss": 6.4161,
      "mean_token_accuracy": 0.13247158452868463,
      "num_tokens": 4475005.0,
      "step": 2195
    },
    {
      "entropy": 6.5794055461883545,
      "epoch": 0.19225727519007255,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998759286720953,
      "loss": 6.3851,
      "mean_token_accuracy": 0.12563437446951867,
      "num_tokens": 4484606.0,
      "step": 2200
    },
    {
      "entropy": 6.607835817337036,
      "epoch": 0.19269422354277724,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998748918205345,
      "loss": 6.4861,
      "mean_token_accuracy": 0.12292944341897964,
      "num_tokens": 4494930.0,
      "step": 2205
    },
    {
      "entropy": 6.6335704803466795,
      "epoch": 0.19313117189548196,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998738506557554,
      "loss": 6.4963,
      "mean_token_accuracy": 0.12030340284109116,
      "num_tokens": 4505304.0,
      "step": 2210
    },
    {
      "entropy": 6.737761306762695,
      "epoch": 0.19356812024818668,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499872805177778,
      "loss": 6.4433,
      "mean_token_accuracy": 0.1285339929163456,
      "num_tokens": 4514333.0,
      "step": 2215
    },
    {
      "entropy": 6.530314922332764,
      "epoch": 0.19400506860089137,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998717553866225,
      "loss": 6.4858,
      "mean_token_accuracy": 0.13358827754855157,
      "num_tokens": 4525175.0,
      "step": 2220
    },
    {
      "entropy": 6.710753393173218,
      "epoch": 0.1944420169535961,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998707012823087,
      "loss": 6.3632,
      "mean_token_accuracy": 0.12810946255922318,
      "num_tokens": 4534212.0,
      "step": 2225
    },
    {
      "entropy": 6.524042177200317,
      "epoch": 0.1948789653063008,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998696428648573,
      "loss": 6.5338,
      "mean_token_accuracy": 0.12357461825013161,
      "num_tokens": 4545141.0,
      "step": 2230
    },
    {
      "entropy": 6.690764856338501,
      "epoch": 0.1953159136590055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998685801342881,
      "loss": 6.404,
      "mean_token_accuracy": 0.13337788805365564,
      "num_tokens": 4554410.0,
      "step": 2235
    },
    {
      "entropy": 6.552084350585938,
      "epoch": 0.19575286201171022,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998675130906217,
      "loss": 6.4012,
      "mean_token_accuracy": 0.13402877897024154,
      "num_tokens": 4564397.0,
      "step": 2240
    },
    {
      "entropy": 6.539749431610107,
      "epoch": 0.19618981036441493,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998664417338787,
      "loss": 6.3971,
      "mean_token_accuracy": 0.13362254872918128,
      "num_tokens": 4573986.0,
      "step": 2245
    },
    {
      "entropy": 6.70291690826416,
      "epoch": 0.19662675871711963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998653660640794,
      "loss": 6.4073,
      "mean_token_accuracy": 0.1318839006125927,
      "num_tokens": 4583449.0,
      "step": 2250
    },
    {
      "entropy": 6.56984543800354,
      "epoch": 0.19706370706982435,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998642860812447,
      "loss": 6.4369,
      "mean_token_accuracy": 0.12867740020155907,
      "num_tokens": 4593637.0,
      "step": 2255
    },
    {
      "entropy": 6.647236490249634,
      "epoch": 0.19750065542252906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998632017853951,
      "loss": 6.4067,
      "mean_token_accuracy": 0.13120522499084472,
      "num_tokens": 4603865.0,
      "step": 2260
    },
    {
      "entropy": 6.672588539123535,
      "epoch": 0.19793760377523376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998621131765514,
      "loss": 6.4538,
      "mean_token_accuracy": 0.13059509620070459,
      "num_tokens": 4613502.0,
      "step": 2265
    },
    {
      "entropy": 6.535258674621582,
      "epoch": 0.19837455212793847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998610202547345,
      "loss": 6.4143,
      "mean_token_accuracy": 0.13172344267368316,
      "num_tokens": 4623163.0,
      "step": 2270
    },
    {
      "entropy": 6.695293951034546,
      "epoch": 0.1988115004806432,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998599230199654,
      "loss": 6.4193,
      "mean_token_accuracy": 0.13490767627954484,
      "num_tokens": 4632850.0,
      "step": 2275
    },
    {
      "entropy": 6.589443254470825,
      "epoch": 0.19924844883334789,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998588214722653,
      "loss": 6.4396,
      "mean_token_accuracy": 0.1286391481757164,
      "num_tokens": 4642858.0,
      "step": 2280
    },
    {
      "entropy": 6.698593091964722,
      "epoch": 0.1996853971860526,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998577156116551,
      "loss": 6.4627,
      "mean_token_accuracy": 0.11916870698332786,
      "num_tokens": 4653083.0,
      "step": 2285
    },
    {
      "entropy": 6.624360275268555,
      "epoch": 0.20012234553875732,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998566054381561,
      "loss": 6.4798,
      "mean_token_accuracy": 0.1257111832499504,
      "num_tokens": 4663353.0,
      "step": 2290
    },
    {
      "entropy": 6.623066425323486,
      "epoch": 0.20055929389146204,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998554909517895,
      "loss": 6.3213,
      "mean_token_accuracy": 0.13321887254714965,
      "num_tokens": 4673878.0,
      "step": 2295
    },
    {
      "entropy": 6.537002611160278,
      "epoch": 0.20099624224416673,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998543721525769,
      "loss": 6.4024,
      "mean_token_accuracy": 0.12497134804725647,
      "num_tokens": 4683661.0,
      "step": 2300
    },
    {
      "entropy": 6.5673808574676515,
      "epoch": 0.20143319059687145,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998532490405397,
      "loss": 6.4518,
      "mean_token_accuracy": 0.12829202711582183,
      "num_tokens": 4693817.0,
      "step": 2305
    },
    {
      "entropy": 6.611567735671997,
      "epoch": 0.20187013894957617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998521216156991,
      "loss": 6.3772,
      "mean_token_accuracy": 0.12686420530080794,
      "num_tokens": 4703550.0,
      "step": 2310
    },
    {
      "entropy": 6.595733690261841,
      "epoch": 0.20230708730228086,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998509898780771,
      "loss": 6.4934,
      "mean_token_accuracy": 0.12352922856807709,
      "num_tokens": 4714872.0,
      "step": 2315
    },
    {
      "entropy": 6.627568149566651,
      "epoch": 0.20274403565498558,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998498538276954,
      "loss": 6.3898,
      "mean_token_accuracy": 0.13363091200590133,
      "num_tokens": 4725745.0,
      "step": 2320
    },
    {
      "entropy": 6.600788736343384,
      "epoch": 0.2031809840076903,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998487134645755,
      "loss": 6.4799,
      "mean_token_accuracy": 0.1253775343298912,
      "num_tokens": 4735560.0,
      "step": 2325
    },
    {
      "entropy": 6.63616189956665,
      "epoch": 0.203617932360395,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998475687887396,
      "loss": 6.4533,
      "mean_token_accuracy": 0.12529229298233985,
      "num_tokens": 4746304.0,
      "step": 2330
    },
    {
      "entropy": 6.672122478485107,
      "epoch": 0.2040548807130997,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998464198002095,
      "loss": 6.4285,
      "mean_token_accuracy": 0.12777206525206566,
      "num_tokens": 4756200.0,
      "step": 2335
    },
    {
      "entropy": 6.560856103897095,
      "epoch": 0.20449182906580443,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998452664990072,
      "loss": 6.436,
      "mean_token_accuracy": 0.12261992320418358,
      "num_tokens": 4766249.0,
      "step": 2340
    },
    {
      "entropy": 6.714810609817505,
      "epoch": 0.20492877741850912,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998441088851548,
      "loss": 6.4579,
      "mean_token_accuracy": 0.12660123854875566,
      "num_tokens": 4776510.0,
      "step": 2345
    },
    {
      "entropy": 6.561935997009277,
      "epoch": 0.20536572577121384,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998429469586748,
      "loss": 6.4211,
      "mean_token_accuracy": 0.1272496648132801,
      "num_tokens": 4786798.0,
      "step": 2350
    },
    {
      "entropy": 6.608212327957153,
      "epoch": 0.20580267412391856,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499841780719589,
      "loss": 6.3616,
      "mean_token_accuracy": 0.12745508700609207,
      "num_tokens": 4796001.0,
      "step": 2355
    },
    {
      "entropy": 6.519495153427124,
      "epoch": 0.20623962247662325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998406101679202,
      "loss": 6.3445,
      "mean_token_accuracy": 0.13589456528425217,
      "num_tokens": 4807634.0,
      "step": 2360
    },
    {
      "entropy": 6.578785228729248,
      "epoch": 0.20667657082932797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998394353036906,
      "loss": 6.3473,
      "mean_token_accuracy": 0.13589533045887947,
      "num_tokens": 4817840.0,
      "step": 2365
    },
    {
      "entropy": 6.551007890701294,
      "epoch": 0.2071135191820327,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998382561269229,
      "loss": 6.3326,
      "mean_token_accuracy": 0.1385917641222477,
      "num_tokens": 4827186.0,
      "step": 2370
    },
    {
      "entropy": 6.533224439620971,
      "epoch": 0.20755046753473738,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998370726376395,
      "loss": 6.3894,
      "mean_token_accuracy": 0.12424067407846451,
      "num_tokens": 4837140.0,
      "step": 2375
    },
    {
      "entropy": 6.589614820480347,
      "epoch": 0.2079874158874421,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998358848358634,
      "loss": 6.4017,
      "mean_token_accuracy": 0.12754351869225503,
      "num_tokens": 4846907.0,
      "step": 2380
    },
    {
      "entropy": 6.561453104019165,
      "epoch": 0.20842436424014682,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499834692721617,
      "loss": 6.421,
      "mean_token_accuracy": 0.13229633197188378,
      "num_tokens": 4856165.0,
      "step": 2385
    },
    {
      "entropy": 6.578060579299927,
      "epoch": 0.2088613125928515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998334962949237,
      "loss": 6.2911,
      "mean_token_accuracy": 0.14010017663240432,
      "num_tokens": 4865308.0,
      "step": 2390
    },
    {
      "entropy": 6.493227291107178,
      "epoch": 0.20929826094555623,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998322955558059,
      "loss": 6.358,
      "mean_token_accuracy": 0.1417185001075268,
      "num_tokens": 4876286.0,
      "step": 2395
    },
    {
      "entropy": 6.591450119018555,
      "epoch": 0.20973520929826095,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998310905042869,
      "loss": 6.384,
      "mean_token_accuracy": 0.13348621502518654,
      "num_tokens": 4887227.0,
      "step": 2400
    },
    {
      "entropy": 6.593507528305054,
      "epoch": 0.21017215765096567,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998298811403898,
      "loss": 6.3979,
      "mean_token_accuracy": 0.1288406252861023,
      "num_tokens": 4896994.0,
      "step": 2405
    },
    {
      "entropy": 6.533228969573974,
      "epoch": 0.21060910600367036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998286674641378,
      "loss": 6.4045,
      "mean_token_accuracy": 0.12903448343276977,
      "num_tokens": 4907623.0,
      "step": 2410
    },
    {
      "entropy": 6.558895254135132,
      "epoch": 0.21104605435637508,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499827449475554,
      "loss": 6.3204,
      "mean_token_accuracy": 0.12823840528726577,
      "num_tokens": 4917040.0,
      "step": 2415
    },
    {
      "entropy": 6.683346271514893,
      "epoch": 0.2114830027090798,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499826227174662,
      "loss": 6.436,
      "mean_token_accuracy": 0.12896525338292122,
      "num_tokens": 4928390.0,
      "step": 2420
    },
    {
      "entropy": 6.50466251373291,
      "epoch": 0.2119199510617845,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998250005614852,
      "loss": 6.3981,
      "mean_token_accuracy": 0.12751310169696808,
      "num_tokens": 4939040.0,
      "step": 2425
    },
    {
      "entropy": 6.558785009384155,
      "epoch": 0.2123568994144892,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499823769636047,
      "loss": 6.4371,
      "mean_token_accuracy": 0.12819619551301004,
      "num_tokens": 4949915.0,
      "step": 2430
    },
    {
      "entropy": 6.607331132888794,
      "epoch": 0.21279384776719393,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998225343983712,
      "loss": 6.3391,
      "mean_token_accuracy": 0.1329527035355568,
      "num_tokens": 4960127.0,
      "step": 2435
    },
    {
      "entropy": 6.574900817871094,
      "epoch": 0.21323079611989862,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998212948484814,
      "loss": 6.3847,
      "mean_token_accuracy": 0.126432553678751,
      "num_tokens": 4970152.0,
      "step": 2440
    },
    {
      "entropy": 6.571488666534424,
      "epoch": 0.21366774447260334,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998200509864012,
      "loss": 6.3577,
      "mean_token_accuracy": 0.1341804251074791,
      "num_tokens": 4979745.0,
      "step": 2445
    },
    {
      "entropy": 6.594966459274292,
      "epoch": 0.21410469282530806,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998188028121547,
      "loss": 6.3687,
      "mean_token_accuracy": 0.1393337517976761,
      "num_tokens": 4989924.0,
      "step": 2450
    },
    {
      "entropy": 6.498985242843628,
      "epoch": 0.21454164117801275,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998175503257658,
      "loss": 6.316,
      "mean_token_accuracy": 0.14141854122281075,
      "num_tokens": 4999698.0,
      "step": 2455
    },
    {
      "entropy": 6.539078855514527,
      "epoch": 0.21497858953071747,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998162935272585,
      "loss": 6.4223,
      "mean_token_accuracy": 0.13204047605395317,
      "num_tokens": 5010625.0,
      "step": 2460
    },
    {
      "entropy": 6.532508611679077,
      "epoch": 0.2154155378834222,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998150324166568,
      "loss": 6.3072,
      "mean_token_accuracy": 0.13948291540145874,
      "num_tokens": 5019740.0,
      "step": 2465
    },
    {
      "entropy": 6.578542041778564,
      "epoch": 0.21585248623612688,
      "grad_norm": 0.78125,
      "learning_rate": 0.000499813766993985,
      "loss": 6.4354,
      "mean_token_accuracy": 0.12458334863185883,
      "num_tokens": 5030601.0,
      "step": 2470
    },
    {
      "entropy": 6.645995283126831,
      "epoch": 0.2162894345888316,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998124972592674,
      "loss": 6.3675,
      "mean_token_accuracy": 0.13052909597754478,
      "num_tokens": 5040567.0,
      "step": 2475
    },
    {
      "entropy": 6.499893093109131,
      "epoch": 0.21672638294153632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998112232125283,
      "loss": 6.2648,
      "mean_token_accuracy": 0.13649773374199867,
      "num_tokens": 5049512.0,
      "step": 2480
    },
    {
      "entropy": 6.491473197937012,
      "epoch": 0.217163331294241,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499809944853792,
      "loss": 6.4316,
      "mean_token_accuracy": 0.1274581141769886,
      "num_tokens": 5060986.0,
      "step": 2485
    },
    {
      "entropy": 6.577273893356323,
      "epoch": 0.21760027964694573,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998086621830833,
      "loss": 6.3573,
      "mean_token_accuracy": 0.13598684519529342,
      "num_tokens": 5071492.0,
      "step": 2490
    },
    {
      "entropy": 6.544683933258057,
      "epoch": 0.21803722799965045,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998073752004267,
      "loss": 6.3905,
      "mean_token_accuracy": 0.12805117219686507,
      "num_tokens": 5081722.0,
      "step": 2495
    },
    {
      "entropy": 6.5953404903411865,
      "epoch": 0.21847417635235516,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998060839058467,
      "loss": 6.3545,
      "mean_token_accuracy": 0.1329400770366192,
      "num_tokens": 5091906.0,
      "step": 2500
    },
    {
      "entropy": 6.5407976627349855,
      "epoch": 0.21891112470505986,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998047882993683,
      "loss": 6.3426,
      "mean_token_accuracy": 0.13389598056674004,
      "num_tokens": 5102025.0,
      "step": 2505
    },
    {
      "entropy": 6.509171485900879,
      "epoch": 0.21934807305776458,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998034883810162,
      "loss": 6.3367,
      "mean_token_accuracy": 0.13235569521784782,
      "num_tokens": 5112354.0,
      "step": 2510
    },
    {
      "entropy": 6.513673925399781,
      "epoch": 0.2197850214104693,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998021841508155,
      "loss": 6.2825,
      "mean_token_accuracy": 0.13621082603931428,
      "num_tokens": 5121695.0,
      "step": 2515
    },
    {
      "entropy": 6.564906120300293,
      "epoch": 0.22022196976317399,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998008756087911,
      "loss": 6.3891,
      "mean_token_accuracy": 0.12776936888694762,
      "num_tokens": 5132561.0,
      "step": 2520
    },
    {
      "entropy": 6.589311695098877,
      "epoch": 0.2206589181158787,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499799562754968,
      "loss": 6.4257,
      "mean_token_accuracy": 0.12993937209248543,
      "num_tokens": 5141981.0,
      "step": 2525
    },
    {
      "entropy": 6.544518756866455,
      "epoch": 0.22109586646858342,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997982455893716,
      "loss": 6.2883,
      "mean_token_accuracy": 0.1357527531683445,
      "num_tokens": 5151550.0,
      "step": 2530
    },
    {
      "entropy": 6.574628019332886,
      "epoch": 0.22153281482128812,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997969241120271,
      "loss": 6.3807,
      "mean_token_accuracy": 0.1260584406554699,
      "num_tokens": 5161679.0,
      "step": 2535
    },
    {
      "entropy": 6.491645574569702,
      "epoch": 0.22196976317399283,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997955983229597,
      "loss": 6.4026,
      "mean_token_accuracy": 0.1277308262884617,
      "num_tokens": 5171575.0,
      "step": 2540
    },
    {
      "entropy": 6.547885131835938,
      "epoch": 0.22240671152669755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499794268222195,
      "loss": 6.2719,
      "mean_token_accuracy": 0.13648097291588784,
      "num_tokens": 5180152.0,
      "step": 2545
    },
    {
      "entropy": 6.491997766494751,
      "epoch": 0.22284365987940224,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997929338097583,
      "loss": 6.3554,
      "mean_token_accuracy": 0.13152262344956397,
      "num_tokens": 5190518.0,
      "step": 2550
    },
    {
      "entropy": 6.607607460021972,
      "epoch": 0.22328060823210696,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997915950856755,
      "loss": 6.3933,
      "mean_token_accuracy": 0.12704384401440622,
      "num_tokens": 5201729.0,
      "step": 2555
    },
    {
      "entropy": 6.559991312026978,
      "epoch": 0.22371755658481168,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997902520499721,
      "loss": 6.3744,
      "mean_token_accuracy": 0.1310870312154293,
      "num_tokens": 5211940.0,
      "step": 2560
    },
    {
      "entropy": 6.471191549301148,
      "epoch": 0.22415450493751637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997889047026738,
      "loss": 6.3049,
      "mean_token_accuracy": 0.13113719150424002,
      "num_tokens": 5221028.0,
      "step": 2565
    },
    {
      "entropy": 6.549423122406006,
      "epoch": 0.2245914532902211,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997875530438066,
      "loss": 6.287,
      "mean_token_accuracy": 0.1366332732141018,
      "num_tokens": 5230498.0,
      "step": 2570
    },
    {
      "entropy": 6.511614179611206,
      "epoch": 0.2250284016429258,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997861970733963,
      "loss": 6.3353,
      "mean_token_accuracy": 0.12960499599575998,
      "num_tokens": 5240112.0,
      "step": 2575
    },
    {
      "entropy": 6.5579187870025635,
      "epoch": 0.2254653499956305,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499784836791469,
      "loss": 6.3078,
      "mean_token_accuracy": 0.13300390243530275,
      "num_tokens": 5249201.0,
      "step": 2580
    },
    {
      "entropy": 6.493053865432739,
      "epoch": 0.22590229834833522,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997834721980507,
      "loss": 6.305,
      "mean_token_accuracy": 0.13774593099951743,
      "num_tokens": 5258541.0,
      "step": 2585
    },
    {
      "entropy": 6.616273880004883,
      "epoch": 0.22633924670103994,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004997821032931676,
      "loss": 6.4101,
      "mean_token_accuracy": 0.1332601450383663,
      "num_tokens": 5268524.0,
      "step": 2590
    },
    {
      "entropy": 6.509102439880371,
      "epoch": 0.22677619505374463,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499780730076846,
      "loss": 6.3307,
      "mean_token_accuracy": 0.1330648258328438,
      "num_tokens": 5280702.0,
      "step": 2595
    },
    {
      "entropy": 6.528602409362793,
      "epoch": 0.22721314340644935,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997793525491122,
      "loss": 6.31,
      "mean_token_accuracy": 0.1283218316733837,
      "num_tokens": 5291214.0,
      "step": 2600
    },
    {
      "entropy": 6.432008266448975,
      "epoch": 0.22765009175915407,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997779707099926,
      "loss": 6.2639,
      "mean_token_accuracy": 0.14060572162270546,
      "num_tokens": 5300729.0,
      "step": 2605
    },
    {
      "entropy": 6.540947341918946,
      "epoch": 0.2280870401118588,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997765845595138,
      "loss": 6.4009,
      "mean_token_accuracy": 0.1288476750254631,
      "num_tokens": 5310551.0,
      "step": 2610
    },
    {
      "entropy": 6.5121190547943115,
      "epoch": 0.22852398846456348,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997751940977024,
      "loss": 6.2763,
      "mean_token_accuracy": 0.13367155566811562,
      "num_tokens": 5320038.0,
      "step": 2615
    },
    {
      "entropy": 6.493434906005859,
      "epoch": 0.2289609368172682,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997737993245848,
      "loss": 6.3191,
      "mean_token_accuracy": 0.1320577569305897,
      "num_tokens": 5329753.0,
      "step": 2620
    },
    {
      "entropy": 6.4705726146698,
      "epoch": 0.22939788516997292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997724002401881,
      "loss": 6.304,
      "mean_token_accuracy": 0.13512743636965752,
      "num_tokens": 5340546.0,
      "step": 2625
    },
    {
      "entropy": 6.446058702468872,
      "epoch": 0.2298348335226776,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997709968445391,
      "loss": 6.3795,
      "mean_token_accuracy": 0.12804465666413306,
      "num_tokens": 5352097.0,
      "step": 2630
    },
    {
      "entropy": 6.65869174003601,
      "epoch": 0.23027178187538233,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997695891376644,
      "loss": 6.3208,
      "mean_token_accuracy": 0.13751014918088914,
      "num_tokens": 5361646.0,
      "step": 2635
    },
    {
      "entropy": 6.395161771774292,
      "epoch": 0.23070873022808705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997681771195912,
      "loss": 6.3041,
      "mean_token_accuracy": 0.13794874846935273,
      "num_tokens": 5371695.0,
      "step": 2640
    },
    {
      "entropy": 6.512343025207519,
      "epoch": 0.23114567858079174,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997667607903466,
      "loss": 6.3805,
      "mean_token_accuracy": 0.13010330498218536,
      "num_tokens": 5381978.0,
      "step": 2645
    },
    {
      "entropy": 6.660280466079712,
      "epoch": 0.23158262693349646,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997653401499578,
      "loss": 6.3697,
      "mean_token_accuracy": 0.13411956354975701,
      "num_tokens": 5391891.0,
      "step": 2650
    },
    {
      "entropy": 6.493831157684326,
      "epoch": 0.23201957528620118,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499763915198452,
      "loss": 6.327,
      "mean_token_accuracy": 0.13065533563494683,
      "num_tokens": 5403172.0,
      "step": 2655
    },
    {
      "entropy": 6.456011819839477,
      "epoch": 0.23245652363890587,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997624859358565,
      "loss": 6.3345,
      "mean_token_accuracy": 0.13543095886707307,
      "num_tokens": 5413130.0,
      "step": 2660
    },
    {
      "entropy": 6.504063415527344,
      "epoch": 0.2328934719916106,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997610523621988,
      "loss": 6.3657,
      "mean_token_accuracy": 0.13021062016487123,
      "num_tokens": 5423510.0,
      "step": 2665
    },
    {
      "entropy": 6.5001222610473635,
      "epoch": 0.2333304203443153,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997596144775063,
      "loss": 6.2,
      "mean_token_accuracy": 0.13856842517852783,
      "num_tokens": 5433707.0,
      "step": 2670
    },
    {
      "entropy": 6.316236448287964,
      "epoch": 0.23376736869702,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997581722818067,
      "loss": 6.3145,
      "mean_token_accuracy": 0.13552152067422868,
      "num_tokens": 5444381.0,
      "step": 2675
    },
    {
      "entropy": 6.673515939712525,
      "epoch": 0.23420431704972472,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997567257751275,
      "loss": 6.264,
      "mean_token_accuracy": 0.13442860767245293,
      "num_tokens": 5453905.0,
      "step": 2680
    },
    {
      "entropy": 6.449249124526977,
      "epoch": 0.23464126540242944,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997552749574965,
      "loss": 6.2845,
      "mean_token_accuracy": 0.13393480256199836,
      "num_tokens": 5464839.0,
      "step": 2685
    },
    {
      "entropy": 6.488210773468017,
      "epoch": 0.23507821375513413,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997538198289416,
      "loss": 6.3229,
      "mean_token_accuracy": 0.12909892126917838,
      "num_tokens": 5475459.0,
      "step": 2690
    },
    {
      "entropy": 6.478167104721069,
      "epoch": 0.23551516210783885,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997523603894907,
      "loss": 6.3397,
      "mean_token_accuracy": 0.13458095341920853,
      "num_tokens": 5484886.0,
      "step": 2695
    },
    {
      "entropy": 6.4847033500671385,
      "epoch": 0.23595211046054357,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997508966391717,
      "loss": 6.318,
      "mean_token_accuracy": 0.13543656319379807,
      "num_tokens": 5494803.0,
      "step": 2700
    },
    {
      "entropy": 6.465766191482544,
      "epoch": 0.2363890588132483,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997494285780129,
      "loss": 6.2689,
      "mean_token_accuracy": 0.13447927311062813,
      "num_tokens": 5504997.0,
      "step": 2705
    },
    {
      "entropy": 6.446899366378784,
      "epoch": 0.23682600716595298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997479562060421,
      "loss": 6.2569,
      "mean_token_accuracy": 0.14331548362970353,
      "num_tokens": 5514492.0,
      "step": 2710
    },
    {
      "entropy": 6.511632680892944,
      "epoch": 0.2372629555186577,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997464795232877,
      "loss": 6.3097,
      "mean_token_accuracy": 0.13123842626810073,
      "num_tokens": 5524419.0,
      "step": 2715
    },
    {
      "entropy": 6.525136137008667,
      "epoch": 0.23769990387136242,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997449985297782,
      "loss": 6.323,
      "mean_token_accuracy": 0.1331531934440136,
      "num_tokens": 5534289.0,
      "step": 2720
    },
    {
      "entropy": 6.446746015548706,
      "epoch": 0.2381368522240671,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997435132255418,
      "loss": 6.2524,
      "mean_token_accuracy": 0.13549713790416718,
      "num_tokens": 5543034.0,
      "step": 2725
    },
    {
      "entropy": 6.511647033691406,
      "epoch": 0.23857380057677183,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997420236106071,
      "loss": 6.3347,
      "mean_token_accuracy": 0.12996043935418128,
      "num_tokens": 5553291.0,
      "step": 2730
    },
    {
      "entropy": 6.52672061920166,
      "epoch": 0.23901074892947655,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997405296850026,
      "loss": 6.3147,
      "mean_token_accuracy": 0.1387110285460949,
      "num_tokens": 5563772.0,
      "step": 2735
    },
    {
      "entropy": 6.469681453704834,
      "epoch": 0.23944769728218124,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997390314487569,
      "loss": 6.352,
      "mean_token_accuracy": 0.13162039518356322,
      "num_tokens": 5574348.0,
      "step": 2740
    },
    {
      "entropy": 6.486959552764892,
      "epoch": 0.23988464563488596,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997375289018989,
      "loss": 6.2096,
      "mean_token_accuracy": 0.13282159790396691,
      "num_tokens": 5583491.0,
      "step": 2745
    },
    {
      "entropy": 6.399618673324585,
      "epoch": 0.24032159398759068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997360220444574,
      "loss": 6.3508,
      "mean_token_accuracy": 0.12875445038080216,
      "num_tokens": 5594273.0,
      "step": 2750
    },
    {
      "entropy": 6.516872072219849,
      "epoch": 0.24075854234029537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997345108764611,
      "loss": 6.2636,
      "mean_token_accuracy": 0.13464151471853256,
      "num_tokens": 5604529.0,
      "step": 2755
    },
    {
      "entropy": 6.463331794738769,
      "epoch": 0.2411954906930001,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997329953979391,
      "loss": 6.2527,
      "mean_token_accuracy": 0.14141659960150718,
      "num_tokens": 5614036.0,
      "step": 2760
    },
    {
      "entropy": 6.510278177261353,
      "epoch": 0.2416324390457048,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997314756089206,
      "loss": 6.2963,
      "mean_token_accuracy": 0.13726957514882088,
      "num_tokens": 5624036.0,
      "step": 2765
    },
    {
      "entropy": 6.450707244873047,
      "epoch": 0.2420693873984095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997299515094347,
      "loss": 6.289,
      "mean_token_accuracy": 0.13998497128486634,
      "num_tokens": 5633987.0,
      "step": 2770
    },
    {
      "entropy": 6.427713871002197,
      "epoch": 0.24250633575111422,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997284230995105,
      "loss": 6.166,
      "mean_token_accuracy": 0.14919004142284392,
      "num_tokens": 5643905.0,
      "step": 2775
    },
    {
      "entropy": 6.4683431625366214,
      "epoch": 0.24294328410381894,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997268903791773,
      "loss": 6.3625,
      "mean_token_accuracy": 0.13415857255458832,
      "num_tokens": 5654922.0,
      "step": 2780
    },
    {
      "entropy": 6.475963926315307,
      "epoch": 0.24338023245652363,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997253533484647,
      "loss": 6.1973,
      "mean_token_accuracy": 0.1422443114221096,
      "num_tokens": 5664407.0,
      "step": 2785
    },
    {
      "entropy": 6.472312259674072,
      "epoch": 0.24381718080922835,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997238120074021,
      "loss": 6.2675,
      "mean_token_accuracy": 0.13551050946116447,
      "num_tokens": 5674524.0,
      "step": 2790
    },
    {
      "entropy": 6.439700555801392,
      "epoch": 0.24425412916193306,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997222663560188,
      "loss": 6.3007,
      "mean_token_accuracy": 0.1321689762175083,
      "num_tokens": 5684691.0,
      "step": 2795
    },
    {
      "entropy": 6.4623698711395265,
      "epoch": 0.24469107751463778,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997207163943449,
      "loss": 6.2353,
      "mean_token_accuracy": 0.13899511322379113,
      "num_tokens": 5693683.0,
      "step": 2800
    },
    {
      "entropy": 6.442118167877197,
      "epoch": 0.24512802586734248,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997191621224098,
      "loss": 6.2998,
      "mean_token_accuracy": 0.13590859174728392,
      "num_tokens": 5704394.0,
      "step": 2805
    },
    {
      "entropy": 6.445099401473999,
      "epoch": 0.2455649742200472,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997176035402435,
      "loss": 6.2111,
      "mean_token_accuracy": 0.1418813318014145,
      "num_tokens": 5715480.0,
      "step": 2810
    },
    {
      "entropy": 6.484379482269287,
      "epoch": 0.2460019225727519,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997160406478758,
      "loss": 6.2227,
      "mean_token_accuracy": 0.14038358852267266,
      "num_tokens": 5725973.0,
      "step": 2815
    },
    {
      "entropy": 6.332956981658936,
      "epoch": 0.2464388709254566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997144734453367,
      "loss": 6.2387,
      "mean_token_accuracy": 0.14358900636434554,
      "num_tokens": 5735798.0,
      "step": 2820
    },
    {
      "entropy": 6.579839372634888,
      "epoch": 0.24687581927816132,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997129019326562,
      "loss": 6.3218,
      "mean_token_accuracy": 0.12896375134587287,
      "num_tokens": 5746623.0,
      "step": 2825
    },
    {
      "entropy": 6.478277635574341,
      "epoch": 0.24731276763086604,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997113261098645,
      "loss": 6.2624,
      "mean_token_accuracy": 0.14029808342456818,
      "num_tokens": 5756526.0,
      "step": 2830
    },
    {
      "entropy": 6.400695037841797,
      "epoch": 0.24774971598357073,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997097459769919,
      "loss": 6.2642,
      "mean_token_accuracy": 0.13547210171818733,
      "num_tokens": 5766670.0,
      "step": 2835
    },
    {
      "entropy": 6.524441814422607,
      "epoch": 0.24818666433627545,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997081615340685,
      "loss": 6.3055,
      "mean_token_accuracy": 0.1356332018971443,
      "num_tokens": 5776473.0,
      "step": 2840
    },
    {
      "entropy": 6.552545595169067,
      "epoch": 0.24862361268898017,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997065727811249,
      "loss": 6.3397,
      "mean_token_accuracy": 0.13625893667340278,
      "num_tokens": 5787059.0,
      "step": 2845
    },
    {
      "entropy": 6.441537141799927,
      "epoch": 0.24906056104168486,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997049797181915,
      "loss": 6.3197,
      "mean_token_accuracy": 0.12882847487926483,
      "num_tokens": 5797367.0,
      "step": 2850
    },
    {
      "entropy": 6.392263746261596,
      "epoch": 0.24949750939438958,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997033823452988,
      "loss": 6.2121,
      "mean_token_accuracy": 0.14500029981136323,
      "num_tokens": 5806704.0,
      "step": 2855
    },
    {
      "entropy": 6.455396413803101,
      "epoch": 0.2499344577470943,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997017806624774,
      "loss": 6.2725,
      "mean_token_accuracy": 0.1365084804594517,
      "num_tokens": 5817797.0,
      "step": 2860
    },
    {
      "entropy": 6.506603479385376,
      "epoch": 0.250371406099799,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997001746697582,
      "loss": 6.3181,
      "mean_token_accuracy": 0.13766707554459573,
      "num_tokens": 5829270.0,
      "step": 2865
    },
    {
      "entropy": 6.450180387496948,
      "epoch": 0.2508083544525037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996985643671719,
      "loss": 6.2536,
      "mean_token_accuracy": 0.1362561121582985,
      "num_tokens": 5839214.0,
      "step": 2870
    },
    {
      "entropy": 6.388732814788819,
      "epoch": 0.2512453028052084,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996969497547495,
      "loss": 6.2499,
      "mean_token_accuracy": 0.13788965120911598,
      "num_tokens": 5848171.0,
      "step": 2875
    },
    {
      "entropy": 6.440700387954712,
      "epoch": 0.25168225115791315,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996953308325217,
      "loss": 6.1604,
      "mean_token_accuracy": 0.14541597664356232,
      "num_tokens": 5858399.0,
      "step": 2880
    },
    {
      "entropy": 6.306634092330933,
      "epoch": 0.25211919951061784,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996937076005198,
      "loss": 6.1752,
      "mean_token_accuracy": 0.148306243121624,
      "num_tokens": 5868212.0,
      "step": 2885
    },
    {
      "entropy": 6.465786075592041,
      "epoch": 0.25255614786332253,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996920800587749,
      "loss": 6.2557,
      "mean_token_accuracy": 0.1431330159306526,
      "num_tokens": 5877883.0,
      "step": 2890
    },
    {
      "entropy": 6.490456247329712,
      "epoch": 0.2529930962160273,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996904482073181,
      "loss": 6.2612,
      "mean_token_accuracy": 0.13619642183184624,
      "num_tokens": 5887090.0,
      "step": 2895
    },
    {
      "entropy": 6.379231071472168,
      "epoch": 0.25343004456873197,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996888120461808,
      "loss": 6.299,
      "mean_token_accuracy": 0.1262368731200695,
      "num_tokens": 5897491.0,
      "step": 2900
    },
    {
      "entropy": 6.50172791481018,
      "epoch": 0.25386699292143666,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996871715753943,
      "loss": 6.3195,
      "mean_token_accuracy": 0.13194667026400567,
      "num_tokens": 5908494.0,
      "step": 2905
    },
    {
      "entropy": 6.527822208404541,
      "epoch": 0.2543039412741414,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049968552679499,
      "loss": 6.3587,
      "mean_token_accuracy": 0.12604505568742752,
      "num_tokens": 5919708.0,
      "step": 2910
    },
    {
      "entropy": 6.4339385509490965,
      "epoch": 0.2547408896268461,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996838777049998,
      "loss": 6.2224,
      "mean_token_accuracy": 0.13939420580863954,
      "num_tokens": 5930122.0,
      "step": 2915
    },
    {
      "entropy": 6.35483832359314,
      "epoch": 0.2551778379795508,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499682224305455,
      "loss": 6.2036,
      "mean_token_accuracy": 0.1342392973601818,
      "num_tokens": 5940033.0,
      "step": 2920
    },
    {
      "entropy": 6.388920021057129,
      "epoch": 0.25561478633225554,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996805665963874,
      "loss": 6.2204,
      "mean_token_accuracy": 0.13428228572010995,
      "num_tokens": 5949335.0,
      "step": 2925
    },
    {
      "entropy": 6.490639734268188,
      "epoch": 0.25605173468496023,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996789045778287,
      "loss": 6.264,
      "mean_token_accuracy": 0.13692333027720452,
      "num_tokens": 5959299.0,
      "step": 2930
    },
    {
      "entropy": 6.447759532928467,
      "epoch": 0.2564886830376649,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996772382498111,
      "loss": 6.2174,
      "mean_token_accuracy": 0.13530389219522476,
      "num_tokens": 5970114.0,
      "step": 2935
    },
    {
      "entropy": 6.415350341796875,
      "epoch": 0.25692563139036967,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996755676123662,
      "loss": 6.3127,
      "mean_token_accuracy": 0.134555646777153,
      "num_tokens": 5979630.0,
      "step": 2940
    },
    {
      "entropy": 6.529413795471191,
      "epoch": 0.25736257974307436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996738926655262,
      "loss": 6.289,
      "mean_token_accuracy": 0.13147507831454278,
      "num_tokens": 5990190.0,
      "step": 2945
    },
    {
      "entropy": 6.40431022644043,
      "epoch": 0.25779952809577905,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996722134093231,
      "loss": 6.3126,
      "mean_token_accuracy": 0.13044267743825913,
      "num_tokens": 6000706.0,
      "step": 2950
    },
    {
      "entropy": 6.401627731323242,
      "epoch": 0.2582364764484838,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996705298437894,
      "loss": 6.2655,
      "mean_token_accuracy": 0.1423163965344429,
      "num_tokens": 6011274.0,
      "step": 2955
    },
    {
      "entropy": 6.5021644115448,
      "epoch": 0.2586734248011885,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996688419689572,
      "loss": 6.1833,
      "mean_token_accuracy": 0.1421648696064949,
      "num_tokens": 6021481.0,
      "step": 2960
    },
    {
      "entropy": 6.357604551315307,
      "epoch": 0.25911037315389324,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996671497848589,
      "loss": 6.2652,
      "mean_token_accuracy": 0.1401491306722164,
      "num_tokens": 6032091.0,
      "step": 2965
    },
    {
      "entropy": 6.467347812652588,
      "epoch": 0.25954732150659793,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996654532915269,
      "loss": 6.2658,
      "mean_token_accuracy": 0.1289597101509571,
      "num_tokens": 6043371.0,
      "step": 2970
    },
    {
      "entropy": 6.476518726348877,
      "epoch": 0.2599842698593026,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996637524889937,
      "loss": 6.2193,
      "mean_token_accuracy": 0.1412220261991024,
      "num_tokens": 6052812.0,
      "step": 2975
    },
    {
      "entropy": 6.3784160137176515,
      "epoch": 0.26042121821200737,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996620473772921,
      "loss": 6.2135,
      "mean_token_accuracy": 0.13825001195073128,
      "num_tokens": 6064033.0,
      "step": 2980
    },
    {
      "entropy": 6.435154962539673,
      "epoch": 0.26085816656471206,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996603379564546,
      "loss": 6.3394,
      "mean_token_accuracy": 0.13179437443614006,
      "num_tokens": 6075837.0,
      "step": 2985
    },
    {
      "entropy": 6.565837717056274,
      "epoch": 0.26129511491741675,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996586242265142,
      "loss": 6.244,
      "mean_token_accuracy": 0.14351730942726135,
      "num_tokens": 6085514.0,
      "step": 2990
    },
    {
      "entropy": 6.4207676410675045,
      "epoch": 0.2617320632701215,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996569061875036,
      "loss": 6.1518,
      "mean_token_accuracy": 0.14098117351531983,
      "num_tokens": 6094992.0,
      "step": 2995
    },
    {
      "entropy": 6.362293386459351,
      "epoch": 0.2621690116228262,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996551838394559,
      "loss": 6.2304,
      "mean_token_accuracy": 0.13752028718590736,
      "num_tokens": 6104841.0,
      "step": 3000
    },
    {
      "epoch": 0.2621690116228262,
      "eval_entropy": 6.155421564639298,
      "eval_loss": 6.250469207763672,
      "eval_mean_token_accuracy": 0.14463501781151133,
      "eval_num_tokens": 6104841.0,
      "eval_runtime": 21.3728,
      "eval_samples_per_second": 1731.83,
      "eval_steps_per_second": 216.49,
      "step": 3000
    },
    {
      "entropy": 6.323623704910278,
      "epoch": 0.2626059599755309,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499653457182404,
      "loss": 6.1192,
      "mean_token_accuracy": 0.14590191543102266,
      "num_tokens": 6114830.0,
      "step": 3005
    },
    {
      "entropy": 6.3733845233917235,
      "epoch": 0.2630429083282356,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996517262163811,
      "loss": 6.2357,
      "mean_token_accuracy": 0.13811737820506095,
      "num_tokens": 6125165.0,
      "step": 3010
    },
    {
      "entropy": 6.408387041091919,
      "epoch": 0.2634798566809403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996499909414204,
      "loss": 6.1922,
      "mean_token_accuracy": 0.1431615449488163,
      "num_tokens": 6135670.0,
      "step": 3015
    },
    {
      "entropy": 6.335513734817505,
      "epoch": 0.263916805033645,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996482513575551,
      "loss": 6.1696,
      "mean_token_accuracy": 0.15036366060376166,
      "num_tokens": 6146055.0,
      "step": 3020
    },
    {
      "entropy": 6.312261390686035,
      "epoch": 0.26435375338634975,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996465074648186,
      "loss": 6.1971,
      "mean_token_accuracy": 0.1399642311036587,
      "num_tokens": 6155941.0,
      "step": 3025
    },
    {
      "entropy": 6.462889575958252,
      "epoch": 0.26479070173905445,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996447592632444,
      "loss": 6.2807,
      "mean_token_accuracy": 0.13768690600991249,
      "num_tokens": 6166291.0,
      "step": 3030
    },
    {
      "entropy": 6.402576875686646,
      "epoch": 0.26522765009175914,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996430067528661,
      "loss": 6.2381,
      "mean_token_accuracy": 0.13512784242630005,
      "num_tokens": 6175294.0,
      "step": 3035
    },
    {
      "entropy": 6.445348691940308,
      "epoch": 0.2656645984444639,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996412499337172,
      "loss": 6.2258,
      "mean_token_accuracy": 0.1466323494911194,
      "num_tokens": 6185273.0,
      "step": 3040
    },
    {
      "entropy": 6.305028247833252,
      "epoch": 0.2661015467971686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996394888058314,
      "loss": 6.1867,
      "mean_token_accuracy": 0.14670756012201308,
      "num_tokens": 6195075.0,
      "step": 3045
    },
    {
      "entropy": 6.291659355163574,
      "epoch": 0.26653849514987327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996377233692425,
      "loss": 6.071,
      "mean_token_accuracy": 0.1511462427675724,
      "num_tokens": 6206018.0,
      "step": 3050
    },
    {
      "entropy": 6.42418360710144,
      "epoch": 0.266975443502578,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996359536239843,
      "loss": 6.2151,
      "mean_token_accuracy": 0.1361853964626789,
      "num_tokens": 6216577.0,
      "step": 3055
    },
    {
      "entropy": 6.4269740104675295,
      "epoch": 0.2674123918552827,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996341795700908,
      "loss": 6.2342,
      "mean_token_accuracy": 0.13638746812939645,
      "num_tokens": 6226521.0,
      "step": 3060
    },
    {
      "entropy": 6.414441347122192,
      "epoch": 0.2678493402079874,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996324012075961,
      "loss": 6.2082,
      "mean_token_accuracy": 0.14697427153587342,
      "num_tokens": 6236509.0,
      "step": 3065
    },
    {
      "entropy": 6.406561756134034,
      "epoch": 0.26828628856069214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996306185365341,
      "loss": 6.2053,
      "mean_token_accuracy": 0.14384867027401924,
      "num_tokens": 6246584.0,
      "step": 3070
    },
    {
      "entropy": 6.428066253662109,
      "epoch": 0.26872323691339683,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996288315569393,
      "loss": 6.2306,
      "mean_token_accuracy": 0.1377950392663479,
      "num_tokens": 6256881.0,
      "step": 3075
    },
    {
      "entropy": 6.442715835571289,
      "epoch": 0.2691601852661015,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996270402688457,
      "loss": 6.1965,
      "mean_token_accuracy": 0.13945829793810843,
      "num_tokens": 6266380.0,
      "step": 3080
    },
    {
      "entropy": 6.391282844543457,
      "epoch": 0.2695971336188063,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996252446722877,
      "loss": 6.249,
      "mean_token_accuracy": 0.13977190032601355,
      "num_tokens": 6274872.0,
      "step": 3085
    },
    {
      "entropy": 6.438188314437866,
      "epoch": 0.27003408197151096,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996234447673,
      "loss": 6.2424,
      "mean_token_accuracy": 0.13705142065882683,
      "num_tokens": 6286167.0,
      "step": 3090
    },
    {
      "entropy": 6.3975907325744625,
      "epoch": 0.27047103032421566,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996216405539169,
      "loss": 6.1913,
      "mean_token_accuracy": 0.13941552564501764,
      "num_tokens": 6295898.0,
      "step": 3095
    },
    {
      "entropy": 6.408236789703369,
      "epoch": 0.2709079786769204,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996198320321728,
      "loss": 6.2233,
      "mean_token_accuracy": 0.13398071005940437,
      "num_tokens": 6306543.0,
      "step": 3100
    },
    {
      "entropy": 6.404483842849731,
      "epoch": 0.2713449270296251,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996180192021029,
      "loss": 6.2053,
      "mean_token_accuracy": 0.1326783411204815,
      "num_tokens": 6317709.0,
      "step": 3105
    },
    {
      "entropy": 6.370485639572143,
      "epoch": 0.2717818753823298,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996162020637415,
      "loss": 6.1765,
      "mean_token_accuracy": 0.14306468218564988,
      "num_tokens": 6326874.0,
      "step": 3110
    },
    {
      "entropy": 6.434158134460449,
      "epoch": 0.27221882373503453,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996143806171238,
      "loss": 6.2653,
      "mean_token_accuracy": 0.13487074226140977,
      "num_tokens": 6337943.0,
      "step": 3115
    },
    {
      "entropy": 6.368112993240357,
      "epoch": 0.2726557720877392,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996125548622845,
      "loss": 6.3132,
      "mean_token_accuracy": 0.1386817805469036,
      "num_tokens": 6348442.0,
      "step": 3120
    },
    {
      "entropy": 6.410566329956055,
      "epoch": 0.2730927204404439,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996107247992587,
      "loss": 6.2037,
      "mean_token_accuracy": 0.13915172815322877,
      "num_tokens": 6358675.0,
      "step": 3125
    },
    {
      "entropy": 6.440056848526001,
      "epoch": 0.27352966879314866,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996088904280815,
      "loss": 6.2499,
      "mean_token_accuracy": 0.14083891808986665,
      "num_tokens": 6368631.0,
      "step": 3130
    },
    {
      "entropy": 6.374667739868164,
      "epoch": 0.27396661714585335,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996070517487881,
      "loss": 6.1153,
      "mean_token_accuracy": 0.1427353397011757,
      "num_tokens": 6378709.0,
      "step": 3135
    },
    {
      "entropy": 6.354655408859253,
      "epoch": 0.27440356549855804,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996052087614138,
      "loss": 6.2579,
      "mean_token_accuracy": 0.13231158256530762,
      "num_tokens": 6389686.0,
      "step": 3140
    },
    {
      "entropy": 6.37133002281189,
      "epoch": 0.2748405138512628,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996033614659938,
      "loss": 6.2399,
      "mean_token_accuracy": 0.13986773192882537,
      "num_tokens": 6399447.0,
      "step": 3145
    },
    {
      "entropy": 6.4011969566345215,
      "epoch": 0.2752774622039675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996015098625638,
      "loss": 6.2127,
      "mean_token_accuracy": 0.13879890292882918,
      "num_tokens": 6408891.0,
      "step": 3150
    },
    {
      "entropy": 6.374314546585083,
      "epoch": 0.2757144105566722,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499599653951159,
      "loss": 6.1902,
      "mean_token_accuracy": 0.1419420689344406,
      "num_tokens": 6419471.0,
      "step": 3155
    },
    {
      "entropy": 6.36949577331543,
      "epoch": 0.2761513589093769,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995977937318153,
      "loss": 6.1628,
      "mean_token_accuracy": 0.14073931723833083,
      "num_tokens": 6430445.0,
      "step": 3160
    },
    {
      "entropy": 6.470263910293579,
      "epoch": 0.2765883072620816,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499595929204568,
      "loss": 6.2003,
      "mean_token_accuracy": 0.13983526080846786,
      "num_tokens": 6440821.0,
      "step": 3165
    },
    {
      "entropy": 6.30518856048584,
      "epoch": 0.27702525561478636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995940603694533,
      "loss": 6.1335,
      "mean_token_accuracy": 0.13914740458130836,
      "num_tokens": 6450958.0,
      "step": 3170
    },
    {
      "entropy": 6.367201137542724,
      "epoch": 0.27746220396749105,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995921872265067,
      "loss": 6.2231,
      "mean_token_accuracy": 0.13962776437401772,
      "num_tokens": 6462106.0,
      "step": 3175
    },
    {
      "entropy": 6.391014480590821,
      "epoch": 0.27789915232019574,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995903097757643,
      "loss": 6.1072,
      "mean_token_accuracy": 0.15038541853427886,
      "num_tokens": 6471890.0,
      "step": 3180
    },
    {
      "entropy": 6.315531873703003,
      "epoch": 0.2783361006729005,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995884280172622,
      "loss": 6.1485,
      "mean_token_accuracy": 0.14273526519536972,
      "num_tokens": 6481959.0,
      "step": 3185
    },
    {
      "entropy": 6.428933429718017,
      "epoch": 0.2787730490256052,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995865419510362,
      "loss": 6.1657,
      "mean_token_accuracy": 0.1419281080365181,
      "num_tokens": 6492086.0,
      "step": 3190
    },
    {
      "entropy": 6.296752452850342,
      "epoch": 0.27920999737830987,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995846515771227,
      "loss": 6.197,
      "mean_token_accuracy": 0.14083308726549149,
      "num_tokens": 6502530.0,
      "step": 3195
    },
    {
      "entropy": 6.443424797058105,
      "epoch": 0.2796469457310146,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499582756895558,
      "loss": 6.1915,
      "mean_token_accuracy": 0.1420839510858059,
      "num_tokens": 6511600.0,
      "step": 3200
    },
    {
      "entropy": 6.332727527618408,
      "epoch": 0.2800838940837193,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995808579063782,
      "loss": 6.173,
      "mean_token_accuracy": 0.14028000384569167,
      "num_tokens": 6522712.0,
      "step": 3205
    },
    {
      "entropy": 6.197222900390625,
      "epoch": 0.280520842436424,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049957895460962,
      "loss": 6.0361,
      "mean_token_accuracy": 0.1627690985798836,
      "num_tokens": 6532472.0,
      "step": 3210
    },
    {
      "entropy": 6.32906699180603,
      "epoch": 0.28095779078912875,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995770470053197,
      "loss": 6.0916,
      "mean_token_accuracy": 0.15538571476936341,
      "num_tokens": 6541778.0,
      "step": 3215
    },
    {
      "entropy": 6.362210512161255,
      "epoch": 0.28139473914183344,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995751350935139,
      "loss": 6.1001,
      "mean_token_accuracy": 0.14264210760593415,
      "num_tokens": 6552368.0,
      "step": 3220
    },
    {
      "entropy": 6.332131481170654,
      "epoch": 0.28183168749453813,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995732188742395,
      "loss": 6.1283,
      "mean_token_accuracy": 0.14483243077993393,
      "num_tokens": 6562593.0,
      "step": 3225
    },
    {
      "entropy": 6.334913444519043,
      "epoch": 0.2822686358472429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995712983475329,
      "loss": 6.1907,
      "mean_token_accuracy": 0.14044841825962068,
      "num_tokens": 6572695.0,
      "step": 3230
    },
    {
      "entropy": 6.283826637268066,
      "epoch": 0.28270558419994757,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995693735134313,
      "loss": 6.1624,
      "mean_token_accuracy": 0.14464569389820098,
      "num_tokens": 6581671.0,
      "step": 3235
    },
    {
      "entropy": 6.319446039199829,
      "epoch": 0.28314253255265226,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995674443719713,
      "loss": 6.1454,
      "mean_token_accuracy": 0.13877534121274948,
      "num_tokens": 6590460.0,
      "step": 3240
    },
    {
      "entropy": 6.3591760158538815,
      "epoch": 0.283579480905357,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049956551092319,
      "loss": 6.1708,
      "mean_token_accuracy": 0.1406522735953331,
      "num_tokens": 6599892.0,
      "step": 3245
    },
    {
      "entropy": 6.398914766311646,
      "epoch": 0.2840164292580617,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995635731671247,
      "loss": 6.239,
      "mean_token_accuracy": 0.13753549978137017,
      "num_tokens": 6610530.0,
      "step": 3250
    },
    {
      "entropy": 6.334639930725098,
      "epoch": 0.2844533776107664,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995616311038124,
      "loss": 6.1508,
      "mean_token_accuracy": 0.14377639144659043,
      "num_tokens": 6621117.0,
      "step": 3255
    },
    {
      "entropy": 6.406909799575805,
      "epoch": 0.28489032596347114,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995596847332901,
      "loss": 6.1945,
      "mean_token_accuracy": 0.13885729759931564,
      "num_tokens": 6631521.0,
      "step": 3260
    },
    {
      "entropy": 6.473127269744873,
      "epoch": 0.2853272743161758,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995577340555956,
      "loss": 6.2809,
      "mean_token_accuracy": 0.1377129450440407,
      "num_tokens": 6642180.0,
      "step": 3265
    },
    {
      "entropy": 6.370123815536499,
      "epoch": 0.2857642226688805,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995557790707661,
      "loss": 6.2011,
      "mean_token_accuracy": 0.1410328447818756,
      "num_tokens": 6652735.0,
      "step": 3270
    },
    {
      "entropy": 6.368545866012573,
      "epoch": 0.28620117102158527,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499553819778839,
      "loss": 6.2156,
      "mean_token_accuracy": 0.1393629640340805,
      "num_tokens": 6663135.0,
      "step": 3275
    },
    {
      "entropy": 6.3105274677276615,
      "epoch": 0.28663811937428996,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499551856179852,
      "loss": 6.182,
      "mean_token_accuracy": 0.13699914887547493,
      "num_tokens": 6673375.0,
      "step": 3280
    },
    {
      "entropy": 6.40924563407898,
      "epoch": 0.28707506772699465,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995498882738428,
      "loss": 6.165,
      "mean_token_accuracy": 0.14128110259771348,
      "num_tokens": 6683101.0,
      "step": 3285
    },
    {
      "entropy": 6.398536825180054,
      "epoch": 0.2875120160796994,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995479160608491,
      "loss": 6.2244,
      "mean_token_accuracy": 0.1456000402569771,
      "num_tokens": 6692763.0,
      "step": 3290
    },
    {
      "entropy": 6.375338983535767,
      "epoch": 0.2879489644324041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995459395409085,
      "loss": 6.1933,
      "mean_token_accuracy": 0.138364577293396,
      "num_tokens": 6702833.0,
      "step": 3295
    },
    {
      "entropy": 6.420780420303345,
      "epoch": 0.2883859127851088,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995439587140593,
      "loss": 6.1924,
      "mean_token_accuracy": 0.13553090244531632,
      "num_tokens": 6713083.0,
      "step": 3300
    },
    {
      "entropy": 6.362914037704468,
      "epoch": 0.2888228611378135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995419735803392,
      "loss": 6.1334,
      "mean_token_accuracy": 0.13989951387047767,
      "num_tokens": 6722501.0,
      "step": 3305
    },
    {
      "entropy": 6.35716142654419,
      "epoch": 0.2892598094905182,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995399841397865,
      "loss": 6.1746,
      "mean_token_accuracy": 0.1394180789589882,
      "num_tokens": 6731961.0,
      "step": 3310
    },
    {
      "entropy": 6.4052025318145756,
      "epoch": 0.2896967578432229,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995379903924392,
      "loss": 6.1626,
      "mean_token_accuracy": 0.13826248794794083,
      "num_tokens": 6742576.0,
      "step": 3315
    },
    {
      "entropy": 6.354210472106933,
      "epoch": 0.29013370619592765,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995359923383356,
      "loss": 6.1543,
      "mean_token_accuracy": 0.14654606133699416,
      "num_tokens": 6753066.0,
      "step": 3320
    },
    {
      "entropy": 6.3612380027771,
      "epoch": 0.29057065454863235,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499533989977514,
      "loss": 6.2882,
      "mean_token_accuracy": 0.1354008845984936,
      "num_tokens": 6764455.0,
      "step": 3325
    },
    {
      "entropy": 6.377067708969117,
      "epoch": 0.29100760290133704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995319833100128,
      "loss": 6.1751,
      "mean_token_accuracy": 0.1440010964870453,
      "num_tokens": 6774655.0,
      "step": 3330
    },
    {
      "entropy": 6.313733148574829,
      "epoch": 0.2914445512540418,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995299723358705,
      "loss": 6.1284,
      "mean_token_accuracy": 0.15051591545343398,
      "num_tokens": 6785942.0,
      "step": 3335
    },
    {
      "entropy": 6.3445725440979,
      "epoch": 0.2918814996067465,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995279570551257,
      "loss": 6.1043,
      "mean_token_accuracy": 0.14889372438192366,
      "num_tokens": 6795355.0,
      "step": 3340
    },
    {
      "entropy": 6.33411750793457,
      "epoch": 0.29231844795945117,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499525937467817,
      "loss": 6.1873,
      "mean_token_accuracy": 0.14229143783450127,
      "num_tokens": 6805579.0,
      "step": 3345
    },
    {
      "entropy": 6.335427141189575,
      "epoch": 0.2927553963121559,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995239135739833,
      "loss": 6.21,
      "mean_token_accuracy": 0.13411197140812875,
      "num_tokens": 6814878.0,
      "step": 3350
    },
    {
      "entropy": 6.394797658920288,
      "epoch": 0.2931923446648606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995218853736631,
      "loss": 6.0911,
      "mean_token_accuracy": 0.15279014706611632,
      "num_tokens": 6824533.0,
      "step": 3355
    },
    {
      "entropy": 6.223644113540649,
      "epoch": 0.29362929301756535,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995198528668956,
      "loss": 6.0061,
      "mean_token_accuracy": 0.15976424664258956,
      "num_tokens": 6835529.0,
      "step": 3360
    },
    {
      "entropy": 6.3180726051330565,
      "epoch": 0.29406624137027004,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995178160537197,
      "loss": 6.203,
      "mean_token_accuracy": 0.14279271438717842,
      "num_tokens": 6845679.0,
      "step": 3365
    },
    {
      "entropy": 6.392837190628052,
      "epoch": 0.29450318972297473,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995157749341744,
      "loss": 6.17,
      "mean_token_accuracy": 0.1362040288746357,
      "num_tokens": 6855876.0,
      "step": 3370
    },
    {
      "entropy": 6.338273286819458,
      "epoch": 0.2949401380756795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995137295082989,
      "loss": 6.2352,
      "mean_token_accuracy": 0.13605054542422296,
      "num_tokens": 6866200.0,
      "step": 3375
    },
    {
      "entropy": 6.392386198043823,
      "epoch": 0.2953770864283842,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995116797761324,
      "loss": 6.1145,
      "mean_token_accuracy": 0.14522978886961938,
      "num_tokens": 6875191.0,
      "step": 3380
    },
    {
      "entropy": 6.2830812454223635,
      "epoch": 0.29581403478108886,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995096257377143,
      "loss": 6.0906,
      "mean_token_accuracy": 0.1417955420911312,
      "num_tokens": 6885352.0,
      "step": 3385
    },
    {
      "entropy": 6.347986698150635,
      "epoch": 0.2962509831337936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995075673930839,
      "loss": 6.1666,
      "mean_token_accuracy": 0.140480125695467,
      "num_tokens": 6895585.0,
      "step": 3390
    },
    {
      "entropy": 6.422118520736694,
      "epoch": 0.2966879314864983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995055047422808,
      "loss": 6.1422,
      "mean_token_accuracy": 0.14024515375494956,
      "num_tokens": 6905532.0,
      "step": 3395
    },
    {
      "entropy": 6.189921236038208,
      "epoch": 0.297124879839203,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995034377853443,
      "loss": 6.1406,
      "mean_token_accuracy": 0.14736572355031968,
      "num_tokens": 6915835.0,
      "step": 3400
    },
    {
      "entropy": 6.325133371353149,
      "epoch": 0.29756182819190774,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995013665223143,
      "loss": 6.1851,
      "mean_token_accuracy": 0.1403566814959049,
      "num_tokens": 6925964.0,
      "step": 3405
    },
    {
      "entropy": 6.356245374679565,
      "epoch": 0.29799877654461243,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994992909532305,
      "loss": 6.1326,
      "mean_token_accuracy": 0.14365364983677864,
      "num_tokens": 6936145.0,
      "step": 3410
    },
    {
      "entropy": 6.272683477401733,
      "epoch": 0.2984357248973171,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994972110781326,
      "loss": 6.0961,
      "mean_token_accuracy": 0.1412777103483677,
      "num_tokens": 6945557.0,
      "step": 3415
    },
    {
      "entropy": 6.286343336105347,
      "epoch": 0.29887267325002187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994951268970606,
      "loss": 6.1248,
      "mean_token_accuracy": 0.14249195754528046,
      "num_tokens": 6955147.0,
      "step": 3420
    },
    {
      "entropy": 6.347802257537841,
      "epoch": 0.29930962160272656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994930384100545,
      "loss": 6.0926,
      "mean_token_accuracy": 0.14310672581195832,
      "num_tokens": 6964024.0,
      "step": 3425
    },
    {
      "entropy": 6.309572982788086,
      "epoch": 0.29974656995543125,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994909456171541,
      "loss": 6.154,
      "mean_token_accuracy": 0.13923503085970879,
      "num_tokens": 6974399.0,
      "step": 3430
    },
    {
      "entropy": 6.290287828445434,
      "epoch": 0.300183518308136,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994888485183998,
      "loss": 6.1142,
      "mean_token_accuracy": 0.1422717772424221,
      "num_tokens": 6984324.0,
      "step": 3435
    },
    {
      "entropy": 6.305847883224487,
      "epoch": 0.3006204666608407,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499486747113832,
      "loss": 6.133,
      "mean_token_accuracy": 0.14629677012562753,
      "num_tokens": 6995659.0,
      "step": 3440
    },
    {
      "entropy": 6.379980659484863,
      "epoch": 0.3010574150135454,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994846414034905,
      "loss": 6.2163,
      "mean_token_accuracy": 0.1342682123184204,
      "num_tokens": 7006463.0,
      "step": 3445
    },
    {
      "entropy": 6.322218656539917,
      "epoch": 0.30149436336625013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499482531387416,
      "loss": 6.1273,
      "mean_token_accuracy": 0.14602649956941605,
      "num_tokens": 7016188.0,
      "step": 3450
    },
    {
      "entropy": 6.322809219360352,
      "epoch": 0.3019313117189548,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994804170656489,
      "loss": 6.1706,
      "mean_token_accuracy": 0.13826305493712426,
      "num_tokens": 7028435.0,
      "step": 3455
    },
    {
      "entropy": 6.395906162261963,
      "epoch": 0.3023682600716595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994782984382299,
      "loss": 6.1514,
      "mean_token_accuracy": 0.14209216162562371,
      "num_tokens": 7037992.0,
      "step": 3460
    },
    {
      "entropy": 6.311761283874512,
      "epoch": 0.30280520842436426,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994761755051995,
      "loss": 6.1538,
      "mean_token_accuracy": 0.1375361017882824,
      "num_tokens": 7048743.0,
      "step": 3465
    },
    {
      "entropy": 6.2811463356018065,
      "epoch": 0.30324215677706895,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994740482665982,
      "loss": 6.1306,
      "mean_token_accuracy": 0.14662047624588012,
      "num_tokens": 7059524.0,
      "step": 3470
    },
    {
      "entropy": 6.357618236541748,
      "epoch": 0.30367910512977364,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994719167224672,
      "loss": 6.1196,
      "mean_token_accuracy": 0.14000913128256798,
      "num_tokens": 7069742.0,
      "step": 3475
    },
    {
      "entropy": 6.3101695537567135,
      "epoch": 0.3041160534824784,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994697808728472,
      "loss": 6.1888,
      "mean_token_accuracy": 0.14483822807669638,
      "num_tokens": 7079909.0,
      "step": 3480
    },
    {
      "entropy": 6.37630410194397,
      "epoch": 0.3045530018351831,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994676407177791,
      "loss": 6.194,
      "mean_token_accuracy": 0.13848980888724327,
      "num_tokens": 7089136.0,
      "step": 3485
    },
    {
      "entropy": 6.332058477401733,
      "epoch": 0.30498995018788777,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994654962573043,
      "loss": 6.1845,
      "mean_token_accuracy": 0.14482409805059432,
      "num_tokens": 7099621.0,
      "step": 3490
    },
    {
      "entropy": 6.282474994659424,
      "epoch": 0.3054268985405925,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994633474914634,
      "loss": 6.0498,
      "mean_token_accuracy": 0.15230769515037537,
      "num_tokens": 7109599.0,
      "step": 3495
    },
    {
      "entropy": 6.2929950714111325,
      "epoch": 0.3058638468932972,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994611944202979,
      "loss": 6.0925,
      "mean_token_accuracy": 0.148536317050457,
      "num_tokens": 7120733.0,
      "step": 3500
    },
    {
      "entropy": 6.370125436782837,
      "epoch": 0.3063007952460019,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994590370438492,
      "loss": 6.1835,
      "mean_token_accuracy": 0.1379859611392021,
      "num_tokens": 7129871.0,
      "step": 3505
    },
    {
      "entropy": 6.236745023727417,
      "epoch": 0.30673774359870665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994568753621585,
      "loss": 6.0917,
      "mean_token_accuracy": 0.14422499388456345,
      "num_tokens": 7139167.0,
      "step": 3510
    },
    {
      "entropy": 6.248943614959717,
      "epoch": 0.30717469195141134,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994547093752673,
      "loss": 6.1636,
      "mean_token_accuracy": 0.135141009837389,
      "num_tokens": 7150534.0,
      "step": 3515
    },
    {
      "entropy": 6.35573935508728,
      "epoch": 0.30761164030411603,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994525390832171,
      "loss": 6.1049,
      "mean_token_accuracy": 0.14837786555290222,
      "num_tokens": 7160472.0,
      "step": 3520
    },
    {
      "entropy": 6.327623891830444,
      "epoch": 0.3080485886568208,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994503644860497,
      "loss": 6.1437,
      "mean_token_accuracy": 0.14327712208032609,
      "num_tokens": 7171636.0,
      "step": 3525
    },
    {
      "entropy": 6.293814277648925,
      "epoch": 0.30848553700952547,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994481855838067,
      "loss": 6.0898,
      "mean_token_accuracy": 0.14264843836426735,
      "num_tokens": 7182163.0,
      "step": 3530
    },
    {
      "entropy": 6.240770578384399,
      "epoch": 0.30892248536223016,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994460023765298,
      "loss": 6.1308,
      "mean_token_accuracy": 0.14274099469184875,
      "num_tokens": 7191694.0,
      "step": 3535
    },
    {
      "entropy": 6.328866243362427,
      "epoch": 0.3093594337149349,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000499443814864261,
      "loss": 6.1377,
      "mean_token_accuracy": 0.1470802530646324,
      "num_tokens": 7202813.0,
      "step": 3540
    },
    {
      "entropy": 6.346216440200806,
      "epoch": 0.3097963820676396,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994416230470422,
      "loss": 6.1513,
      "mean_token_accuracy": 0.14442934840917587,
      "num_tokens": 7213226.0,
      "step": 3545
    },
    {
      "entropy": 6.346348524093628,
      "epoch": 0.3102333304203443,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994394269249155,
      "loss": 6.1661,
      "mean_token_accuracy": 0.14010410904884338,
      "num_tokens": 7223520.0,
      "step": 3550
    },
    {
      "entropy": 6.2628617763519285,
      "epoch": 0.31067027877304904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994372264979231,
      "loss": 6.0776,
      "mean_token_accuracy": 0.1526859149336815,
      "num_tokens": 7232651.0,
      "step": 3555
    },
    {
      "entropy": 6.333708190917969,
      "epoch": 0.3111072271257537,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994350217661069,
      "loss": 6.1665,
      "mean_token_accuracy": 0.14107781425118446,
      "num_tokens": 7242064.0,
      "step": 3560
    },
    {
      "entropy": 6.355745840072632,
      "epoch": 0.3115441754784585,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994328127295095,
      "loss": 6.1673,
      "mean_token_accuracy": 0.13943058550357817,
      "num_tokens": 7253048.0,
      "step": 3565
    },
    {
      "entropy": 6.324016427993774,
      "epoch": 0.31198112383116317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994305993881731,
      "loss": 6.128,
      "mean_token_accuracy": 0.14532174542546272,
      "num_tokens": 7262516.0,
      "step": 3570
    },
    {
      "entropy": 6.24652214050293,
      "epoch": 0.31241807218386786,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994283817421402,
      "loss": 6.0457,
      "mean_token_accuracy": 0.14734001606702804,
      "num_tokens": 7272079.0,
      "step": 3575
    },
    {
      "entropy": 6.2355255603790285,
      "epoch": 0.3128550205365726,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994261597914534,
      "loss": 6.0369,
      "mean_token_accuracy": 0.15247876271605493,
      "num_tokens": 7282150.0,
      "step": 3580
    },
    {
      "entropy": 6.269806385040283,
      "epoch": 0.3132919688892773,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994239335361552,
      "loss": 6.1364,
      "mean_token_accuracy": 0.13893278166651726,
      "num_tokens": 7291782.0,
      "step": 3585
    },
    {
      "entropy": 6.329219055175781,
      "epoch": 0.313728917241982,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994217029762884,
      "loss": 6.1038,
      "mean_token_accuracy": 0.14487882032990457,
      "num_tokens": 7301472.0,
      "step": 3590
    },
    {
      "entropy": 6.268067216873169,
      "epoch": 0.31416586559468673,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994194681118957,
      "loss": 6.0686,
      "mean_token_accuracy": 0.1472780004143715,
      "num_tokens": 7311456.0,
      "step": 3595
    },
    {
      "entropy": 6.36771936416626,
      "epoch": 0.3146028139473914,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049941722894302,
      "loss": 6.1642,
      "mean_token_accuracy": 0.1410333700478077,
      "num_tokens": 7321106.0,
      "step": 3600
    },
    {
      "entropy": 6.30884280204773,
      "epoch": 0.3150397623000961,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994149854697042,
      "loss": 6.1497,
      "mean_token_accuracy": 0.14387298971414567,
      "num_tokens": 7331232.0,
      "step": 3605
    },
    {
      "entropy": 6.3294753074646,
      "epoch": 0.31547671065280086,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994127376919916,
      "loss": 6.1401,
      "mean_token_accuracy": 0.1383569210767746,
      "num_tokens": 7341560.0,
      "step": 3610
    },
    {
      "entropy": 6.3129799365997314,
      "epoch": 0.31591365900550555,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994104856099249,
      "loss": 6.1334,
      "mean_token_accuracy": 0.14288947135210037,
      "num_tokens": 7350923.0,
      "step": 3615
    },
    {
      "entropy": 6.246677017211914,
      "epoch": 0.31635060735821025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994082292235476,
      "loss": 6.0293,
      "mean_token_accuracy": 0.14950403571128845,
      "num_tokens": 7360121.0,
      "step": 3620
    },
    {
      "entropy": 6.232157802581787,
      "epoch": 0.316787555710915,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994059685329027,
      "loss": 6.1187,
      "mean_token_accuracy": 0.14253319799900055,
      "num_tokens": 7370823.0,
      "step": 3625
    },
    {
      "entropy": 6.31125545501709,
      "epoch": 0.3172245040636197,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499403703538034,
      "loss": 6.0906,
      "mean_token_accuracy": 0.142649045586586,
      "num_tokens": 7380102.0,
      "step": 3630
    },
    {
      "entropy": 6.270269346237183,
      "epoch": 0.3176614524163244,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994014342389845,
      "loss": 6.0776,
      "mean_token_accuracy": 0.14614664614200593,
      "num_tokens": 7389909.0,
      "step": 3635
    },
    {
      "entropy": 6.2882462501525875,
      "epoch": 0.3180984007690291,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993991606357979,
      "loss": 6.1499,
      "mean_token_accuracy": 0.1452098935842514,
      "num_tokens": 7401092.0,
      "step": 3640
    },
    {
      "entropy": 6.2227781295776365,
      "epoch": 0.3185353491217338,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993968827285179,
      "loss": 6.0548,
      "mean_token_accuracy": 0.14221011474728584,
      "num_tokens": 7412155.0,
      "step": 3645
    },
    {
      "entropy": 6.253341150283814,
      "epoch": 0.3189722974744385,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993946005171881,
      "loss": 6.009,
      "mean_token_accuracy": 0.15619847998023034,
      "num_tokens": 7421981.0,
      "step": 3650
    },
    {
      "entropy": 6.258985996246338,
      "epoch": 0.31940924582714325,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993923140018523,
      "loss": 6.0465,
      "mean_token_accuracy": 0.14668937474489213,
      "num_tokens": 7432298.0,
      "step": 3655
    },
    {
      "entropy": 6.22645845413208,
      "epoch": 0.31984619417984794,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993900231825542,
      "loss": 6.0957,
      "mean_token_accuracy": 0.1497805140912533,
      "num_tokens": 7442575.0,
      "step": 3660
    },
    {
      "entropy": 6.288065576553345,
      "epoch": 0.32028314253255263,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993877280593381,
      "loss": 6.0723,
      "mean_token_accuracy": 0.1488733172416687,
      "num_tokens": 7453072.0,
      "step": 3665
    },
    {
      "entropy": 6.2989182472229,
      "epoch": 0.3207200908852574,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993854286322477,
      "loss": 6.036,
      "mean_token_accuracy": 0.14672954976558686,
      "num_tokens": 7463173.0,
      "step": 3670
    },
    {
      "entropy": 6.268618583679199,
      "epoch": 0.32115703923796207,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993831249013272,
      "loss": 6.1439,
      "mean_token_accuracy": 0.14052271470427513,
      "num_tokens": 7473751.0,
      "step": 3675
    },
    {
      "entropy": 6.2791993618011475,
      "epoch": 0.32159398759066676,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993808168666207,
      "loss": 5.9528,
      "mean_token_accuracy": 0.16023046523332596,
      "num_tokens": 7483233.0,
      "step": 3680
    },
    {
      "entropy": 6.298495483398438,
      "epoch": 0.3220309359433715,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993785045281726,
      "loss": 6.0716,
      "mean_token_accuracy": 0.14089010134339333,
      "num_tokens": 7493152.0,
      "step": 3685
    },
    {
      "entropy": 6.295819568634033,
      "epoch": 0.3224678842960762,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993761878860272,
      "loss": 6.0526,
      "mean_token_accuracy": 0.14754893332719804,
      "num_tokens": 7502463.0,
      "step": 3690
    },
    {
      "entropy": 6.270647859573364,
      "epoch": 0.3229048326487809,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499373866940229,
      "loss": 6.0968,
      "mean_token_accuracy": 0.14904265403747557,
      "num_tokens": 7512301.0,
      "step": 3695
    },
    {
      "entropy": 6.223095989227295,
      "epoch": 0.32334178100148564,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993715416908225,
      "loss": 6.136,
      "mean_token_accuracy": 0.14603498578071594,
      "num_tokens": 7522591.0,
      "step": 3700
    },
    {
      "entropy": 6.325671577453614,
      "epoch": 0.32377872935419033,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993692121378521,
      "loss": 6.1223,
      "mean_token_accuracy": 0.14244823679327964,
      "num_tokens": 7532400.0,
      "step": 3705
    },
    {
      "entropy": 6.294846057891846,
      "epoch": 0.324215677706895,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993668782813629,
      "loss": 6.032,
      "mean_token_accuracy": 0.1447702407836914,
      "num_tokens": 7542163.0,
      "step": 3710
    },
    {
      "entropy": 6.214679718017578,
      "epoch": 0.32465262605959977,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993645401213992,
      "loss": 6.0396,
      "mean_token_accuracy": 0.1523794949054718,
      "num_tokens": 7551638.0,
      "step": 3715
    },
    {
      "entropy": 6.271987628936768,
      "epoch": 0.32508957441230446,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993621976580061,
      "loss": 6.1168,
      "mean_token_accuracy": 0.14809571653604509,
      "num_tokens": 7561230.0,
      "step": 3720
    },
    {
      "entropy": 6.257942390441895,
      "epoch": 0.32552652276500915,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993598508912285,
      "loss": 5.9836,
      "mean_token_accuracy": 0.15367112308740616,
      "num_tokens": 7570962.0,
      "step": 3725
    },
    {
      "entropy": 6.2392877578735355,
      "epoch": 0.3259634711177139,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993574998211114,
      "loss": 6.1723,
      "mean_token_accuracy": 0.1476600542664528,
      "num_tokens": 7581454.0,
      "step": 3730
    },
    {
      "entropy": 6.283637762069702,
      "epoch": 0.3264004194704186,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993551444476998,
      "loss": 6.0484,
      "mean_token_accuracy": 0.15049980580806732,
      "num_tokens": 7591497.0,
      "step": 3735
    },
    {
      "entropy": 6.307848739624023,
      "epoch": 0.3268373678231233,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993527847710392,
      "loss": 6.063,
      "mean_token_accuracy": 0.14827974885702133,
      "num_tokens": 7600719.0,
      "step": 3740
    },
    {
      "entropy": 6.2438582420349125,
      "epoch": 0.32727431617582803,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993504207911744,
      "loss": 6.078,
      "mean_token_accuracy": 0.14993653669953347,
      "num_tokens": 7610550.0,
      "step": 3745
    },
    {
      "entropy": 6.2891895294189455,
      "epoch": 0.3277112645285327,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993480525081511,
      "loss": 6.0816,
      "mean_token_accuracy": 0.149393929541111,
      "num_tokens": 7620472.0,
      "step": 3750
    },
    {
      "entropy": 6.334989738464356,
      "epoch": 0.3281482128812374,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993456799220145,
      "loss": 6.0352,
      "mean_token_accuracy": 0.1458408862352371,
      "num_tokens": 7630354.0,
      "step": 3755
    },
    {
      "entropy": 6.2195631980896,
      "epoch": 0.32858516123394216,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993433030328103,
      "loss": 6.0717,
      "mean_token_accuracy": 0.1445155881345272,
      "num_tokens": 7639689.0,
      "step": 3760
    },
    {
      "entropy": 6.280947875976563,
      "epoch": 0.32902210958664685,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993409218405839,
      "loss": 6.1038,
      "mean_token_accuracy": 0.13953261822462082,
      "num_tokens": 7649319.0,
      "step": 3765
    },
    {
      "entropy": 6.376852893829346,
      "epoch": 0.3294590579393516,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993385363453812,
      "loss": 6.0982,
      "mean_token_accuracy": 0.1464782640337944,
      "num_tokens": 7659008.0,
      "step": 3770
    },
    {
      "entropy": 6.253328609466553,
      "epoch": 0.3298960062920563,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993361465472477,
      "loss": 6.0469,
      "mean_token_accuracy": 0.15120014771819115,
      "num_tokens": 7669064.0,
      "step": 3775
    },
    {
      "entropy": 6.239322471618652,
      "epoch": 0.330332954644761,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993337524462295,
      "loss": 6.1228,
      "mean_token_accuracy": 0.15430048257112502,
      "num_tokens": 7679122.0,
      "step": 3780
    },
    {
      "entropy": 6.17127628326416,
      "epoch": 0.3307699029974657,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993313540423722,
      "loss": 6.1224,
      "mean_token_accuracy": 0.14809983670711518,
      "num_tokens": 7688373.0,
      "step": 3785
    },
    {
      "entropy": 6.304577112197876,
      "epoch": 0.3312068513501704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993289513357221,
      "loss": 6.0405,
      "mean_token_accuracy": 0.14488947838544847,
      "num_tokens": 7697879.0,
      "step": 3790
    },
    {
      "entropy": 6.288220071792603,
      "epoch": 0.3316437997028751,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993265443263251,
      "loss": 6.0891,
      "mean_token_accuracy": 0.1442583419382572,
      "num_tokens": 7708104.0,
      "step": 3795
    },
    {
      "entropy": 6.2732500553131105,
      "epoch": 0.33208074805557986,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993241330142274,
      "loss": 6.0638,
      "mean_token_accuracy": 0.15259820222854614,
      "num_tokens": 7717626.0,
      "step": 3800
    },
    {
      "entropy": 6.220131731033325,
      "epoch": 0.33251769640828455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993217173994754,
      "loss": 6.0937,
      "mean_token_accuracy": 0.14940546303987504,
      "num_tokens": 7728221.0,
      "step": 3805
    },
    {
      "entropy": 6.290520429611206,
      "epoch": 0.33295464476098924,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993192974821154,
      "loss": 6.093,
      "mean_token_accuracy": 0.14939157366752626,
      "num_tokens": 7739083.0,
      "step": 3810
    },
    {
      "entropy": 6.270976781845093,
      "epoch": 0.333391593113694,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993168732621935,
      "loss": 6.1385,
      "mean_token_accuracy": 0.14242835491895675,
      "num_tokens": 7748573.0,
      "step": 3815
    },
    {
      "entropy": 6.280828046798706,
      "epoch": 0.3338285414663987,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993144447397565,
      "loss": 6.0826,
      "mean_token_accuracy": 0.1489911735057831,
      "num_tokens": 7758741.0,
      "step": 3820
    },
    {
      "entropy": 6.2669477462768555,
      "epoch": 0.33426548981910337,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499312011914851,
      "loss": 6.1188,
      "mean_token_accuracy": 0.14098792672157287,
      "num_tokens": 7768504.0,
      "step": 3825
    },
    {
      "entropy": 6.2922595024108885,
      "epoch": 0.3347024381718081,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993095747875234,
      "loss": 6.2201,
      "mean_token_accuracy": 0.14262133985757827,
      "num_tokens": 7779017.0,
      "step": 3830
    },
    {
      "entropy": 6.252893304824829,
      "epoch": 0.3351393865245128,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993071333578207,
      "loss": 6.0454,
      "mean_token_accuracy": 0.15014319717884064,
      "num_tokens": 7787630.0,
      "step": 3835
    },
    {
      "entropy": 6.243571662902832,
      "epoch": 0.3355763348772175,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993046876257898,
      "loss": 6.063,
      "mean_token_accuracy": 0.1484875872731209,
      "num_tokens": 7797105.0,
      "step": 3840
    },
    {
      "entropy": 6.179040956497192,
      "epoch": 0.33601328322992224,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993022375914773,
      "loss": 6.0513,
      "mean_token_accuracy": 0.15276347994804382,
      "num_tokens": 7807070.0,
      "step": 3845
    },
    {
      "entropy": 6.328529930114746,
      "epoch": 0.33645023158262694,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992997832549303,
      "loss": 6.0727,
      "mean_token_accuracy": 0.14834586530923843,
      "num_tokens": 7818090.0,
      "step": 3850
    },
    {
      "entropy": 6.234403991699219,
      "epoch": 0.3368871799353316,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992973246161961,
      "loss": 6.0023,
      "mean_token_accuracy": 0.15412705689668654,
      "num_tokens": 7827562.0,
      "step": 3855
    },
    {
      "entropy": 6.214991044998169,
      "epoch": 0.3373241282880364,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992948616753215,
      "loss": 6.0699,
      "mean_token_accuracy": 0.14271629229187965,
      "num_tokens": 7837581.0,
      "step": 3860
    },
    {
      "entropy": 6.257319402694702,
      "epoch": 0.33776107664074106,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992923944323541,
      "loss": 6.1355,
      "mean_token_accuracy": 0.14508179798722268,
      "num_tokens": 7848729.0,
      "step": 3865
    },
    {
      "entropy": 6.276949787139893,
      "epoch": 0.33819802499344576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499289922887341,
      "loss": 6.0528,
      "mean_token_accuracy": 0.1467779666185379,
      "num_tokens": 7857379.0,
      "step": 3870
    },
    {
      "entropy": 6.248393630981445,
      "epoch": 0.3386349733461505,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992874470403296,
      "loss": 6.0364,
      "mean_token_accuracy": 0.14853208512067795,
      "num_tokens": 7867200.0,
      "step": 3875
    },
    {
      "entropy": 6.229306173324585,
      "epoch": 0.3390719216988552,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992849668913674,
      "loss": 6.025,
      "mean_token_accuracy": 0.14838217496871947,
      "num_tokens": 7876709.0,
      "step": 3880
    },
    {
      "entropy": 6.260332107543945,
      "epoch": 0.3395088700515599,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992824824405021,
      "loss": 6.1168,
      "mean_token_accuracy": 0.14400348365306853,
      "num_tokens": 7886256.0,
      "step": 3885
    },
    {
      "entropy": 6.239607238769532,
      "epoch": 0.33994581840426463,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992799936877813,
      "loss": 6.0613,
      "mean_token_accuracy": 0.14924716725945472,
      "num_tokens": 7896497.0,
      "step": 3890
    },
    {
      "entropy": 6.2658727169036865,
      "epoch": 0.3403827667569693,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992775006332526,
      "loss": 6.0557,
      "mean_token_accuracy": 0.14877313375473022,
      "num_tokens": 7906608.0,
      "step": 3895
    },
    {
      "entropy": 6.30397253036499,
      "epoch": 0.340819715109674,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499275003276964,
      "loss": 6.1385,
      "mean_token_accuracy": 0.1453283652663231,
      "num_tokens": 7916359.0,
      "step": 3900
    },
    {
      "entropy": 6.273185348510742,
      "epoch": 0.34125666346237876,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992725016189632,
      "loss": 6.0197,
      "mean_token_accuracy": 0.15037634670734407,
      "num_tokens": 7925800.0,
      "step": 3905
    },
    {
      "entropy": 6.296712017059326,
      "epoch": 0.34169361181508345,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992699956592984,
      "loss": 6.1087,
      "mean_token_accuracy": 0.14125905185937881,
      "num_tokens": 7935424.0,
      "step": 3910
    },
    {
      "entropy": 6.301828861236572,
      "epoch": 0.34213056016778814,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992674853980175,
      "loss": 6.0887,
      "mean_token_accuracy": 0.1464080199599266,
      "num_tokens": 7945766.0,
      "step": 3915
    },
    {
      "entropy": 6.2670801162719725,
      "epoch": 0.3425675085204929,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992649708351689,
      "loss": 6.1014,
      "mean_token_accuracy": 0.13565851822495462,
      "num_tokens": 7955622.0,
      "step": 3920
    },
    {
      "entropy": 6.259245538711548,
      "epoch": 0.3430044568731976,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992624519708004,
      "loss": 6.1569,
      "mean_token_accuracy": 0.14040575474500655,
      "num_tokens": 7965750.0,
      "step": 3925
    },
    {
      "entropy": 6.282669734954834,
      "epoch": 0.3434414052259023,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992599288049607,
      "loss": 6.1023,
      "mean_token_accuracy": 0.14655931890010834,
      "num_tokens": 7977568.0,
      "step": 3930
    },
    {
      "entropy": 6.2976117610931395,
      "epoch": 0.343878353578607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499257401337698,
      "loss": 6.141,
      "mean_token_accuracy": 0.14332201108336448,
      "num_tokens": 7988411.0,
      "step": 3935
    },
    {
      "entropy": 6.213099670410156,
      "epoch": 0.3443153019313117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992548695690608,
      "loss": 5.9948,
      "mean_token_accuracy": 0.15325959622859955,
      "num_tokens": 7998315.0,
      "step": 3940
    },
    {
      "entropy": 6.220167732238769,
      "epoch": 0.3447522502840164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992523334990977,
      "loss": 6.0708,
      "mean_token_accuracy": 0.14768939912319184,
      "num_tokens": 8007133.0,
      "step": 3945
    },
    {
      "entropy": 6.239842128753662,
      "epoch": 0.34518919863672115,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992497931278574,
      "loss": 5.993,
      "mean_token_accuracy": 0.15506261885166167,
      "num_tokens": 8016882.0,
      "step": 3950
    },
    {
      "entropy": 6.17403621673584,
      "epoch": 0.34562614698942584,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992472484553886,
      "loss": 6.0066,
      "mean_token_accuracy": 0.1535780608654022,
      "num_tokens": 8026893.0,
      "step": 3955
    },
    {
      "entropy": 6.265686178207398,
      "epoch": 0.34606309534213053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049924469948174,
      "loss": 6.0589,
      "mean_token_accuracy": 0.14699437767267226,
      "num_tokens": 8036240.0,
      "step": 3960
    },
    {
      "entropy": 6.254981803894043,
      "epoch": 0.3465000436948353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992421462069606,
      "loss": 6.0264,
      "mean_token_accuracy": 0.14172793477773665,
      "num_tokens": 8046086.0,
      "step": 3965
    },
    {
      "entropy": 6.143713569641113,
      "epoch": 0.34693699204753997,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992395886310993,
      "loss": 6.0421,
      "mean_token_accuracy": 0.15084123313426973,
      "num_tokens": 8056294.0,
      "step": 3970
    },
    {
      "entropy": 6.270413208007812,
      "epoch": 0.3473739404002447,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992370267542052,
      "loss": 6.0585,
      "mean_token_accuracy": 0.14440521076321602,
      "num_tokens": 8067329.0,
      "step": 3975
    },
    {
      "entropy": 6.23886194229126,
      "epoch": 0.3478108887529494,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992344605763275,
      "loss": 6.0222,
      "mean_token_accuracy": 0.1551137462258339,
      "num_tokens": 8077106.0,
      "step": 3980
    },
    {
      "entropy": 6.247228050231934,
      "epoch": 0.3482478371056541,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992318900975153,
      "loss": 5.9315,
      "mean_token_accuracy": 0.15418799221515656,
      "num_tokens": 8087766.0,
      "step": 3985
    },
    {
      "entropy": 6.154195833206177,
      "epoch": 0.34868478545835885,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992293153178179,
      "loss": 6.0426,
      "mean_token_accuracy": 0.15349235832691194,
      "num_tokens": 8098649.0,
      "step": 3990
    },
    {
      "entropy": 6.229670858383178,
      "epoch": 0.34912173381106354,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992267362372848,
      "loss": 6.0428,
      "mean_token_accuracy": 0.14695343226194382,
      "num_tokens": 8109237.0,
      "step": 3995
    },
    {
      "entropy": 6.190158939361572,
      "epoch": 0.34955868216376823,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992241528559653,
      "loss": 6.0491,
      "mean_token_accuracy": 0.15101227760314942,
      "num_tokens": 8119731.0,
      "step": 4000
    },
    {
      "entropy": 6.3188948154449465,
      "epoch": 0.349995630516473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992215651739092,
      "loss": 6.0762,
      "mean_token_accuracy": 0.1389625683426857,
      "num_tokens": 8129448.0,
      "step": 4005
    },
    {
      "entropy": 6.198659229278564,
      "epoch": 0.35043257886917767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499218973191166,
      "loss": 5.9576,
      "mean_token_accuracy": 0.15365623459219932,
      "num_tokens": 8138345.0,
      "step": 4010
    },
    {
      "entropy": 6.234218788146973,
      "epoch": 0.35086952722188236,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992163769077853,
      "loss": 6.0325,
      "mean_token_accuracy": 0.15174643099308013,
      "num_tokens": 8148380.0,
      "step": 4015
    },
    {
      "entropy": 6.184184837341308,
      "epoch": 0.3513064755745871,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499213776323817,
      "loss": 6.0307,
      "mean_token_accuracy": 0.1515910655260086,
      "num_tokens": 8159359.0,
      "step": 4020
    },
    {
      "entropy": 6.27625584602356,
      "epoch": 0.3517434239272918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499211171439311,
      "loss": 6.13,
      "mean_token_accuracy": 0.14428203105926513,
      "num_tokens": 8168586.0,
      "step": 4025
    },
    {
      "entropy": 6.242440605163575,
      "epoch": 0.3521803722799965,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992085622543172,
      "loss": 6.0295,
      "mean_token_accuracy": 0.1498541384935379,
      "num_tokens": 8178836.0,
      "step": 4030
    },
    {
      "entropy": 6.203923177719116,
      "epoch": 0.35261732063270124,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992059487688859,
      "loss": 6.0256,
      "mean_token_accuracy": 0.14939852058887482,
      "num_tokens": 8188933.0,
      "step": 4035
    },
    {
      "entropy": 6.186490678787232,
      "epoch": 0.35305426898540593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992033309830668,
      "loss": 6.0226,
      "mean_token_accuracy": 0.1565045438706875,
      "num_tokens": 8199366.0,
      "step": 4040
    },
    {
      "entropy": 6.350593328475952,
      "epoch": 0.3534912173381106,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004992007088969105,
      "loss": 5.9916,
      "mean_token_accuracy": 0.15373637303709983,
      "num_tokens": 8209806.0,
      "step": 4045
    },
    {
      "entropy": 6.275001335144043,
      "epoch": 0.35392816569081537,
      "grad_norm": 1.0,
      "learning_rate": 0.000499198082510467,
      "loss": 6.1039,
      "mean_token_accuracy": 0.14267848208546638,
      "num_tokens": 8219367.0,
      "step": 4050
    },
    {
      "entropy": 6.16253433227539,
      "epoch": 0.35436511404352006,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991954518237868,
      "loss": 5.9847,
      "mean_token_accuracy": 0.1520978957414627,
      "num_tokens": 8228918.0,
      "step": 4055
    },
    {
      "entropy": 6.227126121520996,
      "epoch": 0.35480206239622475,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991928168369205,
      "loss": 6.0621,
      "mean_token_accuracy": 0.15055495351552964,
      "num_tokens": 8239380.0,
      "step": 4060
    },
    {
      "entropy": 6.332744073867798,
      "epoch": 0.3552390107489295,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991901775499183,
      "loss": 6.1222,
      "mean_token_accuracy": 0.14949264973402024,
      "num_tokens": 8249131.0,
      "step": 4065
    },
    {
      "entropy": 6.273679733276367,
      "epoch": 0.3556759591016342,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991875339628311,
      "loss": 6.0312,
      "mean_token_accuracy": 0.14499866515398024,
      "num_tokens": 8258550.0,
      "step": 4070
    },
    {
      "entropy": 6.136135101318359,
      "epoch": 0.3561129074543389,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991848860757095,
      "loss": 6.0566,
      "mean_token_accuracy": 0.14953689575195311,
      "num_tokens": 8269089.0,
      "step": 4075
    },
    {
      "entropy": 6.247356796264649,
      "epoch": 0.3565498558070436,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991822338886044,
      "loss": 6.0246,
      "mean_token_accuracy": 0.14642863124608993,
      "num_tokens": 8279735.0,
      "step": 4080
    },
    {
      "entropy": 6.219600105285645,
      "epoch": 0.3569868041597483,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991795774015665,
      "loss": 5.9948,
      "mean_token_accuracy": 0.15589303225278855,
      "num_tokens": 8289693.0,
      "step": 4085
    },
    {
      "entropy": 6.208535385131836,
      "epoch": 0.357423752512453,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991769166146469,
      "loss": 6.0678,
      "mean_token_accuracy": 0.14681545794010162,
      "num_tokens": 8300909.0,
      "step": 4090
    },
    {
      "entropy": 6.214974737167358,
      "epoch": 0.35786070086515775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991742515278965,
      "loss": 6.0435,
      "mean_token_accuracy": 0.1520109787583351,
      "num_tokens": 8311479.0,
      "step": 4095
    },
    {
      "entropy": 6.325199317932129,
      "epoch": 0.35829764921786245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991715821413665,
      "loss": 6.034,
      "mean_token_accuracy": 0.14964867979288102,
      "num_tokens": 8321267.0,
      "step": 4100
    },
    {
      "entropy": 6.17149133682251,
      "epoch": 0.35873459757056714,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991689084551082,
      "loss": 5.9986,
      "mean_token_accuracy": 0.14890123158693314,
      "num_tokens": 8331445.0,
      "step": 4105
    },
    {
      "entropy": 6.183329296112061,
      "epoch": 0.3591715459232719,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991662304691727,
      "loss": 6.0908,
      "mean_token_accuracy": 0.1427320085465908,
      "num_tokens": 8340519.0,
      "step": 4110
    },
    {
      "entropy": 6.3197602272033695,
      "epoch": 0.3596084942759766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991635481836114,
      "loss": 6.0661,
      "mean_token_accuracy": 0.14441095665097237,
      "num_tokens": 8350467.0,
      "step": 4115
    },
    {
      "entropy": 6.202519273757934,
      "epoch": 0.36004544262868127,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991608615984757,
      "loss": 6.0022,
      "mean_token_accuracy": 0.14529087096452714,
      "num_tokens": 8360067.0,
      "step": 4120
    },
    {
      "entropy": 6.245105743408203,
      "epoch": 0.360482390981386,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991581707138175,
      "loss": 6.0681,
      "mean_token_accuracy": 0.15000591427087784,
      "num_tokens": 8370509.0,
      "step": 4125
    },
    {
      "entropy": 6.229617643356323,
      "epoch": 0.3609193393340907,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004991554755296878,
      "loss": 6.0181,
      "mean_token_accuracy": 0.14957005232572557,
      "num_tokens": 8381489.0,
      "step": 4130
    },
    {
      "entropy": 6.133305358886719,
      "epoch": 0.3613562876867954,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991527760461389,
      "loss": 6.0447,
      "mean_token_accuracy": 0.14845455139875413,
      "num_tokens": 8392287.0,
      "step": 4135
    },
    {
      "entropy": 6.175074529647827,
      "epoch": 0.36179323603950014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991500722632223,
      "loss": 6.0192,
      "mean_token_accuracy": 0.15073377788066863,
      "num_tokens": 8402433.0,
      "step": 4140
    },
    {
      "entropy": 6.191783428192139,
      "epoch": 0.36223018439220483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991473641809896,
      "loss": 6.0585,
      "mean_token_accuracy": 0.14769651219248772,
      "num_tokens": 8412944.0,
      "step": 4145
    },
    {
      "entropy": 6.279351711273193,
      "epoch": 0.3626671327449095,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991446517994932,
      "loss": 5.9757,
      "mean_token_accuracy": 0.15047049522399902,
      "num_tokens": 8422511.0,
      "step": 4150
    },
    {
      "entropy": 6.17780818939209,
      "epoch": 0.3631040810976143,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991419351187849,
      "loss": 5.9898,
      "mean_token_accuracy": 0.15758653730154037,
      "num_tokens": 8433094.0,
      "step": 4155
    },
    {
      "entropy": 6.155093050003051,
      "epoch": 0.36354102945031896,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991392141389169,
      "loss": 6.1022,
      "mean_token_accuracy": 0.14809029847383498,
      "num_tokens": 8444430.0,
      "step": 4160
    },
    {
      "entropy": 6.306630706787109,
      "epoch": 0.36397797780302366,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991364888599412,
      "loss": 6.1555,
      "mean_token_accuracy": 0.1399155631661415,
      "num_tokens": 8455171.0,
      "step": 4165
    },
    {
      "entropy": 6.229326391220093,
      "epoch": 0.3644149261557284,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991337592819103,
      "loss": 5.9645,
      "mean_token_accuracy": 0.16167067140340804,
      "num_tokens": 8464724.0,
      "step": 4170
    },
    {
      "entropy": 6.254192352294922,
      "epoch": 0.3648518745084331,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991310254048763,
      "loss": 6.0618,
      "mean_token_accuracy": 0.14853513836860657,
      "num_tokens": 8474732.0,
      "step": 4175
    },
    {
      "entropy": 6.174812602996826,
      "epoch": 0.36528882286113784,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004991282872288919,
      "loss": 5.9886,
      "mean_token_accuracy": 0.15376615524291992,
      "num_tokens": 8484731.0,
      "step": 4180
    },
    {
      "entropy": 6.135093688964844,
      "epoch": 0.36572577121384253,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991255447540095,
      "loss": 5.9994,
      "mean_token_accuracy": 0.15322560518980027,
      "num_tokens": 8494509.0,
      "step": 4185
    },
    {
      "entropy": 6.226243734359741,
      "epoch": 0.3661627195665472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991227979802817,
      "loss": 6.0712,
      "mean_token_accuracy": 0.15194178372621536,
      "num_tokens": 8504260.0,
      "step": 4190
    },
    {
      "entropy": 6.267135906219482,
      "epoch": 0.36659966791925197,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991200469077612,
      "loss": 6.1003,
      "mean_token_accuracy": 0.14540060311555864,
      "num_tokens": 8514222.0,
      "step": 4195
    },
    {
      "entropy": 6.238394927978516,
      "epoch": 0.36703661627195666,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991172915365008,
      "loss": 6.0347,
      "mean_token_accuracy": 0.14810635894536972,
      "num_tokens": 8523979.0,
      "step": 4200
    },
    {
      "entropy": 6.182950115203857,
      "epoch": 0.36747356462466135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991145318665532,
      "loss": 5.9878,
      "mean_token_accuracy": 0.15134034305810928,
      "num_tokens": 8533516.0,
      "step": 4205
    },
    {
      "entropy": 6.24036808013916,
      "epoch": 0.3679105129773661,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991117678979716,
      "loss": 6.0959,
      "mean_token_accuracy": 0.14133178293704987,
      "num_tokens": 8544444.0,
      "step": 4210
    },
    {
      "entropy": 6.229537582397461,
      "epoch": 0.3683474613300708,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991089996308089,
      "loss": 6.014,
      "mean_token_accuracy": 0.15077755153179168,
      "num_tokens": 8555571.0,
      "step": 4215
    },
    {
      "entropy": 6.258168125152588,
      "epoch": 0.3687844096827755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991062270651179,
      "loss": 6.0761,
      "mean_token_accuracy": 0.1440155141055584,
      "num_tokens": 8566486.0,
      "step": 4220
    },
    {
      "entropy": 6.260686540603638,
      "epoch": 0.36922135803548023,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991034502009522,
      "loss": 5.994,
      "mean_token_accuracy": 0.14868567287921905,
      "num_tokens": 8575324.0,
      "step": 4225
    },
    {
      "entropy": 6.178703641891479,
      "epoch": 0.3696583063881849,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991006690383649,
      "loss": 5.9843,
      "mean_token_accuracy": 0.15177585929632187,
      "num_tokens": 8586015.0,
      "step": 4230
    },
    {
      "entropy": 6.180662965774536,
      "epoch": 0.3700952547408896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990978835774093,
      "loss": 5.9716,
      "mean_token_accuracy": 0.15128722190856933,
      "num_tokens": 8594748.0,
      "step": 4235
    },
    {
      "entropy": 6.164231681823731,
      "epoch": 0.37053220309359436,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990950938181389,
      "loss": 6.0357,
      "mean_token_accuracy": 0.1446013018488884,
      "num_tokens": 8605805.0,
      "step": 4240
    },
    {
      "entropy": 6.225905275344848,
      "epoch": 0.37096915144629905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990922997606071,
      "loss": 5.9903,
      "mean_token_accuracy": 0.15047531723976135,
      "num_tokens": 8616591.0,
      "step": 4245
    },
    {
      "entropy": 6.239368629455567,
      "epoch": 0.37140609979900374,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990895014048677,
      "loss": 6.0128,
      "mean_token_accuracy": 0.14804871529340743,
      "num_tokens": 8627063.0,
      "step": 4250
    },
    {
      "entropy": 6.295636320114136,
      "epoch": 0.3718430481517085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990866987509741,
      "loss": 6.1073,
      "mean_token_accuracy": 0.14160066395998,
      "num_tokens": 8637041.0,
      "step": 4255
    },
    {
      "entropy": 6.254852056503296,
      "epoch": 0.3722799965044132,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990838917989802,
      "loss": 6.029,
      "mean_token_accuracy": 0.15455144196748732,
      "num_tokens": 8647485.0,
      "step": 4260
    },
    {
      "entropy": 6.221179723739624,
      "epoch": 0.37271694485711787,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049908108054894,
      "loss": 5.9962,
      "mean_token_accuracy": 0.149809268116951,
      "num_tokens": 8657185.0,
      "step": 4265
    },
    {
      "entropy": 6.21356430053711,
      "epoch": 0.3731538932098226,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990782650009071,
      "loss": 5.9943,
      "mean_token_accuracy": 0.15237897485494614,
      "num_tokens": 8667639.0,
      "step": 4270
    },
    {
      "entropy": 6.234360885620117,
      "epoch": 0.3735908415625273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990754451549358,
      "loss": 6.0223,
      "mean_token_accuracy": 0.1505249857902527,
      "num_tokens": 8678003.0,
      "step": 4275
    },
    {
      "entropy": 6.260536813735962,
      "epoch": 0.374027789915232,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990726210110798,
      "loss": 6.0684,
      "mean_token_accuracy": 0.148902989923954,
      "num_tokens": 8688837.0,
      "step": 4280
    },
    {
      "entropy": 6.17373104095459,
      "epoch": 0.37446473826793675,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990697925693938,
      "loss": 5.8718,
      "mean_token_accuracy": 0.16451138779520988,
      "num_tokens": 8698395.0,
      "step": 4285
    },
    {
      "entropy": 6.20360255241394,
      "epoch": 0.37490168662064144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990669598299316,
      "loss": 5.9927,
      "mean_token_accuracy": 0.14994982928037642,
      "num_tokens": 8708031.0,
      "step": 4290
    },
    {
      "entropy": 6.183550834655762,
      "epoch": 0.37533863497334613,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990641227927478,
      "loss": 5.975,
      "mean_token_accuracy": 0.1573784425854683,
      "num_tokens": 8717825.0,
      "step": 4295
    },
    {
      "entropy": 6.2676128387451175,
      "epoch": 0.3757755833260509,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990612814578967,
      "loss": 6.0417,
      "mean_token_accuracy": 0.14686352163553237,
      "num_tokens": 8728873.0,
      "step": 4300
    },
    {
      "entropy": 6.186302328109742,
      "epoch": 0.37621253167875557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990584358254327,
      "loss": 5.9869,
      "mean_token_accuracy": 0.15025996789336205,
      "num_tokens": 8737917.0,
      "step": 4305
    },
    {
      "entropy": 6.143138313293457,
      "epoch": 0.37664948003146026,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990555858954107,
      "loss": 5.956,
      "mean_token_accuracy": 0.15213080048561095,
      "num_tokens": 8747466.0,
      "step": 4310
    },
    {
      "entropy": 6.199163818359375,
      "epoch": 0.377086428384165,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499052731667885,
      "loss": 5.983,
      "mean_token_accuracy": 0.14877506643533706,
      "num_tokens": 8757037.0,
      "step": 4315
    },
    {
      "entropy": 6.220801067352295,
      "epoch": 0.3775233767368697,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990498731429105,
      "loss": 6.0135,
      "mean_token_accuracy": 0.15214569196105004,
      "num_tokens": 8766780.0,
      "step": 4320
    },
    {
      "entropy": 6.177535533905029,
      "epoch": 0.3779603250895744,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004990470103205421,
      "loss": 5.9905,
      "mean_token_accuracy": 0.15147044509649277,
      "num_tokens": 8777063.0,
      "step": 4325
    },
    {
      "entropy": 6.174711036682129,
      "epoch": 0.37839727344227914,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990441432008347,
      "loss": 6.022,
      "mean_token_accuracy": 0.15331983268260957,
      "num_tokens": 8786479.0,
      "step": 4330
    },
    {
      "entropy": 6.286836624145508,
      "epoch": 0.3788342217949838,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990412717838431,
      "loss": 6.0077,
      "mean_token_accuracy": 0.15067116022109986,
      "num_tokens": 8795663.0,
      "step": 4335
    },
    {
      "entropy": 6.194593238830566,
      "epoch": 0.3792711701476885,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990383960696226,
      "loss": 5.9576,
      "mean_token_accuracy": 0.15578173249959945,
      "num_tokens": 8805854.0,
      "step": 4340
    },
    {
      "entropy": 6.161340379714966,
      "epoch": 0.37970811850039327,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990355160582283,
      "loss": 5.941,
      "mean_token_accuracy": 0.15362798422574997,
      "num_tokens": 8816068.0,
      "step": 4345
    },
    {
      "entropy": 6.109852981567383,
      "epoch": 0.38014506685309796,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990326317497153,
      "loss": 5.9452,
      "mean_token_accuracy": 0.15802857875823975,
      "num_tokens": 8826301.0,
      "step": 4350
    },
    {
      "entropy": 6.22525634765625,
      "epoch": 0.38058201520580265,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990297431441391,
      "loss": 5.9873,
      "mean_token_accuracy": 0.15714989155530928,
      "num_tokens": 8837482.0,
      "step": 4355
    },
    {
      "entropy": 6.203273487091065,
      "epoch": 0.3810189635585074,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499026850241555,
      "loss": 6.0193,
      "mean_token_accuracy": 0.1547929033637047,
      "num_tokens": 8847041.0,
      "step": 4360
    },
    {
      "entropy": 6.156243848800659,
      "epoch": 0.3814559119112121,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990239530420185,
      "loss": 6.0044,
      "mean_token_accuracy": 0.14821452349424363,
      "num_tokens": 8856963.0,
      "step": 4365
    },
    {
      "entropy": 6.224261093139648,
      "epoch": 0.3818928602639168,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990210515455853,
      "loss": 5.9804,
      "mean_token_accuracy": 0.15040574073791504,
      "num_tokens": 8867318.0,
      "step": 4370
    },
    {
      "entropy": 6.172735548019409,
      "epoch": 0.3823298086166215,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990181457523109,
      "loss": 5.9897,
      "mean_token_accuracy": 0.15233128890395164,
      "num_tokens": 8877634.0,
      "step": 4375
    },
    {
      "entropy": 6.164595365524292,
      "epoch": 0.3827667569693262,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499015235662251,
      "loss": 6.043,
      "mean_token_accuracy": 0.14754218459129334,
      "num_tokens": 8887708.0,
      "step": 4380
    },
    {
      "entropy": 6.176977443695068,
      "epoch": 0.38320370532203096,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990123212754616,
      "loss": 5.9539,
      "mean_token_accuracy": 0.15185219049453735,
      "num_tokens": 8897510.0,
      "step": 4385
    },
    {
      "entropy": 6.188373565673828,
      "epoch": 0.38364065367473565,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990094025919985,
      "loss": 6.0033,
      "mean_token_accuracy": 0.14657588303089142,
      "num_tokens": 8908282.0,
      "step": 4390
    },
    {
      "entropy": 6.2857012271881105,
      "epoch": 0.38407760202744035,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990064796119175,
      "loss": 6.0356,
      "mean_token_accuracy": 0.1425875850021839,
      "num_tokens": 8918332.0,
      "step": 4395
    },
    {
      "entropy": 6.173046016693116,
      "epoch": 0.3845145503801451,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990035523352751,
      "loss": 5.9892,
      "mean_token_accuracy": 0.1499785751104355,
      "num_tokens": 8929138.0,
      "step": 4400
    },
    {
      "entropy": 6.1726936340332035,
      "epoch": 0.3849514987328498,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499000620762127,
      "loss": 6.0328,
      "mean_token_accuracy": 0.14553941041231155,
      "num_tokens": 8940871.0,
      "step": 4405
    },
    {
      "entropy": 6.287671709060669,
      "epoch": 0.3853884470855545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989976848925297,
      "loss": 5.9638,
      "mean_token_accuracy": 0.15346103310585021,
      "num_tokens": 8951199.0,
      "step": 4410
    },
    {
      "entropy": 6.165547323226929,
      "epoch": 0.3858253954382592,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989947447265394,
      "loss": 6.0944,
      "mean_token_accuracy": 0.14490815848112107,
      "num_tokens": 8961482.0,
      "step": 4415
    },
    {
      "entropy": 6.293553400039673,
      "epoch": 0.3862623437909639,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989918002642126,
      "loss": 6.0018,
      "mean_token_accuracy": 0.15048256516456604,
      "num_tokens": 8971884.0,
      "step": 4420
    },
    {
      "entropy": 6.137409687042236,
      "epoch": 0.3866992921436686,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004989888515056057,
      "loss": 5.9542,
      "mean_token_accuracy": 0.15159018188714982,
      "num_tokens": 8982086.0,
      "step": 4425
    },
    {
      "entropy": 6.112029314041138,
      "epoch": 0.38713624049637335,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989858984507752,
      "loss": 6.0178,
      "mean_token_accuracy": 0.15148152112960817,
      "num_tokens": 8991555.0,
      "step": 4430
    },
    {
      "entropy": 6.229978370666504,
      "epoch": 0.38757318884907804,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989829410997778,
      "loss": 5.9993,
      "mean_token_accuracy": 0.14953123927116393,
      "num_tokens": 9002459.0,
      "step": 4435
    },
    {
      "entropy": 6.209309053421021,
      "epoch": 0.38801013720178273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989799794526702,
      "loss": 6.0467,
      "mean_token_accuracy": 0.14707298427820206,
      "num_tokens": 9012514.0,
      "step": 4440
    },
    {
      "entropy": 6.207473945617676,
      "epoch": 0.3884470855544875,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989770135095093,
      "loss": 5.9602,
      "mean_token_accuracy": 0.15147794038057327,
      "num_tokens": 9022636.0,
      "step": 4445
    },
    {
      "entropy": 6.140289831161499,
      "epoch": 0.3888840339071922,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989740432703519,
      "loss": 5.9715,
      "mean_token_accuracy": 0.15329394191503526,
      "num_tokens": 9032868.0,
      "step": 4450
    },
    {
      "entropy": 6.165075349807739,
      "epoch": 0.38932098225989686,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498971068735255,
      "loss": 5.9769,
      "mean_token_accuracy": 0.15269885510206221,
      "num_tokens": 9043057.0,
      "step": 4455
    },
    {
      "entropy": 6.17115421295166,
      "epoch": 0.3897579306126016,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004989680899042756,
      "loss": 5.9885,
      "mean_token_accuracy": 0.15040106773376466,
      "num_tokens": 9053179.0,
      "step": 4460
    },
    {
      "entropy": 6.179643630981445,
      "epoch": 0.3901948789653063,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989651067774708,
      "loss": 5.9872,
      "mean_token_accuracy": 0.1546528235077858,
      "num_tokens": 9063112.0,
      "step": 4465
    },
    {
      "entropy": 6.208352756500244,
      "epoch": 0.390631827318011,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989621193548981,
      "loss": 6.008,
      "mean_token_accuracy": 0.1501965716481209,
      "num_tokens": 9073512.0,
      "step": 4470
    },
    {
      "entropy": 6.233810091018677,
      "epoch": 0.39106877567071574,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989591276366145,
      "loss": 6.0081,
      "mean_token_accuracy": 0.14962913393974303,
      "num_tokens": 9083369.0,
      "step": 4475
    },
    {
      "entropy": 6.17062406539917,
      "epoch": 0.39150572402342043,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989561316226774,
      "loss": 5.9956,
      "mean_token_accuracy": 0.14837742149829863,
      "num_tokens": 9092814.0,
      "step": 4480
    },
    {
      "entropy": 6.094864559173584,
      "epoch": 0.3919426723761251,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989531313131444,
      "loss": 5.9173,
      "mean_token_accuracy": 0.1641974166035652,
      "num_tokens": 9102816.0,
      "step": 4485
    },
    {
      "entropy": 6.161947250366211,
      "epoch": 0.39237962072882987,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498950126708073,
      "loss": 6.0505,
      "mean_token_accuracy": 0.14533851146697999,
      "num_tokens": 9114470.0,
      "step": 4490
    },
    {
      "entropy": 6.248921823501587,
      "epoch": 0.39281656908153456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989471178075208,
      "loss": 6.0188,
      "mean_token_accuracy": 0.14610825926065446,
      "num_tokens": 9125604.0,
      "step": 4495
    },
    {
      "entropy": 6.162637090682983,
      "epoch": 0.39325351743423925,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989441046115456,
      "loss": 5.9612,
      "mean_token_accuracy": 0.1528365895152092,
      "num_tokens": 9135088.0,
      "step": 4500
    },
    {
      "entropy": 6.176766443252563,
      "epoch": 0.393690465786944,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989410871202051,
      "loss": 6.0274,
      "mean_token_accuracy": 0.15290433764457703,
      "num_tokens": 9145302.0,
      "step": 4505
    },
    {
      "entropy": 6.120792627334595,
      "epoch": 0.3941274141396487,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989380653335571,
      "loss": 5.9584,
      "mean_token_accuracy": 0.14787636175751687,
      "num_tokens": 9155104.0,
      "step": 4510
    },
    {
      "entropy": 6.179185295104981,
      "epoch": 0.3945643624923534,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989350392516598,
      "loss": 5.8919,
      "mean_token_accuracy": 0.15435189753770828,
      "num_tokens": 9165055.0,
      "step": 4515
    },
    {
      "entropy": 6.181280422210693,
      "epoch": 0.39500131084505813,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989320088745711,
      "loss": 5.9214,
      "mean_token_accuracy": 0.15390761345624923,
      "num_tokens": 9175305.0,
      "step": 4520
    },
    {
      "entropy": 6.148918342590332,
      "epoch": 0.3954382591977628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498928974202349,
      "loss": 6.0067,
      "mean_token_accuracy": 0.14799161106348038,
      "num_tokens": 9185153.0,
      "step": 4525
    },
    {
      "entropy": 6.191053152084351,
      "epoch": 0.3958752075504675,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498925935235052,
      "loss": 5.9578,
      "mean_token_accuracy": 0.14903645291924478,
      "num_tokens": 9195385.0,
      "step": 4530
    },
    {
      "entropy": 6.097781085968018,
      "epoch": 0.39631215590317226,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989228919727382,
      "loss": 5.918,
      "mean_token_accuracy": 0.15367165505886077,
      "num_tokens": 9205009.0,
      "step": 4535
    },
    {
      "entropy": 6.150427293777466,
      "epoch": 0.39674910425587695,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498919844415466,
      "loss": 5.9774,
      "mean_token_accuracy": 0.15060033649206161,
      "num_tokens": 9214684.0,
      "step": 4540
    },
    {
      "entropy": 6.154142951965332,
      "epoch": 0.39718605260858164,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989167925632937,
      "loss": 5.9012,
      "mean_token_accuracy": 0.15627232789993287,
      "num_tokens": 9225721.0,
      "step": 4545
    },
    {
      "entropy": 6.164416837692261,
      "epoch": 0.3976230009612864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989137364162801,
      "loss": 6.0028,
      "mean_token_accuracy": 0.15146558433771135,
      "num_tokens": 9236167.0,
      "step": 4550
    },
    {
      "entropy": 6.14252462387085,
      "epoch": 0.3980599493139911,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989106759744837,
      "loss": 5.9523,
      "mean_token_accuracy": 0.1590500593185425,
      "num_tokens": 9245716.0,
      "step": 4555
    },
    {
      "entropy": 6.20049524307251,
      "epoch": 0.39849689766669577,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989076112379633,
      "loss": 6.0289,
      "mean_token_accuracy": 0.14886338859796525,
      "num_tokens": 9256897.0,
      "step": 4560
    },
    {
      "entropy": 6.239367532730102,
      "epoch": 0.3989338460194005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989045422067775,
      "loss": 6.0297,
      "mean_token_accuracy": 0.13837849721312523,
      "num_tokens": 9266080.0,
      "step": 4565
    },
    {
      "entropy": 6.115847444534301,
      "epoch": 0.3993707943721052,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989014688809853,
      "loss": 5.9555,
      "mean_token_accuracy": 0.1502847209572792,
      "num_tokens": 9276815.0,
      "step": 4570
    },
    {
      "entropy": 6.182366514205933,
      "epoch": 0.3998077427248099,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988983912606454,
      "loss": 6.0139,
      "mean_token_accuracy": 0.15305532068014144,
      "num_tokens": 9287339.0,
      "step": 4575
    },
    {
      "entropy": 6.153854656219482,
      "epoch": 0.40024469107751465,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004988953093458173,
      "loss": 5.8975,
      "mean_token_accuracy": 0.15478773266077042,
      "num_tokens": 9298002.0,
      "step": 4580
    },
    {
      "entropy": 6.152257061004638,
      "epoch": 0.40068163943021934,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988922231365597,
      "loss": 6.0206,
      "mean_token_accuracy": 0.14581730216741562,
      "num_tokens": 9309008.0,
      "step": 4585
    },
    {
      "entropy": 6.159885358810425,
      "epoch": 0.4011185877829241,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498889132632932,
      "loss": 6.015,
      "mean_token_accuracy": 0.14740359485149385,
      "num_tokens": 9319391.0,
      "step": 4590
    },
    {
      "entropy": 6.101526498794556,
      "epoch": 0.4015555361356288,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988860378349934,
      "loss": 5.8749,
      "mean_token_accuracy": 0.164150882512331,
      "num_tokens": 9329808.0,
      "step": 4595
    },
    {
      "entropy": 6.161025524139404,
      "epoch": 0.40199248448833347,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988829387428033,
      "loss": 5.9321,
      "mean_token_accuracy": 0.1556175470352173,
      "num_tokens": 9339580.0,
      "step": 4600
    },
    {
      "entropy": 6.200582599639892,
      "epoch": 0.4024294328410382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988798353564211,
      "loss": 5.9585,
      "mean_token_accuracy": 0.15299488455057145,
      "num_tokens": 9350008.0,
      "step": 4605
    },
    {
      "entropy": 6.187353610992432,
      "epoch": 0.4028663811937429,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988767276759064,
      "loss": 5.9473,
      "mean_token_accuracy": 0.15429434180259705,
      "num_tokens": 9359026.0,
      "step": 4610
    },
    {
      "entropy": 6.177741956710816,
      "epoch": 0.4033033295464476,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988736157013186,
      "loss": 5.9179,
      "mean_token_accuracy": 0.15768380463123322,
      "num_tokens": 9369084.0,
      "step": 4615
    },
    {
      "entropy": 6.142515802383423,
      "epoch": 0.40374027789915234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988704994327178,
      "loss": 5.9171,
      "mean_token_accuracy": 0.15947658866643905,
      "num_tokens": 9378273.0,
      "step": 4620
    },
    {
      "entropy": 6.0832109451293945,
      "epoch": 0.40417722625185704,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988673788701633,
      "loss": 5.9426,
      "mean_token_accuracy": 0.15066118240356446,
      "num_tokens": 9388434.0,
      "step": 4625
    },
    {
      "entropy": 6.1788512706756595,
      "epoch": 0.4046141746045617,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988642540137152,
      "loss": 6.0219,
      "mean_token_accuracy": 0.14581380784511566,
      "num_tokens": 9398484.0,
      "step": 4630
    },
    {
      "entropy": 6.147611856460571,
      "epoch": 0.4050511229572665,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988611248634335,
      "loss": 5.8841,
      "mean_token_accuracy": 0.15747999548912048,
      "num_tokens": 9409544.0,
      "step": 4635
    },
    {
      "entropy": 6.174405765533447,
      "epoch": 0.40548807130997117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498857991419378,
      "loss": 5.9637,
      "mean_token_accuracy": 0.15213829576969146,
      "num_tokens": 9419391.0,
      "step": 4640
    },
    {
      "entropy": 6.193379926681518,
      "epoch": 0.40592501966267586,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988548536816091,
      "loss": 5.9198,
      "mean_token_accuracy": 0.15799257457256316,
      "num_tokens": 9429342.0,
      "step": 4645
    },
    {
      "entropy": 6.1595134258270265,
      "epoch": 0.4063619680153806,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988517116501867,
      "loss": 6.0251,
      "mean_token_accuracy": 0.14357172399759294,
      "num_tokens": 9440643.0,
      "step": 4650
    },
    {
      "entropy": 6.208586025238037,
      "epoch": 0.4067989163680853,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988485653251711,
      "loss": 5.9889,
      "mean_token_accuracy": 0.15398359000682832,
      "num_tokens": 9450674.0,
      "step": 4655
    },
    {
      "entropy": 6.206997299194336,
      "epoch": 0.40723586472079,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988454147066228,
      "loss": 5.9458,
      "mean_token_accuracy": 0.15420523881912232,
      "num_tokens": 9460075.0,
      "step": 4660
    },
    {
      "entropy": 6.141565895080566,
      "epoch": 0.40767281307349473,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988422597946022,
      "loss": 5.9389,
      "mean_token_accuracy": 0.15483299493789673,
      "num_tokens": 9469327.0,
      "step": 4665
    },
    {
      "entropy": 6.190423917770386,
      "epoch": 0.4081097614261994,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988391005891698,
      "loss": 5.9858,
      "mean_token_accuracy": 0.14569978713989257,
      "num_tokens": 9478304.0,
      "step": 4670
    },
    {
      "entropy": 6.149021625518799,
      "epoch": 0.4085467097789041,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498835937090386,
      "loss": 5.9091,
      "mean_token_accuracy": 0.15753501802682876,
      "num_tokens": 9487453.0,
      "step": 4675
    },
    {
      "entropy": 6.090920448303223,
      "epoch": 0.40898365813160886,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988327692983116,
      "loss": 5.8657,
      "mean_token_accuracy": 0.1607446402311325,
      "num_tokens": 9496623.0,
      "step": 4680
    },
    {
      "entropy": 6.17754168510437,
      "epoch": 0.40942060648431355,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988295972130076,
      "loss": 5.9998,
      "mean_token_accuracy": 0.14730391502380372,
      "num_tokens": 9506293.0,
      "step": 4685
    },
    {
      "entropy": 6.16938738822937,
      "epoch": 0.40985755483701825,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988264208345346,
      "loss": 5.906,
      "mean_token_accuracy": 0.1563751608133316,
      "num_tokens": 9516127.0,
      "step": 4690
    },
    {
      "entropy": 6.08460168838501,
      "epoch": 0.410294503189723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988232401629535,
      "loss": 5.8909,
      "mean_token_accuracy": 0.15749040096998215,
      "num_tokens": 9525597.0,
      "step": 4695
    },
    {
      "entropy": 6.131860542297363,
      "epoch": 0.4107314515424277,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988200551983254,
      "loss": 5.8736,
      "mean_token_accuracy": 0.15756741911172867,
      "num_tokens": 9535208.0,
      "step": 4700
    },
    {
      "entropy": 6.093962049484253,
      "epoch": 0.4111683998951324,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988168659407113,
      "loss": 5.9242,
      "mean_token_accuracy": 0.16580573320388795,
      "num_tokens": 9545078.0,
      "step": 4705
    },
    {
      "entropy": 6.175525665283203,
      "epoch": 0.4116053482478371,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988136723901725,
      "loss": 5.954,
      "mean_token_accuracy": 0.14602336138486863,
      "num_tokens": 9555074.0,
      "step": 4710
    },
    {
      "entropy": 6.158622169494629,
      "epoch": 0.4120422966005418,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988104745467702,
      "loss": 5.9282,
      "mean_token_accuracy": 0.15268791913986207,
      "num_tokens": 9564777.0,
      "step": 4715
    },
    {
      "entropy": 6.1089740753173825,
      "epoch": 0.4124792449532465,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988072724105657,
      "loss": 5.887,
      "mean_token_accuracy": 0.15127398818731308,
      "num_tokens": 9574551.0,
      "step": 4720
    },
    {
      "entropy": 6.14306411743164,
      "epoch": 0.41291619330595125,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988040659816203,
      "loss": 6.0032,
      "mean_token_accuracy": 0.141208453476429,
      "num_tokens": 9585162.0,
      "step": 4725
    },
    {
      "entropy": 6.18710560798645,
      "epoch": 0.41335314165865594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988008552599958,
      "loss": 5.9084,
      "mean_token_accuracy": 0.15889257490634917,
      "num_tokens": 9595422.0,
      "step": 4730
    },
    {
      "entropy": 6.127224159240723,
      "epoch": 0.41379009001136063,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987976402457536,
      "loss": 5.9164,
      "mean_token_accuracy": 0.16008937060832978,
      "num_tokens": 9604315.0,
      "step": 4735
    },
    {
      "entropy": 6.111091089248657,
      "epoch": 0.4142270383640654,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987944209389555,
      "loss": 5.9422,
      "mean_token_accuracy": 0.15192492827773094,
      "num_tokens": 9614369.0,
      "step": 4740
    },
    {
      "entropy": 6.1870523452758786,
      "epoch": 0.41466398671677007,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498791197339663,
      "loss": 5.9954,
      "mean_token_accuracy": 0.15036482810974122,
      "num_tokens": 9624154.0,
      "step": 4745
    },
    {
      "entropy": 6.194160509109497,
      "epoch": 0.41510093506947476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498787969447938,
      "loss": 5.9755,
      "mean_token_accuracy": 0.1509677991271019,
      "num_tokens": 9634903.0,
      "step": 4750
    },
    {
      "entropy": 6.108608770370483,
      "epoch": 0.4155378834221795,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987847372638427,
      "loss": 6.009,
      "mean_token_accuracy": 0.15711121559143065,
      "num_tokens": 9644873.0,
      "step": 4755
    },
    {
      "entropy": 6.174947357177734,
      "epoch": 0.4159748317748842,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987815007874388,
      "loss": 5.9109,
      "mean_token_accuracy": 0.1611080899834633,
      "num_tokens": 9654661.0,
      "step": 4760
    },
    {
      "entropy": 6.162149047851562,
      "epoch": 0.4164117801275889,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987782600187884,
      "loss": 5.9268,
      "mean_token_accuracy": 0.1611756145954132,
      "num_tokens": 9664980.0,
      "step": 4765
    },
    {
      "entropy": 6.139432430267334,
      "epoch": 0.41684872848029364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987750149579537,
      "loss": 5.9071,
      "mean_token_accuracy": 0.15827879309654236,
      "num_tokens": 9675117.0,
      "step": 4770
    },
    {
      "entropy": 6.1090240478515625,
      "epoch": 0.41728567683299833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498771765604997,
      "loss": 5.9595,
      "mean_token_accuracy": 0.1518459990620613,
      "num_tokens": 9684900.0,
      "step": 4775
    },
    {
      "entropy": 6.210887861251831,
      "epoch": 0.417722625185703,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987685119599806,
      "loss": 5.9533,
      "mean_token_accuracy": 0.1501029685139656,
      "num_tokens": 9694586.0,
      "step": 4780
    },
    {
      "entropy": 6.1544492721557615,
      "epoch": 0.41815957353840777,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987652540229668,
      "loss": 5.961,
      "mean_token_accuracy": 0.15511146634817125,
      "num_tokens": 9705676.0,
      "step": 4785
    },
    {
      "entropy": 6.174813652038575,
      "epoch": 0.41859652189111246,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987619917940183,
      "loss": 5.9674,
      "mean_token_accuracy": 0.1539301872253418,
      "num_tokens": 9715442.0,
      "step": 4790
    },
    {
      "entropy": 6.244969129562378,
      "epoch": 0.4190334702438172,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004987587252731975,
      "loss": 6.0405,
      "mean_token_accuracy": 0.15172669664025307,
      "num_tokens": 9726621.0,
      "step": 4795
    },
    {
      "entropy": 6.172503280639648,
      "epoch": 0.4194704185965219,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987554544605671,
      "loss": 6.0549,
      "mean_token_accuracy": 0.1454879254102707,
      "num_tokens": 9736688.0,
      "step": 4800
    },
    {
      "entropy": 6.156451320648193,
      "epoch": 0.4199073669492266,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987521793561898,
      "loss": 5.9404,
      "mean_token_accuracy": 0.15448848754167557,
      "num_tokens": 9746478.0,
      "step": 4805
    },
    {
      "entropy": 6.186788892745971,
      "epoch": 0.42034431530193134,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987488999601285,
      "loss": 5.9554,
      "mean_token_accuracy": 0.15627822205424308,
      "num_tokens": 9756370.0,
      "step": 4810
    },
    {
      "entropy": 6.1910998821258545,
      "epoch": 0.42078126365463603,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498745616272446,
      "loss": 5.9713,
      "mean_token_accuracy": 0.14973593205213548,
      "num_tokens": 9766293.0,
      "step": 4815
    },
    {
      "entropy": 6.139960718154907,
      "epoch": 0.4212182120073407,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987423282932054,
      "loss": 5.9448,
      "mean_token_accuracy": 0.15327902287244796,
      "num_tokens": 9777172.0,
      "step": 4820
    },
    {
      "entropy": 6.119137191772461,
      "epoch": 0.42165516036004547,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987390360224697,
      "loss": 5.9174,
      "mean_token_accuracy": 0.15604062974452973,
      "num_tokens": 9787194.0,
      "step": 4825
    },
    {
      "entropy": 6.108503627777099,
      "epoch": 0.42209210871275016,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498735739460302,
      "loss": 5.9529,
      "mean_token_accuracy": 0.15559342354536057,
      "num_tokens": 9797528.0,
      "step": 4830
    },
    {
      "entropy": 6.174628591537475,
      "epoch": 0.42252905706545485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987324386067656,
      "loss": 5.9744,
      "mean_token_accuracy": 0.1515772745013237,
      "num_tokens": 9808412.0,
      "step": 4835
    },
    {
      "entropy": 6.235812091827393,
      "epoch": 0.4229660054181596,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987291334619239,
      "loss": 6.0525,
      "mean_token_accuracy": 0.14427994340658187,
      "num_tokens": 9820047.0,
      "step": 4840
    },
    {
      "entropy": 6.166102123260498,
      "epoch": 0.4234029537708643,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049872582402584,
      "loss": 5.8772,
      "mean_token_accuracy": 0.15683565884828568,
      "num_tokens": 9828812.0,
      "step": 4845
    },
    {
      "entropy": 6.114846515655517,
      "epoch": 0.423839902123569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987225102985778,
      "loss": 5.8991,
      "mean_token_accuracy": 0.15961357057094575,
      "num_tokens": 9838288.0,
      "step": 4850
    },
    {
      "entropy": 6.232522010803223,
      "epoch": 0.4242768504762737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987191922802003,
      "loss": 5.937,
      "mean_token_accuracy": 0.15296330749988557,
      "num_tokens": 9848536.0,
      "step": 4855
    },
    {
      "entropy": 6.092264032363891,
      "epoch": 0.4247137988289784,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987158699707717,
      "loss": 5.9789,
      "mean_token_accuracy": 0.1538124531507492,
      "num_tokens": 9859741.0,
      "step": 4860
    },
    {
      "entropy": 6.21048583984375,
      "epoch": 0.4251507471816831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987125433703554,
      "loss": 5.9311,
      "mean_token_accuracy": 0.15274728387594222,
      "num_tokens": 9869173.0,
      "step": 4865
    },
    {
      "entropy": 6.186399030685425,
      "epoch": 0.42558769553438786,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987092124790155,
      "loss": 5.9571,
      "mean_token_accuracy": 0.1524939864873886,
      "num_tokens": 9879619.0,
      "step": 4870
    },
    {
      "entropy": 6.1053039073944095,
      "epoch": 0.42602464388709255,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987058772968154,
      "loss": 5.8715,
      "mean_token_accuracy": 0.16609814912080764,
      "num_tokens": 9890214.0,
      "step": 4875
    },
    {
      "entropy": 6.105354118347168,
      "epoch": 0.42646159223979724,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987025378238195,
      "loss": 5.9146,
      "mean_token_accuracy": 0.14937784671783447,
      "num_tokens": 9900266.0,
      "step": 4880
    },
    {
      "entropy": 6.208313083648681,
      "epoch": 0.426898540592502,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986991940600917,
      "loss": 5.9327,
      "mean_token_accuracy": 0.15678722262382508,
      "num_tokens": 9910790.0,
      "step": 4885
    },
    {
      "entropy": 6.191554117202759,
      "epoch": 0.4273354889452067,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986958460056961,
      "loss": 5.9584,
      "mean_token_accuracy": 0.15082679465413093,
      "num_tokens": 9921167.0,
      "step": 4890
    },
    {
      "entropy": 6.096292352676391,
      "epoch": 0.42777243729791137,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498692493660697,
      "loss": 5.8915,
      "mean_token_accuracy": 0.15671937689185142,
      "num_tokens": 9931674.0,
      "step": 4895
    },
    {
      "entropy": 6.137964963912964,
      "epoch": 0.4282093856506161,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986891370251587,
      "loss": 5.8499,
      "mean_token_accuracy": 0.159038682281971,
      "num_tokens": 9941484.0,
      "step": 4900
    },
    {
      "entropy": 6.172023105621338,
      "epoch": 0.4286463340033208,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986857760991454,
      "loss": 5.9994,
      "mean_token_accuracy": 0.14832809716463088,
      "num_tokens": 9951885.0,
      "step": 4905
    },
    {
      "entropy": 6.144092607498169,
      "epoch": 0.4290832823560255,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986824108827218,
      "loss": 5.8368,
      "mean_token_accuracy": 0.1598619341850281,
      "num_tokens": 9961658.0,
      "step": 4910
    },
    {
      "entropy": 6.061906051635742,
      "epoch": 0.42952023070873024,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986790413759523,
      "loss": 5.9281,
      "mean_token_accuracy": 0.1530425012111664,
      "num_tokens": 9971971.0,
      "step": 4915
    },
    {
      "entropy": 6.112272930145264,
      "epoch": 0.42995717906143494,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986756675789016,
      "loss": 5.8381,
      "mean_token_accuracy": 0.15551452934741974,
      "num_tokens": 9982211.0,
      "step": 4920
    },
    {
      "entropy": 6.13322172164917,
      "epoch": 0.4303941274141396,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986722894916344,
      "loss": 5.8929,
      "mean_token_accuracy": 0.15656503289937973,
      "num_tokens": 9991935.0,
      "step": 4925
    },
    {
      "entropy": 6.091841220855713,
      "epoch": 0.4308310757668444,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986689071142155,
      "loss": 5.932,
      "mean_token_accuracy": 0.14894939213991165,
      "num_tokens": 10002014.0,
      "step": 4930
    },
    {
      "entropy": 6.134459018707275,
      "epoch": 0.43126802411954906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004986655204467097,
      "loss": 5.9531,
      "mean_token_accuracy": 0.1539854198694229,
      "num_tokens": 10012030.0,
      "step": 4935
    },
    {
      "entropy": 6.146116018295288,
      "epoch": 0.43170497247225376,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986621294891822,
      "loss": 5.9246,
      "mean_token_accuracy": 0.1548383966088295,
      "num_tokens": 10022188.0,
      "step": 4940
    },
    {
      "entropy": 6.159297180175781,
      "epoch": 0.4321419208249585,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986587342416976,
      "loss": 5.831,
      "mean_token_accuracy": 0.16061824560165405,
      "num_tokens": 10032131.0,
      "step": 4945
    },
    {
      "entropy": 6.116433811187744,
      "epoch": 0.4325788691776632,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986553347043214,
      "loss": 5.8737,
      "mean_token_accuracy": 0.15654212534427642,
      "num_tokens": 10043204.0,
      "step": 4950
    },
    {
      "entropy": 6.1559295654296875,
      "epoch": 0.4330158175303679,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986519308771188,
      "loss": 5.9514,
      "mean_token_accuracy": 0.15065933018922806,
      "num_tokens": 10052863.0,
      "step": 4955
    },
    {
      "entropy": 6.13035478591919,
      "epoch": 0.43345276588307263,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986485227601548,
      "loss": 5.9358,
      "mean_token_accuracy": 0.1590741291642189,
      "num_tokens": 10063212.0,
      "step": 4960
    },
    {
      "entropy": 6.174924516677857,
      "epoch": 0.4338897142357773,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498645110353495,
      "loss": 5.9337,
      "mean_token_accuracy": 0.15222507566213608,
      "num_tokens": 10074180.0,
      "step": 4965
    },
    {
      "entropy": 6.0647026062011715,
      "epoch": 0.434326662588482,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986416936572049,
      "loss": 5.8962,
      "mean_token_accuracy": 0.15867372453212739,
      "num_tokens": 10084513.0,
      "step": 4970
    },
    {
      "entropy": 6.128321981430053,
      "epoch": 0.43476361094118676,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986382726713497,
      "loss": 5.9216,
      "mean_token_accuracy": 0.15316855609416963,
      "num_tokens": 10094352.0,
      "step": 4975
    },
    {
      "entropy": 6.034129476547241,
      "epoch": 0.43520055929389145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986348473959955,
      "loss": 5.7804,
      "mean_token_accuracy": 0.16874276250600814,
      "num_tokens": 10103677.0,
      "step": 4980
    },
    {
      "entropy": 6.1139387607574465,
      "epoch": 0.43563750764659614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986314178312076,
      "loss": 5.9922,
      "mean_token_accuracy": 0.14352853894233703,
      "num_tokens": 10113479.0,
      "step": 4985
    },
    {
      "entropy": 6.145417070388794,
      "epoch": 0.4360744559993009,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986279839770519,
      "loss": 5.8885,
      "mean_token_accuracy": 0.15642954260110856,
      "num_tokens": 10123526.0,
      "step": 4990
    },
    {
      "entropy": 6.1418062210083,
      "epoch": 0.4365114043520056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986245458335944,
      "loss": 5.9082,
      "mean_token_accuracy": 0.15344171673059465,
      "num_tokens": 10134176.0,
      "step": 4995
    },
    {
      "entropy": 6.10381669998169,
      "epoch": 0.43694835270471033,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986211034009008,
      "loss": 5.8741,
      "mean_token_accuracy": 0.16505304872989654,
      "num_tokens": 10144763.0,
      "step": 5000
    },
    {
      "entropy": 6.1019261360168455,
      "epoch": 0.437385301057415,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986176566790375,
      "loss": 5.8575,
      "mean_token_accuracy": 0.1642833322286606,
      "num_tokens": 10155011.0,
      "step": 5005
    },
    {
      "entropy": 6.162095165252685,
      "epoch": 0.4378222494101197,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986142056680701,
      "loss": 5.9369,
      "mean_token_accuracy": 0.15561369210481643,
      "num_tokens": 10165474.0,
      "step": 5010
    },
    {
      "entropy": 6.083307552337646,
      "epoch": 0.43825919776282446,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986107503680652,
      "loss": 5.9439,
      "mean_token_accuracy": 0.1566638946533203,
      "num_tokens": 10175137.0,
      "step": 5015
    },
    {
      "entropy": 6.176087284088135,
      "epoch": 0.43869614611552915,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986072907790889,
      "loss": 5.9457,
      "mean_token_accuracy": 0.1491002082824707,
      "num_tokens": 10185623.0,
      "step": 5020
    },
    {
      "entropy": 6.183105421066284,
      "epoch": 0.43913309446823384,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986038269012076,
      "loss": 5.9162,
      "mean_token_accuracy": 0.15544091016054154,
      "num_tokens": 10195201.0,
      "step": 5025
    },
    {
      "entropy": 6.066984462738037,
      "epoch": 0.4395700428209386,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986003587344877,
      "loss": 5.9535,
      "mean_token_accuracy": 0.15503553450107574,
      "num_tokens": 10205951.0,
      "step": 5030
    },
    {
      "entropy": 6.1341217994689945,
      "epoch": 0.4400069911736433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985968862789957,
      "loss": 5.8291,
      "mean_token_accuracy": 0.15822890102863313,
      "num_tokens": 10215665.0,
      "step": 5035
    },
    {
      "entropy": 6.098297786712647,
      "epoch": 0.44044393952634797,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985934095347985,
      "loss": 5.9283,
      "mean_token_accuracy": 0.1514671616256237,
      "num_tokens": 10225037.0,
      "step": 5040
    },
    {
      "entropy": 6.057718515396118,
      "epoch": 0.4408808878790527,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985899285019624,
      "loss": 5.8744,
      "mean_token_accuracy": 0.15146672129631042,
      "num_tokens": 10234581.0,
      "step": 5045
    },
    {
      "entropy": 6.151702833175659,
      "epoch": 0.4413178362317574,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985864431805543,
      "loss": 5.7939,
      "mean_token_accuracy": 0.16405095309019088,
      "num_tokens": 10243716.0,
      "step": 5050
    },
    {
      "entropy": 6.056149816513061,
      "epoch": 0.4417547845844621,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985829535706411,
      "loss": 5.9023,
      "mean_token_accuracy": 0.1601075679063797,
      "num_tokens": 10253198.0,
      "step": 5055
    },
    {
      "entropy": 6.083074235916138,
      "epoch": 0.44219173293716685,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985794596722897,
      "loss": 5.9784,
      "mean_token_accuracy": 0.15283442437648773,
      "num_tokens": 10262912.0,
      "step": 5060
    },
    {
      "entropy": 6.127580738067627,
      "epoch": 0.44262868128987154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498575961485567,
      "loss": 5.9198,
      "mean_token_accuracy": 0.1554662726819515,
      "num_tokens": 10272960.0,
      "step": 5065
    },
    {
      "entropy": 6.175092029571533,
      "epoch": 0.44306562964257623,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985724590105403,
      "loss": 5.9595,
      "mean_token_accuracy": 0.15449291169643403,
      "num_tokens": 10283034.0,
      "step": 5070
    },
    {
      "entropy": 6.10110650062561,
      "epoch": 0.443502577995281,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985689522472767,
      "loss": 5.9254,
      "mean_token_accuracy": 0.15283768624067307,
      "num_tokens": 10292553.0,
      "step": 5075
    },
    {
      "entropy": 6.053380155563355,
      "epoch": 0.44393952634798567,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985654411958434,
      "loss": 5.8794,
      "mean_token_accuracy": 0.15470169633626937,
      "num_tokens": 10302475.0,
      "step": 5080
    },
    {
      "entropy": 6.128862476348877,
      "epoch": 0.44437647470069036,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985619258563078,
      "loss": 5.8967,
      "mean_token_accuracy": 0.15939270108938217,
      "num_tokens": 10311693.0,
      "step": 5085
    },
    {
      "entropy": 6.142067003250122,
      "epoch": 0.4448134230533951,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985584062287372,
      "loss": 5.8934,
      "mean_token_accuracy": 0.16058653146028518,
      "num_tokens": 10323052.0,
      "step": 5090
    },
    {
      "entropy": 6.0882007598876955,
      "epoch": 0.4452503714060998,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985548823131994,
      "loss": 5.8602,
      "mean_token_accuracy": 0.16040343791246414,
      "num_tokens": 10332997.0,
      "step": 5095
    },
    {
      "entropy": 6.212076568603516,
      "epoch": 0.4456873197588045,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985513541097618,
      "loss": 5.9966,
      "mean_token_accuracy": 0.15025889053940772,
      "num_tokens": 10343298.0,
      "step": 5100
    },
    {
      "entropy": 6.206186008453369,
      "epoch": 0.44612426811150924,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498547821618492,
      "loss": 5.879,
      "mean_token_accuracy": 0.15802157968282698,
      "num_tokens": 10352485.0,
      "step": 5105
    },
    {
      "entropy": 5.995059537887573,
      "epoch": 0.44656121646421393,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985442848394579,
      "loss": 5.9236,
      "mean_token_accuracy": 0.15586542040109636,
      "num_tokens": 10362810.0,
      "step": 5110
    },
    {
      "entropy": 6.153477907180786,
      "epoch": 0.4469981648169186,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985407437727273,
      "loss": 5.8714,
      "mean_token_accuracy": 0.1590302214026451,
      "num_tokens": 10372817.0,
      "step": 5115
    },
    {
      "entropy": 6.190938520431518,
      "epoch": 0.44743511316962337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985371984183681,
      "loss": 5.9033,
      "mean_token_accuracy": 0.1612114742398262,
      "num_tokens": 10383046.0,
      "step": 5120
    },
    {
      "entropy": 6.073680734634399,
      "epoch": 0.44787206152232806,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985336487764483,
      "loss": 5.9731,
      "mean_token_accuracy": 0.15040762424468995,
      "num_tokens": 10393498.0,
      "step": 5125
    },
    {
      "entropy": 6.1484394550323485,
      "epoch": 0.44830900987503275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985300948470359,
      "loss": 5.9554,
      "mean_token_accuracy": 0.1554347351193428,
      "num_tokens": 10403595.0,
      "step": 5130
    },
    {
      "entropy": 6.1428296089172365,
      "epoch": 0.4487459582277375,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985265366301992,
      "loss": 5.8883,
      "mean_token_accuracy": 0.15487209260463713,
      "num_tokens": 10413475.0,
      "step": 5135
    },
    {
      "entropy": 6.068651247024536,
      "epoch": 0.4491829065804422,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985229741260065,
      "loss": 5.9181,
      "mean_token_accuracy": 0.16206053793430328,
      "num_tokens": 10423512.0,
      "step": 5140
    },
    {
      "entropy": 6.130358648300171,
      "epoch": 0.4496198549331469,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985194073345259,
      "loss": 5.9421,
      "mean_token_accuracy": 0.15074676126241685,
      "num_tokens": 10434198.0,
      "step": 5145
    },
    {
      "entropy": 6.162382888793945,
      "epoch": 0.4500568032858516,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985158362558261,
      "loss": 5.8602,
      "mean_token_accuracy": 0.1571664348244667,
      "num_tokens": 10444077.0,
      "step": 5150
    },
    {
      "entropy": 6.076663827896118,
      "epoch": 0.4504937516385563,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985122608899752,
      "loss": 5.9198,
      "mean_token_accuracy": 0.15442877262830734,
      "num_tokens": 10454481.0,
      "step": 5155
    },
    {
      "entropy": 6.146946859359741,
      "epoch": 0.450930699991261,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985086812370422,
      "loss": 5.9325,
      "mean_token_accuracy": 0.15687772929668425,
      "num_tokens": 10464453.0,
      "step": 5160
    },
    {
      "entropy": 6.174408197402954,
      "epoch": 0.45136764834396575,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985050972970956,
      "loss": 5.9944,
      "mean_token_accuracy": 0.15035201460123063,
      "num_tokens": 10474932.0,
      "step": 5165
    },
    {
      "entropy": 6.173358964920044,
      "epoch": 0.45180459669667045,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498501509070204,
      "loss": 5.9866,
      "mean_token_accuracy": 0.1478325292468071,
      "num_tokens": 10484719.0,
      "step": 5170
    },
    {
      "entropy": 6.180629014968872,
      "epoch": 0.45224154504937514,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984979165564364,
      "loss": 6.0323,
      "mean_token_accuracy": 0.14743464440107346,
      "num_tokens": 10495808.0,
      "step": 5175
    },
    {
      "entropy": 6.233003568649292,
      "epoch": 0.4526784934020799,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984943197558616,
      "loss": 5.9361,
      "mean_token_accuracy": 0.16189470887184143,
      "num_tokens": 10506528.0,
      "step": 5180
    },
    {
      "entropy": 6.101835107803344,
      "epoch": 0.4531154417547846,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984907186685487,
      "loss": 5.9131,
      "mean_token_accuracy": 0.1625828981399536,
      "num_tokens": 10517117.0,
      "step": 5185
    },
    {
      "entropy": 6.191213703155517,
      "epoch": 0.45355239010748927,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984871132945667,
      "loss": 5.9092,
      "mean_token_accuracy": 0.15359443575143814,
      "num_tokens": 10527394.0,
      "step": 5190
    },
    {
      "entropy": 6.116603422164917,
      "epoch": 0.453989338460194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984835036339848,
      "loss": 5.9176,
      "mean_token_accuracy": 0.15361591726541518,
      "num_tokens": 10536779.0,
      "step": 5195
    },
    {
      "entropy": 6.096918821334839,
      "epoch": 0.4544262868128987,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984798896868722,
      "loss": 5.9923,
      "mean_token_accuracy": 0.1439100369811058,
      "num_tokens": 10547116.0,
      "step": 5200
    },
    {
      "entropy": 6.139439344406128,
      "epoch": 0.45486323516560345,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984762714532981,
      "loss": 5.8604,
      "mean_token_accuracy": 0.1621207997202873,
      "num_tokens": 10556396.0,
      "step": 5205
    },
    {
      "entropy": 6.127708673477173,
      "epoch": 0.45530018351830814,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498472648933332,
      "loss": 6.0415,
      "mean_token_accuracy": 0.15158578306436538,
      "num_tokens": 10568127.0,
      "step": 5210
    },
    {
      "entropy": 6.200006580352783,
      "epoch": 0.45573713187101283,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984690221270436,
      "loss": 5.8647,
      "mean_token_accuracy": 0.15969650596380233,
      "num_tokens": 10578565.0,
      "step": 5215
    },
    {
      "entropy": 6.098275470733642,
      "epoch": 0.4561740802237176,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984653910345021,
      "loss": 5.9067,
      "mean_token_accuracy": 0.1467830255627632,
      "num_tokens": 10588633.0,
      "step": 5220
    },
    {
      "entropy": 6.143524599075318,
      "epoch": 0.4566110285764223,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984617556557774,
      "loss": 5.8586,
      "mean_token_accuracy": 0.16036575734615327,
      "num_tokens": 10597917.0,
      "step": 5225
    },
    {
      "entropy": 6.176416683197021,
      "epoch": 0.45704797692912696,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498458115990939,
      "loss": 5.8991,
      "mean_token_accuracy": 0.14881471544504166,
      "num_tokens": 10608396.0,
      "step": 5230
    },
    {
      "entropy": 6.088785552978516,
      "epoch": 0.4574849252818317,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984544720400568,
      "loss": 5.9131,
      "mean_token_accuracy": 0.15947931855916977,
      "num_tokens": 10618191.0,
      "step": 5235
    },
    {
      "entropy": 6.127312564849854,
      "epoch": 0.4579218736345364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498450823803201,
      "loss": 5.9089,
      "mean_token_accuracy": 0.15233112201094628,
      "num_tokens": 10627594.0,
      "step": 5240
    },
    {
      "entropy": 6.087362957000733,
      "epoch": 0.4583588219872411,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984471712804412,
      "loss": 5.9016,
      "mean_token_accuracy": 0.16228167116641998,
      "num_tokens": 10637260.0,
      "step": 5245
    },
    {
      "entropy": 6.125209045410156,
      "epoch": 0.45879577033994584,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984435144718475,
      "loss": 5.9277,
      "mean_token_accuracy": 0.15509730130434035,
      "num_tokens": 10647968.0,
      "step": 5250
    },
    {
      "entropy": 6.099068117141724,
      "epoch": 0.45923271869265053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984398533774901,
      "loss": 5.8612,
      "mean_token_accuracy": 0.15810431987047197,
      "num_tokens": 10657760.0,
      "step": 5255
    },
    {
      "entropy": 6.104465675354004,
      "epoch": 0.4596696670453552,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984361879974393,
      "loss": 5.8855,
      "mean_token_accuracy": 0.15050318390130996,
      "num_tokens": 10667577.0,
      "step": 5260
    },
    {
      "entropy": 6.0608482837677,
      "epoch": 0.46010661539805997,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004984325183317654,
      "loss": 5.8367,
      "mean_token_accuracy": 0.1569499745965004,
      "num_tokens": 10677984.0,
      "step": 5265
    },
    {
      "entropy": 6.089433002471924,
      "epoch": 0.46054356375076466,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984288443805386,
      "loss": 5.8981,
      "mean_token_accuracy": 0.16103835329413413,
      "num_tokens": 10687700.0,
      "step": 5270
    },
    {
      "entropy": 6.135640239715576,
      "epoch": 0.46098051210346935,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984251661438295,
      "loss": 5.8809,
      "mean_token_accuracy": 0.1602894976735115,
      "num_tokens": 10697351.0,
      "step": 5275
    },
    {
      "entropy": 6.079326105117798,
      "epoch": 0.4614174604561741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984214836217087,
      "loss": 5.909,
      "mean_token_accuracy": 0.15437911823391914,
      "num_tokens": 10707498.0,
      "step": 5280
    },
    {
      "entropy": 6.09013409614563,
      "epoch": 0.4618544088088788,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984177968142467,
      "loss": 5.9378,
      "mean_token_accuracy": 0.15395788848400116,
      "num_tokens": 10717724.0,
      "step": 5285
    },
    {
      "entropy": 6.160751914978027,
      "epoch": 0.4622913571615835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984141057215143,
      "loss": 5.8914,
      "mean_token_accuracy": 0.1560735285282135,
      "num_tokens": 10727030.0,
      "step": 5290
    },
    {
      "entropy": 6.053331899642944,
      "epoch": 0.46272830551428823,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984104103435824,
      "loss": 5.7835,
      "mean_token_accuracy": 0.16150631457567216,
      "num_tokens": 10736861.0,
      "step": 5295
    },
    {
      "entropy": 6.17409873008728,
      "epoch": 0.4631652538669929,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984067106805217,
      "loss": 5.9996,
      "mean_token_accuracy": 0.15002165287733077,
      "num_tokens": 10748310.0,
      "step": 5300
    },
    {
      "entropy": 6.12554702758789,
      "epoch": 0.4636022022196976,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984030067324032,
      "loss": 5.8491,
      "mean_token_accuracy": 0.158063605427742,
      "num_tokens": 10758110.0,
      "step": 5305
    },
    {
      "entropy": 6.061292886734009,
      "epoch": 0.46403915057240236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983992984992979,
      "loss": 5.8965,
      "mean_token_accuracy": 0.1623566895723343,
      "num_tokens": 10767919.0,
      "step": 5310
    },
    {
      "entropy": 6.077050590515137,
      "epoch": 0.46447609892510705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983955859812771,
      "loss": 5.8248,
      "mean_token_accuracy": 0.16128846853971482,
      "num_tokens": 10776888.0,
      "step": 5315
    },
    {
      "entropy": 6.138931465148926,
      "epoch": 0.46491304727781174,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983918691784119,
      "loss": 5.8926,
      "mean_token_accuracy": 0.15311199575662612,
      "num_tokens": 10786806.0,
      "step": 5320
    },
    {
      "entropy": 6.1371315002441404,
      "epoch": 0.4653499956305165,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983881480907736,
      "loss": 5.8822,
      "mean_token_accuracy": 0.15748362690210344,
      "num_tokens": 10796452.0,
      "step": 5325
    },
    {
      "entropy": 6.111987686157226,
      "epoch": 0.4657869439832212,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983844227184334,
      "loss": 5.8497,
      "mean_token_accuracy": 0.15752890110015869,
      "num_tokens": 10806718.0,
      "step": 5330
    },
    {
      "entropy": 6.120180749893189,
      "epoch": 0.46622389233592587,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983806930614631,
      "loss": 5.9378,
      "mean_token_accuracy": 0.15134639739990235,
      "num_tokens": 10817978.0,
      "step": 5335
    },
    {
      "entropy": 6.073476648330688,
      "epoch": 0.4666608406886306,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000498376959119934,
      "loss": 5.8063,
      "mean_token_accuracy": 0.16518545597791673,
      "num_tokens": 10827818.0,
      "step": 5340
    },
    {
      "entropy": 6.137352514266968,
      "epoch": 0.4670977890413353,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983732208939178,
      "loss": 5.9518,
      "mean_token_accuracy": 0.15281863436102866,
      "num_tokens": 10838492.0,
      "step": 5345
    },
    {
      "entropy": 6.134186792373657,
      "epoch": 0.46753473739404,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983694783834863,
      "loss": 5.8519,
      "mean_token_accuracy": 0.1597207710146904,
      "num_tokens": 10847807.0,
      "step": 5350
    },
    {
      "entropy": 6.070673322677612,
      "epoch": 0.46797168574674475,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983657315887109,
      "loss": 5.9378,
      "mean_token_accuracy": 0.15434618294239044,
      "num_tokens": 10858637.0,
      "step": 5355
    },
    {
      "entropy": 6.172582578659058,
      "epoch": 0.46840863409944944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983619805096639,
      "loss": 5.8648,
      "mean_token_accuracy": 0.16046326905488967,
      "num_tokens": 10868042.0,
      "step": 5360
    },
    {
      "entropy": 6.111141014099121,
      "epoch": 0.46884558245215413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983582251464171,
      "loss": 5.8721,
      "mean_token_accuracy": 0.15852262377738952,
      "num_tokens": 10877171.0,
      "step": 5365
    },
    {
      "entropy": 6.051455116271972,
      "epoch": 0.4692825308048589,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983544654990423,
      "loss": 5.8711,
      "mean_token_accuracy": 0.15675054788589476,
      "num_tokens": 10888434.0,
      "step": 5370
    },
    {
      "entropy": 6.120969200134278,
      "epoch": 0.46971947915756357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498350701567612,
      "loss": 5.9098,
      "mean_token_accuracy": 0.1561329334974289,
      "num_tokens": 10898467.0,
      "step": 5375
    },
    {
      "entropy": 6.073193788528442,
      "epoch": 0.47015642751026826,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983469333521983,
      "loss": 5.8794,
      "mean_token_accuracy": 0.16295347958803177,
      "num_tokens": 10908178.0,
      "step": 5380
    },
    {
      "entropy": 6.083098745346069,
      "epoch": 0.470593375862973,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983431608528732,
      "loss": 5.948,
      "mean_token_accuracy": 0.1501172386109829,
      "num_tokens": 10919957.0,
      "step": 5385
    },
    {
      "entropy": 6.165816974639893,
      "epoch": 0.4710303242156777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983393840697092,
      "loss": 5.9192,
      "mean_token_accuracy": 0.15550798326730728,
      "num_tokens": 10929728.0,
      "step": 5390
    },
    {
      "entropy": 6.129124307632447,
      "epoch": 0.4714672725683824,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983356030027789,
      "loss": 5.9484,
      "mean_token_accuracy": 0.15565953701734542,
      "num_tokens": 10940427.0,
      "step": 5395
    },
    {
      "entropy": 6.038702344894409,
      "epoch": 0.47190422092108714,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983318176521546,
      "loss": 5.8998,
      "mean_token_accuracy": 0.15754138231277465,
      "num_tokens": 10950641.0,
      "step": 5400
    },
    {
      "entropy": 6.172031879425049,
      "epoch": 0.4723411692737918,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983280280179092,
      "loss": 5.85,
      "mean_token_accuracy": 0.15937471091747285,
      "num_tokens": 10959341.0,
      "step": 5405
    },
    {
      "entropy": 6.112655210494995,
      "epoch": 0.4727781176264966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498324234100115,
      "loss": 5.8899,
      "mean_token_accuracy": 0.15866801738739014,
      "num_tokens": 10967745.0,
      "step": 5410
    },
    {
      "entropy": 6.0321601867675785,
      "epoch": 0.47321506597920127,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983204358988451,
      "loss": 5.847,
      "mean_token_accuracy": 0.15638446807861328,
      "num_tokens": 10978345.0,
      "step": 5415
    },
    {
      "entropy": 6.149096155166626,
      "epoch": 0.47365201433190596,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983166334141722,
      "loss": 5.9119,
      "mean_token_accuracy": 0.1516627289354801,
      "num_tokens": 10988030.0,
      "step": 5420
    },
    {
      "entropy": 6.156317710876465,
      "epoch": 0.4740889626846107,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983128266461693,
      "loss": 5.8953,
      "mean_token_accuracy": 0.15026577562093735,
      "num_tokens": 10999169.0,
      "step": 5425
    },
    {
      "entropy": 6.077851963043213,
      "epoch": 0.4745259110373154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983090155949093,
      "loss": 5.9727,
      "mean_token_accuracy": 0.1535848781466484,
      "num_tokens": 11009608.0,
      "step": 5430
    },
    {
      "entropy": 6.142884111404419,
      "epoch": 0.4749628593900201,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983052002604653,
      "loss": 5.8983,
      "mean_token_accuracy": 0.15391942262649536,
      "num_tokens": 11020187.0,
      "step": 5435
    },
    {
      "entropy": 6.102903079986572,
      "epoch": 0.47539980774272483,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983013806429107,
      "loss": 5.8729,
      "mean_token_accuracy": 0.15639501586556434,
      "num_tokens": 11030459.0,
      "step": 5440
    },
    {
      "entropy": 6.005454635620117,
      "epoch": 0.4758367560954295,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982975567423185,
      "loss": 5.826,
      "mean_token_accuracy": 0.16008418500423433,
      "num_tokens": 11040880.0,
      "step": 5445
    },
    {
      "entropy": 6.111983442306519,
      "epoch": 0.4762737044481342,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982937285587623,
      "loss": 5.8473,
      "mean_token_accuracy": 0.16507836133241655,
      "num_tokens": 11050901.0,
      "step": 5450
    },
    {
      "entropy": 6.130705165863037,
      "epoch": 0.47671065280083896,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982898960923152,
      "loss": 5.9982,
      "mean_token_accuracy": 0.14802106320858002,
      "num_tokens": 11061274.0,
      "step": 5455
    },
    {
      "entropy": 6.114676237106323,
      "epoch": 0.47714760115354365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982860593430509,
      "loss": 5.8204,
      "mean_token_accuracy": 0.15441938787698745,
      "num_tokens": 11070644.0,
      "step": 5460
    },
    {
      "entropy": 6.078702688217163,
      "epoch": 0.47758454950624835,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982822183110431,
      "loss": 5.8771,
      "mean_token_accuracy": 0.1607626348733902,
      "num_tokens": 11081154.0,
      "step": 5465
    },
    {
      "entropy": 6.141170310974121,
      "epoch": 0.4780214978589531,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982783729963654,
      "loss": 5.9763,
      "mean_token_accuracy": 0.15303587913513184,
      "num_tokens": 11091929.0,
      "step": 5470
    },
    {
      "entropy": 6.007306289672852,
      "epoch": 0.4784584462116578,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982745233990914,
      "loss": 5.8097,
      "mean_token_accuracy": 0.16833598613739015,
      "num_tokens": 11102075.0,
      "step": 5475
    },
    {
      "entropy": 6.106387615203857,
      "epoch": 0.4788953945643625,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982706695192949,
      "loss": 5.9001,
      "mean_token_accuracy": 0.154899200797081,
      "num_tokens": 11111886.0,
      "step": 5480
    },
    {
      "entropy": 6.1208422660827635,
      "epoch": 0.4793323429170672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982668113570503,
      "loss": 5.8967,
      "mean_token_accuracy": 0.15136880427598953,
      "num_tokens": 11122329.0,
      "step": 5485
    },
    {
      "entropy": 6.128077554702759,
      "epoch": 0.4797692912697719,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498262948912431,
      "loss": 5.8202,
      "mean_token_accuracy": 0.16018953621387483,
      "num_tokens": 11132348.0,
      "step": 5490
    },
    {
      "entropy": 6.087352991104126,
      "epoch": 0.4802062396224766,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982590821855115,
      "loss": 5.8882,
      "mean_token_accuracy": 0.15972235947847366,
      "num_tokens": 11141975.0,
      "step": 5495
    },
    {
      "entropy": 6.1300702571868895,
      "epoch": 0.48064318797518135,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982552111763657,
      "loss": 5.8993,
      "mean_token_accuracy": 0.15259432792663574,
      "num_tokens": 11152174.0,
      "step": 5500
    },
    {
      "entropy": 6.0827539443969725,
      "epoch": 0.48108013632788604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982513358850681,
      "loss": 5.8921,
      "mean_token_accuracy": 0.15498456954956055,
      "num_tokens": 11162563.0,
      "step": 5505
    },
    {
      "entropy": 6.1244590282440186,
      "epoch": 0.48151708468059073,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982474563116928,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16450159698724748,
      "num_tokens": 11172572.0,
      "step": 5510
    },
    {
      "entropy": 6.058566474914551,
      "epoch": 0.4819540330332955,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982435724563143,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1621236249804497,
      "num_tokens": 11183404.0,
      "step": 5515
    },
    {
      "entropy": 6.0353001117706295,
      "epoch": 0.4823909813860002,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982396843190072,
      "loss": 5.8934,
      "mean_token_accuracy": 0.15665270984172822,
      "num_tokens": 11193905.0,
      "step": 5520
    },
    {
      "entropy": 6.112768602371216,
      "epoch": 0.48282792973870486,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982357918998459,
      "loss": 5.7899,
      "mean_token_accuracy": 0.17055769711732865,
      "num_tokens": 11203127.0,
      "step": 5525
    },
    {
      "entropy": 6.065280389785767,
      "epoch": 0.4832648780914096,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004982318951989053,
      "loss": 5.8034,
      "mean_token_accuracy": 0.16873722821474074,
      "num_tokens": 11213468.0,
      "step": 5530
    },
    {
      "entropy": 6.103843307495117,
      "epoch": 0.4837018264441143,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982279942162597,
      "loss": 5.8905,
      "mean_token_accuracy": 0.15670445263385774,
      "num_tokens": 11224722.0,
      "step": 5535
    },
    {
      "entropy": 6.1261786937713625,
      "epoch": 0.484138774796819,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982240889519844,
      "loss": 5.8925,
      "mean_token_accuracy": 0.15962006151676178,
      "num_tokens": 11234565.0,
      "step": 5540
    },
    {
      "entropy": 6.0942792892456055,
      "epoch": 0.48457572314952374,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498220179406154,
      "loss": 5.8817,
      "mean_token_accuracy": 0.15751929730176925,
      "num_tokens": 11244496.0,
      "step": 5545
    },
    {
      "entropy": 6.1508289813995365,
      "epoch": 0.48501267150222843,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982162655788436,
      "loss": 5.8568,
      "mean_token_accuracy": 0.16027530580759047,
      "num_tokens": 11254187.0,
      "step": 5550
    },
    {
      "entropy": 6.064441919326782,
      "epoch": 0.4854496198549331,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982123474701283,
      "loss": 5.8421,
      "mean_token_accuracy": 0.16714205592870712,
      "num_tokens": 11264580.0,
      "step": 5555
    },
    {
      "entropy": 6.105088996887207,
      "epoch": 0.48588656820763787,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982084250800832,
      "loss": 5.9733,
      "mean_token_accuracy": 0.15524329617619514,
      "num_tokens": 11275104.0,
      "step": 5560
    },
    {
      "entropy": 6.159511947631836,
      "epoch": 0.48632351656034256,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982044984087836,
      "loss": 5.9885,
      "mean_token_accuracy": 0.15093078166246415,
      "num_tokens": 11285983.0,
      "step": 5565
    },
    {
      "entropy": 6.065144824981689,
      "epoch": 0.48676046491304725,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982005674563048,
      "loss": 5.8752,
      "mean_token_accuracy": 0.16066836416721345,
      "num_tokens": 11295378.0,
      "step": 5570
    },
    {
      "entropy": 6.097844791412354,
      "epoch": 0.487197413265752,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498196632222722,
      "loss": 5.9698,
      "mean_token_accuracy": 0.15424834787845612,
      "num_tokens": 11305500.0,
      "step": 5575
    },
    {
      "entropy": 6.1019445896148685,
      "epoch": 0.4876343616184567,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981926927081109,
      "loss": 5.8791,
      "mean_token_accuracy": 0.15821175277233124,
      "num_tokens": 11316455.0,
      "step": 5580
    },
    {
      "entropy": 6.118305397033692,
      "epoch": 0.4880713099711614,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981887489125471,
      "loss": 5.8455,
      "mean_token_accuracy": 0.1536899395287037,
      "num_tokens": 11325865.0,
      "step": 5585
    },
    {
      "entropy": 6.004519128799439,
      "epoch": 0.48850825832386613,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498184800836106,
      "loss": 5.808,
      "mean_token_accuracy": 0.1714118793606758,
      "num_tokens": 11337368.0,
      "step": 5590
    },
    {
      "entropy": 6.028197860717773,
      "epoch": 0.4889452066765708,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004981808484788635,
      "loss": 5.788,
      "mean_token_accuracy": 0.16869792193174363,
      "num_tokens": 11347129.0,
      "step": 5595
    },
    {
      "entropy": 6.1151933670043945,
      "epoch": 0.48938215502927557,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981768918408955,
      "loss": 5.8516,
      "mean_token_accuracy": 0.16096147298812866,
      "num_tokens": 11356641.0,
      "step": 5600
    },
    {
      "entropy": 6.056279420852661,
      "epoch": 0.48981910338198026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981729309222777,
      "loss": 5.7372,
      "mean_token_accuracy": 0.164483879506588,
      "num_tokens": 11366485.0,
      "step": 5605
    },
    {
      "entropy": 6.024623394012451,
      "epoch": 0.49025605173468495,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981689657230861,
      "loss": 5.8855,
      "mean_token_accuracy": 0.1605944663286209,
      "num_tokens": 11376305.0,
      "step": 5610
    },
    {
      "entropy": 6.0709381103515625,
      "epoch": 0.4906930000873897,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498164996243397,
      "loss": 5.8675,
      "mean_token_accuracy": 0.15576735138893127,
      "num_tokens": 11385837.0,
      "step": 5615
    },
    {
      "entropy": 6.1082054615020756,
      "epoch": 0.4911299484400944,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981610224832861,
      "loss": 5.8884,
      "mean_token_accuracy": 0.1625692754983902,
      "num_tokens": 11396719.0,
      "step": 5620
    },
    {
      "entropy": 6.139967060089111,
      "epoch": 0.4915668967927991,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981570444428299,
      "loss": 5.9111,
      "mean_token_accuracy": 0.15231653898954392,
      "num_tokens": 11408107.0,
      "step": 5625
    },
    {
      "entropy": 6.077764654159546,
      "epoch": 0.4920038451455038,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981530621221048,
      "loss": 5.8293,
      "mean_token_accuracy": 0.1667654886841774,
      "num_tokens": 11417547.0,
      "step": 5630
    },
    {
      "entropy": 6.1007482528686525,
      "epoch": 0.4924407934982085,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981490755211869,
      "loss": 5.9007,
      "mean_token_accuracy": 0.15937500894069673,
      "num_tokens": 11427660.0,
      "step": 5635
    },
    {
      "entropy": 6.045559692382812,
      "epoch": 0.4928777418509132,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981450846401528,
      "loss": 5.7844,
      "mean_token_accuracy": 0.1629929095506668,
      "num_tokens": 11438032.0,
      "step": 5640
    },
    {
      "entropy": 6.092393302917481,
      "epoch": 0.49331469020361796,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498141089479079,
      "loss": 5.8718,
      "mean_token_accuracy": 0.16069578975439072,
      "num_tokens": 11448987.0,
      "step": 5645
    },
    {
      "entropy": 6.057941579818726,
      "epoch": 0.49375163855632265,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981370900380423,
      "loss": 5.8645,
      "mean_token_accuracy": 0.16491677165031432,
      "num_tokens": 11458659.0,
      "step": 5650
    },
    {
      "entropy": 6.123438596725464,
      "epoch": 0.49418858690902734,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981330863171194,
      "loss": 5.8384,
      "mean_token_accuracy": 0.16192056834697724,
      "num_tokens": 11468104.0,
      "step": 5655
    },
    {
      "entropy": 6.091104412078858,
      "epoch": 0.4946255352617321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981290783163867,
      "loss": 5.7977,
      "mean_token_accuracy": 0.16603543758392333,
      "num_tokens": 11479579.0,
      "step": 5660
    },
    {
      "entropy": 6.075469160079956,
      "epoch": 0.4950624836144368,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981250660359216,
      "loss": 5.8784,
      "mean_token_accuracy": 0.1644681215286255,
      "num_tokens": 11489202.0,
      "step": 5665
    },
    {
      "entropy": 6.055730485916138,
      "epoch": 0.49549943196714147,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981210494758007,
      "loss": 5.8927,
      "mean_token_accuracy": 0.15344637930393218,
      "num_tokens": 11499425.0,
      "step": 5670
    },
    {
      "entropy": 6.153387022018433,
      "epoch": 0.4959363803198462,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498117028636101,
      "loss": 5.8743,
      "mean_token_accuracy": 0.16150706708431245,
      "num_tokens": 11508983.0,
      "step": 5675
    },
    {
      "entropy": 6.072080421447754,
      "epoch": 0.4963733286725509,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981130035169,
      "loss": 5.9369,
      "mean_token_accuracy": 0.16237848848104477,
      "num_tokens": 11519791.0,
      "step": 5680
    },
    {
      "entropy": 6.09052529335022,
      "epoch": 0.4968102770252556,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981089741182747,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16089533567428588,
      "num_tokens": 11530537.0,
      "step": 5685
    },
    {
      "entropy": 6.069422388076783,
      "epoch": 0.49724722537796034,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981049404403023,
      "loss": 5.8671,
      "mean_token_accuracy": 0.16157519519329072,
      "num_tokens": 11540591.0,
      "step": 5690
    },
    {
      "entropy": 6.042875909805298,
      "epoch": 0.49768417373066504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981009024830602,
      "loss": 5.8079,
      "mean_token_accuracy": 0.16299513429403306,
      "num_tokens": 11550719.0,
      "step": 5695
    },
    {
      "entropy": 6.042115831375122,
      "epoch": 0.4981211220833697,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498096860246626,
      "loss": 5.8665,
      "mean_token_accuracy": 0.16123356223106383,
      "num_tokens": 11561805.0,
      "step": 5700
    },
    {
      "entropy": 6.19722990989685,
      "epoch": 0.4985580704360745,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498092813731077,
      "loss": 5.8986,
      "mean_token_accuracy": 0.15539520978927612,
      "num_tokens": 11571380.0,
      "step": 5705
    },
    {
      "entropy": 6.036397552490234,
      "epoch": 0.49899501878877917,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498088762936491,
      "loss": 5.8801,
      "mean_token_accuracy": 0.1572152391076088,
      "num_tokens": 11580853.0,
      "step": 5710
    },
    {
      "entropy": 5.938679552078247,
      "epoch": 0.49943196714148386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980847078629457,
      "loss": 5.7742,
      "mean_token_accuracy": 0.17463206201791764,
      "num_tokens": 11591966.0,
      "step": 5715
    },
    {
      "entropy": 6.064130449295044,
      "epoch": 0.4998689154941886,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980806485105186,
      "loss": 5.845,
      "mean_token_accuracy": 0.16573392152786254,
      "num_tokens": 11601239.0,
      "step": 5720
    },
    {
      "entropy": 6.077233600616455,
      "epoch": 0.5003058638468934,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980765848792879,
      "loss": 5.7666,
      "mean_token_accuracy": 0.17577522695064546,
      "num_tokens": 11609857.0,
      "step": 5725
    },
    {
      "entropy": 6.097680330276489,
      "epoch": 0.500742812199598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980725169693314,
      "loss": 5.8202,
      "mean_token_accuracy": 0.16237599551677703,
      "num_tokens": 11618952.0,
      "step": 5730
    },
    {
      "entropy": 6.071645975112915,
      "epoch": 0.5011797605523027,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980684447807272,
      "loss": 5.7513,
      "mean_token_accuracy": 0.16841784864664078,
      "num_tokens": 11627638.0,
      "step": 5735
    },
    {
      "entropy": 6.089624977111816,
      "epoch": 0.5016167089050074,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980643683135533,
      "loss": 5.8707,
      "mean_token_accuracy": 0.1621746838092804,
      "num_tokens": 11637842.0,
      "step": 5740
    },
    {
      "entropy": 5.975975370407104,
      "epoch": 0.5020536572577121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980602875678878,
      "loss": 5.7948,
      "mean_token_accuracy": 0.17163587510585784,
      "num_tokens": 11647217.0,
      "step": 5745
    },
    {
      "entropy": 5.982842636108399,
      "epoch": 0.5024906056104168,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980562025438093,
      "loss": 5.7833,
      "mean_token_accuracy": 0.1593058079481125,
      "num_tokens": 11657558.0,
      "step": 5750
    },
    {
      "entropy": 6.042101001739502,
      "epoch": 0.5029275539631216,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980521132413958,
      "loss": 5.7884,
      "mean_token_accuracy": 0.1595451831817627,
      "num_tokens": 11668054.0,
      "step": 5755
    },
    {
      "entropy": 6.13589997291565,
      "epoch": 0.5033645023158263,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498048019660726,
      "loss": 5.7996,
      "mean_token_accuracy": 0.1616273492574692,
      "num_tokens": 11677779.0,
      "step": 5760
    },
    {
      "entropy": 6.057114553451538,
      "epoch": 0.503801450668531,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980439218018783,
      "loss": 5.8451,
      "mean_token_accuracy": 0.15998862981796264,
      "num_tokens": 11687577.0,
      "step": 5765
    },
    {
      "entropy": 6.139103269577026,
      "epoch": 0.5042383990212357,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004980398196649312,
      "loss": 5.9438,
      "mean_token_accuracy": 0.15554213970899583,
      "num_tokens": 11699630.0,
      "step": 5770
    },
    {
      "entropy": 6.080030202865601,
      "epoch": 0.5046753473739404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980357132499636,
      "loss": 5.8754,
      "mean_token_accuracy": 0.16034791767597198,
      "num_tokens": 11710139.0,
      "step": 5775
    },
    {
      "entropy": 6.091990947723389,
      "epoch": 0.5051122957266451,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980316025570542,
      "loss": 5.8403,
      "mean_token_accuracy": 0.16256014853715897,
      "num_tokens": 11719070.0,
      "step": 5780
    },
    {
      "entropy": 6.076358985900879,
      "epoch": 0.5055492440793499,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980274875862816,
      "loss": 5.8728,
      "mean_token_accuracy": 0.15951588302850722,
      "num_tokens": 11728770.0,
      "step": 5785
    },
    {
      "entropy": 6.065584897994995,
      "epoch": 0.5059861924320546,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498023368337725,
      "loss": 5.8178,
      "mean_token_accuracy": 0.16913911551237107,
      "num_tokens": 11737753.0,
      "step": 5790
    },
    {
      "entropy": 6.011010837554932,
      "epoch": 0.5064231407847593,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980192448114635,
      "loss": 5.8007,
      "mean_token_accuracy": 0.1579100161790848,
      "num_tokens": 11747837.0,
      "step": 5795
    },
    {
      "entropy": 6.055712556838989,
      "epoch": 0.5068600891374639,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980151170075758,
      "loss": 5.9213,
      "mean_token_accuracy": 0.1625312827527523,
      "num_tokens": 11757911.0,
      "step": 5800
    },
    {
      "entropy": 6.130970239639282,
      "epoch": 0.5072970374901686,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980109849261414,
      "loss": 5.7862,
      "mean_token_accuracy": 0.16185480654239653,
      "num_tokens": 11766213.0,
      "step": 5805
    },
    {
      "entropy": 6.055031728744507,
      "epoch": 0.5077339858428733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980068485672394,
      "loss": 5.7594,
      "mean_token_accuracy": 0.16835663914680482,
      "num_tokens": 11775640.0,
      "step": 5810
    },
    {
      "entropy": 6.097806119918824,
      "epoch": 0.5081709341955781,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980027079309493,
      "loss": 5.782,
      "mean_token_accuracy": 0.16621346697211264,
      "num_tokens": 11785664.0,
      "step": 5815
    },
    {
      "entropy": 6.122802639007569,
      "epoch": 0.5086078825482828,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979985630173503,
      "loss": 5.9093,
      "mean_token_accuracy": 0.15274638459086418,
      "num_tokens": 11795402.0,
      "step": 5820
    },
    {
      "entropy": 6.022065305709839,
      "epoch": 0.5090448309009875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979944138265221,
      "loss": 5.7766,
      "mean_token_accuracy": 0.16640096306800842,
      "num_tokens": 11805923.0,
      "step": 5825
    },
    {
      "entropy": 6.070121622085571,
      "epoch": 0.5094817792536922,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979902603585442,
      "loss": 5.8875,
      "mean_token_accuracy": 0.15387427806854248,
      "num_tokens": 11816097.0,
      "step": 5830
    },
    {
      "entropy": 6.080983304977417,
      "epoch": 0.5099187276063969,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979861026134961,
      "loss": 5.788,
      "mean_token_accuracy": 0.1624365270137787,
      "num_tokens": 11826889.0,
      "step": 5835
    },
    {
      "entropy": 6.024716758728028,
      "epoch": 0.5103556759591016,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979819405914579,
      "loss": 5.8488,
      "mean_token_accuracy": 0.162523353099823,
      "num_tokens": 11837935.0,
      "step": 5840
    },
    {
      "entropy": 6.051336288452148,
      "epoch": 0.5107926243118064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979777742925092,
      "loss": 5.846,
      "mean_token_accuracy": 0.1586737349629402,
      "num_tokens": 11847715.0,
      "step": 5845
    },
    {
      "entropy": 6.144127655029297,
      "epoch": 0.5112295726645111,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979736037167298,
      "loss": 5.8579,
      "mean_token_accuracy": 0.158831188082695,
      "num_tokens": 11857841.0,
      "step": 5850
    },
    {
      "entropy": 6.080711126327515,
      "epoch": 0.5116665210172158,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979694288642,
      "loss": 5.8583,
      "mean_token_accuracy": 0.16246726363897324,
      "num_tokens": 11868135.0,
      "step": 5855
    },
    {
      "entropy": 6.075771951675415,
      "epoch": 0.5121034693699205,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979652497349997,
      "loss": 5.785,
      "mean_token_accuracy": 0.15890408009290696,
      "num_tokens": 11877637.0,
      "step": 5860
    },
    {
      "entropy": 5.998299980163575,
      "epoch": 0.5125404177226252,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497961066329209,
      "loss": 5.8513,
      "mean_token_accuracy": 0.15636989027261733,
      "num_tokens": 11887128.0,
      "step": 5865
    },
    {
      "entropy": 6.135884857177734,
      "epoch": 0.5129773660753298,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979568786469083,
      "loss": 5.869,
      "mean_token_accuracy": 0.1521621897816658,
      "num_tokens": 11897885.0,
      "step": 5870
    },
    {
      "entropy": 6.035808944702149,
      "epoch": 0.5134143144280346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979526866881777,
      "loss": 5.801,
      "mean_token_accuracy": 0.16173582524061203,
      "num_tokens": 11907919.0,
      "step": 5875
    },
    {
      "entropy": 5.923618268966675,
      "epoch": 0.5138512627807393,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497948490453098,
      "loss": 5.7782,
      "mean_token_accuracy": 0.16808433681726456,
      "num_tokens": 11918612.0,
      "step": 5880
    },
    {
      "entropy": 6.026860332489013,
      "epoch": 0.514288211133444,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979442899417493,
      "loss": 5.8489,
      "mean_token_accuracy": 0.16112847179174422,
      "num_tokens": 11927759.0,
      "step": 5885
    },
    {
      "entropy": 6.097043371200561,
      "epoch": 0.5147251594861487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979400851542122,
      "loss": 5.8103,
      "mean_token_accuracy": 0.1685948997735977,
      "num_tokens": 11937522.0,
      "step": 5890
    },
    {
      "entropy": 6.0076072216033936,
      "epoch": 0.5151621078388534,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979358760905675,
      "loss": 5.7903,
      "mean_token_accuracy": 0.16276071667671205,
      "num_tokens": 11947787.0,
      "step": 5895
    },
    {
      "entropy": 6.01747145652771,
      "epoch": 0.5155990561915581,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497931662750896,
      "loss": 5.8322,
      "mean_token_accuracy": 0.16161809414625167,
      "num_tokens": 11958929.0,
      "step": 5900
    },
    {
      "entropy": 6.070228338241577,
      "epoch": 0.5160360045442629,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979274451352783,
      "loss": 5.7808,
      "mean_token_accuracy": 0.16201020032167435,
      "num_tokens": 11969242.0,
      "step": 5905
    },
    {
      "entropy": 6.057408142089844,
      "epoch": 0.5164729528969676,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979232232437954,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1614376947283745,
      "num_tokens": 11980149.0,
      "step": 5910
    },
    {
      "entropy": 6.096258878707886,
      "epoch": 0.5169099012496723,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979189970765282,
      "loss": 5.8364,
      "mean_token_accuracy": 0.1574439138174057,
      "num_tokens": 11990636.0,
      "step": 5915
    },
    {
      "entropy": 6.069129228591919,
      "epoch": 0.517346849602377,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979147666335579,
      "loss": 5.8491,
      "mean_token_accuracy": 0.16322307884693146,
      "num_tokens": 12000615.0,
      "step": 5920
    },
    {
      "entropy": 6.038877487182617,
      "epoch": 0.5177837979550817,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979105319149656,
      "loss": 5.8802,
      "mean_token_accuracy": 0.16359985023736953,
      "num_tokens": 12011072.0,
      "step": 5925
    },
    {
      "entropy": 6.024696397781372,
      "epoch": 0.5182207463077865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979062929208324,
      "loss": 5.7764,
      "mean_token_accuracy": 0.16569287329912186,
      "num_tokens": 12021726.0,
      "step": 5930
    },
    {
      "entropy": 6.1139367580413815,
      "epoch": 0.5186576946604912,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979020496512398,
      "loss": 5.9182,
      "mean_token_accuracy": 0.15360669791698456,
      "num_tokens": 12031353.0,
      "step": 5935
    },
    {
      "entropy": 6.086315155029297,
      "epoch": 0.5190946430131959,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497897802106269,
      "loss": 5.8545,
      "mean_token_accuracy": 0.15585757344961165,
      "num_tokens": 12041234.0,
      "step": 5940
    },
    {
      "entropy": 5.954757881164551,
      "epoch": 0.5195315913659005,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978935502860017,
      "loss": 5.7755,
      "mean_token_accuracy": 0.16401639878749846,
      "num_tokens": 12051100.0,
      "step": 5945
    },
    {
      "entropy": 6.086745738983154,
      "epoch": 0.5199685397186052,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978892941905191,
      "loss": 5.8426,
      "mean_token_accuracy": 0.15925330817699432,
      "num_tokens": 12062012.0,
      "step": 5950
    },
    {
      "entropy": 6.13814787864685,
      "epoch": 0.5204054880713099,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978850338199033,
      "loss": 5.8788,
      "mean_token_accuracy": 0.15904145166277886,
      "num_tokens": 12072763.0,
      "step": 5955
    },
    {
      "entropy": 5.998065614700318,
      "epoch": 0.5208424364240147,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978807691742356,
      "loss": 5.7153,
      "mean_token_accuracy": 0.17032663524150848,
      "num_tokens": 12081892.0,
      "step": 5960
    },
    {
      "entropy": 6.015545177459717,
      "epoch": 0.5212793847767194,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497876500253598,
      "loss": 5.7963,
      "mean_token_accuracy": 0.16373672038316728,
      "num_tokens": 12091428.0,
      "step": 5965
    },
    {
      "entropy": 6.048707628250122,
      "epoch": 0.5217163331294241,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978722270580722,
      "loss": 5.8102,
      "mean_token_accuracy": 0.16371404230594636,
      "num_tokens": 12101687.0,
      "step": 5970
    },
    {
      "entropy": 5.966954755783081,
      "epoch": 0.5221532814821288,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978679495877405,
      "loss": 5.7606,
      "mean_token_accuracy": 0.160753433406353,
      "num_tokens": 12111726.0,
      "step": 5975
    },
    {
      "entropy": 6.0778566837310795,
      "epoch": 0.5225902298348335,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978636678426847,
      "loss": 5.7928,
      "mean_token_accuracy": 0.16682983934879303,
      "num_tokens": 12122171.0,
      "step": 5980
    },
    {
      "entropy": 6.119118928909302,
      "epoch": 0.5230271781875382,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978593818229869,
      "loss": 5.8393,
      "mean_token_accuracy": 0.15573921650648118,
      "num_tokens": 12131414.0,
      "step": 5985
    },
    {
      "entropy": 5.977964401245117,
      "epoch": 0.523464126540243,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978550915287294,
      "loss": 5.7908,
      "mean_token_accuracy": 0.15605172738432885,
      "num_tokens": 12141665.0,
      "step": 5990
    },
    {
      "entropy": 6.082938003540039,
      "epoch": 0.5239010748929477,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978507969599945,
      "loss": 5.8799,
      "mean_token_accuracy": 0.15631582140922545,
      "num_tokens": 12151263.0,
      "step": 5995
    },
    {
      "entropy": 6.132467651367188,
      "epoch": 0.5243380232456524,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004978464981168645,
      "loss": 5.8516,
      "mean_token_accuracy": 0.16170431599020957,
      "num_tokens": 12161070.0,
      "step": 6000
    },
    {
      "epoch": 0.5243380232456524,
      "eval_entropy": 5.84592246638278,
      "eval_loss": 5.83839225769043,
      "eval_mean_token_accuracy": 0.16758470430564412,
      "eval_num_tokens": 12161070.0,
      "eval_runtime": 21.3881,
      "eval_samples_per_second": 1730.587,
      "eval_steps_per_second": 216.335,
      "step": 6000
    },
    {
      "entropy": 6.051551675796508,
      "epoch": 0.5247749715983571,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497842194999422,
      "loss": 5.8121,
      "mean_token_accuracy": 0.1631454199552536,
      "num_tokens": 12171197.0,
      "step": 6005
    },
    {
      "entropy": 6.02082085609436,
      "epoch": 0.5252119199510618,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978378876077493,
      "loss": 5.8371,
      "mean_token_accuracy": 0.16132295727729798,
      "num_tokens": 12181462.0,
      "step": 6010
    },
    {
      "entropy": 6.069212293624878,
      "epoch": 0.5256488683037664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978335759419293,
      "loss": 5.8169,
      "mean_token_accuracy": 0.16407132148742676,
      "num_tokens": 12191473.0,
      "step": 6015
    },
    {
      "entropy": 6.1080104351043705,
      "epoch": 0.5260858166564713,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004978292600020444,
      "loss": 5.7749,
      "mean_token_accuracy": 0.16672620177268982,
      "num_tokens": 12200920.0,
      "step": 6020
    },
    {
      "entropy": 6.044141244888306,
      "epoch": 0.5265227650091759,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978249397881776,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16232698559761047,
      "num_tokens": 12210823.0,
      "step": 6025
    },
    {
      "entropy": 6.063474082946778,
      "epoch": 0.5269597133618806,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978206153004117,
      "loss": 5.8578,
      "mean_token_accuracy": 0.16104032844305038,
      "num_tokens": 12221967.0,
      "step": 6030
    },
    {
      "entropy": 6.037987470626831,
      "epoch": 0.5273966617145853,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978162865388297,
      "loss": 5.8138,
      "mean_token_accuracy": 0.1628843292593956,
      "num_tokens": 12232519.0,
      "step": 6035
    },
    {
      "entropy": 6.070814561843872,
      "epoch": 0.52783361006729,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978119535035144,
      "loss": 5.7972,
      "mean_token_accuracy": 0.1579383745789528,
      "num_tokens": 12243193.0,
      "step": 6040
    },
    {
      "entropy": 6.065000867843628,
      "epoch": 0.5282705584199947,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978076161945492,
      "loss": 5.7989,
      "mean_token_accuracy": 0.1628345713019371,
      "num_tokens": 12253640.0,
      "step": 6045
    },
    {
      "entropy": 6.15475172996521,
      "epoch": 0.5287075067726995,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978032746120173,
      "loss": 5.908,
      "mean_token_accuracy": 0.15190446525812148,
      "num_tokens": 12265072.0,
      "step": 6050
    },
    {
      "entropy": 6.068175888061523,
      "epoch": 0.5291444551254042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977989287560016,
      "loss": 5.8435,
      "mean_token_accuracy": 0.15906517803668976,
      "num_tokens": 12274449.0,
      "step": 6055
    },
    {
      "entropy": 6.087897634506225,
      "epoch": 0.5295814034781089,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977945786265857,
      "loss": 5.849,
      "mean_token_accuracy": 0.1522239089012146,
      "num_tokens": 12284910.0,
      "step": 6060
    },
    {
      "entropy": 6.104412317276001,
      "epoch": 0.5300183518308136,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977902242238533,
      "loss": 5.8579,
      "mean_token_accuracy": 0.15928884446620942,
      "num_tokens": 12294046.0,
      "step": 6065
    },
    {
      "entropy": 6.042326354980469,
      "epoch": 0.5304553001835183,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977858655478874,
      "loss": 5.856,
      "mean_token_accuracy": 0.15834434181451798,
      "num_tokens": 12303900.0,
      "step": 6070
    },
    {
      "entropy": 6.091877841949463,
      "epoch": 0.530892248536223,
      "grad_norm": 1.0,
      "learning_rate": 0.000497781502598772,
      "loss": 5.7919,
      "mean_token_accuracy": 0.1617860183119774,
      "num_tokens": 12313326.0,
      "step": 6075
    },
    {
      "entropy": 6.070764589309692,
      "epoch": 0.5313291968889278,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977771353765906,
      "loss": 5.785,
      "mean_token_accuracy": 0.16128855794668198,
      "num_tokens": 12322792.0,
      "step": 6080
    },
    {
      "entropy": 6.08227276802063,
      "epoch": 0.5317661452416325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497772763881427,
      "loss": 5.8952,
      "mean_token_accuracy": 0.1567649260163307,
      "num_tokens": 12333417.0,
      "step": 6085
    },
    {
      "entropy": 6.0735241889953615,
      "epoch": 0.5322030935943372,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977683881133651,
      "loss": 5.7917,
      "mean_token_accuracy": 0.16096222549676895,
      "num_tokens": 12343944.0,
      "step": 6090
    },
    {
      "entropy": 6.0614039421081545,
      "epoch": 0.5326400419470418,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004977640080724888,
      "loss": 5.8044,
      "mean_token_accuracy": 0.15370135754346848,
      "num_tokens": 12353838.0,
      "step": 6095
    },
    {
      "entropy": 6.106985330581665,
      "epoch": 0.5330769902997465,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977596237588821,
      "loss": 5.8786,
      "mean_token_accuracy": 0.15366269052028655,
      "num_tokens": 12364996.0,
      "step": 6100
    },
    {
      "entropy": 6.073399066925049,
      "epoch": 0.5335139386524512,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497755235172629,
      "loss": 5.8541,
      "mean_token_accuracy": 0.15862328559160233,
      "num_tokens": 12375111.0,
      "step": 6105
    },
    {
      "entropy": 5.990836477279663,
      "epoch": 0.533950887005156,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977508423138139,
      "loss": 5.8304,
      "mean_token_accuracy": 0.15765419453382493,
      "num_tokens": 12384376.0,
      "step": 6110
    },
    {
      "entropy": 6.101342487335205,
      "epoch": 0.5343878353578607,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977464451825208,
      "loss": 5.879,
      "mean_token_accuracy": 0.1540617272257805,
      "num_tokens": 12395248.0,
      "step": 6115
    },
    {
      "entropy": 6.112041854858399,
      "epoch": 0.5348247837105654,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977420437788341,
      "loss": 5.8669,
      "mean_token_accuracy": 0.15835232734680177,
      "num_tokens": 12405083.0,
      "step": 6120
    },
    {
      "entropy": 5.976082134246826,
      "epoch": 0.5352617320632701,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977376381028385,
      "loss": 5.7552,
      "mean_token_accuracy": 0.17444901913404465,
      "num_tokens": 12415787.0,
      "step": 6125
    },
    {
      "entropy": 6.101401948928833,
      "epoch": 0.5356986804159748,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977332281546181,
      "loss": 5.81,
      "mean_token_accuracy": 0.16010264605283736,
      "num_tokens": 12425911.0,
      "step": 6130
    },
    {
      "entropy": 6.056052255630493,
      "epoch": 0.5361356287686796,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977288139342578,
      "loss": 5.8326,
      "mean_token_accuracy": 0.1626307561993599,
      "num_tokens": 12437270.0,
      "step": 6135
    },
    {
      "entropy": 6.01606011390686,
      "epoch": 0.5365725771213843,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977243954418422,
      "loss": 5.7971,
      "mean_token_accuracy": 0.16111206114292145,
      "num_tokens": 12446778.0,
      "step": 6140
    },
    {
      "entropy": 5.963247442245484,
      "epoch": 0.537009525474089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977199726774558,
      "loss": 5.7736,
      "mean_token_accuracy": 0.1644420325756073,
      "num_tokens": 12456550.0,
      "step": 6145
    },
    {
      "entropy": 6.079967308044433,
      "epoch": 0.5374464738267937,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977155456411837,
      "loss": 5.7671,
      "mean_token_accuracy": 0.16329844295978546,
      "num_tokens": 12465562.0,
      "step": 6150
    },
    {
      "entropy": 6.0883838653564455,
      "epoch": 0.5378834221794984,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977111143331108,
      "loss": 5.8283,
      "mean_token_accuracy": 0.15746358335018157,
      "num_tokens": 12475277.0,
      "step": 6155
    },
    {
      "entropy": 6.001813697814941,
      "epoch": 0.538320370532203,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977066787533219,
      "loss": 5.7661,
      "mean_token_accuracy": 0.16389472633600236,
      "num_tokens": 12485478.0,
      "step": 6160
    },
    {
      "entropy": 6.1166510581970215,
      "epoch": 0.5387573188849079,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977022389019023,
      "loss": 5.8654,
      "mean_token_accuracy": 0.15942558273673058,
      "num_tokens": 12495714.0,
      "step": 6165
    },
    {
      "entropy": 6.10480432510376,
      "epoch": 0.5391942672376125,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976977947789371,
      "loss": 5.808,
      "mean_token_accuracy": 0.16558642089366912,
      "num_tokens": 12505125.0,
      "step": 6170
    },
    {
      "entropy": 6.000069046020508,
      "epoch": 0.5396312155903172,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976933463845115,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1634201005101204,
      "num_tokens": 12515051.0,
      "step": 6175
    },
    {
      "entropy": 6.018291854858399,
      "epoch": 0.5400681639430219,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976888937187107,
      "loss": 5.7536,
      "mean_token_accuracy": 0.1705291748046875,
      "num_tokens": 12525389.0,
      "step": 6180
    },
    {
      "entropy": 6.077795028686523,
      "epoch": 0.5405051122957266,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976844367816203,
      "loss": 5.8642,
      "mean_token_accuracy": 0.1491462528705597,
      "num_tokens": 12535454.0,
      "step": 6185
    },
    {
      "entropy": 6.03967661857605,
      "epoch": 0.5409420606484313,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976799755733257,
      "loss": 5.6827,
      "mean_token_accuracy": 0.16808735430240632,
      "num_tokens": 12546092.0,
      "step": 6190
    },
    {
      "entropy": 6.026363039016724,
      "epoch": 0.5413790090011361,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976755100939126,
      "loss": 5.8168,
      "mean_token_accuracy": 0.16169485151767732,
      "num_tokens": 12555462.0,
      "step": 6195
    },
    {
      "entropy": 6.062251234054566,
      "epoch": 0.5418159573538408,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976710403434663,
      "loss": 5.8289,
      "mean_token_accuracy": 0.1632273569703102,
      "num_tokens": 12564506.0,
      "step": 6200
    },
    {
      "entropy": 6.076796960830689,
      "epoch": 0.5422529057065455,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497666566322073,
      "loss": 5.8094,
      "mean_token_accuracy": 0.15859714150428772,
      "num_tokens": 12574887.0,
      "step": 6205
    },
    {
      "entropy": 6.138982725143433,
      "epoch": 0.5426898540592502,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976620880298182,
      "loss": 5.826,
      "mean_token_accuracy": 0.16167740821838378,
      "num_tokens": 12584499.0,
      "step": 6210
    },
    {
      "entropy": 6.064753150939941,
      "epoch": 0.5431268024119549,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976576054667878,
      "loss": 5.7699,
      "mean_token_accuracy": 0.15929168462753296,
      "num_tokens": 12593796.0,
      "step": 6215
    },
    {
      "entropy": 5.972159910202026,
      "epoch": 0.5435637507646596,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497653118633068,
      "loss": 5.8233,
      "mean_token_accuracy": 0.1529412552714348,
      "num_tokens": 12603719.0,
      "step": 6220
    },
    {
      "entropy": 6.017406606674195,
      "epoch": 0.5440006991173644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976486275287446,
      "loss": 5.8118,
      "mean_token_accuracy": 0.16069236546754836,
      "num_tokens": 12614407.0,
      "step": 6225
    },
    {
      "entropy": 6.091812086105347,
      "epoch": 0.5444376474700691,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976441321539038,
      "loss": 5.8119,
      "mean_token_accuracy": 0.1701043128967285,
      "num_tokens": 12624414.0,
      "step": 6230
    },
    {
      "entropy": 6.039082098007202,
      "epoch": 0.5448745958227738,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976396325086321,
      "loss": 5.8201,
      "mean_token_accuracy": 0.1647835999727249,
      "num_tokens": 12635679.0,
      "step": 6235
    },
    {
      "entropy": 6.060734033584595,
      "epoch": 0.5453115441754784,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976351285930154,
      "loss": 5.7674,
      "mean_token_accuracy": 0.15868058204650878,
      "num_tokens": 12645253.0,
      "step": 6240
    },
    {
      "entropy": 5.953960847854614,
      "epoch": 0.5457484925281831,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976306204071404,
      "loss": 5.7224,
      "mean_token_accuracy": 0.17618757784366607,
      "num_tokens": 12655815.0,
      "step": 6245
    },
    {
      "entropy": 5.99358720779419,
      "epoch": 0.5461854408808878,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976261079510934,
      "loss": 5.8181,
      "mean_token_accuracy": 0.16412648409605027,
      "num_tokens": 12666173.0,
      "step": 6250
    },
    {
      "entropy": 6.04542818069458,
      "epoch": 0.5466223892335926,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497621591224961,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1651145949959755,
      "num_tokens": 12676008.0,
      "step": 6255
    },
    {
      "entropy": 6.084331512451172,
      "epoch": 0.5470593375862973,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976170702288298,
      "loss": 5.8213,
      "mean_token_accuracy": 0.1627424970269203,
      "num_tokens": 12686217.0,
      "step": 6260
    },
    {
      "entropy": 6.011774969100952,
      "epoch": 0.547496285939002,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004976125449627866,
      "loss": 5.83,
      "mean_token_accuracy": 0.15726649314165114,
      "num_tokens": 12696809.0,
      "step": 6265
    },
    {
      "entropy": 6.102260637283325,
      "epoch": 0.5479332342917067,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004976080154269181,
      "loss": 5.8955,
      "mean_token_accuracy": 0.15118584036827087,
      "num_tokens": 12708328.0,
      "step": 6270
    },
    {
      "entropy": 6.027799320220947,
      "epoch": 0.5483701826444114,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976034816213112,
      "loss": 5.7141,
      "mean_token_accuracy": 0.16318979412317275,
      "num_tokens": 12718478.0,
      "step": 6275
    },
    {
      "entropy": 5.9194399356842045,
      "epoch": 0.5488071309971161,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975989435460528,
      "loss": 5.8032,
      "mean_token_accuracy": 0.162892384827137,
      "num_tokens": 12728853.0,
      "step": 6280
    },
    {
      "entropy": 6.046306610107422,
      "epoch": 0.5492440793498209,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975944012012302,
      "loss": 5.8553,
      "mean_token_accuracy": 0.15959452092647552,
      "num_tokens": 12739036.0,
      "step": 6285
    },
    {
      "entropy": 6.0251960277557375,
      "epoch": 0.5496810277025256,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975898545869303,
      "loss": 5.7562,
      "mean_token_accuracy": 0.16758886128664016,
      "num_tokens": 12748908.0,
      "step": 6290
    },
    {
      "entropy": 6.0087062358856205,
      "epoch": 0.5501179760552303,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975853037032403,
      "loss": 5.8346,
      "mean_token_accuracy": 0.15466514453291894,
      "num_tokens": 12758671.0,
      "step": 6295
    },
    {
      "entropy": 6.012726211547852,
      "epoch": 0.550554924407935,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975807485502476,
      "loss": 5.805,
      "mean_token_accuracy": 0.16215717792510986,
      "num_tokens": 12769552.0,
      "step": 6300
    },
    {
      "entropy": 6.105425882339477,
      "epoch": 0.5509918727606397,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975761891280394,
      "loss": 5.9356,
      "mean_token_accuracy": 0.15056938230991362,
      "num_tokens": 12781032.0,
      "step": 6305
    },
    {
      "entropy": 6.167303895950317,
      "epoch": 0.5514288211133443,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975716254367033,
      "loss": 5.8471,
      "mean_token_accuracy": 0.15739578381180763,
      "num_tokens": 12790818.0,
      "step": 6310
    },
    {
      "entropy": 5.9673717498779295,
      "epoch": 0.5518657694660491,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497567057476327,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1678812563419342,
      "num_tokens": 12799701.0,
      "step": 6315
    },
    {
      "entropy": 5.981673002243042,
      "epoch": 0.5523027178187538,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975624852469976,
      "loss": 5.8716,
      "mean_token_accuracy": 0.16354651004076004,
      "num_tokens": 12809575.0,
      "step": 6320
    },
    {
      "entropy": 6.096052551269532,
      "epoch": 0.5527396661714585,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975579087488033,
      "loss": 5.771,
      "mean_token_accuracy": 0.1671651154756546,
      "num_tokens": 12820278.0,
      "step": 6325
    },
    {
      "entropy": 6.113509225845337,
      "epoch": 0.5531766145241632,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975533279818317,
      "loss": 5.8071,
      "mean_token_accuracy": 0.1612452432513237,
      "num_tokens": 12830774.0,
      "step": 6330
    },
    {
      "entropy": 5.9825366020202635,
      "epoch": 0.5536135628768679,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975487429461706,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16761625409126282,
      "num_tokens": 12839487.0,
      "step": 6335
    },
    {
      "entropy": 5.980599498748779,
      "epoch": 0.5540505112295727,
      "grad_norm": 0.84375,
      "learning_rate": 0.000497544153641908,
      "loss": 5.7203,
      "mean_token_accuracy": 0.16633592545986176,
      "num_tokens": 12849614.0,
      "step": 6340
    },
    {
      "entropy": 6.053549337387085,
      "epoch": 0.5544874595822774,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975395600691319,
      "loss": 5.8452,
      "mean_token_accuracy": 0.1624772861599922,
      "num_tokens": 12859428.0,
      "step": 6345
    },
    {
      "entropy": 6.014547872543335,
      "epoch": 0.5549244079349821,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975349622279305,
      "loss": 5.8259,
      "mean_token_accuracy": 0.16293375492095946,
      "num_tokens": 12869167.0,
      "step": 6350
    },
    {
      "entropy": 6.031063747406006,
      "epoch": 0.5553613562876868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975303601183919,
      "loss": 5.7773,
      "mean_token_accuracy": 0.16692850291728972,
      "num_tokens": 12878410.0,
      "step": 6355
    },
    {
      "entropy": 6.034283590316773,
      "epoch": 0.5557983046403915,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975257537406043,
      "loss": 5.7904,
      "mean_token_accuracy": 0.158176751434803,
      "num_tokens": 12888372.0,
      "step": 6360
    },
    {
      "entropy": 6.014430284500122,
      "epoch": 0.5562352529930962,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975211430946563,
      "loss": 5.8304,
      "mean_token_accuracy": 0.16271426677703857,
      "num_tokens": 12897958.0,
      "step": 6365
    },
    {
      "entropy": 6.048214292526245,
      "epoch": 0.556672201345801,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497516528180636,
      "loss": 5.8401,
      "mean_token_accuracy": 0.1621402770280838,
      "num_tokens": 12907890.0,
      "step": 6370
    },
    {
      "entropy": 6.068901777267456,
      "epoch": 0.5571091496985057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975119089986322,
      "loss": 5.769,
      "mean_token_accuracy": 0.16217422038316726,
      "num_tokens": 12917914.0,
      "step": 6375
    },
    {
      "entropy": 6.003548049926758,
      "epoch": 0.5575460980512104,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975072855487333,
      "loss": 5.8164,
      "mean_token_accuracy": 0.16145033836364747,
      "num_tokens": 12927090.0,
      "step": 6380
    },
    {
      "entropy": 6.04446930885315,
      "epoch": 0.557983046403915,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975026578310282,
      "loss": 5.7956,
      "mean_token_accuracy": 0.16031267493963242,
      "num_tokens": 12937236.0,
      "step": 6385
    },
    {
      "entropy": 6.054268741607666,
      "epoch": 0.5584199947566197,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974980258456055,
      "loss": 5.7586,
      "mean_token_accuracy": 0.1658152610063553,
      "num_tokens": 12947635.0,
      "step": 6390
    },
    {
      "entropy": 6.073834800720215,
      "epoch": 0.5588569431093244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497493389592554,
      "loss": 5.6953,
      "mean_token_accuracy": 0.16651600003242492,
      "num_tokens": 12957236.0,
      "step": 6395
    },
    {
      "entropy": 6.019908809661866,
      "epoch": 0.5592938914620292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974887490719627,
      "loss": 5.8179,
      "mean_token_accuracy": 0.16579750925302505,
      "num_tokens": 12967497.0,
      "step": 6400
    },
    {
      "entropy": 5.970498514175415,
      "epoch": 0.5597308398147339,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974841042839207,
      "loss": 5.7514,
      "mean_token_accuracy": 0.16714754849672317,
      "num_tokens": 12977467.0,
      "step": 6405
    },
    {
      "entropy": 6.068846273422241,
      "epoch": 0.5601677881674386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497479455228517,
      "loss": 5.8051,
      "mean_token_accuracy": 0.15970347225666046,
      "num_tokens": 12987867.0,
      "step": 6410
    },
    {
      "entropy": 6.026486110687256,
      "epoch": 0.5606047365201433,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974748019058407,
      "loss": 5.8153,
      "mean_token_accuracy": 0.15720094591379166,
      "num_tokens": 12998081.0,
      "step": 6415
    },
    {
      "entropy": 6.005708646774292,
      "epoch": 0.561041684872848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974701443159812,
      "loss": 5.8118,
      "mean_token_accuracy": 0.1625289648771286,
      "num_tokens": 13008047.0,
      "step": 6420
    },
    {
      "entropy": 6.007196378707886,
      "epoch": 0.5614786332255527,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974654824590277,
      "loss": 5.7747,
      "mean_token_accuracy": 0.16239859908819199,
      "num_tokens": 13018048.0,
      "step": 6425
    },
    {
      "entropy": 6.001573276519776,
      "epoch": 0.5619155815782575,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974608163350697,
      "loss": 5.7887,
      "mean_token_accuracy": 0.15910013020038605,
      "num_tokens": 13028156.0,
      "step": 6430
    },
    {
      "entropy": 6.072333478927613,
      "epoch": 0.5623525299309622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974561459441967,
      "loss": 5.7916,
      "mean_token_accuracy": 0.15919577330350876,
      "num_tokens": 13038191.0,
      "step": 6435
    },
    {
      "entropy": 5.9851783275604244,
      "epoch": 0.5627894782836669,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974514712864981,
      "loss": 5.7984,
      "mean_token_accuracy": 0.16196192055940628,
      "num_tokens": 13049085.0,
      "step": 6440
    },
    {
      "entropy": 5.994463205337524,
      "epoch": 0.5632264266363716,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974467923620639,
      "loss": 5.8206,
      "mean_token_accuracy": 0.16359213888645172,
      "num_tokens": 13060010.0,
      "step": 6445
    },
    {
      "entropy": 6.143889951705932,
      "epoch": 0.5636633749890763,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974421091709836,
      "loss": 5.8093,
      "mean_token_accuracy": 0.15848844796419143,
      "num_tokens": 13069768.0,
      "step": 6450
    },
    {
      "entropy": 6.030797624588013,
      "epoch": 0.564100323341781,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974374217133471,
      "loss": 5.8142,
      "mean_token_accuracy": 0.1624893218278885,
      "num_tokens": 13079842.0,
      "step": 6455
    },
    {
      "entropy": 5.989433717727661,
      "epoch": 0.5645372716944858,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974327299892443,
      "loss": 5.7393,
      "mean_token_accuracy": 0.16453002244234086,
      "num_tokens": 13089235.0,
      "step": 6460
    },
    {
      "entropy": 6.03585410118103,
      "epoch": 0.5649742200471904,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974280339987651,
      "loss": 5.8001,
      "mean_token_accuracy": 0.15564166158437728,
      "num_tokens": 13098364.0,
      "step": 6465
    },
    {
      "entropy": 6.0769377708435055,
      "epoch": 0.5654111683998951,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974233337419997,
      "loss": 5.922,
      "mean_token_accuracy": 0.15257405638694763,
      "num_tokens": 13108876.0,
      "step": 6470
    },
    {
      "entropy": 6.098965406417847,
      "epoch": 0.5658481167525998,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974186292190383,
      "loss": 5.9063,
      "mean_token_accuracy": 0.15851333886384963,
      "num_tokens": 13119591.0,
      "step": 6475
    },
    {
      "entropy": 6.082044172286987,
      "epoch": 0.5662850651053045,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497413920429971,
      "loss": 5.798,
      "mean_token_accuracy": 0.1702195480465889,
      "num_tokens": 13129796.0,
      "step": 6480
    },
    {
      "entropy": 6.018700838088989,
      "epoch": 0.5667220134580092,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974092073748882,
      "loss": 5.7472,
      "mean_token_accuracy": 0.16556535363197328,
      "num_tokens": 13139302.0,
      "step": 6485
    },
    {
      "entropy": 5.997605228424073,
      "epoch": 0.567158961810714,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974044900538802,
      "loss": 5.8179,
      "mean_token_accuracy": 0.1617901384830475,
      "num_tokens": 13148983.0,
      "step": 6490
    },
    {
      "entropy": 6.074260568618774,
      "epoch": 0.5675959101634187,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973997684670375,
      "loss": 5.7897,
      "mean_token_accuracy": 0.16193607151508332,
      "num_tokens": 13159016.0,
      "step": 6495
    },
    {
      "entropy": 6.045578336715698,
      "epoch": 0.5680328585161234,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973950426144508,
      "loss": 5.8197,
      "mean_token_accuracy": 0.1565999999642372,
      "num_tokens": 13169705.0,
      "step": 6500
    },
    {
      "entropy": 6.08186149597168,
      "epoch": 0.5684698068688281,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973903124962106,
      "loss": 5.8059,
      "mean_token_accuracy": 0.154049614071846,
      "num_tokens": 13181372.0,
      "step": 6505
    },
    {
      "entropy": 5.992932987213135,
      "epoch": 0.5689067552215328,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973855781124077,
      "loss": 5.8521,
      "mean_token_accuracy": 0.1572001427412033,
      "num_tokens": 13192581.0,
      "step": 6510
    },
    {
      "entropy": 6.075012969970703,
      "epoch": 0.5693437035742375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973808394631329,
      "loss": 5.8597,
      "mean_token_accuracy": 0.15612375736236572,
      "num_tokens": 13203269.0,
      "step": 6515
    },
    {
      "entropy": 6.040838146209717,
      "epoch": 0.5697806519269423,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497376096548477,
      "loss": 5.7384,
      "mean_token_accuracy": 0.1658146023750305,
      "num_tokens": 13213183.0,
      "step": 6520
    },
    {
      "entropy": 6.025716781616211,
      "epoch": 0.570217600279647,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973713493685311,
      "loss": 5.831,
      "mean_token_accuracy": 0.16171689182519913,
      "num_tokens": 13223196.0,
      "step": 6525
    },
    {
      "entropy": 6.002033138275147,
      "epoch": 0.5706545486323517,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973665979233861,
      "loss": 5.761,
      "mean_token_accuracy": 0.16273187100887299,
      "num_tokens": 13233915.0,
      "step": 6530
    },
    {
      "entropy": 5.9987036228179935,
      "epoch": 0.5710914969850563,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973618422131334,
      "loss": 5.785,
      "mean_token_accuracy": 0.16429124772548676,
      "num_tokens": 13245108.0,
      "step": 6535
    },
    {
      "entropy": 5.980170583724975,
      "epoch": 0.571528445337761,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497357082237864,
      "loss": 5.8019,
      "mean_token_accuracy": 0.16626695841550826,
      "num_tokens": 13255508.0,
      "step": 6540
    },
    {
      "entropy": 6.060469150543213,
      "epoch": 0.5719653936904658,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973523179976692,
      "loss": 5.8261,
      "mean_token_accuracy": 0.16567044854164123,
      "num_tokens": 13265118.0,
      "step": 6545
    },
    {
      "entropy": 6.0430152893066404,
      "epoch": 0.5724023420431705,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973475494926404,
      "loss": 5.7439,
      "mean_token_accuracy": 0.1659119501709938,
      "num_tokens": 13275947.0,
      "step": 6550
    },
    {
      "entropy": 6.00872802734375,
      "epoch": 0.5728392903958752,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973427767228692,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16719794422388076,
      "num_tokens": 13284662.0,
      "step": 6555
    },
    {
      "entropy": 6.034257078170777,
      "epoch": 0.5732762387485799,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497337999688447,
      "loss": 5.8243,
      "mean_token_accuracy": 0.16217348575592042,
      "num_tokens": 13295627.0,
      "step": 6560
    },
    {
      "entropy": 6.033489847183228,
      "epoch": 0.5737131871012846,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973332183894653,
      "loss": 5.7332,
      "mean_token_accuracy": 0.16402023732662202,
      "num_tokens": 13305205.0,
      "step": 6565
    },
    {
      "entropy": 5.938464498519897,
      "epoch": 0.5741501354539893,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973284328260162,
      "loss": 5.7685,
      "mean_token_accuracy": 0.16244693845510483,
      "num_tokens": 13315464.0,
      "step": 6570
    },
    {
      "entropy": 5.962944173812867,
      "epoch": 0.5745870838066941,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973236429981912,
      "loss": 5.6816,
      "mean_token_accuracy": 0.16979431509971618,
      "num_tokens": 13325468.0,
      "step": 6575
    },
    {
      "entropy": 6.046483421325684,
      "epoch": 0.5750240321593988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973188489060821,
      "loss": 5.8045,
      "mean_token_accuracy": 0.16544649451971055,
      "num_tokens": 13334719.0,
      "step": 6580
    },
    {
      "entropy": 6.004626083374023,
      "epoch": 0.5754609805121035,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973140505497812,
      "loss": 5.7492,
      "mean_token_accuracy": 0.1596517950296402,
      "num_tokens": 13344545.0,
      "step": 6585
    },
    {
      "entropy": 6.034828853607178,
      "epoch": 0.5758979288648082,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973092479293802,
      "loss": 5.8422,
      "mean_token_accuracy": 0.15655328184366227,
      "num_tokens": 13355212.0,
      "step": 6590
    },
    {
      "entropy": 6.017830896377563,
      "epoch": 0.5763348772175129,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973044410449713,
      "loss": 5.7379,
      "mean_token_accuracy": 0.1642249494791031,
      "num_tokens": 13364869.0,
      "step": 6595
    },
    {
      "entropy": 6.018236780166626,
      "epoch": 0.5767718255702176,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972996298966468,
      "loss": 5.7715,
      "mean_token_accuracy": 0.1615628108382225,
      "num_tokens": 13376511.0,
      "step": 6600
    },
    {
      "entropy": 5.988927555084229,
      "epoch": 0.5772087739229224,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497294814484499,
      "loss": 5.8136,
      "mean_token_accuracy": 0.15847112536430358,
      "num_tokens": 13385470.0,
      "step": 6605
    },
    {
      "entropy": 6.0295729637146,
      "epoch": 0.577645722275627,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049728999480862,
      "loss": 5.8575,
      "mean_token_accuracy": 0.1639319285750389,
      "num_tokens": 13395328.0,
      "step": 6610
    },
    {
      "entropy": 6.044540071487427,
      "epoch": 0.5780826706283317,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972851708691025,
      "loss": 5.7666,
      "mean_token_accuracy": 0.1704891562461853,
      "num_tokens": 13405319.0,
      "step": 6615
    },
    {
      "entropy": 5.931734704971314,
      "epoch": 0.5785196189810364,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972803426660389,
      "loss": 5.764,
      "mean_token_accuracy": 0.16114667057991028,
      "num_tokens": 13416119.0,
      "step": 6620
    },
    {
      "entropy": 6.040623903274536,
      "epoch": 0.5789565673337411,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972755101995219,
      "loss": 5.7677,
      "mean_token_accuracy": 0.1594271883368492,
      "num_tokens": 13425824.0,
      "step": 6625
    },
    {
      "entropy": 5.955241966247558,
      "epoch": 0.5793935156864458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972706734696442,
      "loss": 5.69,
      "mean_token_accuracy": 0.1713080585002899,
      "num_tokens": 13435506.0,
      "step": 6630
    },
    {
      "entropy": 5.9656919002532955,
      "epoch": 0.5798304640391506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972658324764984,
      "loss": 5.7223,
      "mean_token_accuracy": 0.17440096586942672,
      "num_tokens": 13445313.0,
      "step": 6635
    },
    {
      "entropy": 5.9706391334533695,
      "epoch": 0.5802674123918553,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972609872201775,
      "loss": 5.6758,
      "mean_token_accuracy": 0.17269710451364517,
      "num_tokens": 13455312.0,
      "step": 6640
    },
    {
      "entropy": 6.0071448802948,
      "epoch": 0.58070436074456,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972561377007744,
      "loss": 5.8009,
      "mean_token_accuracy": 0.15826916098594665,
      "num_tokens": 13465721.0,
      "step": 6645
    },
    {
      "entropy": 6.044702482223511,
      "epoch": 0.5811413090972647,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972512839183822,
      "loss": 5.8232,
      "mean_token_accuracy": 0.1616437941789627,
      "num_tokens": 13477185.0,
      "step": 6650
    },
    {
      "entropy": 5.944295978546142,
      "epoch": 0.5815782574499694,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972464258730939,
      "loss": 5.7305,
      "mean_token_accuracy": 0.17324860095977784,
      "num_tokens": 13487846.0,
      "step": 6655
    },
    {
      "entropy": 5.950640869140625,
      "epoch": 0.5820152058026741,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972415635650027,
      "loss": 5.7962,
      "mean_token_accuracy": 0.17441733479499816,
      "num_tokens": 13498175.0,
      "step": 6660
    },
    {
      "entropy": 6.110525417327881,
      "epoch": 0.5824521541553789,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972366969942017,
      "loss": 5.7517,
      "mean_token_accuracy": 0.16434309780597686,
      "num_tokens": 13506864.0,
      "step": 6665
    },
    {
      "entropy": 5.958086347579956,
      "epoch": 0.5828891025080836,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972318261607845,
      "loss": 5.7629,
      "mean_token_accuracy": 0.16661737561225892,
      "num_tokens": 13516208.0,
      "step": 6670
    },
    {
      "entropy": 6.049145126342774,
      "epoch": 0.5833260508607883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972269510648444,
      "loss": 5.817,
      "mean_token_accuracy": 0.1596122846007347,
      "num_tokens": 13526902.0,
      "step": 6675
    },
    {
      "entropy": 6.039386129379272,
      "epoch": 0.583762999213493,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497222071706475,
      "loss": 5.7332,
      "mean_token_accuracy": 0.17275967001914977,
      "num_tokens": 13537001.0,
      "step": 6680
    },
    {
      "entropy": 5.895462369918823,
      "epoch": 0.5841999475661976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972171880857698,
      "loss": 5.7342,
      "mean_token_accuracy": 0.16966630518436432,
      "num_tokens": 13546971.0,
      "step": 6685
    },
    {
      "entropy": 5.978184461593628,
      "epoch": 0.5846368959189023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972123002028226,
      "loss": 5.7027,
      "mean_token_accuracy": 0.17341143935918807,
      "num_tokens": 13557202.0,
      "step": 6690
    },
    {
      "entropy": 5.938102197647095,
      "epoch": 0.5850738442716071,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972074080577269,
      "loss": 5.7191,
      "mean_token_accuracy": 0.1736176684498787,
      "num_tokens": 13567291.0,
      "step": 6695
    },
    {
      "entropy": 5.975471305847168,
      "epoch": 0.5855107926243118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972025116505767,
      "loss": 5.7814,
      "mean_token_accuracy": 0.17222594171762468,
      "num_tokens": 13576864.0,
      "step": 6700
    },
    {
      "entropy": 6.000371646881104,
      "epoch": 0.5859477409770165,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004971976109814658,
      "loss": 5.7048,
      "mean_token_accuracy": 0.17231739312410355,
      "num_tokens": 13586728.0,
      "step": 6705
    },
    {
      "entropy": 6.0185778617858885,
      "epoch": 0.5863846893297212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971927060504885,
      "loss": 5.7614,
      "mean_token_accuracy": 0.16785011440515518,
      "num_tokens": 13596875.0,
      "step": 6710
    },
    {
      "entropy": 6.029244041442871,
      "epoch": 0.5868216376824259,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004971877968577386,
      "loss": 5.786,
      "mean_token_accuracy": 0.1565221890807152,
      "num_tokens": 13606483.0,
      "step": 6715
    },
    {
      "entropy": 6.0178032398223875,
      "epoch": 0.5872585860351307,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971828834033102,
      "loss": 5.8273,
      "mean_token_accuracy": 0.16335055679082872,
      "num_tokens": 13616780.0,
      "step": 6720
    },
    {
      "entropy": 6.00572099685669,
      "epoch": 0.5876955343878354,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971779656872978,
      "loss": 5.7863,
      "mean_token_accuracy": 0.1611931726336479,
      "num_tokens": 13626418.0,
      "step": 6725
    },
    {
      "entropy": 6.040483665466309,
      "epoch": 0.5881324827405401,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971730437097955,
      "loss": 5.8493,
      "mean_token_accuracy": 0.16322196424007415,
      "num_tokens": 13636956.0,
      "step": 6730
    },
    {
      "entropy": 6.041242551803589,
      "epoch": 0.5885694310932448,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497168117470898,
      "loss": 5.7542,
      "mean_token_accuracy": 0.16862192153930664,
      "num_tokens": 13647458.0,
      "step": 6735
    },
    {
      "entropy": 5.960151958465576,
      "epoch": 0.5890063794459495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971631869706995,
      "loss": 5.7244,
      "mean_token_accuracy": 0.1706304743885994,
      "num_tokens": 13657655.0,
      "step": 6740
    },
    {
      "entropy": 6.029302740097046,
      "epoch": 0.5894433277986542,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971582522092946,
      "loss": 5.8157,
      "mean_token_accuracy": 0.16030220985412597,
      "num_tokens": 13668472.0,
      "step": 6745
    },
    {
      "entropy": 6.012929153442383,
      "epoch": 0.589880276151359,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497153313186778,
      "loss": 5.6884,
      "mean_token_accuracy": 0.17067762464284897,
      "num_tokens": 13677319.0,
      "step": 6750
    },
    {
      "entropy": 5.943634176254273,
      "epoch": 0.5903172245040637,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971483699032445,
      "loss": 5.7233,
      "mean_token_accuracy": 0.16815248876810074,
      "num_tokens": 13688037.0,
      "step": 6755
    },
    {
      "entropy": 6.066489267349243,
      "epoch": 0.5907541728567683,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971434223587888,
      "loss": 5.8413,
      "mean_token_accuracy": 0.1588745504617691,
      "num_tokens": 13698441.0,
      "step": 6760
    },
    {
      "entropy": 6.036239671707153,
      "epoch": 0.591191121209473,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497138470553506,
      "loss": 5.7615,
      "mean_token_accuracy": 0.17169107496738434,
      "num_tokens": 13708140.0,
      "step": 6765
    },
    {
      "entropy": 6.028025484085083,
      "epoch": 0.5916280695621777,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971335144874909,
      "loss": 5.7446,
      "mean_token_accuracy": 0.1690863698720932,
      "num_tokens": 13718069.0,
      "step": 6770
    },
    {
      "entropy": 5.942646074295044,
      "epoch": 0.5920650179148824,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971285541608385,
      "loss": 5.7172,
      "mean_token_accuracy": 0.1785912811756134,
      "num_tokens": 13727932.0,
      "step": 6775
    },
    {
      "entropy": 6.09266414642334,
      "epoch": 0.5925019662675872,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971235895736441,
      "loss": 5.8083,
      "mean_token_accuracy": 0.15703656375408173,
      "num_tokens": 13738239.0,
      "step": 6780
    },
    {
      "entropy": 6.008364343643189,
      "epoch": 0.5929389146202919,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971186207260029,
      "loss": 5.7605,
      "mean_token_accuracy": 0.16451359838247298,
      "num_tokens": 13748622.0,
      "step": 6785
    },
    {
      "entropy": 5.900318956375122,
      "epoch": 0.5933758629729966,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049711364761801,
      "loss": 5.6856,
      "mean_token_accuracy": 0.17163221538066864,
      "num_tokens": 13759221.0,
      "step": 6790
    },
    {
      "entropy": 6.038178110122681,
      "epoch": 0.5938128113257013,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497108670249761,
      "loss": 5.7098,
      "mean_token_accuracy": 0.1669810891151428,
      "num_tokens": 13768343.0,
      "step": 6795
    },
    {
      "entropy": 5.9695981502532955,
      "epoch": 0.594249759678406,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971036886213515,
      "loss": 5.7018,
      "mean_token_accuracy": 0.16564354598522185,
      "num_tokens": 13778225.0,
      "step": 6800
    },
    {
      "entropy": 6.02821717262268,
      "epoch": 0.5946867080311107,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970987027328767,
      "loss": 5.7593,
      "mean_token_accuracy": 0.16254900097846986,
      "num_tokens": 13788633.0,
      "step": 6805
    },
    {
      "entropy": 6.022892713546753,
      "epoch": 0.5951236563838155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970937125844325,
      "loss": 5.6956,
      "mean_token_accuracy": 0.16897351816296577,
      "num_tokens": 13798172.0,
      "step": 6810
    },
    {
      "entropy": 5.875032138824463,
      "epoch": 0.5955606047365202,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970887181761144,
      "loss": 5.6802,
      "mean_token_accuracy": 0.1754966512322426,
      "num_tokens": 13807284.0,
      "step": 6815
    },
    {
      "entropy": 6.019988870620727,
      "epoch": 0.5959975530892249,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970837195080184,
      "loss": 5.7496,
      "mean_token_accuracy": 0.1685019001364708,
      "num_tokens": 13816807.0,
      "step": 6820
    },
    {
      "entropy": 5.991433811187744,
      "epoch": 0.5964345014419296,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970787165802403,
      "loss": 5.7153,
      "mean_token_accuracy": 0.16499842554330826,
      "num_tokens": 13827635.0,
      "step": 6825
    },
    {
      "entropy": 5.961988353729248,
      "epoch": 0.5968714497946342,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497073709392876,
      "loss": 5.775,
      "mean_token_accuracy": 0.16304441243410112,
      "num_tokens": 13838060.0,
      "step": 6830
    },
    {
      "entropy": 6.066063690185547,
      "epoch": 0.5973083981473389,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970686979460216,
      "loss": 5.7833,
      "mean_token_accuracy": 0.16161904036998748,
      "num_tokens": 13848669.0,
      "step": 6835
    },
    {
      "entropy": 5.945089292526245,
      "epoch": 0.5977453465000437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970636822397733,
      "loss": 5.74,
      "mean_token_accuracy": 0.1709183618426323,
      "num_tokens": 13857949.0,
      "step": 6840
    },
    {
      "entropy": 5.982461023330688,
      "epoch": 0.5981822948527484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970586622742272,
      "loss": 5.7277,
      "mean_token_accuracy": 0.17384159415960312,
      "num_tokens": 13867726.0,
      "step": 6845
    },
    {
      "entropy": 6.078283500671387,
      "epoch": 0.5986192432054531,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970536380494795,
      "loss": 5.8098,
      "mean_token_accuracy": 0.15677576810121535,
      "num_tokens": 13878072.0,
      "step": 6850
    },
    {
      "entropy": 5.914581346511841,
      "epoch": 0.5990561915581578,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970486095656268,
      "loss": 5.6775,
      "mean_token_accuracy": 0.17753330618143082,
      "num_tokens": 13887771.0,
      "step": 6855
    },
    {
      "entropy": 6.009640264511108,
      "epoch": 0.5994931399108625,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970435768227652,
      "loss": 5.7141,
      "mean_token_accuracy": 0.17114368677139283,
      "num_tokens": 13897867.0,
      "step": 6860
    },
    {
      "entropy": 6.022708559036255,
      "epoch": 0.5999300882635672,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970385398209916,
      "loss": 5.7781,
      "mean_token_accuracy": 0.166123603284359,
      "num_tokens": 13909171.0,
      "step": 6865
    },
    {
      "entropy": 6.013170289993286,
      "epoch": 0.600367036616272,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970334985604024,
      "loss": 5.7753,
      "mean_token_accuracy": 0.16740324944257737,
      "num_tokens": 13920374.0,
      "step": 6870
    },
    {
      "entropy": 6.021105098724365,
      "epoch": 0.6008039849689767,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970284530410945,
      "loss": 5.7978,
      "mean_token_accuracy": 0.16199224442243576,
      "num_tokens": 13930150.0,
      "step": 6875
    },
    {
      "entropy": 5.991156482696534,
      "epoch": 0.6012409333216814,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004970234032631644,
      "loss": 5.7287,
      "mean_token_accuracy": 0.1705200642347336,
      "num_tokens": 13940553.0,
      "step": 6880
    },
    {
      "entropy": 5.973852014541626,
      "epoch": 0.6016778816743861,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970183492267091,
      "loss": 5.8187,
      "mean_token_accuracy": 0.16620455533266068,
      "num_tokens": 13951086.0,
      "step": 6885
    },
    {
      "entropy": 6.029850864410401,
      "epoch": 0.6021148300270908,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970132909318254,
      "loss": 5.7366,
      "mean_token_accuracy": 0.16540056467056274,
      "num_tokens": 13960684.0,
      "step": 6890
    },
    {
      "entropy": 5.868367910385132,
      "epoch": 0.6025517783797955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970082283786107,
      "loss": 5.5821,
      "mean_token_accuracy": 0.18669819235801696,
      "num_tokens": 13971409.0,
      "step": 6895
    },
    {
      "entropy": 5.970133543014526,
      "epoch": 0.6029887267325003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970031615671617,
      "loss": 5.7717,
      "mean_token_accuracy": 0.16878408938646317,
      "num_tokens": 13981185.0,
      "step": 6900
    },
    {
      "entropy": 6.0089240074157715,
      "epoch": 0.603425675085205,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969980904975758,
      "loss": 5.792,
      "mean_token_accuracy": 0.16326660066843032,
      "num_tokens": 13991670.0,
      "step": 6905
    },
    {
      "entropy": 5.940281343460083,
      "epoch": 0.6038626234379096,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004969930151699501,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1709388792514801,
      "num_tokens": 14002343.0,
      "step": 6910
    },
    {
      "entropy": 5.992174863815308,
      "epoch": 0.6042995717906143,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496987935584382,
      "loss": 5.7974,
      "mean_token_accuracy": 0.16177349239587785,
      "num_tokens": 14013373.0,
      "step": 6915
    },
    {
      "entropy": 6.03256049156189,
      "epoch": 0.604736520143319,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969828517409693,
      "loss": 5.7124,
      "mean_token_accuracy": 0.17402544915676116,
      "num_tokens": 14023233.0,
      "step": 6920
    },
    {
      "entropy": 5.93929762840271,
      "epoch": 0.6051734684960238,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496977763639809,
      "loss": 5.7788,
      "mean_token_accuracy": 0.1657380759716034,
      "num_tokens": 14033054.0,
      "step": 6925
    },
    {
      "entropy": 5.980549430847168,
      "epoch": 0.6056104168487285,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969726712809988,
      "loss": 5.6947,
      "mean_token_accuracy": 0.1625191316008568,
      "num_tokens": 14042039.0,
      "step": 6930
    },
    {
      "entropy": 6.054283714294433,
      "epoch": 0.6060473652014332,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004969675746646364,
      "loss": 5.843,
      "mean_token_accuracy": 0.15921522229909896,
      "num_tokens": 14051957.0,
      "step": 6935
    },
    {
      "entropy": 6.024473237991333,
      "epoch": 0.6064843135541379,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004969624737908197,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1666837677359581,
      "num_tokens": 14062438.0,
      "step": 6940
    },
    {
      "entropy": 6.0115140914917,
      "epoch": 0.6069212619068426,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004969573686596464,
      "loss": 5.7404,
      "mean_token_accuracy": 0.16234497874975204,
      "num_tokens": 14073196.0,
      "step": 6945
    },
    {
      "entropy": 5.99122576713562,
      "epoch": 0.6073582102595473,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969522592712146,
      "loss": 5.7431,
      "mean_token_accuracy": 0.16431919932365419,
      "num_tokens": 14083462.0,
      "step": 6950
    },
    {
      "entropy": 5.965324831008911,
      "epoch": 0.6077951586122521,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969471456256221,
      "loss": 5.8182,
      "mean_token_accuracy": 0.16682534366846086,
      "num_tokens": 14094194.0,
      "step": 6955
    },
    {
      "entropy": 6.063554525375366,
      "epoch": 0.6082321069649568,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496942027722967,
      "loss": 5.7405,
      "mean_token_accuracy": 0.1664691224694252,
      "num_tokens": 14104553.0,
      "step": 6960
    },
    {
      "entropy": 6.032431411743164,
      "epoch": 0.6086690553176615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004969369055633476,
      "loss": 5.6963,
      "mean_token_accuracy": 0.16351156830787658,
      "num_tokens": 14114140.0,
      "step": 6965
    },
    {
      "entropy": 5.9739494800567625,
      "epoch": 0.6091060036703662,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496931779146862,
      "loss": 5.8235,
      "mean_token_accuracy": 0.1612629547715187,
      "num_tokens": 14124723.0,
      "step": 6970
    },
    {
      "entropy": 5.9937139511108395,
      "epoch": 0.6095429520230708,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969266484736085,
      "loss": 5.6879,
      "mean_token_accuracy": 0.16567064672708512,
      "num_tokens": 14134030.0,
      "step": 6975
    },
    {
      "entropy": 5.995045185089111,
      "epoch": 0.6099799003757755,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004969215135436858,
      "loss": 5.6622,
      "mean_token_accuracy": 0.16846228390932083,
      "num_tokens": 14144691.0,
      "step": 6980
    },
    {
      "entropy": 6.013519906997681,
      "epoch": 0.6104168487284803,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969163743571921,
      "loss": 5.8183,
      "mean_token_accuracy": 0.16563833504915237,
      "num_tokens": 14155114.0,
      "step": 6985
    },
    {
      "entropy": 6.0187455177307125,
      "epoch": 0.610853797081185,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000496911230914226,
      "loss": 5.8344,
      "mean_token_accuracy": 0.1536600798368454,
      "num_tokens": 14166097.0,
      "step": 6990
    },
    {
      "entropy": 5.981110954284668,
      "epoch": 0.6112907454338897,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969060832148862,
      "loss": 5.68,
      "mean_token_accuracy": 0.16358502805233002,
      "num_tokens": 14176538.0,
      "step": 6995
    },
    {
      "entropy": 5.966767024993897,
      "epoch": 0.6117276937865944,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969009312592715,
      "loss": 5.698,
      "mean_token_accuracy": 0.16401465088129044,
      "num_tokens": 14187912.0,
      "step": 7000
    },
    {
      "entropy": 5.886053943634034,
      "epoch": 0.6121646421392991,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968957750474806,
      "loss": 5.604,
      "mean_token_accuracy": 0.18139319717884064,
      "num_tokens": 14198446.0,
      "step": 7005
    },
    {
      "entropy": 6.09174108505249,
      "epoch": 0.6126015904920038,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968906145796126,
      "loss": 5.8418,
      "mean_token_accuracy": 0.16500025242567062,
      "num_tokens": 14207723.0,
      "step": 7010
    },
    {
      "entropy": 6.002839708328247,
      "epoch": 0.6130385388447086,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968854498557662,
      "loss": 5.7152,
      "mean_token_accuracy": 0.16631795316934586,
      "num_tokens": 14217590.0,
      "step": 7015
    },
    {
      "entropy": 5.982351922988892,
      "epoch": 0.6134754871974133,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968802808760407,
      "loss": 5.7843,
      "mean_token_accuracy": 0.1620543345808983,
      "num_tokens": 14226629.0,
      "step": 7020
    },
    {
      "entropy": 6.061549711227417,
      "epoch": 0.613912435550118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968751076405351,
      "loss": 5.8511,
      "mean_token_accuracy": 0.15419451892375946,
      "num_tokens": 14235782.0,
      "step": 7025
    },
    {
      "entropy": 6.006581449508667,
      "epoch": 0.6143493839028227,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968699301493488,
      "loss": 5.7115,
      "mean_token_accuracy": 0.17608898580074311,
      "num_tokens": 14246687.0,
      "step": 7030
    },
    {
      "entropy": 6.040561676025391,
      "epoch": 0.6147863322555274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968647484025808,
      "loss": 5.7554,
      "mean_token_accuracy": 0.16642000526189804,
      "num_tokens": 14255961.0,
      "step": 7035
    },
    {
      "entropy": 6.026982927322388,
      "epoch": 0.6152232806082321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004968595624003307,
      "loss": 5.7603,
      "mean_token_accuracy": 0.16782544702291488,
      "num_tokens": 14266024.0,
      "step": 7040
    },
    {
      "entropy": 5.940685987472534,
      "epoch": 0.6156602289609369,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496854372142698,
      "loss": 5.7804,
      "mean_token_accuracy": 0.1642189458012581,
      "num_tokens": 14276590.0,
      "step": 7045
    },
    {
      "entropy": 6.017156457901001,
      "epoch": 0.6160971773136416,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968491776297821,
      "loss": 5.8049,
      "mean_token_accuracy": 0.16669025868177414,
      "num_tokens": 14285992.0,
      "step": 7050
    },
    {
      "entropy": 5.982899475097656,
      "epoch": 0.6165341256663462,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968439788616829,
      "loss": 5.71,
      "mean_token_accuracy": 0.1662043422460556,
      "num_tokens": 14295843.0,
      "step": 7055
    },
    {
      "entropy": 5.965174198150635,
      "epoch": 0.6169710740190509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004968387758384999,
      "loss": 5.7361,
      "mean_token_accuracy": 0.17207943350076677,
      "num_tokens": 14305643.0,
      "step": 7060
    },
    {
      "entropy": 6.0073771476745605,
      "epoch": 0.6174080223717556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968335685603328,
      "loss": 5.7,
      "mean_token_accuracy": 0.17139963209629058,
      "num_tokens": 14315907.0,
      "step": 7065
    },
    {
      "entropy": 5.992789316177368,
      "epoch": 0.6178449707244603,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968283570272817,
      "loss": 5.7885,
      "mean_token_accuracy": 0.16036105304956436,
      "num_tokens": 14325893.0,
      "step": 7070
    },
    {
      "entropy": 6.004592227935791,
      "epoch": 0.6182819190771651,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968231412394464,
      "loss": 5.7691,
      "mean_token_accuracy": 0.1630422964692116,
      "num_tokens": 14337510.0,
      "step": 7075
    },
    {
      "entropy": 6.010964870452881,
      "epoch": 0.6187188674298698,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968179211969272,
      "loss": 5.7089,
      "mean_token_accuracy": 0.17299352288246156,
      "num_tokens": 14346373.0,
      "step": 7080
    },
    {
      "entropy": 5.878944158554077,
      "epoch": 0.6191558157825745,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004968126968998239,
      "loss": 5.6008,
      "mean_token_accuracy": 0.18944043219089507,
      "num_tokens": 14356303.0,
      "step": 7085
    },
    {
      "entropy": 5.978065109252929,
      "epoch": 0.6195927641352792,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968074683482369,
      "loss": 5.7511,
      "mean_token_accuracy": 0.16676455289125441,
      "num_tokens": 14366171.0,
      "step": 7090
    },
    {
      "entropy": 6.064041090011597,
      "epoch": 0.6200297124879839,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968022355422663,
      "loss": 5.8421,
      "mean_token_accuracy": 0.16274562776088713,
      "num_tokens": 14376133.0,
      "step": 7095
    },
    {
      "entropy": 6.024569272994995,
      "epoch": 0.6204666608406886,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967969984820128,
      "loss": 5.7207,
      "mean_token_accuracy": 0.1661253735423088,
      "num_tokens": 14385670.0,
      "step": 7100
    },
    {
      "entropy": 5.985086059570312,
      "epoch": 0.6209036091933934,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967917571675765,
      "loss": 5.7342,
      "mean_token_accuracy": 0.1697813868522644,
      "num_tokens": 14396002.0,
      "step": 7105
    },
    {
      "entropy": 6.006342267990112,
      "epoch": 0.6213405575460981,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967865115990582,
      "loss": 5.8343,
      "mean_token_accuracy": 0.1584721550345421,
      "num_tokens": 14406718.0,
      "step": 7110
    },
    {
      "entropy": 6.1160389423370365,
      "epoch": 0.6217775058988028,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967812617765583,
      "loss": 5.7415,
      "mean_token_accuracy": 0.16247805655002595,
      "num_tokens": 14417027.0,
      "step": 7115
    },
    {
      "entropy": 5.91852765083313,
      "epoch": 0.6222144542515075,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967760077001775,
      "loss": 5.7146,
      "mean_token_accuracy": 0.17086390554904937,
      "num_tokens": 14426885.0,
      "step": 7120
    },
    {
      "entropy": 5.935893821716308,
      "epoch": 0.6226514026042121,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004967707493700169,
      "loss": 5.7166,
      "mean_token_accuracy": 0.16546571552753447,
      "num_tokens": 14437743.0,
      "step": 7125
    },
    {
      "entropy": 5.953591823577881,
      "epoch": 0.623088350956917,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967654867861769,
      "loss": 5.7294,
      "mean_token_accuracy": 0.17408292442560197,
      "num_tokens": 14448808.0,
      "step": 7130
    },
    {
      "entropy": 5.946217441558838,
      "epoch": 0.6235252993096216,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004967602199487588,
      "loss": 5.7562,
      "mean_token_accuracy": 0.17197027057409286,
      "num_tokens": 14458560.0,
      "step": 7135
    },
    {
      "entropy": 6.0128960609436035,
      "epoch": 0.6239622476623263,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967549488578635,
      "loss": 5.8137,
      "mean_token_accuracy": 0.1625184491276741,
      "num_tokens": 14469375.0,
      "step": 7140
    },
    {
      "entropy": 6.043777799606323,
      "epoch": 0.624399196015031,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496749673513592,
      "loss": 5.7495,
      "mean_token_accuracy": 0.16366735398769378,
      "num_tokens": 14478444.0,
      "step": 7145
    },
    {
      "entropy": 5.946481323242187,
      "epoch": 0.6248361443677357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967443939160456,
      "loss": 5.6192,
      "mean_token_accuracy": 0.17665784657001496,
      "num_tokens": 14487776.0,
      "step": 7150
    },
    {
      "entropy": 5.971036386489868,
      "epoch": 0.6252730927204404,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967391100653255,
      "loss": 5.7361,
      "mean_token_accuracy": 0.16909721344709397,
      "num_tokens": 14497879.0,
      "step": 7155
    },
    {
      "entropy": 5.95577220916748,
      "epoch": 0.6257100410731452,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967338219615331,
      "loss": 5.6797,
      "mean_token_accuracy": 0.16984783709049225,
      "num_tokens": 14507753.0,
      "step": 7160
    },
    {
      "entropy": 6.012106704711914,
      "epoch": 0.6261469894258499,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967285296047699,
      "loss": 5.6508,
      "mean_token_accuracy": 0.17839130461215974,
      "num_tokens": 14517086.0,
      "step": 7165
    },
    {
      "entropy": 5.928661966323853,
      "epoch": 0.6265839377785546,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967232329951372,
      "loss": 5.7213,
      "mean_token_accuracy": 0.1626708909869194,
      "num_tokens": 14527097.0,
      "step": 7170
    },
    {
      "entropy": 5.937558555603028,
      "epoch": 0.6270208861312593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967179321327367,
      "loss": 5.7723,
      "mean_token_accuracy": 0.1668960377573967,
      "num_tokens": 14537924.0,
      "step": 7175
    },
    {
      "entropy": 6.049029016494751,
      "epoch": 0.627457834483964,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967126270176702,
      "loss": 5.7633,
      "mean_token_accuracy": 0.16293965727090837,
      "num_tokens": 14547542.0,
      "step": 7180
    },
    {
      "entropy": 5.977898406982422,
      "epoch": 0.6278947828366687,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967073176500392,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1737346902489662,
      "num_tokens": 14557150.0,
      "step": 7185
    },
    {
      "entropy": 6.022466087341309,
      "epoch": 0.6283317311893735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004967020040299458,
      "loss": 5.7082,
      "mean_token_accuracy": 0.17263602912425996,
      "num_tokens": 14567242.0,
      "step": 7190
    },
    {
      "entropy": 5.899930810928344,
      "epoch": 0.6287686795420782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966966861574917,
      "loss": 5.6733,
      "mean_token_accuracy": 0.17001620978116988,
      "num_tokens": 14576631.0,
      "step": 7195
    },
    {
      "entropy": 5.9912927627563475,
      "epoch": 0.6292056278947828,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496691364032779,
      "loss": 5.7718,
      "mean_token_accuracy": 0.16450242400169374,
      "num_tokens": 14587731.0,
      "step": 7200
    },
    {
      "entropy": 6.078956651687622,
      "epoch": 0.6296425762474875,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966860376559097,
      "loss": 5.7722,
      "mean_token_accuracy": 0.16303039491176605,
      "num_tokens": 14598136.0,
      "step": 7205
    },
    {
      "entropy": 5.940353441238403,
      "epoch": 0.6300795246001922,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966807070269862,
      "loss": 5.7508,
      "mean_token_accuracy": 0.1646815523505211,
      "num_tokens": 14607243.0,
      "step": 7210
    },
    {
      "entropy": 5.98602294921875,
      "epoch": 0.6305164729528969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966753721461105,
      "loss": 5.7467,
      "mean_token_accuracy": 0.16294849663972855,
      "num_tokens": 14617094.0,
      "step": 7215
    },
    {
      "entropy": 6.042633581161499,
      "epoch": 0.6309534213056017,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966700330133849,
      "loss": 5.7809,
      "mean_token_accuracy": 0.1587809607386589,
      "num_tokens": 14626486.0,
      "step": 7220
    },
    {
      "entropy": 5.840781307220459,
      "epoch": 0.6313903696583064,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966646896289121,
      "loss": 5.6253,
      "mean_token_accuracy": 0.18178863525390626,
      "num_tokens": 14637170.0,
      "step": 7225
    },
    {
      "entropy": 5.97496600151062,
      "epoch": 0.6318273180110111,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966593419927944,
      "loss": 5.7148,
      "mean_token_accuracy": 0.16624498814344407,
      "num_tokens": 14646396.0,
      "step": 7230
    },
    {
      "entropy": 6.0501340389251705,
      "epoch": 0.6322642663637158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004966539901051343,
      "loss": 5.769,
      "mean_token_accuracy": 0.1566547393798828,
      "num_tokens": 14656097.0,
      "step": 7235
    },
    {
      "entropy": 5.9358161926269535,
      "epoch": 0.6327012147164205,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966486339660345,
      "loss": 5.7082,
      "mean_token_accuracy": 0.1680993154644966,
      "num_tokens": 14667124.0,
      "step": 7240
    },
    {
      "entropy": 5.922321081161499,
      "epoch": 0.6331381630691252,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966432735755978,
      "loss": 5.809,
      "mean_token_accuracy": 0.16217757910490035,
      "num_tokens": 14677427.0,
      "step": 7245
    },
    {
      "entropy": 5.939199638366699,
      "epoch": 0.63357511142183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496637908933927,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1744445726275444,
      "num_tokens": 14686429.0,
      "step": 7250
    },
    {
      "entropy": 6.007258749008178,
      "epoch": 0.6340120597745347,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496632540041125,
      "loss": 5.7201,
      "mean_token_accuracy": 0.17226899713277816,
      "num_tokens": 14696264.0,
      "step": 7255
    },
    {
      "entropy": 5.989048480987549,
      "epoch": 0.6344490081272394,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966271668972946,
      "loss": 5.7363,
      "mean_token_accuracy": 0.165238256752491,
      "num_tokens": 14705935.0,
      "step": 7260
    },
    {
      "entropy": 6.006023263931274,
      "epoch": 0.6348859564799441,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966217895025391,
      "loss": 5.7053,
      "mean_token_accuracy": 0.16480019241571425,
      "num_tokens": 14715978.0,
      "step": 7265
    },
    {
      "entropy": 6.012892055511474,
      "epoch": 0.6353229048326487,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004966164078569616,
      "loss": 5.6945,
      "mean_token_accuracy": 0.1627400055527687,
      "num_tokens": 14726170.0,
      "step": 7270
    },
    {
      "entropy": 5.937580633163452,
      "epoch": 0.6357598531853534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966110219606652,
      "loss": 5.7127,
      "mean_token_accuracy": 0.16892754435539245,
      "num_tokens": 14735775.0,
      "step": 7275
    },
    {
      "entropy": 5.9612280368804935,
      "epoch": 0.6361968015380582,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966056318137532,
      "loss": 5.6873,
      "mean_token_accuracy": 0.16577343493700028,
      "num_tokens": 14746143.0,
      "step": 7280
    },
    {
      "entropy": 6.010287666320801,
      "epoch": 0.6366337498907629,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966002374163293,
      "loss": 5.7384,
      "mean_token_accuracy": 0.17045456022024155,
      "num_tokens": 14756707.0,
      "step": 7285
    },
    {
      "entropy": 5.969405221939087,
      "epoch": 0.6370706982434676,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965948387684964,
      "loss": 5.7273,
      "mean_token_accuracy": 0.16712143421173095,
      "num_tokens": 14765778.0,
      "step": 7290
    },
    {
      "entropy": 5.9678810119628904,
      "epoch": 0.6375076465961723,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965894358703587,
      "loss": 5.7436,
      "mean_token_accuracy": 0.1642889380455017,
      "num_tokens": 14775264.0,
      "step": 7295
    },
    {
      "entropy": 5.984958028793335,
      "epoch": 0.637944594948877,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004965840287220193,
      "loss": 5.712,
      "mean_token_accuracy": 0.16904931366443635,
      "num_tokens": 14785754.0,
      "step": 7300
    },
    {
      "entropy": 5.980724430084228,
      "epoch": 0.6383815433015817,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965786173235821,
      "loss": 5.7618,
      "mean_token_accuracy": 0.16549545377492905,
      "num_tokens": 14795936.0,
      "step": 7305
    },
    {
      "entropy": 6.049427032470703,
      "epoch": 0.6388184916542865,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004965732016751511,
      "loss": 5.7578,
      "mean_token_accuracy": 0.16190102845430374,
      "num_tokens": 14805721.0,
      "step": 7310
    },
    {
      "entropy": 5.954456758499146,
      "epoch": 0.6392554400069912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965677817768299,
      "loss": 5.7654,
      "mean_token_accuracy": 0.16673341244459153,
      "num_tokens": 14816610.0,
      "step": 7315
    },
    {
      "entropy": 6.025742530822754,
      "epoch": 0.6396923883596959,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965623576287225,
      "loss": 5.752,
      "mean_token_accuracy": 0.16269836574792862,
      "num_tokens": 14826126.0,
      "step": 7320
    },
    {
      "entropy": 5.939180612564087,
      "epoch": 0.6401293367124006,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965569292309331,
      "loss": 5.6404,
      "mean_token_accuracy": 0.17330277413129808,
      "num_tokens": 14835657.0,
      "step": 7325
    },
    {
      "entropy": 5.988603782653809,
      "epoch": 0.6405662850651053,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965514965835655,
      "loss": 5.7281,
      "mean_token_accuracy": 0.16551056802272796,
      "num_tokens": 14845430.0,
      "step": 7330
    },
    {
      "entropy": 5.990453672409058,
      "epoch": 0.6410032334178101,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965460596867243,
      "loss": 5.7003,
      "mean_token_accuracy": 0.17055307775735856,
      "num_tokens": 14855459.0,
      "step": 7335
    },
    {
      "entropy": 5.9845317840576175,
      "epoch": 0.6414401817705148,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965406185405135,
      "loss": 5.7208,
      "mean_token_accuracy": 0.16966748535633086,
      "num_tokens": 14865375.0,
      "step": 7340
    },
    {
      "entropy": 5.975282621383667,
      "epoch": 0.6418771301232195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965351731450376,
      "loss": 5.6973,
      "mean_token_accuracy": 0.17416606545448304,
      "num_tokens": 14875211.0,
      "step": 7345
    },
    {
      "entropy": 5.9431600093841555,
      "epoch": 0.6423140784759241,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496529723500401,
      "loss": 5.7569,
      "mean_token_accuracy": 0.16756468266248703,
      "num_tokens": 14886221.0,
      "step": 7350
    },
    {
      "entropy": 6.012370157241821,
      "epoch": 0.6427510268286288,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965242696067081,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1724283680319786,
      "num_tokens": 14896002.0,
      "step": 7355
    },
    {
      "entropy": 6.036994695663452,
      "epoch": 0.6431879751813335,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004965188114640637,
      "loss": 5.7842,
      "mean_token_accuracy": 0.16416873931884765,
      "num_tokens": 14906311.0,
      "step": 7360
    },
    {
      "entropy": 6.02338809967041,
      "epoch": 0.6436249235340383,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965133490725725,
      "loss": 5.727,
      "mean_token_accuracy": 0.16458918005228043,
      "num_tokens": 14915890.0,
      "step": 7365
    },
    {
      "entropy": 5.981739807128906,
      "epoch": 0.644061871886743,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004965078824323391,
      "loss": 5.709,
      "mean_token_accuracy": 0.16600379347801208,
      "num_tokens": 14925891.0,
      "step": 7370
    },
    {
      "entropy": 5.966194343566895,
      "epoch": 0.6444988202394477,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004965024115434685,
      "loss": 5.7425,
      "mean_token_accuracy": 0.16522887200117112,
      "num_tokens": 14937272.0,
      "step": 7375
    },
    {
      "entropy": 6.024422883987427,
      "epoch": 0.6449357685921524,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004964969364060656,
      "loss": 5.7473,
      "mean_token_accuracy": 0.16929561644792557,
      "num_tokens": 14948135.0,
      "step": 7380
    },
    {
      "entropy": 6.054916667938232,
      "epoch": 0.6453727169448571,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964914570202353,
      "loss": 5.826,
      "mean_token_accuracy": 0.16069403886795045,
      "num_tokens": 14959254.0,
      "step": 7385
    },
    {
      "entropy": 5.989320135116577,
      "epoch": 0.6458096652975618,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964859733860829,
      "loss": 5.7367,
      "mean_token_accuracy": 0.1644895166158676,
      "num_tokens": 14969313.0,
      "step": 7390
    },
    {
      "entropy": 5.971711921691894,
      "epoch": 0.6462466136502666,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964804855037134,
      "loss": 5.7219,
      "mean_token_accuracy": 0.16340662389993668,
      "num_tokens": 14978963.0,
      "step": 7395
    },
    {
      "entropy": 6.063074064254761,
      "epoch": 0.6466835620029713,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964749933732322,
      "loss": 5.7475,
      "mean_token_accuracy": 0.16808250099420546,
      "num_tokens": 14989355.0,
      "step": 7400
    },
    {
      "entropy": 5.9691853523254395,
      "epoch": 0.647120510355676,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964694969947445,
      "loss": 5.7023,
      "mean_token_accuracy": 0.16891448199748993,
      "num_tokens": 14998995.0,
      "step": 7405
    },
    {
      "entropy": 5.974824905395508,
      "epoch": 0.6475574587083807,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964639963683558,
      "loss": 5.7511,
      "mean_token_accuracy": 0.1641545072197914,
      "num_tokens": 15008762.0,
      "step": 7410
    },
    {
      "entropy": 6.080143880844116,
      "epoch": 0.6479944070610854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964584914941716,
      "loss": 5.7045,
      "mean_token_accuracy": 0.1718972146511078,
      "num_tokens": 15020062.0,
      "step": 7415
    },
    {
      "entropy": 6.016230726242066,
      "epoch": 0.64843135541379,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964529823722975,
      "loss": 5.8585,
      "mean_token_accuracy": 0.16162025183439255,
      "num_tokens": 15030440.0,
      "step": 7420
    },
    {
      "entropy": 5.970029067993164,
      "epoch": 0.6488683037664948,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964474690028393,
      "loss": 5.7109,
      "mean_token_accuracy": 0.16440936774015427,
      "num_tokens": 15041709.0,
      "step": 7425
    },
    {
      "entropy": 5.996074438095093,
      "epoch": 0.6493052521191995,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964419513859025,
      "loss": 5.8177,
      "mean_token_accuracy": 0.16574871242046357,
      "num_tokens": 15052277.0,
      "step": 7430
    },
    {
      "entropy": 6.015154695510864,
      "epoch": 0.6497422004719042,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496436429521593,
      "loss": 5.7315,
      "mean_token_accuracy": 0.16495052576065064,
      "num_tokens": 15061870.0,
      "step": 7435
    },
    {
      "entropy": 5.886715698242187,
      "epoch": 0.6501791488246089,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964309034100168,
      "loss": 5.5993,
      "mean_token_accuracy": 0.17312809973955154,
      "num_tokens": 15071403.0,
      "step": 7440
    },
    {
      "entropy": 5.984305143356323,
      "epoch": 0.6506160971773136,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004964253730512799,
      "loss": 5.7007,
      "mean_token_accuracy": 0.17175365686416627,
      "num_tokens": 15081651.0,
      "step": 7445
    },
    {
      "entropy": 6.018542051315308,
      "epoch": 0.6510530455300183,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964198384454882,
      "loss": 5.7494,
      "mean_token_accuracy": 0.16727405786514282,
      "num_tokens": 15093403.0,
      "step": 7450
    },
    {
      "entropy": 6.021938276290894,
      "epoch": 0.6514899938827231,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496414299592748,
      "loss": 5.8318,
      "mean_token_accuracy": 0.15873448252677919,
      "num_tokens": 15103618.0,
      "step": 7455
    },
    {
      "entropy": 5.977236318588257,
      "epoch": 0.6519269422354278,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964087564931656,
      "loss": 5.7573,
      "mean_token_accuracy": 0.1736075311899185,
      "num_tokens": 15113926.0,
      "step": 7460
    },
    {
      "entropy": 6.038077974319458,
      "epoch": 0.6523638905881325,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964032091468472,
      "loss": 5.783,
      "mean_token_accuracy": 0.15996074229478835,
      "num_tokens": 15125302.0,
      "step": 7465
    },
    {
      "entropy": 6.00434889793396,
      "epoch": 0.6528008389408372,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004963976575538991,
      "loss": 5.6765,
      "mean_token_accuracy": 0.1734301656484604,
      "num_tokens": 15135943.0,
      "step": 7470
    },
    {
      "entropy": 5.948629665374756,
      "epoch": 0.6532377872935419,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496392101714428,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1737793743610382,
      "num_tokens": 15145001.0,
      "step": 7475
    },
    {
      "entropy": 5.964487361907959,
      "epoch": 0.6536747356462466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963865416285405,
      "loss": 5.6501,
      "mean_token_accuracy": 0.16480645686388015,
      "num_tokens": 15154567.0,
      "step": 7480
    },
    {
      "entropy": 6.038545703887939,
      "epoch": 0.6541116839989514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963809772963431,
      "loss": 5.8043,
      "mean_token_accuracy": 0.16554083079099655,
      "num_tokens": 15164311.0,
      "step": 7485
    },
    {
      "entropy": 5.989369297027588,
      "epoch": 0.6545486323516561,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963754087179425,
      "loss": 5.6959,
      "mean_token_accuracy": 0.16746420711278914,
      "num_tokens": 15175217.0,
      "step": 7490
    },
    {
      "entropy": 5.992207336425781,
      "epoch": 0.6549855807043607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963698358934455,
      "loss": 5.7139,
      "mean_token_accuracy": 0.16923602521419526,
      "num_tokens": 15184895.0,
      "step": 7495
    },
    {
      "entropy": 5.921563625335693,
      "epoch": 0.6554225290570654,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004963642588229591,
      "loss": 5.7035,
      "mean_token_accuracy": 0.1675943061709404,
      "num_tokens": 15194380.0,
      "step": 7500
    },
    {
      "entropy": 6.025741004943848,
      "epoch": 0.6558594774097701,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963586775065902,
      "loss": 5.7295,
      "mean_token_accuracy": 0.16358117759227753,
      "num_tokens": 15204364.0,
      "step": 7505
    },
    {
      "entropy": 6.021099281311035,
      "epoch": 0.6562964257624748,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963530919444459,
      "loss": 5.6901,
      "mean_token_accuracy": 0.1670904591679573,
      "num_tokens": 15214485.0,
      "step": 7510
    },
    {
      "entropy": 6.022384119033814,
      "epoch": 0.6567333741151796,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004963475021366333,
      "loss": 5.8354,
      "mean_token_accuracy": 0.1575244665145874,
      "num_tokens": 15224764.0,
      "step": 7515
    },
    {
      "entropy": 6.041271543502807,
      "epoch": 0.6571703224678843,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004963419080832597,
      "loss": 5.8341,
      "mean_token_accuracy": 0.16026235818862916,
      "num_tokens": 15235481.0,
      "step": 7520
    },
    {
      "entropy": 6.060359382629395,
      "epoch": 0.657607270820589,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963363097844323,
      "loss": 5.7971,
      "mean_token_accuracy": 0.16509583592414856,
      "num_tokens": 15245332.0,
      "step": 7525
    },
    {
      "entropy": 6.057863807678222,
      "epoch": 0.6580442191732937,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963307072402584,
      "loss": 5.78,
      "mean_token_accuracy": 0.16286325305700303,
      "num_tokens": 15255562.0,
      "step": 7530
    },
    {
      "entropy": 6.00372633934021,
      "epoch": 0.6584811675259984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963251004508456,
      "loss": 5.7082,
      "mean_token_accuracy": 0.16614110320806502,
      "num_tokens": 15265655.0,
      "step": 7535
    },
    {
      "entropy": 5.932410669326782,
      "epoch": 0.6589181158787032,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963194894163015,
      "loss": 5.6616,
      "mean_token_accuracy": 0.17542049437761306,
      "num_tokens": 15276311.0,
      "step": 7540
    },
    {
      "entropy": 5.988630867004394,
      "epoch": 0.6593550642314079,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963138741367334,
      "loss": 5.8016,
      "mean_token_accuracy": 0.1621630907058716,
      "num_tokens": 15287142.0,
      "step": 7545
    },
    {
      "entropy": 6.101692533493042,
      "epoch": 0.6597920125841126,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963082546122495,
      "loss": 5.7897,
      "mean_token_accuracy": 0.1637630730867386,
      "num_tokens": 15297484.0,
      "step": 7550
    },
    {
      "entropy": 5.973578596115113,
      "epoch": 0.6602289609368173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963026308429571,
      "loss": 5.7333,
      "mean_token_accuracy": 0.16768308430910112,
      "num_tokens": 15307468.0,
      "step": 7555
    },
    {
      "entropy": 5.897718811035157,
      "epoch": 0.660665909289522,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004962970028289644,
      "loss": 5.6933,
      "mean_token_accuracy": 0.17241705060005189,
      "num_tokens": 15318487.0,
      "step": 7560
    },
    {
      "entropy": 6.016544485092163,
      "epoch": 0.6611028576422266,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004962913705703791,
      "loss": 5.7496,
      "mean_token_accuracy": 0.16684581339359283,
      "num_tokens": 15329193.0,
      "step": 7565
    },
    {
      "entropy": 5.98777585029602,
      "epoch": 0.6615398059949315,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962857340673095,
      "loss": 5.6905,
      "mean_token_accuracy": 0.16755833774805068,
      "num_tokens": 15338886.0,
      "step": 7570
    },
    {
      "entropy": 5.88642110824585,
      "epoch": 0.6619767543476361,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004962800933198635,
      "loss": 5.6249,
      "mean_token_accuracy": 0.17535429298877717,
      "num_tokens": 15348629.0,
      "step": 7575
    },
    {
      "entropy": 5.9668224334716795,
      "epoch": 0.6624137027003408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962744483281495,
      "loss": 5.736,
      "mean_token_accuracy": 0.17385070472955705,
      "num_tokens": 15358004.0,
      "step": 7580
    },
    {
      "entropy": 6.008972024917602,
      "epoch": 0.6628506510530455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962687990922754,
      "loss": 5.7295,
      "mean_token_accuracy": 0.1628757745027542,
      "num_tokens": 15368018.0,
      "step": 7585
    },
    {
      "entropy": 6.086166954040527,
      "epoch": 0.6632875994057502,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049626314561235,
      "loss": 5.7127,
      "mean_token_accuracy": 0.16821101009845735,
      "num_tokens": 15377443.0,
      "step": 7590
    },
    {
      "entropy": 5.998335647583008,
      "epoch": 0.6637245477584549,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004962574878884814,
      "loss": 5.7349,
      "mean_token_accuracy": 0.16841058880090715,
      "num_tokens": 15388278.0,
      "step": 7595
    },
    {
      "entropy": 5.997668647766114,
      "epoch": 0.6641614961111597,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962518259207782,
      "loss": 5.7373,
      "mean_token_accuracy": 0.16834888458251954,
      "num_tokens": 15398619.0,
      "step": 7600
    },
    {
      "entropy": 5.870355129241943,
      "epoch": 0.6645984444638644,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962461597093491,
      "loss": 5.6451,
      "mean_token_accuracy": 0.17491673529148102,
      "num_tokens": 15408635.0,
      "step": 7605
    },
    {
      "entropy": 6.003155136108399,
      "epoch": 0.6650353928165691,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962404892543028,
      "loss": 5.7522,
      "mean_token_accuracy": 0.16205220967531203,
      "num_tokens": 15419113.0,
      "step": 7610
    },
    {
      "entropy": 6.070563411712646,
      "epoch": 0.6654723411692738,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496234814555748,
      "loss": 5.7264,
      "mean_token_accuracy": 0.16315551102161407,
      "num_tokens": 15428959.0,
      "step": 7615
    },
    {
      "entropy": 5.931528234481812,
      "epoch": 0.6659092895219785,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962291356137934,
      "loss": 5.8039,
      "mean_token_accuracy": 0.15894747525453568,
      "num_tokens": 15439116.0,
      "step": 7620
    },
    {
      "entropy": 6.018376970291138,
      "epoch": 0.6663462378746832,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962234524285481,
      "loss": 5.7736,
      "mean_token_accuracy": 0.16772264540195464,
      "num_tokens": 15449715.0,
      "step": 7625
    },
    {
      "entropy": 5.9224193572998045,
      "epoch": 0.666783186227388,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962177650001212,
      "loss": 5.5525,
      "mean_token_accuracy": 0.17777632623910905,
      "num_tokens": 15459170.0,
      "step": 7630
    },
    {
      "entropy": 5.878329563140869,
      "epoch": 0.6672201345800927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962120733286215,
      "loss": 5.6728,
      "mean_token_accuracy": 0.1722118690609932,
      "num_tokens": 15469497.0,
      "step": 7635
    },
    {
      "entropy": 5.974024534225464,
      "epoch": 0.6676570829327974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962063774141583,
      "loss": 5.8176,
      "mean_token_accuracy": 0.16801078617572784,
      "num_tokens": 15481298.0,
      "step": 7640
    },
    {
      "entropy": 6.036898374557495,
      "epoch": 0.668094031285502,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004962006772568409,
      "loss": 5.7534,
      "mean_token_accuracy": 0.1671036437153816,
      "num_tokens": 15491696.0,
      "step": 7645
    },
    {
      "entropy": 6.015488004684448,
      "epoch": 0.6685309796382067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961949728567787,
      "loss": 5.7652,
      "mean_token_accuracy": 0.16258450150489806,
      "num_tokens": 15502412.0,
      "step": 7650
    },
    {
      "entropy": 5.934196853637696,
      "epoch": 0.6689679279909114,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961892642140809,
      "loss": 5.7292,
      "mean_token_accuracy": 0.16624183654785157,
      "num_tokens": 15513436.0,
      "step": 7655
    },
    {
      "entropy": 5.974036359786988,
      "epoch": 0.6694048763436162,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961835513288572,
      "loss": 5.7334,
      "mean_token_accuracy": 0.17304501831531524,
      "num_tokens": 15524564.0,
      "step": 7660
    },
    {
      "entropy": 5.990243864059448,
      "epoch": 0.6698418246963209,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496177834201217,
      "loss": 5.7011,
      "mean_token_accuracy": 0.16360101848840714,
      "num_tokens": 15534539.0,
      "step": 7665
    },
    {
      "entropy": 6.0429980754852295,
      "epoch": 0.6702787730490256,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961721128312702,
      "loss": 5.7782,
      "mean_token_accuracy": 0.15800073444843293,
      "num_tokens": 15544571.0,
      "step": 7670
    },
    {
      "entropy": 5.997656679153442,
      "epoch": 0.6707157214017303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961663872191263,
      "loss": 5.7051,
      "mean_token_accuracy": 0.17574951946735382,
      "num_tokens": 15555066.0,
      "step": 7675
    },
    {
      "entropy": 5.9694513320922855,
      "epoch": 0.671152669754435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961606573648951,
      "loss": 5.7348,
      "mean_token_accuracy": 0.17001061290502548,
      "num_tokens": 15565128.0,
      "step": 7680
    },
    {
      "entropy": 5.975708913803101,
      "epoch": 0.6715896181071397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961549232686867,
      "loss": 5.7948,
      "mean_token_accuracy": 0.16248804479837417,
      "num_tokens": 15576394.0,
      "step": 7685
    },
    {
      "entropy": 6.054337501525879,
      "epoch": 0.6720265664598445,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496149184930611,
      "loss": 5.7396,
      "mean_token_accuracy": 0.16242391616106033,
      "num_tokens": 15585806.0,
      "step": 7690
    },
    {
      "entropy": 5.9783642292022705,
      "epoch": 0.6724635148125492,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961434423507779,
      "loss": 5.7271,
      "mean_token_accuracy": 0.16910791099071504,
      "num_tokens": 15596471.0,
      "step": 7695
    },
    {
      "entropy": 5.941711473464966,
      "epoch": 0.6729004631652539,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961376955292979,
      "loss": 5.6415,
      "mean_token_accuracy": 0.17608566731214523,
      "num_tokens": 15605720.0,
      "step": 7700
    },
    {
      "entropy": 5.969656610488892,
      "epoch": 0.6733374115179586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961319444662809,
      "loss": 5.709,
      "mean_token_accuracy": 0.1725141406059265,
      "num_tokens": 15616167.0,
      "step": 7705
    },
    {
      "entropy": 6.0012486457824705,
      "epoch": 0.6737743598706633,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961261891618374,
      "loss": 5.7995,
      "mean_token_accuracy": 0.15934359803795814,
      "num_tokens": 15627033.0,
      "step": 7710
    },
    {
      "entropy": 6.01514368057251,
      "epoch": 0.6742113082233679,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004961204296160777,
      "loss": 5.7528,
      "mean_token_accuracy": 0.17062305063009262,
      "num_tokens": 15637460.0,
      "step": 7715
    },
    {
      "entropy": 6.024073743820191,
      "epoch": 0.6746482565760727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961146658291123,
      "loss": 5.6738,
      "mean_token_accuracy": 0.17450377494096755,
      "num_tokens": 15646027.0,
      "step": 7720
    },
    {
      "entropy": 5.980246734619141,
      "epoch": 0.6750852049287774,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961088978010518,
      "loss": 5.7037,
      "mean_token_accuracy": 0.17008821964263915,
      "num_tokens": 15655241.0,
      "step": 7725
    },
    {
      "entropy": 5.9257152557373045,
      "epoch": 0.6755221532814821,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961031255320067,
      "loss": 5.6284,
      "mean_token_accuracy": 0.16974418014287948,
      "num_tokens": 15666213.0,
      "step": 7730
    },
    {
      "entropy": 6.035758638381958,
      "epoch": 0.6759591016341868,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004960973490220878,
      "loss": 5.7465,
      "mean_token_accuracy": 0.16223134547472,
      "num_tokens": 15676985.0,
      "step": 7735
    },
    {
      "entropy": 5.999268674850464,
      "epoch": 0.6763960499868915,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960915682714059,
      "loss": 5.772,
      "mean_token_accuracy": 0.16179870516061784,
      "num_tokens": 15687519.0,
      "step": 7740
    },
    {
      "entropy": 6.057408428192138,
      "epoch": 0.6768329983395963,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960857832800719,
      "loss": 5.7455,
      "mean_token_accuracy": 0.1668146952986717,
      "num_tokens": 15697401.0,
      "step": 7745
    },
    {
      "entropy": 5.9785669326782225,
      "epoch": 0.677269946692301,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960799940481968,
      "loss": 5.767,
      "mean_token_accuracy": 0.1599273204803467,
      "num_tokens": 15708159.0,
      "step": 7750
    },
    {
      "entropy": 5.98688759803772,
      "epoch": 0.6777068950450057,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960742005758914,
      "loss": 5.7097,
      "mean_token_accuracy": 0.165096415579319,
      "num_tokens": 15717749.0,
      "step": 7755
    },
    {
      "entropy": 5.956978845596313,
      "epoch": 0.6781438433977104,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960684028632671,
      "loss": 5.6711,
      "mean_token_accuracy": 0.16679559797048568,
      "num_tokens": 15726868.0,
      "step": 7760
    },
    {
      "entropy": 6.029999113082885,
      "epoch": 0.6785807917504151,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960626009104349,
      "loss": 5.7016,
      "mean_token_accuracy": 0.17021865546703338,
      "num_tokens": 15736038.0,
      "step": 7765
    },
    {
      "entropy": 5.991040229797363,
      "epoch": 0.6790177401031198,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960567947175064,
      "loss": 5.7155,
      "mean_token_accuracy": 0.16486626714468003,
      "num_tokens": 15746676.0,
      "step": 7770
    },
    {
      "entropy": 6.00670223236084,
      "epoch": 0.6794546884558246,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960509842845926,
      "loss": 5.6918,
      "mean_token_accuracy": 0.1672493815422058,
      "num_tokens": 15756064.0,
      "step": 7775
    },
    {
      "entropy": 5.97620267868042,
      "epoch": 0.6798916368085293,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960451696118049,
      "loss": 5.7091,
      "mean_token_accuracy": 0.17270632535219194,
      "num_tokens": 15767414.0,
      "step": 7780
    },
    {
      "entropy": 5.99161434173584,
      "epoch": 0.680328585161234,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960393506992553,
      "loss": 5.683,
      "mean_token_accuracy": 0.16803106814622878,
      "num_tokens": 15778276.0,
      "step": 7785
    },
    {
      "entropy": 5.923576498031617,
      "epoch": 0.6807655335139386,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960335275470551,
      "loss": 5.628,
      "mean_token_accuracy": 0.17321360111236572,
      "num_tokens": 15788365.0,
      "step": 7790
    },
    {
      "entropy": 5.978618907928467,
      "epoch": 0.6812024818666433,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960277001553159,
      "loss": 5.6294,
      "mean_token_accuracy": 0.17433976382017136,
      "num_tokens": 15799021.0,
      "step": 7795
    },
    {
      "entropy": 5.886345243453979,
      "epoch": 0.681639430219348,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004960218685241498,
      "loss": 5.709,
      "mean_token_accuracy": 0.17278380393981935,
      "num_tokens": 15808779.0,
      "step": 7800
    },
    {
      "entropy": 6.008462858200073,
      "epoch": 0.6820763785720528,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960160326536681,
      "loss": 5.7218,
      "mean_token_accuracy": 0.17279966324567794,
      "num_tokens": 15818515.0,
      "step": 7805
    },
    {
      "entropy": 6.030579710006714,
      "epoch": 0.6825133269247575,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960101925439834,
      "loss": 5.6767,
      "mean_token_accuracy": 0.16943463832139968,
      "num_tokens": 15828843.0,
      "step": 7810
    },
    {
      "entropy": 6.015769052505493,
      "epoch": 0.6829502752774622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960043481952072,
      "loss": 5.6742,
      "mean_token_accuracy": 0.16773785203695296,
      "num_tokens": 15839323.0,
      "step": 7815
    },
    {
      "entropy": 5.924883031845093,
      "epoch": 0.6833872236301669,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000495998499607452,
      "loss": 5.7251,
      "mean_token_accuracy": 0.16792866587638855,
      "num_tokens": 15848399.0,
      "step": 7820
    },
    {
      "entropy": 5.960476493835449,
      "epoch": 0.6838241719828716,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959926467808296,
      "loss": 5.5993,
      "mean_token_accuracy": 0.1735384628176689,
      "num_tokens": 15858137.0,
      "step": 7825
    },
    {
      "entropy": 5.92253303527832,
      "epoch": 0.6842611203355763,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004959867897154526,
      "loss": 5.7066,
      "mean_token_accuracy": 0.1705910488963127,
      "num_tokens": 15867643.0,
      "step": 7830
    },
    {
      "entropy": 5.959700679779052,
      "epoch": 0.6846980686882811,
      "grad_norm": 0.96875,
      "learning_rate": 0.000495980928411433,
      "loss": 5.6017,
      "mean_token_accuracy": 0.16989771872758866,
      "num_tokens": 15877523.0,
      "step": 7835
    },
    {
      "entropy": 6.000572586059571,
      "epoch": 0.6851350170409858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959750628688834,
      "loss": 5.7875,
      "mean_token_accuracy": 0.16270564496517181,
      "num_tokens": 15886946.0,
      "step": 7840
    },
    {
      "entropy": 5.930234432220459,
      "epoch": 0.6855719653936905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959691930879164,
      "loss": 5.6115,
      "mean_token_accuracy": 0.17378821074962617,
      "num_tokens": 15896646.0,
      "step": 7845
    },
    {
      "entropy": 5.9268018245697025,
      "epoch": 0.6860089137463952,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959633190686445,
      "loss": 5.6535,
      "mean_token_accuracy": 0.17022318094968797,
      "num_tokens": 15907204.0,
      "step": 7850
    },
    {
      "entropy": 5.956172466278076,
      "epoch": 0.6864458620990999,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959574408111804,
      "loss": 5.7038,
      "mean_token_accuracy": 0.17531090676784516,
      "num_tokens": 15917634.0,
      "step": 7855
    },
    {
      "entropy": 5.99599061012268,
      "epoch": 0.6868828104518045,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959515583156367,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1748895525932312,
      "num_tokens": 15926584.0,
      "step": 7860
    },
    {
      "entropy": 5.95475025177002,
      "epoch": 0.6873197588045094,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959456715821264,
      "loss": 5.6379,
      "mean_token_accuracy": 0.17134907841682434,
      "num_tokens": 15936636.0,
      "step": 7865
    },
    {
      "entropy": 5.946742153167724,
      "epoch": 0.687756707157214,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959397806107624,
      "loss": 5.6634,
      "mean_token_accuracy": 0.17933844178915023,
      "num_tokens": 15946113.0,
      "step": 7870
    },
    {
      "entropy": 6.003996086120606,
      "epoch": 0.6881936555099187,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959338854016574,
      "loss": 5.7019,
      "mean_token_accuracy": 0.16820037961006165,
      "num_tokens": 15956630.0,
      "step": 7875
    },
    {
      "entropy": 5.91122989654541,
      "epoch": 0.6886306038626234,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959279859549249,
      "loss": 5.6431,
      "mean_token_accuracy": 0.17330818325281144,
      "num_tokens": 15967129.0,
      "step": 7880
    },
    {
      "entropy": 5.982016372680664,
      "epoch": 0.6890675522153281,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004959220822706779,
      "loss": 5.6395,
      "mean_token_accuracy": 0.1716708391904831,
      "num_tokens": 15976663.0,
      "step": 7885
    },
    {
      "entropy": 5.973646020889282,
      "epoch": 0.6895045005680328,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959161743490295,
      "loss": 5.6373,
      "mean_token_accuracy": 0.17136787325143815,
      "num_tokens": 15986666.0,
      "step": 7890
    },
    {
      "entropy": 5.853838682174683,
      "epoch": 0.6899414489207376,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004959102621900931,
      "loss": 5.5956,
      "mean_token_accuracy": 0.18484989404678345,
      "num_tokens": 15996233.0,
      "step": 7895
    },
    {
      "entropy": 5.961171579360962,
      "epoch": 0.6903783972734423,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959043457939822,
      "loss": 5.7707,
      "mean_token_accuracy": 0.1650151267647743,
      "num_tokens": 16008188.0,
      "step": 7900
    },
    {
      "entropy": 6.000286340713501,
      "epoch": 0.690815345626147,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958984251608101,
      "loss": 5.679,
      "mean_token_accuracy": 0.1679936818778515,
      "num_tokens": 16018494.0,
      "step": 7905
    },
    {
      "entropy": 5.960310983657837,
      "epoch": 0.6912522939788517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958925002906905,
      "loss": 5.6985,
      "mean_token_accuracy": 0.17179163843393325,
      "num_tokens": 16028531.0,
      "step": 7910
    },
    {
      "entropy": 5.9893628597259525,
      "epoch": 0.6916892423315564,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495886571183737,
      "loss": 5.6739,
      "mean_token_accuracy": 0.167478708922863,
      "num_tokens": 16039050.0,
      "step": 7915
    },
    {
      "entropy": 5.957552289962768,
      "epoch": 0.6921261906842611,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958806378400634,
      "loss": 5.7593,
      "mean_token_accuracy": 0.16104569137096406,
      "num_tokens": 16050751.0,
      "step": 7920
    },
    {
      "entropy": 6.008441829681397,
      "epoch": 0.6925631390369659,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958747002597834,
      "loss": 5.6994,
      "mean_token_accuracy": 0.1658043384552002,
      "num_tokens": 16060758.0,
      "step": 7925
    },
    {
      "entropy": 5.992664480209351,
      "epoch": 0.6930000873896706,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004958687584430109,
      "loss": 5.7131,
      "mean_token_accuracy": 0.16733440160751342,
      "num_tokens": 16070852.0,
      "step": 7930
    },
    {
      "entropy": 5.920446825027466,
      "epoch": 0.6934370357423753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958628123898597,
      "loss": 5.6305,
      "mean_token_accuracy": 0.17284321337938308,
      "num_tokens": 16080081.0,
      "step": 7935
    },
    {
      "entropy": 5.956729316711426,
      "epoch": 0.6938739840950799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958568621004442,
      "loss": 5.649,
      "mean_token_accuracy": 0.16790755093097687,
      "num_tokens": 16090483.0,
      "step": 7940
    },
    {
      "entropy": 6.043986511230469,
      "epoch": 0.6943109324477846,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958509075748785,
      "loss": 5.7572,
      "mean_token_accuracy": 0.1634690061211586,
      "num_tokens": 16101136.0,
      "step": 7945
    },
    {
      "entropy": 5.961344814300537,
      "epoch": 0.6947478808004894,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958449488132764,
      "loss": 5.6448,
      "mean_token_accuracy": 0.17701392471790314,
      "num_tokens": 16110124.0,
      "step": 7950
    },
    {
      "entropy": 5.891111087799072,
      "epoch": 0.6951848291531941,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958389858157526,
      "loss": 5.6107,
      "mean_token_accuracy": 0.18005365133285522,
      "num_tokens": 16119581.0,
      "step": 7955
    },
    {
      "entropy": 5.89972562789917,
      "epoch": 0.6956217775058988,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958330185824213,
      "loss": 5.7074,
      "mean_token_accuracy": 0.16786114275455474,
      "num_tokens": 16130119.0,
      "step": 7960
    },
    {
      "entropy": 5.978169107437134,
      "epoch": 0.6960587258586035,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495827047113397,
      "loss": 5.7429,
      "mean_token_accuracy": 0.16373222172260285,
      "num_tokens": 16140702.0,
      "step": 7965
    },
    {
      "entropy": 5.972009992599487,
      "epoch": 0.6964956742113082,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958210714087942,
      "loss": 5.6972,
      "mean_token_accuracy": 0.17101324647665023,
      "num_tokens": 16151107.0,
      "step": 7970
    },
    {
      "entropy": 5.968778896331787,
      "epoch": 0.6969326225640129,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958150914687276,
      "loss": 5.6358,
      "mean_token_accuracy": 0.1746847301721573,
      "num_tokens": 16160779.0,
      "step": 7975
    },
    {
      "entropy": 5.9044920921325685,
      "epoch": 0.6973695709167177,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958091072933118,
      "loss": 5.5777,
      "mean_token_accuracy": 0.1756403401494026,
      "num_tokens": 16169624.0,
      "step": 7980
    },
    {
      "entropy": 5.869996643066406,
      "epoch": 0.6978065192694224,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958031188826616,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17786190360784532,
      "num_tokens": 16179192.0,
      "step": 7985
    },
    {
      "entropy": 5.920633935928345,
      "epoch": 0.6982434676221271,
      "grad_norm": 1.125,
      "learning_rate": 0.0004957971262368919,
      "loss": 5.6593,
      "mean_token_accuracy": 0.17606053799390792,
      "num_tokens": 16188847.0,
      "step": 7990
    },
    {
      "entropy": 6.009812498092652,
      "epoch": 0.6986804159748318,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957911293561176,
      "loss": 5.7673,
      "mean_token_accuracy": 0.1681905835866928,
      "num_tokens": 16199480.0,
      "step": 7995
    },
    {
      "entropy": 5.963663721084595,
      "epoch": 0.6991173643275365,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957851282404538,
      "loss": 5.6529,
      "mean_token_accuracy": 0.18029617816209792,
      "num_tokens": 16208818.0,
      "step": 8000
    },
    {
      "entropy": 6.006942653656006,
      "epoch": 0.6995543126802412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957791228900154,
      "loss": 5.6343,
      "mean_token_accuracy": 0.16877323091030122,
      "num_tokens": 16219298.0,
      "step": 8005
    },
    {
      "entropy": 5.924262666702271,
      "epoch": 0.699991261032946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957731133049179,
      "loss": 5.6371,
      "mean_token_accuracy": 0.17696456909179686,
      "num_tokens": 16228170.0,
      "step": 8010
    },
    {
      "entropy": 5.977137517929077,
      "epoch": 0.7004282093856506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004957670994852764,
      "loss": 5.7153,
      "mean_token_accuracy": 0.16793975234031677,
      "num_tokens": 16237552.0,
      "step": 8015
    },
    {
      "entropy": 6.024384212493897,
      "epoch": 0.7008651577383553,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004957610814312062,
      "loss": 5.7533,
      "mean_token_accuracy": 0.16528825014829635,
      "num_tokens": 16248157.0,
      "step": 8020
    },
    {
      "entropy": 5.951394033432007,
      "epoch": 0.70130210609106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004957550591428229,
      "loss": 5.6289,
      "mean_token_accuracy": 0.17249557822942735,
      "num_tokens": 16258266.0,
      "step": 8025
    },
    {
      "entropy": 5.920563554763794,
      "epoch": 0.7017390544437647,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004957490326202418,
      "loss": 5.7326,
      "mean_token_accuracy": 0.16811388731002808,
      "num_tokens": 16268769.0,
      "step": 8030
    },
    {
      "entropy": 6.0277608871459964,
      "epoch": 0.7021760027964694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957430018635785,
      "loss": 5.6908,
      "mean_token_accuracy": 0.16715661585330963,
      "num_tokens": 16279688.0,
      "step": 8035
    },
    {
      "entropy": 5.9778642654418945,
      "epoch": 0.7026129511491742,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957369668729488,
      "loss": 5.6666,
      "mean_token_accuracy": 0.17562907338142394,
      "num_tokens": 16289441.0,
      "step": 8040
    },
    {
      "entropy": 5.937805461883545,
      "epoch": 0.7030498995018789,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957309276484685,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1692324161529541,
      "num_tokens": 16299802.0,
      "step": 8045
    },
    {
      "entropy": 5.9242242813110355,
      "epoch": 0.7034868478545836,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957248841902533,
      "loss": 5.634,
      "mean_token_accuracy": 0.17715984135866164,
      "num_tokens": 16309577.0,
      "step": 8050
    },
    {
      "entropy": 5.9485162734985355,
      "epoch": 0.7039237962072883,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957188364984191,
      "loss": 5.5899,
      "mean_token_accuracy": 0.17622884809970857,
      "num_tokens": 16319205.0,
      "step": 8055
    },
    {
      "entropy": 5.85321044921875,
      "epoch": 0.704360744559993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495712784573082,
      "loss": 5.6519,
      "mean_token_accuracy": 0.17543867975473404,
      "num_tokens": 16329650.0,
      "step": 8060
    },
    {
      "entropy": 5.8825126647949215,
      "epoch": 0.7047976929126977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957067284143581,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1714753121137619,
      "num_tokens": 16339242.0,
      "step": 8065
    },
    {
      "entropy": 5.998453903198242,
      "epoch": 0.7052346412654025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957006680223634,
      "loss": 5.6452,
      "mean_token_accuracy": 0.17831729650497435,
      "num_tokens": 16348441.0,
      "step": 8070
    },
    {
      "entropy": 5.929697179794312,
      "epoch": 0.7056715896181072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956946033972143,
      "loss": 5.6603,
      "mean_token_accuracy": 0.1724653422832489,
      "num_tokens": 16357425.0,
      "step": 8075
    },
    {
      "entropy": 5.928646755218506,
      "epoch": 0.7061085379708119,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495688534539027,
      "loss": 5.6166,
      "mean_token_accuracy": 0.1716718390583992,
      "num_tokens": 16366180.0,
      "step": 8080
    },
    {
      "entropy": 5.9152756690979,
      "epoch": 0.7065454863235165,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956824614479181,
      "loss": 5.6725,
      "mean_token_accuracy": 0.16842027753591537,
      "num_tokens": 16375714.0,
      "step": 8085
    },
    {
      "entropy": 6.000022888183594,
      "epoch": 0.7069824346762212,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956763841240038,
      "loss": 5.6664,
      "mean_token_accuracy": 0.1660715103149414,
      "num_tokens": 16386051.0,
      "step": 8090
    },
    {
      "entropy": 5.987563800811768,
      "epoch": 0.7074193830289259,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495670302567401,
      "loss": 5.7255,
      "mean_token_accuracy": 0.16674379557371138,
      "num_tokens": 16396616.0,
      "step": 8095
    },
    {
      "entropy": 5.927264785766601,
      "epoch": 0.7078563313816307,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495664216778226,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17981154173612596,
      "num_tokens": 16406301.0,
      "step": 8100
    },
    {
      "entropy": 5.959307289123535,
      "epoch": 0.7082932797343354,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956581267565957,
      "loss": 5.7167,
      "mean_token_accuracy": 0.16652360558509827,
      "num_tokens": 16417107.0,
      "step": 8105
    },
    {
      "entropy": 5.968394994735718,
      "epoch": 0.7087302280870401,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495652032502627,
      "loss": 5.6346,
      "mean_token_accuracy": 0.17637471854686737,
      "num_tokens": 16427292.0,
      "step": 8110
    },
    {
      "entropy": 5.900726556777954,
      "epoch": 0.7091671764397448,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956459340164366,
      "loss": 5.6414,
      "mean_token_accuracy": 0.17783666253089905,
      "num_tokens": 16438191.0,
      "step": 8115
    },
    {
      "entropy": 5.947930383682251,
      "epoch": 0.7096041247924495,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004956398312981417,
      "loss": 5.6109,
      "mean_token_accuracy": 0.17297908365726472,
      "num_tokens": 16448237.0,
      "step": 8120
    },
    {
      "entropy": 5.932255506515503,
      "epoch": 0.7100410731451542,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956337243478592,
      "loss": 5.7026,
      "mean_token_accuracy": 0.17288212329149247,
      "num_tokens": 16458947.0,
      "step": 8125
    },
    {
      "entropy": 5.999772691726685,
      "epoch": 0.710478021497859,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956276131657061,
      "loss": 5.6378,
      "mean_token_accuracy": 0.17141077667474747,
      "num_tokens": 16467464.0,
      "step": 8130
    },
    {
      "entropy": 5.950952863693237,
      "epoch": 0.7109149698505637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956214977518,
      "loss": 5.6543,
      "mean_token_accuracy": 0.17681879550218582,
      "num_tokens": 16477747.0,
      "step": 8135
    },
    {
      "entropy": 5.968418741226197,
      "epoch": 0.7113519182032684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956153781062578,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1672035723924637,
      "num_tokens": 16488111.0,
      "step": 8140
    },
    {
      "entropy": 5.947294139862061,
      "epoch": 0.7117888665559731,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004956092542291971,
      "loss": 5.6582,
      "mean_token_accuracy": 0.17239930480718613,
      "num_tokens": 16497689.0,
      "step": 8145
    },
    {
      "entropy": 5.975532484054566,
      "epoch": 0.7122258149086778,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004956031261207353,
      "loss": 5.7704,
      "mean_token_accuracy": 0.16860691905021669,
      "num_tokens": 16508515.0,
      "step": 8150
    },
    {
      "entropy": 5.846520519256591,
      "epoch": 0.7126627632613826,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049559699378099,
      "loss": 5.6427,
      "mean_token_accuracy": 0.17767313867807388,
      "num_tokens": 16519184.0,
      "step": 8155
    },
    {
      "entropy": 5.990766620635986,
      "epoch": 0.7130997116140873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955908572100787,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16518724113702773,
      "num_tokens": 16529917.0,
      "step": 8160
    },
    {
      "entropy": 5.984876680374145,
      "epoch": 0.7135366599667919,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955847164081192,
      "loss": 5.6307,
      "mean_token_accuracy": 0.1702073931694031,
      "num_tokens": 16539687.0,
      "step": 8165
    },
    {
      "entropy": 5.92875108718872,
      "epoch": 0.7139736083194966,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955785713752292,
      "loss": 5.7213,
      "mean_token_accuracy": 0.16826012283563613,
      "num_tokens": 16550273.0,
      "step": 8170
    },
    {
      "entropy": 6.037826919555664,
      "epoch": 0.7144105566722013,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004955724221115267,
      "loss": 5.7828,
      "mean_token_accuracy": 0.16326912343502045,
      "num_tokens": 16562215.0,
      "step": 8175
    },
    {
      "entropy": 6.032857942581177,
      "epoch": 0.714847505024906,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955662686171295,
      "loss": 5.7351,
      "mean_token_accuracy": 0.17116176187992097,
      "num_tokens": 16573218.0,
      "step": 8180
    },
    {
      "entropy": 6.0085594177246096,
      "epoch": 0.7152844533776108,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004955601108921557,
      "loss": 5.7187,
      "mean_token_accuracy": 0.16171132028102875,
      "num_tokens": 16582740.0,
      "step": 8185
    },
    {
      "entropy": 5.9259802341461185,
      "epoch": 0.7157214017303155,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004955539489367235,
      "loss": 5.6732,
      "mean_token_accuracy": 0.17611398547887802,
      "num_tokens": 16592169.0,
      "step": 8190
    },
    {
      "entropy": 5.953530120849609,
      "epoch": 0.7161583500830202,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955477827509509,
      "loss": 5.644,
      "mean_token_accuracy": 0.16816934943199158,
      "num_tokens": 16602706.0,
      "step": 8195
    },
    {
      "entropy": 5.945762872695923,
      "epoch": 0.7165952984357249,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955416123349562,
      "loss": 5.633,
      "mean_token_accuracy": 0.1719654083251953,
      "num_tokens": 16612012.0,
      "step": 8200
    },
    {
      "entropy": 5.966068506240845,
      "epoch": 0.7170322467884296,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495535437688858,
      "loss": 5.7572,
      "mean_token_accuracy": 0.16732777804136276,
      "num_tokens": 16621798.0,
      "step": 8205
    },
    {
      "entropy": 5.985923862457275,
      "epoch": 0.7174691951411343,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004955292588127744,
      "loss": 5.6674,
      "mean_token_accuracy": 0.17234889268875123,
      "num_tokens": 16631792.0,
      "step": 8210
    },
    {
      "entropy": 5.97695779800415,
      "epoch": 0.7179061434938391,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495523075706824,
      "loss": 5.6269,
      "mean_token_accuracy": 0.17342378050088883,
      "num_tokens": 16641655.0,
      "step": 8215
    },
    {
      "entropy": 5.958208799362183,
      "epoch": 0.7183430918465438,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955168883711257,
      "loss": 5.7152,
      "mean_token_accuracy": 0.16085369288921356,
      "num_tokens": 16651210.0,
      "step": 8220
    },
    {
      "entropy": 6.018463182449341,
      "epoch": 0.7187800401992485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955106968057977,
      "loss": 5.6805,
      "mean_token_accuracy": 0.16819168627262115,
      "num_tokens": 16660078.0,
      "step": 8225
    },
    {
      "entropy": 5.99434461593628,
      "epoch": 0.7192169885519532,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955045010109592,
      "loss": 5.6969,
      "mean_token_accuracy": 0.16938378661870956,
      "num_tokens": 16671211.0,
      "step": 8230
    },
    {
      "entropy": 5.960981225967407,
      "epoch": 0.7196539369046578,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954983009867286,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17598955929279328,
      "num_tokens": 16680850.0,
      "step": 8235
    },
    {
      "entropy": 5.902716016769409,
      "epoch": 0.7200908852573625,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954920967332252,
      "loss": 5.6023,
      "mean_token_accuracy": 0.17169032692909242,
      "num_tokens": 16690447.0,
      "step": 8240
    },
    {
      "entropy": 5.909363698959351,
      "epoch": 0.7205278336100673,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954858882505678,
      "loss": 5.7054,
      "mean_token_accuracy": 0.1682112231850624,
      "num_tokens": 16701061.0,
      "step": 8245
    },
    {
      "entropy": 5.985194683074951,
      "epoch": 0.720964781962772,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954796755388755,
      "loss": 5.605,
      "mean_token_accuracy": 0.17755259275436402,
      "num_tokens": 16711139.0,
      "step": 8250
    },
    {
      "entropy": 5.9732200622558596,
      "epoch": 0.7214017303154767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954734585982675,
      "loss": 5.7462,
      "mean_token_accuracy": 0.15968185365200044,
      "num_tokens": 16722343.0,
      "step": 8255
    },
    {
      "entropy": 5.971442747116089,
      "epoch": 0.7218386786681814,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954672374288631,
      "loss": 5.7317,
      "mean_token_accuracy": 0.16381252855062484,
      "num_tokens": 16733475.0,
      "step": 8260
    },
    {
      "entropy": 5.991418170928955,
      "epoch": 0.7222756270208861,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004954610120307815,
      "loss": 5.7104,
      "mean_token_accuracy": 0.16640409231185913,
      "num_tokens": 16744356.0,
      "step": 8265
    },
    {
      "entropy": 5.920919418334961,
      "epoch": 0.7227125753735908,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954547824041422,
      "loss": 5.7155,
      "mean_token_accuracy": 0.16554230600595474,
      "num_tokens": 16755275.0,
      "step": 8270
    },
    {
      "entropy": 5.946273565292358,
      "epoch": 0.7231495237262956,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954485485490647,
      "loss": 5.6471,
      "mean_token_accuracy": 0.17469835728406907,
      "num_tokens": 16766114.0,
      "step": 8275
    },
    {
      "entropy": 5.9773406982421875,
      "epoch": 0.7235864720790003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004954423104656685,
      "loss": 5.6227,
      "mean_token_accuracy": 0.18017650097608567,
      "num_tokens": 16776391.0,
      "step": 8280
    },
    {
      "entropy": 5.859285306930542,
      "epoch": 0.724023420431705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954360681540732,
      "loss": 5.6057,
      "mean_token_accuracy": 0.17592469304800035,
      "num_tokens": 16784845.0,
      "step": 8285
    },
    {
      "entropy": 5.912224578857422,
      "epoch": 0.7244603687844097,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954298216143986,
      "loss": 5.6227,
      "mean_token_accuracy": 0.1781073823571205,
      "num_tokens": 16794735.0,
      "step": 8290
    },
    {
      "entropy": 6.026312255859375,
      "epoch": 0.7248973171371144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954235708467647,
      "loss": 5.7135,
      "mean_token_accuracy": 0.17188996076583862,
      "num_tokens": 16803643.0,
      "step": 8295
    },
    {
      "entropy": 5.9373516082763675,
      "epoch": 0.725334265489819,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495417315851291,
      "loss": 5.7033,
      "mean_token_accuracy": 0.1706904500722885,
      "num_tokens": 16813064.0,
      "step": 8300
    },
    {
      "entropy": 5.976805257797241,
      "epoch": 0.7257712138425239,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004954110566280978,
      "loss": 5.703,
      "mean_token_accuracy": 0.16932612359523774,
      "num_tokens": 16823166.0,
      "step": 8305
    },
    {
      "entropy": 5.993662643432617,
      "epoch": 0.7262081621952285,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495404793177305,
      "loss": 5.7221,
      "mean_token_accuracy": 0.1667390540242195,
      "num_tokens": 16833546.0,
      "step": 8310
    },
    {
      "entropy": 5.92490701675415,
      "epoch": 0.7266451105479332,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953985254990328,
      "loss": 5.5921,
      "mean_token_accuracy": 0.1778773248195648,
      "num_tokens": 16843794.0,
      "step": 8315
    },
    {
      "entropy": 5.972211599349976,
      "epoch": 0.7270820589006379,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953922535934014,
      "loss": 5.7333,
      "mean_token_accuracy": 0.16743092685937883,
      "num_tokens": 16855677.0,
      "step": 8320
    },
    {
      "entropy": 6.0065654754638675,
      "epoch": 0.7275190072533426,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495385977460531,
      "loss": 5.6813,
      "mean_token_accuracy": 0.1640188977122307,
      "num_tokens": 16866137.0,
      "step": 8325
    },
    {
      "entropy": 5.968684720993042,
      "epoch": 0.7279559556060473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953796971005422,
      "loss": 5.6032,
      "mean_token_accuracy": 0.17238486111164092,
      "num_tokens": 16875997.0,
      "step": 8330
    },
    {
      "entropy": 5.8611455917358395,
      "epoch": 0.7283929039587521,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004953734125135552,
      "loss": 5.6559,
      "mean_token_accuracy": 0.1759241536259651,
      "num_tokens": 16885838.0,
      "step": 8335
    },
    {
      "entropy": 5.900574016571045,
      "epoch": 0.7288298523114568,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953671236996909,
      "loss": 5.5856,
      "mean_token_accuracy": 0.1783141925930977,
      "num_tokens": 16895533.0,
      "step": 8340
    },
    {
      "entropy": 5.998710203170776,
      "epoch": 0.7292668006641615,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953608306590696,
      "loss": 5.71,
      "mean_token_accuracy": 0.16809077188372612,
      "num_tokens": 16906258.0,
      "step": 8345
    },
    {
      "entropy": 5.970078706741333,
      "epoch": 0.7297037490168662,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495354533391812,
      "loss": 5.6672,
      "mean_token_accuracy": 0.17465408593416215,
      "num_tokens": 16917123.0,
      "step": 8350
    },
    {
      "entropy": 5.9451827049255375,
      "epoch": 0.7301406973695709,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953482318980389,
      "loss": 5.6826,
      "mean_token_accuracy": 0.17426684349775315,
      "num_tokens": 16927533.0,
      "step": 8355
    },
    {
      "entropy": 5.995517110824585,
      "epoch": 0.7305776457222757,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953419261778714,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1718117207288742,
      "num_tokens": 16937763.0,
      "step": 8360
    },
    {
      "entropy": 5.92606554031372,
      "epoch": 0.7310145940749804,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953356162314304,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17840242087841035,
      "num_tokens": 16947173.0,
      "step": 8365
    },
    {
      "entropy": 5.961096239089966,
      "epoch": 0.7314515424276851,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953293020588368,
      "loss": 5.6777,
      "mean_token_accuracy": 0.17512092739343643,
      "num_tokens": 16956856.0,
      "step": 8370
    },
    {
      "entropy": 5.93956446647644,
      "epoch": 0.7318884907803898,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953229836602115,
      "loss": 5.6675,
      "mean_token_accuracy": 0.17096923142671586,
      "num_tokens": 16966705.0,
      "step": 8375
    },
    {
      "entropy": 5.950593948364258,
      "epoch": 0.7323254391330944,
      "grad_norm": 0.90625,
      "learning_rate": 0.000495316661035676,
      "loss": 5.6781,
      "mean_token_accuracy": 0.17147308588027954,
      "num_tokens": 16977026.0,
      "step": 8380
    },
    {
      "entropy": 5.968172454833985,
      "epoch": 0.7327623874857991,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953103341853517,
      "loss": 5.5994,
      "mean_token_accuracy": 0.18063658475875854,
      "num_tokens": 16986800.0,
      "step": 8385
    },
    {
      "entropy": 5.949002027511597,
      "epoch": 0.7331993358385039,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004953040031093596,
      "loss": 5.5874,
      "mean_token_accuracy": 0.17797926664352418,
      "num_tokens": 16996392.0,
      "step": 8390
    },
    {
      "entropy": 5.929271268844604,
      "epoch": 0.7336362841912086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952976678078213,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1738160252571106,
      "num_tokens": 17006430.0,
      "step": 8395
    },
    {
      "entropy": 5.994189071655273,
      "epoch": 0.7340732325439133,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952913282808584,
      "loss": 5.6717,
      "mean_token_accuracy": 0.16316924691200257,
      "num_tokens": 17017168.0,
      "step": 8400
    },
    {
      "entropy": 5.8662636280059814,
      "epoch": 0.734510180896618,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952849845285922,
      "loss": 5.5882,
      "mean_token_accuracy": 0.18116444796323777,
      "num_tokens": 17027337.0,
      "step": 8405
    },
    {
      "entropy": 5.930755662918091,
      "epoch": 0.7349471292493227,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952786365511447,
      "loss": 5.571,
      "mean_token_accuracy": 0.1814669668674469,
      "num_tokens": 17036839.0,
      "step": 8410
    },
    {
      "entropy": 5.951767110824585,
      "epoch": 0.7353840776020274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952722843486375,
      "loss": 5.7703,
      "mean_token_accuracy": 0.1637505754828453,
      "num_tokens": 17048221.0,
      "step": 8415
    },
    {
      "entropy": 5.917473077774048,
      "epoch": 0.7358210259547322,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952659279211924,
      "loss": 5.5956,
      "mean_token_accuracy": 0.17329713702201843,
      "num_tokens": 17057694.0,
      "step": 8420
    },
    {
      "entropy": 5.959954023361206,
      "epoch": 0.7362579743074369,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952595672689315,
      "loss": 5.6878,
      "mean_token_accuracy": 0.1606788843870163,
      "num_tokens": 17066492.0,
      "step": 8425
    },
    {
      "entropy": 6.000223827362061,
      "epoch": 0.7366949226601416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952532023919766,
      "loss": 5.6912,
      "mean_token_accuracy": 0.16865280270576477,
      "num_tokens": 17077211.0,
      "step": 8430
    },
    {
      "entropy": 5.9528453826904295,
      "epoch": 0.7371318710128463,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004952468332904498,
      "loss": 5.6518,
      "mean_token_accuracy": 0.17314266115427018,
      "num_tokens": 17088665.0,
      "step": 8435
    },
    {
      "entropy": 5.932890319824219,
      "epoch": 0.737568819365551,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952404599644735,
      "loss": 5.6466,
      "mean_token_accuracy": 0.17729318290948867,
      "num_tokens": 17099767.0,
      "step": 8440
    },
    {
      "entropy": 5.937790489196777,
      "epoch": 0.7380057677182557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952340824141698,
      "loss": 5.5712,
      "mean_token_accuracy": 0.18339237719774246,
      "num_tokens": 17109414.0,
      "step": 8445
    },
    {
      "entropy": 5.92702693939209,
      "epoch": 0.7384427160709605,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952277006396609,
      "loss": 5.6618,
      "mean_token_accuracy": 0.16808170229196548,
      "num_tokens": 17119107.0,
      "step": 8450
    },
    {
      "entropy": 5.941473817825317,
      "epoch": 0.7388796644236652,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952213146410693,
      "loss": 5.6854,
      "mean_token_accuracy": 0.16574075371026992,
      "num_tokens": 17129010.0,
      "step": 8455
    },
    {
      "entropy": 5.994534063339233,
      "epoch": 0.7393166127763698,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952149244185174,
      "loss": 5.7002,
      "mean_token_accuracy": 0.17138614654541015,
      "num_tokens": 17138951.0,
      "step": 8460
    },
    {
      "entropy": 5.993372917175293,
      "epoch": 0.7397535611290745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952085299721279,
      "loss": 5.6463,
      "mean_token_accuracy": 0.1688268318772316,
      "num_tokens": 17149416.0,
      "step": 8465
    },
    {
      "entropy": 5.911770439147949,
      "epoch": 0.7401905094817792,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952021313020235,
      "loss": 5.5943,
      "mean_token_accuracy": 0.18211120516061782,
      "num_tokens": 17160068.0,
      "step": 8470
    },
    {
      "entropy": 5.932995700836182,
      "epoch": 0.7406274578344839,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951957284083269,
      "loss": 5.6511,
      "mean_token_accuracy": 0.17048127055168152,
      "num_tokens": 17168972.0,
      "step": 8475
    },
    {
      "entropy": 5.90786919593811,
      "epoch": 0.7410644061871887,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951893212911607,
      "loss": 5.6063,
      "mean_token_accuracy": 0.1807121753692627,
      "num_tokens": 17178239.0,
      "step": 8480
    },
    {
      "entropy": 5.95505919456482,
      "epoch": 0.7415013545398934,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000495182909950648,
      "loss": 5.7695,
      "mean_token_accuracy": 0.16541819423437118,
      "num_tokens": 17190619.0,
      "step": 8485
    },
    {
      "entropy": 5.862196969985962,
      "epoch": 0.7419383028925981,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951764943869117,
      "loss": 5.5579,
      "mean_token_accuracy": 0.17387183308601378,
      "num_tokens": 17201041.0,
      "step": 8490
    },
    {
      "entropy": 5.936975002288818,
      "epoch": 0.7423752512453028,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951700746000749,
      "loss": 5.6862,
      "mean_token_accuracy": 0.1688358172774315,
      "num_tokens": 17211779.0,
      "step": 8495
    },
    {
      "entropy": 5.986603116989135,
      "epoch": 0.7428121995980075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951636505902606,
      "loss": 5.636,
      "mean_token_accuracy": 0.17799067199230195,
      "num_tokens": 17221249.0,
      "step": 8500
    },
    {
      "entropy": 5.958147573471069,
      "epoch": 0.7432491479507122,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951572223575922,
      "loss": 5.6879,
      "mean_token_accuracy": 0.16987977623939515,
      "num_tokens": 17231634.0,
      "step": 8505
    },
    {
      "entropy": 5.947153759002686,
      "epoch": 0.743686096303417,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951507899021929,
      "loss": 5.7118,
      "mean_token_accuracy": 0.16771858781576157,
      "num_tokens": 17241751.0,
      "step": 8510
    },
    {
      "entropy": 5.992804765701294,
      "epoch": 0.7441230446561217,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495144353224186,
      "loss": 5.691,
      "mean_token_accuracy": 0.16678913086652755,
      "num_tokens": 17251962.0,
      "step": 8515
    },
    {
      "entropy": 5.9510657787323,
      "epoch": 0.7445599930088264,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951379123236952,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17395014315843582,
      "num_tokens": 17261895.0,
      "step": 8520
    },
    {
      "entropy": 5.908085632324219,
      "epoch": 0.744996941361531,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951314672008439,
      "loss": 5.6526,
      "mean_token_accuracy": 0.17893627285957336,
      "num_tokens": 17272025.0,
      "step": 8525
    },
    {
      "entropy": 5.942544174194336,
      "epoch": 0.7454338897142357,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951250178557556,
      "loss": 5.6567,
      "mean_token_accuracy": 0.17837535440921784,
      "num_tokens": 17282705.0,
      "step": 8530
    },
    {
      "entropy": 5.959393072128296,
      "epoch": 0.7458708380669404,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951185642885542,
      "loss": 5.5949,
      "mean_token_accuracy": 0.17821435034275054,
      "num_tokens": 17293045.0,
      "step": 8535
    },
    {
      "entropy": 5.966700315475464,
      "epoch": 0.7463077864196452,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951121064993633,
      "loss": 5.677,
      "mean_token_accuracy": 0.16884415447711945,
      "num_tokens": 17303893.0,
      "step": 8540
    },
    {
      "entropy": 5.951896381378174,
      "epoch": 0.7467447347723499,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951056444883069,
      "loss": 5.6536,
      "mean_token_accuracy": 0.16734211295843124,
      "num_tokens": 17314250.0,
      "step": 8545
    },
    {
      "entropy": 5.931076622009277,
      "epoch": 0.7471816831250546,
      "grad_norm": 1.0,
      "learning_rate": 0.000495099178255509,
      "loss": 5.5893,
      "mean_token_accuracy": 0.1729494720697403,
      "num_tokens": 17324268.0,
      "step": 8550
    },
    {
      "entropy": 5.959023571014404,
      "epoch": 0.7476186314777593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004950927078010934,
      "loss": 5.7086,
      "mean_token_accuracy": 0.16032904386520386,
      "num_tokens": 17334512.0,
      "step": 8555
    },
    {
      "entropy": 5.97030873298645,
      "epoch": 0.748055579830464,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004950862331251845,
      "loss": 5.6428,
      "mean_token_accuracy": 0.1734593242406845,
      "num_tokens": 17346244.0,
      "step": 8560
    },
    {
      "entropy": 5.939942979812622,
      "epoch": 0.7484925281831688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004950797542279061,
      "loss": 5.6434,
      "mean_token_accuracy": 0.17366824001073838,
      "num_tokens": 17355832.0,
      "step": 8565
    },
    {
      "entropy": 5.91514835357666,
      "epoch": 0.7489294765358735,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950732711093828,
      "loss": 5.6461,
      "mean_token_accuracy": 0.16829000413417816,
      "num_tokens": 17366871.0,
      "step": 8570
    },
    {
      "entropy": 5.991394329071045,
      "epoch": 0.7493664248885782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950667837697388,
      "loss": 5.7293,
      "mean_token_accuracy": 0.16241607815027237,
      "num_tokens": 17377242.0,
      "step": 8575
    },
    {
      "entropy": 5.997432279586792,
      "epoch": 0.7498033732412829,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004950602922090988,
      "loss": 5.6814,
      "mean_token_accuracy": 0.16915558129549027,
      "num_tokens": 17387942.0,
      "step": 8580
    },
    {
      "entropy": 5.9788885593414305,
      "epoch": 0.7502403215939876,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950537964275868,
      "loss": 5.6436,
      "mean_token_accuracy": 0.16623931229114533,
      "num_tokens": 17397413.0,
      "step": 8585
    },
    {
      "entropy": 5.932490873336792,
      "epoch": 0.7506772699466923,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950472964253278,
      "loss": 5.5908,
      "mean_token_accuracy": 0.17091099619865419,
      "num_tokens": 17408456.0,
      "step": 8590
    },
    {
      "entropy": 5.956561517715454,
      "epoch": 0.7511142182993971,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004950407922024463,
      "loss": 5.6198,
      "mean_token_accuracy": 0.17130438089370728,
      "num_tokens": 17417805.0,
      "step": 8595
    },
    {
      "entropy": 5.914918565750122,
      "epoch": 0.7515511666521018,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950342837590671,
      "loss": 5.5529,
      "mean_token_accuracy": 0.17918556928634644,
      "num_tokens": 17426932.0,
      "step": 8600
    },
    {
      "entropy": 5.87897138595581,
      "epoch": 0.7519881150048064,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950277710953152,
      "loss": 5.6482,
      "mean_token_accuracy": 0.1722386434674263,
      "num_tokens": 17436770.0,
      "step": 8605
    },
    {
      "entropy": 5.904389333724976,
      "epoch": 0.7524250633575111,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004950212542113153,
      "loss": 5.5702,
      "mean_token_accuracy": 0.17524368613958358,
      "num_tokens": 17447852.0,
      "step": 8610
    },
    {
      "entropy": 5.875369882583618,
      "epoch": 0.7528620117102158,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950147331071924,
      "loss": 5.5715,
      "mean_token_accuracy": 0.1711200311779976,
      "num_tokens": 17458020.0,
      "step": 8615
    },
    {
      "entropy": 5.977892351150513,
      "epoch": 0.7532989600629205,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004950082077830717,
      "loss": 5.684,
      "mean_token_accuracy": 0.16434485465288162,
      "num_tokens": 17468604.0,
      "step": 8620
    },
    {
      "entropy": 5.986811256408691,
      "epoch": 0.7537359084156253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004950016782390782,
      "loss": 5.6415,
      "mean_token_accuracy": 0.17272921949625014,
      "num_tokens": 17478374.0,
      "step": 8625
    },
    {
      "entropy": 5.964027214050293,
      "epoch": 0.75417285676833,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004949951444753372,
      "loss": 5.6362,
      "mean_token_accuracy": 0.1751129597425461,
      "num_tokens": 17487894.0,
      "step": 8630
    },
    {
      "entropy": 5.938914346694946,
      "epoch": 0.7546098051210347,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949886064919741,
      "loss": 5.6181,
      "mean_token_accuracy": 0.17235160022974014,
      "num_tokens": 17498473.0,
      "step": 8635
    },
    {
      "entropy": 5.922804117202759,
      "epoch": 0.7550467534737394,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004949820642891142,
      "loss": 5.6207,
      "mean_token_accuracy": 0.172652904689312,
      "num_tokens": 17507796.0,
      "step": 8640
    },
    {
      "entropy": 5.90608959197998,
      "epoch": 0.7554837018264441,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949755178668832,
      "loss": 5.6279,
      "mean_token_accuracy": 0.17529148161411284,
      "num_tokens": 17517629.0,
      "step": 8645
    },
    {
      "entropy": 5.9683849811553955,
      "epoch": 0.7559206501791488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949689672254064,
      "loss": 5.7032,
      "mean_token_accuracy": 0.16178804785013198,
      "num_tokens": 17528726.0,
      "step": 8650
    },
    {
      "entropy": 5.918129396438599,
      "epoch": 0.7563575985318536,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949624123648096,
      "loss": 5.6039,
      "mean_token_accuracy": 0.17397680580615998,
      "num_tokens": 17539022.0,
      "step": 8655
    },
    {
      "entropy": 5.958914470672608,
      "epoch": 0.7567945468845583,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949558532852185,
      "loss": 5.6537,
      "mean_token_accuracy": 0.17225720584392548,
      "num_tokens": 17549293.0,
      "step": 8660
    },
    {
      "entropy": 5.949682998657226,
      "epoch": 0.757231495237263,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949492899867587,
      "loss": 5.6464,
      "mean_token_accuracy": 0.16983148008584975,
      "num_tokens": 17559072.0,
      "step": 8665
    },
    {
      "entropy": 5.924809360504151,
      "epoch": 0.7576684435899677,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004949427224695564,
      "loss": 5.6179,
      "mean_token_accuracy": 0.17111153304576873,
      "num_tokens": 17568757.0,
      "step": 8670
    },
    {
      "entropy": 5.986868858337402,
      "epoch": 0.7581053919426723,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949361507337374,
      "loss": 5.6147,
      "mean_token_accuracy": 0.1758776366710663,
      "num_tokens": 17578288.0,
      "step": 8675
    },
    {
      "entropy": 5.956007099151611,
      "epoch": 0.758542340295377,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949295747794277,
      "loss": 5.7018,
      "mean_token_accuracy": 0.16617322564125062,
      "num_tokens": 17588864.0,
      "step": 8680
    },
    {
      "entropy": 5.936603975296021,
      "epoch": 0.7589792886480818,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004949229946067537,
      "loss": 5.7229,
      "mean_token_accuracy": 0.16777040362358092,
      "num_tokens": 17599103.0,
      "step": 8685
    },
    {
      "entropy": 5.893248176574707,
      "epoch": 0.7594162370007865,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949164102158413,
      "loss": 5.6055,
      "mean_token_accuracy": 0.175451098382473,
      "num_tokens": 17608608.0,
      "step": 8690
    },
    {
      "entropy": 5.982941722869873,
      "epoch": 0.7598531853534912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494909821606817,
      "loss": 5.692,
      "mean_token_accuracy": 0.17046781182289122,
      "num_tokens": 17618604.0,
      "step": 8695
    },
    {
      "entropy": 5.997108411788941,
      "epoch": 0.7602901337061959,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949032287798069,
      "loss": 5.6464,
      "mean_token_accuracy": 0.17591447681188582,
      "num_tokens": 17628479.0,
      "step": 8700
    },
    {
      "entropy": 5.910526132583618,
      "epoch": 0.7607270820589006,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948966317349379,
      "loss": 5.6541,
      "mean_token_accuracy": 0.17624739557504654,
      "num_tokens": 17637519.0,
      "step": 8705
    },
    {
      "entropy": 5.938403606414795,
      "epoch": 0.7611640304116053,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948900304723361,
      "loss": 5.6027,
      "mean_token_accuracy": 0.18379036337137222,
      "num_tokens": 17647902.0,
      "step": 8710
    },
    {
      "entropy": 5.98733081817627,
      "epoch": 0.7616009787643101,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948834249921285,
      "loss": 5.6105,
      "mean_token_accuracy": 0.17630842626094817,
      "num_tokens": 17658576.0,
      "step": 8715
    },
    {
      "entropy": 5.940467691421508,
      "epoch": 0.7620379271170148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948768152944414,
      "loss": 5.6629,
      "mean_token_accuracy": 0.17111257910728456,
      "num_tokens": 17667059.0,
      "step": 8720
    },
    {
      "entropy": 6.041933822631836,
      "epoch": 0.7624748754697195,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948702013794018,
      "loss": 5.7683,
      "mean_token_accuracy": 0.16528119295835494,
      "num_tokens": 17677447.0,
      "step": 8725
    },
    {
      "entropy": 5.9593089580535885,
      "epoch": 0.7629118238224242,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948635832471366,
      "loss": 5.6409,
      "mean_token_accuracy": 0.1806150943040848,
      "num_tokens": 17687893.0,
      "step": 8730
    },
    {
      "entropy": 5.907771396636963,
      "epoch": 0.7633487721751289,
      "grad_norm": 0.875,
      "learning_rate": 0.0004948569608977726,
      "loss": 5.6527,
      "mean_token_accuracy": 0.1764674738049507,
      "num_tokens": 17698599.0,
      "step": 8735
    },
    {
      "entropy": 5.972860050201416,
      "epoch": 0.7637857205278336,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948503343314369,
      "loss": 5.6513,
      "mean_token_accuracy": 0.1688207656145096,
      "num_tokens": 17709711.0,
      "step": 8740
    },
    {
      "entropy": 5.957196998596191,
      "epoch": 0.7642226688805384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004948437035482566,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17716830521821975,
      "num_tokens": 17719613.0,
      "step": 8745
    },
    {
      "entropy": 5.940670537948608,
      "epoch": 0.764659617233243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948370685483588,
      "loss": 5.6082,
      "mean_token_accuracy": 0.17935529053211213,
      "num_tokens": 17729829.0,
      "step": 8750
    },
    {
      "entropy": 5.897859287261963,
      "epoch": 0.7650965655859477,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948304293318709,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17040786892175674,
      "num_tokens": 17739254.0,
      "step": 8755
    },
    {
      "entropy": 5.963268136978149,
      "epoch": 0.7655335139386524,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948237858989201,
      "loss": 5.6375,
      "mean_token_accuracy": 0.16988625228405,
      "num_tokens": 17749294.0,
      "step": 8760
    },
    {
      "entropy": 5.894342088699341,
      "epoch": 0.7659704622913571,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494817138249634,
      "loss": 5.5689,
      "mean_token_accuracy": 0.18054407685995102,
      "num_tokens": 17758866.0,
      "step": 8765
    },
    {
      "entropy": 5.993446731567383,
      "epoch": 0.7664074106440619,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049481048638414,
      "loss": 5.7107,
      "mean_token_accuracy": 0.1637560099363327,
      "num_tokens": 17769202.0,
      "step": 8770
    },
    {
      "entropy": 5.918931913375855,
      "epoch": 0.7668443589967666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948038303025657,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1721897006034851,
      "num_tokens": 17779125.0,
      "step": 8775
    },
    {
      "entropy": 5.9706220626831055,
      "epoch": 0.7672813073494713,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947971700050387,
      "loss": 5.6574,
      "mean_token_accuracy": 0.16995926648378373,
      "num_tokens": 17788696.0,
      "step": 8780
    },
    {
      "entropy": 5.949014806747437,
      "epoch": 0.767718255702176,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947905054916869,
      "loss": 5.6398,
      "mean_token_accuracy": 0.1715715229511261,
      "num_tokens": 17799867.0,
      "step": 8785
    },
    {
      "entropy": 6.001501226425171,
      "epoch": 0.7681552040548807,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947838367626379,
      "loss": 5.6554,
      "mean_token_accuracy": 0.16742841601371766,
      "num_tokens": 17809418.0,
      "step": 8790
    },
    {
      "entropy": 5.916331958770752,
      "epoch": 0.7685921524075854,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947771638180198,
      "loss": 5.5413,
      "mean_token_accuracy": 0.1838361442089081,
      "num_tokens": 17818554.0,
      "step": 8795
    },
    {
      "entropy": 5.863605594635009,
      "epoch": 0.7690291007602902,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004947704866579605,
      "loss": 5.5929,
      "mean_token_accuracy": 0.1765707790851593,
      "num_tokens": 17828918.0,
      "step": 8800
    },
    {
      "entropy": 5.973348760604859,
      "epoch": 0.7694660491129949,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947638052825883,
      "loss": 5.7479,
      "mean_token_accuracy": 0.16142649352550506,
      "num_tokens": 17840791.0,
      "step": 8805
    },
    {
      "entropy": 6.010628414154053,
      "epoch": 0.7699029974656996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004947571196920309,
      "loss": 5.6535,
      "mean_token_accuracy": 0.1653089165687561,
      "num_tokens": 17850675.0,
      "step": 8810
    },
    {
      "entropy": 5.887283897399902,
      "epoch": 0.7703399458184043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947504298864169,
      "loss": 5.6119,
      "mean_token_accuracy": 0.1758929669857025,
      "num_tokens": 17860684.0,
      "step": 8815
    },
    {
      "entropy": 5.904197072982788,
      "epoch": 0.770776894171109,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004947437358658746,
      "loss": 5.6285,
      "mean_token_accuracy": 0.17459355145692826,
      "num_tokens": 17870394.0,
      "step": 8820
    },
    {
      "entropy": 5.91097617149353,
      "epoch": 0.7712138425238136,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947370376305321,
      "loss": 5.6008,
      "mean_token_accuracy": 0.1767829418182373,
      "num_tokens": 17881256.0,
      "step": 8825
    },
    {
      "entropy": 5.948611545562744,
      "epoch": 0.7716507908765184,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004947303351805183,
      "loss": 5.6504,
      "mean_token_accuracy": 0.17333670556545258,
      "num_tokens": 17890888.0,
      "step": 8830
    },
    {
      "entropy": 5.969726037979126,
      "epoch": 0.7720877392292231,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947236285159615,
      "loss": 5.6175,
      "mean_token_accuracy": 0.1730385646224022,
      "num_tokens": 17900507.0,
      "step": 8835
    },
    {
      "entropy": 5.890967082977295,
      "epoch": 0.7725246875819278,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947169176369902,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17292495518922807,
      "num_tokens": 17911333.0,
      "step": 8840
    },
    {
      "entropy": 5.98371992111206,
      "epoch": 0.7729616359346325,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947102025437334,
      "loss": 5.6432,
      "mean_token_accuracy": 0.17004254162311555,
      "num_tokens": 17921540.0,
      "step": 8845
    },
    {
      "entropy": 5.976136970520019,
      "epoch": 0.7733985842873372,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947034832363199,
      "loss": 5.7176,
      "mean_token_accuracy": 0.16572336107492447,
      "num_tokens": 17931820.0,
      "step": 8850
    },
    {
      "entropy": 5.96554217338562,
      "epoch": 0.7738355326400419,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004946967597148782,
      "loss": 5.6384,
      "mean_token_accuracy": 0.17082494199275972,
      "num_tokens": 17942717.0,
      "step": 8855
    },
    {
      "entropy": 5.9511066436767575,
      "epoch": 0.7742724809927467,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946900319795377,
      "loss": 5.6177,
      "mean_token_accuracy": 0.17906149327754975,
      "num_tokens": 17953775.0,
      "step": 8860
    },
    {
      "entropy": 5.938663291931152,
      "epoch": 0.7747094293454514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946833000304274,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1702869564294815,
      "num_tokens": 17963877.0,
      "step": 8865
    },
    {
      "entropy": 5.93529577255249,
      "epoch": 0.7751463776981561,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946765638676761,
      "loss": 5.6058,
      "mean_token_accuracy": 0.17915844470262526,
      "num_tokens": 17974015.0,
      "step": 8870
    },
    {
      "entropy": 5.88549427986145,
      "epoch": 0.7755833260508608,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946698234914132,
      "loss": 5.619,
      "mean_token_accuracy": 0.17810313403606415,
      "num_tokens": 17984189.0,
      "step": 8875
    },
    {
      "entropy": 5.885636615753174,
      "epoch": 0.7760202744035655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494663078901768,
      "loss": 5.5703,
      "mean_token_accuracy": 0.18005340695381164,
      "num_tokens": 17994615.0,
      "step": 8880
    },
    {
      "entropy": 5.958274221420288,
      "epoch": 0.7764572227562702,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946563300988699,
      "loss": 5.6278,
      "mean_token_accuracy": 0.17140327244997025,
      "num_tokens": 18004727.0,
      "step": 8885
    },
    {
      "entropy": 5.953384017944336,
      "epoch": 0.776894171108975,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946495770828482,
      "loss": 5.6946,
      "mean_token_accuracy": 0.1725299909710884,
      "num_tokens": 18014576.0,
      "step": 8890
    },
    {
      "entropy": 5.9294633865356445,
      "epoch": 0.7773311194616797,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946428198538326,
      "loss": 5.6606,
      "mean_token_accuracy": 0.17412893623113632,
      "num_tokens": 18024709.0,
      "step": 8895
    },
    {
      "entropy": 5.973984050750732,
      "epoch": 0.7777680678143843,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946360584119526,
      "loss": 5.6059,
      "mean_token_accuracy": 0.17208971828222275,
      "num_tokens": 18034673.0,
      "step": 8900
    },
    {
      "entropy": 5.9540468692779545,
      "epoch": 0.778205016167089,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494629292757338,
      "loss": 5.6147,
      "mean_token_accuracy": 0.17223875671625138,
      "num_tokens": 18044774.0,
      "step": 8905
    },
    {
      "entropy": 5.898390483856201,
      "epoch": 0.7786419645197937,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946225228901184,
      "loss": 5.6103,
      "mean_token_accuracy": 0.16923121809959413,
      "num_tokens": 18055244.0,
      "step": 8910
    },
    {
      "entropy": 5.95582947731018,
      "epoch": 0.7790789128724984,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004946157488104236,
      "loss": 5.7008,
      "mean_token_accuracy": 0.16536194533109666,
      "num_tokens": 18066051.0,
      "step": 8915
    },
    {
      "entropy": 5.995481681823731,
      "epoch": 0.7795158612252032,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004946089705183839,
      "loss": 5.7186,
      "mean_token_accuracy": 0.16110515147447585,
      "num_tokens": 18077693.0,
      "step": 8920
    },
    {
      "entropy": 5.939252138137817,
      "epoch": 0.7799528095779079,
      "grad_norm": 0.875,
      "learning_rate": 0.0004946021880141289,
      "loss": 5.6458,
      "mean_token_accuracy": 0.17277394384145736,
      "num_tokens": 18088032.0,
      "step": 8925
    },
    {
      "entropy": 5.950494813919067,
      "epoch": 0.7803897579306126,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494595401297789,
      "loss": 5.6185,
      "mean_token_accuracy": 0.17533794194459915,
      "num_tokens": 18097998.0,
      "step": 8930
    },
    {
      "entropy": 5.933243036270142,
      "epoch": 0.7808267062833173,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004945886103694941,
      "loss": 5.5736,
      "mean_token_accuracy": 0.17785312235355377,
      "num_tokens": 18107538.0,
      "step": 8935
    },
    {
      "entropy": 5.909284257888794,
      "epoch": 0.781263654636022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945818152293747,
      "loss": 5.662,
      "mean_token_accuracy": 0.17331621348857879,
      "num_tokens": 18117792.0,
      "step": 8940
    },
    {
      "entropy": 5.969169998168946,
      "epoch": 0.7817006029887267,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945750158775609,
      "loss": 5.6646,
      "mean_token_accuracy": 0.1717840924859047,
      "num_tokens": 18128005.0,
      "step": 8945
    },
    {
      "entropy": 5.909918355941772,
      "epoch": 0.7821375513414315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945682123141833,
      "loss": 5.643,
      "mean_token_accuracy": 0.17730848491191864,
      "num_tokens": 18138398.0,
      "step": 8950
    },
    {
      "entropy": 5.975086116790772,
      "epoch": 0.7825744996941362,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004945614045393724,
      "loss": 5.6146,
      "mean_token_accuracy": 0.17089865654706954,
      "num_tokens": 18147676.0,
      "step": 8955
    },
    {
      "entropy": 5.838088035583496,
      "epoch": 0.7830114480468409,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945545925532587,
      "loss": 5.527,
      "mean_token_accuracy": 0.18753522336483003,
      "num_tokens": 18156842.0,
      "step": 8960
    },
    {
      "entropy": 5.908769178390503,
      "epoch": 0.7834483963995456,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945477763559728,
      "loss": 5.6449,
      "mean_token_accuracy": 0.16971031725406646,
      "num_tokens": 18166902.0,
      "step": 8965
    },
    {
      "entropy": 5.960187578201294,
      "epoch": 0.7838853447522502,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945409559476455,
      "loss": 5.6325,
      "mean_token_accuracy": 0.1739122226834297,
      "num_tokens": 18177561.0,
      "step": 8970
    },
    {
      "entropy": 5.860386800765991,
      "epoch": 0.784322293104955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945341313284077,
      "loss": 5.6226,
      "mean_token_accuracy": 0.177939672768116,
      "num_tokens": 18187739.0,
      "step": 8975
    },
    {
      "entropy": 5.881693601608276,
      "epoch": 0.7847592414576597,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004945273024983902,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17986249625682832,
      "num_tokens": 18198435.0,
      "step": 8980
    },
    {
      "entropy": 5.945936918258667,
      "epoch": 0.7851961898103644,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494520469457724,
      "loss": 5.5648,
      "mean_token_accuracy": 0.18187845200300218,
      "num_tokens": 18208374.0,
      "step": 8985
    },
    {
      "entropy": 5.814629077911377,
      "epoch": 0.7856331381630691,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004945136322065402,
      "loss": 5.6475,
      "mean_token_accuracy": 0.17407305389642716,
      "num_tokens": 18219229.0,
      "step": 8990
    },
    {
      "entropy": 5.929801940917969,
      "epoch": 0.7860700865157738,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945067907449699,
      "loss": 5.598,
      "mean_token_accuracy": 0.17482235580682753,
      "num_tokens": 18229208.0,
      "step": 8995
    },
    {
      "entropy": 6.0301402568817135,
      "epoch": 0.7865070348684785,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944999450731442,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1576134629547596,
      "num_tokens": 18238454.0,
      "step": 9000
    },
    {
      "epoch": 0.7865070348684785,
      "eval_entropy": 5.815047346852577,
      "eval_loss": 5.638741493225098,
      "eval_mean_token_accuracy": 0.18099814363633915,
      "eval_num_tokens": 18238454.0,
      "eval_runtime": 21.3295,
      "eval_samples_per_second": 1735.345,
      "eval_steps_per_second": 216.93,
      "step": 9000
    },
    {
      "entropy": 5.991175603866577,
      "epoch": 0.7869439832211833,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944930951911947,
      "loss": 5.6294,
      "mean_token_accuracy": 0.17374134957790374,
      "num_tokens": 18248845.0,
      "step": 9005
    },
    {
      "entropy": 5.993597173690796,
      "epoch": 0.787380931573888,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944862410992526,
      "loss": 5.6339,
      "mean_token_accuracy": 0.16740865856409073,
      "num_tokens": 18258912.0,
      "step": 9010
    },
    {
      "entropy": 5.961672782897949,
      "epoch": 0.7878178799265927,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944793827974494,
      "loss": 5.6659,
      "mean_token_accuracy": 0.1684281572699547,
      "num_tokens": 18269330.0,
      "step": 9015
    },
    {
      "entropy": 5.949158716201782,
      "epoch": 0.7882548282792974,
      "grad_norm": 0.875,
      "learning_rate": 0.0004944725202859165,
      "loss": 5.646,
      "mean_token_accuracy": 0.17535251379013062,
      "num_tokens": 18279721.0,
      "step": 9020
    },
    {
      "entropy": 5.884190082550049,
      "epoch": 0.7886917766320021,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944656535647858,
      "loss": 5.5675,
      "mean_token_accuracy": 0.18021733909845353,
      "num_tokens": 18288581.0,
      "step": 9025
    },
    {
      "entropy": 5.9276820659637455,
      "epoch": 0.7891287249847068,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944587826341888,
      "loss": 5.6136,
      "mean_token_accuracy": 0.17172063440084456,
      "num_tokens": 18298179.0,
      "step": 9030
    },
    {
      "entropy": 5.946098470687867,
      "epoch": 0.7895656733374116,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944519074942574,
      "loss": 5.6208,
      "mean_token_accuracy": 0.16878654658794404,
      "num_tokens": 18307957.0,
      "step": 9035
    },
    {
      "entropy": 5.849981737136841,
      "epoch": 0.7900026216901163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004944450281451234,
      "loss": 5.6006,
      "mean_token_accuracy": 0.1710733339190483,
      "num_tokens": 18317799.0,
      "step": 9040
    },
    {
      "entropy": 5.991630125045776,
      "epoch": 0.790439570042821,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004944381445869188,
      "loss": 5.6572,
      "mean_token_accuracy": 0.17650736421346663,
      "num_tokens": 18327889.0,
      "step": 9045
    },
    {
      "entropy": 5.943081855773926,
      "epoch": 0.7908765183955256,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004944312568197756,
      "loss": 5.6282,
      "mean_token_accuracy": 0.17297194451093673,
      "num_tokens": 18338288.0,
      "step": 9050
    },
    {
      "entropy": 5.932630825042724,
      "epoch": 0.7913134667482303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944243648438258,
      "loss": 5.6902,
      "mean_token_accuracy": 0.16534946709871293,
      "num_tokens": 18349070.0,
      "step": 9055
    },
    {
      "entropy": 5.992298889160156,
      "epoch": 0.791750415100935,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004944174686592019,
      "loss": 5.5997,
      "mean_token_accuracy": 0.17692964375019074,
      "num_tokens": 18359683.0,
      "step": 9060
    },
    {
      "entropy": 5.92846450805664,
      "epoch": 0.7921873634536398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004944105682660358,
      "loss": 5.5991,
      "mean_token_accuracy": 0.17372017502784728,
      "num_tokens": 18371355.0,
      "step": 9065
    },
    {
      "entropy": 5.853173112869262,
      "epoch": 0.7926243118063445,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944036636644601,
      "loss": 5.6397,
      "mean_token_accuracy": 0.17553496211767197,
      "num_tokens": 18381591.0,
      "step": 9070
    },
    {
      "entropy": 5.933531999588013,
      "epoch": 0.7930612601590492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004943967548546071,
      "loss": 5.6156,
      "mean_token_accuracy": 0.17261812686920167,
      "num_tokens": 18390760.0,
      "step": 9075
    },
    {
      "entropy": 5.954724740982056,
      "epoch": 0.7934982085117539,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943898418366096,
      "loss": 5.6575,
      "mean_token_accuracy": 0.1722110092639923,
      "num_tokens": 18400643.0,
      "step": 9080
    },
    {
      "entropy": 5.971835470199585,
      "epoch": 0.7939351568644586,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004943829246105997,
      "loss": 5.699,
      "mean_token_accuracy": 0.1661178410053253,
      "num_tokens": 18411559.0,
      "step": 9085
    },
    {
      "entropy": 5.964496421813965,
      "epoch": 0.7943721052171633,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943760031767106,
      "loss": 5.6587,
      "mean_token_accuracy": 0.1780708521604538,
      "num_tokens": 18422645.0,
      "step": 9090
    },
    {
      "entropy": 5.931650400161743,
      "epoch": 0.7948090535698681,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943690775350746,
      "loss": 5.6077,
      "mean_token_accuracy": 0.17388636618852615,
      "num_tokens": 18432369.0,
      "step": 9095
    },
    {
      "entropy": 5.934646081924439,
      "epoch": 0.7952460019225728,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943621476858249,
      "loss": 5.5923,
      "mean_token_accuracy": 0.17633734345436097,
      "num_tokens": 18442143.0,
      "step": 9100
    },
    {
      "entropy": 5.885641050338745,
      "epoch": 0.7956829502752775,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943552136290942,
      "loss": 5.639,
      "mean_token_accuracy": 0.1716434806585312,
      "num_tokens": 18452161.0,
      "step": 9105
    },
    {
      "entropy": 5.995611619949341,
      "epoch": 0.7961198986279822,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943482753650154,
      "loss": 5.7352,
      "mean_token_accuracy": 0.1679619863629341,
      "num_tokens": 18461873.0,
      "step": 9110
    },
    {
      "entropy": 5.966622066497803,
      "epoch": 0.7965568469806869,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943413328937221,
      "loss": 5.6104,
      "mean_token_accuracy": 0.17477046251296996,
      "num_tokens": 18472584.0,
      "step": 9115
    },
    {
      "entropy": 5.942466783523559,
      "epoch": 0.7969937953333915,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943343862153467,
      "loss": 5.5788,
      "mean_token_accuracy": 0.17355528473854065,
      "num_tokens": 18482174.0,
      "step": 9120
    },
    {
      "entropy": 5.943079090118408,
      "epoch": 0.7974307436860963,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004943274353300231,
      "loss": 5.731,
      "mean_token_accuracy": 0.1694606900215149,
      "num_tokens": 18492861.0,
      "step": 9125
    },
    {
      "entropy": 5.99499773979187,
      "epoch": 0.797867692038801,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943204802378843,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17418318688869477,
      "num_tokens": 18502830.0,
      "step": 9130
    },
    {
      "entropy": 5.968546915054321,
      "epoch": 0.7983046403915057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943135209390638,
      "loss": 5.6298,
      "mean_token_accuracy": 0.16796215027570724,
      "num_tokens": 18513063.0,
      "step": 9135
    },
    {
      "entropy": 5.832057142257691,
      "epoch": 0.7987415887442104,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943065574336949,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1782997280359268,
      "num_tokens": 18523773.0,
      "step": 9140
    },
    {
      "entropy": 5.961545753479004,
      "epoch": 0.7991785370969151,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942995897219115,
      "loss": 5.6224,
      "mean_token_accuracy": 0.17051804065704346,
      "num_tokens": 18533101.0,
      "step": 9145
    },
    {
      "entropy": 5.985107803344727,
      "epoch": 0.7996154854496198,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942926178038469,
      "loss": 5.6459,
      "mean_token_accuracy": 0.178301540017128,
      "num_tokens": 18543863.0,
      "step": 9150
    },
    {
      "entropy": 5.8952085971832275,
      "epoch": 0.8000524338023246,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942856416796352,
      "loss": 5.6435,
      "mean_token_accuracy": 0.17575568854808807,
      "num_tokens": 18554249.0,
      "step": 9155
    },
    {
      "entropy": 5.905969953536987,
      "epoch": 0.8004893821550293,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942786613494098,
      "loss": 5.6203,
      "mean_token_accuracy": 0.17257825136184693,
      "num_tokens": 18563525.0,
      "step": 9160
    },
    {
      "entropy": 5.946135473251343,
      "epoch": 0.800926330507734,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004942716768133048,
      "loss": 5.6039,
      "mean_token_accuracy": 0.176299287378788,
      "num_tokens": 18572734.0,
      "step": 9165
    },
    {
      "entropy": 5.910472106933594,
      "epoch": 0.8013632788604387,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942646880714543,
      "loss": 5.6679,
      "mean_token_accuracy": 0.17312319129705428,
      "num_tokens": 18581939.0,
      "step": 9170
    },
    {
      "entropy": 5.934965991973877,
      "epoch": 0.8018002272131434,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494257695123992,
      "loss": 5.6532,
      "mean_token_accuracy": 0.17528002411127092,
      "num_tokens": 18591635.0,
      "step": 9175
    },
    {
      "entropy": 5.935646867752075,
      "epoch": 0.8022371755658482,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942506979710524,
      "loss": 5.5491,
      "mean_token_accuracy": 0.1734914541244507,
      "num_tokens": 18602060.0,
      "step": 9180
    },
    {
      "entropy": 5.93180832862854,
      "epoch": 0.8026741239185529,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942436966127693,
      "loss": 5.6001,
      "mean_token_accuracy": 0.1750059559941292,
      "num_tokens": 18611187.0,
      "step": 9185
    },
    {
      "entropy": 5.963546705245972,
      "epoch": 0.8031110722712576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942366910492775,
      "loss": 5.5466,
      "mean_token_accuracy": 0.18096266835927963,
      "num_tokens": 18620329.0,
      "step": 9190
    },
    {
      "entropy": 5.953118419647216,
      "epoch": 0.8035480206239622,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004942296812807109,
      "loss": 5.6302,
      "mean_token_accuracy": 0.17430526316165923,
      "num_tokens": 18629600.0,
      "step": 9195
    },
    {
      "entropy": 5.925580978393555,
      "epoch": 0.8039849689766669,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942226673072041,
      "loss": 5.6854,
      "mean_token_accuracy": 0.16640919744968413,
      "num_tokens": 18640771.0,
      "step": 9200
    },
    {
      "entropy": 5.908667659759521,
      "epoch": 0.8044219173293716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942156491288917,
      "loss": 5.5869,
      "mean_token_accuracy": 0.17425418496131898,
      "num_tokens": 18650606.0,
      "step": 9205
    },
    {
      "entropy": 5.930092239379883,
      "epoch": 0.8048588656820764,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942086267459083,
      "loss": 5.6288,
      "mean_token_accuracy": 0.17468114644289018,
      "num_tokens": 18660487.0,
      "step": 9210
    },
    {
      "entropy": 5.9295285701751705,
      "epoch": 0.8052958140347811,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942016001583886,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17546120285987854,
      "num_tokens": 18670987.0,
      "step": 9215
    },
    {
      "entropy": 5.949113225936889,
      "epoch": 0.8057327623874858,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941945693664674,
      "loss": 5.6969,
      "mean_token_accuracy": 0.16998145431280137,
      "num_tokens": 18682560.0,
      "step": 9220
    },
    {
      "entropy": 5.973329162597656,
      "epoch": 0.8061697107401905,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004941875343702795,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1697494775056839,
      "num_tokens": 18692743.0,
      "step": 9225
    },
    {
      "entropy": 5.928528499603272,
      "epoch": 0.8066066590928952,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941804951699597,
      "loss": 5.5907,
      "mean_token_accuracy": 0.17869127541780472,
      "num_tokens": 18703039.0,
      "step": 9230
    },
    {
      "entropy": 5.840265607833862,
      "epoch": 0.8070436074455999,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941734517656432,
      "loss": 5.6442,
      "mean_token_accuracy": 0.17035192996263504,
      "num_tokens": 18713777.0,
      "step": 9235
    },
    {
      "entropy": 5.918832778930664,
      "epoch": 0.8074805557983047,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941664041574651,
      "loss": 5.6511,
      "mean_token_accuracy": 0.17143665701150895,
      "num_tokens": 18723422.0,
      "step": 9240
    },
    {
      "entropy": 5.944339275360107,
      "epoch": 0.8079175041510094,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941593523455605,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17621628940105438,
      "num_tokens": 18734369.0,
      "step": 9245
    },
    {
      "entropy": 5.9933618068695065,
      "epoch": 0.8083544525037141,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004941522963300647,
      "loss": 5.645,
      "mean_token_accuracy": 0.17366275191307068,
      "num_tokens": 18744647.0,
      "step": 9250
    },
    {
      "entropy": 5.977871894836426,
      "epoch": 0.8087914008564188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494145236111113,
      "loss": 5.5979,
      "mean_token_accuracy": 0.17265502810478212,
      "num_tokens": 18753929.0,
      "step": 9255
    },
    {
      "entropy": 5.952688980102539,
      "epoch": 0.8092283492091235,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941381716888409,
      "loss": 5.6243,
      "mean_token_accuracy": 0.17520538121461868,
      "num_tokens": 18764103.0,
      "step": 9260
    },
    {
      "entropy": 5.947621583938599,
      "epoch": 0.8096652975618281,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004941311030633838,
      "loss": 5.6547,
      "mean_token_accuracy": 0.16614218205213546,
      "num_tokens": 18773756.0,
      "step": 9265
    },
    {
      "entropy": 5.8903004169464115,
      "epoch": 0.810102245914533,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941240302348773,
      "loss": 5.5133,
      "mean_token_accuracy": 0.19039669036865234,
      "num_tokens": 18783778.0,
      "step": 9270
    },
    {
      "entropy": 5.9246738910675045,
      "epoch": 0.8105391942672376,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941169532034571,
      "loss": 5.6403,
      "mean_token_accuracy": 0.17501290291547775,
      "num_tokens": 18794664.0,
      "step": 9275
    },
    {
      "entropy": 5.880994558334351,
      "epoch": 0.8109761426199423,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941098719692589,
      "loss": 5.6191,
      "mean_token_accuracy": 0.17048076391220093,
      "num_tokens": 18803773.0,
      "step": 9280
    },
    {
      "entropy": 5.9081196784973145,
      "epoch": 0.811413090972647,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941027865324186,
      "loss": 5.657,
      "mean_token_accuracy": 0.17150548249483108,
      "num_tokens": 18814419.0,
      "step": 9285
    },
    {
      "entropy": 6.027905654907227,
      "epoch": 0.8118500393253517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494095696893072,
      "loss": 5.6309,
      "mean_token_accuracy": 0.17258746027946473,
      "num_tokens": 18825011.0,
      "step": 9290
    },
    {
      "entropy": 5.908768510818481,
      "epoch": 0.8122869876780564,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940886030513551,
      "loss": 5.5332,
      "mean_token_accuracy": 0.18093588203191757,
      "num_tokens": 18835003.0,
      "step": 9295
    },
    {
      "entropy": 5.897600555419922,
      "epoch": 0.8127239360307612,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940815050074041,
      "loss": 5.6327,
      "mean_token_accuracy": 0.1715462699532509,
      "num_tokens": 18845072.0,
      "step": 9300
    },
    {
      "entropy": 5.983028793334961,
      "epoch": 0.8131608843834659,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940744027613549,
      "loss": 5.619,
      "mean_token_accuracy": 0.17670044004917146,
      "num_tokens": 18854583.0,
      "step": 9305
    },
    {
      "entropy": 5.948778343200684,
      "epoch": 0.8135978327361706,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940672963133439,
      "loss": 5.6327,
      "mean_token_accuracy": 0.1762762725353241,
      "num_tokens": 18865460.0,
      "step": 9310
    },
    {
      "entropy": 5.885643148422242,
      "epoch": 0.8140347810888753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940601856635074,
      "loss": 5.6334,
      "mean_token_accuracy": 0.17521110773086548,
      "num_tokens": 18875568.0,
      "step": 9315
    },
    {
      "entropy": 5.891025543212891,
      "epoch": 0.81447172944158,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940530708119818,
      "loss": 5.5314,
      "mean_token_accuracy": 0.17806996107101442,
      "num_tokens": 18884894.0,
      "step": 9320
    },
    {
      "entropy": 5.900966024398803,
      "epoch": 0.8149086777942847,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004940459517589034,
      "loss": 5.5945,
      "mean_token_accuracy": 0.17829823046922683,
      "num_tokens": 18895353.0,
      "step": 9325
    },
    {
      "entropy": 5.919049167633057,
      "epoch": 0.8153456261469895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494038828504409,
      "loss": 5.6404,
      "mean_token_accuracy": 0.17730499804019928,
      "num_tokens": 18905474.0,
      "step": 9330
    },
    {
      "entropy": 5.9567101955413815,
      "epoch": 0.8157825744996942,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940317010486351,
      "loss": 5.5706,
      "mean_token_accuracy": 0.1754348322749138,
      "num_tokens": 18914797.0,
      "step": 9335
    },
    {
      "entropy": 5.935894012451172,
      "epoch": 0.8162195228523988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940245693917183,
      "loss": 5.6409,
      "mean_token_accuracy": 0.17083175033330916,
      "num_tokens": 18924642.0,
      "step": 9340
    },
    {
      "entropy": 5.854722166061402,
      "epoch": 0.8166564712051035,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940174335337955,
      "loss": 5.5286,
      "mean_token_accuracy": 0.18536011576652528,
      "num_tokens": 18934026.0,
      "step": 9345
    },
    {
      "entropy": 5.9182079315185545,
      "epoch": 0.8170934195578082,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004940102934750037,
      "loss": 5.5371,
      "mean_token_accuracy": 0.1883857801556587,
      "num_tokens": 18943262.0,
      "step": 9350
    },
    {
      "entropy": 5.893864583969116,
      "epoch": 0.8175303679105129,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940031492154795,
      "loss": 5.5855,
      "mean_token_accuracy": 0.16936717927455902,
      "num_tokens": 18953082.0,
      "step": 9355
    },
    {
      "entropy": 5.880031108856201,
      "epoch": 0.8179673162632177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939960007553603,
      "loss": 5.5999,
      "mean_token_accuracy": 0.17679183930158615,
      "num_tokens": 18962408.0,
      "step": 9360
    },
    {
      "entropy": 6.0023472785949705,
      "epoch": 0.8184042646159224,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493988848094783,
      "loss": 5.662,
      "mean_token_accuracy": 0.16287591606378554,
      "num_tokens": 18972918.0,
      "step": 9365
    },
    {
      "entropy": 5.974187898635864,
      "epoch": 0.8188412129686271,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004939816912338848,
      "loss": 5.6369,
      "mean_token_accuracy": 0.17226020395755767,
      "num_tokens": 18983152.0,
      "step": 9370
    },
    {
      "entropy": 5.8858983516693115,
      "epoch": 0.8192781613213318,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004939745301728031,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17253263890743256,
      "num_tokens": 18993613.0,
      "step": 9375
    },
    {
      "entropy": 5.8906982421875,
      "epoch": 0.8197151096740365,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004939673649116752,
      "loss": 5.5368,
      "mean_token_accuracy": 0.17805820256471633,
      "num_tokens": 19004079.0,
      "step": 9380
    },
    {
      "entropy": 5.888335704803467,
      "epoch": 0.8201520580267413,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004939601954506383,
      "loss": 5.6789,
      "mean_token_accuracy": 0.17126546800136566,
      "num_tokens": 19013994.0,
      "step": 9385
    },
    {
      "entropy": 5.936763620376587,
      "epoch": 0.820589006379446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004939530217898303,
      "loss": 5.6659,
      "mean_token_accuracy": 0.16886581778526305,
      "num_tokens": 19024674.0,
      "step": 9390
    },
    {
      "entropy": 5.9656651496887205,
      "epoch": 0.8210259547321507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939458439293886,
      "loss": 5.6398,
      "mean_token_accuracy": 0.1725035697221756,
      "num_tokens": 19035333.0,
      "step": 9395
    },
    {
      "entropy": 6.025409936904907,
      "epoch": 0.8214629030848554,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004939386618694508,
      "loss": 5.6681,
      "mean_token_accuracy": 0.17290944755077362,
      "num_tokens": 19046279.0,
      "step": 9400
    },
    {
      "entropy": 5.915491342544556,
      "epoch": 0.8218998514375601,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004939314756101547,
      "loss": 5.6257,
      "mean_token_accuracy": 0.16786998212337495,
      "num_tokens": 19056506.0,
      "step": 9405
    },
    {
      "entropy": 5.919381999969483,
      "epoch": 0.8223367997902647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004939242851516384,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17436128109693527,
      "num_tokens": 19067217.0,
      "step": 9410
    },
    {
      "entropy": 5.934492301940918,
      "epoch": 0.8227737481429696,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004939170904940395,
      "loss": 5.6687,
      "mean_token_accuracy": 0.17424398064613342,
      "num_tokens": 19079028.0,
      "step": 9415
    },
    {
      "entropy": 5.8372962951660154,
      "epoch": 0.8232106964956742,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493909891637496,
      "loss": 5.6428,
      "mean_token_accuracy": 0.1786332130432129,
      "num_tokens": 19088236.0,
      "step": 9420
    },
    {
      "entropy": 5.893320178985595,
      "epoch": 0.8236476448483789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939026885821461,
      "loss": 5.594,
      "mean_token_accuracy": 0.18137700706720353,
      "num_tokens": 19099598.0,
      "step": 9425
    },
    {
      "entropy": 5.960141277313232,
      "epoch": 0.8240845932010836,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493895481328128,
      "loss": 5.6494,
      "mean_token_accuracy": 0.1792005032300949,
      "num_tokens": 19110347.0,
      "step": 9430
    },
    {
      "entropy": 5.998931741714477,
      "epoch": 0.8245215415537883,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938882698755798,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17032975256443023,
      "num_tokens": 19120643.0,
      "step": 9435
    },
    {
      "entropy": 5.871499872207641,
      "epoch": 0.824958489906493,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049388105422464,
      "loss": 5.5932,
      "mean_token_accuracy": 0.17987883239984512,
      "num_tokens": 19130209.0,
      "step": 9440
    },
    {
      "entropy": 5.942246389389038,
      "epoch": 0.8253954382591978,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938738343754467,
      "loss": 5.5706,
      "mean_token_accuracy": 0.1778632327914238,
      "num_tokens": 19139196.0,
      "step": 9445
    },
    {
      "entropy": 5.941312217712403,
      "epoch": 0.8258323866119025,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004938666103281389,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17546364217996596,
      "num_tokens": 19149618.0,
      "step": 9450
    },
    {
      "entropy": 5.959366035461426,
      "epoch": 0.8262693349646072,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004938593820828545,
      "loss": 5.5724,
      "mean_token_accuracy": 0.17675871551036834,
      "num_tokens": 19159875.0,
      "step": 9455
    },
    {
      "entropy": 5.912749290466309,
      "epoch": 0.8267062833173119,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004938521496397326,
      "loss": 5.6114,
      "mean_token_accuracy": 0.1724828377366066,
      "num_tokens": 19170403.0,
      "step": 9460
    },
    {
      "entropy": 5.902295780181885,
      "epoch": 0.8271432316700166,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004938449129989117,
      "loss": 5.5749,
      "mean_token_accuracy": 0.18494125902652742,
      "num_tokens": 19181187.0,
      "step": 9465
    },
    {
      "entropy": 5.91620659828186,
      "epoch": 0.8275801800227213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938376721605308,
      "loss": 5.639,
      "mean_token_accuracy": 0.17147258520126343,
      "num_tokens": 19191328.0,
      "step": 9470
    },
    {
      "entropy": 5.908253812789917,
      "epoch": 0.8280171283754261,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938304271247286,
      "loss": 5.5911,
      "mean_token_accuracy": 0.17617038637399673,
      "num_tokens": 19202041.0,
      "step": 9475
    },
    {
      "entropy": 5.905862760543823,
      "epoch": 0.8284540767281308,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938231778916442,
      "loss": 5.5835,
      "mean_token_accuracy": 0.17448510676622392,
      "num_tokens": 19212016.0,
      "step": 9480
    },
    {
      "entropy": 5.933022499084473,
      "epoch": 0.8288910250808355,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938159244614167,
      "loss": 5.5767,
      "mean_token_accuracy": 0.17353864908218383,
      "num_tokens": 19222111.0,
      "step": 9485
    },
    {
      "entropy": 5.9617821216583256,
      "epoch": 0.8293279734335401,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004938086668341849,
      "loss": 5.5601,
      "mean_token_accuracy": 0.17571896463632583,
      "num_tokens": 19231865.0,
      "step": 9490
    },
    {
      "entropy": 5.829162645339966,
      "epoch": 0.8297649217862448,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938014050100884,
      "loss": 5.564,
      "mean_token_accuracy": 0.18120294213294982,
      "num_tokens": 19242993.0,
      "step": 9495
    },
    {
      "entropy": 5.877204561233521,
      "epoch": 0.8302018701389495,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937941389892662,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17481131255626678,
      "num_tokens": 19252776.0,
      "step": 9500
    },
    {
      "entropy": 5.929868221282959,
      "epoch": 0.8306388184916543,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937868687718577,
      "loss": 5.5812,
      "mean_token_accuracy": 0.17314620018005372,
      "num_tokens": 19263022.0,
      "step": 9505
    },
    {
      "entropy": 5.947000455856323,
      "epoch": 0.831075766844359,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937795943580026,
      "loss": 5.6109,
      "mean_token_accuracy": 0.17933596223592757,
      "num_tokens": 19271859.0,
      "step": 9510
    },
    {
      "entropy": 5.841096448898315,
      "epoch": 0.8315127151970637,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049377231574784,
      "loss": 5.5507,
      "mean_token_accuracy": 0.182484370470047,
      "num_tokens": 19282771.0,
      "step": 9515
    },
    {
      "entropy": 5.922836065292358,
      "epoch": 0.8319496635497684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004937650329415099,
      "loss": 5.6096,
      "mean_token_accuracy": 0.16887279152870177,
      "num_tokens": 19292985.0,
      "step": 9520
    },
    {
      "entropy": 5.902521324157715,
      "epoch": 0.8323866119024731,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004937577459391519,
      "loss": 5.6327,
      "mean_token_accuracy": 0.1754354476928711,
      "num_tokens": 19303409.0,
      "step": 9525
    },
    {
      "entropy": 5.922741556167603,
      "epoch": 0.8328235602551778,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937504547409056,
      "loss": 5.611,
      "mean_token_accuracy": 0.17542541772127151,
      "num_tokens": 19312632.0,
      "step": 9530
    },
    {
      "entropy": 5.907842636108398,
      "epoch": 0.8332605086078826,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004937431593469109,
      "loss": 5.5041,
      "mean_token_accuracy": 0.18104026168584825,
      "num_tokens": 19322759.0,
      "step": 9535
    },
    {
      "entropy": 5.902064180374145,
      "epoch": 0.8336974569605873,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937358597573079,
      "loss": 5.5557,
      "mean_token_accuracy": 0.18379427194595338,
      "num_tokens": 19333307.0,
      "step": 9540
    },
    {
      "entropy": 5.831543397903443,
      "epoch": 0.834134405313292,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937285559722364,
      "loss": 5.6536,
      "mean_token_accuracy": 0.17159395515918732,
      "num_tokens": 19342528.0,
      "step": 9545
    },
    {
      "entropy": 5.951305675506592,
      "epoch": 0.8345713536659967,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937212479918367,
      "loss": 5.6138,
      "mean_token_accuracy": 0.1725657731294632,
      "num_tokens": 19352640.0,
      "step": 9550
    },
    {
      "entropy": 5.959615755081177,
      "epoch": 0.8350083020187014,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937139358162488,
      "loss": 5.5743,
      "mean_token_accuracy": 0.1781083568930626,
      "num_tokens": 19363285.0,
      "step": 9555
    },
    {
      "entropy": 5.906778573989868,
      "epoch": 0.835445250371406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493706619445613,
      "loss": 5.5523,
      "mean_token_accuracy": 0.18480108827352523,
      "num_tokens": 19373141.0,
      "step": 9560
    },
    {
      "entropy": 5.851489734649658,
      "epoch": 0.8358821987241108,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004936992988800697,
      "loss": 5.6358,
      "mean_token_accuracy": 0.16956086009740828,
      "num_tokens": 19383091.0,
      "step": 9565
    },
    {
      "entropy": 5.908646821975708,
      "epoch": 0.8363191470768155,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004936919741197591,
      "loss": 5.5639,
      "mean_token_accuracy": 0.17686478793621063,
      "num_tokens": 19393163.0,
      "step": 9570
    },
    {
      "entropy": 5.909564685821533,
      "epoch": 0.8367560954295202,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936846451648219,
      "loss": 5.5297,
      "mean_token_accuracy": 0.18357791602611542,
      "num_tokens": 19403475.0,
      "step": 9575
    },
    {
      "entropy": 5.824192190170288,
      "epoch": 0.8371930437822249,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936773120153987,
      "loss": 5.5647,
      "mean_token_accuracy": 0.17593838423490524,
      "num_tokens": 19413302.0,
      "step": 9580
    },
    {
      "entropy": 5.9002760887146,
      "epoch": 0.8376299921349296,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049366997467163,
      "loss": 5.541,
      "mean_token_accuracy": 0.1878916174173355,
      "num_tokens": 19423473.0,
      "step": 9585
    },
    {
      "entropy": 5.98995943069458,
      "epoch": 0.8380669404876344,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004936626331336567,
      "loss": 5.6566,
      "mean_token_accuracy": 0.17001886665821075,
      "num_tokens": 19433289.0,
      "step": 9590
    },
    {
      "entropy": 5.979811954498291,
      "epoch": 0.8385038888403391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004936552874016194,
      "loss": 5.6398,
      "mean_token_accuracy": 0.17233089357614517,
      "num_tokens": 19442188.0,
      "step": 9595
    },
    {
      "entropy": 5.901090478897094,
      "epoch": 0.8389408371930438,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936479374756592,
      "loss": 5.6442,
      "mean_token_accuracy": 0.1695869728922844,
      "num_tokens": 19452971.0,
      "step": 9600
    },
    {
      "entropy": 5.9801191806793215,
      "epoch": 0.8393777855457485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936405833559169,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17049231380224228,
      "num_tokens": 19463263.0,
      "step": 9605
    },
    {
      "entropy": 5.9991552352905275,
      "epoch": 0.8398147338984532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936332250425337,
      "loss": 5.6174,
      "mean_token_accuracy": 0.17386815696954727,
      "num_tokens": 19473234.0,
      "step": 9610
    },
    {
      "entropy": 5.87638669013977,
      "epoch": 0.8402516822511579,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936258625356507,
      "loss": 5.616,
      "mean_token_accuracy": 0.17171816080808638,
      "num_tokens": 19483714.0,
      "step": 9615
    },
    {
      "entropy": 5.85729570388794,
      "epoch": 0.8406886306038627,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004936184958354091,
      "loss": 5.5085,
      "mean_token_accuracy": 0.18749085813760757,
      "num_tokens": 19493128.0,
      "step": 9620
    },
    {
      "entropy": 5.970263862609864,
      "epoch": 0.8411255789565674,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00049361112494195,
      "loss": 5.6548,
      "mean_token_accuracy": 0.16689304858446122,
      "num_tokens": 19503401.0,
      "step": 9625
    },
    {
      "entropy": 5.937927961349487,
      "epoch": 0.8415625273092721,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936037498554153,
      "loss": 5.6919,
      "mean_token_accuracy": 0.17494331002235414,
      "num_tokens": 19513050.0,
      "step": 9630
    },
    {
      "entropy": 5.940392875671387,
      "epoch": 0.8419994756619767,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493596370575946,
      "loss": 5.5837,
      "mean_token_accuracy": 0.17768284529447556,
      "num_tokens": 19524534.0,
      "step": 9635
    },
    {
      "entropy": 5.90689377784729,
      "epoch": 0.8424364240146814,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935889871036837,
      "loss": 5.5536,
      "mean_token_accuracy": 0.17877138257026673,
      "num_tokens": 19534863.0,
      "step": 9640
    },
    {
      "entropy": 5.935074424743652,
      "epoch": 0.8428733723673861,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004935815994387701,
      "loss": 5.6499,
      "mean_token_accuracy": 0.1724499613046646,
      "num_tokens": 19545037.0,
      "step": 9645
    },
    {
      "entropy": 5.942518520355224,
      "epoch": 0.8433103207200909,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493574207581347,
      "loss": 5.6231,
      "mean_token_accuracy": 0.1652944415807724,
      "num_tokens": 19555377.0,
      "step": 9650
    },
    {
      "entropy": 5.962093114852905,
      "epoch": 0.8437472690727956,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935668115315559,
      "loss": 5.6332,
      "mean_token_accuracy": 0.17055765241384507,
      "num_tokens": 19565874.0,
      "step": 9655
    },
    {
      "entropy": 5.797784948348999,
      "epoch": 0.8441842174255003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493559411289539,
      "loss": 5.4761,
      "mean_token_accuracy": 0.19141393899917603,
      "num_tokens": 19575002.0,
      "step": 9660
    },
    {
      "entropy": 5.887431621551514,
      "epoch": 0.844621165778205,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493552006855438,
      "loss": 5.5785,
      "mean_token_accuracy": 0.17952554374933244,
      "num_tokens": 19585863.0,
      "step": 9665
    },
    {
      "entropy": 5.922807884216309,
      "epoch": 0.8450581141309097,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935445982293949,
      "loss": 5.5453,
      "mean_token_accuracy": 0.17589958608150483,
      "num_tokens": 19596787.0,
      "step": 9670
    },
    {
      "entropy": 5.895658731460571,
      "epoch": 0.8454950624836144,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935371854115519,
      "loss": 5.6489,
      "mean_token_accuracy": 0.1712544620037079,
      "num_tokens": 19607114.0,
      "step": 9675
    },
    {
      "entropy": 5.903927755355835,
      "epoch": 0.8459320108363192,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935297684020513,
      "loss": 5.5195,
      "mean_token_accuracy": 0.1887320190668106,
      "num_tokens": 19617334.0,
      "step": 9680
    },
    {
      "entropy": 5.928301239013672,
      "epoch": 0.8463689591890239,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935223472010352,
      "loss": 5.6171,
      "mean_token_accuracy": 0.17529973834753038,
      "num_tokens": 19626844.0,
      "step": 9685
    },
    {
      "entropy": 5.889812040328979,
      "epoch": 0.8468059075417286,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935149218086459,
      "loss": 5.605,
      "mean_token_accuracy": 0.17277255207300185,
      "num_tokens": 19637897.0,
      "step": 9690
    },
    {
      "entropy": 5.9307637214660645,
      "epoch": 0.8472428558944333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493507492225026,
      "loss": 5.676,
      "mean_token_accuracy": 0.17092891186475753,
      "num_tokens": 19649161.0,
      "step": 9695
    },
    {
      "entropy": 5.977705240249634,
      "epoch": 0.847679804247138,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004935000584503178,
      "loss": 5.5823,
      "mean_token_accuracy": 0.17431258261203766,
      "num_tokens": 19659328.0,
      "step": 9700
    },
    {
      "entropy": 5.895052337646485,
      "epoch": 0.8481167525998426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934926204846641,
      "loss": 5.6687,
      "mean_token_accuracy": 0.1724779948592186,
      "num_tokens": 19669196.0,
      "step": 9705
    },
    {
      "entropy": 5.893037557601929,
      "epoch": 0.8485537009525475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934851783282073,
      "loss": 5.5579,
      "mean_token_accuracy": 0.17454802989959717,
      "num_tokens": 19680019.0,
      "step": 9710
    },
    {
      "entropy": 5.936587333679199,
      "epoch": 0.8489906493052521,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934777319810904,
      "loss": 5.6787,
      "mean_token_accuracy": 0.1752319648861885,
      "num_tokens": 19689391.0,
      "step": 9715
    },
    {
      "entropy": 5.974117422103882,
      "epoch": 0.8494275976579568,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934702814434561,
      "loss": 5.6553,
      "mean_token_accuracy": 0.17137505263090133,
      "num_tokens": 19699683.0,
      "step": 9720
    },
    {
      "entropy": 5.994554281234741,
      "epoch": 0.8498645460106615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934628267154474,
      "loss": 5.7057,
      "mean_token_accuracy": 0.16390338838100432,
      "num_tokens": 19709844.0,
      "step": 9725
    },
    {
      "entropy": 6.004380226135254,
      "epoch": 0.8503014943633662,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493455367797207,
      "loss": 5.6405,
      "mean_token_accuracy": 0.17020069509744645,
      "num_tokens": 19719926.0,
      "step": 9730
    },
    {
      "entropy": 5.91009292602539,
      "epoch": 0.8507384427160709,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004934479046888784,
      "loss": 5.5883,
      "mean_token_accuracy": 0.17239587381482124,
      "num_tokens": 19730548.0,
      "step": 9735
    },
    {
      "entropy": 5.960269165039063,
      "epoch": 0.8511753910687757,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004934404373906045,
      "loss": 5.6435,
      "mean_token_accuracy": 0.1743390753865242,
      "num_tokens": 19740471.0,
      "step": 9740
    },
    {
      "entropy": 5.950755167007446,
      "epoch": 0.8516123394214804,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004934329659025283,
      "loss": 5.5647,
      "mean_token_accuracy": 0.1792108103632927,
      "num_tokens": 19749395.0,
      "step": 9745
    },
    {
      "entropy": 5.9147051811218265,
      "epoch": 0.8520492877741851,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934254902247936,
      "loss": 5.657,
      "mean_token_accuracy": 0.1680002048611641,
      "num_tokens": 19759203.0,
      "step": 9750
    },
    {
      "entropy": 5.959056997299195,
      "epoch": 0.8524862361268898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934180103575436,
      "loss": 5.5048,
      "mean_token_accuracy": 0.1852800339460373,
      "num_tokens": 19768977.0,
      "step": 9755
    },
    {
      "entropy": 5.893286085128784,
      "epoch": 0.8529231844795945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934105263009216,
      "loss": 5.5157,
      "mean_token_accuracy": 0.17823946177959443,
      "num_tokens": 19778569.0,
      "step": 9760
    },
    {
      "entropy": 5.906665468215943,
      "epoch": 0.8533601328322992,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004934030380550713,
      "loss": 5.6217,
      "mean_token_accuracy": 0.17958581447601318,
      "num_tokens": 19787939.0,
      "step": 9765
    },
    {
      "entropy": 5.892000389099121,
      "epoch": 0.853797081185004,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004933955456201362,
      "loss": 5.5888,
      "mean_token_accuracy": 0.17416672110557557,
      "num_tokens": 19798407.0,
      "step": 9770
    },
    {
      "entropy": 5.976316738128662,
      "epoch": 0.8542340295377087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933880489962602,
      "loss": 5.6339,
      "mean_token_accuracy": 0.17542823553085327,
      "num_tokens": 19808495.0,
      "step": 9775
    },
    {
      "entropy": 5.875558710098266,
      "epoch": 0.8546709778904134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004933805481835868,
      "loss": 5.5606,
      "mean_token_accuracy": 0.17470662593841552,
      "num_tokens": 19818709.0,
      "step": 9780
    },
    {
      "entropy": 5.936044788360595,
      "epoch": 0.855107926243118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933730431822603,
      "loss": 5.6508,
      "mean_token_accuracy": 0.17628105133771896,
      "num_tokens": 19828865.0,
      "step": 9785
    },
    {
      "entropy": 5.946932411193847,
      "epoch": 0.8555448745958227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933655339924243,
      "loss": 5.5654,
      "mean_token_accuracy": 0.1740651398897171,
      "num_tokens": 19838757.0,
      "step": 9790
    },
    {
      "entropy": 5.838632249832154,
      "epoch": 0.8559818229485275,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493358020614223,
      "loss": 5.6151,
      "mean_token_accuracy": 0.17729836106300353,
      "num_tokens": 19848967.0,
      "step": 9795
    },
    {
      "entropy": 5.966057205200196,
      "epoch": 0.8564187713012322,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004933505030478003,
      "loss": 5.6233,
      "mean_token_accuracy": 0.17819051593542098,
      "num_tokens": 19859487.0,
      "step": 9800
    },
    {
      "entropy": 5.968180322647095,
      "epoch": 0.8568557196539369,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933429812933005,
      "loss": 5.6842,
      "mean_token_accuracy": 0.17039379924535752,
      "num_tokens": 19870170.0,
      "step": 9805
    },
    {
      "entropy": 5.868817281723023,
      "epoch": 0.8572926680066416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933354553508681,
      "loss": 5.5621,
      "mean_token_accuracy": 0.1769936889410019,
      "num_tokens": 19881031.0,
      "step": 9810
    },
    {
      "entropy": 5.932675695419311,
      "epoch": 0.8577296163593463,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004933279252206472,
      "loss": 5.5977,
      "mean_token_accuracy": 0.17033511847257615,
      "num_tokens": 19891082.0,
      "step": 9815
    },
    {
      "entropy": 5.9507276058197025,
      "epoch": 0.858166564712051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004933203909027823,
      "loss": 5.5905,
      "mean_token_accuracy": 0.17416820228099822,
      "num_tokens": 19900626.0,
      "step": 9820
    },
    {
      "entropy": 5.86556487083435,
      "epoch": 0.8586035130647558,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933128523974179,
      "loss": 5.5701,
      "mean_token_accuracy": 0.18434133529663085,
      "num_tokens": 19910660.0,
      "step": 9825
    },
    {
      "entropy": 5.900490713119507,
      "epoch": 0.8590404614174605,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933053097046986,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17775578498840333,
      "num_tokens": 19920476.0,
      "step": 9830
    },
    {
      "entropy": 6.026728296279908,
      "epoch": 0.8594774097701652,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493297762824769,
      "loss": 5.6922,
      "mean_token_accuracy": 0.1648374617099762,
      "num_tokens": 19931151.0,
      "step": 9835
    },
    {
      "entropy": 5.912132692337036,
      "epoch": 0.8599143581228699,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493290211757774,
      "loss": 5.5799,
      "mean_token_accuracy": 0.17368682771921157,
      "num_tokens": 19942469.0,
      "step": 9840
    },
    {
      "entropy": 5.8857168674469,
      "epoch": 0.8603513064755746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004932826565038584,
      "loss": 5.6467,
      "mean_token_accuracy": 0.17525100111961364,
      "num_tokens": 19953055.0,
      "step": 9845
    },
    {
      "entropy": 5.869904470443726,
      "epoch": 0.8607882548282793,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004932750970631669,
      "loss": 5.6129,
      "mean_token_accuracy": 0.17792824804782867,
      "num_tokens": 19963944.0,
      "step": 9850
    },
    {
      "entropy": 5.893194103240967,
      "epoch": 0.8612252031809841,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932675334358448,
      "loss": 5.5306,
      "mean_token_accuracy": 0.17732472270727156,
      "num_tokens": 19973913.0,
      "step": 9855
    },
    {
      "entropy": 5.886653661727905,
      "epoch": 0.8616621515336887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493259965622037,
      "loss": 5.5891,
      "mean_token_accuracy": 0.17764730900526046,
      "num_tokens": 19984130.0,
      "step": 9860
    },
    {
      "entropy": 5.9153646469116214,
      "epoch": 0.8620990998863934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004932523936218887,
      "loss": 5.518,
      "mean_token_accuracy": 0.1798068627715111,
      "num_tokens": 19993840.0,
      "step": 9865
    },
    {
      "entropy": 5.874785470962524,
      "epoch": 0.8625360482390981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493244817435545,
      "loss": 5.6066,
      "mean_token_accuracy": 0.17545456141233445,
      "num_tokens": 20004228.0,
      "step": 9870
    },
    {
      "entropy": 5.919204998016357,
      "epoch": 0.8629729965918028,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004932372370631515,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1789783239364624,
      "num_tokens": 20012620.0,
      "step": 9875
    },
    {
      "entropy": 5.9102904319763185,
      "epoch": 0.8634099449445075,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932296525048535,
      "loss": 5.6863,
      "mean_token_accuracy": 0.17338194847106933,
      "num_tokens": 20022892.0,
      "step": 9880
    },
    {
      "entropy": 5.909966564178466,
      "epoch": 0.8638468932972123,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004932220637607962,
      "loss": 5.5511,
      "mean_token_accuracy": 0.18368324339389802,
      "num_tokens": 20032788.0,
      "step": 9885
    },
    {
      "entropy": 5.8619097709655765,
      "epoch": 0.864283841649917,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004932144708311256,
      "loss": 5.5997,
      "mean_token_accuracy": 0.1773289680480957,
      "num_tokens": 20043176.0,
      "step": 9890
    },
    {
      "entropy": 5.9188567161560055,
      "epoch": 0.8647207900026217,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000493206873715987,
      "loss": 5.5946,
      "mean_token_accuracy": 0.17481564730405807,
      "num_tokens": 20052298.0,
      "step": 9895
    },
    {
      "entropy": 5.9084343910217285,
      "epoch": 0.8651577383553264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004931992724155262,
      "loss": 5.5432,
      "mean_token_accuracy": 0.17618019431829451,
      "num_tokens": 20063163.0,
      "step": 9900
    },
    {
      "entropy": 5.951395320892334,
      "epoch": 0.8655946867080311,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493191666929889,
      "loss": 5.6257,
      "mean_token_accuracy": 0.17645299285650254,
      "num_tokens": 20073644.0,
      "step": 9905
    },
    {
      "entropy": 5.9283952713012695,
      "epoch": 0.8660316350607358,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931840572592214,
      "loss": 5.5616,
      "mean_token_accuracy": 0.17878945022821427,
      "num_tokens": 20084457.0,
      "step": 9910
    },
    {
      "entropy": 5.8439257621765135,
      "epoch": 0.8664685834134406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931764434036692,
      "loss": 5.4623,
      "mean_token_accuracy": 0.18516533821821213,
      "num_tokens": 20094153.0,
      "step": 9915
    },
    {
      "entropy": 5.8700785636901855,
      "epoch": 0.8669055317661453,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004931688253633784,
      "loss": 5.6604,
      "mean_token_accuracy": 0.1716906577348709,
      "num_tokens": 20104031.0,
      "step": 9920
    },
    {
      "entropy": 5.925810146331787,
      "epoch": 0.86734248011885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931612031384953,
      "loss": 5.5756,
      "mean_token_accuracy": 0.18125031441450118,
      "num_tokens": 20114431.0,
      "step": 9925
    },
    {
      "entropy": 5.861438751220703,
      "epoch": 0.8677794284715546,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493153576729166,
      "loss": 5.5045,
      "mean_token_accuracy": 0.18025831431150435,
      "num_tokens": 20124161.0,
      "step": 9930
    },
    {
      "entropy": 5.890688037872314,
      "epoch": 0.8682163768242593,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004931459461355368,
      "loss": 5.5418,
      "mean_token_accuracy": 0.1832140177488327,
      "num_tokens": 20134227.0,
      "step": 9935
    },
    {
      "entropy": 5.8961992263793945,
      "epoch": 0.868653325176964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931383113577541,
      "loss": 5.5616,
      "mean_token_accuracy": 0.17609407305717467,
      "num_tokens": 20143912.0,
      "step": 9940
    },
    {
      "entropy": 5.871687126159668,
      "epoch": 0.8690902735296688,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004931306723959641,
      "loss": 5.4895,
      "mean_token_accuracy": 0.18133915066719056,
      "num_tokens": 20153858.0,
      "step": 9945
    },
    {
      "entropy": 5.973691844940186,
      "epoch": 0.8695272218823735,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004931230292503137,
      "loss": 5.609,
      "mean_token_accuracy": 0.1737261101603508,
      "num_tokens": 20164309.0,
      "step": 9950
    },
    {
      "entropy": 5.893842983245849,
      "epoch": 0.8699641702350782,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004931153819209492,
      "loss": 5.591,
      "mean_token_accuracy": 0.1796587735414505,
      "num_tokens": 20173949.0,
      "step": 9955
    },
    {
      "entropy": 5.90356969833374,
      "epoch": 0.8704011185877829,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931077304080175,
      "loss": 5.5738,
      "mean_token_accuracy": 0.17333794385194778,
      "num_tokens": 20184206.0,
      "step": 9960
    },
    {
      "entropy": 6.005241537094117,
      "epoch": 0.8708380669404876,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004931000747116651,
      "loss": 5.6892,
      "mean_token_accuracy": 0.1690976068377495,
      "num_tokens": 20194430.0,
      "step": 9965
    },
    {
      "entropy": 5.986863470077514,
      "epoch": 0.8712750152931923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493092414832039,
      "loss": 5.5499,
      "mean_token_accuracy": 0.17988206446170807,
      "num_tokens": 20205004.0,
      "step": 9970
    },
    {
      "entropy": 5.852977561950683,
      "epoch": 0.8717119636458971,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004930847507692862,
      "loss": 5.5259,
      "mean_token_accuracy": 0.17615356296300888,
      "num_tokens": 20214698.0,
      "step": 9975
    },
    {
      "entropy": 5.885163688659668,
      "epoch": 0.8721489119986018,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004930770825235536,
      "loss": 5.6101,
      "mean_token_accuracy": 0.17292344719171523,
      "num_tokens": 20226534.0,
      "step": 9980
    },
    {
      "entropy": 5.966446781158448,
      "epoch": 0.8725858603513065,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004930694100949882,
      "loss": 5.5989,
      "mean_token_accuracy": 0.17275148630142212,
      "num_tokens": 20237414.0,
      "step": 9985
    },
    {
      "entropy": 5.876318645477295,
      "epoch": 0.8730228087040112,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930617334837373,
      "loss": 5.4465,
      "mean_token_accuracy": 0.18286702185869216,
      "num_tokens": 20246948.0,
      "step": 9990
    },
    {
      "entropy": 5.887368488311767,
      "epoch": 0.8734597570567159,
      "grad_norm": 1.0,
      "learning_rate": 0.000493054052689948,
      "loss": 5.6701,
      "mean_token_accuracy": 0.17102539986371995,
      "num_tokens": 20257281.0,
      "step": 9995
    },
    {
      "entropy": 5.897332382202149,
      "epoch": 0.8738967054094207,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004930463677137679,
      "loss": 5.6396,
      "mean_token_accuracy": 0.1718030408024788,
      "num_tokens": 20269041.0,
      "step": 10000
    },
    {
      "entropy": 5.937284135818482,
      "epoch": 0.8743336537621254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004930386785553441,
      "loss": 5.6772,
      "mean_token_accuracy": 0.1677208811044693,
      "num_tokens": 20279688.0,
      "step": 10005
    },
    {
      "entropy": 5.9362547397613525,
      "epoch": 0.87477060211483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004930309852148242,
      "loss": 5.5902,
      "mean_token_accuracy": 0.17964806407690048,
      "num_tokens": 20288913.0,
      "step": 10010
    },
    {
      "entropy": 5.95678825378418,
      "epoch": 0.8752075504675347,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004930232876923558,
      "loss": 5.6122,
      "mean_token_accuracy": 0.1773628756403923,
      "num_tokens": 20299009.0,
      "step": 10015
    },
    {
      "entropy": 5.902712631225586,
      "epoch": 0.8756444988202394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930155859880865,
      "loss": 5.5946,
      "mean_token_accuracy": 0.17121389210224153,
      "num_tokens": 20309478.0,
      "step": 10020
    },
    {
      "entropy": 5.92737603187561,
      "epoch": 0.8760814471729441,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493007880102164,
      "loss": 5.6158,
      "mean_token_accuracy": 0.17546933591365815,
      "num_tokens": 20319451.0,
      "step": 10025
    },
    {
      "entropy": 5.928910779953003,
      "epoch": 0.8765183955256489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493000170034736,
      "loss": 5.5718,
      "mean_token_accuracy": 0.17806826233863832,
      "num_tokens": 20328497.0,
      "step": 10030
    },
    {
      "entropy": 5.855160665512085,
      "epoch": 0.8769553438783536,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929924557859507,
      "loss": 5.4871,
      "mean_token_accuracy": 0.19004450738430023,
      "num_tokens": 20338584.0,
      "step": 10035
    },
    {
      "entropy": 5.888304853439331,
      "epoch": 0.8773922922310583,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929847373559557,
      "loss": 5.5597,
      "mean_token_accuracy": 0.17796892672777176,
      "num_tokens": 20348808.0,
      "step": 10040
    },
    {
      "entropy": 5.867915344238281,
      "epoch": 0.877829240583763,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929770147448992,
      "loss": 5.5547,
      "mean_token_accuracy": 0.17686384469270705,
      "num_tokens": 20358301.0,
      "step": 10045
    },
    {
      "entropy": 5.896743249893189,
      "epoch": 0.8782661889364677,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929692879529295,
      "loss": 5.5539,
      "mean_token_accuracy": 0.17456185221672058,
      "num_tokens": 20369032.0,
      "step": 10050
    },
    {
      "entropy": 5.922493600845337,
      "epoch": 0.8787031372891724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929615569801946,
      "loss": 5.5957,
      "mean_token_accuracy": 0.17627866864204406,
      "num_tokens": 20379101.0,
      "step": 10055
    },
    {
      "entropy": 5.862710046768188,
      "epoch": 0.8791400856418772,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929538218268428,
      "loss": 5.5554,
      "mean_token_accuracy": 0.18122673481702806,
      "num_tokens": 20388864.0,
      "step": 10060
    },
    {
      "entropy": 5.999557781219482,
      "epoch": 0.8795770339945819,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929460824930226,
      "loss": 5.6766,
      "mean_token_accuracy": 0.17017377018928528,
      "num_tokens": 20398402.0,
      "step": 10065
    },
    {
      "entropy": 5.830965042114258,
      "epoch": 0.8800139823472866,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929383389788821,
      "loss": 5.5606,
      "mean_token_accuracy": 0.17866561859846114,
      "num_tokens": 20408209.0,
      "step": 10070
    },
    {
      "entropy": 5.890618133544922,
      "epoch": 0.8804509306999913,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929305912845703,
      "loss": 5.5971,
      "mean_token_accuracy": 0.17705920189619065,
      "num_tokens": 20418117.0,
      "step": 10075
    },
    {
      "entropy": 6.02279691696167,
      "epoch": 0.8808878790526959,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004929228394102355,
      "loss": 5.6254,
      "mean_token_accuracy": 0.17277174890041352,
      "num_tokens": 20428347.0,
      "step": 10080
    },
    {
      "entropy": 5.972275829315185,
      "epoch": 0.8813248274054006,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004929150833560263,
      "loss": 5.5662,
      "mean_token_accuracy": 0.17821199148893357,
      "num_tokens": 20437804.0,
      "step": 10085
    },
    {
      "entropy": 5.947888469696045,
      "epoch": 0.8817617757581054,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929073231220917,
      "loss": 5.6026,
      "mean_token_accuracy": 0.16869329512119294,
      "num_tokens": 20447600.0,
      "step": 10090
    },
    {
      "entropy": 5.9611570835113525,
      "epoch": 0.8821987241108101,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928995587085806,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1756485030055046,
      "num_tokens": 20458023.0,
      "step": 10095
    },
    {
      "entropy": 5.930154705047608,
      "epoch": 0.8826356724635148,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928917901156416,
      "loss": 5.5445,
      "mean_token_accuracy": 0.18173832446336746,
      "num_tokens": 20468195.0,
      "step": 10100
    },
    {
      "entropy": 5.895699501037598,
      "epoch": 0.8830726208162195,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492884017343424,
      "loss": 5.6702,
      "mean_token_accuracy": 0.17013559192419053,
      "num_tokens": 20479250.0,
      "step": 10105
    },
    {
      "entropy": 5.937417268753052,
      "epoch": 0.8835095691689242,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004928762403920767,
      "loss": 5.5766,
      "mean_token_accuracy": 0.17519285380840302,
      "num_tokens": 20489587.0,
      "step": 10110
    },
    {
      "entropy": 5.9462944030761715,
      "epoch": 0.8839465175216289,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492868459261749,
      "loss": 5.5754,
      "mean_token_accuracy": 0.17869874238967895,
      "num_tokens": 20499788.0,
      "step": 10115
    },
    {
      "entropy": 5.887098598480224,
      "epoch": 0.8843834658743337,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049286067395259,
      "loss": 5.6118,
      "mean_token_accuracy": 0.17397077828645707,
      "num_tokens": 20509716.0,
      "step": 10120
    },
    {
      "entropy": 5.903341150283813,
      "epoch": 0.8848204142270384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928528844647493,
      "loss": 5.5483,
      "mean_token_accuracy": 0.17645130604505538,
      "num_tokens": 20519020.0,
      "step": 10125
    },
    {
      "entropy": 5.919848918914795,
      "epoch": 0.8852573625797431,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492845090798376,
      "loss": 5.6164,
      "mean_token_accuracy": 0.17449308037757874,
      "num_tokens": 20528997.0,
      "step": 10130
    },
    {
      "entropy": 5.912308263778686,
      "epoch": 0.8856943109324478,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928372929536197,
      "loss": 5.6149,
      "mean_token_accuracy": 0.1784501627087593,
      "num_tokens": 20538599.0,
      "step": 10135
    },
    {
      "entropy": 6.0072098731994625,
      "epoch": 0.8861312592851525,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928294909306301,
      "loss": 5.5488,
      "mean_token_accuracy": 0.17926603853702544,
      "num_tokens": 20547664.0,
      "step": 10140
    },
    {
      "entropy": 5.94606580734253,
      "epoch": 0.8865682076378572,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928216847295565,
      "loss": 5.6669,
      "mean_token_accuracy": 0.17260199338197707,
      "num_tokens": 20557860.0,
      "step": 10145
    },
    {
      "entropy": 5.873510074615479,
      "epoch": 0.887005155990562,
      "grad_norm": 0.953125,
      "learning_rate": 0.000492813874350549,
      "loss": 5.5975,
      "mean_token_accuracy": 0.17570080310106279,
      "num_tokens": 20569003.0,
      "step": 10150
    },
    {
      "entropy": 5.974510669708252,
      "epoch": 0.8874421043432666,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004928060597937573,
      "loss": 5.5775,
      "mean_token_accuracy": 0.17916307151317595,
      "num_tokens": 20578402.0,
      "step": 10155
    },
    {
      "entropy": 5.916538667678833,
      "epoch": 0.8878790526959713,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927982410593312,
      "loss": 5.4888,
      "mean_token_accuracy": 0.18262031376361848,
      "num_tokens": 20587874.0,
      "step": 10160
    },
    {
      "entropy": 5.898109722137451,
      "epoch": 0.888316001048676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927904181474207,
      "loss": 5.5231,
      "mean_token_accuracy": 0.17661795914173126,
      "num_tokens": 20597270.0,
      "step": 10165
    },
    {
      "entropy": 5.929809474945069,
      "epoch": 0.8887529494013807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927825910581758,
      "loss": 5.6334,
      "mean_token_accuracy": 0.17567212432622908,
      "num_tokens": 20607898.0,
      "step": 10170
    },
    {
      "entropy": 5.93554949760437,
      "epoch": 0.8891898977540854,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927747597917467,
      "loss": 5.5303,
      "mean_token_accuracy": 0.1854040026664734,
      "num_tokens": 20617300.0,
      "step": 10175
    },
    {
      "entropy": 5.917754030227661,
      "epoch": 0.8896268461067902,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927669243482836,
      "loss": 5.6379,
      "mean_token_accuracy": 0.17308132946491242,
      "num_tokens": 20626582.0,
      "step": 10180
    },
    {
      "entropy": 5.880548667907715,
      "epoch": 0.8900637944594949,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927590847279369,
      "loss": 5.5642,
      "mean_token_accuracy": 0.17716656625270844,
      "num_tokens": 20635802.0,
      "step": 10185
    },
    {
      "entropy": 5.93758602142334,
      "epoch": 0.8905007428121996,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927512409308568,
      "loss": 5.5709,
      "mean_token_accuracy": 0.18241159170866011,
      "num_tokens": 20646057.0,
      "step": 10190
    },
    {
      "entropy": 5.944963121414185,
      "epoch": 0.8909376911649043,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004927433929571937,
      "loss": 5.6744,
      "mean_token_accuracy": 0.16531140357255936,
      "num_tokens": 20656511.0,
      "step": 10195
    },
    {
      "entropy": 5.939849853515625,
      "epoch": 0.891374639517609,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927355408070982,
      "loss": 5.5727,
      "mean_token_accuracy": 0.18208911269903183,
      "num_tokens": 20666479.0,
      "step": 10200
    },
    {
      "entropy": 5.891167688369751,
      "epoch": 0.8918115878703138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004927276844807211,
      "loss": 5.5034,
      "mean_token_accuracy": 0.18244811296463012,
      "num_tokens": 20676830.0,
      "step": 10205
    },
    {
      "entropy": 5.854977369308472,
      "epoch": 0.8922485362230185,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927198239782129,
      "loss": 5.5501,
      "mean_token_accuracy": 0.17880351543426515,
      "num_tokens": 20687316.0,
      "step": 10210
    },
    {
      "entropy": 5.9428058624267575,
      "epoch": 0.8926854845757232,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927119592997242,
      "loss": 5.6257,
      "mean_token_accuracy": 0.1723097175359726,
      "num_tokens": 20697901.0,
      "step": 10215
    },
    {
      "entropy": 5.914193868637085,
      "epoch": 0.8931224329284279,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004927040904454062,
      "loss": 5.5444,
      "mean_token_accuracy": 0.1796039015054703,
      "num_tokens": 20707468.0,
      "step": 10220
    },
    {
      "entropy": 5.889121961593628,
      "epoch": 0.8935593812811325,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004926962174154097,
      "loss": 5.5919,
      "mean_token_accuracy": 0.17720798999071122,
      "num_tokens": 20717098.0,
      "step": 10225
    },
    {
      "entropy": 5.998091697692871,
      "epoch": 0.8939963296338372,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926883402098857,
      "loss": 5.6863,
      "mean_token_accuracy": 0.16581256687641144,
      "num_tokens": 20727217.0,
      "step": 10230
    },
    {
      "entropy": 5.959588813781738,
      "epoch": 0.894433277986542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926804588289852,
      "loss": 5.6581,
      "mean_token_accuracy": 0.17077122032642364,
      "num_tokens": 20737397.0,
      "step": 10235
    },
    {
      "entropy": 5.809716987609863,
      "epoch": 0.8948702263392467,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926725732728595,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18513499200344086,
      "num_tokens": 20746712.0,
      "step": 10240
    },
    {
      "entropy": 5.909274911880493,
      "epoch": 0.8953071746919514,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926646835416597,
      "loss": 5.608,
      "mean_token_accuracy": 0.17638629376888276,
      "num_tokens": 20757161.0,
      "step": 10245
    },
    {
      "entropy": 5.9238685131072994,
      "epoch": 0.8957441230446561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004926567896355372,
      "loss": 5.5077,
      "mean_token_accuracy": 0.1893069013953209,
      "num_tokens": 20766478.0,
      "step": 10250
    },
    {
      "entropy": 5.858358669281006,
      "epoch": 0.8961810713973608,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926488915546435,
      "loss": 5.5606,
      "mean_token_accuracy": 0.18310255706310272,
      "num_tokens": 20776317.0,
      "step": 10255
    },
    {
      "entropy": 5.8967273235321045,
      "epoch": 0.8966180197500655,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049264098929913,
      "loss": 5.493,
      "mean_token_accuracy": 0.1837944746017456,
      "num_tokens": 20786534.0,
      "step": 10260
    },
    {
      "entropy": 5.881051969528198,
      "epoch": 0.8970549681027703,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004926330828691482,
      "loss": 5.5218,
      "mean_token_accuracy": 0.17565801590681077,
      "num_tokens": 20796791.0,
      "step": 10265
    },
    {
      "entropy": 5.924180746078491,
      "epoch": 0.897491916455475,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004926251722648499,
      "loss": 5.5452,
      "mean_token_accuracy": 0.1799195557832718,
      "num_tokens": 20806834.0,
      "step": 10270
    },
    {
      "entropy": 5.820607900619507,
      "epoch": 0.8979288648081797,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004926172574863868,
      "loss": 5.5567,
      "mean_token_accuracy": 0.17625196576118468,
      "num_tokens": 20816664.0,
      "step": 10275
    },
    {
      "entropy": 5.894027996063232,
      "epoch": 0.8983658131608844,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926093385339105,
      "loss": 5.678,
      "mean_token_accuracy": 0.17128410339355468,
      "num_tokens": 20826936.0,
      "step": 10280
    },
    {
      "entropy": 5.934953069686889,
      "epoch": 0.8988027615135891,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004926014154075732,
      "loss": 5.5394,
      "mean_token_accuracy": 0.18017439544200897,
      "num_tokens": 20836732.0,
      "step": 10285
    },
    {
      "entropy": 5.931967687606812,
      "epoch": 0.8992397098662938,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004925934881075268,
      "loss": 5.5739,
      "mean_token_accuracy": 0.1818123623728752,
      "num_tokens": 20848093.0,
      "step": 10290
    },
    {
      "entropy": 5.925895929336548,
      "epoch": 0.8996766582189986,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925855566339232,
      "loss": 5.546,
      "mean_token_accuracy": 0.17547869831323623,
      "num_tokens": 20858678.0,
      "step": 10295
    },
    {
      "entropy": 5.8717084407806395,
      "epoch": 0.9001136065717033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925776209869146,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18107181936502456,
      "num_tokens": 20869319.0,
      "step": 10300
    },
    {
      "entropy": 5.8794549942016605,
      "epoch": 0.9005505549244079,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925696811666532,
      "loss": 5.5056,
      "mean_token_accuracy": 0.18415997326374053,
      "num_tokens": 20879475.0,
      "step": 10305
    },
    {
      "entropy": 5.882885360717774,
      "epoch": 0.9009875032771126,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004925617371732914,
      "loss": 5.5418,
      "mean_token_accuracy": 0.18497056365013123,
      "num_tokens": 20889479.0,
      "step": 10310
    },
    {
      "entropy": 5.906025838851929,
      "epoch": 0.9014244516298173,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004925537890069815,
      "loss": 5.6223,
      "mean_token_accuracy": 0.17553223967552184,
      "num_tokens": 20899864.0,
      "step": 10315
    },
    {
      "entropy": 5.948844242095947,
      "epoch": 0.901861399982522,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004925458366678758,
      "loss": 5.5899,
      "mean_token_accuracy": 0.17915809005498887,
      "num_tokens": 20909006.0,
      "step": 10320
    },
    {
      "entropy": 5.921598052978515,
      "epoch": 0.9022983483352268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004925378801561269,
      "loss": 5.5472,
      "mean_token_accuracy": 0.17831066995859146,
      "num_tokens": 20917998.0,
      "step": 10325
    },
    {
      "entropy": 5.838485813140869,
      "epoch": 0.9027352966879315,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925299194718876,
      "loss": 5.5589,
      "mean_token_accuracy": 0.18051585853099822,
      "num_tokens": 20928407.0,
      "step": 10330
    },
    {
      "entropy": 5.845799112319947,
      "epoch": 0.9031722450406362,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925219546153105,
      "loss": 5.5793,
      "mean_token_accuracy": 0.17675303667783737,
      "num_tokens": 20938016.0,
      "step": 10335
    },
    {
      "entropy": 5.954553413391113,
      "epoch": 0.9036091933933409,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925139855865483,
      "loss": 5.6304,
      "mean_token_accuracy": 0.16834931969642639,
      "num_tokens": 20948578.0,
      "step": 10340
    },
    {
      "entropy": 5.93409161567688,
      "epoch": 0.9040461417460456,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004925060123857539,
      "loss": 5.4885,
      "mean_token_accuracy": 0.17862484455108643,
      "num_tokens": 20958193.0,
      "step": 10345
    },
    {
      "entropy": 5.896431875228882,
      "epoch": 0.9044830900987503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004924980350130802,
      "loss": 5.5518,
      "mean_token_accuracy": 0.1863422453403473,
      "num_tokens": 20968092.0,
      "step": 10350
    },
    {
      "entropy": 5.875405931472779,
      "epoch": 0.9049200384514551,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004924900534686801,
      "loss": 5.5751,
      "mean_token_accuracy": 0.17294687032699585,
      "num_tokens": 20978303.0,
      "step": 10355
    },
    {
      "entropy": 5.989206743240357,
      "epoch": 0.9053569868041598,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004924820677527069,
      "loss": 5.5824,
      "mean_token_accuracy": 0.177727872133255,
      "num_tokens": 20988331.0,
      "step": 10360
    },
    {
      "entropy": 5.83052659034729,
      "epoch": 0.9057939351568645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924740778653137,
      "loss": 5.5032,
      "mean_token_accuracy": 0.17685336917638778,
      "num_tokens": 20997785.0,
      "step": 10365
    },
    {
      "entropy": 5.911478328704834,
      "epoch": 0.9062308835095692,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924660838066538,
      "loss": 5.6092,
      "mean_token_accuracy": 0.17738166749477385,
      "num_tokens": 21008026.0,
      "step": 10370
    },
    {
      "entropy": 5.918211460113525,
      "epoch": 0.9066678318622738,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924580855768803,
      "loss": 5.564,
      "mean_token_accuracy": 0.18100535720586777,
      "num_tokens": 21017993.0,
      "step": 10375
    },
    {
      "entropy": 5.890488576889038,
      "epoch": 0.9071047802149785,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004924500831761468,
      "loss": 5.5663,
      "mean_token_accuracy": 0.18100601434707642,
      "num_tokens": 21028149.0,
      "step": 10380
    },
    {
      "entropy": 5.986132574081421,
      "epoch": 0.9075417285676833,
      "grad_norm": 0.875,
      "learning_rate": 0.0004924420766046069,
      "loss": 5.6656,
      "mean_token_accuracy": 0.16899911165237427,
      "num_tokens": 21039229.0,
      "step": 10385
    },
    {
      "entropy": 5.9748475551605225,
      "epoch": 0.907978676920388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004924340658624139,
      "loss": 5.6031,
      "mean_token_accuracy": 0.17533838450908662,
      "num_tokens": 21049358.0,
      "step": 10390
    },
    {
      "entropy": 5.933449792861938,
      "epoch": 0.9084156252730927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004924260509497217,
      "loss": 5.629,
      "mean_token_accuracy": 0.17539625167846679,
      "num_tokens": 21060385.0,
      "step": 10395
    },
    {
      "entropy": 5.947511911392212,
      "epoch": 0.9088525736257974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004924180318666838,
      "loss": 5.6348,
      "mean_token_accuracy": 0.17260485887527466,
      "num_tokens": 21070423.0,
      "step": 10400
    },
    {
      "entropy": 6.017297172546387,
      "epoch": 0.9092895219785021,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924100086134541,
      "loss": 5.6055,
      "mean_token_accuracy": 0.1788026824593544,
      "num_tokens": 21080505.0,
      "step": 10405
    },
    {
      "entropy": 5.938314056396484,
      "epoch": 0.9097264703312069,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004924019811901866,
      "loss": 5.5854,
      "mean_token_accuracy": 0.17371623665094377,
      "num_tokens": 21090875.0,
      "step": 10410
    },
    {
      "entropy": 5.894816780090332,
      "epoch": 0.9101634186839116,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923939495970352,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1797473683953285,
      "num_tokens": 21101427.0,
      "step": 10415
    },
    {
      "entropy": 5.935783195495605,
      "epoch": 0.9106003670366163,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004923859138341539,
      "loss": 5.6619,
      "mean_token_accuracy": 0.1719607576727867,
      "num_tokens": 21112691.0,
      "step": 10420
    },
    {
      "entropy": 5.97493577003479,
      "epoch": 0.911037315389321,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923778739016969,
      "loss": 5.6464,
      "mean_token_accuracy": 0.16765112429857254,
      "num_tokens": 21123932.0,
      "step": 10425
    },
    {
      "entropy": 5.988424301147461,
      "epoch": 0.9114742637420257,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923698297998183,
      "loss": 5.6531,
      "mean_token_accuracy": 0.1681464821100235,
      "num_tokens": 21135380.0,
      "step": 10430
    },
    {
      "entropy": 5.8985082626342775,
      "epoch": 0.9119112120947304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923617815286725,
      "loss": 5.5359,
      "mean_token_accuracy": 0.18132148832082748,
      "num_tokens": 21144951.0,
      "step": 10435
    },
    {
      "entropy": 5.868592834472656,
      "epoch": 0.9123481604474352,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923537290884138,
      "loss": 5.4645,
      "mean_token_accuracy": 0.1792405903339386,
      "num_tokens": 21155919.0,
      "step": 10440
    },
    {
      "entropy": 5.8961952209472654,
      "epoch": 0.9127851088001399,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004923456724791967,
      "loss": 5.5208,
      "mean_token_accuracy": 0.18211438357830048,
      "num_tokens": 21166186.0,
      "step": 10445
    },
    {
      "entropy": 5.8236864566802975,
      "epoch": 0.9132220571528445,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923376117011758,
      "loss": 5.5026,
      "mean_token_accuracy": 0.18899850249290467,
      "num_tokens": 21176043.0,
      "step": 10450
    },
    {
      "entropy": 5.840754222869873,
      "epoch": 0.9136590055055492,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004923295467545055,
      "loss": 5.5068,
      "mean_token_accuracy": 0.18121806532144547,
      "num_tokens": 21185677.0,
      "step": 10455
    },
    {
      "entropy": 5.94188117980957,
      "epoch": 0.9140959538582539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923214776393407,
      "loss": 5.6412,
      "mean_token_accuracy": 0.16796546727418898,
      "num_tokens": 21195977.0,
      "step": 10460
    },
    {
      "entropy": 5.991411352157593,
      "epoch": 0.9145329022109586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492313404355836,
      "loss": 5.5954,
      "mean_token_accuracy": 0.1750256448984146,
      "num_tokens": 21205454.0,
      "step": 10465
    },
    {
      "entropy": 5.92635498046875,
      "epoch": 0.9149698505636634,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923053269041464,
      "loss": 5.4953,
      "mean_token_accuracy": 0.18481633067131042,
      "num_tokens": 21215700.0,
      "step": 10470
    },
    {
      "entropy": 5.8680510997772215,
      "epoch": 0.9154067989163681,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922972452844267,
      "loss": 5.5909,
      "mean_token_accuracy": 0.17599838823080063,
      "num_tokens": 21226544.0,
      "step": 10475
    },
    {
      "entropy": 5.895989036560058,
      "epoch": 0.9158437472690728,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922891594968321,
      "loss": 5.6297,
      "mean_token_accuracy": 0.1716434344649315,
      "num_tokens": 21236792.0,
      "step": 10480
    },
    {
      "entropy": 5.922092580795288,
      "epoch": 0.9162806956217775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922810695415175,
      "loss": 5.5373,
      "mean_token_accuracy": 0.1778067633509636,
      "num_tokens": 21247039.0,
      "step": 10485
    },
    {
      "entropy": 5.893904066085815,
      "epoch": 0.9167176439744822,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922729754186381,
      "loss": 5.557,
      "mean_token_accuracy": 0.18377964049577714,
      "num_tokens": 21257008.0,
      "step": 10490
    },
    {
      "entropy": 5.908799409866333,
      "epoch": 0.9171545923271869,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004922648771283492,
      "loss": 5.6853,
      "mean_token_accuracy": 0.17611668407917022,
      "num_tokens": 21267638.0,
      "step": 10495
    },
    {
      "entropy": 5.971077156066895,
      "epoch": 0.9175915406798917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004922567746708062,
      "loss": 5.526,
      "mean_token_accuracy": 0.18789663761854172,
      "num_tokens": 21277327.0,
      "step": 10500
    },
    {
      "entropy": 5.905153465270996,
      "epoch": 0.9180284890325964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922486680461642,
      "loss": 5.6104,
      "mean_token_accuracy": 0.17451077103614807,
      "num_tokens": 21287579.0,
      "step": 10505
    },
    {
      "entropy": 5.897438097000122,
      "epoch": 0.9184654373853011,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004922405572545792,
      "loss": 5.5213,
      "mean_token_accuracy": 0.17812347412109375,
      "num_tokens": 21297888.0,
      "step": 10510
    },
    {
      "entropy": 5.927792358398437,
      "epoch": 0.9189023857380058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922324422962063,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17634132504463196,
      "num_tokens": 21308506.0,
      "step": 10515
    },
    {
      "entropy": 5.943510484695435,
      "epoch": 0.9193393340907104,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922243231712013,
      "loss": 5.6658,
      "mean_token_accuracy": 0.1685685470700264,
      "num_tokens": 21319252.0,
      "step": 10520
    },
    {
      "entropy": 5.937555551528931,
      "epoch": 0.9197762824434151,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049221619987972,
      "loss": 5.6061,
      "mean_token_accuracy": 0.18234832137823104,
      "num_tokens": 21328829.0,
      "step": 10525
    },
    {
      "entropy": 5.864519834518433,
      "epoch": 0.9202132307961199,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922080724219182,
      "loss": 5.4201,
      "mean_token_accuracy": 0.1867707535624504,
      "num_tokens": 21338524.0,
      "step": 10530
    },
    {
      "entropy": 5.915755653381348,
      "epoch": 0.9206501791488246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004921999407979518,
      "loss": 5.5697,
      "mean_token_accuracy": 0.180117167532444,
      "num_tokens": 21348559.0,
      "step": 10535
    },
    {
      "entropy": 5.93333888053894,
      "epoch": 0.9210871275015293,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921918050079766,
      "loss": 5.5794,
      "mean_token_accuracy": 0.17783726900815963,
      "num_tokens": 21360763.0,
      "step": 10540
    },
    {
      "entropy": 5.949997282028198,
      "epoch": 0.921524075854234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921836650521489,
      "loss": 5.6019,
      "mean_token_accuracy": 0.173680679500103,
      "num_tokens": 21370048.0,
      "step": 10545
    },
    {
      "entropy": 5.891751146316528,
      "epoch": 0.9219610242069387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921755209306247,
      "loss": 5.5818,
      "mean_token_accuracy": 0.1812844157218933,
      "num_tokens": 21379760.0,
      "step": 10550
    },
    {
      "entropy": 5.972139930725097,
      "epoch": 0.9223979725596434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921673726435601,
      "loss": 5.5566,
      "mean_token_accuracy": 0.1700993150472641,
      "num_tokens": 21389712.0,
      "step": 10555
    },
    {
      "entropy": 5.9318116188049315,
      "epoch": 0.9228349209123482,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921592201911115,
      "loss": 5.5381,
      "mean_token_accuracy": 0.17435239553451537,
      "num_tokens": 21399515.0,
      "step": 10560
    },
    {
      "entropy": 5.853435230255127,
      "epoch": 0.9232718692650529,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921510635734353,
      "loss": 5.5161,
      "mean_token_accuracy": 0.1826414704322815,
      "num_tokens": 21409704.0,
      "step": 10565
    },
    {
      "entropy": 5.830459356307983,
      "epoch": 0.9237088176177576,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000492142902790688,
      "loss": 5.5098,
      "mean_token_accuracy": 0.18374867737293243,
      "num_tokens": 21419182.0,
      "step": 10570
    },
    {
      "entropy": 5.871296834945679,
      "epoch": 0.9241457659704623,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921347378430259,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18679474145174027,
      "num_tokens": 21429749.0,
      "step": 10575
    },
    {
      "entropy": 6.002449893951416,
      "epoch": 0.924582714323167,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921265687306059,
      "loss": 5.6685,
      "mean_token_accuracy": 0.1714085802435875,
      "num_tokens": 21440514.0,
      "step": 10580
    },
    {
      "entropy": 5.897093868255615,
      "epoch": 0.9250196626758717,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921183954535844,
      "loss": 5.5335,
      "mean_token_accuracy": 0.18145685493946076,
      "num_tokens": 21450647.0,
      "step": 10585
    },
    {
      "entropy": 5.873497343063354,
      "epoch": 0.9254566110285765,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921102180121184,
      "loss": 5.5699,
      "mean_token_accuracy": 0.17948605120182037,
      "num_tokens": 21460233.0,
      "step": 10590
    },
    {
      "entropy": 5.917082977294922,
      "epoch": 0.9258935593812812,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004921020364063646,
      "loss": 5.5632,
      "mean_token_accuracy": 0.17725660055875778,
      "num_tokens": 21470207.0,
      "step": 10595
    },
    {
      "entropy": 5.875304889678955,
      "epoch": 0.9263305077339858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004920938506364801,
      "loss": 5.5919,
      "mean_token_accuracy": 0.17556622326374055,
      "num_tokens": 21480658.0,
      "step": 10600
    },
    {
      "entropy": 5.866287708282471,
      "epoch": 0.9267674560866905,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004920856607026216,
      "loss": 5.4839,
      "mean_token_accuracy": 0.1922851413488388,
      "num_tokens": 21490566.0,
      "step": 10605
    },
    {
      "entropy": 5.931895208358765,
      "epoch": 0.9272044044393952,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920774666049465,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17869549095630646,
      "num_tokens": 21500752.0,
      "step": 10610
    },
    {
      "entropy": 5.879027509689331,
      "epoch": 0.9276413527921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004920692683436118,
      "loss": 5.5698,
      "mean_token_accuracy": 0.1718427747488022,
      "num_tokens": 21509802.0,
      "step": 10615
    },
    {
      "entropy": 5.917226696014405,
      "epoch": 0.9280783011448047,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920610659187748,
      "loss": 5.6624,
      "mean_token_accuracy": 0.16983651518821716,
      "num_tokens": 21520962.0,
      "step": 10620
    },
    {
      "entropy": 5.996553039550781,
      "epoch": 0.9285152494975094,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004920528593305929,
      "loss": 5.629,
      "mean_token_accuracy": 0.1737879529595375,
      "num_tokens": 21530580.0,
      "step": 10625
    },
    {
      "entropy": 5.880358028411865,
      "epoch": 0.9289521978502141,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920446485792233,
      "loss": 5.5472,
      "mean_token_accuracy": 0.17741750031709672,
      "num_tokens": 21541431.0,
      "step": 10630
    },
    {
      "entropy": 5.915652704238892,
      "epoch": 0.9293891462029188,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920364336648236,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18236221373081207,
      "num_tokens": 21551012.0,
      "step": 10635
    },
    {
      "entropy": 5.877630424499512,
      "epoch": 0.9298260945556235,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920282145875514,
      "loss": 5.5383,
      "mean_token_accuracy": 0.18218722194433212,
      "num_tokens": 21560676.0,
      "step": 10640
    },
    {
      "entropy": 5.8190398693084715,
      "epoch": 0.9302630429083283,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004920199913475644,
      "loss": 5.5015,
      "mean_token_accuracy": 0.18570253252983093,
      "num_tokens": 21570848.0,
      "step": 10645
    },
    {
      "entropy": 5.907581758499146,
      "epoch": 0.930699991261033,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004920117639450202,
      "loss": 5.5847,
      "mean_token_accuracy": 0.17645841985940933,
      "num_tokens": 21582358.0,
      "step": 10650
    },
    {
      "entropy": 5.95600848197937,
      "epoch": 0.9311369396137377,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920035323800767,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17551915496587753,
      "num_tokens": 21591709.0,
      "step": 10655
    },
    {
      "entropy": 5.94644832611084,
      "epoch": 0.9315738879664424,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004919952966528917,
      "loss": 5.502,
      "mean_token_accuracy": 0.18150367736816406,
      "num_tokens": 21601092.0,
      "step": 10660
    },
    {
      "entropy": 5.829319810867309,
      "epoch": 0.932010836319147,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919870567636233,
      "loss": 5.5179,
      "mean_token_accuracy": 0.17998500317335128,
      "num_tokens": 21611003.0,
      "step": 10665
    },
    {
      "entropy": 5.909210538864135,
      "epoch": 0.9324477846718517,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919788127124293,
      "loss": 5.5989,
      "mean_token_accuracy": 0.17410380244255066,
      "num_tokens": 21621528.0,
      "step": 10670
    },
    {
      "entropy": 5.880142545700073,
      "epoch": 0.9328847330245565,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491970564499468,
      "loss": 5.5054,
      "mean_token_accuracy": 0.1845574125647545,
      "num_tokens": 21631431.0,
      "step": 10675
    },
    {
      "entropy": 5.886252307891846,
      "epoch": 0.9333216813772612,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919623121248976,
      "loss": 5.543,
      "mean_token_accuracy": 0.1803442120552063,
      "num_tokens": 21640717.0,
      "step": 10680
    },
    {
      "entropy": 5.885072803497314,
      "epoch": 0.9337586297299659,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919540555888765,
      "loss": 5.5303,
      "mean_token_accuracy": 0.18293066322803497,
      "num_tokens": 21650493.0,
      "step": 10685
    },
    {
      "entropy": 5.9086448669433596,
      "epoch": 0.9341955780826706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004919457948915628,
      "loss": 5.5244,
      "mean_token_accuracy": 0.1883617326617241,
      "num_tokens": 21660062.0,
      "step": 10690
    },
    {
      "entropy": 5.86251392364502,
      "epoch": 0.9346325264353753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004919375300331151,
      "loss": 5.4962,
      "mean_token_accuracy": 0.1818907991051674,
      "num_tokens": 21669364.0,
      "step": 10695
    },
    {
      "entropy": 5.875346946716308,
      "epoch": 0.93506947478808,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919292610136919,
      "loss": 5.6132,
      "mean_token_accuracy": 0.17106032371520996,
      "num_tokens": 21680793.0,
      "step": 10700
    },
    {
      "entropy": 5.910564994812011,
      "epoch": 0.9355064231407848,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004919209878334518,
      "loss": 5.6322,
      "mean_token_accuracy": 0.17522368133068084,
      "num_tokens": 21690862.0,
      "step": 10705
    },
    {
      "entropy": 5.926380920410156,
      "epoch": 0.9359433714934895,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919127104925535,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18074657917022705,
      "num_tokens": 21700665.0,
      "step": 10710
    },
    {
      "entropy": 5.8943873882293705,
      "epoch": 0.9363803198461942,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004919044289911557,
      "loss": 5.5524,
      "mean_token_accuracy": 0.18183531761169433,
      "num_tokens": 21711128.0,
      "step": 10715
    },
    {
      "entropy": 5.9501885890960695,
      "epoch": 0.9368172681988989,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918961433294173,
      "loss": 5.618,
      "mean_token_accuracy": 0.17075878381729126,
      "num_tokens": 21720233.0,
      "step": 10720
    },
    {
      "entropy": 5.910007953643799,
      "epoch": 0.9372542165516036,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918878535074973,
      "loss": 5.5509,
      "mean_token_accuracy": 0.18501711189746856,
      "num_tokens": 21730980.0,
      "step": 10725
    },
    {
      "entropy": 5.830601644515991,
      "epoch": 0.9376911649043083,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004918795595255545,
      "loss": 5.5623,
      "mean_token_accuracy": 0.18201815634965895,
      "num_tokens": 21742802.0,
      "step": 10730
    },
    {
      "entropy": 5.9125946998596195,
      "epoch": 0.9381281132570131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918712613837483,
      "loss": 5.5777,
      "mean_token_accuracy": 0.17735107988119125,
      "num_tokens": 21752743.0,
      "step": 10735
    },
    {
      "entropy": 5.912267780303955,
      "epoch": 0.9385650616097178,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918629590822375,
      "loss": 5.5516,
      "mean_token_accuracy": 0.18049676716327667,
      "num_tokens": 21763198.0,
      "step": 10740
    },
    {
      "entropy": 5.912556982040405,
      "epoch": 0.9390020099624224,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004918546526211816,
      "loss": 5.5552,
      "mean_token_accuracy": 0.18263645321130753,
      "num_tokens": 21773030.0,
      "step": 10745
    },
    {
      "entropy": 5.890395164489746,
      "epoch": 0.9394389583151271,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004918463420007398,
      "loss": 5.5073,
      "mean_token_accuracy": 0.18501332849264146,
      "num_tokens": 21783796.0,
      "step": 10750
    },
    {
      "entropy": 5.920323085784912,
      "epoch": 0.9398759066678318,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918380272210715,
      "loss": 5.549,
      "mean_token_accuracy": 0.18007050603628158,
      "num_tokens": 21792504.0,
      "step": 10755
    },
    {
      "entropy": 5.932688999176025,
      "epoch": 0.9403128550205365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918297082823362,
      "loss": 5.5888,
      "mean_token_accuracy": 0.1780189797282219,
      "num_tokens": 21802764.0,
      "step": 10760
    },
    {
      "entropy": 5.93286714553833,
      "epoch": 0.9407498033732413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918213851846935,
      "loss": 5.563,
      "mean_token_accuracy": 0.17292096018791198,
      "num_tokens": 21813051.0,
      "step": 10765
    },
    {
      "entropy": 5.86859016418457,
      "epoch": 0.941186751725946,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918130579283031,
      "loss": 5.6159,
      "mean_token_accuracy": 0.1784030854701996,
      "num_tokens": 21822941.0,
      "step": 10770
    },
    {
      "entropy": 5.908771085739136,
      "epoch": 0.9416237000786507,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918047265133246,
      "loss": 5.5369,
      "mean_token_accuracy": 0.17559186965227128,
      "num_tokens": 21833225.0,
      "step": 10775
    },
    {
      "entropy": 5.88128342628479,
      "epoch": 0.9420606484313554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004917963909399178,
      "loss": 5.5428,
      "mean_token_accuracy": 0.17875492125749587,
      "num_tokens": 21842360.0,
      "step": 10780
    },
    {
      "entropy": 5.906546258926392,
      "epoch": 0.9424975967840601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004917880512082427,
      "loss": 5.5521,
      "mean_token_accuracy": 0.18058207035064697,
      "num_tokens": 21852515.0,
      "step": 10785
    },
    {
      "entropy": 5.940371561050415,
      "epoch": 0.9429345451367648,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917797073184591,
      "loss": 5.6379,
      "mean_token_accuracy": 0.16890014559030533,
      "num_tokens": 21863040.0,
      "step": 10790
    },
    {
      "entropy": 5.962755823135376,
      "epoch": 0.9433714934894696,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004917713592707271,
      "loss": 5.5801,
      "mean_token_accuracy": 0.1764754831790924,
      "num_tokens": 21872101.0,
      "step": 10795
    },
    {
      "entropy": 5.95067343711853,
      "epoch": 0.9438084418421743,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917630070652069,
      "loss": 5.5516,
      "mean_token_accuracy": 0.1808815598487854,
      "num_tokens": 21882148.0,
      "step": 10800
    },
    {
      "entropy": 5.858833503723145,
      "epoch": 0.944245390194879,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917546507020587,
      "loss": 5.5438,
      "mean_token_accuracy": 0.18813767582178115,
      "num_tokens": 21891832.0,
      "step": 10805
    },
    {
      "entropy": 5.8951520919799805,
      "epoch": 0.9446823385475837,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917462901814427,
      "loss": 5.5522,
      "mean_token_accuracy": 0.17712648659944535,
      "num_tokens": 21901544.0,
      "step": 10810
    },
    {
      "entropy": 5.926298379898071,
      "epoch": 0.9451192869002883,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917379255035192,
      "loss": 5.6066,
      "mean_token_accuracy": 0.17619642466306687,
      "num_tokens": 21911690.0,
      "step": 10815
    },
    {
      "entropy": 5.956694030761719,
      "epoch": 0.9455562352529931,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004917295566684487,
      "loss": 5.6071,
      "mean_token_accuracy": 0.17387891560792923,
      "num_tokens": 21922968.0,
      "step": 10820
    },
    {
      "entropy": 5.872158050537109,
      "epoch": 0.9459931836056978,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004917211836763919,
      "loss": 5.4462,
      "mean_token_accuracy": 0.1854881078004837,
      "num_tokens": 21932904.0,
      "step": 10825
    },
    {
      "entropy": 5.833469486236572,
      "epoch": 0.9464301319584025,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917128065275091,
      "loss": 5.4698,
      "mean_token_accuracy": 0.1914062723517418,
      "num_tokens": 21943065.0,
      "step": 10830
    },
    {
      "entropy": 5.922138595581055,
      "epoch": 0.9468670803111072,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917044252219612,
      "loss": 5.5616,
      "mean_token_accuracy": 0.16796028167009353,
      "num_tokens": 21953448.0,
      "step": 10835
    },
    {
      "entropy": 5.94380750656128,
      "epoch": 0.9473040286638119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004916960397599088,
      "loss": 5.5454,
      "mean_token_accuracy": 0.17900159507989882,
      "num_tokens": 21963374.0,
      "step": 10840
    },
    {
      "entropy": 5.876992845535279,
      "epoch": 0.9477409770165166,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916876501415129,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1710141643881798,
      "num_tokens": 21973726.0,
      "step": 10845
    },
    {
      "entropy": 5.8551147937774655,
      "epoch": 0.9481779253692214,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004916792563669342,
      "loss": 5.4547,
      "mean_token_accuracy": 0.18366723358631135,
      "num_tokens": 21983203.0,
      "step": 10850
    },
    {
      "entropy": 5.900910997390747,
      "epoch": 0.9486148737219261,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916708584363339,
      "loss": 5.5684,
      "mean_token_accuracy": 0.17740870714187623,
      "num_tokens": 21993435.0,
      "step": 10855
    },
    {
      "entropy": 5.974231100082397,
      "epoch": 0.9490518220746308,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491662456349873,
      "loss": 5.6075,
      "mean_token_accuracy": 0.17139323949813842,
      "num_tokens": 22003230.0,
      "step": 10860
    },
    {
      "entropy": 5.9436078548431395,
      "epoch": 0.9494887704273355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916540501077127,
      "loss": 5.5343,
      "mean_token_accuracy": 0.18220920413732528,
      "num_tokens": 22012910.0,
      "step": 10865
    },
    {
      "entropy": 5.858736801147461,
      "epoch": 0.9499257187800402,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004916456397100141,
      "loss": 5.4977,
      "mean_token_accuracy": 0.18110527992248535,
      "num_tokens": 22022802.0,
      "step": 10870
    },
    {
      "entropy": 5.8991954803466795,
      "epoch": 0.9503626671327449,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916372251569386,
      "loss": 5.6067,
      "mean_token_accuracy": 0.17622564882040023,
      "num_tokens": 22032669.0,
      "step": 10875
    },
    {
      "entropy": 5.904579114913941,
      "epoch": 0.9507996154854497,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916288064486477,
      "loss": 5.5141,
      "mean_token_accuracy": 0.1863548055291176,
      "num_tokens": 22042477.0,
      "step": 10880
    },
    {
      "entropy": 5.934581756591797,
      "epoch": 0.9512365638381544,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004916203835853027,
      "loss": 5.5607,
      "mean_token_accuracy": 0.1824100986123085,
      "num_tokens": 22052891.0,
      "step": 10885
    },
    {
      "entropy": 5.855764389038086,
      "epoch": 0.951673512190859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916119565670652,
      "loss": 5.5014,
      "mean_token_accuracy": 0.19058071672916413,
      "num_tokens": 22062413.0,
      "step": 10890
    },
    {
      "entropy": 5.928059911727905,
      "epoch": 0.9521104605435637,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491603525394097,
      "loss": 5.599,
      "mean_token_accuracy": 0.1718723326921463,
      "num_tokens": 22072565.0,
      "step": 10895
    },
    {
      "entropy": 5.880865955352784,
      "epoch": 0.9525474088962684,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004915950900665596,
      "loss": 5.5585,
      "mean_token_accuracy": 0.17846859395503997,
      "num_tokens": 22082910.0,
      "step": 10900
    },
    {
      "entropy": 5.793889904022217,
      "epoch": 0.9529843572489731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000491586650584615,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1895809918642044,
      "num_tokens": 22092636.0,
      "step": 10905
    },
    {
      "entropy": 5.941836738586426,
      "epoch": 0.9534213056016779,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915782069484248,
      "loss": 5.5748,
      "mean_token_accuracy": 0.176946721971035,
      "num_tokens": 22102232.0,
      "step": 10910
    },
    {
      "entropy": 5.988768148422241,
      "epoch": 0.9538582539543826,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004915697591581511,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17055184692144393,
      "num_tokens": 22113592.0,
      "step": 10915
    },
    {
      "entropy": 5.8944254398345945,
      "epoch": 0.9542952023070873,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000491561307213956,
      "loss": 5.5784,
      "mean_token_accuracy": 0.17907858192920684,
      "num_tokens": 22123731.0,
      "step": 10920
    },
    {
      "entropy": 5.906564331054687,
      "epoch": 0.954732150659792,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004915528511160016,
      "loss": 5.6192,
      "mean_token_accuracy": 0.16760052144527435,
      "num_tokens": 22133801.0,
      "step": 10925
    },
    {
      "entropy": 5.964372539520264,
      "epoch": 0.9551690990124967,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004915443908644501,
      "loss": 5.5795,
      "mean_token_accuracy": 0.1801023527979851,
      "num_tokens": 22143939.0,
      "step": 10930
    },
    {
      "entropy": 5.8579347133636475,
      "epoch": 0.9556060473652014,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004915359264594636,
      "loss": 5.5813,
      "mean_token_accuracy": 0.17690520733594894,
      "num_tokens": 22154954.0,
      "step": 10935
    },
    {
      "entropy": 5.8935013771057125,
      "epoch": 0.9560429957179062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915274579012045,
      "loss": 5.579,
      "mean_token_accuracy": 0.1790354773402214,
      "num_tokens": 22165225.0,
      "step": 10940
    },
    {
      "entropy": 6.004147386550903,
      "epoch": 0.9564799440706109,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004915189851898355,
      "loss": 5.6029,
      "mean_token_accuracy": 0.16933415085077286,
      "num_tokens": 22174891.0,
      "step": 10945
    },
    {
      "entropy": 5.866418027877808,
      "epoch": 0.9569168924233156,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915105083255187,
      "loss": 5.5578,
      "mean_token_accuracy": 0.18404623866081238,
      "num_tokens": 22185255.0,
      "step": 10950
    },
    {
      "entropy": 5.86105318069458,
      "epoch": 0.9573538407760203,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000491502027308417,
      "loss": 5.6254,
      "mean_token_accuracy": 0.17540458887815474,
      "num_tokens": 22196124.0,
      "step": 10955
    },
    {
      "entropy": 5.9394989013671875,
      "epoch": 0.957790789128725,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491493542138693,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17908347845077516,
      "num_tokens": 22207563.0,
      "step": 10960
    },
    {
      "entropy": 5.915901327133179,
      "epoch": 0.9582277374814296,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914850528165094,
      "loss": 5.5843,
      "mean_token_accuracy": 0.17504732310771942,
      "num_tokens": 22218527.0,
      "step": 10965
    },
    {
      "entropy": 5.903222942352295,
      "epoch": 0.9586646858341344,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491476559342029,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1762381002306938,
      "num_tokens": 22229538.0,
      "step": 10970
    },
    {
      "entropy": 5.918375539779663,
      "epoch": 0.9591016341868391,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004914680617154148,
      "loss": 5.4684,
      "mean_token_accuracy": 0.18881391733884811,
      "num_tokens": 22240047.0,
      "step": 10975
    },
    {
      "entropy": 5.860059928894043,
      "epoch": 0.9595385825395438,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004914595599368297,
      "loss": 5.4506,
      "mean_token_accuracy": 0.18423337787389754,
      "num_tokens": 22250440.0,
      "step": 10980
    },
    {
      "entropy": 5.903094720840454,
      "epoch": 0.9599755308922485,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004914510540064369,
      "loss": 5.5267,
      "mean_token_accuracy": 0.1817627102136612,
      "num_tokens": 22259909.0,
      "step": 10985
    },
    {
      "entropy": 5.9023425579071045,
      "epoch": 0.9604124792449532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004914425439243994,
      "loss": 5.5891,
      "mean_token_accuracy": 0.17534127384424208,
      "num_tokens": 22270358.0,
      "step": 10990
    },
    {
      "entropy": 5.938672113418579,
      "epoch": 0.9608494275976579,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914340296908805,
      "loss": 5.6204,
      "mean_token_accuracy": 0.17126917093992233,
      "num_tokens": 22280591.0,
      "step": 10995
    },
    {
      "entropy": 5.981192922592163,
      "epoch": 0.9612863759503627,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004914255113060435,
      "loss": 5.5667,
      "mean_token_accuracy": 0.18286112695932388,
      "num_tokens": 22291067.0,
      "step": 11000
    },
    {
      "entropy": 5.881573820114136,
      "epoch": 0.9617233243030674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914169887700519,
      "loss": 5.5014,
      "mean_token_accuracy": 0.18740880638360977,
      "num_tokens": 22300919.0,
      "step": 11005
    },
    {
      "entropy": 5.9030348777771,
      "epoch": 0.9621602726557721,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914084620830689,
      "loss": 5.6047,
      "mean_token_accuracy": 0.17840543985366822,
      "num_tokens": 22311562.0,
      "step": 11010
    },
    {
      "entropy": 5.924402904510498,
      "epoch": 0.9625972210084768,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004913999312452583,
      "loss": 5.5601,
      "mean_token_accuracy": 0.1741113319993019,
      "num_tokens": 22322980.0,
      "step": 11015
    },
    {
      "entropy": 5.906336402893066,
      "epoch": 0.9630341693611815,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004913913962567836,
      "loss": 5.4686,
      "mean_token_accuracy": 0.18108532279729844,
      "num_tokens": 22333242.0,
      "step": 11020
    },
    {
      "entropy": 5.932784605026245,
      "epoch": 0.9634711177138863,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004913828571178085,
      "loss": 5.5724,
      "mean_token_accuracy": 0.1755265086889267,
      "num_tokens": 22343618.0,
      "step": 11025
    },
    {
      "entropy": 5.8681824684143065,
      "epoch": 0.963908066066591,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913743138284969,
      "loss": 5.5709,
      "mean_token_accuracy": 0.17418719232082366,
      "num_tokens": 22354249.0,
      "step": 11030
    },
    {
      "entropy": 5.992564630508423,
      "epoch": 0.9643450144192957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913657663890124,
      "loss": 5.5379,
      "mean_token_accuracy": 0.17548028528690338,
      "num_tokens": 22363824.0,
      "step": 11035
    },
    {
      "entropy": 5.907478618621826,
      "epoch": 0.9647819627720003,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004913572147995193,
      "loss": 5.5252,
      "mean_token_accuracy": 0.17883699089288713,
      "num_tokens": 22373470.0,
      "step": 11040
    },
    {
      "entropy": 5.815442991256714,
      "epoch": 0.965218911124705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004913486590601814,
      "loss": 5.4716,
      "mean_token_accuracy": 0.17863886803388596,
      "num_tokens": 22382999.0,
      "step": 11045
    },
    {
      "entropy": 5.88115963935852,
      "epoch": 0.9656558594774097,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913400991711629,
      "loss": 5.5282,
      "mean_token_accuracy": 0.18449266105890275,
      "num_tokens": 22393415.0,
      "step": 11050
    },
    {
      "entropy": 5.896071481704712,
      "epoch": 0.9660928078301145,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913315351326279,
      "loss": 5.4838,
      "mean_token_accuracy": 0.18078666180372238,
      "num_tokens": 22403040.0,
      "step": 11055
    },
    {
      "entropy": 5.882252502441406,
      "epoch": 0.9665297561828192,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913229669447406,
      "loss": 5.5099,
      "mean_token_accuracy": 0.18341243714094163,
      "num_tokens": 22413434.0,
      "step": 11060
    },
    {
      "entropy": 5.910290956497192,
      "epoch": 0.9669667045355239,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913143946076656,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1747012421488762,
      "num_tokens": 22424446.0,
      "step": 11065
    },
    {
      "entropy": 5.898932647705078,
      "epoch": 0.9674036528882286,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913058181215671,
      "loss": 5.5588,
      "mean_token_accuracy": 0.1793728217482567,
      "num_tokens": 22433928.0,
      "step": 11070
    },
    {
      "entropy": 5.932872295379639,
      "epoch": 0.9678406012409333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912972374866096,
      "loss": 5.5289,
      "mean_token_accuracy": 0.1809048220515251,
      "num_tokens": 22443778.0,
      "step": 11075
    },
    {
      "entropy": 5.924958276748657,
      "epoch": 0.968277549593638,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912886527029579,
      "loss": 5.6126,
      "mean_token_accuracy": 0.17867953926324845,
      "num_tokens": 22453423.0,
      "step": 11080
    },
    {
      "entropy": 5.946870517730713,
      "epoch": 0.9687144979463428,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004912800637707764,
      "loss": 5.5237,
      "mean_token_accuracy": 0.181604366004467,
      "num_tokens": 22464067.0,
      "step": 11085
    },
    {
      "entropy": 5.861277103424072,
      "epoch": 0.9691514462990475,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049127147069023,
      "loss": 5.5276,
      "mean_token_accuracy": 0.18210440874099731,
      "num_tokens": 22474332.0,
      "step": 11090
    },
    {
      "entropy": 5.888250780105591,
      "epoch": 0.9695883946517522,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004912628734614834,
      "loss": 5.5389,
      "mean_token_accuracy": 0.17990748435258866,
      "num_tokens": 22485304.0,
      "step": 11095
    },
    {
      "entropy": 5.873070764541626,
      "epoch": 0.9700253430044569,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912542720847016,
      "loss": 5.4692,
      "mean_token_accuracy": 0.18105277568101882,
      "num_tokens": 22494447.0,
      "step": 11100
    },
    {
      "entropy": 5.921037817001343,
      "epoch": 0.9704622913571616,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912456665600495,
      "loss": 5.5478,
      "mean_token_accuracy": 0.1824273571372032,
      "num_tokens": 22503936.0,
      "step": 11105
    },
    {
      "entropy": 5.8725508689880375,
      "epoch": 0.9708992397098662,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004912370568876923,
      "loss": 5.4461,
      "mean_token_accuracy": 0.1818022161722183,
      "num_tokens": 22513960.0,
      "step": 11110
    },
    {
      "entropy": 5.90219030380249,
      "epoch": 0.971336188062571,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912284430677949,
      "loss": 5.5926,
      "mean_token_accuracy": 0.1777928963303566,
      "num_tokens": 22522309.0,
      "step": 11115
    },
    {
      "entropy": 5.9252372741699215,
      "epoch": 0.9717731364152757,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004912198251005228,
      "loss": 5.4911,
      "mean_token_accuracy": 0.18331363797187805,
      "num_tokens": 22533510.0,
      "step": 11120
    },
    {
      "entropy": 5.8896294116973875,
      "epoch": 0.9722100847679804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912112029860411,
      "loss": 5.5046,
      "mean_token_accuracy": 0.17951876372098924,
      "num_tokens": 22543644.0,
      "step": 11125
    },
    {
      "entropy": 5.946816492080688,
      "epoch": 0.9726470331206851,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004912025767245153,
      "loss": 5.6777,
      "mean_token_accuracy": 0.1669557124376297,
      "num_tokens": 22553475.0,
      "step": 11130
    },
    {
      "entropy": 5.861800765991211,
      "epoch": 0.9730839814733898,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004911939463161108,
      "loss": 5.4865,
      "mean_token_accuracy": 0.18603063821792604,
      "num_tokens": 22563190.0,
      "step": 11135
    },
    {
      "entropy": 5.922503089904785,
      "epoch": 0.9735209298260945,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491185311760993,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18471471220254898,
      "num_tokens": 22572069.0,
      "step": 11140
    },
    {
      "entropy": 5.952958774566651,
      "epoch": 0.9739578781787993,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004911766730593276,
      "loss": 5.6199,
      "mean_token_accuracy": 0.16885922998189926,
      "num_tokens": 22581597.0,
      "step": 11145
    },
    {
      "entropy": 5.964114999771118,
      "epoch": 0.974394826531504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911680302112806,
      "loss": 5.506,
      "mean_token_accuracy": 0.18669941425323486,
      "num_tokens": 22591486.0,
      "step": 11150
    },
    {
      "entropy": 5.955888652801514,
      "epoch": 0.9748317748842087,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004911593832170173,
      "loss": 5.5574,
      "mean_token_accuracy": 0.17579471170902253,
      "num_tokens": 22601615.0,
      "step": 11155
    },
    {
      "entropy": 5.911087894439698,
      "epoch": 0.9752687232369134,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911507320767038,
      "loss": 5.5464,
      "mean_token_accuracy": 0.17761144638061524,
      "num_tokens": 22611397.0,
      "step": 11160
    },
    {
      "entropy": 5.8037097454071045,
      "epoch": 0.9757056715896181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000491142076790506,
      "loss": 5.5242,
      "mean_token_accuracy": 0.18653350323438644,
      "num_tokens": 22621269.0,
      "step": 11165
    },
    {
      "entropy": 5.937428617477417,
      "epoch": 0.9761426199423228,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911334173585899,
      "loss": 5.5014,
      "mean_token_accuracy": 0.18220175802707672,
      "num_tokens": 22631782.0,
      "step": 11170
    },
    {
      "entropy": 5.914890909194947,
      "epoch": 0.9765795682950276,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911247537811216,
      "loss": 5.594,
      "mean_token_accuracy": 0.17610469311475754,
      "num_tokens": 22642240.0,
      "step": 11175
    },
    {
      "entropy": 5.939141130447387,
      "epoch": 0.9770165166477323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004911160860582674,
      "loss": 5.5528,
      "mean_token_accuracy": 0.17803838104009628,
      "num_tokens": 22651977.0,
      "step": 11180
    },
    {
      "entropy": 5.940335369110107,
      "epoch": 0.977453465000437,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911074141901933,
      "loss": 5.5783,
      "mean_token_accuracy": 0.18274836540222167,
      "num_tokens": 22662871.0,
      "step": 11185
    },
    {
      "entropy": 5.84832878112793,
      "epoch": 0.9778904133531416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910987381770657,
      "loss": 5.4937,
      "mean_token_accuracy": 0.19075645953416825,
      "num_tokens": 22672083.0,
      "step": 11190
    },
    {
      "entropy": 5.851566743850708,
      "epoch": 0.9783273617058463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004910900580190512,
      "loss": 5.5156,
      "mean_token_accuracy": 0.18377392143011093,
      "num_tokens": 22681247.0,
      "step": 11195
    },
    {
      "entropy": 5.910662984848022,
      "epoch": 0.9787643100585511,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910813737163161,
      "loss": 5.55,
      "mean_token_accuracy": 0.18341336101293565,
      "num_tokens": 22691231.0,
      "step": 11200
    },
    {
      "entropy": 5.916065835952759,
      "epoch": 0.9792012584112558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491072685269027,
      "loss": 5.5401,
      "mean_token_accuracy": 0.17965136766433715,
      "num_tokens": 22700712.0,
      "step": 11205
    },
    {
      "entropy": 5.952419710159302,
      "epoch": 0.9796382067639605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910639926773506,
      "loss": 5.6624,
      "mean_token_accuracy": 0.17476628422737123,
      "num_tokens": 22710961.0,
      "step": 11210
    },
    {
      "entropy": 5.907360458374024,
      "epoch": 0.9800751551166652,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004910552959414538,
      "loss": 5.5179,
      "mean_token_accuracy": 0.18152714073657988,
      "num_tokens": 22720450.0,
      "step": 11215
    },
    {
      "entropy": 5.804834413528442,
      "epoch": 0.9805121034693699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910465950615029,
      "loss": 5.4805,
      "mean_token_accuracy": 0.18100492507219315,
      "num_tokens": 22730337.0,
      "step": 11220
    },
    {
      "entropy": 5.912643480300903,
      "epoch": 0.9809490518220746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910378900376651,
      "loss": 5.5795,
      "mean_token_accuracy": 0.17732733786106109,
      "num_tokens": 22742379.0,
      "step": 11225
    },
    {
      "entropy": 5.98975887298584,
      "epoch": 0.9813860001747794,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004910291808701077,
      "loss": 5.6426,
      "mean_token_accuracy": 0.17421705275774002,
      "num_tokens": 22753176.0,
      "step": 11230
    },
    {
      "entropy": 5.9204565525054935,
      "epoch": 0.9818229485274841,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004910204675589972,
      "loss": 5.5573,
      "mean_token_accuracy": 0.1789683222770691,
      "num_tokens": 22763263.0,
      "step": 11235
    },
    {
      "entropy": 5.855878067016602,
      "epoch": 0.9822598968801888,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910117501045009,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1792857825756073,
      "num_tokens": 22774630.0,
      "step": 11240
    },
    {
      "entropy": 5.98476734161377,
      "epoch": 0.9826968452328935,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004910030285067861,
      "loss": 5.5785,
      "mean_token_accuracy": 0.17534877210855485,
      "num_tokens": 22785027.0,
      "step": 11245
    },
    {
      "entropy": 5.931835746765136,
      "epoch": 0.9831337935855982,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909943027660201,
      "loss": 5.6102,
      "mean_token_accuracy": 0.1702704355120659,
      "num_tokens": 22796399.0,
      "step": 11250
    },
    {
      "entropy": 5.889214611053466,
      "epoch": 0.9835707419383029,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049098557288237,
      "loss": 5.5484,
      "mean_token_accuracy": 0.17622375190258027,
      "num_tokens": 22806352.0,
      "step": 11255
    },
    {
      "entropy": 5.955852699279785,
      "epoch": 0.9840076902910077,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909768388560036,
      "loss": 5.6076,
      "mean_token_accuracy": 0.17351708114147185,
      "num_tokens": 22818369.0,
      "step": 11260
    },
    {
      "entropy": 5.921997737884522,
      "epoch": 0.9844446386437123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004909681006870882,
      "loss": 5.5293,
      "mean_token_accuracy": 0.17806522995233537,
      "num_tokens": 22828063.0,
      "step": 11265
    },
    {
      "entropy": 5.880717802047729,
      "epoch": 0.984881586996417,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909593583757915,
      "loss": 5.5932,
      "mean_token_accuracy": 0.17599959671497345,
      "num_tokens": 22839265.0,
      "step": 11270
    },
    {
      "entropy": 5.909173107147216,
      "epoch": 0.9853185353491217,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004909506119222812,
      "loss": 5.527,
      "mean_token_accuracy": 0.1859760880470276,
      "num_tokens": 22848501.0,
      "step": 11275
    },
    {
      "entropy": 5.898053932189941,
      "epoch": 0.9857554837018264,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909418613267248,
      "loss": 5.4922,
      "mean_token_accuracy": 0.18248178511857988,
      "num_tokens": 22859176.0,
      "step": 11280
    },
    {
      "entropy": 5.876533889770508,
      "epoch": 0.9861924320545311,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909331065892905,
      "loss": 5.4641,
      "mean_token_accuracy": 0.1866614490747452,
      "num_tokens": 22869312.0,
      "step": 11285
    },
    {
      "entropy": 5.904917430877686,
      "epoch": 0.9866293804072359,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490924347710146,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17592183947563172,
      "num_tokens": 22878699.0,
      "step": 11290
    },
    {
      "entropy": 5.8852437973022464,
      "epoch": 0.9870663287599406,
      "grad_norm": 0.875,
      "learning_rate": 0.0004909155846894595,
      "loss": 5.6036,
      "mean_token_accuracy": 0.17381807118654252,
      "num_tokens": 22889147.0,
      "step": 11295
    },
    {
      "entropy": 5.9588782787323,
      "epoch": 0.9875032771126453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909068175273989,
      "loss": 5.52,
      "mean_token_accuracy": 0.18157160878181458,
      "num_tokens": 22899085.0,
      "step": 11300
    },
    {
      "entropy": 5.924248695373535,
      "epoch": 0.98794022546535,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004908980462241323,
      "loss": 5.6397,
      "mean_token_accuracy": 0.1734975039958954,
      "num_tokens": 22909924.0,
      "step": 11305
    },
    {
      "entropy": 5.91921854019165,
      "epoch": 0.9883771738180547,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908892707798281,
      "loss": 5.6075,
      "mean_token_accuracy": 0.17534982711076735,
      "num_tokens": 22920426.0,
      "step": 11310
    },
    {
      "entropy": 5.936602306365967,
      "epoch": 0.9888141221707594,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908804911946546,
      "loss": 5.5242,
      "mean_token_accuracy": 0.17777251601219177,
      "num_tokens": 22929929.0,
      "step": 11315
    },
    {
      "entropy": 5.866807270050049,
      "epoch": 0.9892510705234642,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908717074687801,
      "loss": 5.4982,
      "mean_token_accuracy": 0.18510529398918152,
      "num_tokens": 22940102.0,
      "step": 11320
    },
    {
      "entropy": 5.9118030071258545,
      "epoch": 0.9896880188761689,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004908629196023732,
      "loss": 5.5536,
      "mean_token_accuracy": 0.18451205343008042,
      "num_tokens": 22949490.0,
      "step": 11325
    },
    {
      "entropy": 5.912836313247681,
      "epoch": 0.9901249672288736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908541275956024,
      "loss": 5.4962,
      "mean_token_accuracy": 0.18373917490243913,
      "num_tokens": 22958119.0,
      "step": 11330
    },
    {
      "entropy": 5.929424476623535,
      "epoch": 0.9905619155815782,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908453314486363,
      "loss": 5.5167,
      "mean_token_accuracy": 0.17851549983024598,
      "num_tokens": 22969447.0,
      "step": 11335
    },
    {
      "entropy": 5.840465450286866,
      "epoch": 0.9909988639342829,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004908365311616437,
      "loss": 5.4162,
      "mean_token_accuracy": 0.1889231249690056,
      "num_tokens": 22978925.0,
      "step": 11340
    },
    {
      "entropy": 5.927017879486084,
      "epoch": 0.9914358122869876,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004908277267347931,
      "loss": 5.5434,
      "mean_token_accuracy": 0.17981443405151368,
      "num_tokens": 22988861.0,
      "step": 11345
    },
    {
      "entropy": 5.875063705444336,
      "epoch": 0.9918727606396924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908189181682538,
      "loss": 5.5691,
      "mean_token_accuracy": 0.17605106383562089,
      "num_tokens": 22998700.0,
      "step": 11350
    },
    {
      "entropy": 5.853112745285034,
      "epoch": 0.9923097089923971,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004908101054621946,
      "loss": 5.5198,
      "mean_token_accuracy": 0.18113692849874496,
      "num_tokens": 23009095.0,
      "step": 11355
    },
    {
      "entropy": 5.870637607574463,
      "epoch": 0.9927466573451018,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908012886167845,
      "loss": 5.5307,
      "mean_token_accuracy": 0.18393958806991578,
      "num_tokens": 23019484.0,
      "step": 11360
    },
    {
      "entropy": 5.875974369049072,
      "epoch": 0.9931836056978065,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907924676321924,
      "loss": 5.4996,
      "mean_token_accuracy": 0.1801249474287033,
      "num_tokens": 23029867.0,
      "step": 11365
    },
    {
      "entropy": 5.8762627124786375,
      "epoch": 0.9936205540505112,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490783642508588,
      "loss": 5.4896,
      "mean_token_accuracy": 0.1885976329445839,
      "num_tokens": 23040179.0,
      "step": 11370
    },
    {
      "entropy": 5.886495351791382,
      "epoch": 0.9940575024032159,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049077481324614,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1900678977370262,
      "num_tokens": 23050389.0,
      "step": 11375
    },
    {
      "entropy": 5.771474361419678,
      "epoch": 0.9944944507559207,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490765979845018,
      "loss": 5.4939,
      "mean_token_accuracy": 0.1885676696896553,
      "num_tokens": 23059984.0,
      "step": 11380
    },
    {
      "entropy": 5.921848249435425,
      "epoch": 0.9949313991086254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907571423053917,
      "loss": 5.5096,
      "mean_token_accuracy": 0.17872909158468248,
      "num_tokens": 23069828.0,
      "step": 11385
    },
    {
      "entropy": 5.916821241378784,
      "epoch": 0.9953683474613301,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907483006274302,
      "loss": 5.5678,
      "mean_token_accuracy": 0.175344580411911,
      "num_tokens": 23079223.0,
      "step": 11390
    },
    {
      "entropy": 5.831023120880127,
      "epoch": 0.9958052958140348,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907394548113033,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17578522562980653,
      "num_tokens": 23088540.0,
      "step": 11395
    },
    {
      "entropy": 5.933796548843384,
      "epoch": 0.9962422441667395,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004907306048571806,
      "loss": 5.6235,
      "mean_token_accuracy": 0.17322255671024323,
      "num_tokens": 23098617.0,
      "step": 11400
    },
    {
      "entropy": 5.957620143890381,
      "epoch": 0.9966791925194443,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004907217507652317,
      "loss": 5.5328,
      "mean_token_accuracy": 0.18058438897132872,
      "num_tokens": 23109702.0,
      "step": 11405
    },
    {
      "entropy": 5.943484449386597,
      "epoch": 0.997116140872149,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907128925356268,
      "loss": 5.5461,
      "mean_token_accuracy": 0.18414741903543472,
      "num_tokens": 23120295.0,
      "step": 11410
    },
    {
      "entropy": 5.87626461982727,
      "epoch": 0.9975530892248536,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004907040301685355,
      "loss": 5.5704,
      "mean_token_accuracy": 0.1771283358335495,
      "num_tokens": 23131999.0,
      "step": 11415
    },
    {
      "entropy": 5.927407217025757,
      "epoch": 0.9979900375775583,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004906951636641279,
      "loss": 5.5886,
      "mean_token_accuracy": 0.17850421965122223,
      "num_tokens": 23141060.0,
      "step": 11420
    },
    {
      "entropy": 5.856444883346557,
      "epoch": 0.998426985930263,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000490686293022574,
      "loss": 5.5243,
      "mean_token_accuracy": 0.18086791336536406,
      "num_tokens": 23151607.0,
      "step": 11425
    },
    {
      "entropy": 5.917354726791382,
      "epoch": 0.9988639342829677,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490677418244044,
      "loss": 5.6002,
      "mean_token_accuracy": 0.1704199105501175,
      "num_tokens": 23161892.0,
      "step": 11430
    },
    {
      "entropy": 5.931758689880371,
      "epoch": 0.9993008826356725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906685393287081,
      "loss": 5.5717,
      "mean_token_accuracy": 0.179206582903862,
      "num_tokens": 23172112.0,
      "step": 11435
    },
    {
      "entropy": 5.932423448562622,
      "epoch": 0.9997378309883772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004906596562767367,
      "loss": 5.4679,
      "mean_token_accuracy": 0.181186480820179,
      "num_tokens": 23181435.0,
      "step": 11440
    },
    {
      "entropy": 5.859612131118775,
      "epoch": 1.0001747793410818,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906507690883,
      "loss": 5.4172,
      "mean_token_accuracy": 0.18776923567056655,
      "num_tokens": 23191084.0,
      "step": 11445
    },
    {
      "entropy": 5.847065830230713,
      "epoch": 1.0006117276937867,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906418777635685,
      "loss": 5.4708,
      "mean_token_accuracy": 0.18683699667453765,
      "num_tokens": 23202372.0,
      "step": 11450
    },
    {
      "entropy": 5.907804775238037,
      "epoch": 1.0010486760464914,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004906329823027129,
      "loss": 5.5114,
      "mean_token_accuracy": 0.18416837751865386,
      "num_tokens": 23214113.0,
      "step": 11455
    },
    {
      "entropy": 5.871933364868164,
      "epoch": 1.001485624399196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906240827059036,
      "loss": 5.4258,
      "mean_token_accuracy": 0.18356582373380662,
      "num_tokens": 23223562.0,
      "step": 11460
    },
    {
      "entropy": 5.926493120193482,
      "epoch": 1.0019225727519008,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004906151789733114,
      "loss": 5.5096,
      "mean_token_accuracy": 0.18116679340600966,
      "num_tokens": 23232901.0,
      "step": 11465
    },
    {
      "entropy": 5.92733850479126,
      "epoch": 1.0023595211046055,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906062711051073,
      "loss": 5.5067,
      "mean_token_accuracy": 0.17773209661245346,
      "num_tokens": 23242869.0,
      "step": 11470
    },
    {
      "entropy": 5.8551473140716555,
      "epoch": 1.0027964694573102,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004905973591014617,
      "loss": 5.4304,
      "mean_token_accuracy": 0.18736587464809418,
      "num_tokens": 23253653.0,
      "step": 11475
    },
    {
      "entropy": 5.823553562164307,
      "epoch": 1.0032334178100148,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905884429625459,
      "loss": 5.4367,
      "mean_token_accuracy": 0.18477417528629303,
      "num_tokens": 23263530.0,
      "step": 11480
    },
    {
      "entropy": 5.8986368656158445,
      "epoch": 1.0036703661627195,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905795226885308,
      "loss": 5.4915,
      "mean_token_accuracy": 0.1809002786874771,
      "num_tokens": 23274946.0,
      "step": 11485
    },
    {
      "entropy": 5.951287174224854,
      "epoch": 1.0041073145154242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905705982795874,
      "loss": 5.4817,
      "mean_token_accuracy": 0.18488264381885527,
      "num_tokens": 23283811.0,
      "step": 11490
    },
    {
      "entropy": 5.834171962738037,
      "epoch": 1.004544262868129,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490561669735887,
      "loss": 5.5111,
      "mean_token_accuracy": 0.18083866387605668,
      "num_tokens": 23294554.0,
      "step": 11495
    },
    {
      "entropy": 5.841756200790405,
      "epoch": 1.0049812112208336,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905527370576009,
      "loss": 5.3612,
      "mean_token_accuracy": 0.19127428978681565,
      "num_tokens": 23304245.0,
      "step": 11500
    },
    {
      "entropy": 5.971625185012817,
      "epoch": 1.0054181595735383,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905438002449004,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1740695834159851,
      "num_tokens": 23315169.0,
      "step": 11505
    },
    {
      "entropy": 5.846873044967651,
      "epoch": 1.0058551079262432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905348592979567,
      "loss": 5.4124,
      "mean_token_accuracy": 0.1898711383342743,
      "num_tokens": 23325528.0,
      "step": 11510
    },
    {
      "entropy": 5.932880496978759,
      "epoch": 1.006292056278948,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004905259142169415,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1792541489005089,
      "num_tokens": 23336744.0,
      "step": 11515
    },
    {
      "entropy": 5.886677646636963,
      "epoch": 1.0067290046316526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004905169650020264,
      "loss": 5.4259,
      "mean_token_accuracy": 0.1892029419541359,
      "num_tokens": 23346092.0,
      "step": 11520
    },
    {
      "entropy": 5.829683876037597,
      "epoch": 1.0071659529843573,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905080116533831,
      "loss": 5.4323,
      "mean_token_accuracy": 0.18020815700292586,
      "num_tokens": 23357326.0,
      "step": 11525
    },
    {
      "entropy": 5.864407730102539,
      "epoch": 1.007602901337062,
      "grad_norm": 0.921875,
      "learning_rate": 0.000490499054171183,
      "loss": 5.4655,
      "mean_token_accuracy": 0.1850619226694107,
      "num_tokens": 23367789.0,
      "step": 11530
    },
    {
      "entropy": 5.921274662017822,
      "epoch": 1.0080398496897667,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904900925555982,
      "loss": 5.4644,
      "mean_token_accuracy": 0.1834110736846924,
      "num_tokens": 23377499.0,
      "step": 11535
    },
    {
      "entropy": 5.926740026473999,
      "epoch": 1.0084767980424714,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004904811268068005,
      "loss": 5.4403,
      "mean_token_accuracy": 0.17745636850595475,
      "num_tokens": 23387867.0,
      "step": 11540
    },
    {
      "entropy": 5.858212947845459,
      "epoch": 1.008913746395176,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004904721569249618,
      "loss": 5.5366,
      "mean_token_accuracy": 0.17533748894929885,
      "num_tokens": 23398179.0,
      "step": 11545
    },
    {
      "entropy": 5.935226774215698,
      "epoch": 1.0093506947478808,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904631829102543,
      "loss": 5.536,
      "mean_token_accuracy": 0.18613833487033843,
      "num_tokens": 23407407.0,
      "step": 11550
    },
    {
      "entropy": 5.91339864730835,
      "epoch": 1.0097876431005854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904542047628501,
      "loss": 5.3736,
      "mean_token_accuracy": 0.19325260072946548,
      "num_tokens": 23417324.0,
      "step": 11555
    },
    {
      "entropy": 5.885081720352173,
      "epoch": 1.0102245914532901,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904452224829212,
      "loss": 5.4303,
      "mean_token_accuracy": 0.1875145897269249,
      "num_tokens": 23427157.0,
      "step": 11560
    },
    {
      "entropy": 5.910502767562866,
      "epoch": 1.0106615398059948,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904362360706403,
      "loss": 5.5125,
      "mean_token_accuracy": 0.17979926615953445,
      "num_tokens": 23437712.0,
      "step": 11565
    },
    {
      "entropy": 5.919670152664184,
      "epoch": 1.0110984881586997,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904272455261793,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18295658975839615,
      "num_tokens": 23447949.0,
      "step": 11570
    },
    {
      "entropy": 5.842067670822144,
      "epoch": 1.0115354365114044,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004904182508497109,
      "loss": 5.41,
      "mean_token_accuracy": 0.18000745624303818,
      "num_tokens": 23457319.0,
      "step": 11575
    },
    {
      "entropy": 5.900247859954834,
      "epoch": 1.0119723848641091,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904092520414075,
      "loss": 5.5169,
      "mean_token_accuracy": 0.17866748422384263,
      "num_tokens": 23468245.0,
      "step": 11580
    },
    {
      "entropy": 5.955326795578003,
      "epoch": 1.0124093332168138,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004904002491014418,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1907447025179863,
      "num_tokens": 23477049.0,
      "step": 11585
    },
    {
      "entropy": 5.896306562423706,
      "epoch": 1.0128462815695185,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903912420299866,
      "loss": 5.5568,
      "mean_token_accuracy": 0.17288741171360017,
      "num_tokens": 23488574.0,
      "step": 11590
    },
    {
      "entropy": 5.920158004760742,
      "epoch": 1.0132832299222232,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903822308272144,
      "loss": 5.4765,
      "mean_token_accuracy": 0.1853225201368332,
      "num_tokens": 23499282.0,
      "step": 11595
    },
    {
      "entropy": 5.914318990707398,
      "epoch": 1.0137201782749279,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903732154932982,
      "loss": 5.5015,
      "mean_token_accuracy": 0.18940852135419844,
      "num_tokens": 23508777.0,
      "step": 11600
    },
    {
      "entropy": 5.878818416595459,
      "epoch": 1.0141571266276326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903641960284108,
      "loss": 5.4453,
      "mean_token_accuracy": 0.1856512874364853,
      "num_tokens": 23518818.0,
      "step": 11605
    },
    {
      "entropy": 5.882084465026855,
      "epoch": 1.0145940749803373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004903551724327254,
      "loss": 5.4633,
      "mean_token_accuracy": 0.1856216236948967,
      "num_tokens": 23529668.0,
      "step": 11610
    },
    {
      "entropy": 5.957795286178589,
      "epoch": 1.015031023333042,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903461447064147,
      "loss": 5.4328,
      "mean_token_accuracy": 0.18890648186206818,
      "num_tokens": 23540418.0,
      "step": 11615
    },
    {
      "entropy": 5.914322090148926,
      "epoch": 1.0154679716857467,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004903371128496523,
      "loss": 5.4377,
      "mean_token_accuracy": 0.18363402485847474,
      "num_tokens": 23550035.0,
      "step": 11620
    },
    {
      "entropy": 5.877568626403809,
      "epoch": 1.0159049200384516,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903280768626113,
      "loss": 5.4977,
      "mean_token_accuracy": 0.18040652424097062,
      "num_tokens": 23559487.0,
      "step": 11625
    },
    {
      "entropy": 5.901291465759277,
      "epoch": 1.0163418683911563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004903190367454648,
      "loss": 5.4128,
      "mean_token_accuracy": 0.19097988158464432,
      "num_tokens": 23569146.0,
      "step": 11630
    },
    {
      "entropy": 5.827565383911133,
      "epoch": 1.016778816743861,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004903099924983864,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18649528920650482,
      "num_tokens": 23579458.0,
      "step": 11635
    },
    {
      "entropy": 5.86078290939331,
      "epoch": 1.0172157650965656,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903009441215495,
      "loss": 5.4205,
      "mean_token_accuracy": 0.18631469905376435,
      "num_tokens": 23589416.0,
      "step": 11640
    },
    {
      "entropy": 5.906558227539063,
      "epoch": 1.0176527134492703,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004902918916151278,
      "loss": 5.5162,
      "mean_token_accuracy": 0.1851527363061905,
      "num_tokens": 23599588.0,
      "step": 11645
    },
    {
      "entropy": 5.932866716384888,
      "epoch": 1.018089661801975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004902828349792947,
      "loss": 5.4909,
      "mean_token_accuracy": 0.18301643580198287,
      "num_tokens": 23610116.0,
      "step": 11650
    },
    {
      "entropy": 5.857520532608032,
      "epoch": 1.0185266101546797,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490273774214224,
      "loss": 5.4058,
      "mean_token_accuracy": 0.19086018800735474,
      "num_tokens": 23619906.0,
      "step": 11655
    },
    {
      "entropy": 5.912260341644287,
      "epoch": 1.0189635585073844,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902647093200896,
      "loss": 5.4612,
      "mean_token_accuracy": 0.18651404976844788,
      "num_tokens": 23630351.0,
      "step": 11660
    },
    {
      "entropy": 5.906693363189698,
      "epoch": 1.019400506860089,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004902556402970652,
      "loss": 5.5195,
      "mean_token_accuracy": 0.18198747634887696,
      "num_tokens": 23639696.0,
      "step": 11665
    },
    {
      "entropy": 5.904260206222534,
      "epoch": 1.0198374552127938,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902465671453248,
      "loss": 5.4842,
      "mean_token_accuracy": 0.18192328065633773,
      "num_tokens": 23649388.0,
      "step": 11670
    },
    {
      "entropy": 5.972312164306641,
      "epoch": 1.0202744035654985,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004902374898650425,
      "loss": 5.5097,
      "mean_token_accuracy": 0.17963872849941254,
      "num_tokens": 23659125.0,
      "step": 11675
    },
    {
      "entropy": 6.008585596084595,
      "epoch": 1.0207113519182032,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004902284084563924,
      "loss": 5.4917,
      "mean_token_accuracy": 0.17945277392864228,
      "num_tokens": 23668817.0,
      "step": 11680
    },
    {
      "entropy": 5.88599739074707,
      "epoch": 1.021148300270908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902193229195487,
      "loss": 5.4909,
      "mean_token_accuracy": 0.1827844575047493,
      "num_tokens": 23679378.0,
      "step": 11685
    },
    {
      "entropy": 5.916742992401123,
      "epoch": 1.0215852486236128,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902102332546855,
      "loss": 5.4576,
      "mean_token_accuracy": 0.1828152135014534,
      "num_tokens": 23689400.0,
      "step": 11690
    },
    {
      "entropy": 6.002728509902954,
      "epoch": 1.0220221969763175,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004902011394619772,
      "loss": 5.5886,
      "mean_token_accuracy": 0.17276642471551895,
      "num_tokens": 23699503.0,
      "step": 11695
    },
    {
      "entropy": 5.902543449401856,
      "epoch": 1.0224591453290222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004901920415415985,
      "loss": 5.5089,
      "mean_token_accuracy": 0.1779915675520897,
      "num_tokens": 23710913.0,
      "step": 11700
    },
    {
      "entropy": 5.849460077285767,
      "epoch": 1.0228960936817268,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004901829394937234,
      "loss": 5.4224,
      "mean_token_accuracy": 0.18690364062786102,
      "num_tokens": 23720730.0,
      "step": 11705
    },
    {
      "entropy": 5.918215560913086,
      "epoch": 1.0233330420344315,
      "grad_norm": 0.96875,
      "learning_rate": 0.000490173833318527,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1790463224053383,
      "num_tokens": 23730902.0,
      "step": 11710
    },
    {
      "entropy": 5.8863818645477295,
      "epoch": 1.0237699903871362,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901647230161837,
      "loss": 5.517,
      "mean_token_accuracy": 0.17904226928949357,
      "num_tokens": 23741901.0,
      "step": 11715
    },
    {
      "entropy": 5.9155456066131595,
      "epoch": 1.024206938739841,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004901556085868681,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1879327565431595,
      "num_tokens": 23752029.0,
      "step": 11720
    },
    {
      "entropy": 5.843714809417724,
      "epoch": 1.0246438870925456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004901464900307553,
      "loss": 5.4243,
      "mean_token_accuracy": 0.19069891571998596,
      "num_tokens": 23762125.0,
      "step": 11725
    },
    {
      "entropy": 5.850928544998169,
      "epoch": 1.0250808354452503,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901373673480201,
      "loss": 5.462,
      "mean_token_accuracy": 0.18915123790502547,
      "num_tokens": 23772056.0,
      "step": 11730
    },
    {
      "entropy": 5.890672588348389,
      "epoch": 1.025517783797955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004901282405388375,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1897692710161209,
      "num_tokens": 23781441.0,
      "step": 11735
    },
    {
      "entropy": 5.879596281051636,
      "epoch": 1.0259547321506597,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004901191096033825,
      "loss": 5.4069,
      "mean_token_accuracy": 0.1855199545621872,
      "num_tokens": 23791472.0,
      "step": 11740
    },
    {
      "entropy": 5.897236728668213,
      "epoch": 1.0263916805033646,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901099745418302,
      "loss": 5.4351,
      "mean_token_accuracy": 0.17926075607538222,
      "num_tokens": 23801940.0,
      "step": 11745
    },
    {
      "entropy": 5.846773862838745,
      "epoch": 1.0268286288560693,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004901008353543559,
      "loss": 5.3766,
      "mean_token_accuracy": 0.19364284723997116,
      "num_tokens": 23811515.0,
      "step": 11750
    },
    {
      "entropy": 5.840851783752441,
      "epoch": 1.027265577208774,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004900916920411349,
      "loss": 5.4654,
      "mean_token_accuracy": 0.1782319501042366,
      "num_tokens": 23821871.0,
      "step": 11755
    },
    {
      "entropy": 5.886299991607666,
      "epoch": 1.0277025255614787,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004900825446023426,
      "loss": 5.407,
      "mean_token_accuracy": 0.1923300176858902,
      "num_tokens": 23831677.0,
      "step": 11760
    },
    {
      "entropy": 5.899836587905884,
      "epoch": 1.0281394739141834,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900733930381543,
      "loss": 5.4484,
      "mean_token_accuracy": 0.18579370379447938,
      "num_tokens": 23841415.0,
      "step": 11765
    },
    {
      "entropy": 5.924854898452759,
      "epoch": 1.028576422266888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900642373487455,
      "loss": 5.5149,
      "mean_token_accuracy": 0.18254859000444412,
      "num_tokens": 23851809.0,
      "step": 11770
    },
    {
      "entropy": 5.948577785491944,
      "epoch": 1.0290133706195927,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004900550775342921,
      "loss": 5.457,
      "mean_token_accuracy": 0.1772794708609581,
      "num_tokens": 23861981.0,
      "step": 11775
    },
    {
      "entropy": 5.854751300811768,
      "epoch": 1.0294503189722974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004900459135949697,
      "loss": 5.4255,
      "mean_token_accuracy": 0.18492085635662078,
      "num_tokens": 23872437.0,
      "step": 11780
    },
    {
      "entropy": 5.863952589035034,
      "epoch": 1.0298872673250021,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004900367455309537,
      "loss": 5.4843,
      "mean_token_accuracy": 0.1790282443165779,
      "num_tokens": 23881650.0,
      "step": 11785
    },
    {
      "entropy": 5.896143531799316,
      "epoch": 1.0303242156777068,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900275733424205,
      "loss": 5.4685,
      "mean_token_accuracy": 0.1839817062020302,
      "num_tokens": 23891631.0,
      "step": 11790
    },
    {
      "entropy": 5.863079023361206,
      "epoch": 1.0307611640304115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900183970295457,
      "loss": 5.4223,
      "mean_token_accuracy": 0.18929210007190705,
      "num_tokens": 23901725.0,
      "step": 11795
    },
    {
      "entropy": 5.881108999252319,
      "epoch": 1.0311981123831164,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900092165925053,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18038008660078048,
      "num_tokens": 23911850.0,
      "step": 11800
    },
    {
      "entropy": 5.942317724227905,
      "epoch": 1.0316350607358211,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004900000320314757,
      "loss": 5.5382,
      "mean_token_accuracy": 0.1756463125348091,
      "num_tokens": 23922571.0,
      "step": 11805
    },
    {
      "entropy": 5.88503885269165,
      "epoch": 1.0320720090885258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899908433466326,
      "loss": 5.436,
      "mean_token_accuracy": 0.18777815103530884,
      "num_tokens": 23932637.0,
      "step": 11810
    },
    {
      "entropy": 5.969953107833862,
      "epoch": 1.0325089574412305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899816505381525,
      "loss": 5.5376,
      "mean_token_accuracy": 0.18061421662569047,
      "num_tokens": 23943288.0,
      "step": 11815
    },
    {
      "entropy": 5.914894866943359,
      "epoch": 1.0329459057939352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899724536062117,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18086837679147721,
      "num_tokens": 23952853.0,
      "step": 11820
    },
    {
      "entropy": 5.849908018112183,
      "epoch": 1.0333828541466399,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899632525509867,
      "loss": 5.4835,
      "mean_token_accuracy": 0.17847428768873214,
      "num_tokens": 23963111.0,
      "step": 11825
    },
    {
      "entropy": 5.868580913543701,
      "epoch": 1.0338198024993446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899540473726538,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1938989907503128,
      "num_tokens": 23972468.0,
      "step": 11830
    },
    {
      "entropy": 5.953849649429321,
      "epoch": 1.0342567508520493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004899448380713897,
      "loss": 5.4356,
      "mean_token_accuracy": 0.18786438107490538,
      "num_tokens": 23982567.0,
      "step": 11835
    },
    {
      "entropy": 5.86208062171936,
      "epoch": 1.034693699204754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899356246473709,
      "loss": 5.4306,
      "mean_token_accuracy": 0.1883201628923416,
      "num_tokens": 23993283.0,
      "step": 11840
    },
    {
      "entropy": 5.827916288375855,
      "epoch": 1.0351306475574586,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004899264071007741,
      "loss": 5.4048,
      "mean_token_accuracy": 0.1868772402405739,
      "num_tokens": 24004504.0,
      "step": 11845
    },
    {
      "entropy": 5.883149003982544,
      "epoch": 1.0355675959101633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899171854317764,
      "loss": 5.4729,
      "mean_token_accuracy": 0.18332068175077437,
      "num_tokens": 24014557.0,
      "step": 11850
    },
    {
      "entropy": 5.889545631408692,
      "epoch": 1.036004544262868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899079596405543,
      "loss": 5.4589,
      "mean_token_accuracy": 0.17968026399612427,
      "num_tokens": 24024495.0,
      "step": 11855
    },
    {
      "entropy": 5.869028377532959,
      "epoch": 1.036441492615573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489898729727285,
      "loss": 5.4568,
      "mean_token_accuracy": 0.18376406878232956,
      "num_tokens": 24034984.0,
      "step": 11860
    },
    {
      "entropy": 5.879584693908692,
      "epoch": 1.0368784409682776,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004898894956921455,
      "loss": 5.4086,
      "mean_token_accuracy": 0.18835099935531616,
      "num_tokens": 24044379.0,
      "step": 11865
    },
    {
      "entropy": 5.92590365409851,
      "epoch": 1.0373153893209823,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004898802575353129,
      "loss": 5.4699,
      "mean_token_accuracy": 0.18236050456762315,
      "num_tokens": 24054133.0,
      "step": 11870
    },
    {
      "entropy": 5.820121765136719,
      "epoch": 1.037752337673687,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004898710152569642,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18652730882167817,
      "num_tokens": 24064868.0,
      "step": 11875
    },
    {
      "entropy": 5.917542219161987,
      "epoch": 1.0381892860263917,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004898617688572769,
      "loss": 5.462,
      "mean_token_accuracy": 0.18240187168121338,
      "num_tokens": 24074977.0,
      "step": 11880
    },
    {
      "entropy": 5.918176317214966,
      "epoch": 1.0386262343790964,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004898525183364284,
      "loss": 5.5113,
      "mean_token_accuracy": 0.17984176576137542,
      "num_tokens": 24085673.0,
      "step": 11885
    },
    {
      "entropy": 5.926577472686768,
      "epoch": 1.039063182731801,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004898432636945959,
      "loss": 5.4867,
      "mean_token_accuracy": 0.18342998921871184,
      "num_tokens": 24095465.0,
      "step": 11890
    },
    {
      "entropy": 5.831007957458496,
      "epoch": 1.0395001310845058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000489834004931957,
      "loss": 5.435,
      "mean_token_accuracy": 0.18487913608551027,
      "num_tokens": 24105576.0,
      "step": 11895
    },
    {
      "entropy": 5.906735515594482,
      "epoch": 1.0399370794372105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898247420486895,
      "loss": 5.4259,
      "mean_token_accuracy": 0.18272297382354735,
      "num_tokens": 24114998.0,
      "step": 11900
    },
    {
      "entropy": 5.883747959136963,
      "epoch": 1.0403740277899152,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898154750449707,
      "loss": 5.4187,
      "mean_token_accuracy": 0.19017506539821624,
      "num_tokens": 24125059.0,
      "step": 11905
    },
    {
      "entropy": 5.861986112594605,
      "epoch": 1.0408109761426199,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004898062039209786,
      "loss": 5.4496,
      "mean_token_accuracy": 0.19033259600400926,
      "num_tokens": 24134062.0,
      "step": 11910
    },
    {
      "entropy": 5.925160455703735,
      "epoch": 1.0412479244953246,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000489796928676891,
      "loss": 5.471,
      "mean_token_accuracy": 0.18420426100492476,
      "num_tokens": 24144016.0,
      "step": 11915
    },
    {
      "entropy": 5.912639236450195,
      "epoch": 1.0416848728480295,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897876493128856,
      "loss": 5.4359,
      "mean_token_accuracy": 0.18811661303043364,
      "num_tokens": 24154337.0,
      "step": 11920
    },
    {
      "entropy": 5.850432348251343,
      "epoch": 1.0421218212007342,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004897783658291406,
      "loss": 5.414,
      "mean_token_accuracy": 0.1881126880645752,
      "num_tokens": 24164440.0,
      "step": 11925
    },
    {
      "entropy": 5.913686800003052,
      "epoch": 1.0425587695534388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489769078225834,
      "loss": 5.5962,
      "mean_token_accuracy": 0.17447670102119445,
      "num_tokens": 24174388.0,
      "step": 11930
    },
    {
      "entropy": 5.8806956768035885,
      "epoch": 1.0429957179061435,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004897597865031441,
      "loss": 5.3977,
      "mean_token_accuracy": 0.18701106011867524,
      "num_tokens": 24183543.0,
      "step": 11935
    },
    {
      "entropy": 5.867287302017212,
      "epoch": 1.0434326662588482,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004897504906612487,
      "loss": 5.4564,
      "mean_token_accuracy": 0.18627522736787797,
      "num_tokens": 24193847.0,
      "step": 11940
    },
    {
      "entropy": 5.86517105102539,
      "epoch": 1.043869614611553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004897411907003265,
      "loss": 5.4419,
      "mean_token_accuracy": 0.1822678864002228,
      "num_tokens": 24203072.0,
      "step": 11945
    },
    {
      "entropy": 5.873770046234131,
      "epoch": 1.0443065629642576,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897318866205558,
      "loss": 5.3695,
      "mean_token_accuracy": 0.19114270359277724,
      "num_tokens": 24213610.0,
      "step": 11950
    },
    {
      "entropy": 5.817074489593506,
      "epoch": 1.0447435113169623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897225784221149,
      "loss": 5.4353,
      "mean_token_accuracy": 0.189871583878994,
      "num_tokens": 24223571.0,
      "step": 11955
    },
    {
      "entropy": 5.859742736816406,
      "epoch": 1.045180459669667,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897132661051826,
      "loss": 5.4688,
      "mean_token_accuracy": 0.1816618800163269,
      "num_tokens": 24233867.0,
      "step": 11960
    },
    {
      "entropy": 5.876847600936889,
      "epoch": 1.0456174080223717,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004897039496699372,
      "loss": 5.4778,
      "mean_token_accuracy": 0.1839870259165764,
      "num_tokens": 24243853.0,
      "step": 11965
    },
    {
      "entropy": 5.8494175434112545,
      "epoch": 1.0460543563750764,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004896946291165575,
      "loss": 5.4228,
      "mean_token_accuracy": 0.18227136582136155,
      "num_tokens": 24254049.0,
      "step": 11970
    },
    {
      "entropy": 5.842693901062011,
      "epoch": 1.0464913047277813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004896853044452224,
      "loss": 5.4,
      "mean_token_accuracy": 0.1889360547065735,
      "num_tokens": 24265166.0,
      "step": 11975
    },
    {
      "entropy": 5.901191186904907,
      "epoch": 1.046928253080486,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004896759756561106,
      "loss": 5.4345,
      "mean_token_accuracy": 0.18943883180618287,
      "num_tokens": 24274981.0,
      "step": 11980
    },
    {
      "entropy": 5.9355250835418705,
      "epoch": 1.0473652014331907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004896666427494012,
      "loss": 5.4748,
      "mean_token_accuracy": 0.18587528318166732,
      "num_tokens": 24285009.0,
      "step": 11985
    },
    {
      "entropy": 5.944714117050171,
      "epoch": 1.0478021497858954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489657305725273,
      "loss": 5.5284,
      "mean_token_accuracy": 0.17514412850141525,
      "num_tokens": 24294669.0,
      "step": 11990
    },
    {
      "entropy": 5.924590826034546,
      "epoch": 1.0482390981386,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004896479645839052,
      "loss": 5.5003,
      "mean_token_accuracy": 0.1818630263209343,
      "num_tokens": 24304647.0,
      "step": 11995
    },
    {
      "entropy": 5.858507442474365,
      "epoch": 1.0486760464913047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004896386193254771,
      "loss": 5.3837,
      "mean_token_accuracy": 0.1892987534403801,
      "num_tokens": 24314266.0,
      "step": 12000
    },
    {
      "epoch": 1.0486760464913047,
      "eval_entropy": 5.610042168817382,
      "eval_loss": 5.526420593261719,
      "eval_mean_token_accuracy": 0.18946720602569297,
      "eval_num_tokens": 24314266.0,
      "eval_runtime": 21.4504,
      "eval_samples_per_second": 1725.562,
      "eval_steps_per_second": 215.707,
      "step": 12000
    }
  ],
  "logging_steps": 5,
  "max_steps": 114420,
  "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": 5405040117043200.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}