swa-latn-10mb-10mb_seed10/checkpoint-14000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.958156629912355,
  "eval_steps": 1000,
  "global_step": 14000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.74207363128662,
      "epoch": 0.0014136273678258412,
      "grad_norm": 6.1875,
      "learning_rate": 2e-06,
      "loss": 10.5511,
      "mean_token_accuracy": 0.0,
      "num_tokens": 4340.0,
      "step": 5
    },
    {
      "entropy": 10.742022514343262,
      "epoch": 0.0028272547356516823,
      "grad_norm": 6.71875,
      "learning_rate": 4.5e-06,
      "loss": 10.5265,
      "mean_token_accuracy": 0.0,
      "num_tokens": 8692.0,
      "step": 10
    },
    {
      "entropy": 10.742014980316162,
      "epoch": 0.004240882103477523,
      "grad_norm": 6.78125,
      "learning_rate": 7e-06,
      "loss": 10.4477,
      "mean_token_accuracy": 0.0,
      "num_tokens": 12453.0,
      "step": 15
    },
    {
      "entropy": 10.742022132873535,
      "epoch": 0.005654509471303365,
      "grad_norm": 6.71875,
      "learning_rate": 9.5e-06,
      "loss": 10.4201,
      "mean_token_accuracy": 0.0,
      "num_tokens": 16439.0,
      "step": 20
    },
    {
      "entropy": 10.741988182067871,
      "epoch": 0.007068136839129205,
      "grad_norm": 6.8125,
      "learning_rate": 1.2e-05,
      "loss": 10.295,
      "mean_token_accuracy": 0.0014771121321246028,
      "num_tokens": 20515.0,
      "step": 25
    },
    {
      "entropy": 10.74176025390625,
      "epoch": 0.008481764206955046,
      "grad_norm": 5.5,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.1401,
      "mean_token_accuracy": 0.029323107562959194,
      "num_tokens": 24336.0,
      "step": 30
    },
    {
      "entropy": 10.741047763824463,
      "epoch": 0.009895391574780888,
      "grad_norm": 4.5625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.0398,
      "mean_token_accuracy": 0.05071598924696445,
      "num_tokens": 28465.0,
      "step": 35
    },
    {
      "entropy": 10.739372825622558,
      "epoch": 0.01130901894260673,
      "grad_norm": 3.90625,
      "learning_rate": 1.95e-05,
      "loss": 9.828,
      "mean_token_accuracy": 0.06147046238183975,
      "num_tokens": 32217.0,
      "step": 40
    },
    {
      "entropy": 10.73618106842041,
      "epoch": 0.01272264631043257,
      "grad_norm": 3.015625,
      "learning_rate": 2.2e-05,
      "loss": 9.7393,
      "mean_token_accuracy": 0.05857079476118088,
      "num_tokens": 36369.0,
      "step": 45
    },
    {
      "entropy": 10.732595252990723,
      "epoch": 0.01413627367825841,
      "grad_norm": 2.859375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.6251,
      "mean_token_accuracy": 0.0617414329200983,
      "num_tokens": 40227.0,
      "step": 50
    },
    {
      "entropy": 10.730756187438965,
      "epoch": 0.015549901046084252,
      "grad_norm": 2.5,
      "learning_rate": 2.7e-05,
      "loss": 9.6041,
      "mean_token_accuracy": 0.058244453743100166,
      "num_tokens": 44266.0,
      "step": 55
    },
    {
      "entropy": 10.728981113433838,
      "epoch": 0.016963528413910092,
      "grad_norm": 2.640625,
      "learning_rate": 2.95e-05,
      "loss": 9.5679,
      "mean_token_accuracy": 0.06274866946041584,
      "num_tokens": 48187.0,
      "step": 60
    },
    {
      "entropy": 10.726116180419922,
      "epoch": 0.018377155781735936,
      "grad_norm": 2.453125,
      "learning_rate": 3.2e-05,
      "loss": 9.4557,
      "mean_token_accuracy": 0.05981031022965908,
      "num_tokens": 52252.0,
      "step": 65
    },
    {
      "entropy": 10.721884059906007,
      "epoch": 0.019790783149561775,
      "grad_norm": 2.53125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.4336,
      "mean_token_accuracy": 0.05559282228350639,
      "num_tokens": 56421.0,
      "step": 70
    },
    {
      "entropy": 10.7160475730896,
      "epoch": 0.021204410517387615,
      "grad_norm": 2.421875,
      "learning_rate": 3.7e-05,
      "loss": 9.3486,
      "mean_token_accuracy": 0.06220139227807522,
      "num_tokens": 60341.0,
      "step": 75
    },
    {
      "entropy": 10.70057315826416,
      "epoch": 0.02261803788521346,
      "grad_norm": 2.421875,
      "learning_rate": 3.95e-05,
      "loss": 9.216,
      "mean_token_accuracy": 0.06268311068415641,
      "num_tokens": 64465.0,
      "step": 80
    },
    {
      "entropy": 10.682071495056153,
      "epoch": 0.0240316652530393,
      "grad_norm": 2.21875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.1917,
      "mean_token_accuracy": 0.06404139809310436,
      "num_tokens": 68514.0,
      "step": 85
    },
    {
      "entropy": 10.654492664337159,
      "epoch": 0.02544529262086514,
      "grad_norm": 2.265625,
      "learning_rate": 4.45e-05,
      "loss": 9.1028,
      "mean_token_accuracy": 0.06569066680967808,
      "num_tokens": 72661.0,
      "step": 90
    },
    {
      "entropy": 10.621818923950196,
      "epoch": 0.02685891998869098,
      "grad_norm": 2.203125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.9739,
      "mean_token_accuracy": 0.07454013973474502,
      "num_tokens": 76698.0,
      "step": 95
    },
    {
      "entropy": 10.587986373901368,
      "epoch": 0.02827254735651682,
      "grad_norm": 2.109375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.8935,
      "mean_token_accuracy": 0.07267784662544727,
      "num_tokens": 80732.0,
      "step": 100
    },
    {
      "entropy": 10.545151424407958,
      "epoch": 0.029686174724342665,
      "grad_norm": 2.15625,
      "learning_rate": 5.2e-05,
      "loss": 8.8022,
      "mean_token_accuracy": 0.06954946778714657,
      "num_tokens": 84797.0,
      "step": 105
    },
    {
      "entropy": 10.516861534118652,
      "epoch": 0.031099802092168505,
      "grad_norm": 2.234375,
      "learning_rate": 5.45e-05,
      "loss": 8.722,
      "mean_token_accuracy": 0.07358665019273758,
      "num_tokens": 88758.0,
      "step": 110
    },
    {
      "entropy": 10.459677600860596,
      "epoch": 0.032513429459994345,
      "grad_norm": 1.953125,
      "learning_rate": 5.7e-05,
      "loss": 8.6039,
      "mean_token_accuracy": 0.07231484800577163,
      "num_tokens": 92722.0,
      "step": 115
    },
    {
      "entropy": 10.407901668548584,
      "epoch": 0.033927056827820185,
      "grad_norm": 2.125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.4476,
      "mean_token_accuracy": 0.07675158120691776,
      "num_tokens": 96602.0,
      "step": 120
    },
    {
      "entropy": 10.325055503845215,
      "epoch": 0.035340684195646024,
      "grad_norm": 1.9453125,
      "learning_rate": 6.2e-05,
      "loss": 8.3606,
      "mean_token_accuracy": 0.07958476431667805,
      "num_tokens": 100598.0,
      "step": 125
    },
    {
      "entropy": 10.222711753845214,
      "epoch": 0.03675431156347187,
      "grad_norm": 1.875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.2593,
      "mean_token_accuracy": 0.07609154582023621,
      "num_tokens": 104460.0,
      "step": 130
    },
    {
      "entropy": 10.145068645477295,
      "epoch": 0.03816793893129771,
      "grad_norm": 2.109375,
      "learning_rate": 6.7e-05,
      "loss": 8.1373,
      "mean_token_accuracy": 0.07444728389382363,
      "num_tokens": 108360.0,
      "step": 135
    },
    {
      "entropy": 10.019375801086426,
      "epoch": 0.03958156629912355,
      "grad_norm": 1.5703125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.9991,
      "mean_token_accuracy": 0.07835616618394851,
      "num_tokens": 112176.0,
      "step": 140
    },
    {
      "entropy": 9.877924728393555,
      "epoch": 0.04099519366694939,
      "grad_norm": 1.671875,
      "learning_rate": 7.2e-05,
      "loss": 8.0465,
      "mean_token_accuracy": 0.07128159664571285,
      "num_tokens": 116192.0,
      "step": 145
    },
    {
      "entropy": 9.766163635253907,
      "epoch": 0.04240882103477523,
      "grad_norm": 1.5390625,
      "learning_rate": 7.45e-05,
      "loss": 7.9125,
      "mean_token_accuracy": 0.07681304663419723,
      "num_tokens": 120221.0,
      "step": 150
    },
    {
      "entropy": 9.596385955810547,
      "epoch": 0.04382244840260108,
      "grad_norm": 1.3515625,
      "learning_rate": 7.7e-05,
      "loss": 7.8471,
      "mean_token_accuracy": 0.07160197533667087,
      "num_tokens": 124364.0,
      "step": 155
    },
    {
      "entropy": 9.397625064849853,
      "epoch": 0.04523607577042692,
      "grad_norm": 1.453125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.6486,
      "mean_token_accuracy": 0.08006830923259259,
      "num_tokens": 128266.0,
      "step": 160
    },
    {
      "entropy": 9.197708892822266,
      "epoch": 0.04664970313825276,
      "grad_norm": 1.234375,
      "learning_rate": 8.2e-05,
      "loss": 7.5695,
      "mean_token_accuracy": 0.08508073762059212,
      "num_tokens": 131985.0,
      "step": 165
    },
    {
      "entropy": 9.003838157653808,
      "epoch": 0.0480633305060786,
      "grad_norm": 1.1640625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.4586,
      "mean_token_accuracy": 0.08466509208083153,
      "num_tokens": 135892.0,
      "step": 170
    },
    {
      "entropy": 8.81080961227417,
      "epoch": 0.04947695787390444,
      "grad_norm": 1.28125,
      "learning_rate": 8.7e-05,
      "loss": 7.5228,
      "mean_token_accuracy": 0.075396229326725,
      "num_tokens": 139937.0,
      "step": 175
    },
    {
      "entropy": 8.659114837646484,
      "epoch": 0.05089058524173028,
      "grad_norm": 1.1484375,
      "learning_rate": 8.95e-05,
      "loss": 7.3426,
      "mean_token_accuracy": 0.08097245395183564,
      "num_tokens": 143659.0,
      "step": 180
    },
    {
      "entropy": 8.553722858428955,
      "epoch": 0.052304212609556124,
      "grad_norm": 1.3515625,
      "learning_rate": 9.2e-05,
      "loss": 7.4259,
      "mean_token_accuracy": 0.08198288567364216,
      "num_tokens": 147394.0,
      "step": 185
    },
    {
      "entropy": 8.326274490356445,
      "epoch": 0.05371783997738196,
      "grad_norm": 1.1015625,
      "learning_rate": 9.45e-05,
      "loss": 7.2628,
      "mean_token_accuracy": 0.08642936870455742,
      "num_tokens": 151082.0,
      "step": 190
    },
    {
      "entropy": 8.238064765930176,
      "epoch": 0.0551314673452078,
      "grad_norm": 1.359375,
      "learning_rate": 9.7e-05,
      "loss": 7.25,
      "mean_token_accuracy": 0.08761365041136741,
      "num_tokens": 155099.0,
      "step": 195
    },
    {
      "entropy": 8.188647174835205,
      "epoch": 0.05654509471303364,
      "grad_norm": 0.9921875,
      "learning_rate": 9.95e-05,
      "loss": 7.4052,
      "mean_token_accuracy": 0.08035821095108986,
      "num_tokens": 159297.0,
      "step": 200
    },
    {
      "entropy": 8.190512943267823,
      "epoch": 0.05795872208085948,
      "grad_norm": 1.21875,
      "learning_rate": 0.000102,
      "loss": 7.268,
      "mean_token_accuracy": 0.08253459706902504,
      "num_tokens": 163337.0,
      "step": 205
    },
    {
      "entropy": 8.12247543334961,
      "epoch": 0.05937234944868533,
      "grad_norm": 1.1875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.3087,
      "mean_token_accuracy": 0.08554964698851109,
      "num_tokens": 167440.0,
      "step": 210
    },
    {
      "entropy": 8.021088409423829,
      "epoch": 0.06078597681651117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000107,
      "loss": 7.1661,
      "mean_token_accuracy": 0.08840125799179077,
      "num_tokens": 171238.0,
      "step": 215
    },
    {
      "entropy": 8.036321210861207,
      "epoch": 0.06219960418433701,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001095,
      "loss": 7.2139,
      "mean_token_accuracy": 0.0908890649676323,
      "num_tokens": 175476.0,
      "step": 220
    },
    {
      "entropy": 7.951605987548828,
      "epoch": 0.06361323155216285,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000112,
      "loss": 7.1261,
      "mean_token_accuracy": 0.08661228194832801,
      "num_tokens": 179398.0,
      "step": 225
    },
    {
      "entropy": 7.849934148788452,
      "epoch": 0.06502685891998869,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001145,
      "loss": 7.1123,
      "mean_token_accuracy": 0.09166858419775963,
      "num_tokens": 183355.0,
      "step": 230
    },
    {
      "entropy": 7.945307731628418,
      "epoch": 0.06644048628781453,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.1135,
      "mean_token_accuracy": 0.08504104316234588,
      "num_tokens": 187278.0,
      "step": 235
    },
    {
      "entropy": 7.837193965911865,
      "epoch": 0.06785411365564037,
      "grad_norm": 1.484375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1333,
      "mean_token_accuracy": 0.09098687320947647,
      "num_tokens": 191136.0,
      "step": 240
    },
    {
      "entropy": 7.820778036117554,
      "epoch": 0.06926774102346621,
      "grad_norm": 1.671875,
      "learning_rate": 0.000122,
      "loss": 7.1943,
      "mean_token_accuracy": 0.0956549420952797,
      "num_tokens": 195096.0,
      "step": 245
    },
    {
      "entropy": 7.859644556045533,
      "epoch": 0.07068136839129205,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0001245,
      "loss": 7.1855,
      "mean_token_accuracy": 0.08993390724062919,
      "num_tokens": 199483.0,
      "step": 250
    },
    {
      "entropy": 7.814686632156372,
      "epoch": 0.0720949957591179,
      "grad_norm": 1.53125,
      "learning_rate": 0.000127,
      "loss": 7.1139,
      "mean_token_accuracy": 0.09194719940423965,
      "num_tokens": 203515.0,
      "step": 255
    },
    {
      "entropy": 7.748371124267578,
      "epoch": 0.07350862312694374,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001295,
      "loss": 7.116,
      "mean_token_accuracy": 0.09589750841259956,
      "num_tokens": 207451.0,
      "step": 260
    },
    {
      "entropy": 7.774532127380371,
      "epoch": 0.07492225049476958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000132,
      "loss": 7.0308,
      "mean_token_accuracy": 0.1008292630314827,
      "num_tokens": 211507.0,
      "step": 265
    },
    {
      "entropy": 7.691170597076416,
      "epoch": 0.07633587786259542,
      "grad_norm": 1.640625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.09,
      "mean_token_accuracy": 0.09460079073905944,
      "num_tokens": 215236.0,
      "step": 270
    },
    {
      "entropy": 7.701872444152832,
      "epoch": 0.07774950523042126,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.0577,
      "mean_token_accuracy": 0.10146123990416527,
      "num_tokens": 218934.0,
      "step": 275
    },
    {
      "entropy": 7.6655951023101805,
      "epoch": 0.0791631325982471,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0001395,
      "loss": 6.9994,
      "mean_token_accuracy": 0.10914008393883705,
      "num_tokens": 223185.0,
      "step": 280
    },
    {
      "entropy": 7.753876638412476,
      "epoch": 0.08057675996607294,
      "grad_norm": 1.5,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1232,
      "mean_token_accuracy": 0.09365936666727066,
      "num_tokens": 227044.0,
      "step": 285
    },
    {
      "entropy": 7.66315860748291,
      "epoch": 0.08199038733389878,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001445,
      "loss": 7.0308,
      "mean_token_accuracy": 0.10147744789719582,
      "num_tokens": 231359.0,
      "step": 290
    },
    {
      "entropy": 7.710978984832764,
      "epoch": 0.08340401470172462,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000147,
      "loss": 7.0018,
      "mean_token_accuracy": 0.11066402792930603,
      "num_tokens": 235224.0,
      "step": 295
    },
    {
      "entropy": 7.5690467834472654,
      "epoch": 0.08481764206955046,
      "grad_norm": 1.484375,
      "learning_rate": 0.0001495,
      "loss": 6.8653,
      "mean_token_accuracy": 0.11461045518517494,
      "num_tokens": 239119.0,
      "step": 300
    },
    {
      "entropy": 7.518357563018799,
      "epoch": 0.0862312694373763,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000152,
      "loss": 6.9653,
      "mean_token_accuracy": 0.10772728994488716,
      "num_tokens": 243275.0,
      "step": 305
    },
    {
      "entropy": 7.694993162155152,
      "epoch": 0.08764489680520215,
      "grad_norm": 1.4375,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9709,
      "mean_token_accuracy": 0.11304818466305733,
      "num_tokens": 247415.0,
      "step": 310
    },
    {
      "entropy": 7.528787612915039,
      "epoch": 0.089058524173028,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000157,
      "loss": 6.9654,
      "mean_token_accuracy": 0.10677048042416573,
      "num_tokens": 251575.0,
      "step": 315
    },
    {
      "entropy": 7.5503821849823,
      "epoch": 0.09047215154085383,
      "grad_norm": 1.65625,
      "learning_rate": 0.0001595,
      "loss": 6.9441,
      "mean_token_accuracy": 0.11381662338972091,
      "num_tokens": 255686.0,
      "step": 320
    },
    {
      "entropy": 7.599794197082519,
      "epoch": 0.09188577890867967,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000162,
      "loss": 6.9401,
      "mean_token_accuracy": 0.11095880568027497,
      "num_tokens": 259850.0,
      "step": 325
    },
    {
      "entropy": 7.6137940883636475,
      "epoch": 0.09329940627650551,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.9622,
      "mean_token_accuracy": 0.10780008137226105,
      "num_tokens": 264043.0,
      "step": 330
    },
    {
      "entropy": 7.65163402557373,
      "epoch": 0.09471303364433135,
      "grad_norm": 1.40625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0215,
      "mean_token_accuracy": 0.1125954583287239,
      "num_tokens": 268296.0,
      "step": 335
    },
    {
      "entropy": 7.420255088806153,
      "epoch": 0.0961266610121572,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.836,
      "mean_token_accuracy": 0.11534447073936463,
      "num_tokens": 272062.0,
      "step": 340
    },
    {
      "entropy": 7.355507898330688,
      "epoch": 0.09754028837998303,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8008,
      "mean_token_accuracy": 0.12810793220996858,
      "num_tokens": 275936.0,
      "step": 345
    },
    {
      "entropy": 7.469290637969971,
      "epoch": 0.09895391574780887,
      "grad_norm": 1.75,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.8213,
      "mean_token_accuracy": 0.1251584582030773,
      "num_tokens": 279526.0,
      "step": 350
    },
    {
      "entropy": 7.5056859970092775,
      "epoch": 0.10036754311563471,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000177,
      "loss": 6.8874,
      "mean_token_accuracy": 0.11884394437074661,
      "num_tokens": 283577.0,
      "step": 355
    },
    {
      "entropy": 7.507189226150513,
      "epoch": 0.10178117048346055,
      "grad_norm": 1.5,
      "learning_rate": 0.0001795,
      "loss": 6.7619,
      "mean_token_accuracy": 0.112952870875597,
      "num_tokens": 287377.0,
      "step": 360
    },
    {
      "entropy": 7.379469537734986,
      "epoch": 0.10319479785128641,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000182,
      "loss": 6.9101,
      "mean_token_accuracy": 0.1114804707467556,
      "num_tokens": 291582.0,
      "step": 365
    },
    {
      "entropy": 7.508454275131226,
      "epoch": 0.10460842521911225,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0001845,
      "loss": 6.7891,
      "mean_token_accuracy": 0.12308030501008034,
      "num_tokens": 295608.0,
      "step": 370
    },
    {
      "entropy": 7.4715001583099365,
      "epoch": 0.10602205258693809,
      "grad_norm": 1.671875,
      "learning_rate": 0.000187,
      "loss": 6.8013,
      "mean_token_accuracy": 0.12096530124545098,
      "num_tokens": 299660.0,
      "step": 375
    },
    {
      "entropy": 7.332260084152222,
      "epoch": 0.10743567995476393,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0001895,
      "loss": 6.7183,
      "mean_token_accuracy": 0.13139156624674797,
      "num_tokens": 303412.0,
      "step": 380
    },
    {
      "entropy": 7.269100999832153,
      "epoch": 0.10884930732258977,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000192,
      "loss": 6.7782,
      "mean_token_accuracy": 0.12529514580965043,
      "num_tokens": 307352.0,
      "step": 385
    },
    {
      "entropy": 7.428966236114502,
      "epoch": 0.1102629346904156,
      "grad_norm": 1.75,
      "learning_rate": 0.0001945,
      "loss": 6.8533,
      "mean_token_accuracy": 0.1281038463115692,
      "num_tokens": 311418.0,
      "step": 390
    },
    {
      "entropy": 7.324939203262329,
      "epoch": 0.11167656205824145,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.6783,
      "mean_token_accuracy": 0.13713634759187698,
      "num_tokens": 315150.0,
      "step": 395
    },
    {
      "entropy": 7.591144180297851,
      "epoch": 0.11309018942606729,
      "grad_norm": 1.6875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8727,
      "mean_token_accuracy": 0.11698746904730797,
      "num_tokens": 318898.0,
      "step": 400
    },
    {
      "entropy": 7.242686319351196,
      "epoch": 0.11450381679389313,
      "grad_norm": 1.484375,
      "learning_rate": 0.000202,
      "loss": 6.6935,
      "mean_token_accuracy": 0.13000213503837585,
      "num_tokens": 322495.0,
      "step": 405
    },
    {
      "entropy": 7.333932304382325,
      "epoch": 0.11591744416171897,
      "grad_norm": 1.375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8398,
      "mean_token_accuracy": 0.12843888401985168,
      "num_tokens": 326473.0,
      "step": 410
    },
    {
      "entropy": 7.408268976211548,
      "epoch": 0.1173310715295448,
      "grad_norm": 1.5625,
      "learning_rate": 0.000207,
      "loss": 6.7863,
      "mean_token_accuracy": 0.12803610786795616,
      "num_tokens": 330393.0,
      "step": 415
    },
    {
      "entropy": 7.3277123928070065,
      "epoch": 0.11874469889737066,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002095,
      "loss": 6.7054,
      "mean_token_accuracy": 0.13229965120553971,
      "num_tokens": 334449.0,
      "step": 420
    },
    {
      "entropy": 7.445288515090942,
      "epoch": 0.1201583262651965,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000212,
      "loss": 6.8181,
      "mean_token_accuracy": 0.11966062635183335,
      "num_tokens": 338740.0,
      "step": 425
    },
    {
      "entropy": 7.247825288772583,
      "epoch": 0.12157195363302234,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002145,
      "loss": 6.6722,
      "mean_token_accuracy": 0.13701122775673866,
      "num_tokens": 342429.0,
      "step": 430
    },
    {
      "entropy": 7.355695915222168,
      "epoch": 0.12298558100084818,
      "grad_norm": 1.5,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6994,
      "mean_token_accuracy": 0.13420747891068457,
      "num_tokens": 346218.0,
      "step": 435
    },
    {
      "entropy": 7.303587532043457,
      "epoch": 0.12439920836867402,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002195,
      "loss": 6.696,
      "mean_token_accuracy": 0.1381567671895027,
      "num_tokens": 350310.0,
      "step": 440
    },
    {
      "entropy": 7.251111030578613,
      "epoch": 0.12581283573649985,
      "grad_norm": 1.59375,
      "learning_rate": 0.000222,
      "loss": 6.6712,
      "mean_token_accuracy": 0.13844901993870734,
      "num_tokens": 353952.0,
      "step": 445
    },
    {
      "entropy": 7.367745399475098,
      "epoch": 0.1272264631043257,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002245,
      "loss": 6.851,
      "mean_token_accuracy": 0.1284424804151058,
      "num_tokens": 357667.0,
      "step": 450
    },
    {
      "entropy": 7.356730890274048,
      "epoch": 0.12864009047215155,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6561,
      "mean_token_accuracy": 0.138033926486969,
      "num_tokens": 361309.0,
      "step": 455
    },
    {
      "entropy": 7.101002645492554,
      "epoch": 0.13005371783997738,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6954,
      "mean_token_accuracy": 0.1395917169749737,
      "num_tokens": 365099.0,
      "step": 460
    },
    {
      "entropy": 7.374086189270019,
      "epoch": 0.13146734520780323,
      "grad_norm": 1.4375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7496,
      "mean_token_accuracy": 0.13763813823461532,
      "num_tokens": 369141.0,
      "step": 465
    },
    {
      "entropy": 7.359116935729981,
      "epoch": 0.13288097257562906,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6675,
      "mean_token_accuracy": 0.1343040108680725,
      "num_tokens": 373128.0,
      "step": 470
    },
    {
      "entropy": 7.0868512153625485,
      "epoch": 0.1342945999434549,
      "grad_norm": 1.53125,
      "learning_rate": 0.000237,
      "loss": 6.6654,
      "mean_token_accuracy": 0.14124030545353888,
      "num_tokens": 377103.0,
      "step": 475
    },
    {
      "entropy": 7.250053071975708,
      "epoch": 0.13570822731128074,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002395,
      "loss": 6.7219,
      "mean_token_accuracy": 0.13121307715773584,
      "num_tokens": 381573.0,
      "step": 480
    },
    {
      "entropy": 7.313079786300659,
      "epoch": 0.1371218546791066,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000242,
      "loss": 6.7674,
      "mean_token_accuracy": 0.13335683792829514,
      "num_tokens": 385759.0,
      "step": 485
    },
    {
      "entropy": 7.177998399734497,
      "epoch": 0.13853548204693242,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002445,
      "loss": 6.6308,
      "mean_token_accuracy": 0.14019295275211335,
      "num_tokens": 389700.0,
      "step": 490
    },
    {
      "entropy": 7.207444953918457,
      "epoch": 0.13994910941475827,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000247,
      "loss": 6.5526,
      "mean_token_accuracy": 0.1400491252541542,
      "num_tokens": 393279.0,
      "step": 495
    },
    {
      "entropy": 7.098467206954956,
      "epoch": 0.1413627367825841,
      "grad_norm": 1.625,
      "learning_rate": 0.0002495,
      "loss": 6.639,
      "mean_token_accuracy": 0.13535916954278945,
      "num_tokens": 397379.0,
      "step": 500
    },
    {
      "entropy": 7.206504201889038,
      "epoch": 0.14277636415040995,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000252,
      "loss": 6.6984,
      "mean_token_accuracy": 0.13445916026830673,
      "num_tokens": 401349.0,
      "step": 505
    },
    {
      "entropy": 7.222359657287598,
      "epoch": 0.1441899915182358,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002545,
      "loss": 6.6017,
      "mean_token_accuracy": 0.1456646129488945,
      "num_tokens": 405361.0,
      "step": 510
    },
    {
      "entropy": 7.192652940750122,
      "epoch": 0.14560361888606163,
      "grad_norm": 1.5,
      "learning_rate": 0.000257,
      "loss": 6.6375,
      "mean_token_accuracy": 0.14064312353730202,
      "num_tokens": 409614.0,
      "step": 515
    },
    {
      "entropy": 7.041819143295288,
      "epoch": 0.14701724625388748,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002595,
      "loss": 6.5656,
      "mean_token_accuracy": 0.14632121846079826,
      "num_tokens": 413529.0,
      "step": 520
    },
    {
      "entropy": 7.031795978546143,
      "epoch": 0.1484308736217133,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000262,
      "loss": 6.5605,
      "mean_token_accuracy": 0.14058664664626122,
      "num_tokens": 417365.0,
      "step": 525
    },
    {
      "entropy": 7.1277703762054445,
      "epoch": 0.14984450098953916,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5435,
      "mean_token_accuracy": 0.1460522949695587,
      "num_tokens": 421232.0,
      "step": 530
    },
    {
      "entropy": 7.189267683029175,
      "epoch": 0.151258128357365,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6245,
      "mean_token_accuracy": 0.1368570201098919,
      "num_tokens": 425503.0,
      "step": 535
    },
    {
      "entropy": 7.121035575866699,
      "epoch": 0.15267175572519084,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6376,
      "mean_token_accuracy": 0.15090529546141623,
      "num_tokens": 429770.0,
      "step": 540
    },
    {
      "entropy": 7.14065523147583,
      "epoch": 0.15408538309301667,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5419,
      "mean_token_accuracy": 0.14979021847248078,
      "num_tokens": 433657.0,
      "step": 545
    },
    {
      "entropy": 7.119353485107422,
      "epoch": 0.15549901046084252,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002745,
      "loss": 6.6487,
      "mean_token_accuracy": 0.1421989232301712,
      "num_tokens": 437732.0,
      "step": 550
    },
    {
      "entropy": 6.953150367736816,
      "epoch": 0.15691263782866835,
      "grad_norm": 1.515625,
      "learning_rate": 0.000277,
      "loss": 6.3952,
      "mean_token_accuracy": 0.15552812665700913,
      "num_tokens": 441866.0,
      "step": 555
    },
    {
      "entropy": 6.980320453643799,
      "epoch": 0.1583262651964942,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002795,
      "loss": 6.4783,
      "mean_token_accuracy": 0.14983998835086823,
      "num_tokens": 445437.0,
      "step": 560
    },
    {
      "entropy": 7.0185740947723385,
      "epoch": 0.15973989256432006,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.4586,
      "mean_token_accuracy": 0.15214325338602067,
      "num_tokens": 449602.0,
      "step": 565
    },
    {
      "entropy": 6.880740737915039,
      "epoch": 0.16115351993214588,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002845,
      "loss": 6.3774,
      "mean_token_accuracy": 0.14884328693151475,
      "num_tokens": 453140.0,
      "step": 570
    },
    {
      "entropy": 7.237173557281494,
      "epoch": 0.16256714729997174,
      "grad_norm": 1.5625,
      "learning_rate": 0.000287,
      "loss": 6.6234,
      "mean_token_accuracy": 0.13322453647851945,
      "num_tokens": 457210.0,
      "step": 575
    },
    {
      "entropy": 7.069291353225708,
      "epoch": 0.16398077466779756,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002895,
      "loss": 6.5821,
      "mean_token_accuracy": 0.14147427529096604,
      "num_tokens": 461534.0,
      "step": 580
    },
    {
      "entropy": 7.033083820343018,
      "epoch": 0.16539440203562342,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000292,
      "loss": 6.4198,
      "mean_token_accuracy": 0.14874927774071695,
      "num_tokens": 465055.0,
      "step": 585
    },
    {
      "entropy": 7.058513498306274,
      "epoch": 0.16680802940344924,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002945,
      "loss": 6.444,
      "mean_token_accuracy": 0.14927454963326453,
      "num_tokens": 469133.0,
      "step": 590
    },
    {
      "entropy": 7.091696453094483,
      "epoch": 0.1682216567712751,
      "grad_norm": 1.546875,
      "learning_rate": 0.000297,
      "loss": 6.5729,
      "mean_token_accuracy": 0.14638903141021728,
      "num_tokens": 473078.0,
      "step": 595
    },
    {
      "entropy": 7.0970518589019775,
      "epoch": 0.16963528413910092,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002995,
      "loss": 6.5146,
      "mean_token_accuracy": 0.14790180325508118,
      "num_tokens": 476988.0,
      "step": 600
    },
    {
      "entropy": 6.953248834609985,
      "epoch": 0.17104891150692678,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000302,
      "loss": 6.3748,
      "mean_token_accuracy": 0.15677656084299088,
      "num_tokens": 481170.0,
      "step": 605
    },
    {
      "entropy": 6.832110261917114,
      "epoch": 0.1724625388747526,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003045,
      "loss": 6.3622,
      "mean_token_accuracy": 0.1458218924701214,
      "num_tokens": 484853.0,
      "step": 610
    },
    {
      "entropy": 7.015735912322998,
      "epoch": 0.17387616624257846,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000307,
      "loss": 6.4019,
      "mean_token_accuracy": 0.15347179323434829,
      "num_tokens": 488550.0,
      "step": 615
    },
    {
      "entropy": 6.82008581161499,
      "epoch": 0.1752897936104043,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003095,
      "loss": 6.4353,
      "mean_token_accuracy": 0.15502589270472528,
      "num_tokens": 492621.0,
      "step": 620
    },
    {
      "entropy": 6.970222520828247,
      "epoch": 0.17670342097823014,
      "grad_norm": 1.671875,
      "learning_rate": 0.000312,
      "loss": 6.4046,
      "mean_token_accuracy": 0.14566168785095215,
      "num_tokens": 496627.0,
      "step": 625
    },
    {
      "entropy": 6.795285272598266,
      "epoch": 0.178117048346056,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003145,
      "loss": 6.4612,
      "mean_token_accuracy": 0.14936821013689042,
      "num_tokens": 500794.0,
      "step": 630
    },
    {
      "entropy": 7.059206104278564,
      "epoch": 0.17953067571388182,
      "grad_norm": 1.578125,
      "learning_rate": 0.000317,
      "loss": 6.4831,
      "mean_token_accuracy": 0.14905075654387473,
      "num_tokens": 504672.0,
      "step": 635
    },
    {
      "entropy": 6.794796943664551,
      "epoch": 0.18094430308170767,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003195,
      "loss": 6.4242,
      "mean_token_accuracy": 0.16338785141706466,
      "num_tokens": 508834.0,
      "step": 640
    },
    {
      "entropy": 7.082453823089599,
      "epoch": 0.1823579304495335,
      "grad_norm": 1.546875,
      "learning_rate": 0.000322,
      "loss": 6.388,
      "mean_token_accuracy": 0.15189965814352036,
      "num_tokens": 512722.0,
      "step": 645
    },
    {
      "entropy": 6.872324419021607,
      "epoch": 0.18377155781735935,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.4785,
      "mean_token_accuracy": 0.14856359511613845,
      "num_tokens": 517270.0,
      "step": 650
    },
    {
      "entropy": 7.005723524093628,
      "epoch": 0.18518518518518517,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.4087,
      "mean_token_accuracy": 0.15056324899196624,
      "num_tokens": 521438.0,
      "step": 655
    },
    {
      "entropy": 6.857723665237427,
      "epoch": 0.18659881255301103,
      "grad_norm": 1.421875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.3936,
      "mean_token_accuracy": 0.14879612773656845,
      "num_tokens": 525890.0,
      "step": 660
    },
    {
      "entropy": 7.036524677276612,
      "epoch": 0.18801243992083685,
      "grad_norm": 1.46875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4232,
      "mean_token_accuracy": 0.149623654037714,
      "num_tokens": 529775.0,
      "step": 665
    },
    {
      "entropy": 6.84811635017395,
      "epoch": 0.1894260672886627,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4021,
      "mean_token_accuracy": 0.1505955122411251,
      "num_tokens": 533889.0,
      "step": 670
    },
    {
      "entropy": 6.79489049911499,
      "epoch": 0.19083969465648856,
      "grad_norm": 1.453125,
      "learning_rate": 0.000337,
      "loss": 6.3423,
      "mean_token_accuracy": 0.16277977973222732,
      "num_tokens": 537701.0,
      "step": 675
    },
    {
      "entropy": 6.869364881515503,
      "epoch": 0.1922533220243144,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003395,
      "loss": 6.3895,
      "mean_token_accuracy": 0.15102225691080093,
      "num_tokens": 542160.0,
      "step": 680
    },
    {
      "entropy": 6.848371076583862,
      "epoch": 0.19366694939214024,
      "grad_norm": 1.4375,
      "learning_rate": 0.000342,
      "loss": 6.3673,
      "mean_token_accuracy": 0.1594443902373314,
      "num_tokens": 546317.0,
      "step": 685
    },
    {
      "entropy": 6.84291672706604,
      "epoch": 0.19508057675996607,
      "grad_norm": 1.59375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.2123,
      "mean_token_accuracy": 0.1704620197415352,
      "num_tokens": 550304.0,
      "step": 690
    },
    {
      "entropy": 6.779688024520874,
      "epoch": 0.19649420412779192,
      "grad_norm": 1.65625,
      "learning_rate": 0.000347,
      "loss": 6.3247,
      "mean_token_accuracy": 0.1501743510365486,
      "num_tokens": 554276.0,
      "step": 695
    },
    {
      "entropy": 6.807392883300781,
      "epoch": 0.19790783149561775,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003495,
      "loss": 6.3376,
      "mean_token_accuracy": 0.15815870016813277,
      "num_tokens": 558044.0,
      "step": 700
    },
    {
      "entropy": 6.69501748085022,
      "epoch": 0.1993214588634436,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000352,
      "loss": 6.3189,
      "mean_token_accuracy": 0.15602569431066513,
      "num_tokens": 561667.0,
      "step": 705
    },
    {
      "entropy": 6.884977722167969,
      "epoch": 0.20073508623126943,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003545,
      "loss": 6.3569,
      "mean_token_accuracy": 0.14620595276355744,
      "num_tokens": 565738.0,
      "step": 710
    },
    {
      "entropy": 6.865601634979248,
      "epoch": 0.20214871359909528,
      "grad_norm": 1.578125,
      "learning_rate": 0.000357,
      "loss": 6.3868,
      "mean_token_accuracy": 0.15818096846342086,
      "num_tokens": 569629.0,
      "step": 715
    },
    {
      "entropy": 6.779451847076416,
      "epoch": 0.2035623409669211,
      "grad_norm": 1.5,
      "learning_rate": 0.0003595,
      "loss": 6.2616,
      "mean_token_accuracy": 0.1626145839691162,
      "num_tokens": 573272.0,
      "step": 720
    },
    {
      "entropy": 6.587831258773804,
      "epoch": 0.20497596833474696,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000362,
      "loss": 6.3088,
      "mean_token_accuracy": 0.15946338027715684,
      "num_tokens": 577224.0,
      "step": 725
    },
    {
      "entropy": 6.784856128692627,
      "epoch": 0.20638959570257281,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003645,
      "loss": 6.2521,
      "mean_token_accuracy": 0.16293586939573287,
      "num_tokens": 580919.0,
      "step": 730
    },
    {
      "entropy": 6.7658926486969,
      "epoch": 0.20780322307039864,
      "grad_norm": 1.609375,
      "learning_rate": 0.000367,
      "loss": 6.3651,
      "mean_token_accuracy": 0.15333567410707474,
      "num_tokens": 584985.0,
      "step": 735
    },
    {
      "entropy": 6.826333856582641,
      "epoch": 0.2092168504382245,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003695,
      "loss": 6.3085,
      "mean_token_accuracy": 0.16295087337493896,
      "num_tokens": 588979.0,
      "step": 740
    },
    {
      "entropy": 6.648220825195312,
      "epoch": 0.21063047780605032,
      "grad_norm": 1.421875,
      "learning_rate": 0.000372,
      "loss": 6.2787,
      "mean_token_accuracy": 0.16390681266784668,
      "num_tokens": 592930.0,
      "step": 745
    },
    {
      "entropy": 6.663014554977417,
      "epoch": 0.21204410517387617,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003745,
      "loss": 6.3273,
      "mean_token_accuracy": 0.16125326007604598,
      "num_tokens": 596832.0,
      "step": 750
    },
    {
      "entropy": 6.937385177612304,
      "epoch": 0.213457732541702,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000377,
      "loss": 6.3617,
      "mean_token_accuracy": 0.15750757604837418,
      "num_tokens": 600934.0,
      "step": 755
    },
    {
      "entropy": 6.72532844543457,
      "epoch": 0.21487135990952785,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003795,
      "loss": 6.2373,
      "mean_token_accuracy": 0.15954903662204742,
      "num_tokens": 604868.0,
      "step": 760
    },
    {
      "entropy": 6.742078399658203,
      "epoch": 0.21628498727735368,
      "grad_norm": 1.59375,
      "learning_rate": 0.000382,
      "loss": 6.3596,
      "mean_token_accuracy": 0.150635027885437,
      "num_tokens": 608747.0,
      "step": 765
    },
    {
      "entropy": 6.704383659362793,
      "epoch": 0.21769861464517953,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003845,
      "loss": 6.3572,
      "mean_token_accuracy": 0.15894525349140168,
      "num_tokens": 612627.0,
      "step": 770
    },
    {
      "entropy": 6.658515930175781,
      "epoch": 0.21911224201300536,
      "grad_norm": 1.5,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.2251,
      "mean_token_accuracy": 0.17111287266016006,
      "num_tokens": 616455.0,
      "step": 775
    },
    {
      "entropy": 6.6746655941009525,
      "epoch": 0.2205258693808312,
      "grad_norm": 1.375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.326,
      "mean_token_accuracy": 0.15957299321889878,
      "num_tokens": 620617.0,
      "step": 780
    },
    {
      "entropy": 6.608477210998535,
      "epoch": 0.22193949674865707,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2692,
      "mean_token_accuracy": 0.1650322839617729,
      "num_tokens": 624863.0,
      "step": 785
    },
    {
      "entropy": 6.672313022613525,
      "epoch": 0.2233531241164829,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.2819,
      "mean_token_accuracy": 0.165033021569252,
      "num_tokens": 629073.0,
      "step": 790
    },
    {
      "entropy": 6.696346044540405,
      "epoch": 0.22476675148430875,
      "grad_norm": 1.53125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3095,
      "mean_token_accuracy": 0.16314611732959747,
      "num_tokens": 633550.0,
      "step": 795
    },
    {
      "entropy": 6.60796046257019,
      "epoch": 0.22618037885213457,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003995,
      "loss": 6.2888,
      "mean_token_accuracy": 0.1602451756596565,
      "num_tokens": 637459.0,
      "step": 800
    },
    {
      "entropy": 6.599572229385376,
      "epoch": 0.22759400621996043,
      "grad_norm": 1.3125,
      "learning_rate": 0.000402,
      "loss": 6.2433,
      "mean_token_accuracy": 0.15662450045347215,
      "num_tokens": 641587.0,
      "step": 805
    },
    {
      "entropy": 6.646263837814331,
      "epoch": 0.22900763358778625,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004045,
      "loss": 6.421,
      "mean_token_accuracy": 0.14964068084955215,
      "num_tokens": 645726.0,
      "step": 810
    },
    {
      "entropy": 6.735714673995972,
      "epoch": 0.2304212609556121,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2004,
      "mean_token_accuracy": 0.16830723136663436,
      "num_tokens": 649688.0,
      "step": 815
    },
    {
      "entropy": 6.588094997406006,
      "epoch": 0.23183488832343793,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004095,
      "loss": 6.2291,
      "mean_token_accuracy": 0.15805790424346924,
      "num_tokens": 653745.0,
      "step": 820
    },
    {
      "entropy": 6.485406589508057,
      "epoch": 0.23324851569126379,
      "grad_norm": 1.5,
      "learning_rate": 0.000412,
      "loss": 6.1713,
      "mean_token_accuracy": 0.1704267144203186,
      "num_tokens": 657650.0,
      "step": 825
    },
    {
      "entropy": 6.624208736419678,
      "epoch": 0.2346621430590896,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004145,
      "loss": 6.25,
      "mean_token_accuracy": 0.1644523784518242,
      "num_tokens": 661614.0,
      "step": 830
    },
    {
      "entropy": 6.544505929946899,
      "epoch": 0.23607577042691547,
      "grad_norm": 1.4375,
      "learning_rate": 0.000417,
      "loss": 6.1953,
      "mean_token_accuracy": 0.1738327383995056,
      "num_tokens": 665484.0,
      "step": 835
    },
    {
      "entropy": 6.586504888534546,
      "epoch": 0.23748939779474132,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004195,
      "loss": 6.1872,
      "mean_token_accuracy": 0.16504454612731934,
      "num_tokens": 669718.0,
      "step": 840
    },
    {
      "entropy": 6.625461673736572,
      "epoch": 0.23890302516256715,
      "grad_norm": 1.375,
      "learning_rate": 0.000422,
      "loss": 6.2064,
      "mean_token_accuracy": 0.16486725360155105,
      "num_tokens": 673913.0,
      "step": 845
    },
    {
      "entropy": 6.616315221786499,
      "epoch": 0.240316652530393,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004245,
      "loss": 6.2657,
      "mean_token_accuracy": 0.15634770691394806,
      "num_tokens": 677911.0,
      "step": 850
    },
    {
      "entropy": 6.552777242660523,
      "epoch": 0.24173027989821882,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000427,
      "loss": 6.0792,
      "mean_token_accuracy": 0.16897292137145997,
      "num_tokens": 681629.0,
      "step": 855
    },
    {
      "entropy": 6.523988723754883,
      "epoch": 0.24314390726604468,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004295,
      "loss": 6.1307,
      "mean_token_accuracy": 0.16514982432126998,
      "num_tokens": 685675.0,
      "step": 860
    },
    {
      "entropy": 6.562773180007935,
      "epoch": 0.2445575346338705,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000432,
      "loss": 6.0325,
      "mean_token_accuracy": 0.17125135958194732,
      "num_tokens": 689630.0,
      "step": 865
    },
    {
      "entropy": 6.422208642959594,
      "epoch": 0.24597116200169636,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004345,
      "loss": 6.1486,
      "mean_token_accuracy": 0.17256792038679122,
      "num_tokens": 693580.0,
      "step": 870
    },
    {
      "entropy": 6.469058227539063,
      "epoch": 0.24738478936952218,
      "grad_norm": 1.578125,
      "learning_rate": 0.000437,
      "loss": 6.2098,
      "mean_token_accuracy": 0.1618221327662468,
      "num_tokens": 697531.0,
      "step": 875
    },
    {
      "entropy": 6.530948734283447,
      "epoch": 0.24879841673734804,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004395,
      "loss": 6.1695,
      "mean_token_accuracy": 0.1632336474955082,
      "num_tokens": 701421.0,
      "step": 880
    },
    {
      "entropy": 6.483053684234619,
      "epoch": 0.2502120441051739,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000442,
      "loss": 6.1415,
      "mean_token_accuracy": 0.16531084030866622,
      "num_tokens": 705557.0,
      "step": 885
    },
    {
      "entropy": 6.433886241912842,
      "epoch": 0.2516256714729997,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004445,
      "loss": 6.0636,
      "mean_token_accuracy": 0.1743975192308426,
      "num_tokens": 709173.0,
      "step": 890
    },
    {
      "entropy": 6.3847051620483395,
      "epoch": 0.25303929884082554,
      "grad_norm": 1.515625,
      "learning_rate": 0.000447,
      "loss": 6.0107,
      "mean_token_accuracy": 0.1733053870499134,
      "num_tokens": 713144.0,
      "step": 895
    },
    {
      "entropy": 6.407188081741333,
      "epoch": 0.2544529262086514,
      "grad_norm": 1.375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3262,
      "mean_token_accuracy": 0.15888329446315766,
      "num_tokens": 717187.0,
      "step": 900
    },
    {
      "entropy": 6.68216462135315,
      "epoch": 0.25586655357647725,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.1792,
      "mean_token_accuracy": 0.16036764532327652,
      "num_tokens": 721311.0,
      "step": 905
    },
    {
      "entropy": 6.426914548873901,
      "epoch": 0.2572801809443031,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.1407,
      "mean_token_accuracy": 0.17487923800945282,
      "num_tokens": 725193.0,
      "step": 910
    },
    {
      "entropy": 6.396089220046997,
      "epoch": 0.2586938083121289,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.1083,
      "mean_token_accuracy": 0.16730499863624573,
      "num_tokens": 729208.0,
      "step": 915
    },
    {
      "entropy": 6.535867500305176,
      "epoch": 0.26010743567995476,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.1726,
      "mean_token_accuracy": 0.17124995291233064,
      "num_tokens": 733364.0,
      "step": 920
    },
    {
      "entropy": 6.3439921855926515,
      "epoch": 0.2615210630477806,
      "grad_norm": 1.484375,
      "learning_rate": 0.000462,
      "loss": 6.1221,
      "mean_token_accuracy": 0.16099516451358795,
      "num_tokens": 737549.0,
      "step": 925
    },
    {
      "entropy": 6.488129186630249,
      "epoch": 0.26293469041560646,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004645,
      "loss": 6.1294,
      "mean_token_accuracy": 0.16554402112960814,
      "num_tokens": 742009.0,
      "step": 930
    },
    {
      "entropy": 6.518905687332153,
      "epoch": 0.26434831778343226,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000467,
      "loss": 6.2303,
      "mean_token_accuracy": 0.16439489349722863,
      "num_tokens": 746258.0,
      "step": 935
    },
    {
      "entropy": 6.532473945617676,
      "epoch": 0.2657619451512581,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004695,
      "loss": 6.1433,
      "mean_token_accuracy": 0.1789155274629593,
      "num_tokens": 750027.0,
      "step": 940
    },
    {
      "entropy": 6.243679428100586,
      "epoch": 0.26717557251908397,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000472,
      "loss": 6.0869,
      "mean_token_accuracy": 0.16812117397785187,
      "num_tokens": 754030.0,
      "step": 945
    },
    {
      "entropy": 6.291708707809448,
      "epoch": 0.2685891998869098,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004745,
      "loss": 6.0033,
      "mean_token_accuracy": 0.17290493547916413,
      "num_tokens": 757663.0,
      "step": 950
    },
    {
      "entropy": 6.357456398010254,
      "epoch": 0.2700028272547357,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000477,
      "loss": 6.1225,
      "mean_token_accuracy": 0.16563030779361726,
      "num_tokens": 761806.0,
      "step": 955
    },
    {
      "entropy": 6.466005516052246,
      "epoch": 0.2714164546225615,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004795,
      "loss": 6.1282,
      "mean_token_accuracy": 0.16799217388033866,
      "num_tokens": 765900.0,
      "step": 960
    },
    {
      "entropy": 6.314237213134765,
      "epoch": 0.27283008199038733,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000482,
      "loss": 6.0802,
      "mean_token_accuracy": 0.16931777596473693,
      "num_tokens": 769842.0,
      "step": 965
    },
    {
      "entropy": 6.361982011795044,
      "epoch": 0.2742437093582132,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004845,
      "loss": 6.2017,
      "mean_token_accuracy": 0.16852450519800186,
      "num_tokens": 773717.0,
      "step": 970
    },
    {
      "entropy": 6.245587730407715,
      "epoch": 0.27565733672603904,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000487,
      "loss": 5.96,
      "mean_token_accuracy": 0.17952348291873932,
      "num_tokens": 777720.0,
      "step": 975
    },
    {
      "entropy": 6.214972496032715,
      "epoch": 0.27707096409386484,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004895,
      "loss": 5.9622,
      "mean_token_accuracy": 0.17511400282382966,
      "num_tokens": 781482.0,
      "step": 980
    },
    {
      "entropy": 6.306806802749634,
      "epoch": 0.2784845914616907,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000492,
      "loss": 6.0785,
      "mean_token_accuracy": 0.16686865836381912,
      "num_tokens": 785390.0,
      "step": 985
    },
    {
      "entropy": 6.173373794555664,
      "epoch": 0.27989821882951654,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004945,
      "loss": 6.017,
      "mean_token_accuracy": 0.18208521008491516,
      "num_tokens": 789464.0,
      "step": 990
    },
    {
      "entropy": 6.59720025062561,
      "epoch": 0.2813118461973424,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000497,
      "loss": 6.1232,
      "mean_token_accuracy": 0.16400160863995553,
      "num_tokens": 793229.0,
      "step": 995
    },
    {
      "entropy": 6.182021236419677,
      "epoch": 0.2827254735651682,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004995,
      "loss": 6.0322,
      "mean_token_accuracy": 0.17415720224380493,
      "num_tokens": 797490.0,
      "step": 1000
    },
    {
      "epoch": 0.2827254735651682,
      "eval_entropy": 6.051144407963093,
      "eval_loss": 6.294586181640625,
      "eval_mean_token_accuracy": 0.16992966667567475,
      "eval_num_tokens": 797490.0,
      "eval_runtime": 4.8981,
      "eval_samples_per_second": 2303.769,
      "eval_steps_per_second": 288.073,
      "step": 1000
    },
    {
      "entropy": 6.32871265411377,
      "epoch": 0.28413910093299405,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999999849524451,
      "loss": 6.0691,
      "mean_token_accuracy": 0.17283144295215608,
      "num_tokens": 801242.0,
      "step": 1005
    },
    {
      "entropy": 6.356661939620972,
      "epoch": 0.2855527283008199,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000499999923821757,
      "loss": 6.1618,
      "mean_token_accuracy": 0.16092283874750138,
      "num_tokens": 805285.0,
      "step": 1010
    },
    {
      "entropy": 6.229091024398803,
      "epoch": 0.28696635566864576,
      "grad_norm": 1.421875,
      "learning_rate": 0.000499999815667476,
      "loss": 5.9852,
      "mean_token_accuracy": 0.17933639883995056,
      "num_tokens": 809278.0,
      "step": 1015
    },
    {
      "entropy": 6.252400302886963,
      "epoch": 0.2883799830364716,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999996604896251,
      "loss": 6.1391,
      "mean_token_accuracy": 0.1700467899441719,
      "num_tokens": 813519.0,
      "step": 1020
    },
    {
      "entropy": 6.289431047439575,
      "epoch": 0.2897936104042974,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999994582882363,
      "loss": 6.0622,
      "mean_token_accuracy": 0.1710103139281273,
      "num_tokens": 817549.0,
      "step": 1025
    },
    {
      "entropy": 6.267086839675903,
      "epoch": 0.29120723777212326,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999992090633522,
      "loss": 6.0473,
      "mean_token_accuracy": 0.17169943898916246,
      "num_tokens": 821608.0,
      "step": 1030
    },
    {
      "entropy": 6.152920913696289,
      "epoch": 0.2926208651399491,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999989128150248,
      "loss": 5.9653,
      "mean_token_accuracy": 0.18662210255861283,
      "num_tokens": 825795.0,
      "step": 1035
    },
    {
      "entropy": 6.3972742557525635,
      "epoch": 0.29403449250777497,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999985695433159,
      "loss": 6.0705,
      "mean_token_accuracy": 0.17289165109395982,
      "num_tokens": 829693.0,
      "step": 1040
    },
    {
      "entropy": 6.2147924423217775,
      "epoch": 0.29544811987560077,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999981792482973,
      "loss": 6.0918,
      "mean_token_accuracy": 0.1711990311741829,
      "num_tokens": 833665.0,
      "step": 1045
    },
    {
      "entropy": 6.21585865020752,
      "epoch": 0.2968617472434266,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004999977419300507,
      "loss": 6.0505,
      "mean_token_accuracy": 0.16072292476892472,
      "num_tokens": 837107.0,
      "step": 1050
    },
    {
      "entropy": 6.245480632781982,
      "epoch": 0.2982753746112525,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999972575886673,
      "loss": 6.1374,
      "mean_token_accuracy": 0.16482933461666108,
      "num_tokens": 841182.0,
      "step": 1055
    },
    {
      "entropy": 6.287858152389527,
      "epoch": 0.29968900197907833,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004999967262242483,
      "loss": 6.0854,
      "mean_token_accuracy": 0.1722053959965706,
      "num_tokens": 845266.0,
      "step": 1060
    },
    {
      "entropy": 6.216232633590698,
      "epoch": 0.3011026293469042,
      "grad_norm": 1.453125,
      "learning_rate": 0.000499996147836905,
      "loss": 6.037,
      "mean_token_accuracy": 0.1735639214515686,
      "num_tokens": 849118.0,
      "step": 1065
    },
    {
      "entropy": 6.361587762832642,
      "epoch": 0.30251625671473,
      "grad_norm": 1.453125,
      "learning_rate": 0.000499995522426758,
      "loss": 6.144,
      "mean_token_accuracy": 0.16536193639039992,
      "num_tokens": 853101.0,
      "step": 1070
    },
    {
      "entropy": 6.167893266677856,
      "epoch": 0.30392988408255583,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004999948499939383,
      "loss": 5.9842,
      "mean_token_accuracy": 0.18141428083181382,
      "num_tokens": 856671.0,
      "step": 1075
    },
    {
      "entropy": 6.260188484191895,
      "epoch": 0.3053435114503817,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999941305385863,
      "loss": 5.9521,
      "mean_token_accuracy": 0.18240092396736146,
      "num_tokens": 860639.0,
      "step": 1080
    },
    {
      "entropy": 6.1321197032928465,
      "epoch": 0.30675713881820754,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004999933640608521,
      "loss": 5.984,
      "mean_token_accuracy": 0.17295702844858168,
      "num_tokens": 864270.0,
      "step": 1085
    },
    {
      "entropy": 6.153362846374511,
      "epoch": 0.30817076618603334,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004999925505608963,
      "loss": 5.9524,
      "mean_token_accuracy": 0.16724486351013185,
      "num_tokens": 868069.0,
      "step": 1090
    },
    {
      "entropy": 6.131827163696289,
      "epoch": 0.3095843935538592,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999916900388887,
      "loss": 5.9961,
      "mean_token_accuracy": 0.1748175099492073,
      "num_tokens": 871943.0,
      "step": 1095
    },
    {
      "entropy": 6.255445575714111,
      "epoch": 0.31099802092168505,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999907824950093,
      "loss": 6.0249,
      "mean_token_accuracy": 0.17832081466913224,
      "num_tokens": 876168.0,
      "step": 1100
    },
    {
      "entropy": 6.095445013046264,
      "epoch": 0.3124116482895109,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004999898279294475,
      "loss": 6.0014,
      "mean_token_accuracy": 0.17347852513194084,
      "num_tokens": 880134.0,
      "step": 1105
    },
    {
      "entropy": 6.133862781524658,
      "epoch": 0.3138252756573367,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004999888263424031,
      "loss": 6.0011,
      "mean_token_accuracy": 0.17168307155370713,
      "num_tokens": 883970.0,
      "step": 1110
    },
    {
      "entropy": 6.260912227630615,
      "epoch": 0.31523890302516255,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004999877777340852,
      "loss": 6.0514,
      "mean_token_accuracy": 0.16243282705545425,
      "num_tokens": 887980.0,
      "step": 1115
    },
    {
      "entropy": 6.018231344223023,
      "epoch": 0.3166525303929884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499986682104713,
      "loss": 5.8732,
      "mean_token_accuracy": 0.18358443528413773,
      "num_tokens": 891867.0,
      "step": 1120
    },
    {
      "entropy": 6.121918344497681,
      "epoch": 0.31806615776081426,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999855394545157,
      "loss": 5.967,
      "mean_token_accuracy": 0.1776360958814621,
      "num_tokens": 895769.0,
      "step": 1125
    },
    {
      "entropy": 6.1299745559692385,
      "epoch": 0.3194797851286401,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999843497837317,
      "loss": 5.9112,
      "mean_token_accuracy": 0.17768835574388503,
      "num_tokens": 899640.0,
      "step": 1130
    },
    {
      "entropy": 6.127333307266236,
      "epoch": 0.3208934124964659,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00049998311309261,
      "loss": 5.9222,
      "mean_token_accuracy": 0.18061408549547195,
      "num_tokens": 903598.0,
      "step": 1135
    },
    {
      "entropy": 6.185299062728882,
      "epoch": 0.32230703986429177,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004999818293814089,
      "loss": 5.9639,
      "mean_token_accuracy": 0.17317982614040375,
      "num_tokens": 907720.0,
      "step": 1140
    },
    {
      "entropy": 5.960520887374878,
      "epoch": 0.3237206672321176,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004999804986503966,
      "loss": 5.9432,
      "mean_token_accuracy": 0.1775066375732422,
      "num_tokens": 911531.0,
      "step": 1145
    },
    {
      "entropy": 6.173041248321534,
      "epoch": 0.3251342945999435,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004999791208998513,
      "loss": 5.8607,
      "mean_token_accuracy": 0.18050449788570405,
      "num_tokens": 915428.0,
      "step": 1150
    },
    {
      "entropy": 6.007273769378662,
      "epoch": 0.3265479219677693,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999776961300611,
      "loss": 5.875,
      "mean_token_accuracy": 0.17531459927558898,
      "num_tokens": 919499.0,
      "step": 1155
    },
    {
      "entropy": 6.183330106735229,
      "epoch": 0.3279615493355951,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999762243413236,
      "loss": 5.8994,
      "mean_token_accuracy": 0.18333160728216172,
      "num_tokens": 923530.0,
      "step": 1160
    },
    {
      "entropy": 6.1380363464355465,
      "epoch": 0.329375176703421,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999747055339464,
      "loss": 5.846,
      "mean_token_accuracy": 0.1873629868030548,
      "num_tokens": 927770.0,
      "step": 1165
    },
    {
      "entropy": 5.994632005691528,
      "epoch": 0.33078880407124683,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004999731397082469,
      "loss": 5.8898,
      "mean_token_accuracy": 0.1810222864151001,
      "num_tokens": 932008.0,
      "step": 1170
    },
    {
      "entropy": 6.158831453323364,
      "epoch": 0.3322024314390727,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004999715268645524,
      "loss": 5.9115,
      "mean_token_accuracy": 0.1809718891978264,
      "num_tokens": 936388.0,
      "step": 1175
    },
    {
      "entropy": 6.054045629501343,
      "epoch": 0.3336160588068985,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999698670032,
      "loss": 5.9492,
      "mean_token_accuracy": 0.18082664608955384,
      "num_tokens": 940436.0,
      "step": 1180
    },
    {
      "entropy": 5.983764600753784,
      "epoch": 0.33502968617472434,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999681601245366,
      "loss": 5.9589,
      "mean_token_accuracy": 0.17784594148397445,
      "num_tokens": 944296.0,
      "step": 1185
    },
    {
      "entropy": 6.146846437454224,
      "epoch": 0.3364433135425502,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004999664062289189,
      "loss": 5.989,
      "mean_token_accuracy": 0.17501545697450638,
      "num_tokens": 948318.0,
      "step": 1190
    },
    {
      "entropy": 6.1085621356964115,
      "epoch": 0.33785694091037605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999646053167134,
      "loss": 5.8718,
      "mean_token_accuracy": 0.18369389176368714,
      "num_tokens": 952147.0,
      "step": 1195
    },
    {
      "entropy": 6.098910617828369,
      "epoch": 0.33927056827820185,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004999627573882966,
      "loss": 6.0154,
      "mean_token_accuracy": 0.17044340819120407,
      "num_tokens": 956445.0,
      "step": 1200
    },
    {
      "entropy": 6.067022705078125,
      "epoch": 0.3406841956460277,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999608624440546,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1799109995365143,
      "num_tokens": 960327.0,
      "step": 1205
    },
    {
      "entropy": 6.08806881904602,
      "epoch": 0.34209782301385355,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999589204843834,
      "loss": 5.9898,
      "mean_token_accuracy": 0.18153459578752518,
      "num_tokens": 964387.0,
      "step": 1210
    },
    {
      "entropy": 6.120113325119019,
      "epoch": 0.3435114503816794,
      "grad_norm": 1.390625,
      "learning_rate": 0.000499956931509689,
      "loss": 5.9261,
      "mean_token_accuracy": 0.18374113589525223,
      "num_tokens": 968327.0,
      "step": 1215
    },
    {
      "entropy": 6.132302856445312,
      "epoch": 0.3449250777495052,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000499954895520387,
      "loss": 5.9704,
      "mean_token_accuracy": 0.1799431711435318,
      "num_tokens": 972358.0,
      "step": 1220
    },
    {
      "entropy": 6.008096218109131,
      "epoch": 0.34633870511733106,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999528125169028,
      "loss": 5.8406,
      "mean_token_accuracy": 0.175810743868351,
      "num_tokens": 976266.0,
      "step": 1225
    },
    {
      "entropy": 6.0000334739685055,
      "epoch": 0.3477523324851569,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000499950682499672,
      "loss": 5.9705,
      "mean_token_accuracy": 0.17845053225755692,
      "num_tokens": 980521.0,
      "step": 1230
    },
    {
      "entropy": 6.040744686126709,
      "epoch": 0.34916595985298277,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004999485054691395,
      "loss": 5.8551,
      "mean_token_accuracy": 0.18057717978954316,
      "num_tokens": 984413.0,
      "step": 1235
    },
    {
      "entropy": 6.139350652694702,
      "epoch": 0.3505795872208086,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004999462814257604,
      "loss": 5.9464,
      "mean_token_accuracy": 0.16705757826566697,
      "num_tokens": 988080.0,
      "step": 1240
    },
    {
      "entropy": 6.139582967758178,
      "epoch": 0.3519932145886344,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999440103699996,
      "loss": 5.8926,
      "mean_token_accuracy": 0.1826186239719391,
      "num_tokens": 992010.0,
      "step": 1245
    },
    {
      "entropy": 5.878952169418335,
      "epoch": 0.35340684195646027,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999416923023316,
      "loss": 5.7834,
      "mean_token_accuracy": 0.194471637904644,
      "num_tokens": 995721.0,
      "step": 1250
    },
    {
      "entropy": 5.920779228210449,
      "epoch": 0.3548204693242861,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999393272232409,
      "loss": 5.8949,
      "mean_token_accuracy": 0.18157693296670913,
      "num_tokens": 999763.0,
      "step": 1255
    },
    {
      "entropy": 6.085085344314575,
      "epoch": 0.356234096692112,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999369151332218,
      "loss": 5.938,
      "mean_token_accuracy": 0.18311794847249985,
      "num_tokens": 1003610.0,
      "step": 1260
    },
    {
      "entropy": 6.084949398040772,
      "epoch": 0.3576477240599378,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999344560327784,
      "loss": 5.8817,
      "mean_token_accuracy": 0.1808584988117218,
      "num_tokens": 1007699.0,
      "step": 1265
    },
    {
      "entropy": 6.034257507324218,
      "epoch": 0.35906135142776363,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004999319499224247,
      "loss": 5.8248,
      "mean_token_accuracy": 0.19244421720504762,
      "num_tokens": 1011573.0,
      "step": 1270
    },
    {
      "entropy": 5.997032213211059,
      "epoch": 0.3604749787955895,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999293968026843,
      "loss": 5.9599,
      "mean_token_accuracy": 0.17845921069383622,
      "num_tokens": 1015569.0,
      "step": 1275
    },
    {
      "entropy": 6.032209300994873,
      "epoch": 0.36188860616341534,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004999267966740909,
      "loss": 5.8013,
      "mean_token_accuracy": 0.1848297104239464,
      "num_tokens": 1019451.0,
      "step": 1280
    },
    {
      "entropy": 5.926502704620361,
      "epoch": 0.3633022335312412,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000499924149537188,
      "loss": 5.7224,
      "mean_token_accuracy": 0.19073605984449388,
      "num_tokens": 1023084.0,
      "step": 1285
    },
    {
      "entropy": 6.071679019927979,
      "epoch": 0.364715860899067,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004999214553925287,
      "loss": 5.841,
      "mean_token_accuracy": 0.18661287277936936,
      "num_tokens": 1026941.0,
      "step": 1290
    },
    {
      "entropy": 6.050941610336304,
      "epoch": 0.36612948826689284,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999187142406761,
      "loss": 5.7795,
      "mean_token_accuracy": 0.19633757770061494,
      "num_tokens": 1030872.0,
      "step": 1295
    },
    {
      "entropy": 5.994774961471558,
      "epoch": 0.3675431156347187,
      "grad_norm": 1.359375,
      "learning_rate": 0.000499915926082203,
      "loss": 5.8838,
      "mean_token_accuracy": 0.17777508050203322,
      "num_tokens": 1035024.0,
      "step": 1300
    },
    {
      "entropy": 5.969749546051025,
      "epoch": 0.36895674300254455,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999130909176923,
      "loss": 5.7872,
      "mean_token_accuracy": 0.18587531447410582,
      "num_tokens": 1038988.0,
      "step": 1305
    },
    {
      "entropy": 5.884610033035278,
      "epoch": 0.37037037037037035,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004999102087477362,
      "loss": 5.7024,
      "mean_token_accuracy": 0.19922589361667634,
      "num_tokens": 1043068.0,
      "step": 1310
    },
    {
      "entropy": 5.866126918792725,
      "epoch": 0.3717839977381962,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999072795729376,
      "loss": 5.776,
      "mean_token_accuracy": 0.18842145353555678,
      "num_tokens": 1046931.0,
      "step": 1315
    },
    {
      "entropy": 5.8829490661621096,
      "epoch": 0.37319762510602206,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999043033939081,
      "loss": 5.7317,
      "mean_token_accuracy": 0.19113447070121764,
      "num_tokens": 1050718.0,
      "step": 1320
    },
    {
      "entropy": 5.90353946685791,
      "epoch": 0.3746112524738479,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999012802112701,
      "loss": 5.8071,
      "mean_token_accuracy": 0.1775677904486656,
      "num_tokens": 1054796.0,
      "step": 1325
    },
    {
      "entropy": 5.928818321228027,
      "epoch": 0.3760248798416737,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004998982100256552,
      "loss": 5.7504,
      "mean_token_accuracy": 0.18737074881792068,
      "num_tokens": 1058657.0,
      "step": 1330
    },
    {
      "entropy": 5.908525562286377,
      "epoch": 0.37743850720949956,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004998950928377052,
      "loss": 5.8693,
      "mean_token_accuracy": 0.18807025104761124,
      "num_tokens": 1062532.0,
      "step": 1335
    },
    {
      "entropy": 5.892481327056885,
      "epoch": 0.3788521345773254,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004998919286480713,
      "loss": 5.7168,
      "mean_token_accuracy": 0.19382492154836656,
      "num_tokens": 1066274.0,
      "step": 1340
    },
    {
      "entropy": 5.8984781265258786,
      "epoch": 0.38026576194515127,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004998887174574153,
      "loss": 5.8189,
      "mean_token_accuracy": 0.18274522572755814,
      "num_tokens": 1070179.0,
      "step": 1345
    },
    {
      "entropy": 5.870694589614868,
      "epoch": 0.3816793893129771,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004998854592664079,
      "loss": 5.7916,
      "mean_token_accuracy": 0.18562129884958267,
      "num_tokens": 1073892.0,
      "step": 1350
    },
    {
      "entropy": 5.9831644058227536,
      "epoch": 0.3830930166808029,
      "grad_norm": 1.25,
      "learning_rate": 0.0004998821540757301,
      "loss": 5.8439,
      "mean_token_accuracy": 0.18298695832490922,
      "num_tokens": 1078399.0,
      "step": 1355
    },
    {
      "entropy": 5.8720427513122555,
      "epoch": 0.3845066440486288,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004998788018860728,
      "loss": 5.754,
      "mean_token_accuracy": 0.19132645279169083,
      "num_tokens": 1082371.0,
      "step": 1360
    },
    {
      "entropy": 5.921624422073364,
      "epoch": 0.38592027141645463,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004998754026981364,
      "loss": 5.8936,
      "mean_token_accuracy": 0.17152681052684784,
      "num_tokens": 1086474.0,
      "step": 1365
    },
    {
      "entropy": 5.971354389190674,
      "epoch": 0.3873338987842805,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004998719565126316,
      "loss": 5.8067,
      "mean_token_accuracy": 0.19351842254400253,
      "num_tokens": 1090519.0,
      "step": 1370
    },
    {
      "entropy": 5.819832897186279,
      "epoch": 0.3887475261521063,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004998684633302783,
      "loss": 5.7492,
      "mean_token_accuracy": 0.18426936864852905,
      "num_tokens": 1094450.0,
      "step": 1375
    },
    {
      "entropy": 6.021639633178711,
      "epoch": 0.39016115351993214,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004998649231518068,
      "loss": 5.7939,
      "mean_token_accuracy": 0.1839954972267151,
      "num_tokens": 1098249.0,
      "step": 1380
    },
    {
      "entropy": 5.936412715911866,
      "epoch": 0.391574780887758,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004998613359779568,
      "loss": 5.7502,
      "mean_token_accuracy": 0.1965183287858963,
      "num_tokens": 1101951.0,
      "step": 1385
    },
    {
      "entropy": 5.852648210525513,
      "epoch": 0.39298840825558384,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004998577018094782,
      "loss": 5.7843,
      "mean_token_accuracy": 0.1870306134223938,
      "num_tokens": 1106097.0,
      "step": 1390
    },
    {
      "entropy": 5.901752901077271,
      "epoch": 0.3944020356234097,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004998540206471304,
      "loss": 5.7213,
      "mean_token_accuracy": 0.1883508324623108,
      "num_tokens": 1110031.0,
      "step": 1395
    },
    {
      "entropy": 5.926095390319825,
      "epoch": 0.3958156629912355,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004998502924916826,
      "loss": 5.8754,
      "mean_token_accuracy": 0.1840156227350235,
      "num_tokens": 1114146.0,
      "step": 1400
    },
    {
      "entropy": 6.012360382080078,
      "epoch": 0.39722929035906135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004998465173439142,
      "loss": 5.8229,
      "mean_token_accuracy": 0.18926550894975663,
      "num_tokens": 1118161.0,
      "step": 1405
    },
    {
      "entropy": 5.941721773147583,
      "epoch": 0.3986429177268872,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004998426952046142,
      "loss": 5.6464,
      "mean_token_accuracy": 0.20289266854524612,
      "num_tokens": 1122014.0,
      "step": 1410
    },
    {
      "entropy": 5.76336145401001,
      "epoch": 0.40005654509471306,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004998388260745811,
      "loss": 5.8007,
      "mean_token_accuracy": 0.17852477133274078,
      "num_tokens": 1125929.0,
      "step": 1415
    },
    {
      "entropy": 5.954787492752075,
      "epoch": 0.40147017246253885,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004998349099546238,
      "loss": 5.7595,
      "mean_token_accuracy": 0.18921004384756088,
      "num_tokens": 1129781.0,
      "step": 1420
    },
    {
      "entropy": 5.847860431671142,
      "epoch": 0.4028837998303647,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004998309468455606,
      "loss": 5.6644,
      "mean_token_accuracy": 0.20904712975025178,
      "num_tokens": 1133591.0,
      "step": 1425
    },
    {
      "entropy": 5.821159505844117,
      "epoch": 0.40429742719819056,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998269367482199,
      "loss": 5.7892,
      "mean_token_accuracy": 0.1861359179019928,
      "num_tokens": 1137406.0,
      "step": 1430
    },
    {
      "entropy": 5.842456293106079,
      "epoch": 0.4057110545660164,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004998228796634396,
      "loss": 5.7103,
      "mean_token_accuracy": 0.19130536019802094,
      "num_tokens": 1141393.0,
      "step": 1435
    },
    {
      "entropy": 5.929346418380737,
      "epoch": 0.4071246819338422,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004998187755920677,
      "loss": 5.8373,
      "mean_token_accuracy": 0.18991080820560455,
      "num_tokens": 1145155.0,
      "step": 1440
    },
    {
      "entropy": 5.8391400337219235,
      "epoch": 0.40853830930166807,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004998146245349621,
      "loss": 5.7896,
      "mean_token_accuracy": 0.18923215866088866,
      "num_tokens": 1149104.0,
      "step": 1445
    },
    {
      "entropy": 5.861079597473145,
      "epoch": 0.4099519366694939,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004998104264929901,
      "loss": 5.7447,
      "mean_token_accuracy": 0.18697023391723633,
      "num_tokens": 1153022.0,
      "step": 1450
    },
    {
      "entropy": 5.918407154083252,
      "epoch": 0.4113655640373198,
      "grad_norm": 1.375,
      "learning_rate": 0.0004998061814670291,
      "loss": 5.7603,
      "mean_token_accuracy": 0.184474678337574,
      "num_tokens": 1156965.0,
      "step": 1455
    },
    {
      "entropy": 5.811736297607422,
      "epoch": 0.41277919140514563,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004998018894579664,
      "loss": 5.7111,
      "mean_token_accuracy": 0.1976108655333519,
      "num_tokens": 1160893.0,
      "step": 1460
    },
    {
      "entropy": 5.9467895984649655,
      "epoch": 0.4141928187729714,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004997975504666989,
      "loss": 5.8613,
      "mean_token_accuracy": 0.17591351866722107,
      "num_tokens": 1165037.0,
      "step": 1465
    },
    {
      "entropy": 5.983978176116944,
      "epoch": 0.4156064461407973,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004997931644941335,
      "loss": 5.7535,
      "mean_token_accuracy": 0.1833250492811203,
      "num_tokens": 1168911.0,
      "step": 1470
    },
    {
      "entropy": 5.925879240036011,
      "epoch": 0.41702007350862313,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004997887315411867,
      "loss": 5.752,
      "mean_token_accuracy": 0.1955208569765091,
      "num_tokens": 1172861.0,
      "step": 1475
    },
    {
      "entropy": 5.804015302658081,
      "epoch": 0.418433700876449,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004997842516087852,
      "loss": 5.7843,
      "mean_token_accuracy": 0.18648661077022552,
      "num_tokens": 1177361.0,
      "step": 1480
    },
    {
      "entropy": 5.9768593311309814,
      "epoch": 0.4198473282442748,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004997797246978651,
      "loss": 5.7778,
      "mean_token_accuracy": 0.17635598480701448,
      "num_tokens": 1181549.0,
      "step": 1485
    },
    {
      "entropy": 5.921354866027832,
      "epoch": 0.42126095561210064,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004997751508093724,
      "loss": 5.7786,
      "mean_token_accuracy": 0.18104448914527893,
      "num_tokens": 1185685.0,
      "step": 1490
    },
    {
      "entropy": 5.8056823253631595,
      "epoch": 0.4226745829799265,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004997705299442632,
      "loss": 5.8771,
      "mean_token_accuracy": 0.17976228892803192,
      "num_tokens": 1189557.0,
      "step": 1495
    },
    {
      "entropy": 5.953405332565308,
      "epoch": 0.42408821034775235,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004997658621035033,
      "loss": 5.7987,
      "mean_token_accuracy": 0.19059500545263292,
      "num_tokens": 1193969.0,
      "step": 1500
    },
    {
      "entropy": 5.930359935760498,
      "epoch": 0.42550183771557815,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004997611472880681,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1988312780857086,
      "num_tokens": 1197843.0,
      "step": 1505
    },
    {
      "entropy": 5.812527418136597,
      "epoch": 0.426915465083404,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000499756385498943,
      "loss": 5.7792,
      "mean_token_accuracy": 0.18155284225940704,
      "num_tokens": 1201938.0,
      "step": 1510
    },
    {
      "entropy": 5.8252583026885985,
      "epoch": 0.42832909245122985,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004997515767371231,
      "loss": 5.6347,
      "mean_token_accuracy": 0.1871896967291832,
      "num_tokens": 1205757.0,
      "step": 1515
    },
    {
      "entropy": 5.889902544021607,
      "epoch": 0.4297427198190557,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004997467210036135,
      "loss": 5.7705,
      "mean_token_accuracy": 0.18698835819959642,
      "num_tokens": 1209779.0,
      "step": 1520
    },
    {
      "entropy": 5.821274614334106,
      "epoch": 0.43115634718688156,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000499741818299429,
      "loss": 5.7323,
      "mean_token_accuracy": 0.19483786523342134,
      "num_tokens": 1213598.0,
      "step": 1525
    },
    {
      "entropy": 5.744606351852417,
      "epoch": 0.43256997455470736,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004997368686255943,
      "loss": 5.6435,
      "mean_token_accuracy": 0.19433416724205016,
      "num_tokens": 1217452.0,
      "step": 1530
    },
    {
      "entropy": 5.858005809783935,
      "epoch": 0.4339836019225332,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004997318719831438,
      "loss": 5.7862,
      "mean_token_accuracy": 0.18418201208114623,
      "num_tokens": 1221354.0,
      "step": 1535
    },
    {
      "entropy": 5.793173360824585,
      "epoch": 0.43539722929035907,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004997268283731216,
      "loss": 5.6892,
      "mean_token_accuracy": 0.19399462640285492,
      "num_tokens": 1225217.0,
      "step": 1540
    },
    {
      "entropy": 5.891930007934571,
      "epoch": 0.4368108566581849,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004997217377965821,
      "loss": 5.6947,
      "mean_token_accuracy": 0.18456600606441498,
      "num_tokens": 1229016.0,
      "step": 1545
    },
    {
      "entropy": 5.927923154830933,
      "epoch": 0.4382244840260107,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000499716600254589,
      "loss": 5.7846,
      "mean_token_accuracy": 0.19131413698196412,
      "num_tokens": 1232825.0,
      "step": 1550
    },
    {
      "entropy": 5.911680030822754,
      "epoch": 0.4396381113938366,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499711415748216,
      "loss": 5.7284,
      "mean_token_accuracy": 0.1870203584432602,
      "num_tokens": 1237032.0,
      "step": 1555
    },
    {
      "entropy": 5.6728795051574705,
      "epoch": 0.4410517387616624,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004997061842785468,
      "loss": 5.6221,
      "mean_token_accuracy": 0.19849239736795427,
      "num_tokens": 1241043.0,
      "step": 1560
    },
    {
      "entropy": 5.756577777862549,
      "epoch": 0.4424653661294883,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004997009058466745,
      "loss": 5.7198,
      "mean_token_accuracy": 0.196329565346241,
      "num_tokens": 1245175.0,
      "step": 1565
    },
    {
      "entropy": 5.840588140487671,
      "epoch": 0.44387899349731413,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004996955804537024,
      "loss": 5.6716,
      "mean_token_accuracy": 0.19366080611944197,
      "num_tokens": 1249540.0,
      "step": 1570
    },
    {
      "entropy": 5.783361577987671,
      "epoch": 0.44529262086513993,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004996902081007435,
      "loss": 5.6577,
      "mean_token_accuracy": 0.186581951379776,
      "num_tokens": 1253532.0,
      "step": 1575
    },
    {
      "entropy": 5.89839916229248,
      "epoch": 0.4467062482329658,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004996847887889205,
      "loss": 5.6562,
      "mean_token_accuracy": 0.18789928704500197,
      "num_tokens": 1257492.0,
      "step": 1580
    },
    {
      "entropy": 5.844485521316528,
      "epoch": 0.44811987560079164,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004996793225193662,
      "loss": 5.7554,
      "mean_token_accuracy": 0.18641360402107238,
      "num_tokens": 1261388.0,
      "step": 1585
    },
    {
      "entropy": 5.823833084106445,
      "epoch": 0.4495335029686175,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004996738092932227,
      "loss": 5.6579,
      "mean_token_accuracy": 0.1978958413004875,
      "num_tokens": 1265417.0,
      "step": 1590
    },
    {
      "entropy": 5.7972126483917235,
      "epoch": 0.4509471303364433,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004996682491116425,
      "loss": 5.6477,
      "mean_token_accuracy": 0.19896170347929001,
      "num_tokens": 1269454.0,
      "step": 1595
    },
    {
      "entropy": 5.7864405632019045,
      "epoch": 0.45236075770426915,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004996626419757875,
      "loss": 5.7422,
      "mean_token_accuracy": 0.19003152698278428,
      "num_tokens": 1273677.0,
      "step": 1600
    },
    {
      "entropy": 5.809814310073852,
      "epoch": 0.453774385072095,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004996569878868296,
      "loss": 5.5735,
      "mean_token_accuracy": 0.20513303130865096,
      "num_tokens": 1277683.0,
      "step": 1605
    },
    {
      "entropy": 5.726590728759765,
      "epoch": 0.45518801243992085,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004996512868459505,
      "loss": 5.725,
      "mean_token_accuracy": 0.18426367342472078,
      "num_tokens": 1281945.0,
      "step": 1610
    },
    {
      "entropy": 5.715907955169678,
      "epoch": 0.45660163980774665,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004996455388543416,
      "loss": 5.6331,
      "mean_token_accuracy": 0.1939169466495514,
      "num_tokens": 1285709.0,
      "step": 1615
    },
    {
      "entropy": 5.965580797195434,
      "epoch": 0.4580152671755725,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004996397439132042,
      "loss": 5.795,
      "mean_token_accuracy": 0.17828181833028794,
      "num_tokens": 1289772.0,
      "step": 1620
    },
    {
      "entropy": 5.7739866256713865,
      "epoch": 0.45942889454339836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004996339020237496,
      "loss": 5.6691,
      "mean_token_accuracy": 0.19340460002422333,
      "num_tokens": 1293671.0,
      "step": 1625
    },
    {
      "entropy": 5.850045537948608,
      "epoch": 0.4608425219112242,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004996280131871985,
      "loss": 5.5937,
      "mean_token_accuracy": 0.19490807950496675,
      "num_tokens": 1297334.0,
      "step": 1630
    },
    {
      "entropy": 5.788289546966553,
      "epoch": 0.46225614927905007,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004996220774047816,
      "loss": 5.6794,
      "mean_token_accuracy": 0.1935423344373703,
      "num_tokens": 1301219.0,
      "step": 1635
    },
    {
      "entropy": 5.798729944229126,
      "epoch": 0.46366977664687586,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004996160946777398,
      "loss": 5.6636,
      "mean_token_accuracy": 0.19195165783166884,
      "num_tokens": 1305166.0,
      "step": 1640
    },
    {
      "entropy": 5.729593896865845,
      "epoch": 0.4650834040147017,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004996100650073229,
      "loss": 5.5989,
      "mean_token_accuracy": 0.20028469562530518,
      "num_tokens": 1308958.0,
      "step": 1645
    },
    {
      "entropy": 5.800077962875366,
      "epoch": 0.46649703138252757,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004996039883947915,
      "loss": 5.817,
      "mean_token_accuracy": 0.18166496753692626,
      "num_tokens": 1313289.0,
      "step": 1650
    },
    {
      "entropy": 5.939286613464356,
      "epoch": 0.4679106587503534,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004995978648414154,
      "loss": 5.6835,
      "mean_token_accuracy": 0.19870763123035431,
      "num_tokens": 1317366.0,
      "step": 1655
    },
    {
      "entropy": 5.767664241790771,
      "epoch": 0.4693242861181792,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004995916943484745,
      "loss": 5.6659,
      "mean_token_accuracy": 0.1947951689362526,
      "num_tokens": 1321362.0,
      "step": 1660
    },
    {
      "entropy": 5.746976137161255,
      "epoch": 0.4707379134860051,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004995854769172583,
      "loss": 5.6213,
      "mean_token_accuracy": 0.19329093098640443,
      "num_tokens": 1325147.0,
      "step": 1665
    },
    {
      "entropy": 5.845152759552002,
      "epoch": 0.47215154085383093,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004995792125490662,
      "loss": 5.6702,
      "mean_token_accuracy": 0.1898410737514496,
      "num_tokens": 1329072.0,
      "step": 1670
    },
    {
      "entropy": 5.709325075149536,
      "epoch": 0.4735651682216568,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004995729012452074,
      "loss": 5.6052,
      "mean_token_accuracy": 0.198711159825325,
      "num_tokens": 1332947.0,
      "step": 1675
    },
    {
      "entropy": 5.627379941940307,
      "epoch": 0.47497879558948264,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004995665430070009,
      "loss": 5.621,
      "mean_token_accuracy": 0.19632248133420943,
      "num_tokens": 1336903.0,
      "step": 1680
    },
    {
      "entropy": 5.819957590103149,
      "epoch": 0.47639242295730844,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004995601378357757,
      "loss": 5.6884,
      "mean_token_accuracy": 0.19369249790906906,
      "num_tokens": 1340989.0,
      "step": 1685
    },
    {
      "entropy": 5.899555778503418,
      "epoch": 0.4778060503251343,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004995536857328702,
      "loss": 5.7288,
      "mean_token_accuracy": 0.17993859052658082,
      "num_tokens": 1345251.0,
      "step": 1690
    },
    {
      "entropy": 5.822159147262573,
      "epoch": 0.47921967769296014,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004995471866996332,
      "loss": 5.8217,
      "mean_token_accuracy": 0.18204482644796371,
      "num_tokens": 1349167.0,
      "step": 1695
    },
    {
      "entropy": 5.724724149703979,
      "epoch": 0.480633305060786,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004995406407374226,
      "loss": 5.5332,
      "mean_token_accuracy": 0.2037088915705681,
      "num_tokens": 1353309.0,
      "step": 1700
    },
    {
      "entropy": 5.69287486076355,
      "epoch": 0.4820469324286118,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004995340478476067,
      "loss": 5.656,
      "mean_token_accuracy": 0.20406957417726518,
      "num_tokens": 1357129.0,
      "step": 1705
    },
    {
      "entropy": 5.6713849067687985,
      "epoch": 0.48346055979643765,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004995274080315631,
      "loss": 5.6198,
      "mean_token_accuracy": 0.19456420093774796,
      "num_tokens": 1360824.0,
      "step": 1710
    },
    {
      "entropy": 5.721224403381347,
      "epoch": 0.4848741871642635,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004995207212906798,
      "loss": 5.6495,
      "mean_token_accuracy": 0.1971166342496872,
      "num_tokens": 1364688.0,
      "step": 1715
    },
    {
      "entropy": 5.750689172744751,
      "epoch": 0.48628781453208936,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000499513987626354,
      "loss": 5.5703,
      "mean_token_accuracy": 0.20018759667873381,
      "num_tokens": 1368889.0,
      "step": 1720
    },
    {
      "entropy": 5.674837827682495,
      "epoch": 0.48770144189991516,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004995072070399933,
      "loss": 5.6836,
      "mean_token_accuracy": 0.2023005336523056,
      "num_tokens": 1372912.0,
      "step": 1725
    },
    {
      "entropy": 5.6823383331298825,
      "epoch": 0.489115069267741,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004995003795330146,
      "loss": 5.5365,
      "mean_token_accuracy": 0.20185312926769255,
      "num_tokens": 1376566.0,
      "step": 1730
    },
    {
      "entropy": 5.636135816574097,
      "epoch": 0.49052869663556686,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004994935051068447,
      "loss": 5.5587,
      "mean_token_accuracy": 0.1983194187283516,
      "num_tokens": 1380710.0,
      "step": 1735
    },
    {
      "entropy": 5.698077821731568,
      "epoch": 0.4919423240033927,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004994865837629206,
      "loss": 5.6284,
      "mean_token_accuracy": 0.2035332128405571,
      "num_tokens": 1384485.0,
      "step": 1740
    },
    {
      "entropy": 5.785005521774292,
      "epoch": 0.49335595137121857,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004994796155026887,
      "loss": 5.6135,
      "mean_token_accuracy": 0.1922136127948761,
      "num_tokens": 1388297.0,
      "step": 1745
    },
    {
      "entropy": 5.679270887374878,
      "epoch": 0.49476957873904437,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004994726003276053,
      "loss": 5.6053,
      "mean_token_accuracy": 0.20150333791971206,
      "num_tokens": 1392390.0,
      "step": 1750
    },
    {
      "entropy": 5.657969236373901,
      "epoch": 0.4961832061068702,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004994655382391365,
      "loss": 5.641,
      "mean_token_accuracy": 0.19058468639850618,
      "num_tokens": 1396561.0,
      "step": 1755
    },
    {
      "entropy": 5.740747547149658,
      "epoch": 0.4975968334746961,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004994584292387582,
      "loss": 5.6525,
      "mean_token_accuracy": 0.19100899398326873,
      "num_tokens": 1400556.0,
      "step": 1760
    },
    {
      "entropy": 5.835249662399292,
      "epoch": 0.49901046084252193,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004994512733279562,
      "loss": 5.7011,
      "mean_token_accuracy": 0.18798956274986267,
      "num_tokens": 1404503.0,
      "step": 1765
    },
    {
      "entropy": 5.733709383010864,
      "epoch": 0.5004240882103478,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004994440705082261,
      "loss": 5.6529,
      "mean_token_accuracy": 0.1851186528801918,
      "num_tokens": 1408504.0,
      "step": 1770
    },
    {
      "entropy": 5.705524873733521,
      "epoch": 0.5018377155781736,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004994368207810731,
      "loss": 5.4874,
      "mean_token_accuracy": 0.2035467341542244,
      "num_tokens": 1412429.0,
      "step": 1775
    },
    {
      "entropy": 5.629953145980835,
      "epoch": 0.5032513429459994,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004994295241480126,
      "loss": 5.5837,
      "mean_token_accuracy": 0.19919728189706803,
      "num_tokens": 1416466.0,
      "step": 1780
    },
    {
      "entropy": 5.678175735473633,
      "epoch": 0.5046649703138253,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004994221806105692,
      "loss": 5.5531,
      "mean_token_accuracy": 0.20242667198181152,
      "num_tokens": 1420226.0,
      "step": 1785
    },
    {
      "entropy": 5.559377288818359,
      "epoch": 0.5060785976816511,
      "grad_norm": 1.375,
      "learning_rate": 0.000499414790170278,
      "loss": 5.5673,
      "mean_token_accuracy": 0.19314760267734526,
      "num_tokens": 1423987.0,
      "step": 1790
    },
    {
      "entropy": 5.732563257217407,
      "epoch": 0.507492225049477,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004994073528286833,
      "loss": 5.6066,
      "mean_token_accuracy": 0.2005213737487793,
      "num_tokens": 1428264.0,
      "step": 1795
    },
    {
      "entropy": 5.717577838897705,
      "epoch": 0.5089058524173028,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993998685873395,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1915834978222847,
      "num_tokens": 1432350.0,
      "step": 1800
    },
    {
      "entropy": 5.589537763595581,
      "epoch": 0.5103194797851286,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004993923374478109,
      "loss": 5.5276,
      "mean_token_accuracy": 0.20159524530172349,
      "num_tokens": 1436298.0,
      "step": 1805
    },
    {
      "entropy": 5.542631483078003,
      "epoch": 0.5117331071529545,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993847594116713,
      "loss": 5.4777,
      "mean_token_accuracy": 0.20509825497865677,
      "num_tokens": 1440206.0,
      "step": 1810
    },
    {
      "entropy": 5.758960771560669,
      "epoch": 0.5131467345207803,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993771344805046,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1927124723792076,
      "num_tokens": 1444566.0,
      "step": 1815
    },
    {
      "entropy": 5.707762908935547,
      "epoch": 0.5145603618886062,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993694626559043,
      "loss": 5.6075,
      "mean_token_accuracy": 0.19674081355333328,
      "num_tokens": 1448792.0,
      "step": 1820
    },
    {
      "entropy": 5.755503702163696,
      "epoch": 0.515973989256432,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004993617439394737,
      "loss": 5.6768,
      "mean_token_accuracy": 0.1900164932012558,
      "num_tokens": 1452456.0,
      "step": 1825
    },
    {
      "entropy": 5.683109760284424,
      "epoch": 0.5173876166242578,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004993539783328261,
      "loss": 5.5949,
      "mean_token_accuracy": 0.19960361272096633,
      "num_tokens": 1456415.0,
      "step": 1830
    },
    {
      "entropy": 5.583623313903809,
      "epoch": 0.5188012439920837,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004993461658375845,
      "loss": 5.5242,
      "mean_token_accuracy": 0.20874519199132918,
      "num_tokens": 1460390.0,
      "step": 1835
    },
    {
      "entropy": 5.659956359863282,
      "epoch": 0.5202148713599095,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993383064553814,
      "loss": 5.6479,
      "mean_token_accuracy": 0.19412834346294403,
      "num_tokens": 1464363.0,
      "step": 1840
    },
    {
      "entropy": 5.733077812194824,
      "epoch": 0.5216284987277354,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993304001878595,
      "loss": 5.5266,
      "mean_token_accuracy": 0.20524653643369675,
      "num_tokens": 1468372.0,
      "step": 1845
    },
    {
      "entropy": 5.567001724243164,
      "epoch": 0.5230421260955612,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004993224470366712,
      "loss": 5.6421,
      "mean_token_accuracy": 0.19311347603797913,
      "num_tokens": 1472272.0,
      "step": 1850
    },
    {
      "entropy": 5.545955991744995,
      "epoch": 0.524455753463387,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004993144470034787,
      "loss": 5.4196,
      "mean_token_accuracy": 0.21298958063125611,
      "num_tokens": 1476195.0,
      "step": 1855
    },
    {
      "entropy": 5.790099716186523,
      "epoch": 0.5258693808312129,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993064000899538,
      "loss": 5.6113,
      "mean_token_accuracy": 0.1957776889204979,
      "num_tokens": 1480364.0,
      "step": 1860
    },
    {
      "entropy": 5.554702568054199,
      "epoch": 0.5272830081990387,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004992983062977784,
      "loss": 5.6126,
      "mean_token_accuracy": 0.2019720569252968,
      "num_tokens": 1484523.0,
      "step": 1865
    },
    {
      "entropy": 5.681436681747437,
      "epoch": 0.5286966355668645,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004992901656286439,
      "loss": 5.6293,
      "mean_token_accuracy": 0.19366944432258607,
      "num_tokens": 1488456.0,
      "step": 1870
    },
    {
      "entropy": 5.638179159164428,
      "epoch": 0.5301102629346904,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000499281978084252,
      "loss": 5.4644,
      "mean_token_accuracy": 0.20922451019287108,
      "num_tokens": 1492599.0,
      "step": 1875
    },
    {
      "entropy": 5.57268214225769,
      "epoch": 0.5315238903025162,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004992737436663134,
      "loss": 5.4751,
      "mean_token_accuracy": 0.22505437433719636,
      "num_tokens": 1496373.0,
      "step": 1880
    },
    {
      "entropy": 5.473353672027588,
      "epoch": 0.5329375176703421,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004992654623765493,
      "loss": 5.5507,
      "mean_token_accuracy": 0.20358589738607408,
      "num_tokens": 1500288.0,
      "step": 1885
    },
    {
      "entropy": 5.735796165466309,
      "epoch": 0.5343511450381679,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004992571342166904,
      "loss": 5.6745,
      "mean_token_accuracy": 0.19157627820968628,
      "num_tokens": 1504169.0,
      "step": 1890
    },
    {
      "entropy": 5.760134601593018,
      "epoch": 0.5357647724059937,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004992487591884772,
      "loss": 5.6245,
      "mean_token_accuracy": 0.19079590141773223,
      "num_tokens": 1508292.0,
      "step": 1895
    },
    {
      "entropy": 5.669271850585938,
      "epoch": 0.5371783997738196,
      "grad_norm": 1.296875,
      "learning_rate": 0.00049924033729366,
      "loss": 5.6237,
      "mean_token_accuracy": 0.2010587587952614,
      "num_tokens": 1512366.0,
      "step": 1900
    },
    {
      "entropy": 5.646955680847168,
      "epoch": 0.5385920271416454,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499231868533999,
      "loss": 5.6106,
      "mean_token_accuracy": 0.1991715833544731,
      "num_tokens": 1516627.0,
      "step": 1905
    },
    {
      "entropy": 5.649567794799805,
      "epoch": 0.5400056545094714,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004992233529112642,
      "loss": 5.4597,
      "mean_token_accuracy": 0.20987114906311036,
      "num_tokens": 1520541.0,
      "step": 1910
    },
    {
      "entropy": 5.664584159851074,
      "epoch": 0.5414192818772972,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004992147904272349,
      "loss": 5.5956,
      "mean_token_accuracy": 0.18337357342243193,
      "num_tokens": 1524505.0,
      "step": 1915
    },
    {
      "entropy": 5.587951421737671,
      "epoch": 0.542832909245123,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004992061810837012,
      "loss": 5.5333,
      "mean_token_accuracy": 0.20285723954439164,
      "num_tokens": 1528658.0,
      "step": 1920
    },
    {
      "entropy": 5.579516839981079,
      "epoch": 0.5442465366129489,
      "grad_norm": 1.296875,
      "learning_rate": 0.000499197524882462,
      "loss": 5.4955,
      "mean_token_accuracy": 0.2082669347524643,
      "num_tokens": 1532620.0,
      "step": 1925
    },
    {
      "entropy": 5.605240058898926,
      "epoch": 0.5456601639807747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991888218253265,
      "loss": 5.5917,
      "mean_token_accuracy": 0.1849711701273918,
      "num_tokens": 1536636.0,
      "step": 1930
    },
    {
      "entropy": 5.720357370376587,
      "epoch": 0.5470737913486005,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004991800719141136,
      "loss": 5.6327,
      "mean_token_accuracy": 0.1938754215836525,
      "num_tokens": 1540821.0,
      "step": 1935
    },
    {
      "entropy": 5.5586694240570065,
      "epoch": 0.5484874187164264,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991712751506518,
      "loss": 5.5268,
      "mean_token_accuracy": 0.21065980643033982,
      "num_tokens": 1544700.0,
      "step": 1940
    },
    {
      "entropy": 5.649634504318238,
      "epoch": 0.5499010460842522,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004991624315367798,
      "loss": 5.5222,
      "mean_token_accuracy": 0.20534564256668092,
      "num_tokens": 1548770.0,
      "step": 1945
    },
    {
      "entropy": 5.693707847595215,
      "epoch": 0.5513146734520781,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004991535410743459,
      "loss": 5.6152,
      "mean_token_accuracy": 0.1894238770008087,
      "num_tokens": 1553015.0,
      "step": 1950
    },
    {
      "entropy": 5.646745252609253,
      "epoch": 0.5527283008199039,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004991446037652079,
      "loss": 5.5949,
      "mean_token_accuracy": 0.1843682825565338,
      "num_tokens": 1556849.0,
      "step": 1955
    },
    {
      "entropy": 5.533147239685059,
      "epoch": 0.5541419281877297,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004991356196112339,
      "loss": 5.5687,
      "mean_token_accuracy": 0.20423490107059478,
      "num_tokens": 1560407.0,
      "step": 1960
    },
    {
      "entropy": 5.639888763427734,
      "epoch": 0.5555555555555556,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991265886143014,
      "loss": 5.4596,
      "mean_token_accuracy": 0.20686074197292328,
      "num_tokens": 1564473.0,
      "step": 1965
    },
    {
      "entropy": 5.58202748298645,
      "epoch": 0.5569691829233814,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004991175107762978,
      "loss": 5.6361,
      "mean_token_accuracy": 0.19261953085660935,
      "num_tokens": 1568141.0,
      "step": 1970
    },
    {
      "entropy": 5.683831691741943,
      "epoch": 0.5583828102912073,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004991083860991204,
      "loss": 5.6186,
      "mean_token_accuracy": 0.20144907236099244,
      "num_tokens": 1572130.0,
      "step": 1975
    },
    {
      "entropy": 5.693622064590454,
      "epoch": 0.5597964376590331,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000499099214584676,
      "loss": 5.6006,
      "mean_token_accuracy": 0.20176017433404922,
      "num_tokens": 1576243.0,
      "step": 1980
    },
    {
      "entropy": 5.718212985992432,
      "epoch": 0.5612100650268589,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004990899962348817,
      "loss": 5.5458,
      "mean_token_accuracy": 0.19472669214010238,
      "num_tokens": 1579916.0,
      "step": 1985
    },
    {
      "entropy": 5.483148622512817,
      "epoch": 0.5626236923946848,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004990807310516638,
      "loss": 5.3971,
      "mean_token_accuracy": 0.22061461508274077,
      "num_tokens": 1584075.0,
      "step": 1990
    },
    {
      "entropy": 5.628428792953491,
      "epoch": 0.5640373197625106,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004990714190369588,
      "loss": 5.5766,
      "mean_token_accuracy": 0.1951795771718025,
      "num_tokens": 1588026.0,
      "step": 1995
    },
    {
      "entropy": 5.706118106842041,
      "epoch": 0.5654509471303364,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499062060192713,
      "loss": 5.5487,
      "mean_token_accuracy": 0.19893766939640045,
      "num_tokens": 1592420.0,
      "step": 2000
    },
    {
      "epoch": 0.5654509471303364,
      "eval_entropy": 5.4442686896050425,
      "eval_loss": 5.730021953582764,
      "eval_mean_token_accuracy": 0.20073206712175648,
      "eval_num_tokens": 1592420.0,
      "eval_runtime": 4.8588,
      "eval_samples_per_second": 2322.381,
      "eval_steps_per_second": 290.401,
      "step": 2000
    },
    {
      "entropy": 5.5566449642181395,
      "epoch": 0.5668645744981623,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000499052654520882,
      "loss": 5.4582,
      "mean_token_accuracy": 0.21683346033096312,
      "num_tokens": 1596409.0,
      "step": 2005
    },
    {
      "entropy": 5.625777530670166,
      "epoch": 0.5682782018659881,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990432020234317,
      "loss": 5.5315,
      "mean_token_accuracy": 0.20139055997133254,
      "num_tokens": 1600549.0,
      "step": 2010
    },
    {
      "entropy": 5.454067277908325,
      "epoch": 0.569691829233814,
      "grad_norm": 1.375,
      "learning_rate": 0.0004990337027023378,
      "loss": 5.4343,
      "mean_token_accuracy": 0.20616025179624559,
      "num_tokens": 1604145.0,
      "step": 2015
    },
    {
      "entropy": 5.501142358779907,
      "epoch": 0.5711054566016398,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004990241565595853,
      "loss": 5.5455,
      "mean_token_accuracy": 0.19608652591705322,
      "num_tokens": 1608078.0,
      "step": 2020
    },
    {
      "entropy": 5.545711851119995,
      "epoch": 0.5725190839694656,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004990145635971694,
      "loss": 5.4784,
      "mean_token_accuracy": 0.19842927604913713,
      "num_tokens": 1612038.0,
      "step": 2025
    },
    {
      "entropy": 5.568185329437256,
      "epoch": 0.5739327113372915,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004990049238170949,
      "loss": 5.5643,
      "mean_token_accuracy": 0.18239825963974,
      "num_tokens": 1616059.0,
      "step": 2030
    },
    {
      "entropy": 5.484708166122436,
      "epoch": 0.5753463387051173,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004989952372213767,
      "loss": 5.4215,
      "mean_token_accuracy": 0.217547644674778,
      "num_tokens": 1619653.0,
      "step": 2035
    },
    {
      "entropy": 5.592620325088501,
      "epoch": 0.5767599660729432,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004989855038120389,
      "loss": 5.4529,
      "mean_token_accuracy": 0.210155089199543,
      "num_tokens": 1623820.0,
      "step": 2040
    },
    {
      "entropy": 5.673013305664062,
      "epoch": 0.578173593440769,
      "grad_norm": 1.265625,
      "learning_rate": 0.000498975723591116,
      "loss": 5.6388,
      "mean_token_accuracy": 0.2011185497045517,
      "num_tokens": 1627606.0,
      "step": 2045
    },
    {
      "entropy": 5.710173654556274,
      "epoch": 0.5795872208085948,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004989658965606519,
      "loss": 5.6078,
      "mean_token_accuracy": 0.19209033399820327,
      "num_tokens": 1631581.0,
      "step": 2050
    },
    {
      "entropy": 5.637240791320801,
      "epoch": 0.5810008481764207,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004989560227227003,
      "loss": 5.5653,
      "mean_token_accuracy": 0.19754677414894103,
      "num_tokens": 1635670.0,
      "step": 2055
    },
    {
      "entropy": 5.533895635604859,
      "epoch": 0.5824144755442465,
      "grad_norm": 1.25,
      "learning_rate": 0.0004989461020793248,
      "loss": 5.5605,
      "mean_token_accuracy": 0.2057570233941078,
      "num_tokens": 1639860.0,
      "step": 2060
    },
    {
      "entropy": 5.632710218429565,
      "epoch": 0.5838281029120724,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004989361346325989,
      "loss": 5.5747,
      "mean_token_accuracy": 0.1988660603761673,
      "num_tokens": 1644119.0,
      "step": 2065
    },
    {
      "entropy": 5.631876420974732,
      "epoch": 0.5852417302798982,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989261203846056,
      "loss": 5.5129,
      "mean_token_accuracy": 0.19564798325300217,
      "num_tokens": 1648106.0,
      "step": 2070
    },
    {
      "entropy": 5.628188610076904,
      "epoch": 0.586655357647724,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004989160593374379,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1943550795316696,
      "num_tokens": 1652430.0,
      "step": 2075
    },
    {
      "entropy": 5.649709939956665,
      "epoch": 0.5880689850155499,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004989059514931985,
      "loss": 5.533,
      "mean_token_accuracy": 0.20396776497364044,
      "num_tokens": 1656775.0,
      "step": 2080
    },
    {
      "entropy": 5.654207420349121,
      "epoch": 0.5894826123833757,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004988957968539997,
      "loss": 5.6254,
      "mean_token_accuracy": 0.19971878081560135,
      "num_tokens": 1661158.0,
      "step": 2085
    },
    {
      "entropy": 5.573638296127319,
      "epoch": 0.5908962397512015,
      "grad_norm": 1.296875,
      "learning_rate": 0.000498885595421964,
      "loss": 5.4981,
      "mean_token_accuracy": 0.206602144241333,
      "num_tokens": 1664839.0,
      "step": 2090
    },
    {
      "entropy": 5.506303167343139,
      "epoch": 0.5923098671190274,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004988753471992233,
      "loss": 5.3477,
      "mean_token_accuracy": 0.21307512819767,
      "num_tokens": 1668493.0,
      "step": 2095
    },
    {
      "entropy": 5.5169202327728275,
      "epoch": 0.5937234944868532,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004988650521879195,
      "loss": 5.5681,
      "mean_token_accuracy": 0.1938401862978935,
      "num_tokens": 1672376.0,
      "step": 2100
    },
    {
      "entropy": 5.6174476623535154,
      "epoch": 0.5951371218546792,
      "grad_norm": 1.3125,
      "learning_rate": 0.000498854710390204,
      "loss": 5.4617,
      "mean_token_accuracy": 0.20711683630943298,
      "num_tokens": 1676315.0,
      "step": 2105
    },
    {
      "entropy": 5.556337833404541,
      "epoch": 0.596550749222505,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004988443218082384,
      "loss": 5.5785,
      "mean_token_accuracy": 0.20104349106550218,
      "num_tokens": 1680195.0,
      "step": 2110
    },
    {
      "entropy": 5.4387730121612545,
      "epoch": 0.5979643765903307,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004988338864441937,
      "loss": 5.4152,
      "mean_token_accuracy": 0.20739189386367798,
      "num_tokens": 1684194.0,
      "step": 2115
    },
    {
      "entropy": 5.642705869674683,
      "epoch": 0.5993780039581567,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004988234043002508,
      "loss": 5.4491,
      "mean_token_accuracy": 0.2025332361459732,
      "num_tokens": 1688303.0,
      "step": 2120
    },
    {
      "entropy": 5.581110000610352,
      "epoch": 0.6007916313259825,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004988128753786006,
      "loss": 5.6538,
      "mean_token_accuracy": 0.18712435364723207,
      "num_tokens": 1692806.0,
      "step": 2125
    },
    {
      "entropy": 5.687484979629517,
      "epoch": 0.6022052586938084,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004988022996814435,
      "loss": 5.573,
      "mean_token_accuracy": 0.1949217587709427,
      "num_tokens": 1697028.0,
      "step": 2130
    },
    {
      "entropy": 5.558356523513794,
      "epoch": 0.6036188860616342,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004987916772109897,
      "loss": 5.5124,
      "mean_token_accuracy": 0.20491410195827484,
      "num_tokens": 1701117.0,
      "step": 2135
    },
    {
      "entropy": 5.507727575302124,
      "epoch": 0.60503251342946,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004987810079694591,
      "loss": 5.546,
      "mean_token_accuracy": 0.20128675401210785,
      "num_tokens": 1705157.0,
      "step": 2140
    },
    {
      "entropy": 5.534308624267578,
      "epoch": 0.6064461407972859,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987702919590819,
      "loss": 5.5511,
      "mean_token_accuracy": 0.210813270509243,
      "num_tokens": 1709476.0,
      "step": 2145
    },
    {
      "entropy": 5.639748430252075,
      "epoch": 0.6078597681651117,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004987595291820972,
      "loss": 5.5881,
      "mean_token_accuracy": 0.19741401970386505,
      "num_tokens": 1713700.0,
      "step": 2150
    },
    {
      "entropy": 5.590295839309692,
      "epoch": 0.6092733955329375,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004987487196407547,
      "loss": 5.4362,
      "mean_token_accuracy": 0.2136401653289795,
      "num_tokens": 1717628.0,
      "step": 2155
    },
    {
      "entropy": 5.4449460983276365,
      "epoch": 0.6106870229007634,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004987378633373134,
      "loss": 5.351,
      "mean_token_accuracy": 0.2092408284544945,
      "num_tokens": 1721618.0,
      "step": 2160
    },
    {
      "entropy": 5.513040208816529,
      "epoch": 0.6121006502685892,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004987269602740423,
      "loss": 5.3616,
      "mean_token_accuracy": 0.21337435394525528,
      "num_tokens": 1725527.0,
      "step": 2165
    },
    {
      "entropy": 5.4719398498535154,
      "epoch": 0.6135142776364151,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004987160104532198,
      "loss": 5.4253,
      "mean_token_accuracy": 0.20791638344526292,
      "num_tokens": 1729562.0,
      "step": 2170
    },
    {
      "entropy": 5.54003643989563,
      "epoch": 0.6149279050042409,
      "grad_norm": 1.25,
      "learning_rate": 0.0004987050138771347,
      "loss": 5.5033,
      "mean_token_accuracy": 0.20253001302480697,
      "num_tokens": 1733491.0,
      "step": 2175
    },
    {
      "entropy": 5.54003643989563,
      "epoch": 0.6163415323720667,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498693970548085,
      "loss": 5.3609,
      "mean_token_accuracy": 0.21518508344888687,
      "num_tokens": 1737309.0,
      "step": 2180
    },
    {
      "entropy": 5.562676477432251,
      "epoch": 0.6177551597398926,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986828804683787,
      "loss": 5.6043,
      "mean_token_accuracy": 0.19376907795667647,
      "num_tokens": 1741514.0,
      "step": 2185
    },
    {
      "entropy": 5.620116949081421,
      "epoch": 0.6191687871077184,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004986717436403337,
      "loss": 5.4595,
      "mean_token_accuracy": 0.21228516697883607,
      "num_tokens": 1745249.0,
      "step": 2190
    },
    {
      "entropy": 5.419637966156006,
      "epoch": 0.6205824144755443,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004986605600662773,
      "loss": 5.47,
      "mean_token_accuracy": 0.20662501752376555,
      "num_tokens": 1749318.0,
      "step": 2195
    },
    {
      "entropy": 5.462861013412476,
      "epoch": 0.6219960418433701,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004986493297485469,
      "loss": 5.5437,
      "mean_token_accuracy": 0.19386183023452758,
      "num_tokens": 1753419.0,
      "step": 2200
    },
    {
      "entropy": 5.6106353282928465,
      "epoch": 0.6234096692111959,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986380526894896,
      "loss": 5.4658,
      "mean_token_accuracy": 0.20832172632217408,
      "num_tokens": 1757337.0,
      "step": 2205
    },
    {
      "entropy": 5.447007656097412,
      "epoch": 0.6248232965790218,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004986267288914623,
      "loss": 5.4496,
      "mean_token_accuracy": 0.20539478212594986,
      "num_tokens": 1761335.0,
      "step": 2210
    },
    {
      "entropy": 5.396089172363281,
      "epoch": 0.6262369239468476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004986153583568315,
      "loss": 5.429,
      "mean_token_accuracy": 0.2138877660036087,
      "num_tokens": 1765370.0,
      "step": 2215
    },
    {
      "entropy": 5.696418952941895,
      "epoch": 0.6276505513146734,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004986039410879737,
      "loss": 5.5815,
      "mean_token_accuracy": 0.1938244104385376,
      "num_tokens": 1769272.0,
      "step": 2220
    },
    {
      "entropy": 5.5408367156982425,
      "epoch": 0.6290641786824993,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004985924770872748,
      "loss": 5.4517,
      "mean_token_accuracy": 0.20318606048822402,
      "num_tokens": 1773329.0,
      "step": 2225
    },
    {
      "entropy": 5.4070816993713375,
      "epoch": 0.6304778060503251,
      "grad_norm": 1.375,
      "learning_rate": 0.0004985809663571309,
      "loss": 5.4137,
      "mean_token_accuracy": 0.22061175405979155,
      "num_tokens": 1777223.0,
      "step": 2230
    },
    {
      "entropy": 5.562509918212891,
      "epoch": 0.631891433418151,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004985694088999476,
      "loss": 5.5138,
      "mean_token_accuracy": 0.2007150322198868,
      "num_tokens": 1781359.0,
      "step": 2235
    },
    {
      "entropy": 5.520141315460205,
      "epoch": 0.6333050607859768,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004985578047181404,
      "loss": 5.4645,
      "mean_token_accuracy": 0.20261273682117462,
      "num_tokens": 1785605.0,
      "step": 2240
    },
    {
      "entropy": 5.530501174926758,
      "epoch": 0.6347186881538026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004985461538141344,
      "loss": 5.4674,
      "mean_token_accuracy": 0.2011677071452141,
      "num_tokens": 1789187.0,
      "step": 2245
    },
    {
      "entropy": 5.574052000045777,
      "epoch": 0.6361323155216285,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004985344561903646,
      "loss": 5.5618,
      "mean_token_accuracy": 0.20011836439371108,
      "num_tokens": 1792971.0,
      "step": 2250
    },
    {
      "entropy": 5.608518457412719,
      "epoch": 0.6375459428894543,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004985227118492758,
      "loss": 5.5191,
      "mean_token_accuracy": 0.20964549332857133,
      "num_tokens": 1796938.0,
      "step": 2255
    },
    {
      "entropy": 5.61181960105896,
      "epoch": 0.6389595702572802,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985109207933224,
      "loss": 5.5303,
      "mean_token_accuracy": 0.20179047882556916,
      "num_tokens": 1801017.0,
      "step": 2260
    },
    {
      "entropy": 5.59629225730896,
      "epoch": 0.640373197625106,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004984990830249687,
      "loss": 5.5526,
      "mean_token_accuracy": 0.1967478796839714,
      "num_tokens": 1804819.0,
      "step": 2265
    },
    {
      "entropy": 5.543248891830444,
      "epoch": 0.6417868249929318,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984871985466888,
      "loss": 5.2833,
      "mean_token_accuracy": 0.22129721641540528,
      "num_tokens": 1808570.0,
      "step": 2270
    },
    {
      "entropy": 5.519654655456543,
      "epoch": 0.6432004523607577,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004984752673609663,
      "loss": 5.5729,
      "mean_token_accuracy": 0.19660354554653167,
      "num_tokens": 1812762.0,
      "step": 2275
    },
    {
      "entropy": 5.472464466094971,
      "epoch": 0.6446140797285835,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004984632894702949,
      "loss": 5.4289,
      "mean_token_accuracy": 0.20329504907131196,
      "num_tokens": 1816836.0,
      "step": 2280
    },
    {
      "entropy": 5.550273275375366,
      "epoch": 0.6460277070964093,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004984512648771779,
      "loss": 5.3748,
      "mean_token_accuracy": 0.2208165168762207,
      "num_tokens": 1820663.0,
      "step": 2285
    },
    {
      "entropy": 5.413545417785644,
      "epoch": 0.6474413344642352,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004984391935841283,
      "loss": 5.5055,
      "mean_token_accuracy": 0.2082822859287262,
      "num_tokens": 1825018.0,
      "step": 2290
    },
    {
      "entropy": 5.4688355922698975,
      "epoch": 0.648854961832061,
      "grad_norm": 1.40625,
      "learning_rate": 0.000498427075593669,
      "loss": 5.4444,
      "mean_token_accuracy": 0.2056533843278885,
      "num_tokens": 1829068.0,
      "step": 2295
    },
    {
      "entropy": 5.58225564956665,
      "epoch": 0.650268589199887,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004984149109083325,
      "loss": 5.406,
      "mean_token_accuracy": 0.21179707050323487,
      "num_tokens": 1832971.0,
      "step": 2300
    },
    {
      "entropy": 5.427003192901611,
      "epoch": 0.6516822165677127,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000498402699530661,
      "loss": 5.4303,
      "mean_token_accuracy": 0.20738290697336198,
      "num_tokens": 1837021.0,
      "step": 2305
    },
    {
      "entropy": 5.402704811096191,
      "epoch": 0.6530958439355385,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498390441463207,
      "loss": 5.3714,
      "mean_token_accuracy": 0.21459479480981827,
      "num_tokens": 1841161.0,
      "step": 2310
    },
    {
      "entropy": 5.4311668395996096,
      "epoch": 0.6545094713033645,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498378136708532,
      "loss": 5.4507,
      "mean_token_accuracy": 0.20626376271247865,
      "num_tokens": 1845219.0,
      "step": 2315
    },
    {
      "entropy": 5.5389612197875975,
      "epoch": 0.6559230986711903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004983657852692079,
      "loss": 5.4873,
      "mean_token_accuracy": 0.20499810874462127,
      "num_tokens": 1849101.0,
      "step": 2320
    },
    {
      "entropy": 5.606302070617676,
      "epoch": 0.6573367260390162,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004983533871478159,
      "loss": 5.5586,
      "mean_token_accuracy": 0.19572011679410933,
      "num_tokens": 1853160.0,
      "step": 2325
    },
    {
      "entropy": 5.409348201751709,
      "epoch": 0.658750353406842,
      "grad_norm": 1.28125,
      "learning_rate": 0.000498340942346947,
      "loss": 5.4022,
      "mean_token_accuracy": 0.2098221644759178,
      "num_tokens": 1857002.0,
      "step": 2330
    },
    {
      "entropy": 5.392949056625366,
      "epoch": 0.6601639807746678,
      "grad_norm": 1.375,
      "learning_rate": 0.0004983284508692025,
      "loss": 5.364,
      "mean_token_accuracy": 0.21255889534950256,
      "num_tokens": 1860393.0,
      "step": 2335
    },
    {
      "entropy": 5.473573160171509,
      "epoch": 0.6615776081424937,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004983159127171926,
      "loss": 5.4675,
      "mean_token_accuracy": 0.20936144888401031,
      "num_tokens": 1864240.0,
      "step": 2340
    },
    {
      "entropy": 5.47200608253479,
      "epoch": 0.6629912355103195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004983033278935379,
      "loss": 5.4102,
      "mean_token_accuracy": 0.20349428206682205,
      "num_tokens": 1868158.0,
      "step": 2345
    },
    {
      "entropy": 5.3972259044647215,
      "epoch": 0.6644048628781454,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004982906964008686,
      "loss": 5.3302,
      "mean_token_accuracy": 0.21207149028778077,
      "num_tokens": 1872136.0,
      "step": 2350
    },
    {
      "entropy": 5.50235013961792,
      "epoch": 0.6658184902459712,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982780182418245,
      "loss": 5.4773,
      "mean_token_accuracy": 0.20729459077119827,
      "num_tokens": 1876278.0,
      "step": 2355
    },
    {
      "entropy": 5.567453908920288,
      "epoch": 0.667232117613797,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004982652934190553,
      "loss": 5.486,
      "mean_token_accuracy": 0.20608954280614852,
      "num_tokens": 1880477.0,
      "step": 2360
    },
    {
      "entropy": 5.323032474517822,
      "epoch": 0.6686457449816229,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004982525219352204,
      "loss": 5.2575,
      "mean_token_accuracy": 0.22075929641723632,
      "num_tokens": 1884636.0,
      "step": 2365
    },
    {
      "entropy": 5.422608804702759,
      "epoch": 0.6700593723494487,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004982397037929891,
      "loss": 5.4597,
      "mean_token_accuracy": 0.2153452843427658,
      "num_tokens": 1888473.0,
      "step": 2370
    },
    {
      "entropy": 5.483827829360962,
      "epoch": 0.6714729997172745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004982268389950399,
      "loss": 5.412,
      "mean_token_accuracy": 0.20630247592926027,
      "num_tokens": 1892235.0,
      "step": 2375
    },
    {
      "entropy": 5.440791273117066,
      "epoch": 0.6728866270851004,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498213927544062,
      "loss": 5.4818,
      "mean_token_accuracy": 0.21008707135915755,
      "num_tokens": 1896294.0,
      "step": 2380
    },
    {
      "entropy": 5.5512748718261715,
      "epoch": 0.6743002544529262,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982009694427535,
      "loss": 5.4943,
      "mean_token_accuracy": 0.2023972362279892,
      "num_tokens": 1900517.0,
      "step": 2385
    },
    {
      "entropy": 5.477317762374878,
      "epoch": 0.6757138818207521,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004981879646938225,
      "loss": 5.406,
      "mean_token_accuracy": 0.21320468336343765,
      "num_tokens": 1904385.0,
      "step": 2390
    },
    {
      "entropy": 5.405479955673218,
      "epoch": 0.6771275091885779,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004981749132999872,
      "loss": 5.5958,
      "mean_token_accuracy": 0.19751414954662322,
      "num_tokens": 1908378.0,
      "step": 2395
    },
    {
      "entropy": 5.559717321395874,
      "epoch": 0.6785411365564037,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004981618152639749,
      "loss": 5.5465,
      "mean_token_accuracy": 0.2008774310350418,
      "num_tokens": 1912287.0,
      "step": 2400
    },
    {
      "entropy": 5.542268133163452,
      "epoch": 0.6799547639242296,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004981486705885234,
      "loss": 5.3446,
      "mean_token_accuracy": 0.21260107159614564,
      "num_tokens": 1916489.0,
      "step": 2405
    },
    {
      "entropy": 5.443733596801758,
      "epoch": 0.6813683912920554,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004981354792763796,
      "loss": 5.3265,
      "mean_token_accuracy": 0.21611038148403167,
      "num_tokens": 1920443.0,
      "step": 2410
    },
    {
      "entropy": 5.47453498840332,
      "epoch": 0.6827820186598813,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004981222413303004,
      "loss": 5.4629,
      "mean_token_accuracy": 0.21295766681432723,
      "num_tokens": 1924483.0,
      "step": 2415
    },
    {
      "entropy": 5.448499298095703,
      "epoch": 0.6841956460277071,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004981089567530525,
      "loss": 5.4585,
      "mean_token_accuracy": 0.20760513991117477,
      "num_tokens": 1928307.0,
      "step": 2420
    },
    {
      "entropy": 5.524633836746216,
      "epoch": 0.6856092733955329,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004980956255474124,
      "loss": 5.4341,
      "mean_token_accuracy": 0.21599441021680832,
      "num_tokens": 1932279.0,
      "step": 2425
    },
    {
      "entropy": 5.503360843658447,
      "epoch": 0.6870229007633588,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000498082247716166,
      "loss": 5.3663,
      "mean_token_accuracy": 0.20277948826551437,
      "num_tokens": 1936420.0,
      "step": 2430
    },
    {
      "entropy": 5.411725807189941,
      "epoch": 0.6884365281311846,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004980688232621095,
      "loss": 5.3858,
      "mean_token_accuracy": 0.2142704486846924,
      "num_tokens": 1940252.0,
      "step": 2435
    },
    {
      "entropy": 5.480589294433594,
      "epoch": 0.6898501554990104,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004980553521880482,
      "loss": 5.4548,
      "mean_token_accuracy": 0.19498221576213837,
      "num_tokens": 1944346.0,
      "step": 2440
    },
    {
      "entropy": 5.4414145946502686,
      "epoch": 0.6912637828668363,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004980418344967978,
      "loss": 5.3983,
      "mean_token_accuracy": 0.2134234771132469,
      "num_tokens": 1948330.0,
      "step": 2445
    },
    {
      "entropy": 5.435844564437867,
      "epoch": 0.6926774102346621,
      "grad_norm": 1.234375,
      "learning_rate": 0.000498028270191183,
      "loss": 5.402,
      "mean_token_accuracy": 0.20972336530685426,
      "num_tokens": 1952489.0,
      "step": 2450
    },
    {
      "entropy": 5.479365015029908,
      "epoch": 0.694091037602488,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004980146592740391,
      "loss": 5.397,
      "mean_token_accuracy": 0.20905091166496276,
      "num_tokens": 1956850.0,
      "step": 2455
    },
    {
      "entropy": 5.433628702163697,
      "epoch": 0.6955046649703138,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004980010017482104,
      "loss": 5.4489,
      "mean_token_accuracy": 0.2088480368256569,
      "num_tokens": 1960901.0,
      "step": 2460
    },
    {
      "entropy": 5.33675332069397,
      "epoch": 0.6969182923381396,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004979872976165513,
      "loss": 5.4547,
      "mean_token_accuracy": 0.2013069435954094,
      "num_tokens": 1964730.0,
      "step": 2465
    },
    {
      "entropy": 5.7374622344970705,
      "epoch": 0.6983319197059655,
      "grad_norm": 1.375,
      "learning_rate": 0.000497973546881926,
      "loss": 5.4922,
      "mean_token_accuracy": 0.19955756217241288,
      "num_tokens": 1968489.0,
      "step": 2470
    },
    {
      "entropy": 5.62591814994812,
      "epoch": 0.6997455470737913,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004979597495472081,
      "loss": 5.6041,
      "mean_token_accuracy": 0.19394168257713318,
      "num_tokens": 1972536.0,
      "step": 2475
    },
    {
      "entropy": 5.542226314544678,
      "epoch": 0.7011591744416172,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004979459056152814,
      "loss": 5.5563,
      "mean_token_accuracy": 0.1934622511267662,
      "num_tokens": 1976410.0,
      "step": 2480
    },
    {
      "entropy": 5.546902847290039,
      "epoch": 0.702572801809443,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000497932015089039,
      "loss": 5.3735,
      "mean_token_accuracy": 0.2068051278591156,
      "num_tokens": 1980294.0,
      "step": 2485
    },
    {
      "entropy": 5.400882911682129,
      "epoch": 0.7039864291772688,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004979180779713841,
      "loss": 5.3773,
      "mean_token_accuracy": 0.21438602060079576,
      "num_tokens": 1984096.0,
      "step": 2490
    },
    {
      "entropy": 5.387070846557617,
      "epoch": 0.7054000565450947,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004979040942652294,
      "loss": 5.3429,
      "mean_token_accuracy": 0.21412564367055892,
      "num_tokens": 1987714.0,
      "step": 2495
    },
    {
      "entropy": 5.4379284381866455,
      "epoch": 0.7068136839129205,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004978900639734974,
      "loss": 5.4291,
      "mean_token_accuracy": 0.2055034726858139,
      "num_tokens": 1991695.0,
      "step": 2500
    },
    {
      "entropy": 5.602289438247681,
      "epoch": 0.7082273112807463,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978759870991203,
      "loss": 5.5607,
      "mean_token_accuracy": 0.18915494084358214,
      "num_tokens": 1995735.0,
      "step": 2505
    },
    {
      "entropy": 5.417237043380737,
      "epoch": 0.7096409386485723,
      "grad_norm": 1.375,
      "learning_rate": 0.0004978618636450401,
      "loss": 5.3789,
      "mean_token_accuracy": 0.2134407415986061,
      "num_tokens": 1999642.0,
      "step": 2510
    },
    {
      "entropy": 5.299266576766968,
      "epoch": 0.711054566016398,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004978476936142085,
      "loss": 5.4159,
      "mean_token_accuracy": 0.2092890202999115,
      "num_tokens": 2003374.0,
      "step": 2515
    },
    {
      "entropy": 5.344558954238892,
      "epoch": 0.712468193384224,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000497833477009587,
      "loss": 5.3025,
      "mean_token_accuracy": 0.2184336155653,
      "num_tokens": 2007286.0,
      "step": 2520
    },
    {
      "entropy": 5.4438793659210205,
      "epoch": 0.7138818207520498,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004978192138341468,
      "loss": 5.3299,
      "mean_token_accuracy": 0.2091567873954773,
      "num_tokens": 2011245.0,
      "step": 2525
    },
    {
      "entropy": 5.371488094329834,
      "epoch": 0.7152954481198756,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978049040908688,
      "loss": 5.3136,
      "mean_token_accuracy": 0.21774005740880967,
      "num_tokens": 2015535.0,
      "step": 2530
    },
    {
      "entropy": 5.462795925140381,
      "epoch": 0.7167090754877015,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004977905477827437,
      "loss": 5.4923,
      "mean_token_accuracy": 0.19267662912607192,
      "num_tokens": 2019531.0,
      "step": 2535
    },
    {
      "entropy": 5.458331871032715,
      "epoch": 0.7181227028555273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004977761449127716,
      "loss": 5.3604,
      "mean_token_accuracy": 0.20466733276844024,
      "num_tokens": 2023665.0,
      "step": 2540
    },
    {
      "entropy": 5.399077320098877,
      "epoch": 0.7195363302233532,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497761695483963,
      "loss": 5.3461,
      "mean_token_accuracy": 0.2058019906282425,
      "num_tokens": 2027718.0,
      "step": 2545
    },
    {
      "entropy": 5.425813341140747,
      "epoch": 0.720949957591179,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004977471994993375,
      "loss": 5.4704,
      "mean_token_accuracy": 0.1954423427581787,
      "num_tokens": 2031752.0,
      "step": 2550
    },
    {
      "entropy": 5.447065830230713,
      "epoch": 0.7223635849590048,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977326569619247,
      "loss": 5.4989,
      "mean_token_accuracy": 0.20671794414520264,
      "num_tokens": 2035678.0,
      "step": 2555
    },
    {
      "entropy": 5.500088357925415,
      "epoch": 0.7237772123268307,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004977180678747639,
      "loss": 5.4082,
      "mean_token_accuracy": 0.1994786888360977,
      "num_tokens": 2039659.0,
      "step": 2560
    },
    {
      "entropy": 5.377178764343261,
      "epoch": 0.7251908396946565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004977034322409043,
      "loss": 5.3022,
      "mean_token_accuracy": 0.22200674265623094,
      "num_tokens": 2043451.0,
      "step": 2565
    },
    {
      "entropy": 5.408650875091553,
      "epoch": 0.7266044670624824,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004976887500634044,
      "loss": 5.3769,
      "mean_token_accuracy": 0.20982315689325332,
      "num_tokens": 2047388.0,
      "step": 2570
    },
    {
      "entropy": 5.363976049423218,
      "epoch": 0.7280180944303082,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004976740213453329,
      "loss": 5.3214,
      "mean_token_accuracy": 0.21539135426282882,
      "num_tokens": 2051153.0,
      "step": 2575
    },
    {
      "entropy": 5.343084049224854,
      "epoch": 0.729431721798134,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004976592460897678,
      "loss": 5.3005,
      "mean_token_accuracy": 0.22188264429569243,
      "num_tokens": 2055150.0,
      "step": 2580
    },
    {
      "entropy": 5.382208204269409,
      "epoch": 0.7308453491659599,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004976444242997973,
      "loss": 5.2374,
      "mean_token_accuracy": 0.2150484710931778,
      "num_tokens": 2059065.0,
      "step": 2585
    },
    {
      "entropy": 5.382781219482422,
      "epoch": 0.7322589765337857,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004976295559785187,
      "loss": 5.3745,
      "mean_token_accuracy": 0.198026043176651,
      "num_tokens": 2063137.0,
      "step": 2590
    },
    {
      "entropy": 5.454232454299927,
      "epoch": 0.7336726039016115,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004976146411290398,
      "loss": 5.4954,
      "mean_token_accuracy": 0.20750914961099626,
      "num_tokens": 2067112.0,
      "step": 2595
    },
    {
      "entropy": 5.412012767791748,
      "epoch": 0.7350862312694374,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004975996797544774,
      "loss": 5.4636,
      "mean_token_accuracy": 0.2137315034866333,
      "num_tokens": 2071336.0,
      "step": 2600
    },
    {
      "entropy": 5.416234493255615,
      "epoch": 0.7364998586372632,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004975846718579585,
      "loss": 5.3089,
      "mean_token_accuracy": 0.2125646710395813,
      "num_tokens": 2075072.0,
      "step": 2605
    },
    {
      "entropy": 5.363396263122558,
      "epoch": 0.7379134860050891,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004975696174426196,
      "loss": 5.3686,
      "mean_token_accuracy": 0.21096211373806,
      "num_tokens": 2079177.0,
      "step": 2610
    },
    {
      "entropy": 5.520514678955078,
      "epoch": 0.7393271133729149,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004975545165116069,
      "loss": 5.4923,
      "mean_token_accuracy": 0.20160481184720994,
      "num_tokens": 2083010.0,
      "step": 2615
    },
    {
      "entropy": 5.39557147026062,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004975393690680765,
      "loss": 5.3944,
      "mean_token_accuracy": 0.21329910457134246,
      "num_tokens": 2087225.0,
      "step": 2620
    },
    {
      "entropy": 5.355341625213623,
      "epoch": 0.7421543681085666,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975241751151942,
      "loss": 5.3517,
      "mean_token_accuracy": 0.2054790213704109,
      "num_tokens": 2091053.0,
      "step": 2625
    },
    {
      "entropy": 5.476370191574096,
      "epoch": 0.7435679954763924,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004975089346561354,
      "loss": 5.4972,
      "mean_token_accuracy": 0.20132934898138047,
      "num_tokens": 2095152.0,
      "step": 2630
    },
    {
      "entropy": 5.485976266860962,
      "epoch": 0.7449816228442183,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004974936476940851,
      "loss": 5.3576,
      "mean_token_accuracy": 0.21123178601264953,
      "num_tokens": 2099282.0,
      "step": 2635
    },
    {
      "entropy": 5.487665700912475,
      "epoch": 0.7463952502120441,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004974783142322383,
      "loss": 5.4251,
      "mean_token_accuracy": 0.20355846732854843,
      "num_tokens": 2103358.0,
      "step": 2640
    },
    {
      "entropy": 5.468116903305054,
      "epoch": 0.7478088775798699,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004974629342737995,
      "loss": 5.4438,
      "mean_token_accuracy": 0.2065952703356743,
      "num_tokens": 2107180.0,
      "step": 2645
    },
    {
      "entropy": 5.247257423400879,
      "epoch": 0.7492225049476958,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004974475078219833,
      "loss": 5.3647,
      "mean_token_accuracy": 0.21316832154989243,
      "num_tokens": 2111002.0,
      "step": 2650
    },
    {
      "entropy": 5.357695865631103,
      "epoch": 0.7506361323155216,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004974320348800134,
      "loss": 5.2408,
      "mean_token_accuracy": 0.20852022171020507,
      "num_tokens": 2114925.0,
      "step": 2655
    },
    {
      "entropy": 5.451169013977051,
      "epoch": 0.7520497596833474,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004974165154511237,
      "loss": 5.3472,
      "mean_token_accuracy": 0.21330751180648805,
      "num_tokens": 2119039.0,
      "step": 2660
    },
    {
      "entropy": 5.410284042358398,
      "epoch": 0.7534633870511733,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004974009495385578,
      "loss": 5.353,
      "mean_token_accuracy": 0.21694692373275756,
      "num_tokens": 2123215.0,
      "step": 2665
    },
    {
      "entropy": 5.412848854064942,
      "epoch": 0.7548770144189991,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004973853371455686,
      "loss": 5.3483,
      "mean_token_accuracy": 0.20225527435541152,
      "num_tokens": 2127013.0,
      "step": 2670
    },
    {
      "entropy": 5.468294858932495,
      "epoch": 0.756290641786825,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973696782754191,
      "loss": 5.4464,
      "mean_token_accuracy": 0.20311361104249953,
      "num_tokens": 2131310.0,
      "step": 2675
    },
    {
      "entropy": 5.42818603515625,
      "epoch": 0.7577042691546508,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004973539729313819,
      "loss": 5.4026,
      "mean_token_accuracy": 0.2121235892176628,
      "num_tokens": 2135203.0,
      "step": 2680
    },
    {
      "entropy": 5.377758741378784,
      "epoch": 0.7591178965224766,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004973382211167394,
      "loss": 5.3306,
      "mean_token_accuracy": 0.21791625618934632,
      "num_tokens": 2139212.0,
      "step": 2685
    },
    {
      "entropy": 5.377042150497436,
      "epoch": 0.7605315238903025,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004973224228347837,
      "loss": 5.3246,
      "mean_token_accuracy": 0.20743905156850814,
      "num_tokens": 2143215.0,
      "step": 2690
    },
    {
      "entropy": 5.336807489395142,
      "epoch": 0.7619451512581283,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004973065780888164,
      "loss": 5.3244,
      "mean_token_accuracy": 0.2237856090068817,
      "num_tokens": 2146984.0,
      "step": 2695
    },
    {
      "entropy": 5.466645050048828,
      "epoch": 0.7633587786259542,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004972906868821489,
      "loss": 5.4005,
      "mean_token_accuracy": 0.21173439621925355,
      "num_tokens": 2151107.0,
      "step": 2700
    },
    {
      "entropy": 5.389848041534424,
      "epoch": 0.76477240599378,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004972747492181024,
      "loss": 5.3846,
      "mean_token_accuracy": 0.20756343454122544,
      "num_tokens": 2155036.0,
      "step": 2705
    },
    {
      "entropy": 5.4115753173828125,
      "epoch": 0.7661860333616058,
      "grad_norm": 1.28125,
      "learning_rate": 0.000497258765100008,
      "loss": 5.3651,
      "mean_token_accuracy": 0.20176379829645158,
      "num_tokens": 2159043.0,
      "step": 2710
    },
    {
      "entropy": 5.4078697681427,
      "epoch": 0.7675996607294318,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000497242734531206,
      "loss": 5.381,
      "mean_token_accuracy": 0.2030152916908264,
      "num_tokens": 2163089.0,
      "step": 2715
    },
    {
      "entropy": 5.416038370132446,
      "epoch": 0.7690132880972576,
      "grad_norm": 1.25,
      "learning_rate": 0.0004972266575150468,
      "loss": 5.312,
      "mean_token_accuracy": 0.2068483904004097,
      "num_tokens": 2167089.0,
      "step": 2720
    },
    {
      "entropy": 5.413974142074585,
      "epoch": 0.7704269154650834,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004972105340548903,
      "loss": 5.3581,
      "mean_token_accuracy": 0.21817459911108017,
      "num_tokens": 2170899.0,
      "step": 2725
    },
    {
      "entropy": 5.337532997131348,
      "epoch": 0.7718405428329093,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004971943641541065,
      "loss": 5.4063,
      "mean_token_accuracy": 0.2055883526802063,
      "num_tokens": 2174820.0,
      "step": 2730
    },
    {
      "entropy": 5.404606342315674,
      "epoch": 0.7732541702007351,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004971781478160743,
      "loss": 5.3472,
      "mean_token_accuracy": 0.21686217188835144,
      "num_tokens": 2179240.0,
      "step": 2735
    },
    {
      "entropy": 5.41935715675354,
      "epoch": 0.774667797568561,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004971618850441833,
      "loss": 5.4785,
      "mean_token_accuracy": 0.2156490206718445,
      "num_tokens": 2183302.0,
      "step": 2740
    },
    {
      "entropy": 5.568381071090698,
      "epoch": 0.7760814249363868,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004971455758418322,
      "loss": 5.405,
      "mean_token_accuracy": 0.1920212149620056,
      "num_tokens": 2187397.0,
      "step": 2745
    },
    {
      "entropy": 5.414394855499268,
      "epoch": 0.7774950523042126,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971292202124293,
      "loss": 5.3526,
      "mean_token_accuracy": 0.21126374155282973,
      "num_tokens": 2191475.0,
      "step": 2750
    },
    {
      "entropy": 5.378679180145264,
      "epoch": 0.7789086796720385,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004971128181593931,
      "loss": 5.4867,
      "mean_token_accuracy": 0.21251560300588607,
      "num_tokens": 2195952.0,
      "step": 2755
    },
    {
      "entropy": 5.4398126125335695,
      "epoch": 0.7803223070398643,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004970963696861512,
      "loss": 5.3682,
      "mean_token_accuracy": 0.20482446253299713,
      "num_tokens": 2200555.0,
      "step": 2760
    },
    {
      "entropy": 5.51300368309021,
      "epoch": 0.7817359344076902,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004970798747961417,
      "loss": 5.3824,
      "mean_token_accuracy": 0.2078801453113556,
      "num_tokens": 2204477.0,
      "step": 2765
    },
    {
      "entropy": 5.400747537612915,
      "epoch": 0.783149561775516,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004970633334928116,
      "loss": 5.4376,
      "mean_token_accuracy": 0.19922342598438264,
      "num_tokens": 2208252.0,
      "step": 2770
    },
    {
      "entropy": 5.42206335067749,
      "epoch": 0.7845631891433418,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004970467457796179,
      "loss": 5.3334,
      "mean_token_accuracy": 0.20905719548463822,
      "num_tokens": 2211933.0,
      "step": 2775
    },
    {
      "entropy": 5.404748868942261,
      "epoch": 0.7859768165111677,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970301116600274,
      "loss": 5.3501,
      "mean_token_accuracy": 0.21212174743413925,
      "num_tokens": 2215966.0,
      "step": 2780
    },
    {
      "entropy": 5.34576644897461,
      "epoch": 0.7873904438789935,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004970134311375168,
      "loss": 5.369,
      "mean_token_accuracy": 0.20176939517259598,
      "num_tokens": 2219976.0,
      "step": 2785
    },
    {
      "entropy": 5.504501152038574,
      "epoch": 0.7888040712468194,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004969967042155717,
      "loss": 5.5131,
      "mean_token_accuracy": 0.20712943375110626,
      "num_tokens": 2224236.0,
      "step": 2790
    },
    {
      "entropy": 5.545808506011963,
      "epoch": 0.7902176986146452,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004969799308976883,
      "loss": 5.3114,
      "mean_token_accuracy": 0.22004663199186325,
      "num_tokens": 2228137.0,
      "step": 2795
    },
    {
      "entropy": 5.274337720870972,
      "epoch": 0.791631325982471,
      "grad_norm": 1.296875,
      "learning_rate": 0.000496963111187372,
      "loss": 5.1913,
      "mean_token_accuracy": 0.21963750272989274,
      "num_tokens": 2232186.0,
      "step": 2800
    },
    {
      "entropy": 5.311163997650146,
      "epoch": 0.7930449533502969,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004969462450881382,
      "loss": 5.3606,
      "mean_token_accuracy": 0.20587599128484727,
      "num_tokens": 2236183.0,
      "step": 2805
    },
    {
      "entropy": 5.4265374660491945,
      "epoch": 0.7944585807181227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004969293326035115,
      "loss": 5.3162,
      "mean_token_accuracy": 0.21735673695802687,
      "num_tokens": 2240087.0,
      "step": 2810
    },
    {
      "entropy": 5.384152603149414,
      "epoch": 0.7958722080859485,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004969123737370267,
      "loss": 5.3452,
      "mean_token_accuracy": 0.21657740622758864,
      "num_tokens": 2243905.0,
      "step": 2815
    },
    {
      "entropy": 5.386789321899414,
      "epoch": 0.7972858354537744,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004968953684922281,
      "loss": 5.3324,
      "mean_token_accuracy": 0.21099380999803544,
      "num_tokens": 2248190.0,
      "step": 2820
    },
    {
      "entropy": 5.51895809173584,
      "epoch": 0.7986994628216002,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004968783168726697,
      "loss": 5.4215,
      "mean_token_accuracy": 0.20733469277620314,
      "num_tokens": 2252375.0,
      "step": 2825
    },
    {
      "entropy": 5.399526977539063,
      "epoch": 0.8001130901894261,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004968612188819149,
      "loss": 5.2524,
      "mean_token_accuracy": 0.2202279359102249,
      "num_tokens": 2256152.0,
      "step": 2830
    },
    {
      "entropy": 5.323817729949951,
      "epoch": 0.8015267175572519,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004968440745235375,
      "loss": 5.4054,
      "mean_token_accuracy": 0.20645098388195038,
      "num_tokens": 2260023.0,
      "step": 2835
    },
    {
      "entropy": 5.362152194976806,
      "epoch": 0.8029403449250777,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004968268838011203,
      "loss": 5.2363,
      "mean_token_accuracy": 0.22098526656627654,
      "num_tokens": 2264116.0,
      "step": 2840
    },
    {
      "entropy": 5.393295431137085,
      "epoch": 0.8043539722929036,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004968096467182561,
      "loss": 5.331,
      "mean_token_accuracy": 0.21994270533323287,
      "num_tokens": 2268113.0,
      "step": 2845
    },
    {
      "entropy": 5.323547506332398,
      "epoch": 0.8057675996607294,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004967923632785473,
      "loss": 5.3022,
      "mean_token_accuracy": 0.2200825721025467,
      "num_tokens": 2271781.0,
      "step": 2850
    },
    {
      "entropy": 5.337121534347534,
      "epoch": 0.8071812270285553,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004967750334856063,
      "loss": 5.2763,
      "mean_token_accuracy": 0.21979268193244933,
      "num_tokens": 2275662.0,
      "step": 2855
    },
    {
      "entropy": 5.371746683120728,
      "epoch": 0.8085948543963811,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004967576573430546,
      "loss": 5.3663,
      "mean_token_accuracy": 0.21294018477201462,
      "num_tokens": 2279753.0,
      "step": 2860
    },
    {
      "entropy": 5.373874664306641,
      "epoch": 0.8100084817642069,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004967402348545237,
      "loss": 5.3664,
      "mean_token_accuracy": 0.21907803565263748,
      "num_tokens": 2283684.0,
      "step": 2865
    },
    {
      "entropy": 5.5256712436676025,
      "epoch": 0.8114221091320328,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004967227660236551,
      "loss": 5.4241,
      "mean_token_accuracy": 0.1930916875600815,
      "num_tokens": 2287894.0,
      "step": 2870
    },
    {
      "entropy": 5.345724725723267,
      "epoch": 0.8128357364998586,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004967052508540995,
      "loss": 5.2391,
      "mean_token_accuracy": 0.21856693029403687,
      "num_tokens": 2291833.0,
      "step": 2875
    },
    {
      "entropy": 5.253313446044922,
      "epoch": 0.8142493638676844,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004966876893495174,
      "loss": 5.233,
      "mean_token_accuracy": 0.2188587084412575,
      "num_tokens": 2295800.0,
      "step": 2880
    },
    {
      "entropy": 5.382266187667847,
      "epoch": 0.8156629912355103,
      "grad_norm": 1.359375,
      "learning_rate": 0.000496670081513579,
      "loss": 5.3665,
      "mean_token_accuracy": 0.20460723340511322,
      "num_tokens": 2299704.0,
      "step": 2885
    },
    {
      "entropy": 5.46207594871521,
      "epoch": 0.8170766186033361,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004966524273499644,
      "loss": 5.4699,
      "mean_token_accuracy": 0.1964840516448021,
      "num_tokens": 2303646.0,
      "step": 2890
    },
    {
      "entropy": 5.35508623123169,
      "epoch": 0.818490245971162,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004966347268623632,
      "loss": 5.2776,
      "mean_token_accuracy": 0.22075464427471161,
      "num_tokens": 2307587.0,
      "step": 2895
    },
    {
      "entropy": 5.351888513565063,
      "epoch": 0.8199038733389878,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004966169800544746,
      "loss": 5.3383,
      "mean_token_accuracy": 0.2160826101899147,
      "num_tokens": 2311372.0,
      "step": 2900
    },
    {
      "entropy": 5.394636487960815,
      "epoch": 0.8213175007068136,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004965991869300076,
      "loss": 5.4191,
      "mean_token_accuracy": 0.21245351135730745,
      "num_tokens": 2315804.0,
      "step": 2905
    },
    {
      "entropy": 5.418462181091309,
      "epoch": 0.8227311280746396,
      "grad_norm": 1.25,
      "learning_rate": 0.0004965813474926809,
      "loss": 5.3683,
      "mean_token_accuracy": 0.20156540721654892,
      "num_tokens": 2320095.0,
      "step": 2910
    },
    {
      "entropy": 5.445220184326172,
      "epoch": 0.8241447554424653,
      "grad_norm": 1.25,
      "learning_rate": 0.0004965634617462227,
      "loss": 5.5311,
      "mean_token_accuracy": 0.19646522551774978,
      "num_tokens": 2324324.0,
      "step": 2915
    },
    {
      "entropy": 5.497080945968628,
      "epoch": 0.8255583828102913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004965455296943712,
      "loss": 5.4288,
      "mean_token_accuracy": 0.21110617220401764,
      "num_tokens": 2328419.0,
      "step": 2920
    },
    {
      "entropy": 5.508669900894165,
      "epoch": 0.8269720101781171,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004965275513408739,
      "loss": 5.3289,
      "mean_token_accuracy": 0.21663168221712112,
      "num_tokens": 2332338.0,
      "step": 2925
    },
    {
      "entropy": 5.248326587677002,
      "epoch": 0.8283856375459429,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004965095266894884,
      "loss": 5.2622,
      "mean_token_accuracy": 0.20836566537618637,
      "num_tokens": 2336139.0,
      "step": 2930
    },
    {
      "entropy": 5.3068788051605225,
      "epoch": 0.8297992649137688,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004964914557439815,
      "loss": 5.4313,
      "mean_token_accuracy": 0.20343007743358613,
      "num_tokens": 2340222.0,
      "step": 2935
    },
    {
      "entropy": 5.519611406326294,
      "epoch": 0.8312128922815946,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004964733385081299,
      "loss": 5.4636,
      "mean_token_accuracy": 0.18783076256513595,
      "num_tokens": 2343895.0,
      "step": 2940
    },
    {
      "entropy": 5.369992256164551,
      "epoch": 0.8326265196494204,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004964551749857203,
      "loss": 5.196,
      "mean_token_accuracy": 0.22638077437877654,
      "num_tokens": 2347924.0,
      "step": 2945
    },
    {
      "entropy": 5.372833156585694,
      "epoch": 0.8340401470172463,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004964369651805484,
      "loss": 5.4979,
      "mean_token_accuracy": 0.19390685856342316,
      "num_tokens": 2352261.0,
      "step": 2950
    },
    {
      "entropy": 5.355490064620971,
      "epoch": 0.8354537743850721,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004964187090964202,
      "loss": 5.389,
      "mean_token_accuracy": 0.2094727098941803,
      "num_tokens": 2356127.0,
      "step": 2955
    },
    {
      "entropy": 5.4388844013214115,
      "epoch": 0.836867401752898,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004964004067371509,
      "loss": 5.2777,
      "mean_token_accuracy": 0.2158604696393013,
      "num_tokens": 2360318.0,
      "step": 2960
    },
    {
      "entropy": 5.473987436294555,
      "epoch": 0.8382810291207238,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004963820581065659,
      "loss": 5.3526,
      "mean_token_accuracy": 0.20864890217781068,
      "num_tokens": 2364421.0,
      "step": 2965
    },
    {
      "entropy": 5.406693792343139,
      "epoch": 0.8396946564885496,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004963636632084996,
      "loss": 5.3444,
      "mean_token_accuracy": 0.21230589747428893,
      "num_tokens": 2368360.0,
      "step": 2970
    },
    {
      "entropy": 5.381004810333252,
      "epoch": 0.8411082838563755,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004963452220467966,
      "loss": 5.2351,
      "mean_token_accuracy": 0.2133149892091751,
      "num_tokens": 2372233.0,
      "step": 2975
    },
    {
      "entropy": 5.3749185562133786,
      "epoch": 0.8425219112242013,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004963267346253108,
      "loss": 5.228,
      "mean_token_accuracy": 0.2158984884619713,
      "num_tokens": 2376175.0,
      "step": 2980
    },
    {
      "entropy": 5.213108348846435,
      "epoch": 0.8439355385920272,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963082009479063,
      "loss": 5.3595,
      "mean_token_accuracy": 0.21129231303930282,
      "num_tokens": 2380052.0,
      "step": 2985
    },
    {
      "entropy": 5.3238255977630615,
      "epoch": 0.845349165959853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004962896210184563,
      "loss": 5.4534,
      "mean_token_accuracy": 0.2019223690032959,
      "num_tokens": 2384273.0,
      "step": 2990
    },
    {
      "entropy": 5.401165962219238,
      "epoch": 0.8467627933276788,
      "grad_norm": 1.375,
      "learning_rate": 0.0004962709948408439,
      "loss": 5.3274,
      "mean_token_accuracy": 0.2195882886648178,
      "num_tokens": 2388113.0,
      "step": 2995
    },
    {
      "entropy": 5.442725038528442,
      "epoch": 0.8481764206955047,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004962523224189618,
      "loss": 5.4501,
      "mean_token_accuracy": 0.20411702394485473,
      "num_tokens": 2392345.0,
      "step": 3000
    },
    {
      "epoch": 0.8481764206955047,
      "eval_entropy": 5.174652101299088,
      "eval_loss": 5.490718364715576,
      "eval_mean_token_accuracy": 0.2128728753355781,
      "eval_num_tokens": 2392345.0,
      "eval_runtime": 4.8375,
      "eval_samples_per_second": 2332.633,
      "eval_steps_per_second": 291.682,
      "step": 3000
    },
    {
      "entropy": 5.367189741134643,
      "epoch": 0.8495900480633305,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004962336037567126,
      "loss": 5.2325,
      "mean_token_accuracy": 0.2248583048582077,
      "num_tokens": 2396304.0,
      "step": 3005
    },
    {
      "entropy": 5.238517618179321,
      "epoch": 0.8510036754311563,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004962148388580083,
      "loss": 5.2524,
      "mean_token_accuracy": 0.2163602665066719,
      "num_tokens": 2400137.0,
      "step": 3010
    },
    {
      "entropy": 5.343601322174072,
      "epoch": 0.8524173027989822,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004961960277267705,
      "loss": 5.3324,
      "mean_token_accuracy": 0.21407525837421418,
      "num_tokens": 2404047.0,
      "step": 3015
    },
    {
      "entropy": 5.330806875228882,
      "epoch": 0.853830930166808,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004961771703669308,
      "loss": 5.3214,
      "mean_token_accuracy": 0.2130121484398842,
      "num_tokens": 2408212.0,
      "step": 3020
    },
    {
      "entropy": 5.411522912979126,
      "epoch": 0.8552445575346339,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004961582667824303,
      "loss": 5.3369,
      "mean_token_accuracy": 0.2133579060435295,
      "num_tokens": 2412073.0,
      "step": 3025
    },
    {
      "entropy": 5.462206506729126,
      "epoch": 0.8566581849024597,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004961393169772196,
      "loss": 5.4226,
      "mean_token_accuracy": 0.21138060837984085,
      "num_tokens": 2416432.0,
      "step": 3030
    },
    {
      "entropy": 5.330096626281739,
      "epoch": 0.8580718122702855,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004961203209552591,
      "loss": 5.3764,
      "mean_token_accuracy": 0.21394240856170654,
      "num_tokens": 2420331.0,
      "step": 3035
    },
    {
      "entropy": 5.407485389709473,
      "epoch": 0.8594854396381114,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004961012787205188,
      "loss": 5.3311,
      "mean_token_accuracy": 0.20742064863443374,
      "num_tokens": 2424082.0,
      "step": 3040
    },
    {
      "entropy": 5.260201835632325,
      "epoch": 0.8608990670059372,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960821902769786,
      "loss": 5.2247,
      "mean_token_accuracy": 0.22140268832445145,
      "num_tokens": 2427702.0,
      "step": 3045
    },
    {
      "entropy": 5.364960336685181,
      "epoch": 0.8623126943737631,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004960630556286278,
      "loss": 5.3153,
      "mean_token_accuracy": 0.21259585320949553,
      "num_tokens": 2431589.0,
      "step": 3050
    },
    {
      "entropy": 5.382180643081665,
      "epoch": 0.8637263217415889,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004960438747794654,
      "loss": 5.3648,
      "mean_token_accuracy": 0.21003350019454955,
      "num_tokens": 2435741.0,
      "step": 3055
    },
    {
      "entropy": 5.294457149505615,
      "epoch": 0.8651399491094147,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004960246477334998,
      "loss": 5.4555,
      "mean_token_accuracy": 0.21593728959560393,
      "num_tokens": 2439529.0,
      "step": 3060
    },
    {
      "entropy": 5.353067779541016,
      "epoch": 0.8665535764772406,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960053744947499,
      "loss": 5.2017,
      "mean_token_accuracy": 0.22102957218885422,
      "num_tokens": 2443511.0,
      "step": 3065
    },
    {
      "entropy": 5.446825456619263,
      "epoch": 0.8679672038450664,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004959860550672433,
      "loss": 5.4239,
      "mean_token_accuracy": 0.2073673814535141,
      "num_tokens": 2447369.0,
      "step": 3070
    },
    {
      "entropy": 5.3776548385620115,
      "epoch": 0.8693808312128923,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004959666894550177,
      "loss": 5.4509,
      "mean_token_accuracy": 0.21306181699037552,
      "num_tokens": 2451458.0,
      "step": 3075
    },
    {
      "entropy": 5.325544977188111,
      "epoch": 0.8707944585807181,
      "grad_norm": 1.25,
      "learning_rate": 0.0004959472776621205,
      "loss": 5.2568,
      "mean_token_accuracy": 0.2175483986735344,
      "num_tokens": 2455681.0,
      "step": 3080
    },
    {
      "entropy": 5.348993444442749,
      "epoch": 0.8722080859485439,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004959278196926085,
      "loss": 5.2386,
      "mean_token_accuracy": 0.22383360266685487,
      "num_tokens": 2459887.0,
      "step": 3085
    },
    {
      "entropy": 5.301842069625854,
      "epoch": 0.8736217133163698,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004959083155505483,
      "loss": 5.2391,
      "mean_token_accuracy": 0.22726237922906875,
      "num_tokens": 2463837.0,
      "step": 3090
    },
    {
      "entropy": 5.395781373977661,
      "epoch": 0.8750353406841956,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004958887652400163,
      "loss": 5.3609,
      "mean_token_accuracy": 0.20157104581594468,
      "num_tokens": 2467714.0,
      "step": 3095
    },
    {
      "entropy": 5.279039478302002,
      "epoch": 0.8764489680520214,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004958691687650983,
      "loss": 5.3171,
      "mean_token_accuracy": 0.21099349111318588,
      "num_tokens": 2472016.0,
      "step": 3100
    },
    {
      "entropy": 5.311031198501587,
      "epoch": 0.8778625954198473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004958495261298898,
      "loss": 5.2758,
      "mean_token_accuracy": 0.22611490041017532,
      "num_tokens": 2475898.0,
      "step": 3105
    },
    {
      "entropy": 5.400108575820923,
      "epoch": 0.8792762227876731,
      "grad_norm": 1.375,
      "learning_rate": 0.0004958298373384961,
      "loss": 5.3629,
      "mean_token_accuracy": 0.2152768462896347,
      "num_tokens": 2480144.0,
      "step": 3110
    },
    {
      "entropy": 5.273153829574585,
      "epoch": 0.880689850155499,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004958101023950319,
      "loss": 5.3021,
      "mean_token_accuracy": 0.214484603703022,
      "num_tokens": 2484127.0,
      "step": 3115
    },
    {
      "entropy": 5.267424821853638,
      "epoch": 0.8821034775233249,
      "grad_norm": 1.25,
      "learning_rate": 0.0004957903213036217,
      "loss": 5.3343,
      "mean_token_accuracy": 0.22128357440233232,
      "num_tokens": 2488107.0,
      "step": 3120
    },
    {
      "entropy": 5.193635702133179,
      "epoch": 0.8835171048911507,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004957704940683998,
      "loss": 5.2381,
      "mean_token_accuracy": 0.22014840990304946,
      "num_tokens": 2491981.0,
      "step": 3125
    },
    {
      "entropy": 5.306534147262573,
      "epoch": 0.8849307322589766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957506206935097,
      "loss": 5.2006,
      "mean_token_accuracy": 0.23508219867944719,
      "num_tokens": 2496240.0,
      "step": 3130
    },
    {
      "entropy": 5.378411054611206,
      "epoch": 0.8863443596268024,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004957307011831051,
      "loss": 5.1466,
      "mean_token_accuracy": 0.22752396166324615,
      "num_tokens": 2500175.0,
      "step": 3135
    },
    {
      "entropy": 5.322124004364014,
      "epoch": 0.8877579869946283,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004957107355413489,
      "loss": 5.3099,
      "mean_token_accuracy": 0.2118844509124756,
      "num_tokens": 2503950.0,
      "step": 3140
    },
    {
      "entropy": 5.301897192001343,
      "epoch": 0.8891716143624541,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004956907237724137,
      "loss": 5.309,
      "mean_token_accuracy": 0.21629729270935058,
      "num_tokens": 2507690.0,
      "step": 3145
    },
    {
      "entropy": 5.285606002807617,
      "epoch": 0.8905852417302799,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000495670665880482,
      "loss": 5.3293,
      "mean_token_accuracy": 0.2038235917687416,
      "num_tokens": 2511645.0,
      "step": 3150
    },
    {
      "entropy": 5.364981412887573,
      "epoch": 0.8919988690981058,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004956505618697456,
      "loss": 5.4133,
      "mean_token_accuracy": 0.1998341739177704,
      "num_tokens": 2515465.0,
      "step": 3155
    },
    {
      "entropy": 5.366808462142944,
      "epoch": 0.8934124964659316,
      "grad_norm": 1.25,
      "learning_rate": 0.0004956304117444064,
      "loss": 5.3124,
      "mean_token_accuracy": 0.21686449348926545,
      "num_tokens": 2519428.0,
      "step": 3160
    },
    {
      "entropy": 5.2982847690582275,
      "epoch": 0.8948261238337574,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004956102155086754,
      "loss": 5.22,
      "mean_token_accuracy": 0.21951276957988738,
      "num_tokens": 2523346.0,
      "step": 3165
    },
    {
      "entropy": 5.305537414550781,
      "epoch": 0.8962397512015833,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004955899731667737,
      "loss": 5.3176,
      "mean_token_accuracy": 0.2135823741555214,
      "num_tokens": 2527299.0,
      "step": 3170
    },
    {
      "entropy": 5.402263402938843,
      "epoch": 0.8976533785694091,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004955696847229315,
      "loss": 5.3584,
      "mean_token_accuracy": 0.21289705336093903,
      "num_tokens": 2531492.0,
      "step": 3175
    },
    {
      "entropy": 5.349612045288086,
      "epoch": 0.899067005937235,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004955493501813892,
      "loss": 5.2509,
      "mean_token_accuracy": 0.21643438786268235,
      "num_tokens": 2535641.0,
      "step": 3180
    },
    {
      "entropy": 5.3515684604644775,
      "epoch": 0.9004806333050608,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004955289695463967,
      "loss": 5.3845,
      "mean_token_accuracy": 0.21191352903842925,
      "num_tokens": 2539674.0,
      "step": 3185
    },
    {
      "entropy": 5.308618640899658,
      "epoch": 0.9018942606728866,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004955085428222131,
      "loss": 5.2995,
      "mean_token_accuracy": 0.2106427475810051,
      "num_tokens": 2543574.0,
      "step": 3190
    },
    {
      "entropy": 5.2738823890686035,
      "epoch": 0.9033078880407125,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004954880700131077,
      "loss": 5.3746,
      "mean_token_accuracy": 0.20454230010509492,
      "num_tokens": 2547381.0,
      "step": 3195
    },
    {
      "entropy": 5.357921981811524,
      "epoch": 0.9047215154085383,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954675511233592,
      "loss": 5.3033,
      "mean_token_accuracy": 0.2107496216893196,
      "num_tokens": 2551545.0,
      "step": 3200
    },
    {
      "entropy": 5.306631946563721,
      "epoch": 0.9061351427763642,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004954469861572557,
      "loss": 5.2813,
      "mean_token_accuracy": 0.21209642738103868,
      "num_tokens": 2555336.0,
      "step": 3205
    },
    {
      "entropy": 5.221210670471192,
      "epoch": 0.90754877014419,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004954263751190954,
      "loss": 5.2695,
      "mean_token_accuracy": 0.2048604115843773,
      "num_tokens": 2559301.0,
      "step": 3210
    },
    {
      "entropy": 5.324595499038696,
      "epoch": 0.9089623975120158,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004954057180131857,
      "loss": 5.3095,
      "mean_token_accuracy": 0.20961526334285735,
      "num_tokens": 2563185.0,
      "step": 3215
    },
    {
      "entropy": 5.433643341064453,
      "epoch": 0.9103760248798417,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004953850148438439,
      "loss": 5.3546,
      "mean_token_accuracy": 0.21090478003025054,
      "num_tokens": 2567110.0,
      "step": 3220
    },
    {
      "entropy": 5.271013450622559,
      "epoch": 0.9117896522476675,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004953642656153968,
      "loss": 5.2673,
      "mean_token_accuracy": 0.210799939930439,
      "num_tokens": 2571350.0,
      "step": 3225
    },
    {
      "entropy": 5.34548921585083,
      "epoch": 0.9132032796154933,
      "grad_norm": 1.46875,
      "learning_rate": 0.000495343470332181,
      "loss": 5.2692,
      "mean_token_accuracy": 0.20986772030591966,
      "num_tokens": 2575139.0,
      "step": 3230
    },
    {
      "entropy": 5.3623948097229,
      "epoch": 0.9146169069833192,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953226289985425,
      "loss": 5.2447,
      "mean_token_accuracy": 0.20734981894493104,
      "num_tokens": 2579124.0,
      "step": 3235
    },
    {
      "entropy": 5.396227598190308,
      "epoch": 0.916030534351145,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004953017416188367,
      "loss": 5.2993,
      "mean_token_accuracy": 0.21398847103118895,
      "num_tokens": 2583130.0,
      "step": 3240
    },
    {
      "entropy": 5.31501259803772,
      "epoch": 0.9174441617189709,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004952808081974294,
      "loss": 5.3502,
      "mean_token_accuracy": 0.20861060917377472,
      "num_tokens": 2587210.0,
      "step": 3245
    },
    {
      "entropy": 5.350300121307373,
      "epoch": 0.9188577890867967,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004952598287386954,
      "loss": 5.2156,
      "mean_token_accuracy": 0.2098054587841034,
      "num_tokens": 2591372.0,
      "step": 3250
    },
    {
      "entropy": 5.358992052078247,
      "epoch": 0.9202714164546225,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004952388032470191,
      "loss": 5.3332,
      "mean_token_accuracy": 0.21697407364845275,
      "num_tokens": 2595683.0,
      "step": 3255
    },
    {
      "entropy": 5.265797424316406,
      "epoch": 0.9216850438224484,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004952177317267949,
      "loss": 5.1931,
      "mean_token_accuracy": 0.2227411612868309,
      "num_tokens": 2599472.0,
      "step": 3260
    },
    {
      "entropy": 5.270275402069092,
      "epoch": 0.9230986711902742,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004951966141824265,
      "loss": 5.1879,
      "mean_token_accuracy": 0.21839166730642318,
      "num_tokens": 2603509.0,
      "step": 3265
    },
    {
      "entropy": 5.207490396499634,
      "epoch": 0.9245122985581001,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004951754506183275,
      "loss": 5.2003,
      "mean_token_accuracy": 0.2220302402973175,
      "num_tokens": 2607191.0,
      "step": 3270
    },
    {
      "entropy": 5.3171838283538815,
      "epoch": 0.9259259259259259,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004951542410389208,
      "loss": 5.2894,
      "mean_token_accuracy": 0.2121696650981903,
      "num_tokens": 2611698.0,
      "step": 3275
    },
    {
      "entropy": 5.328981685638428,
      "epoch": 0.9273395532937517,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004951329854486392,
      "loss": 5.2769,
      "mean_token_accuracy": 0.22239854633808137,
      "num_tokens": 2615340.0,
      "step": 3280
    },
    {
      "entropy": 5.251132154464722,
      "epoch": 0.9287531806615776,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004951116838519248,
      "loss": 5.2566,
      "mean_token_accuracy": 0.21864542216062546,
      "num_tokens": 2619465.0,
      "step": 3285
    },
    {
      "entropy": 5.2473451614379885,
      "epoch": 0.9301668080294034,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004950903362532296,
      "loss": 5.1314,
      "mean_token_accuracy": 0.23277716040611268,
      "num_tokens": 2623625.0,
      "step": 3290
    },
    {
      "entropy": 5.268678426742554,
      "epoch": 0.9315804353972293,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004950689426570151,
      "loss": 5.1579,
      "mean_token_accuracy": 0.2243214190006256,
      "num_tokens": 2627368.0,
      "step": 3295
    },
    {
      "entropy": 5.337188482284546,
      "epoch": 0.9329940627650551,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004950475030677525,
      "loss": 5.2888,
      "mean_token_accuracy": 0.22068092375993728,
      "num_tokens": 2631280.0,
      "step": 3300
    },
    {
      "entropy": 5.185908603668213,
      "epoch": 0.9344076901328809,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004950260174899225,
      "loss": 5.2118,
      "mean_token_accuracy": 0.2284255415201187,
      "num_tokens": 2635420.0,
      "step": 3305
    },
    {
      "entropy": 5.215357923507691,
      "epoch": 0.9358213175007069,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004950044859280154,
      "loss": 5.183,
      "mean_token_accuracy": 0.22036392837762833,
      "num_tokens": 2639370.0,
      "step": 3310
    },
    {
      "entropy": 5.429097461700439,
      "epoch": 0.9372349448685326,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004949829083865312,
      "loss": 5.272,
      "mean_token_accuracy": 0.220812326669693,
      "num_tokens": 2643195.0,
      "step": 3315
    },
    {
      "entropy": 5.314804220199585,
      "epoch": 0.9386485722363584,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004949612848699795,
      "loss": 5.3496,
      "mean_token_accuracy": 0.20861096680164337,
      "num_tokens": 2647144.0,
      "step": 3320
    },
    {
      "entropy": 5.267690372467041,
      "epoch": 0.9400621996041844,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949396153828794,
      "loss": 5.1689,
      "mean_token_accuracy": 0.22159842997789383,
      "num_tokens": 2651201.0,
      "step": 3325
    },
    {
      "entropy": 5.304390716552734,
      "epoch": 0.9414758269720102,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004949178999297599,
      "loss": 5.2784,
      "mean_token_accuracy": 0.2142605498433113,
      "num_tokens": 2655281.0,
      "step": 3330
    },
    {
      "entropy": 5.276190900802613,
      "epoch": 0.9428894543398361,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004948961385151592,
      "loss": 5.1167,
      "mean_token_accuracy": 0.2349545329809189,
      "num_tokens": 2659226.0,
      "step": 3335
    },
    {
      "entropy": 5.247615575790405,
      "epoch": 0.9443030817076619,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004948743311436254,
      "loss": 5.1881,
      "mean_token_accuracy": 0.2129216194152832,
      "num_tokens": 2663168.0,
      "step": 3340
    },
    {
      "entropy": 5.2464855194091795,
      "epoch": 0.9457167090754877,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004948524778197159,
      "loss": 5.3619,
      "mean_token_accuracy": 0.20938823372125626,
      "num_tokens": 2667258.0,
      "step": 3345
    },
    {
      "entropy": 5.354072713851929,
      "epoch": 0.9471303364433136,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004948305785479982,
      "loss": 5.2519,
      "mean_token_accuracy": 0.20965614020824433,
      "num_tokens": 2671149.0,
      "step": 3350
    },
    {
      "entropy": 5.336177778244019,
      "epoch": 0.9485439638111394,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004948086333330488,
      "loss": 5.3275,
      "mean_token_accuracy": 0.2187487006187439,
      "num_tokens": 2675287.0,
      "step": 3355
    },
    {
      "entropy": 5.295772695541382,
      "epoch": 0.9499575911789653,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004947866421794545,
      "loss": 5.1894,
      "mean_token_accuracy": 0.21324728280305863,
      "num_tokens": 2679148.0,
      "step": 3360
    },
    {
      "entropy": 5.281546545028687,
      "epoch": 0.9513712185467911,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004947646050918109,
      "loss": 5.1983,
      "mean_token_accuracy": 0.219826440513134,
      "num_tokens": 2683125.0,
      "step": 3365
    },
    {
      "entropy": 5.169071054458618,
      "epoch": 0.9527848459146169,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000494742522074724,
      "loss": 5.2337,
      "mean_token_accuracy": 0.22465005964040757,
      "num_tokens": 2687079.0,
      "step": 3370
    },
    {
      "entropy": 5.372869873046875,
      "epoch": 0.9541984732824428,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004947203931328087,
      "loss": 5.3773,
      "mean_token_accuracy": 0.20540621131658554,
      "num_tokens": 2691065.0,
      "step": 3375
    },
    {
      "entropy": 5.290035629272461,
      "epoch": 0.9556121006502686,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004946982182706901,
      "loss": 5.2072,
      "mean_token_accuracy": 0.22314609438180924,
      "num_tokens": 2694844.0,
      "step": 3380
    },
    {
      "entropy": 5.273801517486572,
      "epoch": 0.9570257280180944,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004946759974930023,
      "loss": 5.3036,
      "mean_token_accuracy": 0.22169433981180192,
      "num_tokens": 2698932.0,
      "step": 3385
    },
    {
      "entropy": 5.1581202983856205,
      "epoch": 0.9584393553859203,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004946537308043895,
      "loss": 5.1332,
      "mean_token_accuracy": 0.21580587178468705,
      "num_tokens": 2702810.0,
      "step": 3390
    },
    {
      "entropy": 5.3055212020874025,
      "epoch": 0.9598529827537461,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004946314182095053,
      "loss": 5.3593,
      "mean_token_accuracy": 0.20648064315319062,
      "num_tokens": 2706701.0,
      "step": 3395
    },
    {
      "entropy": 5.261645174026489,
      "epoch": 0.961266610121572,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004946090597130128,
      "loss": 5.3104,
      "mean_token_accuracy": 0.20457247644662857,
      "num_tokens": 2710533.0,
      "step": 3400
    },
    {
      "entropy": 5.338420963287353,
      "epoch": 0.9626802374893978,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004945866553195848,
      "loss": 5.3013,
      "mean_token_accuracy": 0.2202003166079521,
      "num_tokens": 2714586.0,
      "step": 3405
    },
    {
      "entropy": 5.2334400653839115,
      "epoch": 0.9640938648572236,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004945642050339038,
      "loss": 5.3166,
      "mean_token_accuracy": 0.21974583864212036,
      "num_tokens": 2718357.0,
      "step": 3410
    },
    {
      "entropy": 5.491139984130859,
      "epoch": 0.9655074922250495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004945417088606617,
      "loss": 5.5482,
      "mean_token_accuracy": 0.1914077877998352,
      "num_tokens": 2722811.0,
      "step": 3415
    },
    {
      "entropy": 5.349028253555298,
      "epoch": 0.9669211195928753,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004945191668045601,
      "loss": 5.2235,
      "mean_token_accuracy": 0.22175128310918807,
      "num_tokens": 2726815.0,
      "step": 3420
    },
    {
      "entropy": 5.342743015289306,
      "epoch": 0.9683347469607012,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004944965788703101,
      "loss": 5.3457,
      "mean_token_accuracy": 0.19939375072717666,
      "num_tokens": 2730683.0,
      "step": 3425
    },
    {
      "entropy": 5.289308977127075,
      "epoch": 0.969748374328527,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004944739450626323,
      "loss": 5.1528,
      "mean_token_accuracy": 0.22440018355846406,
      "num_tokens": 2734950.0,
      "step": 3430
    },
    {
      "entropy": 5.2037581443786625,
      "epoch": 0.9711620016963528,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004944512653862572,
      "loss": 5.1708,
      "mean_token_accuracy": 0.23219923079013824,
      "num_tokens": 2738894.0,
      "step": 3435
    },
    {
      "entropy": 5.291185712814331,
      "epoch": 0.9725756290641787,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004944285398459247,
      "loss": 5.2892,
      "mean_token_accuracy": 0.2185189440846443,
      "num_tokens": 2742848.0,
      "step": 3440
    },
    {
      "entropy": 5.288788318634033,
      "epoch": 0.9739892564320045,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004944057684463844,
      "loss": 5.2676,
      "mean_token_accuracy": 0.21666843295097352,
      "num_tokens": 2746809.0,
      "step": 3445
    },
    {
      "entropy": 5.231118106842041,
      "epoch": 0.9754028837998303,
      "grad_norm": 1.375,
      "learning_rate": 0.0004943829511923952,
      "loss": 5.2465,
      "mean_token_accuracy": 0.21768148243427277,
      "num_tokens": 2750450.0,
      "step": 3450
    },
    {
      "entropy": 5.1876380920410154,
      "epoch": 0.9768165111676562,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004943600880887258,
      "loss": 5.2598,
      "mean_token_accuracy": 0.21461962908506393,
      "num_tokens": 2754572.0,
      "step": 3455
    },
    {
      "entropy": 5.342851734161377,
      "epoch": 0.978230138535482,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004943371791401544,
      "loss": 5.3819,
      "mean_token_accuracy": 0.20070458799600602,
      "num_tokens": 2758614.0,
      "step": 3460
    },
    {
      "entropy": 5.406234884262085,
      "epoch": 0.9796437659033079,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000494314224351469,
      "loss": 5.2735,
      "mean_token_accuracy": 0.21284016817808152,
      "num_tokens": 2763031.0,
      "step": 3465
    },
    {
      "entropy": 5.389446210861206,
      "epoch": 0.9810573932711337,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942912237274668,
      "loss": 5.4384,
      "mean_token_accuracy": 0.20352917462587355,
      "num_tokens": 2766953.0,
      "step": 3470
    },
    {
      "entropy": 5.326390171051026,
      "epoch": 0.9824710206389595,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494268177272955,
      "loss": 5.2986,
      "mean_token_accuracy": 0.21028649359941481,
      "num_tokens": 2770822.0,
      "step": 3475
    },
    {
      "entropy": 5.322681999206543,
      "epoch": 0.9838846480067854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00049424508499275,
      "loss": 5.2714,
      "mean_token_accuracy": 0.2100103810429573,
      "num_tokens": 2774599.0,
      "step": 3480
    },
    {
      "entropy": 5.227937269210815,
      "epoch": 0.9852982753746112,
      "grad_norm": 1.328125,
      "learning_rate": 0.000494221946891678,
      "loss": 5.2126,
      "mean_token_accuracy": 0.21849976629018783,
      "num_tokens": 2778508.0,
      "step": 3485
    },
    {
      "entropy": 5.324561738967896,
      "epoch": 0.9867119027424371,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004941987629745747,
      "loss": 5.3229,
      "mean_token_accuracy": 0.20360010415315627,
      "num_tokens": 2782535.0,
      "step": 3490
    },
    {
      "entropy": 5.3999138355255125,
      "epoch": 0.9881255301102629,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004941755332462856,
      "loss": 5.3643,
      "mean_token_accuracy": 0.21058977842330934,
      "num_tokens": 2786339.0,
      "step": 3495
    },
    {
      "entropy": 5.385817527770996,
      "epoch": 0.9895391574780887,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004941522577116652,
      "loss": 5.3058,
      "mean_token_accuracy": 0.2122144103050232,
      "num_tokens": 2790466.0,
      "step": 3500
    },
    {
      "entropy": 5.302184438705444,
      "epoch": 0.9909527848459146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004941289363755782,
      "loss": 5.2854,
      "mean_token_accuracy": 0.2152460917830467,
      "num_tokens": 2794267.0,
      "step": 3505
    },
    {
      "entropy": 5.341381645202636,
      "epoch": 0.9923664122137404,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004941055692428987,
      "loss": 5.3159,
      "mean_token_accuracy": 0.20648507475852967,
      "num_tokens": 2798394.0,
      "step": 3510
    },
    {
      "entropy": 5.291337776184082,
      "epoch": 0.9937800395815664,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00049408215631851,
      "loss": 5.2363,
      "mean_token_accuracy": 0.21810831874608994,
      "num_tokens": 2802429.0,
      "step": 3515
    },
    {
      "entropy": 5.31103663444519,
      "epoch": 0.9951936669493922,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004940586976073054,
      "loss": 5.2944,
      "mean_token_accuracy": 0.20695877224206924,
      "num_tokens": 2806303.0,
      "step": 3520
    },
    {
      "entropy": 5.2627636909484865,
      "epoch": 0.996607294317218,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004940351931141875,
      "loss": 5.2031,
      "mean_token_accuracy": 0.22894581407308578,
      "num_tokens": 2810123.0,
      "step": 3525
    },
    {
      "entropy": 5.210841369628906,
      "epoch": 0.9980209216850439,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004940116428440688,
      "loss": 5.2844,
      "mean_token_accuracy": 0.2093127340078354,
      "num_tokens": 2813991.0,
      "step": 3530
    },
    {
      "entropy": 5.245090055465698,
      "epoch": 0.9994345490528697,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004939880468018712,
      "loss": 5.2436,
      "mean_token_accuracy": 0.21966125816106796,
      "num_tokens": 2818203.0,
      "step": 3535
    },
    {
      "entropy": 5.314150047302246,
      "epoch": 1.0008481764206956,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004939644049925259,
      "loss": 5.1244,
      "mean_token_accuracy": 0.21864545345306396,
      "num_tokens": 2822334.0,
      "step": 3540
    },
    {
      "entropy": 5.473085403442383,
      "epoch": 1.0022618037885214,
      "grad_norm": 1.3125,
      "learning_rate": 0.000493940717420974,
      "loss": 5.2669,
      "mean_token_accuracy": 0.21420305520296096,
      "num_tokens": 2826277.0,
      "step": 3545
    },
    {
      "entropy": 5.315029287338257,
      "epoch": 1.0036754311563472,
      "grad_norm": 1.34375,
      "learning_rate": 0.000493916984092166,
      "loss": 5.0561,
      "mean_token_accuracy": 0.22078744620084761,
      "num_tokens": 2830093.0,
      "step": 3550
    },
    {
      "entropy": 5.174912023544311,
      "epoch": 1.005089058524173,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004938932050110621,
      "loss": 5.1113,
      "mean_token_accuracy": 0.23046851456165313,
      "num_tokens": 2833961.0,
      "step": 3555
    },
    {
      "entropy": 5.188733339309692,
      "epoch": 1.0065026858919988,
      "grad_norm": 1.375,
      "learning_rate": 0.000493869380182632,
      "loss": 5.0074,
      "mean_token_accuracy": 0.23168223947286606,
      "num_tokens": 2837923.0,
      "step": 3560
    },
    {
      "entropy": 5.145092868804932,
      "epoch": 1.0079163132598248,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004938455096118548,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23700931966304778,
      "num_tokens": 2841820.0,
      "step": 3565
    },
    {
      "entropy": 5.0993650436401365,
      "epoch": 1.0093299406276506,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004938215933037195,
      "loss": 5.1002,
      "mean_token_accuracy": 0.22340761125087738,
      "num_tokens": 2845901.0,
      "step": 3570
    },
    {
      "entropy": 5.212271213531494,
      "epoch": 1.0107435679954764,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004937976312632242,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22256288677453995,
      "num_tokens": 2849841.0,
      "step": 3575
    },
    {
      "entropy": 5.203208684921265,
      "epoch": 1.0121571953633022,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000493773623495377,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2299910753965378,
      "num_tokens": 2853601.0,
      "step": 3580
    },
    {
      "entropy": 5.274860095977783,
      "epoch": 1.013570822731128,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004937495700051955,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2231670841574669,
      "num_tokens": 2857967.0,
      "step": 3585
    },
    {
      "entropy": 5.164321994781494,
      "epoch": 1.014984450098954,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004937254707977065,
      "loss": 5.0466,
      "mean_token_accuracy": 0.23029865324497223,
      "num_tokens": 2862009.0,
      "step": 3590
    },
    {
      "entropy": 5.130255174636841,
      "epoch": 1.0163980774667798,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004937013258779465,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22396016120910645,
      "num_tokens": 2866078.0,
      "step": 3595
    },
    {
      "entropy": 5.2307158470153805,
      "epoch": 1.0178117048346056,
      "grad_norm": 1.265625,
      "learning_rate": 0.000493677135250962,
      "loss": 5.0401,
      "mean_token_accuracy": 0.22082847207784653,
      "num_tokens": 2870041.0,
      "step": 3600
    },
    {
      "entropy": 5.201313018798828,
      "epoch": 1.0192253322024314,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004936528989218084,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2307869702577591,
      "num_tokens": 2873901.0,
      "step": 3605
    },
    {
      "entropy": 5.157670402526856,
      "epoch": 1.0206389595702572,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004936286168955511,
      "loss": 5.019,
      "mean_token_accuracy": 0.22578337490558625,
      "num_tokens": 2877916.0,
      "step": 3610
    },
    {
      "entropy": 5.15597505569458,
      "epoch": 1.0220525869380832,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004936042891772647,
      "loss": 5.0217,
      "mean_token_accuracy": 0.22585746347904206,
      "num_tokens": 2882207.0,
      "step": 3615
    },
    {
      "entropy": 5.224420547485352,
      "epoch": 1.023466214305909,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004935799157720338,
      "loss": 5.0045,
      "mean_token_accuracy": 0.23484410494565963,
      "num_tokens": 2886211.0,
      "step": 3620
    },
    {
      "entropy": 5.239304113388061,
      "epoch": 1.0248798416737348,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004935554966849522,
      "loss": 5.2128,
      "mean_token_accuracy": 0.22152378410100937,
      "num_tokens": 2890167.0,
      "step": 3625
    },
    {
      "entropy": 5.135939455032348,
      "epoch": 1.0262934690415606,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004935310319211231,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22932442277669907,
      "num_tokens": 2894214.0,
      "step": 3630
    },
    {
      "entropy": 5.174535322189331,
      "epoch": 1.0277070964093864,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004935065214856599,
      "loss": 5.103,
      "mean_token_accuracy": 0.2289595425128937,
      "num_tokens": 2898423.0,
      "step": 3635
    },
    {
      "entropy": 5.129387426376343,
      "epoch": 1.0291207237772124,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004934819653836848,
      "loss": 5.0081,
      "mean_token_accuracy": 0.23242045640945436,
      "num_tokens": 2902728.0,
      "step": 3640
    },
    {
      "entropy": 5.096001625061035,
      "epoch": 1.0305343511450382,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004934573636203299,
      "loss": 5.016,
      "mean_token_accuracy": 0.22506662905216218,
      "num_tokens": 2906687.0,
      "step": 3645
    },
    {
      "entropy": 5.163566017150879,
      "epoch": 1.031947978512864,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004934327162007371,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21606790870428086,
      "num_tokens": 2910708.0,
      "step": 3650
    },
    {
      "entropy": 5.2279486656188965,
      "epoch": 1.0333616058806898,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004934080231300573,
      "loss": 5.0734,
      "mean_token_accuracy": 0.22245530635118485,
      "num_tokens": 2914416.0,
      "step": 3655
    },
    {
      "entropy": 5.254195880889893,
      "epoch": 1.0347752332485156,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004933832844134511,
      "loss": 5.0802,
      "mean_token_accuracy": 0.230830354988575,
      "num_tokens": 2918738.0,
      "step": 3660
    },
    {
      "entropy": 5.142400598526001,
      "epoch": 1.0361888606163416,
      "grad_norm": 1.25,
      "learning_rate": 0.0004933585000560891,
      "loss": 5.0644,
      "mean_token_accuracy": 0.23004631847143173,
      "num_tokens": 2922575.0,
      "step": 3665
    },
    {
      "entropy": 5.303077745437622,
      "epoch": 1.0376024879841674,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004933336700631508,
      "loss": 5.027,
      "mean_token_accuracy": 0.22817589342594147,
      "num_tokens": 2926707.0,
      "step": 3670
    },
    {
      "entropy": 5.243477392196655,
      "epoch": 1.0390161153519932,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004933087944398256,
      "loss": 5.0686,
      "mean_token_accuracy": 0.22908651530742646,
      "num_tokens": 2930785.0,
      "step": 3675
    },
    {
      "entropy": 5.0108260154724125,
      "epoch": 1.040429742719819,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004932838731913124,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2431721419095993,
      "num_tokens": 2934893.0,
      "step": 3680
    },
    {
      "entropy": 5.130848550796509,
      "epoch": 1.0418433700876448,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004932589063228194,
      "loss": 5.0961,
      "mean_token_accuracy": 0.21659439504146577,
      "num_tokens": 2939034.0,
      "step": 3685
    },
    {
      "entropy": 5.22723536491394,
      "epoch": 1.0432569974554706,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004932338938395647,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22529290467500687,
      "num_tokens": 2943089.0,
      "step": 3690
    },
    {
      "entropy": 5.192731094360352,
      "epoch": 1.0446706248232966,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004932088357467757,
      "loss": 5.0573,
      "mean_token_accuracy": 0.22995467334985734,
      "num_tokens": 2947128.0,
      "step": 3695
    },
    {
      "entropy": 5.1997706413269045,
      "epoch": 1.0460842521911224,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004931837320496894,
      "loss": 5.1278,
      "mean_token_accuracy": 0.21710651665925979,
      "num_tokens": 2951295.0,
      "step": 3700
    },
    {
      "entropy": 5.152848958969116,
      "epoch": 1.0474978795589482,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004931585827535524,
      "loss": 5.1319,
      "mean_token_accuracy": 0.21446435153484344,
      "num_tokens": 2955522.0,
      "step": 3705
    },
    {
      "entropy": 5.167764568328858,
      "epoch": 1.048911506926774,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004931333878636205,
      "loss": 5.0052,
      "mean_token_accuracy": 0.22012700587511064,
      "num_tokens": 2959410.0,
      "step": 3710
    },
    {
      "entropy": 5.125832176208496,
      "epoch": 1.0503251342945998,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004931081473851595,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2211052656173706,
      "num_tokens": 2963411.0,
      "step": 3715
    },
    {
      "entropy": 5.104986000061035,
      "epoch": 1.0517387616624259,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004930828613234446,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2301958531141281,
      "num_tokens": 2967623.0,
      "step": 3720
    },
    {
      "entropy": 5.2205822467803955,
      "epoch": 1.0531523890302517,
      "grad_norm": 1.296875,
      "learning_rate": 0.00049305752968376,
      "loss": 4.9666,
      "mean_token_accuracy": 0.23770354986190795,
      "num_tokens": 2971708.0,
      "step": 3725
    },
    {
      "entropy": 5.120276355743409,
      "epoch": 1.0545660163980775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004930321524714002,
      "loss": 5.0071,
      "mean_token_accuracy": 0.24287337809801102,
      "num_tokens": 2975491.0,
      "step": 3730
    },
    {
      "entropy": 5.174483919143677,
      "epoch": 1.0559796437659033,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004930067296916689,
      "loss": 5.0891,
      "mean_token_accuracy": 0.22007248550653458,
      "num_tokens": 2979368.0,
      "step": 3735
    },
    {
      "entropy": 5.120765161514282,
      "epoch": 1.057393271133729,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004929812613498791,
      "loss": 5.015,
      "mean_token_accuracy": 0.23333019763231277,
      "num_tokens": 2983939.0,
      "step": 3740
    },
    {
      "entropy": 5.077562236785889,
      "epoch": 1.058806898501555,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004929557474513537,
      "loss": 4.9594,
      "mean_token_accuracy": 0.22762384116649628,
      "num_tokens": 2987895.0,
      "step": 3745
    },
    {
      "entropy": 5.128376579284668,
      "epoch": 1.0602205258693809,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004929301880014249,
      "loss": 5.0219,
      "mean_token_accuracy": 0.22629786878824235,
      "num_tokens": 2992017.0,
      "step": 3750
    },
    {
      "entropy": 5.119693613052368,
      "epoch": 1.0616341532372067,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004929045830054344,
      "loss": 5.0029,
      "mean_token_accuracy": 0.2173638328909874,
      "num_tokens": 2995957.0,
      "step": 3755
    },
    {
      "entropy": 5.0717541694641115,
      "epoch": 1.0630477806050325,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004928789324687335,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2484455719590187,
      "num_tokens": 2999816.0,
      "step": 3760
    },
    {
      "entropy": 5.144631576538086,
      "epoch": 1.0644614079728583,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004928532363966831,
      "loss": 5.0167,
      "mean_token_accuracy": 0.22890725880861282,
      "num_tokens": 3003801.0,
      "step": 3765
    },
    {
      "entropy": 5.1410760402679445,
      "epoch": 1.0658750353406843,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004928274947946534,
      "loss": 5.117,
      "mean_token_accuracy": 0.22192999273538588,
      "num_tokens": 3007710.0,
      "step": 3770
    },
    {
      "entropy": 5.215381050109864,
      "epoch": 1.06728866270851,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004928017076680244,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2226067766547203,
      "num_tokens": 3011573.0,
      "step": 3775
    },
    {
      "entropy": 5.158493137359619,
      "epoch": 1.0687022900763359,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004927758750221852,
      "loss": 5.0654,
      "mean_token_accuracy": 0.2298080265522003,
      "num_tokens": 3015832.0,
      "step": 3780
    },
    {
      "entropy": 5.057914161682129,
      "epoch": 1.0701159174441617,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004927499968625349,
      "loss": 4.9694,
      "mean_token_accuracy": 0.23105335980653763,
      "num_tokens": 3019878.0,
      "step": 3785
    },
    {
      "entropy": 5.060720205307007,
      "epoch": 1.0715295448119875,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004927240731944818,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2461293011903763,
      "num_tokens": 3023725.0,
      "step": 3790
    },
    {
      "entropy": 5.2345055103302,
      "epoch": 1.0729431721798135,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004926981040234437,
      "loss": 5.15,
      "mean_token_accuracy": 0.21904803365468978,
      "num_tokens": 3028055.0,
      "step": 3795
    },
    {
      "entropy": 5.071320819854736,
      "epoch": 1.0743567995476393,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004926720893548482,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2409299835562706,
      "num_tokens": 3032214.0,
      "step": 3800
    },
    {
      "entropy": 5.14586877822876,
      "epoch": 1.075770426915465,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004926460291941319,
      "loss": 4.979,
      "mean_token_accuracy": 0.23053666800260544,
      "num_tokens": 3036507.0,
      "step": 3805
    },
    {
      "entropy": 5.149288845062256,
      "epoch": 1.077184054283291,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004926199235467414,
      "loss": 5.1175,
      "mean_token_accuracy": 0.22399256825447084,
      "num_tokens": 3040720.0,
      "step": 3810
    },
    {
      "entropy": 5.175496530532837,
      "epoch": 1.0785976816511167,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004925937724181326,
      "loss": 5.0237,
      "mean_token_accuracy": 0.22914376705884934,
      "num_tokens": 3044826.0,
      "step": 3815
    },
    {
      "entropy": 5.172716474533081,
      "epoch": 1.0800113090189427,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004925675758137711,
      "loss": 5.0337,
      "mean_token_accuracy": 0.22899835407733918,
      "num_tokens": 3048840.0,
      "step": 3820
    },
    {
      "entropy": 5.114548110961914,
      "epoch": 1.0814249363867685,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004925413337391315,
      "loss": 4.9248,
      "mean_token_accuracy": 0.23229692578315736,
      "num_tokens": 3052716.0,
      "step": 3825
    },
    {
      "entropy": 5.117950773239135,
      "epoch": 1.0828385637545943,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925150461996984,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22668319940567017,
      "num_tokens": 3056662.0,
      "step": 3830
    },
    {
      "entropy": 5.154039525985718,
      "epoch": 1.08425219112242,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004924887132009656,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2286270946264267,
      "num_tokens": 3060638.0,
      "step": 3835
    },
    {
      "entropy": 5.152283954620361,
      "epoch": 1.085665818490246,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004924623347484369,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22818633615970613,
      "num_tokens": 3064438.0,
      "step": 3840
    },
    {
      "entropy": 5.098384666442871,
      "epoch": 1.087079445858072,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004924359108476248,
      "loss": 4.9041,
      "mean_token_accuracy": 0.23847402185201644,
      "num_tokens": 3068099.0,
      "step": 3845
    },
    {
      "entropy": 5.026648664474488,
      "epoch": 1.0884930732258977,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004924094415040519,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23882860988378524,
      "num_tokens": 3072039.0,
      "step": 3850
    },
    {
      "entropy": 4.981844711303711,
      "epoch": 1.0899067005937235,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004923829267232503,
      "loss": 5.0414,
      "mean_token_accuracy": 0.23245490938425065,
      "num_tokens": 3076110.0,
      "step": 3855
    },
    {
      "entropy": 5.15345573425293,
      "epoch": 1.0913203279615493,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004923563665107612,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23124676197767258,
      "num_tokens": 3080514.0,
      "step": 3860
    },
    {
      "entropy": 5.18023567199707,
      "epoch": 1.0927339553293751,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004923297608721355,
      "loss": 5.0185,
      "mean_token_accuracy": 0.22580058872699738,
      "num_tokens": 3084537.0,
      "step": 3865
    },
    {
      "entropy": 5.120248746871948,
      "epoch": 1.094147582697201,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004923031098129338,
      "loss": 5.0458,
      "mean_token_accuracy": 0.23244117200374603,
      "num_tokens": 3088107.0,
      "step": 3870
    },
    {
      "entropy": 5.083855247497558,
      "epoch": 1.095561210065027,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004922764133387259,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22794615626335143,
      "num_tokens": 3091790.0,
      "step": 3875
    },
    {
      "entropy": 5.110710859298706,
      "epoch": 1.0969748374328527,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004922496714550912,
      "loss": 4.897,
      "mean_token_accuracy": 0.23963625431060792,
      "num_tokens": 3095757.0,
      "step": 3880
    },
    {
      "entropy": 5.155925369262695,
      "epoch": 1.0983884648006785,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004922228841676186,
      "loss": 5.0613,
      "mean_token_accuracy": 0.22538823187351226,
      "num_tokens": 3100172.0,
      "step": 3885
    },
    {
      "entropy": 5.141041469573975,
      "epoch": 1.0998020921685043,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004921960514819065,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20952586680650712,
      "num_tokens": 3103857.0,
      "step": 3890
    },
    {
      "entropy": 5.1645832538604735,
      "epoch": 1.1012157195363301,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004921691734035627,
      "loss": 5.0788,
      "mean_token_accuracy": 0.22629565447568895,
      "num_tokens": 3107817.0,
      "step": 3895
    },
    {
      "entropy": 5.159869527816772,
      "epoch": 1.1026293469041561,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004921422499382046,
      "loss": 4.9976,
      "mean_token_accuracy": 0.23880043029785156,
      "num_tokens": 3111655.0,
      "step": 3900
    },
    {
      "entropy": 5.2506067752838135,
      "epoch": 1.104042974271982,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000492115281091459,
      "loss": 5.1255,
      "mean_token_accuracy": 0.2335188254714012,
      "num_tokens": 3115959.0,
      "step": 3905
    },
    {
      "entropy": 5.076710176467896,
      "epoch": 1.1054566016398077,
      "grad_norm": 1.375,
      "learning_rate": 0.0004920882668689623,
      "loss": 4.858,
      "mean_token_accuracy": 0.23754922300577164,
      "num_tokens": 3119930.0,
      "step": 3910
    },
    {
      "entropy": 5.011963653564453,
      "epoch": 1.1068702290076335,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004920612072763603,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2260516107082367,
      "num_tokens": 3123806.0,
      "step": 3915
    },
    {
      "entropy": 5.127112245559692,
      "epoch": 1.1082838563754593,
      "grad_norm": 1.375,
      "learning_rate": 0.0004920341023193082,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2268049016594887,
      "num_tokens": 3127942.0,
      "step": 3920
    },
    {
      "entropy": 5.10136833190918,
      "epoch": 1.1096974837432854,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004920069520034708,
      "loss": 5.052,
      "mean_token_accuracy": 0.22814164012670518,
      "num_tokens": 3132014.0,
      "step": 3925
    },
    {
      "entropy": 5.146014165878296,
      "epoch": 1.1111111111111112,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004919797563345224,
      "loss": 5.0784,
      "mean_token_accuracy": 0.23491773158311843,
      "num_tokens": 3135854.0,
      "step": 3930
    },
    {
      "entropy": 5.228779458999634,
      "epoch": 1.112524738478937,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004919525153181466,
      "loss": 5.1213,
      "mean_token_accuracy": 0.2265205755829811,
      "num_tokens": 3139804.0,
      "step": 3935
    },
    {
      "entropy": 5.204964113235474,
      "epoch": 1.1139383658467628,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004919252289600367,
      "loss": 5.1843,
      "mean_token_accuracy": 0.22286111414432525,
      "num_tokens": 3143473.0,
      "step": 3940
    },
    {
      "entropy": 5.157012414932251,
      "epoch": 1.1153519932145886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004918978972658954,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22547006607055664,
      "num_tokens": 3147626.0,
      "step": 3945
    },
    {
      "entropy": 5.1330376148223875,
      "epoch": 1.1167656205824146,
      "grad_norm": 1.375,
      "learning_rate": 0.0004918705202414348,
      "loss": 5.08,
      "mean_token_accuracy": 0.21345378905534745,
      "num_tokens": 3151758.0,
      "step": 3950
    },
    {
      "entropy": 5.0832582950592045,
      "epoch": 1.1181792479502404,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004918430978923768,
      "loss": 4.969,
      "mean_token_accuracy": 0.23154980689287186,
      "num_tokens": 3155490.0,
      "step": 3955
    },
    {
      "entropy": 5.252908706665039,
      "epoch": 1.1195928753180662,
      "grad_norm": 1.390625,
      "learning_rate": 0.000491815630224452,
      "loss": 5.144,
      "mean_token_accuracy": 0.22490071952342988,
      "num_tokens": 3159385.0,
      "step": 3960
    },
    {
      "entropy": 5.264437961578369,
      "epoch": 1.121006502685892,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004917881172434012,
      "loss": 5.1495,
      "mean_token_accuracy": 0.21658464819192885,
      "num_tokens": 3163198.0,
      "step": 3965
    },
    {
      "entropy": 5.203642702102661,
      "epoch": 1.1224201300537178,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004917605589549746,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22007464617490768,
      "num_tokens": 3167140.0,
      "step": 3970
    },
    {
      "entropy": 4.9864466190338135,
      "epoch": 1.1238337574215436,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004917329553649315,
      "loss": 4.8914,
      "mean_token_accuracy": 0.24323611706495285,
      "num_tokens": 3170782.0,
      "step": 3975
    },
    {
      "entropy": 5.085910511016846,
      "epoch": 1.1252473847893696,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004917053064790411,
      "loss": 4.9655,
      "mean_token_accuracy": 0.23082458823919297,
      "num_tokens": 3174678.0,
      "step": 3980
    },
    {
      "entropy": 5.158240079879761,
      "epoch": 1.1266610121571954,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004916776123030815,
      "loss": 5.0579,
      "mean_token_accuracy": 0.22890633344650269,
      "num_tokens": 3178569.0,
      "step": 3985
    },
    {
      "entropy": 5.148614931106567,
      "epoch": 1.1280746395250212,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004916498728428411,
      "loss": 4.9417,
      "mean_token_accuracy": 0.23426750153303147,
      "num_tokens": 3182328.0,
      "step": 3990
    },
    {
      "entropy": 5.047394418716431,
      "epoch": 1.129488266892847,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004916220881041168,
      "loss": 4.9928,
      "mean_token_accuracy": 0.2306484341621399,
      "num_tokens": 3186477.0,
      "step": 3995
    },
    {
      "entropy": 5.06369194984436,
      "epoch": 1.1309018942606728,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004915942580927156,
      "loss": 5.0541,
      "mean_token_accuracy": 0.22101714909076692,
      "num_tokens": 3190517.0,
      "step": 4000
    },
    {
      "epoch": 1.1309018942606728,
      "eval_entropy": 4.887318427304187,
      "eval_loss": 5.3592329025268555,
      "eval_mean_token_accuracy": 0.2222705981426844,
      "eval_num_tokens": 3190517.0,
      "eval_runtime": 4.8277,
      "eval_samples_per_second": 2337.345,
      "eval_steps_per_second": 292.272,
      "step": 4000
    },
    {
      "entropy": 5.089830827713013,
      "epoch": 1.1323155216284988,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000491566382814454,
      "loss": 4.986,
      "mean_token_accuracy": 0.22719856947660447,
      "num_tokens": 3194434.0,
      "step": 4005
    },
    {
      "entropy": 5.1518348217010494,
      "epoch": 1.1337291489963246,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004915384622751575,
      "loss": 5.0251,
      "mean_token_accuracy": 0.23961628675460817,
      "num_tokens": 3198365.0,
      "step": 4010
    },
    {
      "entropy": 5.147364664077759,
      "epoch": 1.1351427763641504,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004915104964806615,
      "loss": 5.0322,
      "mean_token_accuracy": 0.23171922713518142,
      "num_tokens": 3202272.0,
      "step": 4015
    },
    {
      "entropy": 5.124147176742554,
      "epoch": 1.1365564037319762,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004914824854368105,
      "loss": 5.1038,
      "mean_token_accuracy": 0.22330406904220582,
      "num_tokens": 3206151.0,
      "step": 4020
    },
    {
      "entropy": 5.177210521697998,
      "epoch": 1.137970031099802,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004914544291494588,
      "loss": 5.026,
      "mean_token_accuracy": 0.22658679038286209,
      "num_tokens": 3210038.0,
      "step": 4025
    },
    {
      "entropy": 5.114462995529175,
      "epoch": 1.139383658467628,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004914263276244699,
      "loss": 5.0148,
      "mean_token_accuracy": 0.23395077884197235,
      "num_tokens": 3213883.0,
      "step": 4030
    },
    {
      "entropy": 5.121993732452393,
      "epoch": 1.1407972858354538,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004913981808677168,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22724643349647522,
      "num_tokens": 3217861.0,
      "step": 4035
    },
    {
      "entropy": 5.014950227737427,
      "epoch": 1.1422109132032796,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004913699888850822,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2441011130809784,
      "num_tokens": 3221492.0,
      "step": 4040
    },
    {
      "entropy": 5.179648542404175,
      "epoch": 1.1436245405711054,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004913417516824578,
      "loss": 5.0555,
      "mean_token_accuracy": 0.22246990650892257,
      "num_tokens": 3225504.0,
      "step": 4045
    },
    {
      "entropy": 5.206320714950562,
      "epoch": 1.1450381679389312,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004913134692657451,
      "loss": 5.1326,
      "mean_token_accuracy": 0.22411256581544875,
      "num_tokens": 3229722.0,
      "step": 4050
    },
    {
      "entropy": 5.113077163696289,
      "epoch": 1.1464517953067572,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000491285141640855,
      "loss": 5.0394,
      "mean_token_accuracy": 0.22567428946495055,
      "num_tokens": 3233865.0,
      "step": 4055
    },
    {
      "entropy": 5.120617341995239,
      "epoch": 1.147865422674583,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004912567688137078,
      "loss": 5.0259,
      "mean_token_accuracy": 0.22035330682992935,
      "num_tokens": 3237586.0,
      "step": 4060
    },
    {
      "entropy": 5.041660642623901,
      "epoch": 1.1492790500424088,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004912283507902333,
      "loss": 5.0148,
      "mean_token_accuracy": 0.23355903923511506,
      "num_tokens": 3241431.0,
      "step": 4065
    },
    {
      "entropy": 5.196377897262574,
      "epoch": 1.1506926774102346,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004911998875763704,
      "loss": 5.0259,
      "mean_token_accuracy": 0.23113097995519638,
      "num_tokens": 3245427.0,
      "step": 4070
    },
    {
      "entropy": 5.144703388214111,
      "epoch": 1.1521063047780604,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004911713791780681,
      "loss": 4.9791,
      "mean_token_accuracy": 0.23430070281028748,
      "num_tokens": 3249361.0,
      "step": 4075
    },
    {
      "entropy": 5.179360437393188,
      "epoch": 1.1535199321458864,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004911428256012841,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22397041469812393,
      "num_tokens": 3253181.0,
      "step": 4080
    },
    {
      "entropy": 4.99040265083313,
      "epoch": 1.1549335595137122,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004911142268519864,
      "loss": 4.8911,
      "mean_token_accuracy": 0.242074853181839,
      "num_tokens": 3257036.0,
      "step": 4085
    },
    {
      "entropy": 5.019851207733154,
      "epoch": 1.156347186881538,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004910855829361515,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2265746369957924,
      "num_tokens": 3261180.0,
      "step": 4090
    },
    {
      "entropy": 5.064997768402099,
      "epoch": 1.1577608142493638,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004910568938597661,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2436510443687439,
      "num_tokens": 3265502.0,
      "step": 4095
    },
    {
      "entropy": 5.130299806594849,
      "epoch": 1.1591744416171896,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004910281596288259,
      "loss": 5.0922,
      "mean_token_accuracy": 0.22970363199710847,
      "num_tokens": 3269261.0,
      "step": 4100
    },
    {
      "entropy": 5.167954015731811,
      "epoch": 1.1605880689850157,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909993802493362,
      "loss": 4.9864,
      "mean_token_accuracy": 0.23749542534351348,
      "num_tokens": 3273700.0,
      "step": 4105
    },
    {
      "entropy": 5.148645830154419,
      "epoch": 1.1620016963528415,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004909705557273118,
      "loss": 5.0268,
      "mean_token_accuracy": 0.2318521648645401,
      "num_tokens": 3277829.0,
      "step": 4110
    },
    {
      "entropy": 5.140073251724243,
      "epoch": 1.1634153237206672,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004909416860687767,
      "loss": 5.0641,
      "mean_token_accuracy": 0.2342190682888031,
      "num_tokens": 3281833.0,
      "step": 4115
    },
    {
      "entropy": 5.053380632400513,
      "epoch": 1.164828951088493,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004909127712797646,
      "loss": 4.935,
      "mean_token_accuracy": 0.23492969572544098,
      "num_tokens": 3285539.0,
      "step": 4120
    },
    {
      "entropy": 5.14288444519043,
      "epoch": 1.1662425784563188,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004908838113663185,
      "loss": 5.0606,
      "mean_token_accuracy": 0.22649090141057968,
      "num_tokens": 3289199.0,
      "step": 4125
    },
    {
      "entropy": 5.126579427719117,
      "epoch": 1.1676562058241449,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004908548063344907,
      "loss": 4.974,
      "mean_token_accuracy": 0.23450205028057097,
      "num_tokens": 3292892.0,
      "step": 4130
    },
    {
      "entropy": 5.168074131011963,
      "epoch": 1.1690698331919707,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004908257561903433,
      "loss": 4.9945,
      "mean_token_accuracy": 0.2271781012415886,
      "num_tokens": 3297059.0,
      "step": 4135
    },
    {
      "entropy": 5.021758937835694,
      "epoch": 1.1704834605597965,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004907966609399474,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2469888985157013,
      "num_tokens": 3300918.0,
      "step": 4140
    },
    {
      "entropy": 4.976537895202637,
      "epoch": 1.1718970879276223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004907675205893838,
      "loss": 4.8753,
      "mean_token_accuracy": 0.25018723756074907,
      "num_tokens": 3305087.0,
      "step": 4145
    },
    {
      "entropy": 5.09632420539856,
      "epoch": 1.173310715295448,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004907383351447428,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21508849114179612,
      "num_tokens": 3309439.0,
      "step": 4150
    },
    {
      "entropy": 5.125357389450073,
      "epoch": 1.174724342663274,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004907091046121236,
      "loss": 5.0395,
      "mean_token_accuracy": 0.22342185229063033,
      "num_tokens": 3313423.0,
      "step": 4155
    },
    {
      "entropy": 5.1657825946807865,
      "epoch": 1.1761379700310999,
      "grad_norm": 1.375,
      "learning_rate": 0.0004906798289976357,
      "loss": 5.0062,
      "mean_token_accuracy": 0.23468360155820847,
      "num_tokens": 3317300.0,
      "step": 4160
    },
    {
      "entropy": 5.195534896850586,
      "epoch": 1.1775515973989257,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004906505083073971,
      "loss": 5.1056,
      "mean_token_accuracy": 0.2214929237961769,
      "num_tokens": 3321572.0,
      "step": 4165
    },
    {
      "entropy": 5.180498218536377,
      "epoch": 1.1789652247667515,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004906211425475358,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21681768000125884,
      "num_tokens": 3325535.0,
      "step": 4170
    },
    {
      "entropy": 5.159073781967163,
      "epoch": 1.1803788521345773,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004905917317241891,
      "loss": 5.0094,
      "mean_token_accuracy": 0.2324025958776474,
      "num_tokens": 3329260.0,
      "step": 4175
    },
    {
      "entropy": 5.183757495880127,
      "epoch": 1.181792479502403,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004905622758435038,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22742208689451218,
      "num_tokens": 3333492.0,
      "step": 4180
    },
    {
      "entropy": 5.201954412460327,
      "epoch": 1.183206106870229,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004905327749116356,
      "loss": 5.1086,
      "mean_token_accuracy": 0.22647471427917482,
      "num_tokens": 3337730.0,
      "step": 4185
    },
    {
      "entropy": 5.085085058212281,
      "epoch": 1.1846197342380549,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004905032289347504,
      "loss": 5.0962,
      "mean_token_accuracy": 0.2242988422513008,
      "num_tokens": 3341699.0,
      "step": 4190
    },
    {
      "entropy": 5.117003870010376,
      "epoch": 1.1860333616058807,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004904736379190229,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21530511230230331,
      "num_tokens": 3345730.0,
      "step": 4195
    },
    {
      "entropy": 5.154226636886596,
      "epoch": 1.1874469889737065,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004904440018706377,
      "loss": 4.9989,
      "mean_token_accuracy": 0.23489325046539306,
      "num_tokens": 3349534.0,
      "step": 4200
    },
    {
      "entropy": 5.090934085845947,
      "epoch": 1.1888606163415323,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004904143207957882,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23164262771606445,
      "num_tokens": 3353279.0,
      "step": 4205
    },
    {
      "entropy": 4.95575795173645,
      "epoch": 1.1902742437093583,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004903845947006778,
      "loss": 5.0447,
      "mean_token_accuracy": 0.22197383940219878,
      "num_tokens": 3357326.0,
      "step": 4210
    },
    {
      "entropy": 5.136634111404419,
      "epoch": 1.191687871077184,
      "grad_norm": 1.296875,
      "learning_rate": 0.000490354823591519,
      "loss": 5.0307,
      "mean_token_accuracy": 0.22976124435663223,
      "num_tokens": 3361219.0,
      "step": 4215
    },
    {
      "entropy": 5.2035078525543215,
      "epoch": 1.19310149844501,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004903250074745338,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21598300635814666,
      "num_tokens": 3365136.0,
      "step": 4220
    },
    {
      "entropy": 5.111600542068482,
      "epoch": 1.1945151258128357,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004902951463559535,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2354698285460472,
      "num_tokens": 3369020.0,
      "step": 4225
    },
    {
      "entropy": 5.1551628589630125,
      "epoch": 1.1959287531806615,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000490265240242019,
      "loss": 5.1141,
      "mean_token_accuracy": 0.21577179580926895,
      "num_tokens": 3373176.0,
      "step": 4230
    },
    {
      "entropy": 5.060961151123047,
      "epoch": 1.1973423805484873,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004902352891389804,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2261984169483185,
      "num_tokens": 3377033.0,
      "step": 4235
    },
    {
      "entropy": 5.218928480148316,
      "epoch": 1.1987560079163133,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004902052930530975,
      "loss": 5.0497,
      "mean_token_accuracy": 0.22233416140079498,
      "num_tokens": 3381011.0,
      "step": 4240
    },
    {
      "entropy": 5.117517137527466,
      "epoch": 1.2001696352841391,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004901752519906391,
      "loss": 4.9789,
      "mean_token_accuracy": 0.23157614320516587,
      "num_tokens": 3384960.0,
      "step": 4245
    },
    {
      "entropy": 4.970371389389038,
      "epoch": 1.201583262651965,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004901451659578836,
      "loss": 4.8736,
      "mean_token_accuracy": 0.24079060405492783,
      "num_tokens": 3388888.0,
      "step": 4250
    },
    {
      "entropy": 4.967624044418335,
      "epoch": 1.2029968900197907,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004901150349611187,
      "loss": 4.944,
      "mean_token_accuracy": 0.2368533194065094,
      "num_tokens": 3392719.0,
      "step": 4255
    },
    {
      "entropy": 5.187564516067505,
      "epoch": 1.2044105173876165,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004900848590066419,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21811688393354417,
      "num_tokens": 3396884.0,
      "step": 4260
    },
    {
      "entropy": 5.100981140136719,
      "epoch": 1.2058241447554425,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004900546381007595,
      "loss": 4.944,
      "mean_token_accuracy": 0.23413423150777818,
      "num_tokens": 3400764.0,
      "step": 4265
    },
    {
      "entropy": 5.0627936840057375,
      "epoch": 1.2072377721232683,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004900243722497877,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21828858852386473,
      "num_tokens": 3405120.0,
      "step": 4270
    },
    {
      "entropy": 5.05987663269043,
      "epoch": 1.2086513994910941,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004899940614600517,
      "loss": 5.0522,
      "mean_token_accuracy": 0.23300132006406785,
      "num_tokens": 3409200.0,
      "step": 4275
    },
    {
      "entropy": 5.154522848129273,
      "epoch": 1.21006502685892,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004899637057378864,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22696007341146468,
      "num_tokens": 3413360.0,
      "step": 4280
    },
    {
      "entropy": 5.14468297958374,
      "epoch": 1.2114786542267457,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004899333050896358,
      "loss": 5.1198,
      "mean_token_accuracy": 0.21608099490404128,
      "num_tokens": 3417445.0,
      "step": 4285
    },
    {
      "entropy": 5.105215263366699,
      "epoch": 1.2128922815945717,
      "grad_norm": 1.375,
      "learning_rate": 0.0004899028595216536,
      "loss": 5.001,
      "mean_token_accuracy": 0.22085639238357543,
      "num_tokens": 3421289.0,
      "step": 4290
    },
    {
      "entropy": 5.021739721298218,
      "epoch": 1.2143059089623975,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004898723690403027,
      "loss": 4.9349,
      "mean_token_accuracy": 0.23851098269224166,
      "num_tokens": 3425211.0,
      "step": 4295
    },
    {
      "entropy": 5.0703798770904545,
      "epoch": 1.2157195363302233,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004898418336519553,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2258564054965973,
      "num_tokens": 3429101.0,
      "step": 4300
    },
    {
      "entropy": 5.122488689422608,
      "epoch": 1.2171331636980491,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004898112533629933,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22742394208908082,
      "num_tokens": 3433334.0,
      "step": 4305
    },
    {
      "entropy": 5.046411180496216,
      "epoch": 1.218546791065875,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004897806281798076,
      "loss": 5.0657,
      "mean_token_accuracy": 0.23195153027772902,
      "num_tokens": 3437044.0,
      "step": 4310
    },
    {
      "entropy": 5.115883493423462,
      "epoch": 1.219960418433701,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004897499581087989,
      "loss": 5.0659,
      "mean_token_accuracy": 0.2173108845949173,
      "num_tokens": 3440979.0,
      "step": 4315
    },
    {
      "entropy": 5.157653951644898,
      "epoch": 1.2213740458015268,
      "grad_norm": 1.359375,
      "learning_rate": 0.000489719243156377,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21958375573158265,
      "num_tokens": 3445285.0,
      "step": 4320
    },
    {
      "entropy": 5.157428121566772,
      "epoch": 1.2227876731693526,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004896884833289609,
      "loss": 5.0592,
      "mean_token_accuracy": 0.22933107763528823,
      "num_tokens": 3449460.0,
      "step": 4325
    },
    {
      "entropy": 5.130751514434815,
      "epoch": 1.2242013005371783,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004896576786329794,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2368018940091133,
      "num_tokens": 3453193.0,
      "step": 4330
    },
    {
      "entropy": 5.104935503005981,
      "epoch": 1.2256149279050041,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004896268290748704,
      "loss": 5.1057,
      "mean_token_accuracy": 0.2277104601264,
      "num_tokens": 3457099.0,
      "step": 4335
    },
    {
      "entropy": 5.069167995452881,
      "epoch": 1.2270285552728302,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004895959346610815,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21637845337390899,
      "num_tokens": 3461370.0,
      "step": 4340
    },
    {
      "entropy": 5.222411251068115,
      "epoch": 1.228442182640656,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004895649953980691,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2158205047249794,
      "num_tokens": 3465663.0,
      "step": 4345
    },
    {
      "entropy": 5.106366920471191,
      "epoch": 1.2298558100084818,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004895340112922997,
      "loss": 5.0385,
      "mean_token_accuracy": 0.22892773449420928,
      "num_tokens": 3469304.0,
      "step": 4350
    },
    {
      "entropy": 4.950320529937744,
      "epoch": 1.2312694373763076,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004895029823502485,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2480193257331848,
      "num_tokens": 3473199.0,
      "step": 4355
    },
    {
      "entropy": 5.019837427139282,
      "epoch": 1.2326830647441334,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004894719085784005,
      "loss": 4.8772,
      "mean_token_accuracy": 0.2495657578110695,
      "num_tokens": 3476989.0,
      "step": 4360
    },
    {
      "entropy": 5.120286560058593,
      "epoch": 1.2340966921119594,
      "grad_norm": 1.375,
      "learning_rate": 0.0004894407899832497,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2303723081946373,
      "num_tokens": 3481123.0,
      "step": 4365
    },
    {
      "entropy": 4.991329479217529,
      "epoch": 1.2355103194797852,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004894096265713,
      "loss": 4.9015,
      "mean_token_accuracy": 0.2328546240925789,
      "num_tokens": 3484963.0,
      "step": 4370
    },
    {
      "entropy": 4.9451151371002195,
      "epoch": 1.236923946847611,
      "grad_norm": 1.3125,
      "learning_rate": 0.000489378418349064,
      "loss": 4.898,
      "mean_token_accuracy": 0.24232887625694274,
      "num_tokens": 3488754.0,
      "step": 4375
    },
    {
      "entropy": 5.005414867401123,
      "epoch": 1.2383375742154368,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004893471653230644,
      "loss": 4.9015,
      "mean_token_accuracy": 0.23712760210037231,
      "num_tokens": 3493147.0,
      "step": 4380
    },
    {
      "entropy": 5.145499801635742,
      "epoch": 1.2397512015832626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004893158674998328,
      "loss": 4.9837,
      "mean_token_accuracy": 0.23047097772359848,
      "num_tokens": 3497217.0,
      "step": 4385
    },
    {
      "entropy": 5.027615642547607,
      "epoch": 1.2411648289510886,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004892845248859101,
      "loss": 5.0142,
      "mean_token_accuracy": 0.23378553688526155,
      "num_tokens": 3500948.0,
      "step": 4390
    },
    {
      "entropy": 5.092244863510132,
      "epoch": 1.2425784563189144,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004892531374878468,
      "loss": 5.0744,
      "mean_token_accuracy": 0.22575632929801942,
      "num_tokens": 3504997.0,
      "step": 4395
    },
    {
      "entropy": 5.019850635528565,
      "epoch": 1.2439920836867402,
      "grad_norm": 1.25,
      "learning_rate": 0.0004892217053122026,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2305405169725418,
      "num_tokens": 3509198.0,
      "step": 4400
    },
    {
      "entropy": 5.111858129501343,
      "epoch": 1.245405711054566,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004891902283655468,
      "loss": 5.126,
      "mean_token_accuracy": 0.21807885766029358,
      "num_tokens": 3513002.0,
      "step": 4405
    },
    {
      "entropy": 5.11797685623169,
      "epoch": 1.2468193384223918,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004891587066544577,
      "loss": 4.9638,
      "mean_token_accuracy": 0.229783396422863,
      "num_tokens": 3516941.0,
      "step": 4410
    },
    {
      "entropy": 5.121893310546875,
      "epoch": 1.2482329657902178,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004891271401855232,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22427659928798677,
      "num_tokens": 3520788.0,
      "step": 4415
    },
    {
      "entropy": 5.116243505477906,
      "epoch": 1.2496465931580436,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004890955289653407,
      "loss": 5.1321,
      "mean_token_accuracy": 0.22196538299322127,
      "num_tokens": 3524690.0,
      "step": 4420
    },
    {
      "entropy": 5.0702392578125,
      "epoch": 1.2510602205258694,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004890638730005164,
      "loss": 4.9965,
      "mean_token_accuracy": 0.23890410959720612,
      "num_tokens": 3528826.0,
      "step": 4425
    },
    {
      "entropy": 5.122500228881836,
      "epoch": 1.2524738478936952,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004890321722976665,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22608898133039473,
      "num_tokens": 3532660.0,
      "step": 4430
    },
    {
      "entropy": 5.090112972259521,
      "epoch": 1.253887475261521,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000489000426863416,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22756127119064332,
      "num_tokens": 3536635.0,
      "step": 4435
    },
    {
      "entropy": 5.055538082122803,
      "epoch": 1.255301102629347,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004889686367043997,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2354679986834526,
      "num_tokens": 3540468.0,
      "step": 4440
    },
    {
      "entropy": 5.028900384902954,
      "epoch": 1.2567147299971728,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004889368018272614,
      "loss": 4.9491,
      "mean_token_accuracy": 0.23732344508171083,
      "num_tokens": 3544748.0,
      "step": 4445
    },
    {
      "entropy": 5.063275241851807,
      "epoch": 1.2581283573649986,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004889049222386547,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2354758635163307,
      "num_tokens": 3548766.0,
      "step": 4450
    },
    {
      "entropy": 5.177508735656739,
      "epoch": 1.2595419847328244,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004888729979452417,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2300665333867073,
      "num_tokens": 3553137.0,
      "step": 4455
    },
    {
      "entropy": 5.121517610549927,
      "epoch": 1.2609556121006502,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004888410289536949,
      "loss": 5.0627,
      "mean_token_accuracy": 0.22703436613082886,
      "num_tokens": 3557029.0,
      "step": 4460
    },
    {
      "entropy": 5.019227027893066,
      "epoch": 1.2623692394684762,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004888090152706953,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2401907354593277,
      "num_tokens": 3560919.0,
      "step": 4465
    },
    {
      "entropy": 5.031711006164551,
      "epoch": 1.263782866836302,
      "grad_norm": 1.25,
      "learning_rate": 0.0004887769569029336,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2265866681933403,
      "num_tokens": 3565096.0,
      "step": 4470
    },
    {
      "entropy": 5.0650722026824955,
      "epoch": 1.2651964942041278,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048874485385711,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23378395438194274,
      "num_tokens": 3568780.0,
      "step": 4475
    },
    {
      "entropy": 5.034435892105103,
      "epoch": 1.2666101215719536,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004887127061399336,
      "loss": 4.9709,
      "mean_token_accuracy": 0.23217694610357284,
      "num_tokens": 3572743.0,
      "step": 4480
    },
    {
      "entropy": 5.000231456756592,
      "epoch": 1.2680237489397794,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004886805137581232,
      "loss": 5.0388,
      "mean_token_accuracy": 0.2315497562289238,
      "num_tokens": 3576721.0,
      "step": 4485
    },
    {
      "entropy": 4.953033542633056,
      "epoch": 1.2694373763076054,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004886482767184068,
      "loss": 4.7938,
      "mean_token_accuracy": 0.2466677188873291,
      "num_tokens": 3580388.0,
      "step": 4490
    },
    {
      "entropy": 5.112115430831909,
      "epoch": 1.270851003675431,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004886159950275217,
      "loss": 5.0333,
      "mean_token_accuracy": 0.23307109028100967,
      "num_tokens": 3584249.0,
      "step": 4495
    },
    {
      "entropy": 5.100350332260132,
      "epoch": 1.272264631043257,
      "grad_norm": 1.25,
      "learning_rate": 0.0004885836686922147,
      "loss": 4.997,
      "mean_token_accuracy": 0.2359594464302063,
      "num_tokens": 3588539.0,
      "step": 4500
    },
    {
      "entropy": 5.052192735671997,
      "epoch": 1.2736782584110828,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004885512977192416,
      "loss": 5.0337,
      "mean_token_accuracy": 0.22936938852071762,
      "num_tokens": 3592756.0,
      "step": 4505
    },
    {
      "entropy": 5.121135139465332,
      "epoch": 1.2750918857789086,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004885188821153678,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2374897688627243,
      "num_tokens": 3596569.0,
      "step": 4510
    },
    {
      "entropy": 5.130150508880615,
      "epoch": 1.2765055131467344,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000488486421887368,
      "loss": 5.0775,
      "mean_token_accuracy": 0.22795253992080688,
      "num_tokens": 3600541.0,
      "step": 4515
    },
    {
      "entropy": 5.102085018157959,
      "epoch": 1.2779191405145602,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004884539170420263,
      "loss": 5.0189,
      "mean_token_accuracy": 0.22976435720920563,
      "num_tokens": 3604567.0,
      "step": 4520
    },
    {
      "entropy": 5.113991165161133,
      "epoch": 1.2793327678823863,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000488421367586136,
      "loss": 5.05,
      "mean_token_accuracy": 0.221399287879467,
      "num_tokens": 3608528.0,
      "step": 4525
    },
    {
      "entropy": 5.131040000915528,
      "epoch": 1.280746395250212,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004883887735264993,
      "loss": 5.0796,
      "mean_token_accuracy": 0.22173923701047898,
      "num_tokens": 3612408.0,
      "step": 4530
    },
    {
      "entropy": 5.077217292785645,
      "epoch": 1.2821600226180379,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004883561348699287,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23994230329990388,
      "num_tokens": 3616176.0,
      "step": 4535
    },
    {
      "entropy": 5.05971393585205,
      "epoch": 1.2835736499858637,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883234516232452,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2328529253602028,
      "num_tokens": 3620446.0,
      "step": 4540
    },
    {
      "entropy": 5.1482625007629395,
      "epoch": 1.2849872773536894,
      "grad_norm": 1.25,
      "learning_rate": 0.0004882907237932796,
      "loss": 4.9432,
      "mean_token_accuracy": 0.23651572465896606,
      "num_tokens": 3624437.0,
      "step": 4545
    },
    {
      "entropy": 5.127484130859375,
      "epoch": 1.2864009047215155,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004882579513868716,
      "loss": 5.0275,
      "mean_token_accuracy": 0.22705754339694978,
      "num_tokens": 3628418.0,
      "step": 4550
    },
    {
      "entropy": 5.05588059425354,
      "epoch": 1.2878145320893413,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048822513441087046,
      "loss": 4.9243,
      "mean_token_accuracy": 0.23365500718355178,
      "num_tokens": 3632237.0,
      "step": 4555
    },
    {
      "entropy": 5.042345762252808,
      "epoch": 1.289228159457167,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004881922728721348,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22918311059474944,
      "num_tokens": 3636275.0,
      "step": 4560
    },
    {
      "entropy": 5.025683307647705,
      "epoch": 1.2906417868249929,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048815936677753244,
      "loss": 5.0404,
      "mean_token_accuracy": 0.22625458985567093,
      "num_tokens": 3640319.0,
      "step": 4565
    },
    {
      "entropy": 5.2253625869750975,
      "epoch": 1.2920554141928187,
      "grad_norm": 1.375,
      "learning_rate": 0.00048812641613394053,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21494955122470855,
      "num_tokens": 3644200.0,
      "step": 4570
    },
    {
      "entropy": 5.032268857955932,
      "epoch": 1.2934690415606447,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004880934209482457,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23899020105600358,
      "num_tokens": 3648314.0,
      "step": 4575
    },
    {
      "entropy": 5.081238889694214,
      "epoch": 1.2948826689284705,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048806038122734354,
      "loss": 5.0655,
      "mean_token_accuracy": 0.22133136093616484,
      "num_tokens": 3652241.0,
      "step": 4580
    },
    {
      "entropy": 5.142512273788452,
      "epoch": 1.2962962962962963,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048802729697813916,
      "loss": 5.0872,
      "mean_token_accuracy": 0.21524255722761154,
      "num_tokens": 3656162.0,
      "step": 4585
    },
    {
      "entropy": 5.060039281845093,
      "epoch": 1.297709923664122,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004879941682075472,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2330801784992218,
      "num_tokens": 3660168.0,
      "step": 4590
    },
    {
      "entropy": 5.1759899139404295,
      "epoch": 1.2991235510319479,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004879609949224911,
      "loss": 5.0811,
      "mean_token_accuracy": 0.22278601676225662,
      "num_tokens": 3664299.0,
      "step": 4595
    },
    {
      "entropy": 5.122478771209717,
      "epoch": 1.300537178399774,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048792777712990397,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21671532243490219,
      "num_tokens": 3668054.0,
      "step": 4600
    },
    {
      "entropy": 5.035900974273682,
      "epoch": 1.3019508057675997,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004878945148367281,
      "loss": 4.9721,
      "mean_token_accuracy": 0.2366459771990776,
      "num_tokens": 3672161.0,
      "step": 4605
    },
    {
      "entropy": 5.165577840805054,
      "epoch": 1.3033644331354255,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048786120804991526,
      "loss": 5.1406,
      "mean_token_accuracy": 0.21782566010951995,
      "num_tokens": 3676040.0,
      "step": 4610
    },
    {
      "entropy": 5.183472537994385,
      "epoch": 1.3047780605032513,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048782785677642604,
      "loss": 5.0607,
      "mean_token_accuracy": 0.22802962064743043,
      "num_tokens": 3680251.0,
      "step": 4615
    },
    {
      "entropy": 5.074163293838501,
      "epoch": 1.306191687871077,
      "grad_norm": 1.34375,
      "learning_rate": 0.000487794461023231,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22687329202890397,
      "num_tokens": 3684194.0,
      "step": 4620
    },
    {
      "entropy": 5.073023509979248,
      "epoch": 1.307605315238903,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048776102079730937,
      "loss": 4.9775,
      "mean_token_accuracy": 0.2265431821346283,
      "num_tokens": 3688224.0,
      "step": 4625
    },
    {
      "entropy": 5.094221878051758,
      "epoch": 1.309018942606729,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048772753610565014,
      "loss": 4.9952,
      "mean_token_accuracy": 0.2399193525314331,
      "num_tokens": 3691956.0,
      "step": 4630
    },
    {
      "entropy": 5.05341854095459,
      "epoch": 1.3104325699745547,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004876940069552513,
      "loss": 4.9336,
      "mean_token_accuracy": 0.24319173246622086,
      "num_tokens": 3696183.0,
      "step": 4635
    },
    {
      "entropy": 5.060268688201904,
      "epoch": 1.3118461973423805,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048766043335312026,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2282293915748596,
      "num_tokens": 3699886.0,
      "step": 4640
    },
    {
      "entropy": 5.050581312179565,
      "epoch": 1.3132598247102063,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004876268153062738,
      "loss": 5.0487,
      "mean_token_accuracy": 0.22755955159664154,
      "num_tokens": 3703885.0,
      "step": 4645
    },
    {
      "entropy": 5.065429973602295,
      "epoch": 1.3146734520780323,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048759315282173766,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22579556554555893,
      "num_tokens": 3707493.0,
      "step": 4650
    },
    {
      "entropy": 4.995752620697021,
      "epoch": 1.3160870794458581,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004875594459065473,
      "loss": 5.0298,
      "mean_token_accuracy": 0.22605978399515153,
      "num_tokens": 3711466.0,
      "step": 4655
    },
    {
      "entropy": 5.044043684005738,
      "epoch": 1.317500706813684,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048752569456774717,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22801981270313262,
      "num_tokens": 3715270.0,
      "step": 4660
    },
    {
      "entropy": 5.157716369628906,
      "epoch": 1.3189143341815097,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048749189881239105,
      "loss": 5.0323,
      "mean_token_accuracy": 0.23153322488069533,
      "num_tokens": 3719202.0,
      "step": 4665
    },
    {
      "entropy": 4.995691251754761,
      "epoch": 1.3203279615493355,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048745805864754214,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2523768678307533,
      "num_tokens": 3723197.0,
      "step": 4670
    },
    {
      "entropy": 4.996036863327026,
      "epoch": 1.3217415889171615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004874241740802727,
      "loss": 4.9635,
      "mean_token_accuracy": 0.23059980273246766,
      "num_tokens": 3727528.0,
      "step": 4675
    },
    {
      "entropy": 5.040787220001221,
      "epoch": 1.3231552162849873,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004873902451176645,
      "loss": 4.9144,
      "mean_token_accuracy": 0.23785099685192107,
      "num_tokens": 3731410.0,
      "step": 4680
    },
    {
      "entropy": 5.1270466327667235,
      "epoch": 1.3245688436528131,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004873562717668084,
      "loss": 5.1001,
      "mean_token_accuracy": 0.23110138326883317,
      "num_tokens": 3735694.0,
      "step": 4685
    },
    {
      "entropy": 5.04110312461853,
      "epoch": 1.325982471020639,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004873222540348048,
      "loss": 5.0655,
      "mean_token_accuracy": 0.2262193515896797,
      "num_tokens": 3739836.0,
      "step": 4690
    },
    {
      "entropy": 4.978133249282837,
      "epoch": 1.3273960983884647,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004872881919287629,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22875244170427322,
      "num_tokens": 3743748.0,
      "step": 4695
    },
    {
      "entropy": 5.105270481109619,
      "epoch": 1.3288097257562907,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004872540854558018,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2333902969956398,
      "num_tokens": 3748034.0,
      "step": 4700
    },
    {
      "entropy": 5.077601099014283,
      "epoch": 1.3302233531241165,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004872199346230493,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23692454099655152,
      "num_tokens": 3751887.0,
      "step": 4705
    },
    {
      "entropy": 5.120017433166504,
      "epoch": 1.3316369804919423,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048718573943764285,
      "loss": 5.1078,
      "mean_token_accuracy": 0.21115975826978683,
      "num_tokens": 3756305.0,
      "step": 4710
    },
    {
      "entropy": 5.153009366989136,
      "epoch": 1.3330506078597681,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048715149990672887,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21767646223306655,
      "num_tokens": 3760139.0,
      "step": 4715
    },
    {
      "entropy": 5.153610754013061,
      "epoch": 1.334464235227594,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004871172160374634,
      "loss": 4.9791,
      "mean_token_accuracy": 0.23960563093423842,
      "num_tokens": 3763966.0,
      "step": 4720
    },
    {
      "entropy": 5.000993919372559,
      "epoch": 1.33587786259542,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048708288783701144,
      "loss": 4.9639,
      "mean_token_accuracy": 0.23444944471120835,
      "num_tokens": 3767902.0,
      "step": 4725
    },
    {
      "entropy": 5.004330968856811,
      "epoch": 1.3372914899632458,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004870485153125474,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2367618203163147,
      "num_tokens": 3771760.0,
      "step": 4730
    },
    {
      "entropy": 5.107804203033448,
      "epoch": 1.3387051173310716,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048701409847125494,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22552258521318436,
      "num_tokens": 3775543.0,
      "step": 4735
    },
    {
      "entropy": 5.0472547054290775,
      "epoch": 1.3401187446988974,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048697963732032695,
      "loss": 4.9235,
      "mean_token_accuracy": 0.23070451021194457,
      "num_tokens": 3779950.0,
      "step": 4740
    },
    {
      "entropy": 5.077399873733521,
      "epoch": 1.3415323720667232,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048694513186696566,
      "loss": 4.954,
      "mean_token_accuracy": 0.23352414667606353,
      "num_tokens": 3784012.0,
      "step": 4745
    },
    {
      "entropy": 5.124929428100586,
      "epoch": 1.3429459994345492,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004869105821183824,
      "loss": 5.0638,
      "mean_token_accuracy": 0.2216840460896492,
      "num_tokens": 3787973.0,
      "step": 4750
    },
    {
      "entropy": 5.042369604110718,
      "epoch": 1.344359626802375,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048687598808179786,
      "loss": 4.9691,
      "mean_token_accuracy": 0.23382918089628218,
      "num_tokens": 3791576.0,
      "step": 4755
    },
    {
      "entropy": 5.161550283432007,
      "epoch": 1.3457732541702008,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048684134976444203,
      "loss": 5.1033,
      "mean_token_accuracy": 0.21670863926410674,
      "num_tokens": 3795495.0,
      "step": 4760
    },
    {
      "entropy": 5.059696626663208,
      "epoch": 1.3471868815380266,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004868066671735541,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2298843666911125,
      "num_tokens": 3799477.0,
      "step": 4765
    },
    {
      "entropy": 5.0809825420379635,
      "epoch": 1.3486005089058524,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004867719403163825,
      "loss": 5.0696,
      "mean_token_accuracy": 0.22125642895698547,
      "num_tokens": 3803305.0,
      "step": 4770
    },
    {
      "entropy": 5.012603425979615,
      "epoch": 1.3500141362736784,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004867371692001849,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2339460954070091,
      "num_tokens": 3807374.0,
      "step": 4775
    },
    {
      "entropy": 4.925884628295899,
      "epoch": 1.351427763641504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004867023538322283,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22832754105329514,
      "num_tokens": 3811103.0,
      "step": 4780
    },
    {
      "entropy": 4.984415674209595,
      "epoch": 1.35284139100933,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004866674942197889,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23991322070360183,
      "num_tokens": 3814947.0,
      "step": 4785
    },
    {
      "entropy": 5.113937139511108,
      "epoch": 1.3542550183771558,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048663259037015204,
      "loss": 5.045,
      "mean_token_accuracy": 0.2342761844396591,
      "num_tokens": 3818945.0,
      "step": 4790
    },
    {
      "entropy": 4.998484706878662,
      "epoch": 1.3556686457449816,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048659764229061264,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2485048919916153,
      "num_tokens": 3822852.0,
      "step": 4795
    },
    {
      "entropy": 5.0883890151977536,
      "epoch": 1.3570822731128076,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048656264998847435,
      "loss": 5.0139,
      "mean_token_accuracy": 0.2224154829978943,
      "num_tokens": 3826663.0,
      "step": 4800
    },
    {
      "entropy": 5.021430587768554,
      "epoch": 1.3584959004806332,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004865276134710505,
      "loss": 4.9254,
      "mean_token_accuracy": 0.24365831464529036,
      "num_tokens": 3830585.0,
      "step": 4805
    },
    {
      "entropy": 5.097690582275391,
      "epoch": 1.3599095278484592,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004864925327456635,
      "loss": 4.9085,
      "mean_token_accuracy": 0.23187643885612488,
      "num_tokens": 3834432.0,
      "step": 4810
    },
    {
      "entropy": 5.0584817886352536,
      "epoch": 1.361323155216285,
      "grad_norm": 1.25,
      "learning_rate": 0.00048645740781964506,
      "loss": 4.9726,
      "mean_token_accuracy": 0.23237268924713134,
      "num_tokens": 3838416.0,
      "step": 4815
    },
    {
      "entropy": 5.126487588882446,
      "epoch": 1.3627367825841108,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048642223870033584,
      "loss": 5.1137,
      "mean_token_accuracy": 0.21244313865900039,
      "num_tokens": 3842360.0,
      "step": 4820
    },
    {
      "entropy": 5.141035890579223,
      "epoch": 1.3641504099519366,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004863870253950863,
      "loss": 5.0347,
      "mean_token_accuracy": 0.2249879240989685,
      "num_tokens": 3846496.0,
      "step": 4825
    },
    {
      "entropy": 5.10675573348999,
      "epoch": 1.3655640373197624,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004863517679112555,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2388698637485504,
      "num_tokens": 3850550.0,
      "step": 4830
    },
    {
      "entropy": 5.038706588745117,
      "epoch": 1.3669776646875884,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004863164662562122,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2208436459302902,
      "num_tokens": 3854751.0,
      "step": 4835
    },
    {
      "entropy": 5.050574493408203,
      "epoch": 1.3683912920554142,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004862811204373342,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2306164175271988,
      "num_tokens": 3858554.0,
      "step": 4840
    },
    {
      "entropy": 4.96640191078186,
      "epoch": 1.36980491942324,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004862457304620086,
      "loss": 4.8771,
      "mean_token_accuracy": 0.24121769964694978,
      "num_tokens": 3862312.0,
      "step": 4845
    },
    {
      "entropy": 5.054334878921509,
      "epoch": 1.3712185467910658,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048621029633763157,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2349928632378578,
      "num_tokens": 3866595.0,
      "step": 4850
    },
    {
      "entropy": 5.090772724151611,
      "epoch": 1.3726321741588916,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048617481807160874,
      "loss": 4.9823,
      "mean_token_accuracy": 0.24338055849075318,
      "num_tokens": 3870863.0,
      "step": 4855
    },
    {
      "entropy": 5.171737003326416,
      "epoch": 1.3740458015267176,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004861392956713547,
      "loss": 5.0979,
      "mean_token_accuracy": 0.2231503114104271,
      "num_tokens": 3874797.0,
      "step": 4860
    },
    {
      "entropy": 5.051101207733154,
      "epoch": 1.3754594288945434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048610372914429365,
      "loss": 4.9484,
      "mean_token_accuracy": 0.23833859711885452,
      "num_tokens": 3878965.0,
      "step": 4865
    },
    {
      "entropy": 4.987088918685913,
      "epoch": 1.3768730562623692,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004860681184978586,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2359229579567909,
      "num_tokens": 3883299.0,
      "step": 4870
    },
    {
      "entropy": 5.109549474716187,
      "epoch": 1.378286683630195,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048603246373949197,
      "loss": 5.0988,
      "mean_token_accuracy": 0.2201838105916977,
      "num_tokens": 3887616.0,
      "step": 4875
    },
    {
      "entropy": 5.189830923080445,
      "epoch": 1.3797003109980208,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048599676487664535,
      "loss": 5.0595,
      "mean_token_accuracy": 0.2242570251226425,
      "num_tokens": 3892043.0,
      "step": 4880
    },
    {
      "entropy": 5.026374673843383,
      "epoch": 1.3811139383658468,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004859610219167797,
      "loss": 4.9644,
      "mean_token_accuracy": 0.24089155346155167,
      "num_tokens": 3896385.0,
      "step": 4885
    },
    {
      "entropy": 5.068024301528931,
      "epoch": 1.3825275657336726,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000485925234867365,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22480327636003494,
      "num_tokens": 3900277.0,
      "step": 4890
    },
    {
      "entropy": 4.970526313781738,
      "epoch": 1.3839411931014984,
      "grad_norm": 1.375,
      "learning_rate": 0.00048588940373588054,
      "loss": 4.865,
      "mean_token_accuracy": 0.2530467495322227,
      "num_tokens": 3903653.0,
      "step": 4895
    },
    {
      "entropy": 5.081054592132569,
      "epoch": 1.3853548204693242,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004858535285298148,
      "loss": 4.9706,
      "mean_token_accuracy": 0.23895137906074523,
      "num_tokens": 3908003.0,
      "step": 4900
    },
    {
      "entropy": 4.975905513763427,
      "epoch": 1.38676844783715,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048581760925666545,
      "loss": 4.9538,
      "mean_token_accuracy": 0.23338055461645127,
      "num_tokens": 3912120.0,
      "step": 4905
    },
    {
      "entropy": 5.164676237106323,
      "epoch": 1.388182075204976,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004857816459239394,
      "loss": 5.0957,
      "mean_token_accuracy": 0.2190960317850113,
      "num_tokens": 3916078.0,
      "step": 4910
    },
    {
      "entropy": 5.208339643478394,
      "epoch": 1.3895957025728018,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004857456385391528,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2106938049197197,
      "num_tokens": 3920214.0,
      "step": 4915
    },
    {
      "entropy": 5.088286447525024,
      "epoch": 1.3910093299406276,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000485709587109831,
      "loss": 4.9304,
      "mean_token_accuracy": 0.23054065704345703,
      "num_tokens": 3924092.0,
      "step": 4920
    },
    {
      "entropy": 5.016297054290772,
      "epoch": 1.3924229573084534,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004856734916435085,
      "loss": 5.0222,
      "mean_token_accuracy": 0.23953258991241455,
      "num_tokens": 3927999.0,
      "step": 4925
    },
    {
      "entropy": 5.071065759658813,
      "epoch": 1.3938365846762792,
      "grad_norm": 1.375,
      "learning_rate": 0.00048563735214772897,
      "loss": 4.9792,
      "mean_token_accuracy": 0.23855535984039306,
      "num_tokens": 3931804.0,
      "step": 4930
    },
    {
      "entropy": 5.145835876464844,
      "epoch": 1.3952502120441053,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048560116863004537,
      "loss": 5.03,
      "mean_token_accuracy": 0.23220262229442595,
      "num_tokens": 3935569.0,
      "step": 4935
    },
    {
      "entropy": 5.037695360183716,
      "epoch": 1.396663839411931,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048556494109801984,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22855437248945237,
      "num_tokens": 3939450.0,
      "step": 4940
    },
    {
      "entropy": 5.011439847946167,
      "epoch": 1.3980774667797569,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004855286695592238,
      "loss": 4.9746,
      "mean_token_accuracy": 0.22636061161756516,
      "num_tokens": 3943718.0,
      "step": 4945
    },
    {
      "entropy": 5.070659208297729,
      "epoch": 1.3994910941475827,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004854923540212376,
      "loss": 5.005,
      "mean_token_accuracy": 0.2399015799164772,
      "num_tokens": 3947558.0,
      "step": 4950
    },
    {
      "entropy": 5.073257064819336,
      "epoch": 1.4009047215154085,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048545599449165106,
      "loss": 4.9145,
      "mean_token_accuracy": 0.24072454124689102,
      "num_tokens": 3951642.0,
      "step": 4955
    },
    {
      "entropy": 5.073284769058228,
      "epoch": 1.4023183488832345,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048541959097806315,
      "loss": 5.0496,
      "mean_token_accuracy": 0.23663049191236496,
      "num_tokens": 3956038.0,
      "step": 4960
    },
    {
      "entropy": 5.140478610992432,
      "epoch": 1.4037319762510603,
      "grad_norm": 1.5,
      "learning_rate": 0.00048538314348808186,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2274949625134468,
      "num_tokens": 3959855.0,
      "step": 4965
    },
    {
      "entropy": 5.094705247879029,
      "epoch": 1.405145603618886,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004853466520293245,
      "loss": 4.9888,
      "mean_token_accuracy": 0.2246784672141075,
      "num_tokens": 3963819.0,
      "step": 4970
    },
    {
      "entropy": 5.028082466125488,
      "epoch": 1.4065592309867119,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048531011660941784,
      "loss": 4.9038,
      "mean_token_accuracy": 0.23658126443624497,
      "num_tokens": 3967526.0,
      "step": 4975
    },
    {
      "entropy": 4.984070491790772,
      "epoch": 1.4079728583545377,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048527353723599716,
      "loss": 4.9865,
      "mean_token_accuracy": 0.23184578418731688,
      "num_tokens": 3971432.0,
      "step": 4980
    },
    {
      "entropy": 4.939075946807861,
      "epoch": 1.4093864857223637,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004852369139167075,
      "loss": 4.8663,
      "mean_token_accuracy": 0.24537726640701293,
      "num_tokens": 3975225.0,
      "step": 4985
    },
    {
      "entropy": 5.138537979125976,
      "epoch": 1.4108001130901895,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048520024665920286,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22945340424776078,
      "num_tokens": 3979265.0,
      "step": 4990
    },
    {
      "entropy": 5.163413953781128,
      "epoch": 1.4122137404580153,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004851635354711466,
      "loss": 4.9895,
      "mean_token_accuracy": 0.23044971823692323,
      "num_tokens": 3983413.0,
      "step": 4995
    },
    {
      "entropy": 5.1170515537261965,
      "epoch": 1.413627367825841,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048512678036021097,
      "loss": 4.96,
      "mean_token_accuracy": 0.23421283513307573,
      "num_tokens": 3987304.0,
      "step": 5000
    },
    {
      "epoch": 1.413627367825841,
      "eval_entropy": 4.866368312011113,
      "eval_loss": 5.276529312133789,
      "eval_mean_token_accuracy": 0.2279621098932436,
      "eval_num_tokens": 3987304.0,
      "eval_runtime": 4.6492,
      "eval_samples_per_second": 2427.104,
      "eval_steps_per_second": 303.496,
      "step": 5000
    },
    {
      "entropy": 5.192615985870361,
      "epoch": 1.4150409951936669,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048508998133407766,
      "loss": 5.192,
      "mean_token_accuracy": 0.20705655217170715,
      "num_tokens": 3991705.0,
      "step": 5005
    },
    {
      "entropy": 5.001135349273682,
      "epoch": 1.416454622561493,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004850531384004374,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23787226974964143,
      "num_tokens": 3995489.0,
      "step": 5010
    },
    {
      "entropy": 5.016736650466919,
      "epoch": 1.4178682499293187,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048501625156699006,
      "loss": 5.031,
      "mean_token_accuracy": 0.23071810454130173,
      "num_tokens": 3999346.0,
      "step": 5015
    },
    {
      "entropy": 5.072630834579468,
      "epoch": 1.4192818772971445,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000484979320841445,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22651198208332063,
      "num_tokens": 4003374.0,
      "step": 5020
    },
    {
      "entropy": 5.06053318977356,
      "epoch": 1.4206955046649703,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048494234623152027,
      "loss": 5.0276,
      "mean_token_accuracy": 0.22763097435235977,
      "num_tokens": 4007527.0,
      "step": 5025
    },
    {
      "entropy": 4.953833341598511,
      "epoch": 1.422109132032796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048490532774494343,
      "loss": 4.7652,
      "mean_token_accuracy": 0.24919261187314987,
      "num_tokens": 4011447.0,
      "step": 5030
    },
    {
      "entropy": 4.9741466522216795,
      "epoch": 1.4235227594006221,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004848682653894511,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23011216521263123,
      "num_tokens": 4015352.0,
      "step": 5035
    },
    {
      "entropy": 4.982890510559082,
      "epoch": 1.424936386768448,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048483115917278906,
      "loss": 5.045,
      "mean_token_accuracy": 0.22259730249643325,
      "num_tokens": 4019468.0,
      "step": 5040
    },
    {
      "entropy": 5.025899887084961,
      "epoch": 1.4263500141362737,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048479400910271243,
      "loss": 4.9508,
      "mean_token_accuracy": 0.23430838882923127,
      "num_tokens": 4023451.0,
      "step": 5045
    },
    {
      "entropy": 5.005314588546753,
      "epoch": 1.4277636415040995,
      "grad_norm": 1.375,
      "learning_rate": 0.00048475681518698517,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2399151623249054,
      "num_tokens": 4027676.0,
      "step": 5050
    },
    {
      "entropy": 5.056168079376221,
      "epoch": 1.4291772688719253,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048471957743338056,
      "loss": 5.0622,
      "mean_token_accuracy": 0.22606683224439622,
      "num_tokens": 4031590.0,
      "step": 5055
    },
    {
      "entropy": 5.02661919593811,
      "epoch": 1.4305908962397513,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004846822958496812,
      "loss": 4.9202,
      "mean_token_accuracy": 0.24440328180789947,
      "num_tokens": 4035370.0,
      "step": 5060
    },
    {
      "entropy": 5.0589845180511475,
      "epoch": 1.432004523607577,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048464497044367865,
      "loss": 4.9831,
      "mean_token_accuracy": 0.23421025276184082,
      "num_tokens": 4039145.0,
      "step": 5065
    },
    {
      "entropy": 5.102833414077759,
      "epoch": 1.433418150975403,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048460760122317364,
      "loss": 5.0528,
      "mean_token_accuracy": 0.2277243360877037,
      "num_tokens": 4042993.0,
      "step": 5070
    },
    {
      "entropy": 5.105997800827026,
      "epoch": 1.4348317783432287,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004845701881959762,
      "loss": 5.0349,
      "mean_token_accuracy": 0.23160772025585175,
      "num_tokens": 4047313.0,
      "step": 5075
    },
    {
      "entropy": 5.038058280944824,
      "epoch": 1.4362454057110545,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004845327313699053,
      "loss": 5.0499,
      "mean_token_accuracy": 0.22046917974948882,
      "num_tokens": 4051554.0,
      "step": 5080
    },
    {
      "entropy": 5.068963670730591,
      "epoch": 1.4376590330788805,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004844952307527893,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2161948248744011,
      "num_tokens": 4055470.0,
      "step": 5085
    },
    {
      "entropy": 5.0373156547546385,
      "epoch": 1.4390726604467061,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048445768635246557,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2316660076379776,
      "num_tokens": 4059352.0,
      "step": 5090
    },
    {
      "entropy": 5.1284068584442135,
      "epoch": 1.4404862878145321,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004844200981767806,
      "loss": 5.0332,
      "mean_token_accuracy": 0.23470153659582138,
      "num_tokens": 4063523.0,
      "step": 5095
    },
    {
      "entropy": 5.1711712837219235,
      "epoch": 1.441899915182358,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048438246623359014,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22678114026784896,
      "num_tokens": 4067219.0,
      "step": 5100
    },
    {
      "entropy": 5.105133056640625,
      "epoch": 1.4433135425501837,
      "grad_norm": 1.34375,
      "learning_rate": 0.000484344790530759,
      "loss": 5.012,
      "mean_token_accuracy": 0.22977727949619292,
      "num_tokens": 4071351.0,
      "step": 5105
    },
    {
      "entropy": 5.035439157485962,
      "epoch": 1.4447271699180095,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004843070710761613,
      "loss": 4.9659,
      "mean_token_accuracy": 0.23493245393037795,
      "num_tokens": 4075113.0,
      "step": 5110
    },
    {
      "entropy": 5.09015040397644,
      "epoch": 1.4461407972858353,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004842693078776799,
      "loss": 5.0328,
      "mean_token_accuracy": 0.23061479032039642,
      "num_tokens": 4079146.0,
      "step": 5115
    },
    {
      "entropy": 5.005983018875122,
      "epoch": 1.4475544246536614,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048423150094320733,
      "loss": 4.9501,
      "mean_token_accuracy": 0.23514939248561859,
      "num_tokens": 4083195.0,
      "step": 5120
    },
    {
      "entropy": 4.981443119049072,
      "epoch": 1.4489680520214872,
      "grad_norm": 1.25,
      "learning_rate": 0.0004841936502806449,
      "loss": 4.8273,
      "mean_token_accuracy": 0.24832531809806824,
      "num_tokens": 4087251.0,
      "step": 5125
    },
    {
      "entropy": 4.9908167839050295,
      "epoch": 1.450381679389313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048415575589790314,
      "loss": 4.9997,
      "mean_token_accuracy": 0.23131684362888336,
      "num_tokens": 4091477.0,
      "step": 5130
    },
    {
      "entropy": 5.042552089691162,
      "epoch": 1.4517953067571387,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004841178178029018,
      "loss": 5.0075,
      "mean_token_accuracy": 0.23003988415002824,
      "num_tokens": 4095277.0,
      "step": 5135
    },
    {
      "entropy": 5.079316663742065,
      "epoch": 1.4532089341249645,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004840798360035697,
      "loss": 4.8858,
      "mean_token_accuracy": 0.24320492446422576,
      "num_tokens": 4099157.0,
      "step": 5140
    },
    {
      "entropy": 5.1023646831512455,
      "epoch": 1.4546225614927906,
      "grad_norm": 1.375,
      "learning_rate": 0.00048404181050784475,
      "loss": 4.9184,
      "mean_token_accuracy": 0.23112163990736007,
      "num_tokens": 4103183.0,
      "step": 5145
    },
    {
      "entropy": 5.039215612411499,
      "epoch": 1.4560361888606164,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048400374132367415,
      "loss": 4.9049,
      "mean_token_accuracy": 0.23072464913129806,
      "num_tokens": 4107265.0,
      "step": 5150
    },
    {
      "entropy": 4.988401222229004,
      "epoch": 1.4574498162284422,
      "grad_norm": 1.296875,
      "learning_rate": 0.000483965628459014,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2458241268992424,
      "num_tokens": 4110919.0,
      "step": 5155
    },
    {
      "entropy": 4.984357643127441,
      "epoch": 1.458863443596268,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004839274719218297,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23315515667200087,
      "num_tokens": 4114808.0,
      "step": 5160
    },
    {
      "entropy": 5.204904222488404,
      "epoch": 1.4602770709640938,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048388927172009574,
      "loss": 5.1873,
      "mean_token_accuracy": 0.21103968173265458,
      "num_tokens": 4119299.0,
      "step": 5165
    },
    {
      "entropy": 5.09746413230896,
      "epoch": 1.4616906983319198,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004838510278617957,
      "loss": 4.859,
      "mean_token_accuracy": 0.23111955672502518,
      "num_tokens": 4123276.0,
      "step": 5170
    },
    {
      "entropy": 5.028835439682007,
      "epoch": 1.4631043256997456,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004838127403549224,
      "loss": 4.9609,
      "mean_token_accuracy": 0.23025520890951157,
      "num_tokens": 4127286.0,
      "step": 5175
    },
    {
      "entropy": 5.070145320892334,
      "epoch": 1.4645179530675714,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048377440920747755,
      "loss": 4.9765,
      "mean_token_accuracy": 0.2328324109315872,
      "num_tokens": 4131478.0,
      "step": 5180
    },
    {
      "entropy": 5.052413988113403,
      "epoch": 1.4659315804353972,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004837360344274723,
      "loss": 4.8614,
      "mean_token_accuracy": 0.24189515560865402,
      "num_tokens": 4135594.0,
      "step": 5185
    },
    {
      "entropy": 4.9923395156860355,
      "epoch": 1.467345207803223,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004836976160229266,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22817785739898683,
      "num_tokens": 4139444.0,
      "step": 5190
    },
    {
      "entropy": 5.1101799488067625,
      "epoch": 1.468758835171049,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048365915400186963,
      "loss": 4.9996,
      "mean_token_accuracy": 0.2356755644083023,
      "num_tokens": 4143735.0,
      "step": 5195
    },
    {
      "entropy": 5.101583337783813,
      "epoch": 1.4701724625388748,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048362064837233985,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21966368108987808,
      "num_tokens": 4147737.0,
      "step": 5200
    },
    {
      "entropy": 5.036973476409912,
      "epoch": 1.4715860899067006,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048358209914238455,
      "loss": 4.8967,
      "mean_token_accuracy": 0.23562174439430236,
      "num_tokens": 4151707.0,
      "step": 5205
    },
    {
      "entropy": 5.097429943084717,
      "epoch": 1.4729997172745264,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004835435063200605,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2353842079639435,
      "num_tokens": 4155656.0,
      "step": 5210
    },
    {
      "entropy": 4.970907926559448,
      "epoch": 1.4744133446423522,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004835048699134331,
      "loss": 4.8862,
      "mean_token_accuracy": 0.245866858959198,
      "num_tokens": 4159496.0,
      "step": 5215
    },
    {
      "entropy": 5.033643913269043,
      "epoch": 1.4758269720101782,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004834661899305773,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2304802194237709,
      "num_tokens": 4163496.0,
      "step": 5220
    },
    {
      "entropy": 5.174898386001587,
      "epoch": 1.477240599378004,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004834274663795769,
      "loss": 5.107,
      "mean_token_accuracy": 0.2238018736243248,
      "num_tokens": 4167608.0,
      "step": 5225
    },
    {
      "entropy": 5.0143718242645265,
      "epoch": 1.4786542267458298,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048338869926852493,
      "loss": 4.87,
      "mean_token_accuracy": 0.24899086505174636,
      "num_tokens": 4171482.0,
      "step": 5230
    },
    {
      "entropy": 5.0302187442779545,
      "epoch": 1.4800678541136556,
      "grad_norm": 1.25,
      "learning_rate": 0.0004833498886055234,
      "loss": 5.0194,
      "mean_token_accuracy": 0.22697904855012893,
      "num_tokens": 4175461.0,
      "step": 5235
    },
    {
      "entropy": 5.11761155128479,
      "epoch": 1.4814814814814814,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048331103439868355,
      "loss": 5.0656,
      "mean_token_accuracy": 0.2130562037229538,
      "num_tokens": 4179378.0,
      "step": 5240
    },
    {
      "entropy": 5.090002012252808,
      "epoch": 1.4828951088493074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048327213665612574,
      "loss": 4.9617,
      "mean_token_accuracy": 0.23327282518148423,
      "num_tokens": 4183699.0,
      "step": 5245
    },
    {
      "entropy": 5.182643556594849,
      "epoch": 1.4843087362171332,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004832331953859793,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21803289502859116,
      "num_tokens": 4187637.0,
      "step": 5250
    },
    {
      "entropy": 4.980442810058594,
      "epoch": 1.485722363584959,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048319421059638256,
      "loss": 4.942,
      "mean_token_accuracy": 0.23772436380386353,
      "num_tokens": 4191956.0,
      "step": 5255
    },
    {
      "entropy": 5.002339649200439,
      "epoch": 1.4871359909527848,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004831551822954834,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22388166338205337,
      "num_tokens": 4196038.0,
      "step": 5260
    },
    {
      "entropy": 5.135624790191651,
      "epoch": 1.4885496183206106,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004831161104914382,
      "loss": 5.0291,
      "mean_token_accuracy": 0.21558183431625366,
      "num_tokens": 4200122.0,
      "step": 5265
    },
    {
      "entropy": 5.085350179672242,
      "epoch": 1.4899632456884366,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048307699519241294,
      "loss": 4.856,
      "mean_token_accuracy": 0.23699084520339966,
      "num_tokens": 4203923.0,
      "step": 5270
    },
    {
      "entropy": 5.064455604553222,
      "epoch": 1.4913768730562624,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004830378364065824,
      "loss": 4.9318,
      "mean_token_accuracy": 0.23319276720285415,
      "num_tokens": 4207877.0,
      "step": 5275
    },
    {
      "entropy": 4.96812481880188,
      "epoch": 1.4927905004240882,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048299863414213046,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2450646936893463,
      "num_tokens": 4211625.0,
      "step": 5280
    },
    {
      "entropy": 5.0622025489807125,
      "epoch": 1.494204127791914,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048295938840725025,
      "loss": 5.0843,
      "mean_token_accuracy": 0.22682557702064515,
      "num_tokens": 4215837.0,
      "step": 5285
    },
    {
      "entropy": 5.075719785690308,
      "epoch": 1.4956177551597398,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048292009921014386,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2293403208255768,
      "num_tokens": 4219899.0,
      "step": 5290
    },
    {
      "entropy": 4.922153282165527,
      "epoch": 1.4970313825275658,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004828807665590224,
      "loss": 4.7969,
      "mean_token_accuracy": 0.25032604187726976,
      "num_tokens": 4223609.0,
      "step": 5295
    },
    {
      "entropy": 4.972326946258545,
      "epoch": 1.4984450098953916,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048284139046210625,
      "loss": 4.9076,
      "mean_token_accuracy": 0.23496110290288924,
      "num_tokens": 4227552.0,
      "step": 5300
    },
    {
      "entropy": 5.160442161560058,
      "epoch": 1.4998586372632174,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048280197092762475,
      "loss": 5.049,
      "mean_token_accuracy": 0.22921695113182067,
      "num_tokens": 4231500.0,
      "step": 5305
    },
    {
      "entropy": 5.05962438583374,
      "epoch": 1.5012722646310432,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004827625079638164,
      "loss": 4.9752,
      "mean_token_accuracy": 0.23784275650978087,
      "num_tokens": 4235348.0,
      "step": 5310
    },
    {
      "entropy": 5.061267948150634,
      "epoch": 1.502685891998869,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048272300157892856,
      "loss": 5.0116,
      "mean_token_accuracy": 0.22396658211946488,
      "num_tokens": 4239255.0,
      "step": 5315
    },
    {
      "entropy": 5.0244927406311035,
      "epoch": 1.504099519366695,
      "grad_norm": 1.40625,
      "learning_rate": 0.000482683451781218,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2280827581882477,
      "num_tokens": 4243124.0,
      "step": 5320
    },
    {
      "entropy": 5.025907135009765,
      "epoch": 1.5055131467345206,
      "grad_norm": 1.375,
      "learning_rate": 0.0004826438585789503,
      "loss": 4.8828,
      "mean_token_accuracy": 0.23566618859767913,
      "num_tokens": 4247239.0,
      "step": 5325
    },
    {
      "entropy": 5.085031509399414,
      "epoch": 1.5069267741023467,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004826042219804,
      "loss": 5.0267,
      "mean_token_accuracy": 0.22469329386949538,
      "num_tokens": 4251376.0,
      "step": 5330
    },
    {
      "entropy": 5.049340295791626,
      "epoch": 1.5083404014701725,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004825645419938513,
      "loss": 4.8943,
      "mean_token_accuracy": 0.23595458567142485,
      "num_tokens": 4255559.0,
      "step": 5335
    },
    {
      "entropy": 4.97284574508667,
      "epoch": 1.5097540288379983,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004825248186275967,
      "loss": 4.977,
      "mean_token_accuracy": 0.23704527020454408,
      "num_tokens": 4259778.0,
      "step": 5340
    },
    {
      "entropy": 4.967990398406982,
      "epoch": 1.5111676562058243,
      "grad_norm": 1.375,
      "learning_rate": 0.0004824850518899384,
      "loss": 4.9203,
      "mean_token_accuracy": 0.23278773128986358,
      "num_tokens": 4263678.0,
      "step": 5345
    },
    {
      "entropy": 5.117977523803711,
      "epoch": 1.5125812835736498,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004824452417891874,
      "loss": 5.0084,
      "mean_token_accuracy": 0.22468505054712296,
      "num_tokens": 4267487.0,
      "step": 5350
    },
    {
      "entropy": 5.1667650699615475,
      "epoch": 1.5139949109414759,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048240538833366355,
      "loss": 5.0431,
      "mean_token_accuracy": 0.22304808348417282,
      "num_tokens": 4271592.0,
      "step": 5355
    },
    {
      "entropy": 5.073079252243042,
      "epoch": 1.5154085383093017,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048236549153169605,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22755898535251617,
      "num_tokens": 4275378.0,
      "step": 5360
    },
    {
      "entropy": 4.984932470321655,
      "epoch": 1.5168221656771275,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004823255513916232,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2302310660481453,
      "num_tokens": 4279441.0,
      "step": 5365
    },
    {
      "entropy": 5.0882096767425535,
      "epoch": 1.5182357930449535,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048228556792179216,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2314021348953247,
      "num_tokens": 4283320.0,
      "step": 5370
    },
    {
      "entropy": 5.1342579364776615,
      "epoch": 1.519649420412779,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822455411305592,
      "loss": 4.9659,
      "mean_token_accuracy": 0.23380256146192552,
      "num_tokens": 4287380.0,
      "step": 5375
    },
    {
      "entropy": 5.081647729873657,
      "epoch": 1.521063047780605,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004822054710262898,
      "loss": 5.0439,
      "mean_token_accuracy": 0.22692810893058776,
      "num_tokens": 4291541.0,
      "step": 5380
    },
    {
      "entropy": 5.102030849456787,
      "epoch": 1.5224766751484309,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004821653576173582,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2355323314666748,
      "num_tokens": 4295649.0,
      "step": 5385
    },
    {
      "entropy": 5.081121921539307,
      "epoch": 1.5238903025162567,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048212520091214794,
      "loss": 4.9706,
      "mean_token_accuracy": 0.23803273141384124,
      "num_tokens": 4299383.0,
      "step": 5390
    },
    {
      "entropy": 4.9772069454193115,
      "epoch": 1.5253039298840827,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004820850009190516,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2433154284954071,
      "num_tokens": 4303403.0,
      "step": 5395
    },
    {
      "entropy": 5.055086517333985,
      "epoch": 1.5267175572519083,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048204475764647057,
      "loss": 4.9889,
      "mean_token_accuracy": 0.2336754858493805,
      "num_tokens": 4307458.0,
      "step": 5400
    },
    {
      "entropy": 4.980713129043579,
      "epoch": 1.5281311846197343,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048200447110281546,
      "loss": 4.8952,
      "mean_token_accuracy": 0.24659692347049714,
      "num_tokens": 4311130.0,
      "step": 5405
    },
    {
      "entropy": 4.9933535099029545,
      "epoch": 1.52954481198756,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048196414129650605,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2398616001009941,
      "num_tokens": 4315271.0,
      "step": 5410
    },
    {
      "entropy": 5.021324634552002,
      "epoch": 1.530958439355386,
      "grad_norm": 1.3125,
      "learning_rate": 0.000481923768235971,
      "loss": 4.9763,
      "mean_token_accuracy": 0.2372221827507019,
      "num_tokens": 4319044.0,
      "step": 5415
    },
    {
      "entropy": 5.106617259979248,
      "epoch": 1.532372066723212,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048188335192964776,
      "loss": 4.9342,
      "mean_token_accuracy": 0.2337781861424446,
      "num_tokens": 4322980.0,
      "step": 5420
    },
    {
      "entropy": 5.053520011901855,
      "epoch": 1.5337856940910375,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004818428923859834,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22488932758569719,
      "num_tokens": 4326971.0,
      "step": 5425
    },
    {
      "entropy": 5.09512186050415,
      "epoch": 1.5351993214588635,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004818023896134337,
      "loss": 4.8964,
      "mean_token_accuracy": 0.24536970108747483,
      "num_tokens": 4331124.0,
      "step": 5430
    },
    {
      "entropy": 5.040243196487427,
      "epoch": 1.5366129488266893,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048176184362046335,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2351780891418457,
      "num_tokens": 4334945.0,
      "step": 5435
    },
    {
      "entropy": 5.064059209823609,
      "epoch": 1.538026576194515,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004817212544155462,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21822983771562576,
      "num_tokens": 4338989.0,
      "step": 5440
    },
    {
      "entropy": 5.034025812149048,
      "epoch": 1.5394402035623411,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048168062200716524,
      "loss": 4.9965,
      "mean_token_accuracy": 0.23389795869588853,
      "num_tokens": 4342748.0,
      "step": 5445
    },
    {
      "entropy": 5.073337268829346,
      "epoch": 1.5408538309301667,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004816399464038124,
      "loss": 5.0029,
      "mean_token_accuracy": 0.23524331897497178,
      "num_tokens": 4346598.0,
      "step": 5450
    },
    {
      "entropy": 5.008875370025635,
      "epoch": 1.5422674582979927,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048159922761398846,
      "loss": 4.96,
      "mean_token_accuracy": 0.22971295416355134,
      "num_tokens": 4350477.0,
      "step": 5455
    },
    {
      "entropy": 4.971022462844848,
      "epoch": 1.5436810856658185,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004815584656462036,
      "loss": 4.8553,
      "mean_token_accuracy": 0.2356441929936409,
      "num_tokens": 4354556.0,
      "step": 5460
    },
    {
      "entropy": 5.002338075637818,
      "epoch": 1.5450947130336443,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004815176605089767,
      "loss": 4.9061,
      "mean_token_accuracy": 0.2365303412079811,
      "num_tokens": 4358575.0,
      "step": 5465
    },
    {
      "entropy": 5.110936832427979,
      "epoch": 1.5465083404014703,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004814768122108359,
      "loss": 5.0292,
      "mean_token_accuracy": 0.22225579768419265,
      "num_tokens": 4362555.0,
      "step": 5470
    },
    {
      "entropy": 4.964795923233032,
      "epoch": 1.547921967769296,
      "grad_norm": 1.25,
      "learning_rate": 0.00048143592076031803,
      "loss": 4.9084,
      "mean_token_accuracy": 0.23802143931388856,
      "num_tokens": 4366714.0,
      "step": 5475
    },
    {
      "entropy": 5.005723905563355,
      "epoch": 1.549335595137122,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004813949861659694,
      "loss": 4.9217,
      "mean_token_accuracy": 0.229959636926651,
      "num_tokens": 4370591.0,
      "step": 5480
    },
    {
      "entropy": 5.033409500122071,
      "epoch": 1.5507492225049477,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048135400843634497,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22791687846183778,
      "num_tokens": 4374436.0,
      "step": 5485
    },
    {
      "entropy": 5.062553453445434,
      "epoch": 1.5521628498727735,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048131298758000884,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2214925020933151,
      "num_tokens": 4378655.0,
      "step": 5490
    },
    {
      "entropy": 5.054187822341919,
      "epoch": 1.5535764772405993,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004812719236055341,
      "loss": 4.9291,
      "mean_token_accuracy": 0.23569601029157639,
      "num_tokens": 4382457.0,
      "step": 5495
    },
    {
      "entropy": 5.042050075531006,
      "epoch": 1.5549901046084251,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004812308165215028,
      "loss": 5.0119,
      "mean_token_accuracy": 0.22637168765068055,
      "num_tokens": 4386300.0,
      "step": 5500
    },
    {
      "entropy": 4.95188455581665,
      "epoch": 1.5564037319762511,
      "grad_norm": 1.25,
      "learning_rate": 0.00048118966633650613,
      "loss": 4.9347,
      "mean_token_accuracy": 0.24873102009296416,
      "num_tokens": 4390285.0,
      "step": 5505
    },
    {
      "entropy": 5.109093284606933,
      "epoch": 1.557817359344077,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004811484730591444,
      "loss": 4.996,
      "mean_token_accuracy": 0.22325274348258972,
      "num_tokens": 4394268.0,
      "step": 5510
    },
    {
      "entropy": 5.074758386611938,
      "epoch": 1.5592309867119027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004811072366980264,
      "loss": 4.8203,
      "mean_token_accuracy": 0.25320621728897097,
      "num_tokens": 4398310.0,
      "step": 5515
    },
    {
      "entropy": 5.029850053787231,
      "epoch": 1.5606446140797285,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004810659572617705,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2398248240351677,
      "num_tokens": 4402306.0,
      "step": 5520
    },
    {
      "entropy": 5.096603965759277,
      "epoch": 1.5620582414475543,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048102463475900386,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22897083908319474,
      "num_tokens": 4406263.0,
      "step": 5525
    },
    {
      "entropy": 5.116132354736328,
      "epoch": 1.5634718688153804,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048098326919836253,
      "loss": 5.0386,
      "mean_token_accuracy": 0.22826100140810013,
      "num_tokens": 4410383.0,
      "step": 5530
    },
    {
      "entropy": 5.071072912216186,
      "epoch": 1.5648854961832062,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004809418605884917,
      "loss": 4.8928,
      "mean_token_accuracy": 0.24345889836549758,
      "num_tokens": 4414244.0,
      "step": 5535
    },
    {
      "entropy": 4.98833475112915,
      "epoch": 1.566299123551032,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004809004089380456,
      "loss": 4.8832,
      "mean_token_accuracy": 0.246638685464859,
      "num_tokens": 4418315.0,
      "step": 5540
    },
    {
      "entropy": 5.010937929153442,
      "epoch": 1.5677127509188578,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004808589142556872,
      "loss": 5.024,
      "mean_token_accuracy": 0.23410337269306183,
      "num_tokens": 4422333.0,
      "step": 5545
    },
    {
      "entropy": 5.035740184783935,
      "epoch": 1.5691263782866836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004808173765500887,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21526957750320436,
      "num_tokens": 4426255.0,
      "step": 5550
    },
    {
      "entropy": 5.009223461151123,
      "epoch": 1.5705400056545096,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048077579582993127,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2406192496418953,
      "num_tokens": 4430398.0,
      "step": 5555
    },
    {
      "entropy": 5.15426869392395,
      "epoch": 1.5719536330223352,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048073417210390504,
      "loss": 5.1335,
      "mean_token_accuracy": 0.22664235383272172,
      "num_tokens": 4434849.0,
      "step": 5560
    },
    {
      "entropy": 4.961669063568115,
      "epoch": 1.5733672603901612,
      "grad_norm": 1.40625,
      "learning_rate": 0.000480692505380709,
      "loss": 4.9215,
      "mean_token_accuracy": 0.2390369474887848,
      "num_tokens": 4438735.0,
      "step": 5565
    },
    {
      "entropy": 5.0195615768432615,
      "epoch": 1.574780887757987,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048065079566905126,
      "loss": 4.9421,
      "mean_token_accuracy": 0.24880785197019578,
      "num_tokens": 4442453.0,
      "step": 5570
    },
    {
      "entropy": 5.1129461288452145,
      "epoch": 1.5761945151258128,
      "grad_norm": 1.328125,
      "learning_rate": 0.000480609042977649,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2371583327651024,
      "num_tokens": 4446552.0,
      "step": 5575
    },
    {
      "entropy": 5.100673961639404,
      "epoch": 1.5776081424936388,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048056724731522824,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23811424821615218,
      "num_tokens": 4450576.0,
      "step": 5580
    },
    {
      "entropy": 4.986291599273682,
      "epoch": 1.5790217698614644,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048052540869052384,
      "loss": 4.8941,
      "mean_token_accuracy": 0.24235827177762986,
      "num_tokens": 4454667.0,
      "step": 5585
    },
    {
      "entropy": 4.921323204040528,
      "epoch": 1.5804353972292904,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048048352711228007,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23454583585262298,
      "num_tokens": 4458337.0,
      "step": 5590
    },
    {
      "entropy": 5.005747842788696,
      "epoch": 1.5818490245971162,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004804416025892498,
      "loss": 4.9765,
      "mean_token_accuracy": 0.23187257051467897,
      "num_tokens": 4462298.0,
      "step": 5595
    },
    {
      "entropy": 5.015349817276001,
      "epoch": 1.583262651964942,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004803996351301949,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21941987723112105,
      "num_tokens": 4466586.0,
      "step": 5600
    },
    {
      "entropy": 4.927168083190918,
      "epoch": 1.584676279332768,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048035762474388656,
      "loss": 4.8506,
      "mean_token_accuracy": 0.2436788186430931,
      "num_tokens": 4470352.0,
      "step": 5605
    },
    {
      "entropy": 4.999021434783936,
      "epoch": 1.5860899067005936,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004803155714391044,
      "loss": 5.0447,
      "mean_token_accuracy": 0.23322182446718215,
      "num_tokens": 4474741.0,
      "step": 5610
    },
    {
      "entropy": 5.0357763290405275,
      "epoch": 1.5875035340684196,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004802734752246375,
      "loss": 4.9427,
      "mean_token_accuracy": 0.23027453273534776,
      "num_tokens": 4478613.0,
      "step": 5615
    },
    {
      "entropy": 5.127229452133179,
      "epoch": 1.5889171614362454,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004802313361092837,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2283207044005394,
      "num_tokens": 4482375.0,
      "step": 5620
    },
    {
      "entropy": 5.00951738357544,
      "epoch": 1.5903307888040712,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004801891541018497,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23590154945850372,
      "num_tokens": 4486672.0,
      "step": 5625
    },
    {
      "entropy": 5.081471729278564,
      "epoch": 1.5917444161718972,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048014692921115136,
      "loss": 5.1098,
      "mean_token_accuracy": 0.2216850534081459,
      "num_tokens": 4490606.0,
      "step": 5630
    },
    {
      "entropy": 5.060573148727417,
      "epoch": 1.5931580435397228,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048010466144601344,
      "loss": 5.0939,
      "mean_token_accuracy": 0.23171295523643493,
      "num_tokens": 4494690.0,
      "step": 5635
    },
    {
      "entropy": 5.040666007995606,
      "epoch": 1.5945716709075488,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004800623508152696,
      "loss": 4.9135,
      "mean_token_accuracy": 0.23520017117261888,
      "num_tokens": 4498576.0,
      "step": 5640
    },
    {
      "entropy": 4.972749996185303,
      "epoch": 1.5959852982753746,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048001999732776246,
      "loss": 4.8353,
      "mean_token_accuracy": 0.24108946323394775,
      "num_tokens": 4502489.0,
      "step": 5645
    },
    {
      "entropy": 5.018619108200073,
      "epoch": 1.5973989256432004,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004799776009923437,
      "loss": 4.9573,
      "mean_token_accuracy": 0.23463171422481538,
      "num_tokens": 4506270.0,
      "step": 5650
    },
    {
      "entropy": 4.981639003753662,
      "epoch": 1.5988125530110264,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047993516181787395,
      "loss": 4.9644,
      "mean_token_accuracy": 0.23098833113908768,
      "num_tokens": 4510306.0,
      "step": 5655
    },
    {
      "entropy": 4.9591179370880125,
      "epoch": 1.600226180378852,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004798926798132226,
      "loss": 4.8155,
      "mean_token_accuracy": 0.24453141391277314,
      "num_tokens": 4514353.0,
      "step": 5660
    },
    {
      "entropy": 4.982988786697388,
      "epoch": 1.601639807746678,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004798501549872681,
      "loss": 5.0037,
      "mean_token_accuracy": 0.23053833097219467,
      "num_tokens": 4518691.0,
      "step": 5665
    },
    {
      "entropy": 5.049451875686645,
      "epoch": 1.6030534351145038,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004798075873488981,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2379942610859871,
      "num_tokens": 4522625.0,
      "step": 5670
    },
    {
      "entropy": 5.057288026809692,
      "epoch": 1.6044670624823296,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047976497690700883,
      "loss": 4.8379,
      "mean_token_accuracy": 0.24098067432641984,
      "num_tokens": 4526595.0,
      "step": 5675
    },
    {
      "entropy": 5.027168846130371,
      "epoch": 1.6058806898501556,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047972232367050557,
      "loss": 4.9249,
      "mean_token_accuracy": 0.23355893939733505,
      "num_tokens": 4530928.0,
      "step": 5680
    },
    {
      "entropy": 4.997762155532837,
      "epoch": 1.6072943172179812,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004796796276483026,
      "loss": 4.9447,
      "mean_token_accuracy": 0.23614317327737808,
      "num_tokens": 4534812.0,
      "step": 5685
    },
    {
      "entropy": 4.9719751358032225,
      "epoch": 1.6087079445858072,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047963688884932316,
      "loss": 4.9561,
      "mean_token_accuracy": 0.24238796383142472,
      "num_tokens": 4538869.0,
      "step": 5690
    },
    {
      "entropy": 5.067907762527466,
      "epoch": 1.610121571953633,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047959410728249947,
      "loss": 4.9865,
      "mean_token_accuracy": 0.23088856637477875,
      "num_tokens": 4543069.0,
      "step": 5695
    },
    {
      "entropy": 5.090418481826783,
      "epoch": 1.6115351993214588,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004795512829567725,
      "loss": 5.0204,
      "mean_token_accuracy": 0.22371262609958648,
      "num_tokens": 4546990.0,
      "step": 5700
    },
    {
      "entropy": 5.021384572982788,
      "epoch": 1.6129488266892849,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004795084158810922,
      "loss": 4.8764,
      "mean_token_accuracy": 0.24143022298812866,
      "num_tokens": 4550979.0,
      "step": 5705
    },
    {
      "entropy": 5.007719802856445,
      "epoch": 1.6143624540571104,
      "grad_norm": 1.25,
      "learning_rate": 0.0004794655060644177,
      "loss": 4.7851,
      "mean_token_accuracy": 0.23614216595888138,
      "num_tokens": 4555002.0,
      "step": 5710
    },
    {
      "entropy": 4.852123117446899,
      "epoch": 1.6157760814249365,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004794225535157168,
      "loss": 4.7782,
      "mean_token_accuracy": 0.24855546653270721,
      "num_tokens": 4558813.0,
      "step": 5715
    },
    {
      "entropy": 5.028229188919068,
      "epoch": 1.6171897087927622,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047937955824396634,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22896572053432465,
      "num_tokens": 4562941.0,
      "step": 5720
    },
    {
      "entropy": 5.038366174697876,
      "epoch": 1.618603336160588,
      "grad_norm": 1.453125,
      "learning_rate": 0.000479336520258152,
      "loss": 4.9872,
      "mean_token_accuracy": 0.23519830107688905,
      "num_tokens": 4566820.0,
      "step": 5725
    },
    {
      "entropy": 4.97196044921875,
      "epoch": 1.620016963528414,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047929343956726854,
      "loss": 4.8866,
      "mean_token_accuracy": 0.24217161238193513,
      "num_tokens": 4570577.0,
      "step": 5730
    },
    {
      "entropy": 5.019470834732056,
      "epoch": 1.6214305908962396,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047925031618031963,
      "loss": 5.0253,
      "mean_token_accuracy": 0.2278924986720085,
      "num_tokens": 4574721.0,
      "step": 5735
    },
    {
      "entropy": 5.0640359878540036,
      "epoch": 1.6228442182640657,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004792071501063175,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2343967765569687,
      "num_tokens": 4578499.0,
      "step": 5740
    },
    {
      "entropy": 5.0430145263671875,
      "epoch": 1.6242578456318915,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047916394135428386,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2328893169760704,
      "num_tokens": 4582377.0,
      "step": 5745
    },
    {
      "entropy": 5.0419713973999025,
      "epoch": 1.6256714729997173,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047912068993324897,
      "loss": 4.924,
      "mean_token_accuracy": 0.23506866097450257,
      "num_tokens": 4586396.0,
      "step": 5750
    },
    {
      "entropy": 4.999245119094849,
      "epoch": 1.6270851003675433,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047907739585225214,
      "loss": 5.0071,
      "mean_token_accuracy": 0.24333165138959884,
      "num_tokens": 4590605.0,
      "step": 5755
    },
    {
      "entropy": 5.070264387130737,
      "epoch": 1.6284987277353689,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004790340591203416,
      "loss": 4.911,
      "mean_token_accuracy": 0.23459983766078948,
      "num_tokens": 4594759.0,
      "step": 5760
    },
    {
      "entropy": 4.993712711334228,
      "epoch": 1.6299123551031949,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004789906797465744,
      "loss": 4.8285,
      "mean_token_accuracy": 0.25126106292009354,
      "num_tokens": 4598738.0,
      "step": 5765
    },
    {
      "entropy": 5.053193140029907,
      "epoch": 1.6313259824710207,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004789472577400165,
      "loss": 5.0997,
      "mean_token_accuracy": 0.22495850324630737,
      "num_tokens": 4602843.0,
      "step": 5770
    },
    {
      "entropy": 4.945891904830932,
      "epoch": 1.6327396098388465,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047890379310974295,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2434600129723549,
      "num_tokens": 4606826.0,
      "step": 5775
    },
    {
      "entropy": 5.021273851394653,
      "epoch": 1.6341532372066723,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004788602858648375,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22081454396247863,
      "num_tokens": 4610476.0,
      "step": 5780
    },
    {
      "entropy": 5.023816823959351,
      "epoch": 1.635566864574498,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047881673601439306,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2258377805352211,
      "num_tokens": 4614287.0,
      "step": 5785
    },
    {
      "entropy": 5.059210586547851,
      "epoch": 1.636980491942324,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000478773143567511,
      "loss": 4.9646,
      "mean_token_accuracy": 0.23210439234972,
      "num_tokens": 4618279.0,
      "step": 5790
    },
    {
      "entropy": 5.02095627784729,
      "epoch": 1.6383941193101499,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004787295085333021,
      "loss": 4.8858,
      "mean_token_accuracy": 0.24506396651268006,
      "num_tokens": 4622491.0,
      "step": 5795
    },
    {
      "entropy": 4.972891426086425,
      "epoch": 1.6398077466779757,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004786858309208858,
      "loss": 4.8612,
      "mean_token_accuracy": 0.23267780840396882,
      "num_tokens": 4626314.0,
      "step": 5800
    },
    {
      "entropy": 5.041982698440552,
      "epoch": 1.6412213740458015,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047864211073939036,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23292933255434037,
      "num_tokens": 4630593.0,
      "step": 5805
    },
    {
      "entropy": 5.020829439163208,
      "epoch": 1.6426350014136273,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047859834799795305,
      "loss": 4.951,
      "mean_token_accuracy": 0.24297254532575607,
      "num_tokens": 4634562.0,
      "step": 5810
    },
    {
      "entropy": 5.015611410140991,
      "epoch": 1.6440486287814533,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047855454270571995,
      "loss": 4.935,
      "mean_token_accuracy": 0.23390918970108032,
      "num_tokens": 4638689.0,
      "step": 5815
    },
    {
      "entropy": 4.935266351699829,
      "epoch": 1.645462256149279,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004785106948718463,
      "loss": 4.8694,
      "mean_token_accuracy": 0.24739470332860947,
      "num_tokens": 4642544.0,
      "step": 5820
    },
    {
      "entropy": 4.966820621490479,
      "epoch": 1.646875883517105,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004784668045054958,
      "loss": 4.844,
      "mean_token_accuracy": 0.2505292922258377,
      "num_tokens": 4646247.0,
      "step": 5825
    },
    {
      "entropy": 5.0284491062164305,
      "epoch": 1.6482895108849307,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047842287161584135,
      "loss": 4.8941,
      "mean_token_accuracy": 0.2343850925564766,
      "num_tokens": 4650503.0,
      "step": 5830
    },
    {
      "entropy": 5.004018545150757,
      "epoch": 1.6497031382527565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004783788962120647,
      "loss": 4.9799,
      "mean_token_accuracy": 0.23461047410964966,
      "num_tokens": 4654541.0,
      "step": 5835
    },
    {
      "entropy": 5.034882736206055,
      "epoch": 1.6511167656205825,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047833487830335644,
      "loss": 4.9915,
      "mean_token_accuracy": 0.2287354588508606,
      "num_tokens": 4658619.0,
      "step": 5840
    },
    {
      "entropy": 4.850153827667237,
      "epoch": 1.652530392988408,
      "grad_norm": 1.296875,
      "learning_rate": 0.000478290817898916,
      "loss": 4.8516,
      "mean_token_accuracy": 0.24215939193964003,
      "num_tokens": 4662417.0,
      "step": 5845
    },
    {
      "entropy": 4.9653009414672855,
      "epoch": 1.6539440203562341,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047824671500795165,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2231622412800789,
      "num_tokens": 4666779.0,
      "step": 5850
    },
    {
      "entropy": 5.081892013549805,
      "epoch": 1.65535764772406,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047820256963968075,
      "loss": 4.9787,
      "mean_token_accuracy": 0.23246766477823258,
      "num_tokens": 4671010.0,
      "step": 5855
    },
    {
      "entropy": 5.172537612915039,
      "epoch": 1.6567712750918857,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047815838180332936,
      "loss": 5.0342,
      "mean_token_accuracy": 0.22071370333433152,
      "num_tokens": 4675225.0,
      "step": 5860
    },
    {
      "entropy": 5.007997941970825,
      "epoch": 1.6581849024597117,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047811415150813245,
      "loss": 4.8716,
      "mean_token_accuracy": 0.24120320677757262,
      "num_tokens": 4679488.0,
      "step": 5865
    },
    {
      "entropy": 5.002289342880249,
      "epoch": 1.6595985298275373,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047806987876333393,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22495046854019166,
      "num_tokens": 4683271.0,
      "step": 5870
    },
    {
      "entropy": 4.926527070999145,
      "epoch": 1.6610121571953633,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047802556357818643,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23844262808561326,
      "num_tokens": 4687240.0,
      "step": 5875
    },
    {
      "entropy": 4.977902793884278,
      "epoch": 1.6624257845631891,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047798120596195164,
      "loss": 4.8741,
      "mean_token_accuracy": 0.2297168642282486,
      "num_tokens": 4691322.0,
      "step": 5880
    },
    {
      "entropy": 5.015220022201538,
      "epoch": 1.663839411931015,
      "grad_norm": 1.375,
      "learning_rate": 0.0004779368059239,
      "loss": 5.0065,
      "mean_token_accuracy": 0.23096994757652284,
      "num_tokens": 4695369.0,
      "step": 5885
    },
    {
      "entropy": 5.0626311779022215,
      "epoch": 1.665253039298841,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004778923634733109,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2201218768954277,
      "num_tokens": 4699392.0,
      "step": 5890
    },
    {
      "entropy": 5.029302406311035,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004778478786194724,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2236104875802994,
      "num_tokens": 4703112.0,
      "step": 5895
    },
    {
      "entropy": 4.985857200622559,
      "epoch": 1.6680802940344925,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047780335137168176,
      "loss": 4.8581,
      "mean_token_accuracy": 0.24740647822618483,
      "num_tokens": 4707201.0,
      "step": 5900
    },
    {
      "entropy": 5.038682508468628,
      "epoch": 1.6694939214023183,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047775878173924463,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2290230482816696,
      "num_tokens": 4711257.0,
      "step": 5905
    },
    {
      "entropy": 5.010770273208618,
      "epoch": 1.6709075487701441,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047771416973147607,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23835608512163162,
      "num_tokens": 4715438.0,
      "step": 5910
    },
    {
      "entropy": 5.0856860160827635,
      "epoch": 1.6723211761379702,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047766951535769956,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2303079053759575,
      "num_tokens": 4719524.0,
      "step": 5915
    },
    {
      "entropy": 5.0622156143188475,
      "epoch": 1.6737348035057957,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047762481862724754,
      "loss": 4.9335,
      "mean_token_accuracy": 0.23119895607233049,
      "num_tokens": 4723350.0,
      "step": 5920
    },
    {
      "entropy": 5.019683885574341,
      "epoch": 1.6751484308736218,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047758007954946144,
      "loss": 5.0181,
      "mean_token_accuracy": 0.23447049856185914,
      "num_tokens": 4727225.0,
      "step": 5925
    },
    {
      "entropy": 4.948367118835449,
      "epoch": 1.6765620582414476,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047753529813369147,
      "loss": 4.7975,
      "mean_token_accuracy": 0.2506874784827232,
      "num_tokens": 4731252.0,
      "step": 5930
    },
    {
      "entropy": 4.97208890914917,
      "epoch": 1.6779756856092733,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047749047438929663,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2391890376806259,
      "num_tokens": 4735428.0,
      "step": 5935
    },
    {
      "entropy": 4.947220706939698,
      "epoch": 1.6793893129770994,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047744560832564475,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23972647339105607,
      "num_tokens": 4739480.0,
      "step": 5940
    },
    {
      "entropy": 5.142313528060913,
      "epoch": 1.680802940344925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004774006999521127,
      "loss": 5.059,
      "mean_token_accuracy": 0.2171017974615097,
      "num_tokens": 4744145.0,
      "step": 5945
    },
    {
      "entropy": 5.082251358032226,
      "epoch": 1.682216567712751,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004773557492780859,
      "loss": 4.9078,
      "mean_token_accuracy": 0.243063785135746,
      "num_tokens": 4748088.0,
      "step": 5950
    },
    {
      "entropy": 5.012283992767334,
      "epoch": 1.6836301950805768,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004773107563129588,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2438160791993141,
      "num_tokens": 4751874.0,
      "step": 5955
    },
    {
      "entropy": 5.034447574615479,
      "epoch": 1.6850438224484026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004772657210661348,
      "loss": 5.0254,
      "mean_token_accuracy": 0.22636561542749406,
      "num_tokens": 4755934.0,
      "step": 5960
    },
    {
      "entropy": 5.054370641708374,
      "epoch": 1.6864574498162286,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004772206435470258,
      "loss": 4.9592,
      "mean_token_accuracy": 0.2367024913430214,
      "num_tokens": 4759554.0,
      "step": 5965
    },
    {
      "entropy": 5.039686584472657,
      "epoch": 1.6878710771840542,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004771755237650528,
      "loss": 4.883,
      "mean_token_accuracy": 0.24380149692296982,
      "num_tokens": 4763508.0,
      "step": 5970
    },
    {
      "entropy": 4.977809429168701,
      "epoch": 1.6892847045518802,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004771303617296455,
      "loss": 4.7986,
      "mean_token_accuracy": 0.24572733640670777,
      "num_tokens": 4767408.0,
      "step": 5975
    },
    {
      "entropy": 4.975335884094238,
      "epoch": 1.690698331919706,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004770851574502426,
      "loss": 4.9916,
      "mean_token_accuracy": 0.23698644936084748,
      "num_tokens": 4771347.0,
      "step": 5980
    },
    {
      "entropy": 5.003446626663208,
      "epoch": 1.6921119592875318,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047703991093629147,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22792416512966157,
      "num_tokens": 4775541.0,
      "step": 5985
    },
    {
      "entropy": 4.970942163467408,
      "epoch": 1.6935255866553578,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004769946221972483,
      "loss": 4.7575,
      "mean_token_accuracy": 0.25294104367494585,
      "num_tokens": 4779485.0,
      "step": 5990
    },
    {
      "entropy": 5.069515132904053,
      "epoch": 1.6949392140231834,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004769492912425782,
      "loss": 4.986,
      "mean_token_accuracy": 0.2262274518609047,
      "num_tokens": 4783619.0,
      "step": 5995
    },
    {
      "entropy": 5.016628122329712,
      "epoch": 1.6963528413910094,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047690391808175497,
      "loss": 4.9316,
      "mean_token_accuracy": 0.23867592811584473,
      "num_tokens": 4787648.0,
      "step": 6000
    },
    {
      "epoch": 1.6963528413910094,
      "eval_entropy": 4.787501366234772,
      "eval_loss": 5.196343421936035,
      "eval_mean_token_accuracy": 0.23100133312035756,
      "eval_num_tokens": 4787648.0,
      "eval_runtime": 4.6352,
      "eval_samples_per_second": 2434.419,
      "eval_steps_per_second": 304.41,
      "step": 6000
    },
    {
      "entropy": 4.967789888381958,
      "epoch": 1.6977664687588352,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768585027242615,
      "loss": 4.9092,
      "mean_token_accuracy": 0.24383485615253447,
      "num_tokens": 4791302.0,
      "step": 6005
    },
    {
      "entropy": 5.041885566711426,
      "epoch": 1.699180096126661,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047681304517958916,
      "loss": 4.8598,
      "mean_token_accuracy": 0.24095146209001542,
      "num_tokens": 4795481.0,
      "step": 6010
    },
    {
      "entropy": 5.119360303878784,
      "epoch": 1.700593723494487,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047676754545723836,
      "loss": 5.0378,
      "mean_token_accuracy": 0.22744368761777878,
      "num_tokens": 4799380.0,
      "step": 6015
    },
    {
      "entropy": 4.984142589569092,
      "epoch": 1.7020073508623126,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004767220035667183,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22570559233427048,
      "num_tokens": 4802946.0,
      "step": 6020
    },
    {
      "entropy": 5.070523691177368,
      "epoch": 1.7034209782301386,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004766764195175469,
      "loss": 4.9905,
      "mean_token_accuracy": 0.22667042315006256,
      "num_tokens": 4806925.0,
      "step": 6025
    },
    {
      "entropy": 4.94235053062439,
      "epoch": 1.7048346055979644,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047663079331925086,
      "loss": 4.842,
      "mean_token_accuracy": 0.24382590055465697,
      "num_tokens": 4810879.0,
      "step": 6030
    },
    {
      "entropy": 4.911730527877808,
      "epoch": 1.7062482329657902,
      "grad_norm": 1.40625,
      "learning_rate": 0.000476585124981366,
      "loss": 4.7722,
      "mean_token_accuracy": 0.25501230359077454,
      "num_tokens": 4814863.0,
      "step": 6035
    },
    {
      "entropy": 4.926906299591065,
      "epoch": 1.7076618603336162,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004765394145134365,
      "loss": 4.8579,
      "mean_token_accuracy": 0.24565281718969345,
      "num_tokens": 4818548.0,
      "step": 6040
    },
    {
      "entropy": 4.938519716262817,
      "epoch": 1.7090754877014418,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004764936619250156,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22830152064561843,
      "num_tokens": 4822693.0,
      "step": 6045
    },
    {
      "entropy": 5.033918809890747,
      "epoch": 1.7104891150692678,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047644786722566537,
      "loss": 4.9512,
      "mean_token_accuracy": 0.23547366112470627,
      "num_tokens": 4827110.0,
      "step": 6050
    },
    {
      "entropy": 4.992440414428711,
      "epoch": 1.7119027424370936,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047640203042495667,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23438241928815842,
      "num_tokens": 4831327.0,
      "step": 6055
    },
    {
      "entropy": 4.947700119018554,
      "epoch": 1.7133163698049194,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047635615153246907,
      "loss": 4.7957,
      "mean_token_accuracy": 0.24477908611297608,
      "num_tokens": 4835767.0,
      "step": 6060
    },
    {
      "entropy": 5.015694808959961,
      "epoch": 1.7147299971727452,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004763102305577908,
      "loss": 4.9072,
      "mean_token_accuracy": 0.24117992520332338,
      "num_tokens": 4840163.0,
      "step": 6065
    },
    {
      "entropy": 5.094604396820069,
      "epoch": 1.716143624540571,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004762642675105194,
      "loss": 4.951,
      "mean_token_accuracy": 0.22490872591733932,
      "num_tokens": 4843874.0,
      "step": 6070
    },
    {
      "entropy": 5.023176670074463,
      "epoch": 1.717557251908397,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004762182624002606,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23475403040647508,
      "num_tokens": 4848042.0,
      "step": 6075
    },
    {
      "entropy": 4.879738903045654,
      "epoch": 1.7189708792762228,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047617221523662916,
      "loss": 4.8652,
      "mean_token_accuracy": 0.24273412823677062,
      "num_tokens": 4851998.0,
      "step": 6080
    },
    {
      "entropy": 4.993034553527832,
      "epoch": 1.7203845066440486,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047612612602924883,
      "loss": 4.847,
      "mean_token_accuracy": 0.24026925265789031,
      "num_tokens": 4856141.0,
      "step": 6085
    },
    {
      "entropy": 4.958903360366821,
      "epoch": 1.7217981340118744,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004760799947877518,
      "loss": 4.951,
      "mean_token_accuracy": 0.23311926573514938,
      "num_tokens": 4860069.0,
      "step": 6090
    },
    {
      "entropy": 4.958827304840088,
      "epoch": 1.7232117613797002,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004760338215217795,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2262375146150589,
      "num_tokens": 4864042.0,
      "step": 6095
    },
    {
      "entropy": 4.95157675743103,
      "epoch": 1.7246253887475262,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004759876062409815,
      "loss": 4.7692,
      "mean_token_accuracy": 0.25691608935594556,
      "num_tokens": 4868086.0,
      "step": 6100
    },
    {
      "entropy": 5.051073741912842,
      "epoch": 1.726039016115352,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004759413489550167,
      "loss": 5.0479,
      "mean_token_accuracy": 0.21923680901527404,
      "num_tokens": 4872034.0,
      "step": 6105
    },
    {
      "entropy": 5.014369678497315,
      "epoch": 1.7274526434831778,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047589504967355254,
      "loss": 4.9432,
      "mean_token_accuracy": 0.23702065497636796,
      "num_tokens": 4876293.0,
      "step": 6110
    },
    {
      "entropy": 5.017498254776001,
      "epoch": 1.7288662708510036,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004758487084062653,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23894359916448593,
      "num_tokens": 4880573.0,
      "step": 6115
    },
    {
      "entropy": 5.014670133590698,
      "epoch": 1.7302798982188294,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047580232516284,
      "loss": 4.9631,
      "mean_token_accuracy": 0.23239366859197616,
      "num_tokens": 4884598.0,
      "step": 6120
    },
    {
      "entropy": 4.931390333175659,
      "epoch": 1.7316935255866555,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047575589995297045,
      "loss": 4.9437,
      "mean_token_accuracy": 0.24026952981948851,
      "num_tokens": 4888343.0,
      "step": 6125
    },
    {
      "entropy": 5.053763341903687,
      "epoch": 1.733107152954481,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047570943278635937,
      "loss": 4.923,
      "mean_token_accuracy": 0.23619078248739242,
      "num_tokens": 4892876.0,
      "step": 6130
    },
    {
      "entropy": 5.022941446304321,
      "epoch": 1.734520780322307,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047566292367271786,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2390052005648613,
      "num_tokens": 4896589.0,
      "step": 6135
    },
    {
      "entropy": 5.016025543212891,
      "epoch": 1.7359344076901329,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047561637262176623,
      "loss": 4.874,
      "mean_token_accuracy": 0.2377936691045761,
      "num_tokens": 4900868.0,
      "step": 6140
    },
    {
      "entropy": 5.069879341125488,
      "epoch": 1.7373480350579587,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004755697796432333,
      "loss": 4.9921,
      "mean_token_accuracy": 0.23275185078382493,
      "num_tokens": 4904689.0,
      "step": 6145
    },
    {
      "entropy": 4.974972486495972,
      "epoch": 1.7387616624257847,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047552314474685664,
      "loss": 4.7488,
      "mean_token_accuracy": 0.24192290008068085,
      "num_tokens": 4908421.0,
      "step": 6150
    },
    {
      "entropy": 4.8942280292510985,
      "epoch": 1.7401752897936102,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047547646794238277,
      "loss": 4.8337,
      "mean_token_accuracy": 0.24013668596744536,
      "num_tokens": 4912481.0,
      "step": 6155
    },
    {
      "entropy": 4.970977687835694,
      "epoch": 1.7415889171614363,
      "grad_norm": 1.375,
      "learning_rate": 0.0004754297492395668,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2349382609128952,
      "num_tokens": 4916430.0,
      "step": 6160
    },
    {
      "entropy": 5.039912319183349,
      "epoch": 1.743002544529262,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004753829886481727,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2515605241060257,
      "num_tokens": 4920322.0,
      "step": 6165
    },
    {
      "entropy": 4.942368078231811,
      "epoch": 1.7444161718970879,
      "grad_norm": 1.3125,
      "learning_rate": 0.000475336186177973,
      "loss": 4.9031,
      "mean_token_accuracy": 0.24391392469406128,
      "num_tokens": 4924386.0,
      "step": 6170
    },
    {
      "entropy": 5.008260011672974,
      "epoch": 1.7458297992649139,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047528934183874936,
      "loss": 4.998,
      "mean_token_accuracy": 0.2253402978181839,
      "num_tokens": 4928711.0,
      "step": 6175
    },
    {
      "entropy": 5.065802907943725,
      "epoch": 1.7472434266327395,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004752424556402918,
      "loss": 4.844,
      "mean_token_accuracy": 0.24829844385385513,
      "num_tokens": 4932810.0,
      "step": 6180
    },
    {
      "entropy": 4.928239822387695,
      "epoch": 1.7486570540005655,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004751955275923992,
      "loss": 4.7882,
      "mean_token_accuracy": 0.2535097599029541,
      "num_tokens": 4936893.0,
      "step": 6185
    },
    {
      "entropy": 4.873041200637817,
      "epoch": 1.7500706813683913,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047514855770487935,
      "loss": 4.8434,
      "mean_token_accuracy": 0.24244190752506256,
      "num_tokens": 4940942.0,
      "step": 6190
    },
    {
      "entropy": 4.98823504447937,
      "epoch": 1.751484308736217,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047510154598754865,
      "loss": 5.033,
      "mean_token_accuracy": 0.22440304309129716,
      "num_tokens": 4945366.0,
      "step": 6195
    },
    {
      "entropy": 5.046714925765992,
      "epoch": 1.752897936104043,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047505449245023216,
      "loss": 4.8747,
      "mean_token_accuracy": 0.24124807864427567,
      "num_tokens": 4949276.0,
      "step": 6200
    },
    {
      "entropy": 5.025658798217774,
      "epoch": 1.7543115634718687,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004750073971027639,
      "loss": 4.9159,
      "mean_token_accuracy": 0.2375473067164421,
      "num_tokens": 4953068.0,
      "step": 6205
    },
    {
      "entropy": 4.966525745391846,
      "epoch": 1.7557251908396947,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004749602599549865,
      "loss": 4.8088,
      "mean_token_accuracy": 0.2450304239988327,
      "num_tokens": 4957148.0,
      "step": 6210
    },
    {
      "entropy": 5.032566165924072,
      "epoch": 1.7571388182075205,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004749130810167511,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23088853657245637,
      "num_tokens": 4961070.0,
      "step": 6215
    },
    {
      "entropy": 5.057465934753418,
      "epoch": 1.7585524455753463,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047486586029791823,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22729524224996567,
      "num_tokens": 4964883.0,
      "step": 6220
    },
    {
      "entropy": 4.960890197753907,
      "epoch": 1.7599660729431723,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047481859780835637,
      "loss": 5.0208,
      "mean_token_accuracy": 0.23460467755794526,
      "num_tokens": 4968825.0,
      "step": 6225
    },
    {
      "entropy": 5.054586553573609,
      "epoch": 1.7613797003109979,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004747712935579431,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2263317808508873,
      "num_tokens": 4973067.0,
      "step": 6230
    },
    {
      "entropy": 5.078257703781128,
      "epoch": 1.762793327678824,
      "grad_norm": 1.265625,
      "learning_rate": 0.000474723947556565,
      "loss": 4.9655,
      "mean_token_accuracy": 0.24151697605848313,
      "num_tokens": 4977215.0,
      "step": 6235
    },
    {
      "entropy": 5.0086832523345945,
      "epoch": 1.7642069550466497,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004746765598141169,
      "loss": 4.925,
      "mean_token_accuracy": 0.22978031039237976,
      "num_tokens": 4980749.0,
      "step": 6240
    },
    {
      "entropy": 4.986809015274048,
      "epoch": 1.7656205824144755,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047462913034050245,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2421949699521065,
      "num_tokens": 4984757.0,
      "step": 6245
    },
    {
      "entropy": 5.018244981765747,
      "epoch": 1.7670342097823015,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047458165914563436,
      "loss": 4.9799,
      "mean_token_accuracy": 0.23878287374973298,
      "num_tokens": 4988657.0,
      "step": 6250
    },
    {
      "entropy": 5.03305344581604,
      "epoch": 1.768447837150127,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004745341462394336,
      "loss": 4.9742,
      "mean_token_accuracy": 0.24029705673456192,
      "num_tokens": 4992479.0,
      "step": 6255
    },
    {
      "entropy": 5.055025434494018,
      "epoch": 1.7698614645179531,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004744865916318302,
      "loss": 5.061,
      "mean_token_accuracy": 0.2202444091439247,
      "num_tokens": 4996495.0,
      "step": 6260
    },
    {
      "entropy": 5.063772821426392,
      "epoch": 1.771275091885779,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047443899533276276,
      "loss": 4.967,
      "mean_token_accuracy": 0.22889157980680466,
      "num_tokens": 5000503.0,
      "step": 6265
    },
    {
      "entropy": 5.026196002960205,
      "epoch": 1.7726887192536047,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047439135735217845,
      "loss": 4.8573,
      "mean_token_accuracy": 0.24488679319620132,
      "num_tokens": 5004402.0,
      "step": 6270
    },
    {
      "entropy": 4.95560393333435,
      "epoch": 1.7741023466214307,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004743436777000336,
      "loss": 4.9533,
      "mean_token_accuracy": 0.2302405059337616,
      "num_tokens": 5008287.0,
      "step": 6275
    },
    {
      "entropy": 4.931298160552979,
      "epoch": 1.7755159739892563,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047429595638629276,
      "loss": 4.8679,
      "mean_token_accuracy": 0.24718620181083678,
      "num_tokens": 5012217.0,
      "step": 6280
    },
    {
      "entropy": 5.108009672164917,
      "epoch": 1.7769296013570823,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004742481934209295,
      "loss": 5.0242,
      "mean_token_accuracy": 0.22131722271442414,
      "num_tokens": 5016454.0,
      "step": 6285
    },
    {
      "entropy": 5.050626754760742,
      "epoch": 1.7783432287249081,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004742003888139258,
      "loss": 4.9498,
      "mean_token_accuracy": 0.233401158452034,
      "num_tokens": 5020483.0,
      "step": 6290
    },
    {
      "entropy": 5.011825275421143,
      "epoch": 1.779756856092734,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004741525425752727,
      "loss": 4.8241,
      "mean_token_accuracy": 0.24205415248870848,
      "num_tokens": 5024091.0,
      "step": 6295
    },
    {
      "entropy": 4.991188001632691,
      "epoch": 1.78117048346056,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004741046547149699,
      "loss": 4.9766,
      "mean_token_accuracy": 0.23260482400655746,
      "num_tokens": 5028090.0,
      "step": 6300
    },
    {
      "entropy": 5.004785108566284,
      "epoch": 1.7825841108283855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047405672524302524,
      "loss": 4.9533,
      "mean_token_accuracy": 0.23143772184848785,
      "num_tokens": 5032045.0,
      "step": 6305
    },
    {
      "entropy": 4.988901996612549,
      "epoch": 1.7839977381962115,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004740087541694561,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23663478940725327,
      "num_tokens": 5035996.0,
      "step": 6310
    },
    {
      "entropy": 4.971297073364258,
      "epoch": 1.7854113655640373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004739607415042879,
      "loss": 4.8738,
      "mean_token_accuracy": 0.24033638834953308,
      "num_tokens": 5040001.0,
      "step": 6315
    },
    {
      "entropy": 4.828831195831299,
      "epoch": 1.7868249929318631,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004739126872575551,
      "loss": 4.7487,
      "mean_token_accuracy": 0.2623280555009842,
      "num_tokens": 5043888.0,
      "step": 6320
    },
    {
      "entropy": 4.952526378631592,
      "epoch": 1.7882386202996892,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047386459143930064,
      "loss": 4.9026,
      "mean_token_accuracy": 0.24140859842300416,
      "num_tokens": 5047760.0,
      "step": 6325
    },
    {
      "entropy": 5.10832953453064,
      "epoch": 1.7896522476675147,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047381645405957635,
      "loss": 5.0396,
      "mean_token_accuracy": 0.226837220788002,
      "num_tokens": 5051551.0,
      "step": 6330
    },
    {
      "entropy": 5.027993631362915,
      "epoch": 1.7910658750353408,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047376827512844266,
      "loss": 4.871,
      "mean_token_accuracy": 0.2474130004644394,
      "num_tokens": 5055566.0,
      "step": 6335
    },
    {
      "entropy": 5.0467442035675045,
      "epoch": 1.7924795024031666,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004737200546559685,
      "loss": 4.9889,
      "mean_token_accuracy": 0.21631215512752533,
      "num_tokens": 5059599.0,
      "step": 6340
    },
    {
      "entropy": 4.920080709457397,
      "epoch": 1.7938931297709924,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047367179265223183,
      "loss": 4.8425,
      "mean_token_accuracy": 0.23506534993648528,
      "num_tokens": 5063541.0,
      "step": 6345
    },
    {
      "entropy": 4.968846035003662,
      "epoch": 1.7953067571388182,
      "grad_norm": 1.34375,
      "learning_rate": 0.000473623489127319,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2436520501971245,
      "num_tokens": 5067615.0,
      "step": 6350
    },
    {
      "entropy": 5.011948156356811,
      "epoch": 1.796720384506644,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004735751440913253,
      "loss": 4.923,
      "mean_token_accuracy": 0.23049866110086442,
      "num_tokens": 5071825.0,
      "step": 6355
    },
    {
      "entropy": 5.064019632339478,
      "epoch": 1.79813401187447,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047352675755435436,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2319762736558914,
      "num_tokens": 5075806.0,
      "step": 6360
    },
    {
      "entropy": 5.0321760177612305,
      "epoch": 1.7995476392422958,
      "grad_norm": 1.375,
      "learning_rate": 0.0004734783295265188,
      "loss": 4.9051,
      "mean_token_accuracy": 0.23534818440675737,
      "num_tokens": 5079909.0,
      "step": 6365
    },
    {
      "entropy": 5.041381978988648,
      "epoch": 1.8009612666101216,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047342986001793966,
      "loss": 5.1419,
      "mean_token_accuracy": 0.212579907476902,
      "num_tokens": 5083995.0,
      "step": 6370
    },
    {
      "entropy": 4.9673027992248535,
      "epoch": 1.8023748939779474,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004733813490387468,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23824481368064881,
      "num_tokens": 5088086.0,
      "step": 6375
    },
    {
      "entropy": 5.004122972488403,
      "epoch": 1.8037885213457732,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004733327965990788,
      "loss": 4.9325,
      "mean_token_accuracy": 0.233283531665802,
      "num_tokens": 5091816.0,
      "step": 6380
    },
    {
      "entropy": 5.09460072517395,
      "epoch": 1.8052021487135992,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004732842027090828,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22599300146102905,
      "num_tokens": 5095672.0,
      "step": 6385
    },
    {
      "entropy": 5.005027151107788,
      "epoch": 1.806615776081425,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004732355673789145,
      "loss": 5.0196,
      "mean_token_accuracy": 0.22894366085529327,
      "num_tokens": 5099590.0,
      "step": 6390
    },
    {
      "entropy": 4.963682460784912,
      "epoch": 1.8080294034492508,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004731868906187385,
      "loss": 4.8538,
      "mean_token_accuracy": 0.2460229754447937,
      "num_tokens": 5103604.0,
      "step": 6395
    },
    {
      "entropy": 4.975797176361084,
      "epoch": 1.8094430308170766,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047313817243872786,
      "loss": 4.8716,
      "mean_token_accuracy": 0.24058213382959365,
      "num_tokens": 5107548.0,
      "step": 6400
    },
    {
      "entropy": 4.971105575561523,
      "epoch": 1.8108566581849024,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047308941284906456,
      "loss": 4.9212,
      "mean_token_accuracy": 0.24544260352849961,
      "num_tokens": 5111535.0,
      "step": 6405
    },
    {
      "entropy": 5.038210725784301,
      "epoch": 1.8122702855527284,
      "grad_norm": 1.375,
      "learning_rate": 0.00047304061185993884,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2322723686695099,
      "num_tokens": 5115314.0,
      "step": 6410
    },
    {
      "entropy": 5.058740472793579,
      "epoch": 1.813683912920554,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004729917694815499,
      "loss": 4.8748,
      "mean_token_accuracy": 0.23383525758981705,
      "num_tokens": 5119219.0,
      "step": 6415
    },
    {
      "entropy": 4.909460783004761,
      "epoch": 1.81509754028838,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004729428857241055,
      "loss": 4.7351,
      "mean_token_accuracy": 0.25123975723981856,
      "num_tokens": 5122851.0,
      "step": 6420
    },
    {
      "entropy": 4.940452575683594,
      "epoch": 1.8165111676562058,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000472893960597822,
      "loss": 4.9649,
      "mean_token_accuracy": 0.23358141630887985,
      "num_tokens": 5126783.0,
      "step": 6425
    },
    {
      "entropy": 4.9256181716918945,
      "epoch": 1.8179247950240316,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004728449941129244,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2505159735679626,
      "num_tokens": 5130443.0,
      "step": 6430
    },
    {
      "entropy": 5.022846126556397,
      "epoch": 1.8193384223918576,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004727959862796466,
      "loss": 4.8691,
      "mean_token_accuracy": 0.24508552998304367,
      "num_tokens": 5134324.0,
      "step": 6435
    },
    {
      "entropy": 5.034541749954224,
      "epoch": 1.8207520497596832,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047274693710823077,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22270795702934265,
      "num_tokens": 5137950.0,
      "step": 6440
    },
    {
      "entropy": 4.9297192096710205,
      "epoch": 1.8221656771275092,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004726978466089278,
      "loss": 4.751,
      "mean_token_accuracy": 0.24727299809455872,
      "num_tokens": 5141849.0,
      "step": 6445
    },
    {
      "entropy": 4.864014101028443,
      "epoch": 1.823579304495335,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004726487147919976,
      "loss": 4.8974,
      "mean_token_accuracy": 0.24383654445409775,
      "num_tokens": 5145653.0,
      "step": 6450
    },
    {
      "entropy": 4.9667603969573975,
      "epoch": 1.8249929318631608,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004725995416677081,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23920563459396363,
      "num_tokens": 5149517.0,
      "step": 6455
    },
    {
      "entropy": 5.048270034790039,
      "epoch": 1.8264065592309868,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047255032724633633,
      "loss": 5.01,
      "mean_token_accuracy": 0.23020731061697006,
      "num_tokens": 5153525.0,
      "step": 6460
    },
    {
      "entropy": 5.0229823112487795,
      "epoch": 1.8278201865988124,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004725010715381679,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23343727141618728,
      "num_tokens": 5157762.0,
      "step": 6465
    },
    {
      "entropy": 5.072922849655152,
      "epoch": 1.8292338139666384,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047245177455349673,
      "loss": 5.1198,
      "mean_token_accuracy": 0.2288946256041527,
      "num_tokens": 5161740.0,
      "step": 6470
    },
    {
      "entropy": 4.999675273895264,
      "epoch": 1.8306474413344642,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047240243630262567,
      "loss": 4.7998,
      "mean_token_accuracy": 0.24980879873037337,
      "num_tokens": 5165755.0,
      "step": 6475
    },
    {
      "entropy": 5.010202407836914,
      "epoch": 1.83206106870229,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047235305679586614,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2335016831755638,
      "num_tokens": 5169805.0,
      "step": 6480
    },
    {
      "entropy": 4.969574642181397,
      "epoch": 1.833474696070116,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047230363604353813,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23995181769132615,
      "num_tokens": 5173983.0,
      "step": 6485
    },
    {
      "entropy": 5.059214735031128,
      "epoch": 1.8348883234379416,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047225417405597023,
      "loss": 5.0532,
      "mean_token_accuracy": 0.23197215050458908,
      "num_tokens": 5178159.0,
      "step": 6490
    },
    {
      "entropy": 5.040166330337525,
      "epoch": 1.8363019508057676,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047220467084349983,
      "loss": 4.8719,
      "mean_token_accuracy": 0.2477919027209282,
      "num_tokens": 5182204.0,
      "step": 6495
    },
    {
      "entropy": 4.9999885082244875,
      "epoch": 1.8377155781735934,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047215512641647265,
      "loss": 4.8239,
      "mean_token_accuracy": 0.24256523549556733,
      "num_tokens": 5186179.0,
      "step": 6500
    },
    {
      "entropy": 4.914892578125,
      "epoch": 1.8391292055414192,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004721055407852432,
      "loss": 4.8995,
      "mean_token_accuracy": 0.23660919368267058,
      "num_tokens": 5190184.0,
      "step": 6505
    },
    {
      "entropy": 5.039983606338501,
      "epoch": 1.8405428329092453,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004720559139601746,
      "loss": 5.0683,
      "mean_token_accuracy": 0.23612862229347228,
      "num_tokens": 5194126.0,
      "step": 6510
    },
    {
      "entropy": 4.974927997589111,
      "epoch": 1.8419564602770708,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004720062459516385,
      "loss": 4.8809,
      "mean_token_accuracy": 0.24582742154598236,
      "num_tokens": 5198214.0,
      "step": 6515
    },
    {
      "entropy": 4.990711307525634,
      "epoch": 1.8433700876448968,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047195653677001536,
      "loss": 4.8714,
      "mean_token_accuracy": 0.23671845346689224,
      "num_tokens": 5202278.0,
      "step": 6520
    },
    {
      "entropy": 5.082223272323608,
      "epoch": 1.8447837150127226,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004719067864256939,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2357987567782402,
      "num_tokens": 5206475.0,
      "step": 6525
    },
    {
      "entropy": 4.99591326713562,
      "epoch": 1.8461973423805484,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047185699492907174,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23042265474796295,
      "num_tokens": 5210386.0,
      "step": 6530
    },
    {
      "entropy": 4.960571002960205,
      "epoch": 1.8476109697483745,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000471807162290555,
      "loss": 4.8815,
      "mean_token_accuracy": 0.23795896470546724,
      "num_tokens": 5214342.0,
      "step": 6535
    },
    {
      "entropy": 4.909586668014526,
      "epoch": 1.8490245971162,
      "grad_norm": 1.375,
      "learning_rate": 0.00047175728852055825,
      "loss": 4.7721,
      "mean_token_accuracy": 0.2421373501420021,
      "num_tokens": 5218328.0,
      "step": 6540
    },
    {
      "entropy": 4.988396167755127,
      "epoch": 1.850438224484026,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004717073736295051,
      "loss": 4.9307,
      "mean_token_accuracy": 0.23764688372612,
      "num_tokens": 5222580.0,
      "step": 6545
    },
    {
      "entropy": 5.054730749130249,
      "epoch": 1.8518518518518519,
      "grad_norm": 1.375,
      "learning_rate": 0.00047165741762782707,
      "loss": 4.8929,
      "mean_token_accuracy": 0.23106091171503068,
      "num_tokens": 5226568.0,
      "step": 6550
    },
    {
      "entropy": 5.034188890457154,
      "epoch": 1.8532654792196777,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004716074205259649,
      "loss": 4.8611,
      "mean_token_accuracy": 0.24204599410295485,
      "num_tokens": 5230830.0,
      "step": 6555
    },
    {
      "entropy": 5.110712909698487,
      "epoch": 1.8546791065875037,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004715573823343676,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2356141909956932,
      "num_tokens": 5234648.0,
      "step": 6560
    },
    {
      "entropy": 5.036970472335815,
      "epoch": 1.8560927339553293,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004715073030634929,
      "loss": 4.9583,
      "mean_token_accuracy": 0.23680069744586946,
      "num_tokens": 5238592.0,
      "step": 6565
    },
    {
      "entropy": 4.962250089645385,
      "epoch": 1.8575063613231553,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047145718272380706,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2436669573187828,
      "num_tokens": 5242596.0,
      "step": 6570
    },
    {
      "entropy": 5.044175863265991,
      "epoch": 1.858919988690981,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004714070213257849,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21982454508543015,
      "num_tokens": 5246494.0,
      "step": 6575
    },
    {
      "entropy": 5.053348875045776,
      "epoch": 1.8603336160588069,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004713568188799097,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2297771915793419,
      "num_tokens": 5250456.0,
      "step": 6580
    },
    {
      "entropy": 5.028491973876953,
      "epoch": 1.861747243426633,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004713065753966737,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2481822893023491,
      "num_tokens": 5254526.0,
      "step": 6585
    },
    {
      "entropy": 5.032120609283448,
      "epoch": 1.8631608707944585,
      "grad_norm": 1.375,
      "learning_rate": 0.0004712562908865774,
      "loss": 5.0169,
      "mean_token_accuracy": 0.23263430893421172,
      "num_tokens": 5258393.0,
      "step": 6590
    },
    {
      "entropy": 4.915306615829468,
      "epoch": 1.8645744981622845,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047120596536012976,
      "loss": 4.9053,
      "mean_token_accuracy": 0.23999127447605134,
      "num_tokens": 5262272.0,
      "step": 6595
    },
    {
      "entropy": 4.898223924636841,
      "epoch": 1.8659881255301103,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004711555988278488,
      "loss": 4.8196,
      "mean_token_accuracy": 0.24073677361011506,
      "num_tokens": 5266175.0,
      "step": 6600
    },
    {
      "entropy": 4.987274980545044,
      "epoch": 1.867401752897936,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047110519130026066,
      "loss": 4.8499,
      "mean_token_accuracy": 0.2418785497546196,
      "num_tokens": 5269992.0,
      "step": 6605
    },
    {
      "entropy": 5.019068050384521,
      "epoch": 1.868815380265762,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004710547427879002,
      "loss": 4.8552,
      "mean_token_accuracy": 0.23732171207666397,
      "num_tokens": 5274235.0,
      "step": 6610
    },
    {
      "entropy": 4.952259731292725,
      "epoch": 1.8702290076335877,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004710042533013108,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2503652900457382,
      "num_tokens": 5278104.0,
      "step": 6615
    },
    {
      "entropy": 5.0682594776153564,
      "epoch": 1.8716426350014137,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004709537228510446,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2323915183544159,
      "num_tokens": 5282017.0,
      "step": 6620
    },
    {
      "entropy": 5.091192388534546,
      "epoch": 1.8730562623692395,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000470903151447662,
      "loss": 5.0293,
      "mean_token_accuracy": 0.23519181907176973,
      "num_tokens": 5286509.0,
      "step": 6625
    },
    {
      "entropy": 5.040829038619995,
      "epoch": 1.8744698897370653,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004708525391017321,
      "loss": 4.9181,
      "mean_token_accuracy": 0.23326471149921418,
      "num_tokens": 5290297.0,
      "step": 6630
    },
    {
      "entropy": 5.123863267898559,
      "epoch": 1.875883517104891,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047080188582383287,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22867102175951004,
      "num_tokens": 5294738.0,
      "step": 6635
    },
    {
      "entropy": 5.0816779136657715,
      "epoch": 1.877297144472717,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047075119162455015,
      "loss": 4.9345,
      "mean_token_accuracy": 0.24591534286737443,
      "num_tokens": 5298676.0,
      "step": 6640
    },
    {
      "entropy": 4.903414106369018,
      "epoch": 1.878710771840543,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004707004565144789,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2477968379855156,
      "num_tokens": 5302429.0,
      "step": 6645
    },
    {
      "entropy": 4.951305675506592,
      "epoch": 1.8801243992083687,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004706496805042225,
      "loss": 4.8951,
      "mean_token_accuracy": 0.23841262459754944,
      "num_tokens": 5306218.0,
      "step": 6650
    },
    {
      "entropy": 4.952214336395263,
      "epoch": 1.8815380265761945,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047059886360439256,
      "loss": 4.8043,
      "mean_token_accuracy": 0.24402655363082887,
      "num_tokens": 5310294.0,
      "step": 6655
    },
    {
      "entropy": 5.0273655414581295,
      "epoch": 1.8829516539440203,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004705480058256098,
      "loss": 4.9513,
      "mean_token_accuracy": 0.23883764147758485,
      "num_tokens": 5313858.0,
      "step": 6660
    },
    {
      "entropy": 4.939585542678833,
      "epoch": 1.884365281311846,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047049710717850295,
      "loss": 4.8602,
      "mean_token_accuracy": 0.24537258744239807,
      "num_tokens": 5318083.0,
      "step": 6665
    },
    {
      "entropy": 4.962027168273925,
      "epoch": 1.8857789086796721,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047044616767370973,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22506044209003448,
      "num_tokens": 5322024.0,
      "step": 6670
    },
    {
      "entropy": 5.067338132858277,
      "epoch": 1.887192536047498,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000470395187321876,
      "loss": 4.9356,
      "mean_token_accuracy": 0.23359693586826324,
      "num_tokens": 5325903.0,
      "step": 6675
    },
    {
      "entropy": 5.064741230010986,
      "epoch": 1.8886061634153237,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004703441661336565,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2347056746482849,
      "num_tokens": 5329502.0,
      "step": 6680
    },
    {
      "entropy": 5.051345491409302,
      "epoch": 1.8900197907831495,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004702931041197142,
      "loss": 4.9405,
      "mean_token_accuracy": 0.23370400816202164,
      "num_tokens": 5333629.0,
      "step": 6685
    },
    {
      "entropy": 5.020053577423096,
      "epoch": 1.8914334181509753,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047024200129072073,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22949773967266082,
      "num_tokens": 5337551.0,
      "step": 6690
    },
    {
      "entropy": 5.039714050292969,
      "epoch": 1.8928470455188013,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004701908576573564,
      "loss": 4.9406,
      "mean_token_accuracy": 0.2414928048849106,
      "num_tokens": 5341498.0,
      "step": 6695
    },
    {
      "entropy": 5.0743255615234375,
      "epoch": 1.8942606728866271,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047013967323030983,
      "loss": 4.9508,
      "mean_token_accuracy": 0.23676986545324324,
      "num_tokens": 5345574.0,
      "step": 6700
    },
    {
      "entropy": 4.984074354171753,
      "epoch": 1.895674300254453,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004700884480202783,
      "loss": 4.834,
      "mean_token_accuracy": 0.24674287885427476,
      "num_tokens": 5349371.0,
      "step": 6705
    },
    {
      "entropy": 5.031948852539062,
      "epoch": 1.8970879276222787,
      "grad_norm": 1.25,
      "learning_rate": 0.0004700371820379674,
      "loss": 4.9403,
      "mean_token_accuracy": 0.23622692972421647,
      "num_tokens": 5353485.0,
      "step": 6710
    },
    {
      "entropy": 4.990548849105835,
      "epoch": 1.8985015549901045,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004699858752940916,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2502109229564667,
      "num_tokens": 5357320.0,
      "step": 6715
    },
    {
      "entropy": 5.034406995773315,
      "epoch": 1.8999151823579306,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004699345277993737,
      "loss": 4.9998,
      "mean_token_accuracy": 0.23155584633350373,
      "num_tokens": 5360879.0,
      "step": 6720
    },
    {
      "entropy": 5.010735082626343,
      "epoch": 1.9013288097257561,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004698831395645448,
      "loss": 4.9281,
      "mean_token_accuracy": 0.23807407319545745,
      "num_tokens": 5365029.0,
      "step": 6725
    },
    {
      "entropy": 4.973284578323364,
      "epoch": 1.9027424370935822,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046983171060034493,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23535877764225005,
      "num_tokens": 5369020.0,
      "step": 6730
    },
    {
      "entropy": 5.032124090194702,
      "epoch": 1.904156064461408,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004697802409175223,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2322619989514351,
      "num_tokens": 5373092.0,
      "step": 6735
    },
    {
      "entropy": 5.059900379180908,
      "epoch": 1.9055696918292337,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046972873052683384,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2367260128259659,
      "num_tokens": 5376880.0,
      "step": 6740
    },
    {
      "entropy": 4.9693145751953125,
      "epoch": 1.9069833191970598,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004696771794390449,
      "loss": 4.7942,
      "mean_token_accuracy": 0.2499829038977623,
      "num_tokens": 5380654.0,
      "step": 6745
    },
    {
      "entropy": 4.985115718841553,
      "epoch": 1.9083969465648853,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046962558766492924,
      "loss": 4.8915,
      "mean_token_accuracy": 0.24329203367233276,
      "num_tokens": 5384762.0,
      "step": 6750
    },
    {
      "entropy": 4.931437587738037,
      "epoch": 1.9098105739327114,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004695739552152693,
      "loss": 4.7608,
      "mean_token_accuracy": 0.2593420758843422,
      "num_tokens": 5388503.0,
      "step": 6755
    },
    {
      "entropy": 5.049009084701538,
      "epoch": 1.9112242013005372,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046952228210085596,
      "loss": 4.977,
      "mean_token_accuracy": 0.2207857444882393,
      "num_tokens": 5392279.0,
      "step": 6760
    },
    {
      "entropy": 5.049694204330445,
      "epoch": 1.912637828668363,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046947056833248854,
      "loss": 5.0053,
      "mean_token_accuracy": 0.23272380828857422,
      "num_tokens": 5396326.0,
      "step": 6765
    },
    {
      "entropy": 4.9959447383880615,
      "epoch": 1.914051456036189,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004694188139209748,
      "loss": 4.9325,
      "mean_token_accuracy": 0.24052215963602067,
      "num_tokens": 5400557.0,
      "step": 6770
    },
    {
      "entropy": 5.040218210220337,
      "epoch": 1.9154650834040146,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046936701887713135,
      "loss": 4.9701,
      "mean_token_accuracy": 0.22560301274061204,
      "num_tokens": 5404418.0,
      "step": 6775
    },
    {
      "entropy": 5.018356800079346,
      "epoch": 1.9168787107718406,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004693151832117828,
      "loss": 4.7659,
      "mean_token_accuracy": 0.23729924708604813,
      "num_tokens": 5408467.0,
      "step": 6780
    },
    {
      "entropy": 5.025453424453735,
      "epoch": 1.9182923381396664,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004692633069357626,
      "loss": 4.9186,
      "mean_token_accuracy": 0.24248077869415283,
      "num_tokens": 5412603.0,
      "step": 6785
    },
    {
      "entropy": 4.9906693458557125,
      "epoch": 1.9197059655074922,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004692113900599125,
      "loss": 4.9,
      "mean_token_accuracy": 0.24125048369169236,
      "num_tokens": 5416529.0,
      "step": 6790
    },
    {
      "entropy": 4.910904502868652,
      "epoch": 1.9211195928753182,
      "grad_norm": 1.328125,
      "learning_rate": 0.000469159432595083,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2510335996747017,
      "num_tokens": 5420514.0,
      "step": 6795
    },
    {
      "entropy": 4.911242485046387,
      "epoch": 1.9225332202431438,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004691074345521326,
      "loss": 4.8382,
      "mean_token_accuracy": 0.25361909419298173,
      "num_tokens": 5424539.0,
      "step": 6800
    },
    {
      "entropy": 4.899059152603149,
      "epoch": 1.9239468476109698,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004690553959419286,
      "loss": 4.8753,
      "mean_token_accuracy": 0.24057021886110305,
      "num_tokens": 5428507.0,
      "step": 6805
    },
    {
      "entropy": 4.907997274398804,
      "epoch": 1.9253604749787956,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000469003316775347,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2527383163571358,
      "num_tokens": 5432168.0,
      "step": 6810
    },
    {
      "entropy": 5.089661312103272,
      "epoch": 1.9267741023466214,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046895119706327185,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22847762405872346,
      "num_tokens": 5435848.0,
      "step": 6815
    },
    {
      "entropy": 4.981099081039429,
      "epoch": 1.9281877297144474,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046889903681659583,
      "loss": 4.9164,
      "mean_token_accuracy": 0.24020892530679702,
      "num_tokens": 5439823.0,
      "step": 6820
    },
    {
      "entropy": 4.964100360870361,
      "epoch": 1.929601357082273,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004688468360462202,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2427339568734169,
      "num_tokens": 5443836.0,
      "step": 6825
    },
    {
      "entropy": 4.9801661491394045,
      "epoch": 1.931014984450099,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004687945947630546,
      "loss": 4.8993,
      "mean_token_accuracy": 0.24369361698627473,
      "num_tokens": 5447638.0,
      "step": 6830
    },
    {
      "entropy": 4.9256730556488035,
      "epoch": 1.9324286118179248,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046874231297801695,
      "loss": 4.9463,
      "mean_token_accuracy": 0.23757262229919435,
      "num_tokens": 5451444.0,
      "step": 6835
    },
    {
      "entropy": 5.116860055923462,
      "epoch": 1.9338422391857506,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004686899907020341,
      "loss": 5.1126,
      "mean_token_accuracy": 0.2251540720462799,
      "num_tokens": 5455687.0,
      "step": 6840
    },
    {
      "entropy": 5.078496599197388,
      "epoch": 1.9352558665535766,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004686376279460408,
      "loss": 4.8612,
      "mean_token_accuracy": 0.2301710233092308,
      "num_tokens": 5459406.0,
      "step": 6845
    },
    {
      "entropy": 4.907673597335815,
      "epoch": 1.9366694939214022,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046858522472098074,
      "loss": 4.7409,
      "mean_token_accuracy": 0.2636879548430443,
      "num_tokens": 5463227.0,
      "step": 6850
    },
    {
      "entropy": 5.168267679214478,
      "epoch": 1.9380831212892282,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046853278103780585,
      "loss": 5.2384,
      "mean_token_accuracy": 0.20319467335939406,
      "num_tokens": 5467154.0,
      "step": 6855
    },
    {
      "entropy": 5.073860836029053,
      "epoch": 1.939496748657054,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004684802969074764,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2379152998328209,
      "num_tokens": 5470977.0,
      "step": 6860
    },
    {
      "entropy": 5.026559352874756,
      "epoch": 1.9409103760248798,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004684277723409614,
      "loss": 4.8756,
      "mean_token_accuracy": 0.24274382144212722,
      "num_tokens": 5474824.0,
      "step": 6865
    },
    {
      "entropy": 4.897392511367798,
      "epoch": 1.9423240033927058,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000468375207349238,
      "loss": 4.8297,
      "mean_token_accuracy": 0.25318351686000823,
      "num_tokens": 5478873.0,
      "step": 6870
    },
    {
      "entropy": 4.973334932327271,
      "epoch": 1.9437376307605314,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004683226019432922,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23013330698013307,
      "num_tokens": 5483208.0,
      "step": 6875
    },
    {
      "entropy": 4.966379976272583,
      "epoch": 1.9451512581283574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046826995613411806,
      "loss": 4.8379,
      "mean_token_accuracy": 0.22892674803733826,
      "num_tokens": 5487387.0,
      "step": 6880
    },
    {
      "entropy": 4.902877759933472,
      "epoch": 1.9465648854961832,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004682172699327182,
      "loss": 4.7836,
      "mean_token_accuracy": 0.24981795847415925,
      "num_tokens": 5491486.0,
      "step": 6885
    },
    {
      "entropy": 4.953005647659301,
      "epoch": 1.947978512864009,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046816454335010365,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22954194843769074,
      "num_tokens": 5495333.0,
      "step": 6890
    },
    {
      "entropy": 5.004650640487671,
      "epoch": 1.949392140231835,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046811177639729417,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22773503810167312,
      "num_tokens": 5499351.0,
      "step": 6895
    },
    {
      "entropy": 4.9662353038787845,
      "epoch": 1.9508057675996606,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046805896908531756,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23888093382120132,
      "num_tokens": 5503260.0,
      "step": 6900
    },
    {
      "entropy": 5.0105287551879885,
      "epoch": 1.9522193949674866,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004680061214252102,
      "loss": 4.8752,
      "mean_token_accuracy": 0.23643625974655152,
      "num_tokens": 5507223.0,
      "step": 6905
    },
    {
      "entropy": 5.025991296768188,
      "epoch": 1.9536330223353124,
      "grad_norm": 1.21875,
      "learning_rate": 0.000467953233428017,
      "loss": 4.8749,
      "mean_token_accuracy": 0.23839300572872163,
      "num_tokens": 5511347.0,
      "step": 6910
    },
    {
      "entropy": 5.003885889053345,
      "epoch": 1.9550466497031382,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004679003051047912,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22756122797727585,
      "num_tokens": 5515350.0,
      "step": 6915
    },
    {
      "entropy": 4.966977262496949,
      "epoch": 1.9564602770709643,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004678473364665945,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2389771893620491,
      "num_tokens": 5519262.0,
      "step": 6920
    },
    {
      "entropy": 4.950049114227295,
      "epoch": 1.9578739044387898,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046779432752449705,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2323794275522232,
      "num_tokens": 5523209.0,
      "step": 6925
    },
    {
      "entropy": 4.8986893653869625,
      "epoch": 1.9592875318066159,
      "grad_norm": 1.375,
      "learning_rate": 0.0004677412782895773,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2408262237906456,
      "num_tokens": 5526903.0,
      "step": 6930
    },
    {
      "entropy": 4.946874666213989,
      "epoch": 1.9607011591744417,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004676881887729222,
      "loss": 4.8062,
      "mean_token_accuracy": 0.24547500014305115,
      "num_tokens": 5530667.0,
      "step": 6935
    },
    {
      "entropy": 4.95073618888855,
      "epoch": 1.9621147865422675,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046763505898562735,
      "loss": 4.8493,
      "mean_token_accuracy": 0.23796316385269164,
      "num_tokens": 5534413.0,
      "step": 6940
    },
    {
      "entropy": 4.934475612640381,
      "epoch": 1.9635284139100933,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046758188893879633,
      "loss": 4.9296,
      "mean_token_accuracy": 0.24794361740350723,
      "num_tokens": 5538662.0,
      "step": 6945
    },
    {
      "entropy": 4.92299222946167,
      "epoch": 1.964942041277919,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046752867864354136,
      "loss": 4.7564,
      "mean_token_accuracy": 0.2606536105275154,
      "num_tokens": 5542602.0,
      "step": 6950
    },
    {
      "entropy": 5.0195588111877445,
      "epoch": 1.966355668645745,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046747542811098315,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22639136612415314,
      "num_tokens": 5546244.0,
      "step": 6955
    },
    {
      "entropy": 4.949305582046509,
      "epoch": 1.9677692960135709,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004674221373522507,
      "loss": 4.9219,
      "mean_token_accuracy": 0.24098569601774217,
      "num_tokens": 5550116.0,
      "step": 6960
    },
    {
      "entropy": 4.978331708908081,
      "epoch": 1.9691829233813967,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004673688063784815,
      "loss": 4.8621,
      "mean_token_accuracy": 0.24865348935127257,
      "num_tokens": 5554116.0,
      "step": 6965
    },
    {
      "entropy": 4.9058756828308105,
      "epoch": 1.9705965507492225,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046731543520082123,
      "loss": 4.7636,
      "mean_token_accuracy": 0.2541049674153328,
      "num_tokens": 5557911.0,
      "step": 6970
    },
    {
      "entropy": 4.936982345581055,
      "epoch": 1.9720101781170483,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004672620238304243,
      "loss": 4.854,
      "mean_token_accuracy": 0.24166148751974106,
      "num_tokens": 5561828.0,
      "step": 6975
    },
    {
      "entropy": 5.016292667388916,
      "epoch": 1.9734238054848743,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046720857227845324,
      "loss": 4.8292,
      "mean_token_accuracy": 0.24934437721967698,
      "num_tokens": 5565976.0,
      "step": 6980
    },
    {
      "entropy": 4.888951396942138,
      "epoch": 1.9748374328527,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046715508055607925,
      "loss": 4.7385,
      "mean_token_accuracy": 0.2442086562514305,
      "num_tokens": 5569654.0,
      "step": 6985
    },
    {
      "entropy": 4.873531723022461,
      "epoch": 1.9762510602205259,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004671015486744817,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2353682830929756,
      "num_tokens": 5573440.0,
      "step": 6990
    },
    {
      "entropy": 4.976876735687256,
      "epoch": 1.9776646875883517,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004670479766448483,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23563070297241212,
      "num_tokens": 5577667.0,
      "step": 6995
    },
    {
      "entropy": 4.956812238693237,
      "epoch": 1.9790783149561775,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004669943644783754,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23592994213104249,
      "num_tokens": 5581520.0,
      "step": 7000
    },
    {
      "epoch": 1.9790783149561775,
      "eval_entropy": 4.767377365404287,
      "eval_loss": 5.132265567779541,
      "eval_mean_token_accuracy": 0.23601013140569418,
      "eval_num_tokens": 5581520.0,
      "eval_runtime": 4.7094,
      "eval_samples_per_second": 2396.079,
      "eval_steps_per_second": 299.616,
      "step": 7000
    },
    {
      "entropy": 4.901002597808838,
      "epoch": 1.9804919423240035,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004669407121862676,
      "loss": 4.7419,
      "mean_token_accuracy": 0.2551999732851982,
      "num_tokens": 5585338.0,
      "step": 7005
    },
    {
      "entropy": 5.073434209823608,
      "epoch": 1.981905569691829,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046688701977973783,
      "loss": 4.9474,
      "mean_token_accuracy": 0.230147847533226,
      "num_tokens": 5589332.0,
      "step": 7010
    },
    {
      "entropy": 4.958309841156006,
      "epoch": 1.983319197059655,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004668332872700075,
      "loss": 4.8278,
      "mean_token_accuracy": 0.24610604643821715,
      "num_tokens": 5593048.0,
      "step": 7015
    },
    {
      "entropy": 4.970778369903565,
      "epoch": 1.984732824427481,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004667795146683065,
      "loss": 4.8884,
      "mean_token_accuracy": 0.24478642791509628,
      "num_tokens": 5596984.0,
      "step": 7020
    },
    {
      "entropy": 4.996025943756104,
      "epoch": 1.9861464517953067,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046672570198587283,
      "loss": 5.0322,
      "mean_token_accuracy": 0.22356994897127153,
      "num_tokens": 5600661.0,
      "step": 7025
    },
    {
      "entropy": 4.920639181137085,
      "epoch": 1.9875600791631327,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046667184923395306,
      "loss": 4.808,
      "mean_token_accuracy": 0.2435439333319664,
      "num_tokens": 5604479.0,
      "step": 7030
    },
    {
      "entropy": 5.0560688972473145,
      "epoch": 1.9889737065309583,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000466617956423802,
      "loss": 4.8995,
      "mean_token_accuracy": 0.24173268526792527,
      "num_tokens": 5608603.0,
      "step": 7035
    },
    {
      "entropy": 4.920607423782348,
      "epoch": 1.9903873338987843,
      "grad_norm": 1.375,
      "learning_rate": 0.000466564023566683,
      "loss": 4.7982,
      "mean_token_accuracy": 0.2412361368536949,
      "num_tokens": 5612579.0,
      "step": 7040
    },
    {
      "entropy": 5.064804601669311,
      "epoch": 1.99180096126661,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046651005067386767,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2266793057322502,
      "num_tokens": 5616768.0,
      "step": 7045
    },
    {
      "entropy": 4.953020429611206,
      "epoch": 1.993214588634436,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000466456037756636,
      "loss": 4.8097,
      "mean_token_accuracy": 0.250866961479187,
      "num_tokens": 5620574.0,
      "step": 7050
    },
    {
      "entropy": 5.034789323806763,
      "epoch": 1.994628216002262,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004664019848262764,
      "loss": 5.0041,
      "mean_token_accuracy": 0.23023169338703156,
      "num_tokens": 5624740.0,
      "step": 7055
    },
    {
      "entropy": 4.867894458770752,
      "epoch": 1.9960418433700875,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046634789189408554,
      "loss": 4.9008,
      "mean_token_accuracy": 0.24058068543672562,
      "num_tokens": 5628361.0,
      "step": 7060
    },
    {
      "entropy": 4.9638293266296385,
      "epoch": 1.9974554707379135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046629375897136856,
      "loss": 4.8395,
      "mean_token_accuracy": 0.24968526214361192,
      "num_tokens": 5632431.0,
      "step": 7065
    },
    {
      "entropy": 5.092235136032104,
      "epoch": 1.9988690981057393,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004662395860694388,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2363850876688957,
      "num_tokens": 5636495.0,
      "step": 7070
    },
    {
      "entropy": 5.16230320930481,
      "epoch": 2.000282725473565,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004661853731996182,
      "loss": 5.0036,
      "mean_token_accuracy": 0.21839372664690018,
      "num_tokens": 5640304.0,
      "step": 7075
    },
    {
      "entropy": 5.003241157531738,
      "epoch": 2.001696352841391,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046613112037323693,
      "loss": 4.5851,
      "mean_token_accuracy": 0.2621611520648003,
      "num_tokens": 5644231.0,
      "step": 7080
    },
    {
      "entropy": 4.829736518859863,
      "epoch": 2.0031099802092167,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046607682760163333,
      "loss": 4.4944,
      "mean_token_accuracy": 0.2620402738451958,
      "num_tokens": 5648266.0,
      "step": 7085
    },
    {
      "entropy": 4.9413769245147705,
      "epoch": 2.0045236075770427,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004660224948961544,
      "loss": 4.6666,
      "mean_token_accuracy": 0.23823485225439073,
      "num_tokens": 5651856.0,
      "step": 7090
    },
    {
      "entropy": 4.805984306335449,
      "epoch": 2.0059372349448688,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046596812226815516,
      "loss": 4.5125,
      "mean_token_accuracy": 0.2655372381210327,
      "num_tokens": 5655556.0,
      "step": 7095
    },
    {
      "entropy": 4.917257690429688,
      "epoch": 2.0073508623126943,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004659137097289994,
      "loss": 4.6329,
      "mean_token_accuracy": 0.2510543391108513,
      "num_tokens": 5659647.0,
      "step": 7100
    },
    {
      "entropy": 4.967804050445556,
      "epoch": 2.0087644896805203,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046585925729005883,
      "loss": 4.6625,
      "mean_token_accuracy": 0.24263476878404616,
      "num_tokens": 5663240.0,
      "step": 7105
    },
    {
      "entropy": 4.920840835571289,
      "epoch": 2.010178117048346,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004658047649627138,
      "loss": 4.6004,
      "mean_token_accuracy": 0.2559704214334488,
      "num_tokens": 5667333.0,
      "step": 7110
    },
    {
      "entropy": 4.978841495513916,
      "epoch": 2.011591744416172,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046575023275835276,
      "loss": 4.644,
      "mean_token_accuracy": 0.24520733952522278,
      "num_tokens": 5671261.0,
      "step": 7115
    },
    {
      "entropy": 4.885233640670776,
      "epoch": 2.0130053717839975,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046569566068837263,
      "loss": 4.6266,
      "mean_token_accuracy": 0.2514878287911415,
      "num_tokens": 5675454.0,
      "step": 7120
    },
    {
      "entropy": 4.9167218685150145,
      "epoch": 2.0144189991518235,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004656410487641785,
      "loss": 4.6515,
      "mean_token_accuracy": 0.24912064224481584,
      "num_tokens": 5679490.0,
      "step": 7125
    },
    {
      "entropy": 4.987251091003418,
      "epoch": 2.0158326265196496,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004655863969971843,
      "loss": 4.6758,
      "mean_token_accuracy": 0.25598817616701125,
      "num_tokens": 5683574.0,
      "step": 7130
    },
    {
      "entropy": 4.882364749908447,
      "epoch": 2.017246253887475,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046553170539881156,
      "loss": 4.6127,
      "mean_token_accuracy": 0.252235546708107,
      "num_tokens": 5687542.0,
      "step": 7135
    },
    {
      "entropy": 4.8574542045593265,
      "epoch": 2.018659881255301,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004654769739804906,
      "loss": 4.568,
      "mean_token_accuracy": 0.26174445897340776,
      "num_tokens": 5691202.0,
      "step": 7140
    },
    {
      "entropy": 4.857198810577392,
      "epoch": 2.0200735086231267,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046542220275366,
      "loss": 4.6551,
      "mean_token_accuracy": 0.25430774986743926,
      "num_tokens": 5695476.0,
      "step": 7145
    },
    {
      "entropy": 4.832617855072021,
      "epoch": 2.0214871359909528,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046536739172976647,
      "loss": 4.6447,
      "mean_token_accuracy": 0.26446863561868666,
      "num_tokens": 5699530.0,
      "step": 7150
    },
    {
      "entropy": 4.904586696624756,
      "epoch": 2.0229007633587788,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046531254092026535,
      "loss": 4.555,
      "mean_token_accuracy": 0.26620845049619674,
      "num_tokens": 5703573.0,
      "step": 7155
    },
    {
      "entropy": 4.869217205047607,
      "epoch": 2.0243143907266044,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046525765033662007,
      "loss": 4.6218,
      "mean_token_accuracy": 0.2523535996675491,
      "num_tokens": 5707953.0,
      "step": 7160
    },
    {
      "entropy": 4.927516651153565,
      "epoch": 2.0257280180944304,
      "grad_norm": 1.375,
      "learning_rate": 0.00046520271999030225,
      "loss": 4.65,
      "mean_token_accuracy": 0.2607363760471344,
      "num_tokens": 5712018.0,
      "step": 7165
    },
    {
      "entropy": 4.895551061630249,
      "epoch": 2.027141645462256,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004651477498927922,
      "loss": 4.6392,
      "mean_token_accuracy": 0.25268771052360534,
      "num_tokens": 5715785.0,
      "step": 7170
    },
    {
      "entropy": 4.8994299411773685,
      "epoch": 2.028555272830082,
      "grad_norm": 1.375,
      "learning_rate": 0.0004650927400555783,
      "loss": 4.5722,
      "mean_token_accuracy": 0.2569097697734833,
      "num_tokens": 5719699.0,
      "step": 7175
    },
    {
      "entropy": 4.895987844467163,
      "epoch": 2.029968900197908,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046503769049015714,
      "loss": 4.592,
      "mean_token_accuracy": 0.2501364812254906,
      "num_tokens": 5723594.0,
      "step": 7180
    },
    {
      "entropy": 4.882723236083985,
      "epoch": 2.0313825275657336,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046498260120803395,
      "loss": 4.5839,
      "mean_token_accuracy": 0.24766190350055695,
      "num_tokens": 5727474.0,
      "step": 7185
    },
    {
      "entropy": 4.897294902801514,
      "epoch": 2.0327961549335596,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046492747222072185,
      "loss": 4.5332,
      "mean_token_accuracy": 0.2637475818395615,
      "num_tokens": 5731264.0,
      "step": 7190
    },
    {
      "entropy": 4.846499729156494,
      "epoch": 2.034209782301385,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004648723035397425,
      "loss": 4.6754,
      "mean_token_accuracy": 0.25438218712806704,
      "num_tokens": 5735065.0,
      "step": 7195
    },
    {
      "entropy": 4.860324096679688,
      "epoch": 2.035623409669211,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004648170951766259,
      "loss": 4.633,
      "mean_token_accuracy": 0.24662139415740966,
      "num_tokens": 5738974.0,
      "step": 7200
    },
    {
      "entropy": 4.829107809066772,
      "epoch": 2.037037037037037,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004647618471429102,
      "loss": 4.5075,
      "mean_token_accuracy": 0.2675254315137863,
      "num_tokens": 5742782.0,
      "step": 7205
    },
    {
      "entropy": 4.945222187042236,
      "epoch": 2.0384506644048628,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004647065594501418,
      "loss": 4.6442,
      "mean_token_accuracy": 0.23724554628133773,
      "num_tokens": 5746991.0,
      "step": 7210
    },
    {
      "entropy": 4.918517351150513,
      "epoch": 2.039864291772689,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046465123210987576,
      "loss": 4.6706,
      "mean_token_accuracy": 0.24840682595968247,
      "num_tokens": 5750955.0,
      "step": 7215
    },
    {
      "entropy": 4.8616166591644285,
      "epoch": 2.0412779191405144,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046459586513367487,
      "loss": 4.5365,
      "mean_token_accuracy": 0.2655811071395874,
      "num_tokens": 5754809.0,
      "step": 7220
    },
    {
      "entropy": 4.882145595550537,
      "epoch": 2.0426915465083404,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004645404585331106,
      "loss": 4.6566,
      "mean_token_accuracy": 0.252901965379715,
      "num_tokens": 5758658.0,
      "step": 7225
    },
    {
      "entropy": 4.906905746459961,
      "epoch": 2.0441051738761664,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004644850123197626,
      "loss": 4.674,
      "mean_token_accuracy": 0.24458914548158645,
      "num_tokens": 5762965.0,
      "step": 7230
    },
    {
      "entropy": 4.901117372512817,
      "epoch": 2.045518801243992,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004644295265052187,
      "loss": 4.6674,
      "mean_token_accuracy": 0.24658581167459487,
      "num_tokens": 5767204.0,
      "step": 7235
    },
    {
      "entropy": 5.0585778713226315,
      "epoch": 2.046932428611818,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046437400110107527,
      "loss": 4.6811,
      "mean_token_accuracy": 0.24671834260225295,
      "num_tokens": 5771239.0,
      "step": 7240
    },
    {
      "entropy": 4.897186040878296,
      "epoch": 2.0483460559796436,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004643184361189366,
      "loss": 4.6712,
      "mean_token_accuracy": 0.244027179479599,
      "num_tokens": 5775345.0,
      "step": 7245
    },
    {
      "entropy": 4.913565063476563,
      "epoch": 2.0497596833474696,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004642628315704155,
      "loss": 4.5849,
      "mean_token_accuracy": 0.2582767128944397,
      "num_tokens": 5779294.0,
      "step": 7250
    },
    {
      "entropy": 4.856963014602661,
      "epoch": 2.0511733107152956,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046420718746713297,
      "loss": 4.5711,
      "mean_token_accuracy": 0.25621732473373415,
      "num_tokens": 5783362.0,
      "step": 7255
    },
    {
      "entropy": 4.880762100219727,
      "epoch": 2.052586938083121,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004641515038207182,
      "loss": 4.5629,
      "mean_token_accuracy": 0.26701031178236007,
      "num_tokens": 5787134.0,
      "step": 7260
    },
    {
      "entropy": 4.816731643676758,
      "epoch": 2.0540005654509472,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004640957806428089,
      "loss": 4.6273,
      "mean_token_accuracy": 0.25541260093450546,
      "num_tokens": 5791116.0,
      "step": 7265
    },
    {
      "entropy": 4.87394642829895,
      "epoch": 2.055414192818773,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004640400179450508,
      "loss": 4.6341,
      "mean_token_accuracy": 0.2509767562150955,
      "num_tokens": 5795292.0,
      "step": 7270
    },
    {
      "entropy": 4.934666442871094,
      "epoch": 2.056827820186599,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004639842157390979,
      "loss": 4.5507,
      "mean_token_accuracy": 0.26945238262414933,
      "num_tokens": 5799138.0,
      "step": 7275
    },
    {
      "entropy": 4.850317668914795,
      "epoch": 2.058241447554425,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046392837403661245,
      "loss": 4.6571,
      "mean_token_accuracy": 0.2486834242939949,
      "num_tokens": 5803289.0,
      "step": 7280
    },
    {
      "entropy": 4.839475774765015,
      "epoch": 2.0596550749222504,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046387249284926535,
      "loss": 4.5836,
      "mean_token_accuracy": 0.2561303049325943,
      "num_tokens": 5807286.0,
      "step": 7285
    },
    {
      "entropy": 4.829013299942017,
      "epoch": 2.0610687022900764,
      "grad_norm": 1.34375,
      "learning_rate": 0.000463816572188735,
      "loss": 4.6806,
      "mean_token_accuracy": 0.24840794503688812,
      "num_tokens": 5811170.0,
      "step": 7290
    },
    {
      "entropy": 4.8133909702301025,
      "epoch": 2.062482329657902,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004637606120667087,
      "loss": 4.417,
      "mean_token_accuracy": 0.26526159793138504,
      "num_tokens": 5815059.0,
      "step": 7295
    },
    {
      "entropy": 5.002345991134644,
      "epoch": 2.063895957025728,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004637046124948818,
      "loss": 4.7239,
      "mean_token_accuracy": 0.2513126119971275,
      "num_tokens": 5819109.0,
      "step": 7300
    },
    {
      "entropy": 4.866905641555786,
      "epoch": 2.065309584393554,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004636485734849578,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24989327788352966,
      "num_tokens": 5823249.0,
      "step": 7305
    },
    {
      "entropy": 4.887831783294677,
      "epoch": 2.0667232117613796,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004635924950486485,
      "loss": 4.6948,
      "mean_token_accuracy": 0.24162260293960572,
      "num_tokens": 5826936.0,
      "step": 7310
    },
    {
      "entropy": 5.01615629196167,
      "epoch": 2.0681368391292057,
      "grad_norm": 1.359375,
      "learning_rate": 0.000463536377197674,
      "loss": 4.7569,
      "mean_token_accuracy": 0.2407267928123474,
      "num_tokens": 5830884.0,
      "step": 7315
    },
    {
      "entropy": 4.982859182357788,
      "epoch": 2.0695504664970312,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004634802199437625,
      "loss": 4.5999,
      "mean_token_accuracy": 0.2500802740454674,
      "num_tokens": 5834741.0,
      "step": 7320
    },
    {
      "entropy": 4.918830299377442,
      "epoch": 2.0709640938648572,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046342402329865054,
      "loss": 4.6368,
      "mean_token_accuracy": 0.2528371557593346,
      "num_tokens": 5838592.0,
      "step": 7325
    },
    {
      "entropy": 4.861496782302856,
      "epoch": 2.0723777212326833,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046336778727408285,
      "loss": 4.6737,
      "mean_token_accuracy": 0.24914259910583497,
      "num_tokens": 5842745.0,
      "step": 7330
    },
    {
      "entropy": 4.924810123443604,
      "epoch": 2.073791348600509,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004633115118818125,
      "loss": 4.696,
      "mean_token_accuracy": 0.2402736634016037,
      "num_tokens": 5846611.0,
      "step": 7335
    },
    {
      "entropy": 4.939493417739868,
      "epoch": 2.075204975968335,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046325519713360066,
      "loss": 4.6703,
      "mean_token_accuracy": 0.2567262023687363,
      "num_tokens": 5850982.0,
      "step": 7340
    },
    {
      "entropy": 4.930327558517456,
      "epoch": 2.0766186033361604,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046319884304121676,
      "loss": 4.6361,
      "mean_token_accuracy": 0.24650452882051468,
      "num_tokens": 5855034.0,
      "step": 7345
    },
    {
      "entropy": 4.945933437347412,
      "epoch": 2.0780322307039865,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004631424496164384,
      "loss": 4.5748,
      "mean_token_accuracy": 0.26548305451869963,
      "num_tokens": 5858912.0,
      "step": 7350
    },
    {
      "entropy": 4.908020830154419,
      "epoch": 2.0794458580718125,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046308601687105157,
      "loss": 4.7057,
      "mean_token_accuracy": 0.2555872619152069,
      "num_tokens": 5862721.0,
      "step": 7355
    },
    {
      "entropy": 4.904927158355713,
      "epoch": 2.080859485439638,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004630295448168502,
      "loss": 4.6417,
      "mean_token_accuracy": 0.2502488538622856,
      "num_tokens": 5866767.0,
      "step": 7360
    },
    {
      "entropy": 5.001476383209228,
      "epoch": 2.082273112807464,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046297303346563677,
      "loss": 4.7973,
      "mean_token_accuracy": 0.2270250216126442,
      "num_tokens": 5870733.0,
      "step": 7365
    },
    {
      "entropy": 4.907336854934693,
      "epoch": 2.0836867401752897,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046291648282922176,
      "loss": 4.6324,
      "mean_token_accuracy": 0.24673094600439072,
      "num_tokens": 5874571.0,
      "step": 7370
    },
    {
      "entropy": 4.86214542388916,
      "epoch": 2.0851003675431157,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004628598929194239,
      "loss": 4.5392,
      "mean_token_accuracy": 0.2621063247323036,
      "num_tokens": 5878562.0,
      "step": 7375
    },
    {
      "entropy": 4.889634466171264,
      "epoch": 2.0865139949109412,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046280326374807006,
      "loss": 4.6065,
      "mean_token_accuracy": 0.2500217393040657,
      "num_tokens": 5882601.0,
      "step": 7380
    },
    {
      "entropy": 4.8090293407440186,
      "epoch": 2.0879276222787673,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004627465953269955,
      "loss": 4.6163,
      "mean_token_accuracy": 0.24665324985980988,
      "num_tokens": 5886647.0,
      "step": 7385
    },
    {
      "entropy": 4.907321310043335,
      "epoch": 2.0893412496465933,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004626898876680434,
      "loss": 4.638,
      "mean_token_accuracy": 0.24985187649726867,
      "num_tokens": 5890985.0,
      "step": 7390
    },
    {
      "entropy": 4.986405944824218,
      "epoch": 2.090754877014419,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004626331407830655,
      "loss": 4.7798,
      "mean_token_accuracy": 0.2356831759214401,
      "num_tokens": 5894735.0,
      "step": 7395
    },
    {
      "entropy": 4.8798459529876705,
      "epoch": 2.092168504382245,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004625763546839214,
      "loss": 4.5416,
      "mean_token_accuracy": 0.256428436934948,
      "num_tokens": 5898955.0,
      "step": 7400
    },
    {
      "entropy": 4.913504123687744,
      "epoch": 2.093582131750071,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046251952938247915,
      "loss": 4.6012,
      "mean_token_accuracy": 0.25780409276485444,
      "num_tokens": 5902875.0,
      "step": 7405
    },
    {
      "entropy": 4.874858427047729,
      "epoch": 2.0949957591178965,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046246266489061486,
      "loss": 4.6642,
      "mean_token_accuracy": 0.25740828514099123,
      "num_tokens": 5906801.0,
      "step": 7410
    },
    {
      "entropy": 4.871133184432983,
      "epoch": 2.0964093864857225,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004624057612202128,
      "loss": 4.6896,
      "mean_token_accuracy": 0.2517325282096863,
      "num_tokens": 5910930.0,
      "step": 7415
    },
    {
      "entropy": 4.960059595108032,
      "epoch": 2.097823013853548,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046234881838316556,
      "loss": 4.6403,
      "mean_token_accuracy": 0.25102746933698655,
      "num_tokens": 5914963.0,
      "step": 7420
    },
    {
      "entropy": 4.878540945053101,
      "epoch": 2.099236641221374,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046229183639137384,
      "loss": 4.7274,
      "mean_token_accuracy": 0.24386123418807984,
      "num_tokens": 5918919.0,
      "step": 7425
    },
    {
      "entropy": 4.903829050064087,
      "epoch": 2.1006502685891997,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046223481525674645,
      "loss": 4.682,
      "mean_token_accuracy": 0.24366087913513185,
      "num_tokens": 5922914.0,
      "step": 7430
    },
    {
      "entropy": 4.946233797073364,
      "epoch": 2.1020638959570257,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046217775499120045,
      "loss": 4.7544,
      "mean_token_accuracy": 0.23950617164373397,
      "num_tokens": 5926737.0,
      "step": 7435
    },
    {
      "entropy": 4.818638706207276,
      "epoch": 2.1034775233248517,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046212065560666125,
      "loss": 4.6594,
      "mean_token_accuracy": 0.26517554074525834,
      "num_tokens": 5930979.0,
      "step": 7440
    },
    {
      "entropy": 5.087787771224976,
      "epoch": 2.1048911506926773,
      "grad_norm": 1.328125,
      "learning_rate": 0.000462063517115062,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22724665254354476,
      "num_tokens": 5935325.0,
      "step": 7445
    },
    {
      "entropy": 4.98613018989563,
      "epoch": 2.1063047780605033,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004620063395283445,
      "loss": 4.6175,
      "mean_token_accuracy": 0.2500539183616638,
      "num_tokens": 5939445.0,
      "step": 7450
    },
    {
      "entropy": 4.984401750564575,
      "epoch": 2.107718405428329,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046194912285845845,
      "loss": 4.7166,
      "mean_token_accuracy": 0.24352382123470306,
      "num_tokens": 5943125.0,
      "step": 7455
    },
    {
      "entropy": 4.862561130523682,
      "epoch": 2.109132032796155,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004618918671173618,
      "loss": 4.6416,
      "mean_token_accuracy": 0.2500259533524513,
      "num_tokens": 5947342.0,
      "step": 7460
    },
    {
      "entropy": 4.797551965713501,
      "epoch": 2.110545660163981,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046183457231702065,
      "loss": 4.5901,
      "mean_token_accuracy": 0.26116833537817,
      "num_tokens": 5951269.0,
      "step": 7465
    },
    {
      "entropy": 4.8322913646698,
      "epoch": 2.1119592875318065,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004617772384694092,
      "loss": 4.5355,
      "mean_token_accuracy": 0.2675459191203117,
      "num_tokens": 5955280.0,
      "step": 7470
    },
    {
      "entropy": 4.839614391326904,
      "epoch": 2.1133729148996325,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004617198655865099,
      "loss": 4.6188,
      "mean_token_accuracy": 0.25303735435009,
      "num_tokens": 5959662.0,
      "step": 7475
    },
    {
      "entropy": 4.875168561935425,
      "epoch": 2.114786542267458,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046166245368031337,
      "loss": 4.5482,
      "mean_token_accuracy": 0.26120825558900834,
      "num_tokens": 5963471.0,
      "step": 7480
    },
    {
      "entropy": 4.982862281799316,
      "epoch": 2.116200169635284,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046160500276281826,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24568137377500535,
      "num_tokens": 5967678.0,
      "step": 7485
    },
    {
      "entropy": 4.996191549301147,
      "epoch": 2.11761379700311,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004615475128460315,
      "loss": 4.7235,
      "mean_token_accuracy": 0.24324966222047806,
      "num_tokens": 5971900.0,
      "step": 7490
    },
    {
      "entropy": 5.007793664932251,
      "epoch": 2.1190274243709357,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046148998394196825,
      "loss": 4.6855,
      "mean_token_accuracy": 0.2398323342204094,
      "num_tokens": 5975922.0,
      "step": 7495
    },
    {
      "entropy": 4.890924596786499,
      "epoch": 2.1204410517387617,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004614324160626514,
      "loss": 4.6007,
      "mean_token_accuracy": 0.2643897011876106,
      "num_tokens": 5979791.0,
      "step": 7500
    },
    {
      "entropy": 4.884063482284546,
      "epoch": 2.1218546791065873,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004613748092201126,
      "loss": 4.6105,
      "mean_token_accuracy": 0.2667216271162033,
      "num_tokens": 5983698.0,
      "step": 7505
    },
    {
      "entropy": 4.976096963882446,
      "epoch": 2.1232683064744133,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004613171634263911,
      "loss": 4.7214,
      "mean_token_accuracy": 0.24156821370124817,
      "num_tokens": 5987670.0,
      "step": 7510
    },
    {
      "entropy": 4.9562479019165036,
      "epoch": 2.1246819338422394,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004612594786935346,
      "loss": 4.6885,
      "mean_token_accuracy": 0.24046777039766312,
      "num_tokens": 5991592.0,
      "step": 7515
    },
    {
      "entropy": 4.829369258880615,
      "epoch": 2.126095561210065,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046120175503359885,
      "loss": 4.6268,
      "mean_token_accuracy": 0.25349479764699934,
      "num_tokens": 5995743.0,
      "step": 7520
    },
    {
      "entropy": 4.858245086669922,
      "epoch": 2.127509188577891,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004611439924586477,
      "loss": 4.6005,
      "mean_token_accuracy": 0.25550036430358886,
      "num_tokens": 5999789.0,
      "step": 7525
    },
    {
      "entropy": 4.8302079200744625,
      "epoch": 2.1289228159457165,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004610861909807533,
      "loss": 4.5362,
      "mean_token_accuracy": 0.2647275298833847,
      "num_tokens": 6003479.0,
      "step": 7530
    },
    {
      "entropy": 4.895051383972168,
      "epoch": 2.1303364433135425,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046102835061199565,
      "loss": 4.6209,
      "mean_token_accuracy": 0.24479036629199982,
      "num_tokens": 6007396.0,
      "step": 7535
    },
    {
      "entropy": 4.861796522140503,
      "epoch": 2.1317500706813686,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046097047136446314,
      "loss": 4.6343,
      "mean_token_accuracy": 0.2411048859357834,
      "num_tokens": 6011164.0,
      "step": 7540
    },
    {
      "entropy": 4.8608301162719725,
      "epoch": 2.133163698049194,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046091255325025204,
      "loss": 4.6811,
      "mean_token_accuracy": 0.24455875009298325,
      "num_tokens": 6015359.0,
      "step": 7545
    },
    {
      "entropy": 4.92618088722229,
      "epoch": 2.13457732541702,
      "grad_norm": 1.390625,
      "learning_rate": 0.000460854596281467,
      "loss": 4.7273,
      "mean_token_accuracy": 0.24114064127206802,
      "num_tokens": 6019548.0,
      "step": 7550
    },
    {
      "entropy": 4.986301612854004,
      "epoch": 2.1359909527848457,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004607966004702206,
      "loss": 4.7109,
      "mean_token_accuracy": 0.24620176255702972,
      "num_tokens": 6023368.0,
      "step": 7555
    },
    {
      "entropy": 4.94984188079834,
      "epoch": 2.1374045801526718,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046073856582863365,
      "loss": 4.6151,
      "mean_token_accuracy": 0.25271814614534377,
      "num_tokens": 6026917.0,
      "step": 7560
    },
    {
      "entropy": 4.945967531204223,
      "epoch": 2.138818207520498,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004606804923688351,
      "loss": 4.6978,
      "mean_token_accuracy": 0.2511825397610664,
      "num_tokens": 6031241.0,
      "step": 7565
    },
    {
      "entropy": 4.866176795959473,
      "epoch": 2.1402318348883234,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046062238010296174,
      "loss": 4.6384,
      "mean_token_accuracy": 0.24873545318841933,
      "num_tokens": 6035295.0,
      "step": 7570
    },
    {
      "entropy": 4.811161851882934,
      "epoch": 2.1416454622561494,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004605642290431588,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24283711314201356,
      "num_tokens": 6039344.0,
      "step": 7575
    },
    {
      "entropy": 4.899979782104492,
      "epoch": 2.143059089623975,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046050603920157955,
      "loss": 4.6782,
      "mean_token_accuracy": 0.2568929836153984,
      "num_tokens": 6043369.0,
      "step": 7580
    },
    {
      "entropy": 4.93179669380188,
      "epoch": 2.144472716991801,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004604478105903852,
      "loss": 4.6235,
      "mean_token_accuracy": 0.2545727267861366,
      "num_tokens": 6047265.0,
      "step": 7585
    },
    {
      "entropy": 4.896879816055298,
      "epoch": 2.145886344359627,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046038954322174527,
      "loss": 4.7172,
      "mean_token_accuracy": 0.24752796739339827,
      "num_tokens": 6051224.0,
      "step": 7590
    },
    {
      "entropy": 4.90925612449646,
      "epoch": 2.1472999717274526,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046033123710783713,
      "loss": 4.6296,
      "mean_token_accuracy": 0.24693907648324967,
      "num_tokens": 6055139.0,
      "step": 7595
    },
    {
      "entropy": 4.873336839675903,
      "epoch": 2.1487135990952786,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046027289226084655,
      "loss": 4.6208,
      "mean_token_accuracy": 0.24568787217140198,
      "num_tokens": 6059264.0,
      "step": 7600
    },
    {
      "entropy": 4.961430597305298,
      "epoch": 2.150127226463104,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046021450869296717,
      "loss": 4.7556,
      "mean_token_accuracy": 0.2350273385643959,
      "num_tokens": 6062886.0,
      "step": 7605
    },
    {
      "entropy": 4.920861291885376,
      "epoch": 2.15154085383093,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046015608641640076,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24469216167926788,
      "num_tokens": 6066856.0,
      "step": 7610
    },
    {
      "entropy": 4.863957500457763,
      "epoch": 2.152954481198756,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046009762544335734,
      "loss": 4.6594,
      "mean_token_accuracy": 0.24695113599300383,
      "num_tokens": 6070702.0,
      "step": 7615
    },
    {
      "entropy": 4.892665529251099,
      "epoch": 2.154368108566582,
      "grad_norm": 1.375,
      "learning_rate": 0.0004600391257860547,
      "loss": 4.6675,
      "mean_token_accuracy": 0.25194244682788847,
      "num_tokens": 6074867.0,
      "step": 7620
    },
    {
      "entropy": 4.8159223079681395,
      "epoch": 2.155781735934408,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004599805874567191,
      "loss": 4.5561,
      "mean_token_accuracy": 0.2645264759659767,
      "num_tokens": 6078979.0,
      "step": 7625
    },
    {
      "entropy": 4.760105419158935,
      "epoch": 2.1571953633022334,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045992201046758457,
      "loss": 4.4939,
      "mean_token_accuracy": 0.2799355015158653,
      "num_tokens": 6082750.0,
      "step": 7630
    },
    {
      "entropy": 4.858875179290772,
      "epoch": 2.1586089906700594,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004598633948308935,
      "loss": 4.6416,
      "mean_token_accuracy": 0.24431851506233215,
      "num_tokens": 6087001.0,
      "step": 7635
    },
    {
      "entropy": 4.909610319137573,
      "epoch": 2.1600226180378854,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045980474055889585,
      "loss": 4.6308,
      "mean_token_accuracy": 0.2511474221944809,
      "num_tokens": 6091030.0,
      "step": 7640
    },
    {
      "entropy": 4.882517528533936,
      "epoch": 2.161436245405711,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004597460476638504,
      "loss": 4.6327,
      "mean_token_accuracy": 0.2503205433487892,
      "num_tokens": 6095185.0,
      "step": 7645
    },
    {
      "entropy": 4.816834306716919,
      "epoch": 2.162849872773537,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045968731615802334,
      "loss": 4.6538,
      "mean_token_accuracy": 0.2517275080084801,
      "num_tokens": 6099154.0,
      "step": 7650
    },
    {
      "entropy": 4.928690481185913,
      "epoch": 2.1642635001413626,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045962854605368926,
      "loss": 4.7122,
      "mean_token_accuracy": 0.2402295097708702,
      "num_tokens": 6103084.0,
      "step": 7655
    },
    {
      "entropy": 4.871238470077515,
      "epoch": 2.1656771275091886,
      "grad_norm": 1.375,
      "learning_rate": 0.0004595697373631309,
      "loss": 4.495,
      "mean_token_accuracy": 0.26312555372714996,
      "num_tokens": 6107163.0,
      "step": 7660
    },
    {
      "entropy": 4.857384729385376,
      "epoch": 2.167090754877014,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004595108900986386,
      "loss": 4.6289,
      "mean_token_accuracy": 0.2553263932466507,
      "num_tokens": 6110978.0,
      "step": 7665
    },
    {
      "entropy": 4.819251680374146,
      "epoch": 2.16850438224484,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045945200427251133,
      "loss": 4.6323,
      "mean_token_accuracy": 0.26160899847745894,
      "num_tokens": 6115109.0,
      "step": 7670
    },
    {
      "entropy": 4.875520277023315,
      "epoch": 2.1699180096126662,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004593930798970558,
      "loss": 4.7484,
      "mean_token_accuracy": 0.25106675922870636,
      "num_tokens": 6119068.0,
      "step": 7675
    },
    {
      "entropy": 4.944271087646484,
      "epoch": 2.171331636980492,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045933411698458683,
      "loss": 4.6906,
      "mean_token_accuracy": 0.25721260011196134,
      "num_tokens": 6122773.0,
      "step": 7680
    },
    {
      "entropy": 4.945666027069092,
      "epoch": 2.172745264348318,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045927511554742724,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2572709143161774,
      "num_tokens": 6127009.0,
      "step": 7685
    },
    {
      "entropy": 4.935110712051392,
      "epoch": 2.174158891716144,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045921607559790806,
      "loss": 4.6394,
      "mean_token_accuracy": 0.2579881563782692,
      "num_tokens": 6131002.0,
      "step": 7690
    },
    {
      "entropy": 4.8415669918060305,
      "epoch": 2.1755725190839694,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004591569971483682,
      "loss": 4.6116,
      "mean_token_accuracy": 0.24984921813011168,
      "num_tokens": 6134827.0,
      "step": 7695
    },
    {
      "entropy": 4.819009304046631,
      "epoch": 2.1769861464517954,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004590978802111548,
      "loss": 4.673,
      "mean_token_accuracy": 0.24826411306858062,
      "num_tokens": 6138992.0,
      "step": 7700
    },
    {
      "entropy": 4.843704128265381,
      "epoch": 2.178399773819621,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045903872479862286,
      "loss": 4.5885,
      "mean_token_accuracy": 0.25611171424388884,
      "num_tokens": 6143206.0,
      "step": 7705
    },
    {
      "entropy": 4.859374904632569,
      "epoch": 2.179813401187447,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004589795309231354,
      "loss": 4.5517,
      "mean_token_accuracy": 0.2594165772199631,
      "num_tokens": 6146981.0,
      "step": 7710
    },
    {
      "entropy": 4.857029914855957,
      "epoch": 2.1812270285552726,
      "grad_norm": 1.375,
      "learning_rate": 0.00045892029859706377,
      "loss": 4.6662,
      "mean_token_accuracy": 0.24854653626680373,
      "num_tokens": 6151010.0,
      "step": 7715
    },
    {
      "entropy": 4.865697193145752,
      "epoch": 2.1826406559230986,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045886102783278704,
      "loss": 4.6188,
      "mean_token_accuracy": 0.258250430226326,
      "num_tokens": 6154765.0,
      "step": 7720
    },
    {
      "entropy": 4.959687852859497,
      "epoch": 2.1840542832909247,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045880171864269245,
      "loss": 4.6951,
      "mean_token_accuracy": 0.2391132116317749,
      "num_tokens": 6158493.0,
      "step": 7725
    },
    {
      "entropy": 4.9179528713226315,
      "epoch": 2.1854679106587502,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004587423710391753,
      "loss": 4.6198,
      "mean_token_accuracy": 0.2631689369678497,
      "num_tokens": 6162659.0,
      "step": 7730
    },
    {
      "entropy": 4.899079751968384,
      "epoch": 2.1868815380265763,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004586829850346388,
      "loss": 4.6649,
      "mean_token_accuracy": 0.2442048266530037,
      "num_tokens": 6166761.0,
      "step": 7735
    },
    {
      "entropy": 4.819434022903442,
      "epoch": 2.188295165394402,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045862356064149424,
      "loss": 4.6335,
      "mean_token_accuracy": 0.2515073552727699,
      "num_tokens": 6170719.0,
      "step": 7740
    },
    {
      "entropy": 4.906767511367798,
      "epoch": 2.189708792762228,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000458564097872161,
      "loss": 4.701,
      "mean_token_accuracy": 0.23593829423189164,
      "num_tokens": 6175013.0,
      "step": 7745
    },
    {
      "entropy": 4.820659494400024,
      "epoch": 2.191122420130054,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004585045967390665,
      "loss": 4.6237,
      "mean_token_accuracy": 0.255658221244812,
      "num_tokens": 6178897.0,
      "step": 7750
    },
    {
      "entropy": 4.831353044509887,
      "epoch": 2.1925360474978794,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000458445057254646,
      "loss": 4.6374,
      "mean_token_accuracy": 0.262596969306469,
      "num_tokens": 6182685.0,
      "step": 7755
    },
    {
      "entropy": 4.860255575180053,
      "epoch": 2.1939496748657055,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004583854794313429,
      "loss": 4.5834,
      "mean_token_accuracy": 0.2587323307991028,
      "num_tokens": 6186277.0,
      "step": 7760
    },
    {
      "entropy": 4.949373865127564,
      "epoch": 2.195363302233531,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004583258632816087,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24448552131652831,
      "num_tokens": 6190409.0,
      "step": 7765
    },
    {
      "entropy": 4.986118316650391,
      "epoch": 2.196776929601357,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004582662088179027,
      "loss": 4.7031,
      "mean_token_accuracy": 0.2487078532576561,
      "num_tokens": 6194582.0,
      "step": 7770
    },
    {
      "entropy": 4.989755773544312,
      "epoch": 2.198190556969183,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045820651605269223,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2408284530043602,
      "num_tokens": 6198593.0,
      "step": 7775
    },
    {
      "entropy": 4.9848497867584225,
      "epoch": 2.1996041843370087,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004581467849984529,
      "loss": 4.7191,
      "mean_token_accuracy": 0.2494204193353653,
      "num_tokens": 6202639.0,
      "step": 7780
    },
    {
      "entropy": 4.876068496704102,
      "epoch": 2.2010178117048347,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004580870156676681,
      "loss": 4.6964,
      "mean_token_accuracy": 0.2529083460569382,
      "num_tokens": 6206674.0,
      "step": 7785
    },
    {
      "entropy": 4.870501089096069,
      "epoch": 2.2024314390726603,
      "grad_norm": 1.375,
      "learning_rate": 0.00045802720807282916,
      "loss": 4.6531,
      "mean_token_accuracy": 0.25124975889921186,
      "num_tokens": 6210707.0,
      "step": 7790
    },
    {
      "entropy": 4.95373158454895,
      "epoch": 2.2038450664404863,
      "grad_norm": 1.375,
      "learning_rate": 0.0004579673622264356,
      "loss": 4.6896,
      "mean_token_accuracy": 0.24937673658132553,
      "num_tokens": 6214711.0,
      "step": 7795
    },
    {
      "entropy": 4.892318487167358,
      "epoch": 2.2052586938083123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045790747814099464,
      "loss": 4.6671,
      "mean_token_accuracy": 0.24569808840751647,
      "num_tokens": 6219088.0,
      "step": 7800
    },
    {
      "entropy": 4.861852645874023,
      "epoch": 2.206672321176138,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045784755582902186,
      "loss": 4.6548,
      "mean_token_accuracy": 0.23614756762981415,
      "num_tokens": 6223365.0,
      "step": 7805
    },
    {
      "entropy": 4.986469888687134,
      "epoch": 2.208085948543964,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045778759530304067,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23898189514875412,
      "num_tokens": 6227469.0,
      "step": 7810
    },
    {
      "entropy": 4.8986423969268795,
      "epoch": 2.2094995759117895,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045772759657558234,
      "loss": 4.5142,
      "mean_token_accuracy": 0.2694539040327072,
      "num_tokens": 6231428.0,
      "step": 7815
    },
    {
      "entropy": 4.910695028305054,
      "epoch": 2.2109132032796155,
      "grad_norm": 1.375,
      "learning_rate": 0.00045766755965918627,
      "loss": 4.6764,
      "mean_token_accuracy": 0.2442860186100006,
      "num_tokens": 6235244.0,
      "step": 7820
    },
    {
      "entropy": 4.82358751296997,
      "epoch": 2.2123268306474415,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004576074845663999,
      "loss": 4.6337,
      "mean_token_accuracy": 0.2622364431619644,
      "num_tokens": 6239434.0,
      "step": 7825
    },
    {
      "entropy": 4.8598438739776615,
      "epoch": 2.213740458015267,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004575473713097784,
      "loss": 4.6237,
      "mean_token_accuracy": 0.2505569398403168,
      "num_tokens": 6243480.0,
      "step": 7830
    },
    {
      "entropy": 4.84587664604187,
      "epoch": 2.215154085383093,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004574872199018851,
      "loss": 4.6528,
      "mean_token_accuracy": 0.2462034523487091,
      "num_tokens": 6247488.0,
      "step": 7835
    },
    {
      "entropy": 4.818640851974488,
      "epoch": 2.2165677127509187,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004574270303552913,
      "loss": 4.5333,
      "mean_token_accuracy": 0.26765376031398774,
      "num_tokens": 6251505.0,
      "step": 7840
    },
    {
      "entropy": 4.849223804473877,
      "epoch": 2.2179813401187447,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004573668026825763,
      "loss": 4.6327,
      "mean_token_accuracy": 0.24939990043640137,
      "num_tokens": 6255458.0,
      "step": 7845
    },
    {
      "entropy": 4.965388917922974,
      "epoch": 2.2193949674865707,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004573065368963273,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23333799988031387,
      "num_tokens": 6259095.0,
      "step": 7850
    },
    {
      "entropy": 4.958670520782471,
      "epoch": 2.2208085948543963,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045724623300913935,
      "loss": 4.7423,
      "mean_token_accuracy": 0.2411131590604782,
      "num_tokens": 6263014.0,
      "step": 7855
    },
    {
      "entropy": 4.818654918670655,
      "epoch": 2.2222222222222223,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004571858910336157,
      "loss": 4.6148,
      "mean_token_accuracy": 0.25421337485313417,
      "num_tokens": 6266993.0,
      "step": 7860
    },
    {
      "entropy": 5.019851303100586,
      "epoch": 2.223635849590048,
      "grad_norm": 1.375,
      "learning_rate": 0.0004571255109823675,
      "loss": 4.7015,
      "mean_token_accuracy": 0.2364548698067665,
      "num_tokens": 6270957.0,
      "step": 7865
    },
    {
      "entropy": 4.890041065216065,
      "epoch": 2.225049476957874,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004570650928680136,
      "loss": 4.6325,
      "mean_token_accuracy": 0.24749798476696014,
      "num_tokens": 6274928.0,
      "step": 7870
    },
    {
      "entropy": 4.916370439529419,
      "epoch": 2.2264631043257,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004570046367031813,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24893801659345627,
      "num_tokens": 6278955.0,
      "step": 7875
    },
    {
      "entropy": 4.7520676612854,
      "epoch": 2.2278767316935255,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004569441425005053,
      "loss": 4.4987,
      "mean_token_accuracy": 0.26559889763593675,
      "num_tokens": 6283021.0,
      "step": 7880
    },
    {
      "entropy": 4.947638893127442,
      "epoch": 2.2292903590613515,
      "grad_norm": 1.375,
      "learning_rate": 0.00045688361027262865,
      "loss": 4.711,
      "mean_token_accuracy": 0.24464566856622696,
      "num_tokens": 6287372.0,
      "step": 7885
    },
    {
      "entropy": 4.937538099288941,
      "epoch": 2.230703986429177,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004568230400322022,
      "loss": 4.7151,
      "mean_token_accuracy": 0.2397904574871063,
      "num_tokens": 6291754.0,
      "step": 7890
    },
    {
      "entropy": 4.787794876098633,
      "epoch": 2.232117613797003,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045676243179188473,
      "loss": 4.6291,
      "mean_token_accuracy": 0.2612901240587234,
      "num_tokens": 6295405.0,
      "step": 7895
    },
    {
      "entropy": 4.8579833984375,
      "epoch": 2.233531241164829,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004567017855643431,
      "loss": 4.7303,
      "mean_token_accuracy": 0.24914026260375977,
      "num_tokens": 6299409.0,
      "step": 7900
    },
    {
      "entropy": 4.866698312759399,
      "epoch": 2.2349448685326547,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004566411013622518,
      "loss": 4.726,
      "mean_token_accuracy": 0.2437924399971962,
      "num_tokens": 6303678.0,
      "step": 7905
    },
    {
      "entropy": 4.93493390083313,
      "epoch": 2.2363584959004807,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045658037919829353,
      "loss": 4.7478,
      "mean_token_accuracy": 0.2410943925380707,
      "num_tokens": 6307541.0,
      "step": 7910
    },
    {
      "entropy": 4.820869016647339,
      "epoch": 2.2377721232683063,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004565196190851589,
      "loss": 4.505,
      "mean_token_accuracy": 0.2713637962937355,
      "num_tokens": 6311630.0,
      "step": 7915
    },
    {
      "entropy": 4.883880996704102,
      "epoch": 2.2391857506361323,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004564588210355464,
      "loss": 4.6856,
      "mean_token_accuracy": 0.2529781997203827,
      "num_tokens": 6315521.0,
      "step": 7920
    },
    {
      "entropy": 4.893385887145996,
      "epoch": 2.2405993780039584,
      "grad_norm": 1.375,
      "learning_rate": 0.00045639798506216236,
      "loss": 4.6798,
      "mean_token_accuracy": 0.2452431708574295,
      "num_tokens": 6319463.0,
      "step": 7925
    },
    {
      "entropy": 4.911597633361817,
      "epoch": 2.242013005371784,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045633711117772113,
      "loss": 4.6467,
      "mean_token_accuracy": 0.2575483381748199,
      "num_tokens": 6323818.0,
      "step": 7930
    },
    {
      "entropy": 4.876330041885376,
      "epoch": 2.24342663273961,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004562761993949451,
      "loss": 4.7789,
      "mean_token_accuracy": 0.22986506372690202,
      "num_tokens": 6327339.0,
      "step": 7935
    },
    {
      "entropy": 4.880350971221924,
      "epoch": 2.2448402601074355,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045621524972656435,
      "loss": 4.4899,
      "mean_token_accuracy": 0.2678906872868538,
      "num_tokens": 6331549.0,
      "step": 7940
    },
    {
      "entropy": 4.756036853790283,
      "epoch": 2.2462538874752616,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045615426218531703,
      "loss": 4.5827,
      "mean_token_accuracy": 0.26007754504680636,
      "num_tokens": 6335606.0,
      "step": 7945
    },
    {
      "entropy": 4.911910915374756,
      "epoch": 2.247667514843087,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045609323678394914,
      "loss": 4.6598,
      "mean_token_accuracy": 0.2547120630741119,
      "num_tokens": 6339752.0,
      "step": 7950
    },
    {
      "entropy": 4.86850414276123,
      "epoch": 2.249081142210913,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004560321735352146,
      "loss": 4.5556,
      "mean_token_accuracy": 0.25856774151325224,
      "num_tokens": 6343312.0,
      "step": 7955
    },
    {
      "entropy": 4.837004232406616,
      "epoch": 2.250494769578739,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045597107245187523,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24258037358522416,
      "num_tokens": 6347090.0,
      "step": 7960
    },
    {
      "entropy": 4.742466878890991,
      "epoch": 2.2519083969465647,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004559099335467008,
      "loss": 4.5,
      "mean_token_accuracy": 0.2634102150797844,
      "num_tokens": 6350945.0,
      "step": 7965
    },
    {
      "entropy": 4.8074798583984375,
      "epoch": 2.2533220243143908,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000455848756832469,
      "loss": 4.5899,
      "mean_token_accuracy": 0.25583866983652115,
      "num_tokens": 6354830.0,
      "step": 7970
    },
    {
      "entropy": 4.83731517791748,
      "epoch": 2.254735651682217,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045578754232196536,
      "loss": 4.5815,
      "mean_token_accuracy": 0.2616676643490791,
      "num_tokens": 6358659.0,
      "step": 7975
    },
    {
      "entropy": 4.978814268112183,
      "epoch": 2.2561492790500424,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004557262900279834,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24685793817043306,
      "num_tokens": 6362890.0,
      "step": 7980
    },
    {
      "entropy": 4.8424969673156735,
      "epoch": 2.2575629064178684,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045566499996332427,
      "loss": 4.6574,
      "mean_token_accuracy": 0.2483309179544449,
      "num_tokens": 6367212.0,
      "step": 7985
    },
    {
      "entropy": 4.877757215499878,
      "epoch": 2.258976533785694,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045560367214079746,
      "loss": 4.5446,
      "mean_token_accuracy": 0.26867552697658537,
      "num_tokens": 6371058.0,
      "step": 7990
    },
    {
      "entropy": 4.971283054351806,
      "epoch": 2.26039016115352,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004555423065732199,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23049061000347137,
      "num_tokens": 6374931.0,
      "step": 7995
    },
    {
      "entropy": 4.966945648193359,
      "epoch": 2.2618037885213456,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004554809032734168,
      "loss": 4.738,
      "mean_token_accuracy": 0.23721687495708466,
      "num_tokens": 6379069.0,
      "step": 8000
    },
    {
      "epoch": 2.2618037885213456,
      "eval_entropy": 4.678514098039677,
      "eval_loss": 5.113204479217529,
      "eval_mean_token_accuracy": 0.2378401588371339,
      "eval_num_tokens": 6379069.0,
      "eval_runtime": 4.8965,
      "eval_samples_per_second": 2304.526,
      "eval_steps_per_second": 288.168,
      "step": 8000
    },
    {
      "entropy": 4.810961055755615,
      "epoch": 2.2632174158891716,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004554194622542209,
      "loss": 4.5571,
      "mean_token_accuracy": 0.2604352757334709,
      "num_tokens": 6383241.0,
      "step": 8005
    },
    {
      "entropy": 4.826586484909058,
      "epoch": 2.2646310432569976,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045535798352847306,
      "loss": 4.5691,
      "mean_token_accuracy": 0.2579935476183891,
      "num_tokens": 6387303.0,
      "step": 8010
    },
    {
      "entropy": 4.9231236457824705,
      "epoch": 2.266044670624823,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004552964671090219,
      "loss": 4.7041,
      "mean_token_accuracy": 0.2519547000527382,
      "num_tokens": 6391604.0,
      "step": 8015
    },
    {
      "entropy": 4.916821241378784,
      "epoch": 2.267458297992649,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045523491300872423,
      "loss": 4.6791,
      "mean_token_accuracy": 0.24187731891870498,
      "num_tokens": 6395672.0,
      "step": 8020
    },
    {
      "entropy": 4.937158536911011,
      "epoch": 2.268871925360475,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045517332124044414,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2479722633957863,
      "num_tokens": 6399962.0,
      "step": 8025
    },
    {
      "entropy": 4.9073097705841064,
      "epoch": 2.270285552728301,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004551116918170541,
      "loss": 4.5752,
      "mean_token_accuracy": 0.25848977714776994,
      "num_tokens": 6404012.0,
      "step": 8030
    },
    {
      "entropy": 4.779842519760132,
      "epoch": 2.271699180096127,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045505002475143433,
      "loss": 4.6176,
      "mean_token_accuracy": 0.2524653822183609,
      "num_tokens": 6407773.0,
      "step": 8035
    },
    {
      "entropy": 4.782460165023804,
      "epoch": 2.2731128074639524,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045498832005647267,
      "loss": 4.7026,
      "mean_token_accuracy": 0.24428311586380005,
      "num_tokens": 6411681.0,
      "step": 8040
    },
    {
      "entropy": 4.893334150314331,
      "epoch": 2.2745264348317784,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004549265777450652,
      "loss": 4.7107,
      "mean_token_accuracy": 0.24530066698789596,
      "num_tokens": 6415374.0,
      "step": 8045
    },
    {
      "entropy": 4.9495690822601315,
      "epoch": 2.275940062199604,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045486479783011553,
      "loss": 4.6082,
      "mean_token_accuracy": 0.25667900294065477,
      "num_tokens": 6419321.0,
      "step": 8050
    },
    {
      "entropy": 4.879259395599365,
      "epoch": 2.27735368956743,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045480298032453544,
      "loss": 4.5711,
      "mean_token_accuracy": 0.24959348291158676,
      "num_tokens": 6423064.0,
      "step": 8055
    },
    {
      "entropy": 4.8198669910430905,
      "epoch": 2.278767316935256,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004547411252412443,
      "loss": 4.7175,
      "mean_token_accuracy": 0.2330663412809372,
      "num_tokens": 6427130.0,
      "step": 8060
    },
    {
      "entropy": 4.839763164520264,
      "epoch": 2.2801809443030816,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004546792325931695,
      "loss": 4.5773,
      "mean_token_accuracy": 0.24867051541805268,
      "num_tokens": 6431401.0,
      "step": 8065
    },
    {
      "entropy": 4.8087231636047365,
      "epoch": 2.2815945716709076,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004546173023932461,
      "loss": 4.632,
      "mean_token_accuracy": 0.2597040802240372,
      "num_tokens": 6434993.0,
      "step": 8070
    },
    {
      "entropy": 4.907133293151856,
      "epoch": 2.283008199038733,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045455533465441734,
      "loss": 4.6473,
      "mean_token_accuracy": 0.25460434556007383,
      "num_tokens": 6438819.0,
      "step": 8075
    },
    {
      "entropy": 4.901380825042724,
      "epoch": 2.284421826406559,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004544933293896339,
      "loss": 4.6564,
      "mean_token_accuracy": 0.2533603772521019,
      "num_tokens": 6442849.0,
      "step": 8080
    },
    {
      "entropy": 4.853040981292724,
      "epoch": 2.2858354537743852,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045443128661185457,
      "loss": 4.6745,
      "mean_token_accuracy": 0.2509542047977448,
      "num_tokens": 6446853.0,
      "step": 8085
    },
    {
      "entropy": 4.91700291633606,
      "epoch": 2.287249081142211,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045436920633404597,
      "loss": 4.6982,
      "mean_token_accuracy": 0.24499902427196502,
      "num_tokens": 6451312.0,
      "step": 8090
    },
    {
      "entropy": 4.846444892883301,
      "epoch": 2.288662708510037,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004543070885691824,
      "loss": 4.7332,
      "mean_token_accuracy": 0.24791177958250046,
      "num_tokens": 6455356.0,
      "step": 8095
    },
    {
      "entropy": 4.8298914432525635,
      "epoch": 2.2900763358778624,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045424493333024607,
      "loss": 4.5677,
      "mean_token_accuracy": 0.26459922939538955,
      "num_tokens": 6459102.0,
      "step": 8100
    },
    {
      "entropy": 4.800870800018311,
      "epoch": 2.2914899632456884,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045418274063022715,
      "loss": 4.5394,
      "mean_token_accuracy": 0.2659174770116806,
      "num_tokens": 6462864.0,
      "step": 8105
    },
    {
      "entropy": 4.817087459564209,
      "epoch": 2.2929035906135145,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004541205104821235,
      "loss": 4.6176,
      "mean_token_accuracy": 0.2533339515328407,
      "num_tokens": 6466870.0,
      "step": 8110
    },
    {
      "entropy": 4.860741806030274,
      "epoch": 2.29431721798134,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004540582428989408,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2541018173098564,
      "num_tokens": 6470793.0,
      "step": 8115
    },
    {
      "entropy": 4.906598281860352,
      "epoch": 2.295730845349166,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045399593789369246,
      "loss": 4.7374,
      "mean_token_accuracy": 0.245554356276989,
      "num_tokens": 6474851.0,
      "step": 8120
    },
    {
      "entropy": 4.8137119770050045,
      "epoch": 2.2971444727169916,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045393359547940017,
      "loss": 4.5989,
      "mean_token_accuracy": 0.2629430279135704,
      "num_tokens": 6478827.0,
      "step": 8125
    },
    {
      "entropy": 4.963550233840943,
      "epoch": 2.2985581000848176,
      "grad_norm": 1.5,
      "learning_rate": 0.0004538712156690927,
      "loss": 4.7874,
      "mean_token_accuracy": 0.2384938359260559,
      "num_tokens": 6482771.0,
      "step": 8130
    },
    {
      "entropy": 4.8731293201446535,
      "epoch": 2.2999717274526437,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045380879847580743,
      "loss": 4.6318,
      "mean_token_accuracy": 0.2517322227358818,
      "num_tokens": 6486876.0,
      "step": 8135
    },
    {
      "entropy": 4.799281883239746,
      "epoch": 2.3013853548204692,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000453746343912589,
      "loss": 4.5206,
      "mean_token_accuracy": 0.26870024502277373,
      "num_tokens": 6490741.0,
      "step": 8140
    },
    {
      "entropy": 4.80910291671753,
      "epoch": 2.3027989821882953,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045368385199249004,
      "loss": 4.6596,
      "mean_token_accuracy": 0.24854776710271836,
      "num_tokens": 6494387.0,
      "step": 8145
    },
    {
      "entropy": 4.877581596374512,
      "epoch": 2.304212609556121,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045362132272857084,
      "loss": 4.648,
      "mean_token_accuracy": 0.25215546488761903,
      "num_tokens": 6498384.0,
      "step": 8150
    },
    {
      "entropy": 4.907680559158325,
      "epoch": 2.305626236923947,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004535587561338998,
      "loss": 4.7487,
      "mean_token_accuracy": 0.25238660126924517,
      "num_tokens": 6502371.0,
      "step": 8155
    },
    {
      "entropy": 4.90586256980896,
      "epoch": 2.307039864291773,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004534961522215529,
      "loss": 4.711,
      "mean_token_accuracy": 0.2389559656381607,
      "num_tokens": 6506660.0,
      "step": 8160
    },
    {
      "entropy": 4.92593035697937,
      "epoch": 2.3084534916595985,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000453433511004614,
      "loss": 4.6672,
      "mean_token_accuracy": 0.25590457767248154,
      "num_tokens": 6510691.0,
      "step": 8165
    },
    {
      "entropy": 5.002705144882202,
      "epoch": 2.3098671190274245,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004533708324961747,
      "loss": 4.7428,
      "mean_token_accuracy": 0.23928603678941726,
      "num_tokens": 6514525.0,
      "step": 8170
    },
    {
      "entropy": 4.941864252090454,
      "epoch": 2.31128074639525,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045330811670933434,
      "loss": 4.6081,
      "mean_token_accuracy": 0.25387614518404006,
      "num_tokens": 6518532.0,
      "step": 8175
    },
    {
      "entropy": 4.795288038253784,
      "epoch": 2.312694373763076,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045324536365720034,
      "loss": 4.5529,
      "mean_token_accuracy": 0.25909191370010376,
      "num_tokens": 6522600.0,
      "step": 8180
    },
    {
      "entropy": 4.8765160083770756,
      "epoch": 2.3141080011309016,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004531825733528875,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24207843989133834,
      "num_tokens": 6527131.0,
      "step": 8185
    },
    {
      "entropy": 4.9078040599823,
      "epoch": 2.3155216284987277,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004531197458095187,
      "loss": 4.7585,
      "mean_token_accuracy": 0.24226156026124954,
      "num_tokens": 6531400.0,
      "step": 8190
    },
    {
      "entropy": 4.816693592071533,
      "epoch": 2.3169352558665537,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004530568810402245,
      "loss": 4.5239,
      "mean_token_accuracy": 0.2624052092432976,
      "num_tokens": 6535298.0,
      "step": 8195
    },
    {
      "entropy": 4.867839908599853,
      "epoch": 2.3183488832343793,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045299397905814316,
      "loss": 4.6634,
      "mean_token_accuracy": 0.24662096947431564,
      "num_tokens": 6539272.0,
      "step": 8200
    },
    {
      "entropy": 4.871376705169678,
      "epoch": 2.3197625106022053,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045293103987642096,
      "loss": 4.6471,
      "mean_token_accuracy": 0.250042712688446,
      "num_tokens": 6543107.0,
      "step": 8205
    },
    {
      "entropy": 5.013834285736084,
      "epoch": 2.3211761379700313,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004528680635082117,
      "loss": 4.7293,
      "mean_token_accuracy": 0.24051018804311752,
      "num_tokens": 6547096.0,
      "step": 8210
    },
    {
      "entropy": 4.879400300979614,
      "epoch": 2.322589765337857,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045280504996667704,
      "loss": 4.5779,
      "mean_token_accuracy": 0.24956424087285994,
      "num_tokens": 6550863.0,
      "step": 8215
    },
    {
      "entropy": 4.862359142303466,
      "epoch": 2.324003392705683,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045274199926498645,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24737590849399566,
      "num_tokens": 6554403.0,
      "step": 8220
    },
    {
      "entropy": 4.756053352355957,
      "epoch": 2.3254170200735085,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004526789114163172,
      "loss": 4.5961,
      "mean_token_accuracy": 0.25908535718917847,
      "num_tokens": 6558542.0,
      "step": 8225
    },
    {
      "entropy": 4.86450777053833,
      "epoch": 2.3268306474413345,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004526157864338541,
      "loss": 4.5727,
      "mean_token_accuracy": 0.25068659484386446,
      "num_tokens": 6562471.0,
      "step": 8230
    },
    {
      "entropy": 4.819010019302368,
      "epoch": 2.32824427480916,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045255262433079003,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2574344739317894,
      "num_tokens": 6566459.0,
      "step": 8235
    },
    {
      "entropy": 4.826765394210815,
      "epoch": 2.329657902176986,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045248942512032543,
      "loss": 4.6384,
      "mean_token_accuracy": 0.2546679377555847,
      "num_tokens": 6570380.0,
      "step": 8240
    },
    {
      "entropy": 4.783012914657593,
      "epoch": 2.331071529544812,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045242618881566847,
      "loss": 4.6095,
      "mean_token_accuracy": 0.25267250537872316,
      "num_tokens": 6574142.0,
      "step": 8245
    },
    {
      "entropy": 4.849971294403076,
      "epoch": 2.3324851569126377,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004523629154300353,
      "loss": 4.5891,
      "mean_token_accuracy": 0.26194084882736207,
      "num_tokens": 6578286.0,
      "step": 8250
    },
    {
      "entropy": 4.809189128875732,
      "epoch": 2.3338987842804637,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045229960497664964,
      "loss": 4.5873,
      "mean_token_accuracy": 0.2504382863640785,
      "num_tokens": 6582045.0,
      "step": 8255
    },
    {
      "entropy": 4.856415796279907,
      "epoch": 2.3353124116482897,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004522362574687428,
      "loss": 4.6494,
      "mean_token_accuracy": 0.24462329745292663,
      "num_tokens": 6586246.0,
      "step": 8260
    },
    {
      "entropy": 4.92406120300293,
      "epoch": 2.3367260390161153,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004521728729195542,
      "loss": 4.7013,
      "mean_token_accuracy": 0.24874500334262847,
      "num_tokens": 6590274.0,
      "step": 8265
    },
    {
      "entropy": 4.878043556213379,
      "epoch": 2.3381396663839413,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045210945134233077,
      "loss": 4.6936,
      "mean_token_accuracy": 0.25521301925182344,
      "num_tokens": 6594327.0,
      "step": 8270
    },
    {
      "entropy": 4.898689603805542,
      "epoch": 2.339553293751767,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045204599275032724,
      "loss": 4.7092,
      "mean_token_accuracy": 0.2415061801671982,
      "num_tokens": 6597873.0,
      "step": 8275
    },
    {
      "entropy": 4.831254529953003,
      "epoch": 2.340966921119593,
      "grad_norm": 1.296875,
      "learning_rate": 0.000451982497156806,
      "loss": 4.5999,
      "mean_token_accuracy": 0.2557010561227798,
      "num_tokens": 6601919.0,
      "step": 8280
    },
    {
      "entropy": 4.939549112319947,
      "epoch": 2.3423805484874185,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004519189645750373,
      "loss": 4.7108,
      "mean_token_accuracy": 0.258120895922184,
      "num_tokens": 6606008.0,
      "step": 8285
    },
    {
      "entropy": 4.978422784805298,
      "epoch": 2.3437941758552445,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000451855395018299,
      "loss": 4.6617,
      "mean_token_accuracy": 0.2577178627252579,
      "num_tokens": 6609870.0,
      "step": 8290
    },
    {
      "entropy": 4.833541536331177,
      "epoch": 2.3452078032230705,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004517917884998768,
      "loss": 4.6014,
      "mean_token_accuracy": 0.24858151972293854,
      "num_tokens": 6613678.0,
      "step": 8295
    },
    {
      "entropy": 4.720071840286255,
      "epoch": 2.346621430590896,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004517281450330642,
      "loss": 4.6454,
      "mean_token_accuracy": 0.24882385581731797,
      "num_tokens": 6617648.0,
      "step": 8300
    },
    {
      "entropy": 4.8153883934021,
      "epoch": 2.348035057958722,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045166446463116195,
      "loss": 4.5654,
      "mean_token_accuracy": 0.26598226726055146,
      "num_tokens": 6621736.0,
      "step": 8305
    },
    {
      "entropy": 4.823480033874512,
      "epoch": 2.349448685326548,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004516007473074791,
      "loss": 4.5949,
      "mean_token_accuracy": 0.25615633726119996,
      "num_tokens": 6625542.0,
      "step": 8310
    },
    {
      "entropy": 4.899303007125854,
      "epoch": 2.3508623126943737,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004515369930753322,
      "loss": 4.7261,
      "mean_token_accuracy": 0.23804289251565933,
      "num_tokens": 6629790.0,
      "step": 8315
    },
    {
      "entropy": 4.878256034851074,
      "epoch": 2.3522759400621998,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004514732019480453,
      "loss": 4.6059,
      "mean_token_accuracy": 0.2584063529968262,
      "num_tokens": 6633758.0,
      "step": 8320
    },
    {
      "entropy": 4.846408557891846,
      "epoch": 2.3536895674300253,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004514093739389506,
      "loss": 4.6476,
      "mean_token_accuracy": 0.2584442064166069,
      "num_tokens": 6638069.0,
      "step": 8325
    },
    {
      "entropy": 4.786826276779175,
      "epoch": 2.3551031947978514,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004513455090613876,
      "loss": 4.6795,
      "mean_token_accuracy": 0.25553069412708285,
      "num_tokens": 6641811.0,
      "step": 8330
    },
    {
      "entropy": 4.768904829025269,
      "epoch": 2.356516822165677,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004512816073287036,
      "loss": 4.6272,
      "mean_token_accuracy": 0.2619549512863159,
      "num_tokens": 6645839.0,
      "step": 8335
    },
    {
      "entropy": 4.984873867034912,
      "epoch": 2.357930449533503,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045121766875425386,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24839209914207458,
      "num_tokens": 6649878.0,
      "step": 8340
    },
    {
      "entropy": 4.886408185958862,
      "epoch": 2.359344076901329,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045115369335140096,
      "loss": 4.6055,
      "mean_token_accuracy": 0.24616075158119202,
      "num_tokens": 6653775.0,
      "step": 8345
    },
    {
      "entropy": 4.890987396240234,
      "epoch": 2.3607577042691545,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004510896811335155,
      "loss": 4.7036,
      "mean_token_accuracy": 0.2476263478398323,
      "num_tokens": 6657905.0,
      "step": 8350
    },
    {
      "entropy": 4.962121057510376,
      "epoch": 2.3621713316369806,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045102563211397556,
      "loss": 4.7525,
      "mean_token_accuracy": 0.25208212435245514,
      "num_tokens": 6662067.0,
      "step": 8355
    },
    {
      "entropy": 4.85676851272583,
      "epoch": 2.363584959004806,
      "grad_norm": 1.5,
      "learning_rate": 0.0004509615463061671,
      "loss": 4.645,
      "mean_token_accuracy": 0.24859954565763473,
      "num_tokens": 6665774.0,
      "step": 8360
    },
    {
      "entropy": 4.8691332817077635,
      "epoch": 2.364998586372632,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045089742372348345,
      "loss": 4.6056,
      "mean_token_accuracy": 0.2554431453347206,
      "num_tokens": 6669566.0,
      "step": 8365
    },
    {
      "entropy": 4.932449865341186,
      "epoch": 2.366412213740458,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004508332643793261,
      "loss": 4.7583,
      "mean_token_accuracy": 0.2455241337418556,
      "num_tokens": 6673505.0,
      "step": 8370
    },
    {
      "entropy": 4.964992523193359,
      "epoch": 2.3678258411082838,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004507690682871037,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2476774975657463,
      "num_tokens": 6677115.0,
      "step": 8375
    },
    {
      "entropy": 4.906571340560913,
      "epoch": 2.3692394684761098,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000450704835460233,
      "loss": 4.6732,
      "mean_token_accuracy": 0.24876879453659057,
      "num_tokens": 6680763.0,
      "step": 8380
    },
    {
      "entropy": 5.029644584655761,
      "epoch": 2.3706530958439354,
      "grad_norm": 1.375,
      "learning_rate": 0.00045064056591213826,
      "loss": 4.758,
      "mean_token_accuracy": 0.23427406698465347,
      "num_tokens": 6685082.0,
      "step": 8385
    },
    {
      "entropy": 4.900610446929932,
      "epoch": 2.3720667232117614,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045057625965625126,
      "loss": 4.6634,
      "mean_token_accuracy": 0.24435991644859315,
      "num_tokens": 6688925.0,
      "step": 8390
    },
    {
      "entropy": 4.734590816497803,
      "epoch": 2.3734803505795874,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045051191670601184,
      "loss": 4.5195,
      "mean_token_accuracy": 0.2513942688703537,
      "num_tokens": 6692832.0,
      "step": 8395
    },
    {
      "entropy": 4.840996551513672,
      "epoch": 2.374893977947413,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004504475370748671,
      "loss": 4.6534,
      "mean_token_accuracy": 0.23854925483465195,
      "num_tokens": 6696758.0,
      "step": 8400
    },
    {
      "entropy": 4.895843172073365,
      "epoch": 2.376307605315239,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045038312077627207,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2482403576374054,
      "num_tokens": 6700968.0,
      "step": 8405
    },
    {
      "entropy": 4.978205013275146,
      "epoch": 2.3777212326830646,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004503186678236893,
      "loss": 4.7543,
      "mean_token_accuracy": 0.24280380606651306,
      "num_tokens": 6705137.0,
      "step": 8410
    },
    {
      "entropy": 5.012967729568482,
      "epoch": 2.3791348600508906,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045025417823058906,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23150495439767838,
      "num_tokens": 6709051.0,
      "step": 8415
    },
    {
      "entropy": 4.795136213302612,
      "epoch": 2.3805484874187166,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004501896520104494,
      "loss": 4.5139,
      "mean_token_accuracy": 0.2725525125861168,
      "num_tokens": 6713029.0,
      "step": 8420
    },
    {
      "entropy": 4.826746940612793,
      "epoch": 2.381962114786542,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004501250891767558,
      "loss": 4.7065,
      "mean_token_accuracy": 0.24844780266284944,
      "num_tokens": 6717420.0,
      "step": 8425
    },
    {
      "entropy": 4.869444894790649,
      "epoch": 2.383375742154368,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004500604897430015,
      "loss": 4.5438,
      "mean_token_accuracy": 0.2503835693001747,
      "num_tokens": 6721458.0,
      "step": 8430
    },
    {
      "entropy": 4.8097762107849125,
      "epoch": 2.384789369522194,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004499958537226873,
      "loss": 4.6136,
      "mean_token_accuracy": 0.25815854370594027,
      "num_tokens": 6725584.0,
      "step": 8435
    },
    {
      "entropy": 4.876239824295044,
      "epoch": 2.38620299689002,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044993118112932187,
      "loss": 4.6612,
      "mean_token_accuracy": 0.2573897361755371,
      "num_tokens": 6730125.0,
      "step": 8440
    },
    {
      "entropy": 4.8847216129302975,
      "epoch": 2.387616624257846,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004498664719764213,
      "loss": 4.5837,
      "mean_token_accuracy": 0.25412219017744064,
      "num_tokens": 6734182.0,
      "step": 8445
    },
    {
      "entropy": 4.884736490249634,
      "epoch": 2.3890302516256714,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044980172627750954,
      "loss": 4.6287,
      "mean_token_accuracy": 0.2599888741970062,
      "num_tokens": 6738163.0,
      "step": 8450
    },
    {
      "entropy": 4.851342105865479,
      "epoch": 2.3904438789934974,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044973694404611776,
      "loss": 4.6166,
      "mean_token_accuracy": 0.252636456489563,
      "num_tokens": 6741861.0,
      "step": 8455
    },
    {
      "entropy": 4.909964179992675,
      "epoch": 2.391857506361323,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044967212529578533,
      "loss": 4.773,
      "mean_token_accuracy": 0.24714642465114595,
      "num_tokens": 6745776.0,
      "step": 8460
    },
    {
      "entropy": 4.946491718292236,
      "epoch": 2.393271133729149,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004496072700400588,
      "loss": 4.6543,
      "mean_token_accuracy": 0.25064171850681305,
      "num_tokens": 6749775.0,
      "step": 8465
    },
    {
      "entropy": 4.930731630325317,
      "epoch": 2.3946847610969746,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044954237829249264,
      "loss": 4.733,
      "mean_token_accuracy": 0.24851613789796828,
      "num_tokens": 6753535.0,
      "step": 8470
    },
    {
      "entropy": 4.888524103164673,
      "epoch": 2.3960983884648006,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004494774500666487,
      "loss": 4.7023,
      "mean_token_accuracy": 0.22713498771190643,
      "num_tokens": 6757733.0,
      "step": 8475
    },
    {
      "entropy": 4.873904275894165,
      "epoch": 2.3975120158326266,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044941248537609663,
      "loss": 4.7082,
      "mean_token_accuracy": 0.2469261184334755,
      "num_tokens": 6762026.0,
      "step": 8480
    },
    {
      "entropy": 4.922762203216553,
      "epoch": 2.398925643200452,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004493474842344137,
      "loss": 4.6461,
      "mean_token_accuracy": 0.2410658046603203,
      "num_tokens": 6766095.0,
      "step": 8485
    },
    {
      "entropy": 4.843142318725586,
      "epoch": 2.4003392705682782,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044928244665518464,
      "loss": 4.6403,
      "mean_token_accuracy": 0.2455910012125969,
      "num_tokens": 6770185.0,
      "step": 8490
    },
    {
      "entropy": 4.856609630584717,
      "epoch": 2.4017528979361042,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044921737265200194,
      "loss": 4.6742,
      "mean_token_accuracy": 0.250920395553112,
      "num_tokens": 6773735.0,
      "step": 8495
    },
    {
      "entropy": 4.867924404144287,
      "epoch": 2.40316652530393,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004491522622384658,
      "loss": 4.7228,
      "mean_token_accuracy": 0.24794320315122603,
      "num_tokens": 6777546.0,
      "step": 8500
    },
    {
      "entropy": 4.820406484603882,
      "epoch": 2.404580152671756,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004490871154281836,
      "loss": 4.6257,
      "mean_token_accuracy": 0.2634152263402939,
      "num_tokens": 6781181.0,
      "step": 8505
    },
    {
      "entropy": 4.880620527267456,
      "epoch": 2.4059937800395814,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044902193223477096,
      "loss": 4.5466,
      "mean_token_accuracy": 0.25471906661987304,
      "num_tokens": 6785123.0,
      "step": 8510
    },
    {
      "entropy": 4.839361763000488,
      "epoch": 2.4074074074074074,
      "grad_norm": 1.375,
      "learning_rate": 0.0004489567126718505,
      "loss": 4.6103,
      "mean_token_accuracy": 0.2576262906193733,
      "num_tokens": 6789024.0,
      "step": 8515
    },
    {
      "entropy": 4.823822736740112,
      "epoch": 2.408821034775233,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044889145675305284,
      "loss": 4.6613,
      "mean_token_accuracy": 0.2509738117456436,
      "num_tokens": 6792893.0,
      "step": 8520
    },
    {
      "entropy": 4.899869060516357,
      "epoch": 2.410234662143059,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004488261644920161,
      "loss": 4.7658,
      "mean_token_accuracy": 0.24188071191310884,
      "num_tokens": 6796722.0,
      "step": 8525
    },
    {
      "entropy": 4.872188758850098,
      "epoch": 2.411648289510885,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044876083590238575,
      "loss": 4.6567,
      "mean_token_accuracy": 0.26000567376613615,
      "num_tokens": 6800662.0,
      "step": 8530
    },
    {
      "entropy": 4.969104766845703,
      "epoch": 2.4130619168787106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004486954709978153,
      "loss": 4.7163,
      "mean_token_accuracy": 0.23044851422309875,
      "num_tokens": 6804619.0,
      "step": 8535
    },
    {
      "entropy": 4.864474439620972,
      "epoch": 2.4144755442465367,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044863006979196555,
      "loss": 4.6864,
      "mean_token_accuracy": 0.24679640531539918,
      "num_tokens": 6808410.0,
      "step": 8540
    },
    {
      "entropy": 4.930956745147705,
      "epoch": 2.4158891716143627,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004485646322985048,
      "loss": 4.6525,
      "mean_token_accuracy": 0.24700527787208557,
      "num_tokens": 6812304.0,
      "step": 8545
    },
    {
      "entropy": 4.789476442337036,
      "epoch": 2.4173027989821882,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044849915853110924,
      "loss": 4.6023,
      "mean_token_accuracy": 0.25799730569124224,
      "num_tokens": 6816446.0,
      "step": 8550
    },
    {
      "entropy": 4.89752779006958,
      "epoch": 2.4187164263500143,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00044843364850346245,
      "loss": 4.6813,
      "mean_token_accuracy": 0.24706668704748153,
      "num_tokens": 6820714.0,
      "step": 8555
    },
    {
      "entropy": 4.893315935134888,
      "epoch": 2.42013005371784,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044836810222925554,
      "loss": 4.6663,
      "mean_token_accuracy": 0.25243432968854906,
      "num_tokens": 6824647.0,
      "step": 8560
    },
    {
      "entropy": 4.804040479660034,
      "epoch": 2.421543681085666,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044830251972218746,
      "loss": 4.4673,
      "mean_token_accuracy": 0.273172564804554,
      "num_tokens": 6828110.0,
      "step": 8565
    },
    {
      "entropy": 4.9041462421417235,
      "epoch": 2.4229573084534914,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004482369009959644,
      "loss": 4.7075,
      "mean_token_accuracy": 0.23556320518255233,
      "num_tokens": 6832311.0,
      "step": 8570
    },
    {
      "entropy": 4.889885663986206,
      "epoch": 2.4243709358213175,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044817124606430015,
      "loss": 4.6681,
      "mean_token_accuracy": 0.24938600957393647,
      "num_tokens": 6836123.0,
      "step": 8575
    },
    {
      "entropy": 4.979565000534057,
      "epoch": 2.4257845631891435,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004481055549409165,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23892668783664703,
      "num_tokens": 6840214.0,
      "step": 8580
    },
    {
      "entropy": 4.857937622070312,
      "epoch": 2.427198190556969,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004480398276395422,
      "loss": 4.5921,
      "mean_token_accuracy": 0.2610870346426964,
      "num_tokens": 6844321.0,
      "step": 8585
    },
    {
      "entropy": 4.889852905273438,
      "epoch": 2.428611817924795,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044797406417391403,
      "loss": 4.5774,
      "mean_token_accuracy": 0.24895098954439163,
      "num_tokens": 6848365.0,
      "step": 8590
    },
    {
      "entropy": 4.896711826324463,
      "epoch": 2.430025445292621,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044790826455777603,
      "loss": 4.6065,
      "mean_token_accuracy": 0.25399615317583085,
      "num_tokens": 6852179.0,
      "step": 8595
    },
    {
      "entropy": 4.883155822753906,
      "epoch": 2.4314390726604467,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044784242880488006,
      "loss": 4.7015,
      "mean_token_accuracy": 0.24537769109010696,
      "num_tokens": 6856321.0,
      "step": 8600
    },
    {
      "entropy": 4.888510704040527,
      "epoch": 2.4328527000282727,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004477765569289853,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2439849704504013,
      "num_tokens": 6860316.0,
      "step": 8605
    },
    {
      "entropy": 4.819053840637207,
      "epoch": 2.4342663273960983,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044771064894385854,
      "loss": 4.6014,
      "mean_token_accuracy": 0.26226596236228944,
      "num_tokens": 6864119.0,
      "step": 8610
    },
    {
      "entropy": 4.758191728591919,
      "epoch": 2.4356799547639243,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044764470486327406,
      "loss": 4.611,
      "mean_token_accuracy": 0.26351589560508726,
      "num_tokens": 6867939.0,
      "step": 8615
    },
    {
      "entropy": 4.853074359893799,
      "epoch": 2.43709358213175,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000447578724701014,
      "loss": 4.6672,
      "mean_token_accuracy": 0.261250801384449,
      "num_tokens": 6871561.0,
      "step": 8620
    },
    {
      "entropy": 4.914012908935547,
      "epoch": 2.438507209499576,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004475127084708676,
      "loss": 4.6781,
      "mean_token_accuracy": 0.2540192499756813,
      "num_tokens": 6875399.0,
      "step": 8625
    },
    {
      "entropy": 4.902969741821289,
      "epoch": 2.439920836867402,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000447446656186632,
      "loss": 4.6698,
      "mean_token_accuracy": 0.2449094161391258,
      "num_tokens": 6879298.0,
      "step": 8630
    },
    {
      "entropy": 4.909339904785156,
      "epoch": 2.4413344642352275,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004473805678621115,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24226270318031312,
      "num_tokens": 6883204.0,
      "step": 8635
    },
    {
      "entropy": 4.860505485534668,
      "epoch": 2.4427480916030535,
      "grad_norm": 1.375,
      "learning_rate": 0.00044731444351111833,
      "loss": 4.6311,
      "mean_token_accuracy": 0.2503343477845192,
      "num_tokens": 6887469.0,
      "step": 8640
    },
    {
      "entropy": 4.788114881515503,
      "epoch": 2.444161718970879,
      "grad_norm": 1.296875,
      "learning_rate": 0.00044724828314747213,
      "loss": 4.541,
      "mean_token_accuracy": 0.2623747497797012,
      "num_tokens": 6891484.0,
      "step": 8645
    },
    {
      "entropy": 4.885194110870361,
      "epoch": 2.445575346338705,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004471820867849998,
      "loss": 4.7113,
      "mean_token_accuracy": 0.23764886409044267,
      "num_tokens": 6895300.0,
      "step": 8650
    },
    {
      "entropy": 4.872346067428589,
      "epoch": 2.446988973706531,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004471158544375361,
      "loss": 4.6637,
      "mean_token_accuracy": 0.2522799164056778,
      "num_tokens": 6899533.0,
      "step": 8655
    },
    {
      "entropy": 4.900804662704468,
      "epoch": 2.4484026010743567,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004470495861189232,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24291862696409225,
      "num_tokens": 6903364.0,
      "step": 8660
    },
    {
      "entropy": 4.850862073898315,
      "epoch": 2.4498162284421827,
      "grad_norm": 1.375,
      "learning_rate": 0.0004469832818430106,
      "loss": 4.5933,
      "mean_token_accuracy": 0.25578264743089674,
      "num_tokens": 6907500.0,
      "step": 8665
    },
    {
      "entropy": 4.824142742156982,
      "epoch": 2.4512298558100083,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004469169416236557,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24271387308835984,
      "num_tokens": 6911372.0,
      "step": 8670
    },
    {
      "entropy": 4.865387773513794,
      "epoch": 2.4526434831778343,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004468505654747231,
      "loss": 4.6543,
      "mean_token_accuracy": 0.25014936923980713,
      "num_tokens": 6915439.0,
      "step": 8675
    },
    {
      "entropy": 4.969800806045532,
      "epoch": 2.4540571105456603,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044678415341008497,
      "loss": 4.7284,
      "mean_token_accuracy": 0.25008800327777864,
      "num_tokens": 6919654.0,
      "step": 8680
    },
    {
      "entropy": 4.90244779586792,
      "epoch": 2.455470737913486,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044671770544362107,
      "loss": 4.664,
      "mean_token_accuracy": 0.2549007996916771,
      "num_tokens": 6923505.0,
      "step": 8685
    },
    {
      "entropy": 4.941406679153443,
      "epoch": 2.456884365281312,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004466512215892186,
      "loss": 4.6998,
      "mean_token_accuracy": 0.2479606196284294,
      "num_tokens": 6927436.0,
      "step": 8690
    },
    {
      "entropy": 4.813643980026245,
      "epoch": 2.4582979926491375,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044658470186077233,
      "loss": 4.6905,
      "mean_token_accuracy": 0.25939424335956573,
      "num_tokens": 6931680.0,
      "step": 8695
    },
    {
      "entropy": 4.838484001159668,
      "epoch": 2.4597116200169635,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044651814627218433,
      "loss": 4.6191,
      "mean_token_accuracy": 0.2553516626358032,
      "num_tokens": 6935855.0,
      "step": 8700
    },
    {
      "entropy": 4.895722436904907,
      "epoch": 2.4611252473847895,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004464515548373645,
      "loss": 4.6312,
      "mean_token_accuracy": 0.2533067375421524,
      "num_tokens": 6939861.0,
      "step": 8705
    },
    {
      "entropy": 4.865724897384643,
      "epoch": 2.462538874752615,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004463849275702299,
      "loss": 4.6361,
      "mean_token_accuracy": 0.25706158578395844,
      "num_tokens": 6944166.0,
      "step": 8710
    },
    {
      "entropy": 4.793997430801392,
      "epoch": 2.463952502120441,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004463182644847052,
      "loss": 4.6099,
      "mean_token_accuracy": 0.25378126800060274,
      "num_tokens": 6948009.0,
      "step": 8715
    },
    {
      "entropy": 4.8703857421875,
      "epoch": 2.4653661294882667,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004462515655947227,
      "loss": 4.687,
      "mean_token_accuracy": 0.2531363725662231,
      "num_tokens": 6952269.0,
      "step": 8720
    },
    {
      "entropy": 4.8666540622711185,
      "epoch": 2.4667797568560927,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004461848309142219,
      "loss": 4.6671,
      "mean_token_accuracy": 0.24808001518249512,
      "num_tokens": 6956086.0,
      "step": 8725
    },
    {
      "entropy": 5.000027227401733,
      "epoch": 2.4681933842239188,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044611806045715004,
      "loss": 4.7558,
      "mean_token_accuracy": 0.23270391523838044,
      "num_tokens": 6960180.0,
      "step": 8730
    },
    {
      "entropy": 4.953352308273315,
      "epoch": 2.4696070115917443,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004460512542374617,
      "loss": 4.7161,
      "mean_token_accuracy": 0.2411133676767349,
      "num_tokens": 6964352.0,
      "step": 8735
    },
    {
      "entropy": 4.933160829544067,
      "epoch": 2.4710206389595704,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044598441226911903,
      "loss": 4.7847,
      "mean_token_accuracy": 0.23628398925065994,
      "num_tokens": 6968247.0,
      "step": 8740
    },
    {
      "entropy": 4.857762384414673,
      "epoch": 2.472434266327396,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044591753456609147,
      "loss": 4.5883,
      "mean_token_accuracy": 0.25725068598985673,
      "num_tokens": 6972407.0,
      "step": 8745
    },
    {
      "entropy": 4.86319637298584,
      "epoch": 2.473847893695222,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044585062114235616,
      "loss": 4.6823,
      "mean_token_accuracy": 0.24494930803775788,
      "num_tokens": 6976280.0,
      "step": 8750
    },
    {
      "entropy": 4.9038026332855225,
      "epoch": 2.4752615210630475,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044578367201189746,
      "loss": 4.7098,
      "mean_token_accuracy": 0.24665283411741257,
      "num_tokens": 6980322.0,
      "step": 8755
    },
    {
      "entropy": 5.067902231216431,
      "epoch": 2.4766751484308736,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044571668718870744,
      "loss": 4.8334,
      "mean_token_accuracy": 0.22892946302890776,
      "num_tokens": 6984653.0,
      "step": 8760
    },
    {
      "entropy": 4.841150379180908,
      "epoch": 2.4780887757986996,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004456496666867855,
      "loss": 4.7265,
      "mean_token_accuracy": 0.25348948389291764,
      "num_tokens": 6988483.0,
      "step": 8765
    },
    {
      "entropy": 4.902491807937622,
      "epoch": 2.479502403166525,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004455826105201385,
      "loss": 4.6951,
      "mean_token_accuracy": 0.250818096101284,
      "num_tokens": 6992747.0,
      "step": 8770
    },
    {
      "entropy": 4.847514390945435,
      "epoch": 2.480916030534351,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004455155187027807,
      "loss": 4.5467,
      "mean_token_accuracy": 0.25337847918272016,
      "num_tokens": 6996735.0,
      "step": 8775
    },
    {
      "entropy": 4.92073769569397,
      "epoch": 2.482329657902177,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004454483912487339,
      "loss": 4.6652,
      "mean_token_accuracy": 0.2456749141216278,
      "num_tokens": 7000666.0,
      "step": 8780
    },
    {
      "entropy": 4.865755939483643,
      "epoch": 2.4837432852700028,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044538122817202745,
      "loss": 4.5639,
      "mean_token_accuracy": 0.25727399289608,
      "num_tokens": 7004622.0,
      "step": 8785
    },
    {
      "entropy": 4.821947240829468,
      "epoch": 2.485156912637829,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004453140294866979,
      "loss": 4.6287,
      "mean_token_accuracy": 0.2581113576889038,
      "num_tokens": 7008685.0,
      "step": 8790
    },
    {
      "entropy": 4.71886043548584,
      "epoch": 2.4865705400056544,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044524679520678936,
      "loss": 4.4837,
      "mean_token_accuracy": 0.26534110456705096,
      "num_tokens": 7012756.0,
      "step": 8795
    },
    {
      "entropy": 4.794670724868775,
      "epoch": 2.4879841673734804,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004451795253463534,
      "loss": 4.6613,
      "mean_token_accuracy": 0.2536075636744499,
      "num_tokens": 7016869.0,
      "step": 8800
    },
    {
      "entropy": 4.885772323608398,
      "epoch": 2.489397794741306,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044511221991944903,
      "loss": 4.7315,
      "mean_token_accuracy": 0.24940666258335115,
      "num_tokens": 7020594.0,
      "step": 8805
    },
    {
      "entropy": 4.927555561065674,
      "epoch": 2.490811422109132,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044504487894014253,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2510636493563652,
      "num_tokens": 7024674.0,
      "step": 8810
    },
    {
      "entropy": 4.914074182510376,
      "epoch": 2.492225049476958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044497750242250794,
      "loss": 4.6768,
      "mean_token_accuracy": 0.25540348440408706,
      "num_tokens": 7028302.0,
      "step": 8815
    },
    {
      "entropy": 4.861120080947876,
      "epoch": 2.4936386768447836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004449100903806264,
      "loss": 4.611,
      "mean_token_accuracy": 0.2513492465019226,
      "num_tokens": 7032159.0,
      "step": 8820
    },
    {
      "entropy": 4.898788499832153,
      "epoch": 2.4950523042126096,
      "grad_norm": 1.375,
      "learning_rate": 0.00044484264282858673,
      "loss": 4.612,
      "mean_token_accuracy": 0.25256517380475996,
      "num_tokens": 7036309.0,
      "step": 8825
    },
    {
      "entropy": 4.869152498245239,
      "epoch": 2.4964659315804356,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044477515978048497,
      "loss": 4.6484,
      "mean_token_accuracy": 0.2503778338432312,
      "num_tokens": 7040325.0,
      "step": 8830
    },
    {
      "entropy": 4.797275257110596,
      "epoch": 2.497879558948261,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044470764125042464,
      "loss": 4.615,
      "mean_token_accuracy": 0.25789368003606794,
      "num_tokens": 7044335.0,
      "step": 8835
    },
    {
      "entropy": 4.904232120513916,
      "epoch": 2.499293186316087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044464008725251673,
      "loss": 4.6724,
      "mean_token_accuracy": 0.2513721391558647,
      "num_tokens": 7048703.0,
      "step": 8840
    },
    {
      "entropy": 4.930908298492431,
      "epoch": 2.500706813683913,
      "grad_norm": 1.46875,
      "learning_rate": 0.00044457249780087963,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24476093500852586,
      "num_tokens": 7052733.0,
      "step": 8845
    },
    {
      "entropy": 4.961987829208374,
      "epoch": 2.502120441051739,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044450487290963913,
      "loss": 4.7711,
      "mean_token_accuracy": 0.24654352366924287,
      "num_tokens": 7056969.0,
      "step": 8850
    },
    {
      "entropy": 4.969698572158814,
      "epoch": 2.5035340684195644,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004444372125929283,
      "loss": 4.7727,
      "mean_token_accuracy": 0.24706077873706817,
      "num_tokens": 7060769.0,
      "step": 8855
    },
    {
      "entropy": 4.922748851776123,
      "epoch": 2.5049476957873904,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044436951686488787,
      "loss": 4.5891,
      "mean_token_accuracy": 0.25563936978578566,
      "num_tokens": 7064667.0,
      "step": 8860
    },
    {
      "entropy": 4.943627214431762,
      "epoch": 2.5063613231552164,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004443017857396657,
      "loss": 4.7389,
      "mean_token_accuracy": 0.23616517335176468,
      "num_tokens": 7068675.0,
      "step": 8865
    },
    {
      "entropy": 4.862420272827149,
      "epoch": 2.507774950523042,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044423401923141724,
      "loss": 4.7178,
      "mean_token_accuracy": 0.24815622419118882,
      "num_tokens": 7072628.0,
      "step": 8870
    },
    {
      "entropy": 4.812437534332275,
      "epoch": 2.509188577890868,
      "grad_norm": 1.375,
      "learning_rate": 0.0004441662173543054,
      "loss": 4.6414,
      "mean_token_accuracy": 0.25227170586586,
      "num_tokens": 7076879.0,
      "step": 8875
    },
    {
      "entropy": 4.990383338928223,
      "epoch": 2.510602205258694,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004440983801225001,
      "loss": 4.6697,
      "mean_token_accuracy": 0.23636886179447175,
      "num_tokens": 7080620.0,
      "step": 8880
    },
    {
      "entropy": 4.898235845565796,
      "epoch": 2.5120158326265196,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044403050755017907,
      "loss": 4.7363,
      "mean_token_accuracy": 0.25073603838682174,
      "num_tokens": 7084627.0,
      "step": 8885
    },
    {
      "entropy": 4.773361253738403,
      "epoch": 2.5134294599943456,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004439625996515272,
      "loss": 4.4998,
      "mean_token_accuracy": 0.26744371801614764,
      "num_tokens": 7088427.0,
      "step": 8890
    },
    {
      "entropy": 4.865802526473999,
      "epoch": 2.514843087362171,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044389465644073685,
      "loss": 4.6632,
      "mean_token_accuracy": 0.25766807049512863,
      "num_tokens": 7092372.0,
      "step": 8895
    },
    {
      "entropy": 4.94409818649292,
      "epoch": 2.5162567147299972,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044382667793200776,
      "loss": 4.6954,
      "mean_token_accuracy": 0.2526626616716385,
      "num_tokens": 7096150.0,
      "step": 8900
    },
    {
      "entropy": 4.79977240562439,
      "epoch": 2.517670342097823,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004437586641395469,
      "loss": 4.578,
      "mean_token_accuracy": 0.2526792883872986,
      "num_tokens": 7100209.0,
      "step": 8905
    },
    {
      "entropy": 4.896723079681396,
      "epoch": 2.519083969465649,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004436906150775688,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2398255482316017,
      "num_tokens": 7104278.0,
      "step": 8910
    },
    {
      "entropy": 4.796131801605225,
      "epoch": 2.520497596833475,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044362253076029543,
      "loss": 4.5625,
      "mean_token_accuracy": 0.2633028537034988,
      "num_tokens": 7108139.0,
      "step": 8915
    },
    {
      "entropy": 4.870937156677246,
      "epoch": 2.5219112242013004,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004435544112019557,
      "loss": 4.7218,
      "mean_token_accuracy": 0.24145184010267257,
      "num_tokens": 7112214.0,
      "step": 8920
    },
    {
      "entropy": 4.936765384674072,
      "epoch": 2.5233248515691264,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044348625641678634,
      "loss": 4.7019,
      "mean_token_accuracy": 0.2456716224551201,
      "num_tokens": 7116545.0,
      "step": 8925
    },
    {
      "entropy": 4.839352130889893,
      "epoch": 2.5247384789369525,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004434180664190314,
      "loss": 4.4918,
      "mean_token_accuracy": 0.2711749717593193,
      "num_tokens": 7120519.0,
      "step": 8930
    },
    {
      "entropy": 4.856968641281128,
      "epoch": 2.526152106304778,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000443349841222942,
      "loss": 4.6417,
      "mean_token_accuracy": 0.24725841134786605,
      "num_tokens": 7124567.0,
      "step": 8935
    },
    {
      "entropy": 4.875683927536011,
      "epoch": 2.527565733672604,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044328158084277683,
      "loss": 4.704,
      "mean_token_accuracy": 0.24697965979576111,
      "num_tokens": 7128670.0,
      "step": 8940
    },
    {
      "entropy": 4.895025682449341,
      "epoch": 2.5289793610404296,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044321328529280183,
      "loss": 4.6298,
      "mean_token_accuracy": 0.25382993668317794,
      "num_tokens": 7132504.0,
      "step": 8945
    },
    {
      "entropy": 4.835316562652588,
      "epoch": 2.5303929884082557,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004431449545872904,
      "loss": 4.6107,
      "mean_token_accuracy": 0.2573100358247757,
      "num_tokens": 7136504.0,
      "step": 8950
    },
    {
      "entropy": 4.8149748802185055,
      "epoch": 2.5318066157760812,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044307658874052333,
      "loss": 4.6074,
      "mean_token_accuracy": 0.2522086575627327,
      "num_tokens": 7140253.0,
      "step": 8955
    },
    {
      "entropy": 4.826704216003418,
      "epoch": 2.5332202431439073,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004430081877667886,
      "loss": 4.7218,
      "mean_token_accuracy": 0.24741041362285615,
      "num_tokens": 7144699.0,
      "step": 8960
    },
    {
      "entropy": 4.828955793380738,
      "epoch": 2.5346338705117333,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004429397516803814,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2589452862739563,
      "num_tokens": 7148643.0,
      "step": 8965
    },
    {
      "entropy": 4.8674328327178955,
      "epoch": 2.536047497879559,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004428712804956047,
      "loss": 4.7552,
      "mean_token_accuracy": 0.23673311918973922,
      "num_tokens": 7152619.0,
      "step": 8970
    },
    {
      "entropy": 4.922368907928467,
      "epoch": 2.537461125247385,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004428027742267685,
      "loss": 4.6884,
      "mean_token_accuracy": 0.24896937608718872,
      "num_tokens": 7156543.0,
      "step": 8975
    },
    {
      "entropy": 4.878751182556153,
      "epoch": 2.538874752615211,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044273423288819004,
      "loss": 4.6813,
      "mean_token_accuracy": 0.24544369727373122,
      "num_tokens": 7160440.0,
      "step": 8980
    },
    {
      "entropy": 4.922648000717163,
      "epoch": 2.5402883799830365,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004426656564941942,
      "loss": 4.7599,
      "mean_token_accuracy": 0.2501348614692688,
      "num_tokens": 7164585.0,
      "step": 8985
    },
    {
      "entropy": 4.879630422592163,
      "epoch": 2.541702007350862,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000442597045059113,
      "loss": 4.6812,
      "mean_token_accuracy": 0.25576352924108503,
      "num_tokens": 7168527.0,
      "step": 8990
    },
    {
      "entropy": 4.981833839416504,
      "epoch": 2.543115634718688,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004425283985972857,
      "loss": 4.6814,
      "mean_token_accuracy": 0.24809899777173997,
      "num_tokens": 7172494.0,
      "step": 8995
    },
    {
      "entropy": 4.89121379852295,
      "epoch": 2.544529262086514,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004424597171230591,
      "loss": 4.698,
      "mean_token_accuracy": 0.24869736582040786,
      "num_tokens": 7176627.0,
      "step": 9000
    },
    {
      "epoch": 2.544529262086514,
      "eval_entropy": 4.651766316084723,
      "eval_loss": 5.0735063552856445,
      "eval_mean_token_accuracy": 0.24150473008250448,
      "eval_num_tokens": 7176627.0,
      "eval_runtime": 4.8914,
      "eval_samples_per_second": 2306.917,
      "eval_steps_per_second": 288.467,
      "step": 9000
    },
    {
      "entropy": 4.814825820922851,
      "epoch": 2.5459428894543397,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044239100065078714,
      "loss": 4.6328,
      "mean_token_accuracy": 0.25143989324569704,
      "num_tokens": 7180398.0,
      "step": 9005
    },
    {
      "entropy": 4.899396133422852,
      "epoch": 2.5473565168221657,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044232224919483115,
      "loss": 4.6839,
      "mean_token_accuracy": 0.25714144855737686,
      "num_tokens": 7184182.0,
      "step": 9010
    },
    {
      "entropy": 4.866952562332154,
      "epoch": 2.5487701441899917,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004422534627695598,
      "loss": 4.609,
      "mean_token_accuracy": 0.24941940009593963,
      "num_tokens": 7188139.0,
      "step": 9015
    },
    {
      "entropy": 4.846384572982788,
      "epoch": 2.5501837715578173,
      "grad_norm": 1.453125,
      "learning_rate": 0.000442184641389349,
      "loss": 4.7265,
      "mean_token_accuracy": 0.24951156377792358,
      "num_tokens": 7191848.0,
      "step": 9020
    },
    {
      "entropy": 4.957627153396606,
      "epoch": 2.5515973989256433,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004421157850685819,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2342834770679474,
      "num_tokens": 7196039.0,
      "step": 9025
    },
    {
      "entropy": 4.855026960372925,
      "epoch": 2.553011026293469,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044204689382164927,
      "loss": 4.6464,
      "mean_token_accuracy": 0.24672157019376756,
      "num_tokens": 7200075.0,
      "step": 9030
    },
    {
      "entropy": 4.892865419387817,
      "epoch": 2.554424653661295,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004419779676629487,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2422384202480316,
      "num_tokens": 7204236.0,
      "step": 9035
    },
    {
      "entropy": 4.889660978317261,
      "epoch": 2.5558382810291205,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004419090066068856,
      "loss": 4.6657,
      "mean_token_accuracy": 0.24387118369340896,
      "num_tokens": 7208390.0,
      "step": 9040
    },
    {
      "entropy": 4.808865737915039,
      "epoch": 2.5572519083969465,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004418400106678721,
      "loss": 4.5392,
      "mean_token_accuracy": 0.2600240930914879,
      "num_tokens": 7212355.0,
      "step": 9045
    },
    {
      "entropy": 4.867876052856445,
      "epoch": 2.5586655357647725,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044177097986032815,
      "loss": 4.6605,
      "mean_token_accuracy": 0.24762177914381028,
      "num_tokens": 7216336.0,
      "step": 9050
    },
    {
      "entropy": 4.8872462749481205,
      "epoch": 2.560079163132598,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044170191419868064,
      "loss": 4.7589,
      "mean_token_accuracy": 0.2379223108291626,
      "num_tokens": 7220115.0,
      "step": 9055
    },
    {
      "entropy": 4.865660905838013,
      "epoch": 2.561492790500424,
      "grad_norm": 1.34375,
      "learning_rate": 0.000441632813697364,
      "loss": 4.6069,
      "mean_token_accuracy": 0.25413164496421814,
      "num_tokens": 7224230.0,
      "step": 9060
    },
    {
      "entropy": 4.927585506439209,
      "epoch": 2.56290641786825,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044156367837081966,
      "loss": 4.7187,
      "mean_token_accuracy": 0.2507879137992859,
      "num_tokens": 7228186.0,
      "step": 9065
    },
    {
      "entropy": 4.920462369918823,
      "epoch": 2.5643200452360757,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004414945082334966,
      "loss": 4.7009,
      "mean_token_accuracy": 0.24978994727134704,
      "num_tokens": 7231953.0,
      "step": 9070
    },
    {
      "entropy": 4.842449569702149,
      "epoch": 2.5657336726039017,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004414253032998508,
      "loss": 4.5687,
      "mean_token_accuracy": 0.25420765280723573,
      "num_tokens": 7235810.0,
      "step": 9075
    },
    {
      "entropy": 4.880776262283325,
      "epoch": 2.5671472999717273,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004413560635843458,
      "loss": 4.6503,
      "mean_token_accuracy": 0.24720506966114045,
      "num_tokens": 7239879.0,
      "step": 9080
    },
    {
      "entropy": 4.940559053421021,
      "epoch": 2.5685609273395533,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044128678910145225,
      "loss": 4.734,
      "mean_token_accuracy": 0.24640875607728957,
      "num_tokens": 7243948.0,
      "step": 9085
    },
    {
      "entropy": 4.781366586685181,
      "epoch": 2.569974554707379,
      "grad_norm": 1.390625,
      "learning_rate": 0.000441217479865648,
      "loss": 4.4671,
      "mean_token_accuracy": 0.27382553517818453,
      "num_tokens": 7248024.0,
      "step": 9090
    },
    {
      "entropy": 4.88687219619751,
      "epoch": 2.571388182075205,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004411481358914184,
      "loss": 4.64,
      "mean_token_accuracy": 0.24446267634630203,
      "num_tokens": 7252130.0,
      "step": 9095
    },
    {
      "entropy": 4.825691890716553,
      "epoch": 2.572801809443031,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004410787571932558,
      "loss": 4.7012,
      "mean_token_accuracy": 0.2488132268190384,
      "num_tokens": 7256326.0,
      "step": 9100
    },
    {
      "entropy": 4.89831132888794,
      "epoch": 2.5742154368108565,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004410093437856599,
      "loss": 4.7244,
      "mean_token_accuracy": 0.24187981933355332,
      "num_tokens": 7260507.0,
      "step": 9105
    },
    {
      "entropy": 4.9437628269195555,
      "epoch": 2.5756290641786825,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044093989568313774,
      "loss": 4.5726,
      "mean_token_accuracy": 0.25567595213651656,
      "num_tokens": 7264805.0,
      "step": 9110
    },
    {
      "entropy": 4.894422245025635,
      "epoch": 2.5770426915465086,
      "grad_norm": 1.390625,
      "learning_rate": 0.00044087041290020356,
      "loss": 4.5706,
      "mean_token_accuracy": 0.2521216720342636,
      "num_tokens": 7268699.0,
      "step": 9115
    },
    {
      "entropy": 4.795257186889648,
      "epoch": 2.578456318914334,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004408008954513787,
      "loss": 4.52,
      "mean_token_accuracy": 0.2674206256866455,
      "num_tokens": 7272601.0,
      "step": 9120
    },
    {
      "entropy": 4.835134506225586,
      "epoch": 2.57986994628216,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004407313433511921,
      "loss": 4.7392,
      "mean_token_accuracy": 0.2401050567626953,
      "num_tokens": 7276570.0,
      "step": 9125
    },
    {
      "entropy": 4.8011829376220705,
      "epoch": 2.5812835736499857,
      "grad_norm": 1.421875,
      "learning_rate": 0.00044066175661417947,
      "loss": 4.6574,
      "mean_token_accuracy": 0.24744735211133956,
      "num_tokens": 7280215.0,
      "step": 9130
    },
    {
      "entropy": 4.78481502532959,
      "epoch": 2.5826972010178118,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004405921352548842,
      "loss": 4.561,
      "mean_token_accuracy": 0.2702275961637497,
      "num_tokens": 7284148.0,
      "step": 9135
    },
    {
      "entropy": 4.880260229110718,
      "epoch": 2.5841108283856373,
      "grad_norm": 1.375,
      "learning_rate": 0.0004405224792878567,
      "loss": 4.6768,
      "mean_token_accuracy": 0.246889166533947,
      "num_tokens": 7288408.0,
      "step": 9140
    },
    {
      "entropy": 4.906869220733642,
      "epoch": 2.5855244557534633,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004404527887276545,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2409200116991997,
      "num_tokens": 7292157.0,
      "step": 9145
    },
    {
      "entropy": 4.872857332229614,
      "epoch": 2.5869380831212894,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004403830635888426,
      "loss": 4.6362,
      "mean_token_accuracy": 0.2514513969421387,
      "num_tokens": 7296193.0,
      "step": 9150
    },
    {
      "entropy": 4.774008941650391,
      "epoch": 2.588351710489115,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00044031330388599315,
      "loss": 4.6197,
      "mean_token_accuracy": 0.25086003392934797,
      "num_tokens": 7300186.0,
      "step": 9155
    },
    {
      "entropy": 4.807612657546997,
      "epoch": 2.589765337856941,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004402435096336854,
      "loss": 4.6837,
      "mean_token_accuracy": 0.24396176785230636,
      "num_tokens": 7304559.0,
      "step": 9160
    },
    {
      "entropy": 4.806529760360718,
      "epoch": 2.591178965224767,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044017368084650605,
      "loss": 4.6221,
      "mean_token_accuracy": 0.25230983793735506,
      "num_tokens": 7308827.0,
      "step": 9165
    },
    {
      "entropy": 4.834507322311401,
      "epoch": 2.5925925925925926,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004401038175390488,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2552493214607239,
      "num_tokens": 7313038.0,
      "step": 9170
    },
    {
      "entropy": 4.870154094696045,
      "epoch": 2.5940062199604186,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044003391972591456,
      "loss": 4.544,
      "mean_token_accuracy": 0.26403322964906695,
      "num_tokens": 7316913.0,
      "step": 9175
    },
    {
      "entropy": 4.794100427627564,
      "epoch": 2.595419847328244,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004399639874217117,
      "loss": 4.6372,
      "mean_token_accuracy": 0.2534163534641266,
      "num_tokens": 7320634.0,
      "step": 9180
    },
    {
      "entropy": 4.79074273109436,
      "epoch": 2.59683347469607,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043989402064105557,
      "loss": 4.6548,
      "mean_token_accuracy": 0.24706927090883254,
      "num_tokens": 7324663.0,
      "step": 9185
    },
    {
      "entropy": 4.749100017547607,
      "epoch": 2.5982471020638958,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043982401939856875,
      "loss": 4.5378,
      "mean_token_accuracy": 0.2624316975474358,
      "num_tokens": 7328814.0,
      "step": 9190
    },
    {
      "entropy": 4.8486096382141115,
      "epoch": 2.5996607294317218,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043975398370888115,
      "loss": 4.5536,
      "mean_token_accuracy": 0.2561468482017517,
      "num_tokens": 7332780.0,
      "step": 9195
    },
    {
      "entropy": 4.8892378330230715,
      "epoch": 2.601074356799548,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004396839135866298,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23313588798046112,
      "num_tokens": 7336769.0,
      "step": 9200
    },
    {
      "entropy": 4.891507387161255,
      "epoch": 2.6024879841673734,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004396138090464589,
      "loss": 4.6203,
      "mean_token_accuracy": 0.2546934008598328,
      "num_tokens": 7340943.0,
      "step": 9205
    },
    {
      "entropy": 4.82293062210083,
      "epoch": 2.6039016115351994,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004395436701030198,
      "loss": 4.6659,
      "mean_token_accuracy": 0.2534699097275734,
      "num_tokens": 7344944.0,
      "step": 9210
    },
    {
      "entropy": 4.896666574478149,
      "epoch": 2.6053152389030254,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043947349677097116,
      "loss": 4.719,
      "mean_token_accuracy": 0.2388124480843544,
      "num_tokens": 7348778.0,
      "step": 9215
    },
    {
      "entropy": 4.818767070770264,
      "epoch": 2.606728866270851,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043940328906497876,
      "loss": 4.6156,
      "mean_token_accuracy": 0.25418018400669096,
      "num_tokens": 7352878.0,
      "step": 9220
    },
    {
      "entropy": 4.882868051528931,
      "epoch": 2.608142493638677,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004393330469997156,
      "loss": 4.6416,
      "mean_token_accuracy": 0.2535845249891281,
      "num_tokens": 7356825.0,
      "step": 9225
    },
    {
      "entropy": 4.741895961761474,
      "epoch": 2.6095561210065026,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004392627705898619,
      "loss": 4.5531,
      "mean_token_accuracy": 0.2733841747045517,
      "num_tokens": 7360603.0,
      "step": 9230
    },
    {
      "entropy": 4.886696147918701,
      "epoch": 2.6109697483743286,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00043919245985010475,
      "loss": 4.7764,
      "mean_token_accuracy": 0.24133803844451904,
      "num_tokens": 7364757.0,
      "step": 9235
    },
    {
      "entropy": 4.822178840637207,
      "epoch": 2.612383375742154,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043912211479513896,
      "loss": 4.5951,
      "mean_token_accuracy": 0.25403084307909013,
      "num_tokens": 7368581.0,
      "step": 9240
    },
    {
      "entropy": 4.930347204208374,
      "epoch": 2.61379700310998,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000439051735439666,
      "loss": 4.6802,
      "mean_token_accuracy": 0.2557662710547447,
      "num_tokens": 7372448.0,
      "step": 9245
    },
    {
      "entropy": 4.898662996292114,
      "epoch": 2.615210630477806,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004389813217983949,
      "loss": 4.6147,
      "mean_token_accuracy": 0.2542915537953377,
      "num_tokens": 7376464.0,
      "step": 9250
    },
    {
      "entropy": 4.914443159103394,
      "epoch": 2.616624257845632,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004389108738860415,
      "loss": 4.8479,
      "mean_token_accuracy": 0.2297476723790169,
      "num_tokens": 7380445.0,
      "step": 9255
    },
    {
      "entropy": 4.926639223098755,
      "epoch": 2.618037885213458,
      "grad_norm": 1.34375,
      "learning_rate": 0.000438840391717329,
      "loss": 4.7512,
      "mean_token_accuracy": 0.24162334501743316,
      "num_tokens": 7384848.0,
      "step": 9260
    },
    {
      "entropy": 4.857936000823974,
      "epoch": 2.619451512581284,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043876987530698785,
      "loss": 4.6127,
      "mean_token_accuracy": 0.25807690918445586,
      "num_tokens": 7388665.0,
      "step": 9265
    },
    {
      "entropy": 4.8431836605072025,
      "epoch": 2.6208651399491094,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043869932466975546,
      "loss": 4.6041,
      "mean_token_accuracy": 0.25482713282108305,
      "num_tokens": 7392807.0,
      "step": 9270
    },
    {
      "entropy": 4.78021388053894,
      "epoch": 2.622278767316935,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004386287398203766,
      "loss": 4.5382,
      "mean_token_accuracy": 0.2745848923921585,
      "num_tokens": 7396592.0,
      "step": 9275
    },
    {
      "entropy": 4.817171669006347,
      "epoch": 2.623692394684761,
      "grad_norm": 1.375,
      "learning_rate": 0.0004385581207736029,
      "loss": 4.6105,
      "mean_token_accuracy": 0.24991951882839203,
      "num_tokens": 7400540.0,
      "step": 9280
    },
    {
      "entropy": 4.810059642791748,
      "epoch": 2.625106022052587,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043848746754419335,
      "loss": 4.6042,
      "mean_token_accuracy": 0.2541697800159454,
      "num_tokens": 7404467.0,
      "step": 9285
    },
    {
      "entropy": 4.864866447448731,
      "epoch": 2.6265196494204126,
      "grad_norm": 1.421875,
      "learning_rate": 0.000438416780146914,
      "loss": 4.5793,
      "mean_token_accuracy": 0.25637721121311186,
      "num_tokens": 7408377.0,
      "step": 9290
    },
    {
      "entropy": 4.842103433609009,
      "epoch": 2.6279332767882386,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00043834605859653826,
      "loss": 4.5537,
      "mean_token_accuracy": 0.2674950405955315,
      "num_tokens": 7412375.0,
      "step": 9295
    },
    {
      "entropy": 4.916817045211792,
      "epoch": 2.6293469041560646,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043827530290784634,
      "loss": 4.754,
      "mean_token_accuracy": 0.22751364558935167,
      "num_tokens": 7416613.0,
      "step": 9300
    },
    {
      "entropy": 4.885684061050415,
      "epoch": 2.6307605315238902,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00043820451309562576,
      "loss": 4.6172,
      "mean_token_accuracy": 0.2574413910508156,
      "num_tokens": 7420558.0,
      "step": 9305
    },
    {
      "entropy": 4.7922276020050045,
      "epoch": 2.6321741588917162,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004381336891746711,
      "loss": 4.6325,
      "mean_token_accuracy": 0.2563048005104065,
      "num_tokens": 7424394.0,
      "step": 9310
    },
    {
      "entropy": 4.999895906448364,
      "epoch": 2.633587786259542,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004380628311597843,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22801423817873,
      "num_tokens": 7428751.0,
      "step": 9315
    },
    {
      "entropy": 4.984036016464233,
      "epoch": 2.635001413627368,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043799193906577413,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23314179927110673,
      "num_tokens": 7432816.0,
      "step": 9320
    },
    {
      "entropy": 4.895705127716065,
      "epoch": 2.6364150409951934,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043792101290745654,
      "loss": 4.5808,
      "mean_token_accuracy": 0.2637035608291626,
      "num_tokens": 7436754.0,
      "step": 9325
    },
    {
      "entropy": 4.86237154006958,
      "epoch": 2.6378286683630194,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004378500526996548,
      "loss": 4.6525,
      "mean_token_accuracy": 0.2554896458983421,
      "num_tokens": 7440742.0,
      "step": 9330
    },
    {
      "entropy": 4.793785762786865,
      "epoch": 2.6392422957308455,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000437779058457199,
      "loss": 4.6,
      "mean_token_accuracy": 0.261631141602993,
      "num_tokens": 7444565.0,
      "step": 9335
    },
    {
      "entropy": 4.84995470046997,
      "epoch": 2.640655923098671,
      "grad_norm": 1.375,
      "learning_rate": 0.00043770803019492667,
      "loss": 4.692,
      "mean_token_accuracy": 0.23831041008234025,
      "num_tokens": 7448510.0,
      "step": 9340
    },
    {
      "entropy": 4.951950025558472,
      "epoch": 2.642069550466497,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004376369679276821,
      "loss": 4.6137,
      "mean_token_accuracy": 0.26139389127492907,
      "num_tokens": 7452536.0,
      "step": 9345
    },
    {
      "entropy": 4.90651388168335,
      "epoch": 2.643483177834323,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000437565871670317,
      "loss": 4.6711,
      "mean_token_accuracy": 0.25623977482318877,
      "num_tokens": 7456245.0,
      "step": 9350
    },
    {
      "entropy": 4.829598808288575,
      "epoch": 2.6448968052021486,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004374947414376899,
      "loss": 4.605,
      "mean_token_accuracy": 0.25036528557538984,
      "num_tokens": 7460306.0,
      "step": 9355
    },
    {
      "entropy": 4.8269976615905765,
      "epoch": 2.6463104325699747,
      "grad_norm": 1.375,
      "learning_rate": 0.0004374235772446668,
      "loss": 4.6897,
      "mean_token_accuracy": 0.2526609808206558,
      "num_tokens": 7464103.0,
      "step": 9360
    },
    {
      "entropy": 4.930158996582032,
      "epoch": 2.6477240599378002,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004373523791061204,
      "loss": 4.671,
      "mean_token_accuracy": 0.2432880535721779,
      "num_tokens": 7468316.0,
      "step": 9365
    },
    {
      "entropy": 4.985217618942261,
      "epoch": 2.6491376873056263,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043728114703693073,
      "loss": 4.8131,
      "mean_token_accuracy": 0.249200177192688,
      "num_tokens": 7472254.0,
      "step": 9370
    },
    {
      "entropy": 4.880460786819458,
      "epoch": 2.650551314673452,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004372098810519849,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2538872748613358,
      "num_tokens": 7476345.0,
      "step": 9375
    },
    {
      "entropy": 4.792589330673218,
      "epoch": 2.651964942041278,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004371385811661769,
      "loss": 4.6041,
      "mean_token_accuracy": 0.25059485137462617,
      "num_tokens": 7480210.0,
      "step": 9380
    },
    {
      "entropy": 4.836337566375732,
      "epoch": 2.653378569409104,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043706724739440815,
      "loss": 4.6352,
      "mean_token_accuracy": 0.2516987144947052,
      "num_tokens": 7483861.0,
      "step": 9385
    },
    {
      "entropy": 4.755267381668091,
      "epoch": 2.6547921967769295,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004369958797515869,
      "loss": 4.5686,
      "mean_token_accuracy": 0.25968614816665647,
      "num_tokens": 7488041.0,
      "step": 9390
    },
    {
      "entropy": 4.82827672958374,
      "epoch": 2.6562058241447555,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004369244782526284,
      "loss": 4.6486,
      "mean_token_accuracy": 0.25019073486328125,
      "num_tokens": 7492122.0,
      "step": 9395
    },
    {
      "entropy": 4.828032064437866,
      "epoch": 2.6576194515125815,
      "grad_norm": 1.375,
      "learning_rate": 0.00043685304291245545,
      "loss": 4.5224,
      "mean_token_accuracy": 0.26409244537353516,
      "num_tokens": 7496102.0,
      "step": 9400
    },
    {
      "entropy": 4.855035352706909,
      "epoch": 2.659033078880407,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043678157374599724,
      "loss": 4.6996,
      "mean_token_accuracy": 0.24786279648542403,
      "num_tokens": 7500084.0,
      "step": 9405
    },
    {
      "entropy": 4.804668283462524,
      "epoch": 2.660446706248233,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043671007076819066,
      "loss": 4.5726,
      "mean_token_accuracy": 0.25843910723924635,
      "num_tokens": 7504068.0,
      "step": 9410
    },
    {
      "entropy": 4.832597303390503,
      "epoch": 2.6618603336160587,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043663853399397923,
      "loss": 4.5592,
      "mean_token_accuracy": 0.25143383592367174,
      "num_tokens": 7507903.0,
      "step": 9415
    },
    {
      "entropy": 4.851873588562012,
      "epoch": 2.6632739609838847,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043656696343831377,
      "loss": 4.6256,
      "mean_token_accuracy": 0.24687854945659637,
      "num_tokens": 7511694.0,
      "step": 9420
    },
    {
      "entropy": 4.80516881942749,
      "epoch": 2.6646875883517103,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000436495359116152,
      "loss": 4.5701,
      "mean_token_accuracy": 0.26000512987375257,
      "num_tokens": 7516120.0,
      "step": 9425
    },
    {
      "entropy": 4.850139045715332,
      "epoch": 2.6661012157195363,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043642372104245883,
      "loss": 4.5805,
      "mean_token_accuracy": 0.2457722619175911,
      "num_tokens": 7520068.0,
      "step": 9430
    },
    {
      "entropy": 4.844283151626587,
      "epoch": 2.6675148430873623,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004363520492322062,
      "loss": 4.6202,
      "mean_token_accuracy": 0.26166790574789045,
      "num_tokens": 7524136.0,
      "step": 9435
    },
    {
      "entropy": 4.875305986404419,
      "epoch": 2.668928470455188,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000436280343700373,
      "loss": 4.5858,
      "mean_token_accuracy": 0.25276473313570025,
      "num_tokens": 7528026.0,
      "step": 9440
    },
    {
      "entropy": 4.794535684585571,
      "epoch": 2.670342097823014,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004362086044619454,
      "loss": 4.5693,
      "mean_token_accuracy": 0.2539200857281685,
      "num_tokens": 7532149.0,
      "step": 9445
    },
    {
      "entropy": 4.86738395690918,
      "epoch": 2.67175572519084,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004361368315319162,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24012451320886613,
      "num_tokens": 7536270.0,
      "step": 9450
    },
    {
      "entropy": 4.823611927032471,
      "epoch": 2.6731693525586655,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043606502492528576,
      "loss": 4.6646,
      "mean_token_accuracy": 0.2593794628977776,
      "num_tokens": 7540157.0,
      "step": 9455
    },
    {
      "entropy": 4.863368034362793,
      "epoch": 2.6745829799264915,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004359931846570611,
      "loss": 4.6796,
      "mean_token_accuracy": 0.25464999824762347,
      "num_tokens": 7543861.0,
      "step": 9460
    },
    {
      "entropy": 4.839675378799439,
      "epoch": 2.675996607294317,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004359213107422564,
      "loss": 4.5975,
      "mean_token_accuracy": 0.25951915681362153,
      "num_tokens": 7547709.0,
      "step": 9465
    },
    {
      "entropy": 4.929944944381714,
      "epoch": 2.677410234662143,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004358494031958929,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24460129886865617,
      "num_tokens": 7551773.0,
      "step": 9470
    },
    {
      "entropy": 4.832008743286133,
      "epoch": 2.6788238620299687,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00043577746203299865,
      "loss": 4.6812,
      "mean_token_accuracy": 0.2473175898194313,
      "num_tokens": 7555716.0,
      "step": 9475
    },
    {
      "entropy": 5.034928655624389,
      "epoch": 2.6802374893977947,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004357054872686092,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2232452303171158,
      "num_tokens": 7559583.0,
      "step": 9480
    },
    {
      "entropy": 4.832053852081299,
      "epoch": 2.6816511167656207,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004356334789177666,
      "loss": 4.5481,
      "mean_token_accuracy": 0.26879881173372266,
      "num_tokens": 7563484.0,
      "step": 9485
    },
    {
      "entropy": 4.761224699020386,
      "epoch": 2.6830647441334463,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043556143699552034,
      "loss": 4.6223,
      "mean_token_accuracy": 0.24982294887304307,
      "num_tokens": 7567196.0,
      "step": 9490
    },
    {
      "entropy": 4.825017642974854,
      "epoch": 2.6844783715012723,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004354893615169265,
      "loss": 4.7234,
      "mean_token_accuracy": 0.24304340183734893,
      "num_tokens": 7571161.0,
      "step": 9495
    },
    {
      "entropy": 4.920314788818359,
      "epoch": 2.6858919988690984,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043541725249704856,
      "loss": 4.7341,
      "mean_token_accuracy": 0.24500496238470076,
      "num_tokens": 7575226.0,
      "step": 9500
    },
    {
      "entropy": 4.912919664382935,
      "epoch": 2.687305626236924,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004353451099509569,
      "loss": 4.6977,
      "mean_token_accuracy": 0.24860414415597915,
      "num_tokens": 7579444.0,
      "step": 9505
    },
    {
      "entropy": 4.825108289718628,
      "epoch": 2.68871925360475,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004352729338937288,
      "loss": 4.7177,
      "mean_token_accuracy": 0.2513407036662102,
      "num_tokens": 7583298.0,
      "step": 9510
    },
    {
      "entropy": 4.8723170280456545,
      "epoch": 2.6901328809725755,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004352007243404486,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24303073436021805,
      "num_tokens": 7587292.0,
      "step": 9515
    },
    {
      "entropy": 4.8941077709198,
      "epoch": 2.6915465083404015,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043512848130620763,
      "loss": 4.6552,
      "mean_token_accuracy": 0.2505094468593597,
      "num_tokens": 7591427.0,
      "step": 9520
    },
    {
      "entropy": 4.895117855072021,
      "epoch": 2.692960135708227,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043505620480610433,
      "loss": 4.7143,
      "mean_token_accuracy": 0.2441198617219925,
      "num_tokens": 7595121.0,
      "step": 9525
    },
    {
      "entropy": 4.912750768661499,
      "epoch": 2.694373763076053,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043498389485524395,
      "loss": 4.7014,
      "mean_token_accuracy": 0.23493561893701553,
      "num_tokens": 7599072.0,
      "step": 9530
    },
    {
      "entropy": 4.899863290786743,
      "epoch": 2.695787390443879,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043491155146873885,
      "loss": 4.6874,
      "mean_token_accuracy": 0.2449964851140976,
      "num_tokens": 7602860.0,
      "step": 9535
    },
    {
      "entropy": 4.883940744400024,
      "epoch": 2.6972010178117047,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004348391746617083,
      "loss": 4.6791,
      "mean_token_accuracy": 0.23739833384752274,
      "num_tokens": 7606856.0,
      "step": 9540
    },
    {
      "entropy": 4.895227670669556,
      "epoch": 2.6986146451795308,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004347667644492788,
      "loss": 4.7185,
      "mean_token_accuracy": 0.2424500674009323,
      "num_tokens": 7611148.0,
      "step": 9545
    },
    {
      "entropy": 4.914761114120483,
      "epoch": 2.700028272547357,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043469432084658336,
      "loss": 4.6552,
      "mean_token_accuracy": 0.24506007134914398,
      "num_tokens": 7615121.0,
      "step": 9550
    },
    {
      "entropy": 4.814155626296997,
      "epoch": 2.7014418999151824,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043462184386876246,
      "loss": 4.592,
      "mean_token_accuracy": 0.257502855360508,
      "num_tokens": 7619009.0,
      "step": 9555
    },
    {
      "entropy": 4.801906394958496,
      "epoch": 2.702855527283008,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004345493335309632,
      "loss": 4.625,
      "mean_token_accuracy": 0.2462763398885727,
      "num_tokens": 7623089.0,
      "step": 9560
    },
    {
      "entropy": 4.7804030418396,
      "epoch": 2.704269154650834,
      "grad_norm": 1.515625,
      "learning_rate": 0.00043447678984833994,
      "loss": 4.6062,
      "mean_token_accuracy": 0.26039413958787916,
      "num_tokens": 7626880.0,
      "step": 9565
    },
    {
      "entropy": 4.869806957244873,
      "epoch": 2.70568278201866,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004344042128360537,
      "loss": 4.6678,
      "mean_token_accuracy": 0.24891093671321868,
      "num_tokens": 7631050.0,
      "step": 9570
    },
    {
      "entropy": 4.859990167617798,
      "epoch": 2.7070964093864855,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043433160250927277,
      "loss": 4.7115,
      "mean_token_accuracy": 0.25623124539852143,
      "num_tokens": 7635306.0,
      "step": 9575
    },
    {
      "entropy": 4.883833980560302,
      "epoch": 2.7085100367543116,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004342589588831721,
      "loss": 4.6784,
      "mean_token_accuracy": 0.244734226167202,
      "num_tokens": 7639260.0,
      "step": 9580
    },
    {
      "entropy": 4.915551424026489,
      "epoch": 2.7099236641221376,
      "grad_norm": 1.375,
      "learning_rate": 0.000434186281972934,
      "loss": 4.7586,
      "mean_token_accuracy": 0.24660334587097169,
      "num_tokens": 7643241.0,
      "step": 9585
    },
    {
      "entropy": 4.877920341491699,
      "epoch": 2.711337291489963,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004341135717937473,
      "loss": 4.59,
      "mean_token_accuracy": 0.2534252434968948,
      "num_tokens": 7647204.0,
      "step": 9590
    },
    {
      "entropy": 4.886265707015991,
      "epoch": 2.712750918857789,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043404082836080796,
      "loss": 4.6508,
      "mean_token_accuracy": 0.24095214009284974,
      "num_tokens": 7651469.0,
      "step": 9595
    },
    {
      "entropy": 4.841483306884766,
      "epoch": 2.714164546225615,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000433968051689319,
      "loss": 4.6604,
      "mean_token_accuracy": 0.256812684237957,
      "num_tokens": 7655291.0,
      "step": 9600
    },
    {
      "entropy": 4.81247844696045,
      "epoch": 2.715578173593441,
      "grad_norm": 1.453125,
      "learning_rate": 0.00043389524179449025,
      "loss": 4.6105,
      "mean_token_accuracy": 0.2651200845837593,
      "num_tokens": 7659381.0,
      "step": 9605
    },
    {
      "entropy": 4.908606100082397,
      "epoch": 2.7169918009612664,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043382239869153855,
      "loss": 4.664,
      "mean_token_accuracy": 0.24466655105352403,
      "num_tokens": 7663197.0,
      "step": 9610
    },
    {
      "entropy": 4.8344074249267575,
      "epoch": 2.7184054283290924,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043374952239568767,
      "loss": 4.6373,
      "mean_token_accuracy": 0.24169098436832429,
      "num_tokens": 7667445.0,
      "step": 9615
    },
    {
      "entropy": 4.921321296691895,
      "epoch": 2.7198190556969184,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043367661292216823,
      "loss": 4.6926,
      "mean_token_accuracy": 0.24388381987810134,
      "num_tokens": 7671184.0,
      "step": 9620
    },
    {
      "entropy": 4.862404155731201,
      "epoch": 2.721232683064744,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000433603670286218,
      "loss": 4.5957,
      "mean_token_accuracy": 0.258972005546093,
      "num_tokens": 7674911.0,
      "step": 9625
    },
    {
      "entropy": 4.844997072219849,
      "epoch": 2.72264631043257,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043353069450308146,
      "loss": 4.6691,
      "mean_token_accuracy": 0.25384822487831116,
      "num_tokens": 7678751.0,
      "step": 9630
    },
    {
      "entropy": 4.850261068344116,
      "epoch": 2.724059937800396,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00043345768558800997,
      "loss": 4.5652,
      "mean_token_accuracy": 0.2643941193819046,
      "num_tokens": 7682623.0,
      "step": 9635
    },
    {
      "entropy": 4.80199966430664,
      "epoch": 2.7254735651682216,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004333846435562622,
      "loss": 4.566,
      "mean_token_accuracy": 0.2632409855723381,
      "num_tokens": 7686590.0,
      "step": 9640
    },
    {
      "entropy": 4.826428365707398,
      "epoch": 2.7268871925360476,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043331156842310326,
      "loss": 4.6181,
      "mean_token_accuracy": 0.25908993035554884,
      "num_tokens": 7690618.0,
      "step": 9645
    },
    {
      "entropy": 4.954466009140015,
      "epoch": 2.728300819903873,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004332384602038055,
      "loss": 4.7392,
      "mean_token_accuracy": 0.23639548420906067,
      "num_tokens": 7694441.0,
      "step": 9650
    },
    {
      "entropy": 4.80050892829895,
      "epoch": 2.729714447271699,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043316531891364815,
      "loss": 4.6197,
      "mean_token_accuracy": 0.25359245091676713,
      "num_tokens": 7698133.0,
      "step": 9655
    },
    {
      "entropy": 4.837449455261231,
      "epoch": 2.731128074639525,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004330921445679172,
      "loss": 4.5764,
      "mean_token_accuracy": 0.2623891681432724,
      "num_tokens": 7701855.0,
      "step": 9660
    },
    {
      "entropy": 4.872520446777344,
      "epoch": 2.732541702007351,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043301893718190556,
      "loss": 4.6377,
      "mean_token_accuracy": 0.25807993561029435,
      "num_tokens": 7705958.0,
      "step": 9665
    },
    {
      "entropy": 4.865269136428833,
      "epoch": 2.733955329375177,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004329456967709132,
      "loss": 4.6718,
      "mean_token_accuracy": 0.24368327111005783,
      "num_tokens": 7710060.0,
      "step": 9670
    },
    {
      "entropy": 4.733313655853271,
      "epoch": 2.7353689567430024,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00043287242335024706,
      "loss": 4.4635,
      "mean_token_accuracy": 0.2640411272644997,
      "num_tokens": 7713712.0,
      "step": 9675
    },
    {
      "entropy": 4.839530420303345,
      "epoch": 2.7367825841108284,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043279911693522055,
      "loss": 4.6764,
      "mean_token_accuracy": 0.24788110256195067,
      "num_tokens": 7717535.0,
      "step": 9680
    },
    {
      "entropy": 4.8970530986785885,
      "epoch": 2.7381962114786544,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004327257775411545,
      "loss": 4.6889,
      "mean_token_accuracy": 0.24222799837589265,
      "num_tokens": 7721150.0,
      "step": 9685
    },
    {
      "entropy": 4.880186367034912,
      "epoch": 2.73960983884648,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043265240518337617,
      "loss": 4.5875,
      "mean_token_accuracy": 0.2512853741645813,
      "num_tokens": 7724983.0,
      "step": 9690
    },
    {
      "entropy": 4.855270767211914,
      "epoch": 2.741023466214306,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043257899987722017,
      "loss": 4.6906,
      "mean_token_accuracy": 0.24186884313821794,
      "num_tokens": 7729025.0,
      "step": 9695
    },
    {
      "entropy": 4.781070899963379,
      "epoch": 2.7424370935821316,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043250556163802756,
      "loss": 4.5588,
      "mean_token_accuracy": 0.2607974648475647,
      "num_tokens": 7732975.0,
      "step": 9700
    },
    {
      "entropy": 4.826660633087158,
      "epoch": 2.7438507209499576,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004324320904811466,
      "loss": 4.506,
      "mean_token_accuracy": 0.2782688707113266,
      "num_tokens": 7736843.0,
      "step": 9705
    },
    {
      "entropy": 4.79444465637207,
      "epoch": 2.745264348317783,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00043235858642193226,
      "loss": 4.6729,
      "mean_token_accuracy": 0.24772382378578187,
      "num_tokens": 7740476.0,
      "step": 9710
    },
    {
      "entropy": 4.907176351547241,
      "epoch": 2.7466779756856092,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043228504947574644,
      "loss": 4.7319,
      "mean_token_accuracy": 0.23835003972053528,
      "num_tokens": 7744464.0,
      "step": 9715
    },
    {
      "entropy": 4.952947998046875,
      "epoch": 2.7480916030534353,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00043221147965795794,
      "loss": 4.7273,
      "mean_token_accuracy": 0.2381937399506569,
      "num_tokens": 7748325.0,
      "step": 9720
    },
    {
      "entropy": 4.9189318180084225,
      "epoch": 2.749505230421261,
      "grad_norm": 1.4375,
      "learning_rate": 0.00043213787698394225,
      "loss": 4.7579,
      "mean_token_accuracy": 0.23863381147384644,
      "num_tokens": 7752106.0,
      "step": 9725
    },
    {
      "entropy": 4.907739782333374,
      "epoch": 2.750918857789087,
      "grad_norm": 1.46875,
      "learning_rate": 0.000432064241469082,
      "loss": 4.6753,
      "mean_token_accuracy": 0.24603926092386247,
      "num_tokens": 7755927.0,
      "step": 9730
    },
    {
      "entropy": 4.816323566436767,
      "epoch": 2.752332485156913,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004319905731287666,
      "loss": 4.6053,
      "mean_token_accuracy": 0.2565395846962929,
      "num_tokens": 7759780.0,
      "step": 9735
    },
    {
      "entropy": 4.874113273620606,
      "epoch": 2.7537461125247384,
      "grad_norm": 1.375,
      "learning_rate": 0.0004319168719783922,
      "loss": 4.6062,
      "mean_token_accuracy": 0.25769962966442106,
      "num_tokens": 7763854.0,
      "step": 9740
    },
    {
      "entropy": 4.859772777557373,
      "epoch": 2.7551597398925645,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043184313803336187,
      "loss": 4.6988,
      "mean_token_accuracy": 0.26035857051610944,
      "num_tokens": 7767970.0,
      "step": 9745
    },
    {
      "entropy": 4.893213748931885,
      "epoch": 2.75657336726039,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004317693713090856,
      "loss": 4.6129,
      "mean_token_accuracy": 0.2607592597603798,
      "num_tokens": 7771790.0,
      "step": 9750
    },
    {
      "entropy": 4.775887155532837,
      "epoch": 2.757986994628216,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043169557182098017,
      "loss": 4.4647,
      "mean_token_accuracy": 0.278387051820755,
      "num_tokens": 7775591.0,
      "step": 9755
    },
    {
      "entropy": 4.74648642539978,
      "epoch": 2.7594006219960416,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004316217395844692,
      "loss": 4.5515,
      "mean_token_accuracy": 0.25995130836963654,
      "num_tokens": 7779718.0,
      "step": 9760
    },
    {
      "entropy": 4.9524003028869625,
      "epoch": 2.7608142493638677,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004315478746149832,
      "loss": 4.6987,
      "mean_token_accuracy": 0.2502158656716347,
      "num_tokens": 7783734.0,
      "step": 9765
    },
    {
      "entropy": 4.952664422988891,
      "epoch": 2.7622278767316937,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004314739769279594,
      "loss": 4.7478,
      "mean_token_accuracy": 0.24359761029481888,
      "num_tokens": 7787556.0,
      "step": 9770
    },
    {
      "entropy": 4.868218994140625,
      "epoch": 2.7636415040995193,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043140004653884214,
      "loss": 4.5693,
      "mean_token_accuracy": 0.2712606802582741,
      "num_tokens": 7791304.0,
      "step": 9775
    },
    {
      "entropy": 4.812817478179932,
      "epoch": 2.7650551314673453,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004313260834630822,
      "loss": 4.6359,
      "mean_token_accuracy": 0.25051970034837723,
      "num_tokens": 7795262.0,
      "step": 9780
    },
    {
      "entropy": 4.815307712554931,
      "epoch": 2.7664687588351713,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004312520877161375,
      "loss": 4.6669,
      "mean_token_accuracy": 0.2514455422759056,
      "num_tokens": 7799618.0,
      "step": 9785
    },
    {
      "entropy": 4.815370035171509,
      "epoch": 2.767882386202997,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004311780593134727,
      "loss": 4.6215,
      "mean_token_accuracy": 0.2562608256936073,
      "num_tokens": 7803543.0,
      "step": 9790
    },
    {
      "entropy": 4.861965036392212,
      "epoch": 2.769296013570823,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00043110399827055943,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23997777849435806,
      "num_tokens": 7807591.0,
      "step": 9795
    },
    {
      "entropy": 4.81014609336853,
      "epoch": 2.7707096409386485,
      "grad_norm": 1.3125,
      "learning_rate": 0.00043102990460287566,
      "loss": 4.6765,
      "mean_token_accuracy": 0.2528988614678383,
      "num_tokens": 7811621.0,
      "step": 9800
    },
    {
      "entropy": 4.895223474502563,
      "epoch": 2.7721232683064745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004309557783259068,
      "loss": 4.7084,
      "mean_token_accuracy": 0.2513528555631638,
      "num_tokens": 7815862.0,
      "step": 9805
    },
    {
      "entropy": 4.924595022201538,
      "epoch": 2.7735368956743,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004308816194551446,
      "loss": 4.6285,
      "mean_token_accuracy": 0.2485656425356865,
      "num_tokens": 7819727.0,
      "step": 9810
    },
    {
      "entropy": 4.865536880493164,
      "epoch": 2.774950523042126,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004308074280060879,
      "loss": 4.6659,
      "mean_token_accuracy": 0.25097320675849916,
      "num_tokens": 7823488.0,
      "step": 9815
    },
    {
      "entropy": 4.876431179046631,
      "epoch": 2.776364150409952,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004307332039942422,
      "loss": 4.7412,
      "mean_token_accuracy": 0.2485273599624634,
      "num_tokens": 7827767.0,
      "step": 9820
    },
    {
      "entropy": 4.873225402832031,
      "epoch": 2.7777777777777777,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004306589474351199,
      "loss": 4.7536,
      "mean_token_accuracy": 0.24785084873437882,
      "num_tokens": 7831897.0,
      "step": 9825
    },
    {
      "entropy": 4.931944990158081,
      "epoch": 2.7791914051456037,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004305846583442402,
      "loss": 4.6751,
      "mean_token_accuracy": 0.2531357005238533,
      "num_tokens": 7836241.0,
      "step": 9830
    },
    {
      "entropy": 4.8298118114471436,
      "epoch": 2.7806050325134297,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000430510336737129,
      "loss": 4.6146,
      "mean_token_accuracy": 0.2580153986811638,
      "num_tokens": 7840285.0,
      "step": 9835
    },
    {
      "entropy": 4.817322444915772,
      "epoch": 2.7820186598812553,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00043043598262931907,
      "loss": 4.6446,
      "mean_token_accuracy": 0.24673960506916046,
      "num_tokens": 7844383.0,
      "step": 9840
    },
    {
      "entropy": 4.7755186557769775,
      "epoch": 2.783432287249081,
      "grad_norm": 1.4375,
      "learning_rate": 0.00043036159603635,
      "loss": 4.5679,
      "mean_token_accuracy": 0.2668361350893974,
      "num_tokens": 7848213.0,
      "step": 9845
    },
    {
      "entropy": 4.87588849067688,
      "epoch": 2.784845914616907,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00043028717697376795,
      "loss": 4.6952,
      "mean_token_accuracy": 0.24972268790006638,
      "num_tokens": 7852565.0,
      "step": 9850
    },
    {
      "entropy": 4.957758140563965,
      "epoch": 2.786259541984733,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004302127254571263,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2335992380976677,
      "num_tokens": 7856392.0,
      "step": 9855
    },
    {
      "entropy": 4.945465707778931,
      "epoch": 2.7876731693525585,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004301382415019848,
      "loss": 4.7542,
      "mean_token_accuracy": 0.24494190365076066,
      "num_tokens": 7860156.0,
      "step": 9860
    },
    {
      "entropy": 4.883669042587281,
      "epoch": 2.7890867967203845,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004300637251239102,
      "loss": 4.7481,
      "mean_token_accuracy": 0.24594495445489883,
      "num_tokens": 7864189.0,
      "step": 9865
    },
    {
      "entropy": 4.915887069702149,
      "epoch": 2.7905004240882105,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00042998917633847586,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24104094058275222,
      "num_tokens": 7868265.0,
      "step": 9870
    },
    {
      "entropy": 4.943803834915161,
      "epoch": 2.791914051456036,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00042991459516126217,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24904415309429168,
      "num_tokens": 7872150.0,
      "step": 9875
    },
    {
      "entropy": 4.917537021636963,
      "epoch": 2.793327678823862,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000429839981607856,
      "loss": 4.5087,
      "mean_token_accuracy": 0.2617950916290283,
      "num_tokens": 7876161.0,
      "step": 9880
    },
    {
      "entropy": 4.793934297561646,
      "epoch": 2.794741306191688,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042976533569385116,
      "loss": 4.6005,
      "mean_token_accuracy": 0.25740389823913573,
      "num_tokens": 7879971.0,
      "step": 9885
    },
    {
      "entropy": 4.852872753143311,
      "epoch": 2.7961549335595137,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004296906574348483,
      "loss": 4.6737,
      "mean_token_accuracy": 0.24890671223402022,
      "num_tokens": 7884201.0,
      "step": 9890
    },
    {
      "entropy": 4.856833219528198,
      "epoch": 2.7975685609273393,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004296159468464545,
      "loss": 4.596,
      "mean_token_accuracy": 0.251895035803318,
      "num_tokens": 7888344.0,
      "step": 9895
    },
    {
      "entropy": 4.8422754287719725,
      "epoch": 2.7989821882951653,
      "grad_norm": 1.484375,
      "learning_rate": 0.000429541203944284,
      "loss": 4.6423,
      "mean_token_accuracy": 0.24859222620725632,
      "num_tokens": 7892173.0,
      "step": 9900
    },
    {
      "entropy": 4.8172478675842285,
      "epoch": 2.8003958156629913,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004294664287439575,
      "loss": 4.5942,
      "mean_token_accuracy": 0.25541715770959855,
      "num_tokens": 7896234.0,
      "step": 9905
    },
    {
      "entropy": 4.8123067855834964,
      "epoch": 2.801809443030817,
      "grad_norm": 1.53125,
      "learning_rate": 0.00042939162126110256,
      "loss": 4.5952,
      "mean_token_accuracy": 0.261730982363224,
      "num_tokens": 7900189.0,
      "step": 9910
    },
    {
      "entropy": 4.87520842552185,
      "epoch": 2.803223070398643,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004293167815113535,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24916821271181105,
      "num_tokens": 7904147.0,
      "step": 9915
    },
    {
      "entropy": 4.860744953155518,
      "epoch": 2.804636697766469,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00042924190951035144,
      "loss": 4.6513,
      "mean_token_accuracy": 0.25360911935567854,
      "num_tokens": 7908271.0,
      "step": 9920
    },
    {
      "entropy": 4.866771078109741,
      "epoch": 2.8060503251342945,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042916700527374405,
      "loss": 4.692,
      "mean_token_accuracy": 0.24698856770992278,
      "num_tokens": 7912338.0,
      "step": 9925
    },
    {
      "entropy": 4.836369752883911,
      "epoch": 2.8074639525021206,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004290920688171859,
      "loss": 4.612,
      "mean_token_accuracy": 0.2465449944138527,
      "num_tokens": 7916449.0,
      "step": 9930
    },
    {
      "entropy": 4.876132154464722,
      "epoch": 2.808877579869946,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042901710015633826,
      "loss": 4.7331,
      "mean_token_accuracy": 0.2425728425383568,
      "num_tokens": 7920515.0,
      "step": 9935
    },
    {
      "entropy": 4.878490686416626,
      "epoch": 2.810291207237772,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004289420993068692,
      "loss": 4.6339,
      "mean_token_accuracy": 0.24594679474830627,
      "num_tokens": 7924667.0,
      "step": 9940
    },
    {
      "entropy": 4.83689317703247,
      "epoch": 2.8117048346055977,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042886706628445317,
      "loss": 4.5828,
      "mean_token_accuracy": 0.257267427444458,
      "num_tokens": 7928834.0,
      "step": 9945
    },
    {
      "entropy": 4.842553567886353,
      "epoch": 2.8131184619734237,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004287920011047718,
      "loss": 4.6772,
      "mean_token_accuracy": 0.24618014991283416,
      "num_tokens": 7932799.0,
      "step": 9950
    },
    {
      "entropy": 4.762071561813355,
      "epoch": 2.8145320893412498,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004287169037835132,
      "loss": 4.6015,
      "mean_token_accuracy": 0.2539287656545639,
      "num_tokens": 7936853.0,
      "step": 9955
    },
    {
      "entropy": 4.879246759414673,
      "epoch": 2.8159457167090753,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042864177433637237,
      "loss": 4.6948,
      "mean_token_accuracy": 0.25117368102073667,
      "num_tokens": 7940797.0,
      "step": 9960
    },
    {
      "entropy": 4.923811721801758,
      "epoch": 2.8173593440769014,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042856661277905066,
      "loss": 4.6806,
      "mean_token_accuracy": 0.25007946193218233,
      "num_tokens": 7944829.0,
      "step": 9965
    },
    {
      "entropy": 4.812182664871216,
      "epoch": 2.8187729714447274,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042849141912725664,
      "loss": 4.6333,
      "mean_token_accuracy": 0.24886305630207062,
      "num_tokens": 7948885.0,
      "step": 9970
    },
    {
      "entropy": 4.8478111743927,
      "epoch": 2.820186598812553,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004284161933967051,
      "loss": 4.6,
      "mean_token_accuracy": 0.26527938693761827,
      "num_tokens": 7952999.0,
      "step": 9975
    },
    {
      "entropy": 4.787328720092773,
      "epoch": 2.821600226180379,
      "grad_norm": 1.5,
      "learning_rate": 0.00042834093560311786,
      "loss": 4.6059,
      "mean_token_accuracy": 0.25689260810613634,
      "num_tokens": 7957111.0,
      "step": 9980
    },
    {
      "entropy": 4.786101388931274,
      "epoch": 2.8230138535482046,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042826564576222336,
      "loss": 4.5824,
      "mean_token_accuracy": 0.2559566840529442,
      "num_tokens": 7960834.0,
      "step": 9985
    },
    {
      "entropy": 4.879970407485962,
      "epoch": 2.8244274809160306,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004281903238897566,
      "loss": 4.6762,
      "mean_token_accuracy": 0.25245027542114257,
      "num_tokens": 7964863.0,
      "step": 9990
    },
    {
      "entropy": 4.926187467575073,
      "epoch": 2.825841108283856,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004281149700014594,
      "loss": 4.6824,
      "mean_token_accuracy": 0.24631931483745576,
      "num_tokens": 7968738.0,
      "step": 9995
    },
    {
      "entropy": 4.864536714553833,
      "epoch": 2.827254735651682,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00042803958411308045,
      "loss": 4.5899,
      "mean_token_accuracy": 0.25540718585252764,
      "num_tokens": 7972967.0,
      "step": 10000
    },
    {
      "epoch": 2.827254735651682,
      "eval_entropy": 4.674065130471677,
      "eval_loss": 5.041006565093994,
      "eval_mean_token_accuracy": 0.24404867024053334,
      "eval_num_tokens": 7972967.0,
      "eval_runtime": 4.8771,
      "eval_samples_per_second": 2313.658,
      "eval_steps_per_second": 289.31,
      "step": 10000
    },
    {
      "entropy": 4.866463565826416,
      "epoch": 2.828668363019508,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004279641662403746,
      "loss": 4.5993,
      "mean_token_accuracy": 0.25540264397859574,
      "num_tokens": 7976806.0,
      "step": 10005
    },
    {
      "entropy": 4.884882593154908,
      "epoch": 2.8300819903873338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004278887163991042,
      "loss": 4.8138,
      "mean_token_accuracy": 0.2508085876703262,
      "num_tokens": 7980692.0,
      "step": 10010
    },
    {
      "entropy": 4.84144024848938,
      "epoch": 2.83149561775516,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042781323460503725,
      "loss": 4.6211,
      "mean_token_accuracy": 0.2536349192261696,
      "num_tokens": 7984629.0,
      "step": 10015
    },
    {
      "entropy": 4.888455533981324,
      "epoch": 2.832909245122986,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004277377208739494,
      "loss": 4.7152,
      "mean_token_accuracy": 0.2546853706240654,
      "num_tokens": 7988624.0,
      "step": 10020
    },
    {
      "entropy": 4.843786334991455,
      "epoch": 2.8343228724908114,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042766217522162225,
      "loss": 4.6172,
      "mean_token_accuracy": 0.25524652898311617,
      "num_tokens": 7992157.0,
      "step": 10025
    },
    {
      "entropy": 4.873719978332519,
      "epoch": 2.8357364998586374,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042758659766384456,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2578320547938347,
      "num_tokens": 7996388.0,
      "step": 10030
    },
    {
      "entropy": 4.78538498878479,
      "epoch": 2.837150127226463,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004275109882164115,
      "loss": 4.6025,
      "mean_token_accuracy": 0.25137643069028853,
      "num_tokens": 8000341.0,
      "step": 10035
    },
    {
      "entropy": 4.852574491500855,
      "epoch": 2.838563754594289,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004274353468951251,
      "loss": 4.6402,
      "mean_token_accuracy": 0.25642036348581315,
      "num_tokens": 8004656.0,
      "step": 10040
    },
    {
      "entropy": 4.866937303543091,
      "epoch": 2.8399773819621146,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004273596737157937,
      "loss": 4.6268,
      "mean_token_accuracy": 0.256108446419239,
      "num_tokens": 8008548.0,
      "step": 10045
    },
    {
      "entropy": 4.952336692810059,
      "epoch": 2.8413910093299406,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004272839686942327,
      "loss": 4.7301,
      "mean_token_accuracy": 0.24770182520151138,
      "num_tokens": 8012619.0,
      "step": 10050
    },
    {
      "entropy": 4.913769817352295,
      "epoch": 2.8428046366977666,
      "grad_norm": 1.421875,
      "learning_rate": 0.00042720823184626397,
      "loss": 4.7349,
      "mean_token_accuracy": 0.24549670666456222,
      "num_tokens": 8016741.0,
      "step": 10055
    },
    {
      "entropy": 4.819621610641479,
      "epoch": 2.844218264065592,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042713246318771604,
      "loss": 4.5525,
      "mean_token_accuracy": 0.26525541841983796,
      "num_tokens": 8020790.0,
      "step": 10060
    },
    {
      "entropy": 4.877183532714843,
      "epoch": 2.845631891433418,
      "grad_norm": 1.390625,
      "learning_rate": 0.000427056662734424,
      "loss": 4.5473,
      "mean_token_accuracy": 0.2674215078353882,
      "num_tokens": 8024678.0,
      "step": 10065
    },
    {
      "entropy": 4.818094873428345,
      "epoch": 2.8470455188012442,
      "grad_norm": 1.5,
      "learning_rate": 0.00042698083050222974,
      "loss": 4.7098,
      "mean_token_accuracy": 0.2451152890920639,
      "num_tokens": 8028986.0,
      "step": 10070
    },
    {
      "entropy": 4.854453420639038,
      "epoch": 2.84845914616907,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042690496650698174,
      "loss": 4.6595,
      "mean_token_accuracy": 0.24779412746429444,
      "num_tokens": 8033247.0,
      "step": 10075
    },
    {
      "entropy": 4.875601291656494,
      "epoch": 2.849872773536896,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042682907076453506,
      "loss": 4.7202,
      "mean_token_accuracy": 0.25137836635112765,
      "num_tokens": 8036916.0,
      "step": 10080
    },
    {
      "entropy": 4.883870840072632,
      "epoch": 2.8512864009047214,
      "grad_norm": 1.5,
      "learning_rate": 0.00042675314329075154,
      "loss": 4.5742,
      "mean_token_accuracy": 0.2672653943300247,
      "num_tokens": 8040934.0,
      "step": 10085
    },
    {
      "entropy": 4.8745424270629885,
      "epoch": 2.8527000282725474,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00042667718410149954,
      "loss": 4.7246,
      "mean_token_accuracy": 0.24945971518754959,
      "num_tokens": 8044799.0,
      "step": 10090
    },
    {
      "entropy": 4.894460535049438,
      "epoch": 2.854113655640373,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042660119321265386,
      "loss": 4.6669,
      "mean_token_accuracy": 0.2561689779162407,
      "num_tokens": 8048643.0,
      "step": 10095
    },
    {
      "entropy": 4.8751153469085695,
      "epoch": 2.855527283008199,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004265251706400963,
      "loss": 4.5347,
      "mean_token_accuracy": 0.260586841404438,
      "num_tokens": 8052273.0,
      "step": 10100
    },
    {
      "entropy": 4.832471895217895,
      "epoch": 2.856940910376025,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042644911639971514,
      "loss": 4.6308,
      "mean_token_accuracy": 0.2444923222064972,
      "num_tokens": 8056566.0,
      "step": 10105
    },
    {
      "entropy": 4.846393251419068,
      "epoch": 2.8583545377438506,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042637303050740516,
      "loss": 4.7614,
      "mean_token_accuracy": 0.24512107372283937,
      "num_tokens": 8060397.0,
      "step": 10110
    },
    {
      "entropy": 4.925962829589844,
      "epoch": 2.8597681651116766,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004262969129790679,
      "loss": 4.6648,
      "mean_token_accuracy": 0.2479138508439064,
      "num_tokens": 8064258.0,
      "step": 10115
    },
    {
      "entropy": 4.821431732177734,
      "epoch": 2.8611817924795027,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004262207638306114,
      "loss": 4.6796,
      "mean_token_accuracy": 0.25401643961668013,
      "num_tokens": 8068222.0,
      "step": 10120
    },
    {
      "entropy": 4.878140020370483,
      "epoch": 2.8625954198473282,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004261445830779503,
      "loss": 4.6543,
      "mean_token_accuracy": 0.25216632187366483,
      "num_tokens": 8072155.0,
      "step": 10125
    },
    {
      "entropy": 4.85254020690918,
      "epoch": 2.864009047215154,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000426068370737006,
      "loss": 4.5991,
      "mean_token_accuracy": 0.25417042672634127,
      "num_tokens": 8076172.0,
      "step": 10130
    },
    {
      "entropy": 4.839605855941772,
      "epoch": 2.86542267458298,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004259921268237065,
      "loss": 4.6894,
      "mean_token_accuracy": 0.2542981967329979,
      "num_tokens": 8080215.0,
      "step": 10135
    },
    {
      "entropy": 4.751418972015381,
      "epoch": 2.866836301950806,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004259158513539862,
      "loss": 4.55,
      "mean_token_accuracy": 0.26479924768209456,
      "num_tokens": 8084313.0,
      "step": 10140
    },
    {
      "entropy": 4.8596508502960205,
      "epoch": 2.8682499293186314,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042583954434378615,
      "loss": 4.7011,
      "mean_token_accuracy": 0.24612938910722731,
      "num_tokens": 8088356.0,
      "step": 10145
    },
    {
      "entropy": 4.906797885894775,
      "epoch": 2.8696635566864575,
      "grad_norm": 1.375,
      "learning_rate": 0.0004257632058090541,
      "loss": 4.7993,
      "mean_token_accuracy": 0.24169344455003738,
      "num_tokens": 8092569.0,
      "step": 10150
    },
    {
      "entropy": 4.88328070640564,
      "epoch": 2.8710771840542835,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004256868357657443,
      "loss": 4.6143,
      "mean_token_accuracy": 0.255996011197567,
      "num_tokens": 8096849.0,
      "step": 10155
    },
    {
      "entropy": 4.899047422409057,
      "epoch": 2.872490811422109,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042561043422981767,
      "loss": 4.6108,
      "mean_token_accuracy": 0.2659933969378471,
      "num_tokens": 8101000.0,
      "step": 10160
    },
    {
      "entropy": 4.842488861083984,
      "epoch": 2.873904438789935,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004255340012172417,
      "loss": 4.6642,
      "mean_token_accuracy": 0.26194532215595245,
      "num_tokens": 8104844.0,
      "step": 10165
    },
    {
      "entropy": 4.798759889602661,
      "epoch": 2.875318066157761,
      "grad_norm": 1.453125,
      "learning_rate": 0.00042545753674399025,
      "loss": 4.601,
      "mean_token_accuracy": 0.25294087529182435,
      "num_tokens": 8108788.0,
      "step": 10170
    },
    {
      "entropy": 4.913931798934937,
      "epoch": 2.8767316935255867,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000425381040826044,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24005190134048462,
      "num_tokens": 8112700.0,
      "step": 10175
    },
    {
      "entropy": 4.949397373199463,
      "epoch": 2.8781453208934122,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042530451347939024,
      "loss": 4.6329,
      "mean_token_accuracy": 0.248592571914196,
      "num_tokens": 8116647.0,
      "step": 10180
    },
    {
      "entropy": 4.814911365509033,
      "epoch": 2.8795589482612383,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042522795472002263,
      "loss": 4.6077,
      "mean_token_accuracy": 0.25760546028614045,
      "num_tokens": 8120840.0,
      "step": 10185
    },
    {
      "entropy": 4.851918506622314,
      "epoch": 2.8809725756290643,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042515136456394134,
      "loss": 4.6108,
      "mean_token_accuracy": 0.25688067972660067,
      "num_tokens": 8124773.0,
      "step": 10190
    },
    {
      "entropy": 4.92313289642334,
      "epoch": 2.88238620299689,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004250747430271535,
      "loss": 4.6107,
      "mean_token_accuracy": 0.24927992969751359,
      "num_tokens": 8128596.0,
      "step": 10195
    },
    {
      "entropy": 4.8454142093658445,
      "epoch": 2.883799830364716,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042499809012567233,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24783019870519638,
      "num_tokens": 8132596.0,
      "step": 10200
    },
    {
      "entropy": 4.812903261184692,
      "epoch": 2.885213457732542,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042492140587551785,
      "loss": 4.6438,
      "mean_token_accuracy": 0.25576954185962675,
      "num_tokens": 8136384.0,
      "step": 10205
    },
    {
      "entropy": 4.85305814743042,
      "epoch": 2.8866270851003675,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042484469029271666,
      "loss": 4.6908,
      "mean_token_accuracy": 0.25137128084897997,
      "num_tokens": 8140348.0,
      "step": 10210
    },
    {
      "entropy": 4.892729616165161,
      "epoch": 2.8880407124681935,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00042476794339330184,
      "loss": 4.6485,
      "mean_token_accuracy": 0.2534482702612877,
      "num_tokens": 8144248.0,
      "step": 10215
    },
    {
      "entropy": 4.883625841140747,
      "epoch": 2.889454339836019,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042469116519331296,
      "loss": 4.6012,
      "mean_token_accuracy": 0.2531357079744339,
      "num_tokens": 8148035.0,
      "step": 10220
    },
    {
      "entropy": 4.931650924682617,
      "epoch": 2.890867967203845,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00042461435570879626,
      "loss": 4.7808,
      "mean_token_accuracy": 0.2388952061533928,
      "num_tokens": 8151830.0,
      "step": 10225
    },
    {
      "entropy": 4.8512163162231445,
      "epoch": 2.8922815945716707,
      "grad_norm": 1.296875,
      "learning_rate": 0.00042453751495580435,
      "loss": 4.599,
      "mean_token_accuracy": 0.2483886420726776,
      "num_tokens": 8155974.0,
      "step": 10230
    },
    {
      "entropy": 4.84147801399231,
      "epoch": 2.8936952219394967,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004244606429503965,
      "loss": 4.6338,
      "mean_token_accuracy": 0.2516477808356285,
      "num_tokens": 8159954.0,
      "step": 10235
    },
    {
      "entropy": 4.8708446502685545,
      "epoch": 2.8951088493073227,
      "grad_norm": 1.4375,
      "learning_rate": 0.00042438373970863865,
      "loss": 4.6415,
      "mean_token_accuracy": 0.24459604322910308,
      "num_tokens": 8163736.0,
      "step": 10240
    },
    {
      "entropy": 4.892603015899658,
      "epoch": 2.8965224766751483,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004243068052466028,
      "loss": 4.6459,
      "mean_token_accuracy": 0.2422148808836937,
      "num_tokens": 8167579.0,
      "step": 10245
    },
    {
      "entropy": 4.913422155380249,
      "epoch": 2.8979361040429743,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042422983958036804,
      "loss": 4.7028,
      "mean_token_accuracy": 0.23924105167388915,
      "num_tokens": 8171521.0,
      "step": 10250
    },
    {
      "entropy": 4.814581108093262,
      "epoch": 2.8993497314108003,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042415284272601957,
      "loss": 4.6096,
      "mean_token_accuracy": 0.25148614346981046,
      "num_tokens": 8175688.0,
      "step": 10255
    },
    {
      "entropy": 4.792218351364136,
      "epoch": 2.900763358778626,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004240758146996493,
      "loss": 4.585,
      "mean_token_accuracy": 0.25316630601882933,
      "num_tokens": 8179571.0,
      "step": 10260
    },
    {
      "entropy": 4.966167163848877,
      "epoch": 2.902176986146452,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004239987555173557,
      "loss": 4.7353,
      "mean_token_accuracy": 0.2416842982172966,
      "num_tokens": 8183665.0,
      "step": 10265
    },
    {
      "entropy": 4.852408838272095,
      "epoch": 2.9035906135142775,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004239216651952435,
      "loss": 4.6784,
      "mean_token_accuracy": 0.2534638628363609,
      "num_tokens": 8187807.0,
      "step": 10270
    },
    {
      "entropy": 4.814750003814697,
      "epoch": 2.9050042408821035,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042384454374942417,
      "loss": 4.562,
      "mean_token_accuracy": 0.2525628998875618,
      "num_tokens": 8191908.0,
      "step": 10275
    },
    {
      "entropy": 4.835200071334839,
      "epoch": 2.906417868249929,
      "grad_norm": 1.34375,
      "learning_rate": 0.00042376739119601565,
      "loss": 4.5789,
      "mean_token_accuracy": 0.2645233809947968,
      "num_tokens": 8195909.0,
      "step": 10280
    },
    {
      "entropy": 4.940717792510986,
      "epoch": 2.907831495617755,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004236902075511423,
      "loss": 4.687,
      "mean_token_accuracy": 0.2571380078792572,
      "num_tokens": 8199443.0,
      "step": 10285
    },
    {
      "entropy": 4.776969385147095,
      "epoch": 2.909245122985581,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000423612992830935,
      "loss": 4.5836,
      "mean_token_accuracy": 0.2636663198471069,
      "num_tokens": 8203448.0,
      "step": 10290
    },
    {
      "entropy": 4.81067795753479,
      "epoch": 2.9106587503534067,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004235357470515313,
      "loss": 4.6678,
      "mean_token_accuracy": 0.24975768625736236,
      "num_tokens": 8207163.0,
      "step": 10295
    },
    {
      "entropy": 4.8283368110656735,
      "epoch": 2.9120723777212327,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00042345847022907495,
      "loss": 4.6254,
      "mean_token_accuracy": 0.255475115776062,
      "num_tokens": 8210957.0,
      "step": 10300
    },
    {
      "entropy": 4.818454694747925,
      "epoch": 2.9134860050890588,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00042338116237971635,
      "loss": 4.5217,
      "mean_token_accuracy": 0.26829566806554794,
      "num_tokens": 8214555.0,
      "step": 10305
    },
    {
      "entropy": 4.791113328933716,
      "epoch": 2.9148996324568843,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042330382351961235,
      "loss": 4.5623,
      "mean_token_accuracy": 0.2659288436174393,
      "num_tokens": 8218441.0,
      "step": 10310
    },
    {
      "entropy": 4.727895593643188,
      "epoch": 2.9163132598247103,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042322645366492633,
      "loss": 4.4492,
      "mean_token_accuracy": 0.26574035733938217,
      "num_tokens": 8222383.0,
      "step": 10315
    },
    {
      "entropy": 4.889033365249634,
      "epoch": 2.917726887192536,
      "grad_norm": 1.421875,
      "learning_rate": 0.00042314905283182816,
      "loss": 4.7058,
      "mean_token_accuracy": 0.24405403733253478,
      "num_tokens": 8226282.0,
      "step": 10320
    },
    {
      "entropy": 4.87326135635376,
      "epoch": 2.919140514560362,
      "grad_norm": 1.328125,
      "learning_rate": 0.000423071621036494,
      "loss": 4.6136,
      "mean_token_accuracy": 0.2568039208650589,
      "num_tokens": 8230241.0,
      "step": 10325
    },
    {
      "entropy": 4.909646892547608,
      "epoch": 2.9205541419281875,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004229941582951067,
      "loss": 4.6725,
      "mean_token_accuracy": 0.24540476948022844,
      "num_tokens": 8233997.0,
      "step": 10330
    },
    {
      "entropy": 4.924815559387207,
      "epoch": 2.9219677692960135,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004229166646238556,
      "loss": 4.6701,
      "mean_token_accuracy": 0.24900762140750884,
      "num_tokens": 8237873.0,
      "step": 10335
    },
    {
      "entropy": 4.927318572998047,
      "epoch": 2.9233813966638396,
      "grad_norm": 1.40625,
      "learning_rate": 0.000422839140038936,
      "loss": 4.658,
      "mean_token_accuracy": 0.2531849190592766,
      "num_tokens": 8241687.0,
      "step": 10340
    },
    {
      "entropy": 4.850474882125854,
      "epoch": 2.924795024031665,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042276158455655064,
      "loss": 4.6923,
      "mean_token_accuracy": 0.24237418323755264,
      "num_tokens": 8245642.0,
      "step": 10345
    },
    {
      "entropy": 4.84357476234436,
      "epoch": 2.926208651399491,
      "grad_norm": 1.375,
      "learning_rate": 0.0004226839981929077,
      "loss": 4.5458,
      "mean_token_accuracy": 0.2628508388996124,
      "num_tokens": 8249648.0,
      "step": 10350
    },
    {
      "entropy": 4.826624584197998,
      "epoch": 2.927622278767317,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042260638096422237,
      "loss": 4.6618,
      "mean_token_accuracy": 0.260707451403141,
      "num_tokens": 8253722.0,
      "step": 10355
    },
    {
      "entropy": 4.9337362289428714,
      "epoch": 2.9290359061351428,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042252873288671624,
      "loss": 4.7285,
      "mean_token_accuracy": 0.24291004240512848,
      "num_tokens": 8258161.0,
      "step": 10360
    },
    {
      "entropy": 4.814038324356079,
      "epoch": 2.9304495335029688,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042245105397661724,
      "loss": 4.5707,
      "mean_token_accuracy": 0.2583500608801842,
      "num_tokens": 8262015.0,
      "step": 10365
    },
    {
      "entropy": 4.974380683898926,
      "epoch": 2.9318631608707943,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004223733442501597,
      "loss": 4.7831,
      "mean_token_accuracy": 0.23312955349683762,
      "num_tokens": 8266002.0,
      "step": 10370
    },
    {
      "entropy": 4.871824312210083,
      "epoch": 2.9332767882386204,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042229560372358463,
      "loss": 4.6509,
      "mean_token_accuracy": 0.23998259752988815,
      "num_tokens": 8270125.0,
      "step": 10375
    },
    {
      "entropy": 4.90169129371643,
      "epoch": 2.934690415606446,
      "grad_norm": 1.546875,
      "learning_rate": 0.00042221783241313914,
      "loss": 4.6633,
      "mean_token_accuracy": 0.24711370021104812,
      "num_tokens": 8273686.0,
      "step": 10380
    },
    {
      "entropy": 4.86688756942749,
      "epoch": 2.936104042974272,
      "grad_norm": 1.4375,
      "learning_rate": 0.00042214003033507716,
      "loss": 4.6196,
      "mean_token_accuracy": 0.25165033489465716,
      "num_tokens": 8277499.0,
      "step": 10385
    },
    {
      "entropy": 4.822997999191284,
      "epoch": 2.937517670342098,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004220621975056586,
      "loss": 4.6802,
      "mean_token_accuracy": 0.25758338123559954,
      "num_tokens": 8281947.0,
      "step": 10390
    },
    {
      "entropy": 4.840761327743531,
      "epoch": 2.9389312977099236,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004219843339411502,
      "loss": 4.6911,
      "mean_token_accuracy": 0.2457735702395439,
      "num_tokens": 8285869.0,
      "step": 10395
    },
    {
      "entropy": 4.860964155197143,
      "epoch": 2.9403449250777496,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00042190643965782495,
      "loss": 4.6154,
      "mean_token_accuracy": 0.23947821706533431,
      "num_tokens": 8289754.0,
      "step": 10400
    },
    {
      "entropy": 4.92356505393982,
      "epoch": 2.9417585524455756,
      "grad_norm": 1.46875,
      "learning_rate": 0.00042182851467196223,
      "loss": 4.7056,
      "mean_token_accuracy": 0.2515940427780151,
      "num_tokens": 8293424.0,
      "step": 10405
    },
    {
      "entropy": 4.963449907302857,
      "epoch": 2.943172179813401,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042175055899984796,
      "loss": 4.6676,
      "mean_token_accuracy": 0.25106121301651,
      "num_tokens": 8297015.0,
      "step": 10410
    },
    {
      "entropy": 4.7748877048492435,
      "epoch": 2.9445858071812268,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004216725726577744,
      "loss": 4.5928,
      "mean_token_accuracy": 0.26458068639039994,
      "num_tokens": 8300946.0,
      "step": 10415
    },
    {
      "entropy": 4.832032918930054,
      "epoch": 2.9459994345490528,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004215945556620399,
      "loss": 4.6505,
      "mean_token_accuracy": 0.25343901515007017,
      "num_tokens": 8305373.0,
      "step": 10420
    },
    {
      "entropy": 4.910195827484131,
      "epoch": 2.947413061916879,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042151650802895005,
      "loss": 4.686,
      "mean_token_accuracy": 0.24375094771385192,
      "num_tokens": 8309336.0,
      "step": 10425
    },
    {
      "entropy": 4.861594533920288,
      "epoch": 2.9488266892847044,
      "grad_norm": 1.390625,
      "learning_rate": 0.00042143842977481583,
      "loss": 4.6816,
      "mean_token_accuracy": 0.24237390607595444,
      "num_tokens": 8313302.0,
      "step": 10430
    },
    {
      "entropy": 4.731359910964966,
      "epoch": 2.9502403166525304,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004213603209159555,
      "loss": 4.5696,
      "mean_token_accuracy": 0.25912997126579285,
      "num_tokens": 8317193.0,
      "step": 10435
    },
    {
      "entropy": 4.90677227973938,
      "epoch": 2.9516539440203564,
      "grad_norm": 1.375,
      "learning_rate": 0.0004212821814686931,
      "loss": 4.6503,
      "mean_token_accuracy": 0.2518499314785004,
      "num_tokens": 8321139.0,
      "step": 10440
    },
    {
      "entropy": 4.952362155914306,
      "epoch": 2.953067571388182,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004212040114493593,
      "loss": 4.6676,
      "mean_token_accuracy": 0.2516310691833496,
      "num_tokens": 8324862.0,
      "step": 10445
    },
    {
      "entropy": 4.8898955345153805,
      "epoch": 2.954481198756008,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042112581087429133,
      "loss": 4.6448,
      "mean_token_accuracy": 0.2400026649236679,
      "num_tokens": 8329219.0,
      "step": 10450
    },
    {
      "entropy": 4.845518732070923,
      "epoch": 2.955894826123834,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00042104757975983245,
      "loss": 4.6731,
      "mean_token_accuracy": 0.24874717444181443,
      "num_tokens": 8332969.0,
      "step": 10455
    },
    {
      "entropy": 4.833274793624878,
      "epoch": 2.9573084534916596,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004209693181223325,
      "loss": 4.6672,
      "mean_token_accuracy": 0.2556930333375931,
      "num_tokens": 8336976.0,
      "step": 10460
    },
    {
      "entropy": 4.871947717666626,
      "epoch": 2.958722080859485,
      "grad_norm": 1.40625,
      "learning_rate": 0.00042089102597814775,
      "loss": 4.5856,
      "mean_token_accuracy": 0.26356328278779984,
      "num_tokens": 8340982.0,
      "step": 10465
    },
    {
      "entropy": 4.883913660049439,
      "epoch": 2.960135708227311,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004208127033436407,
      "loss": 4.6287,
      "mean_token_accuracy": 0.2533979922533035,
      "num_tokens": 8344860.0,
      "step": 10470
    },
    {
      "entropy": 4.864223194122315,
      "epoch": 2.9615493355951372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042073435023518035,
      "loss": 4.6449,
      "mean_token_accuracy": 0.25758776515722276,
      "num_tokens": 8348926.0,
      "step": 10475
    },
    {
      "entropy": 4.7729896068572994,
      "epoch": 2.962962962962963,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00042065596666914195,
      "loss": 4.5575,
      "mean_token_accuracy": 0.2635492205619812,
      "num_tokens": 8353083.0,
      "step": 10480
    },
    {
      "entropy": 4.894790744781494,
      "epoch": 2.964376590330789,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00042057755266190734,
      "loss": 4.7321,
      "mean_token_accuracy": 0.24883221536874772,
      "num_tokens": 8357627.0,
      "step": 10485
    },
    {
      "entropy": 4.916041517257691,
      "epoch": 2.965790217698615,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004204991082298643,
      "loss": 4.6707,
      "mean_token_accuracy": 0.24380408525466918,
      "num_tokens": 8361492.0,
      "step": 10490
    },
    {
      "entropy": 4.838923597335816,
      "epoch": 2.9672038450664404,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042042063338940745,
      "loss": 4.6013,
      "mean_token_accuracy": 0.2617553934454918,
      "num_tokens": 8365587.0,
      "step": 10495
    },
    {
      "entropy": 4.791348457336426,
      "epoch": 2.9686174724342664,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004203421281569375,
      "loss": 4.6974,
      "mean_token_accuracy": 0.25294720083475114,
      "num_tokens": 8369460.0,
      "step": 10500
    },
    {
      "entropy": 4.832683277130127,
      "epoch": 2.970031099802092,
      "grad_norm": 1.5,
      "learning_rate": 0.00042026359254886147,
      "loss": 4.6242,
      "mean_token_accuracy": 0.2627121612429619,
      "num_tokens": 8373394.0,
      "step": 10505
    },
    {
      "entropy": 4.848023509979248,
      "epoch": 2.971444727169918,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00042018502658159294,
      "loss": 4.5865,
      "mean_token_accuracy": 0.26513745337724687,
      "num_tokens": 8377278.0,
      "step": 10510
    },
    {
      "entropy": 4.991074085235596,
      "epoch": 2.9728583545377436,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042010643027155163,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23700769990682602,
      "num_tokens": 8381335.0,
      "step": 10515
    },
    {
      "entropy": 4.860500001907349,
      "epoch": 2.9742719819055696,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004200278036351638,
      "loss": 4.6739,
      "mean_token_accuracy": 0.2536489814519882,
      "num_tokens": 8385234.0,
      "step": 10520
    },
    {
      "entropy": 4.804272747039795,
      "epoch": 2.9756856092733956,
      "grad_norm": 1.3125,
      "learning_rate": 0.00041994914668886166,
      "loss": 4.6982,
      "mean_token_accuracy": 0.2428934618830681,
      "num_tokens": 8389656.0,
      "step": 10525
    },
    {
      "entropy": 4.863270950317383,
      "epoch": 2.9770992366412212,
      "grad_norm": 1.5,
      "learning_rate": 0.00041987045944908433,
      "loss": 4.6792,
      "mean_token_accuracy": 0.24745032787322999,
      "num_tokens": 8393484.0,
      "step": 10530
    },
    {
      "entropy": 4.912521839141846,
      "epoch": 2.9785128640090472,
      "grad_norm": 1.40625,
      "learning_rate": 0.00041979174193227677,
      "loss": 4.6418,
      "mean_token_accuracy": 0.2515360310673714,
      "num_tokens": 8397672.0,
      "step": 10535
    },
    {
      "entropy": 4.908854198455811,
      "epoch": 2.9799264913768733,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004197129941548905,
      "loss": 4.697,
      "mean_token_accuracy": 0.2467256546020508,
      "num_tokens": 8401607.0,
      "step": 10540
    },
    {
      "entropy": 4.942834091186524,
      "epoch": 2.981340118744699,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004196342161333833,
      "loss": 4.763,
      "mean_token_accuracy": 0.23534928560256957,
      "num_tokens": 8406060.0,
      "step": 10545
    },
    {
      "entropy": 4.948119211196899,
      "epoch": 2.982753746112525,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041955540788421946,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24064231812953948,
      "num_tokens": 8410098.0,
      "step": 10550
    },
    {
      "entropy": 4.921736764907837,
      "epoch": 2.9841673734803504,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004194765694238691,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2543004333972931,
      "num_tokens": 8414306.0,
      "step": 10555
    },
    {
      "entropy": 4.84614405632019,
      "epoch": 2.9855810008481765,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041939770076880936,
      "loss": 4.6639,
      "mean_token_accuracy": 0.2541520789265633,
      "num_tokens": 8418204.0,
      "step": 10560
    },
    {
      "entropy": 4.845655202865601,
      "epoch": 2.986994628216002,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004193188019355229,
      "loss": 4.6829,
      "mean_token_accuracy": 0.2444151148200035,
      "num_tokens": 8421946.0,
      "step": 10565
    },
    {
      "entropy": 4.947361850738526,
      "epoch": 2.988408255583828,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004192398729404993,
      "loss": 4.7861,
      "mean_token_accuracy": 0.23368808329105378,
      "num_tokens": 8426087.0,
      "step": 10570
    },
    {
      "entropy": 4.960837745666504,
      "epoch": 2.989821882951654,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041916091380023423,
      "loss": 4.7098,
      "mean_token_accuracy": 0.25635330826044084,
      "num_tokens": 8430145.0,
      "step": 10575
    },
    {
      "entropy": 4.860358047485351,
      "epoch": 2.9912355103194797,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004190819245312297,
      "loss": 4.5837,
      "mean_token_accuracy": 0.2592990756034851,
      "num_tokens": 8434134.0,
      "step": 10580
    },
    {
      "entropy": 4.863066911697388,
      "epoch": 2.9926491376873057,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041900290514999383,
      "loss": 4.6527,
      "mean_token_accuracy": 0.2433200001716614,
      "num_tokens": 8438464.0,
      "step": 10585
    },
    {
      "entropy": 4.772168350219727,
      "epoch": 2.9940627650551317,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041892385567304135,
      "loss": 4.658,
      "mean_token_accuracy": 0.25139556974172594,
      "num_tokens": 8442391.0,
      "step": 10590
    },
    {
      "entropy": 4.868023681640625,
      "epoch": 2.9954763924229573,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041884477611689295,
      "loss": 4.6709,
      "mean_token_accuracy": 0.24659744054079055,
      "num_tokens": 8446454.0,
      "step": 10595
    },
    {
      "entropy": 4.802685832977295,
      "epoch": 2.9968900197907833,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004187656664980759,
      "loss": 4.6113,
      "mean_token_accuracy": 0.25905914753675463,
      "num_tokens": 8450444.0,
      "step": 10600
    },
    {
      "entropy": 4.964135074615479,
      "epoch": 2.998303647158609,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004186865268331236,
      "loss": 4.7648,
      "mean_token_accuracy": 0.24262286126613616,
      "num_tokens": 8454278.0,
      "step": 10605
    },
    {
      "entropy": 4.9586005210876465,
      "epoch": 2.999717274526435,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004186073571385757,
      "loss": 4.8039,
      "mean_token_accuracy": 0.24477051794528962,
      "num_tokens": 8458573.0,
      "step": 10610
    },
    {
      "entropy": 4.871467638015747,
      "epoch": 3.0011309018942605,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041852815743097824,
      "loss": 4.5405,
      "mean_token_accuracy": 0.2641226649284363,
      "num_tokens": 8462460.0,
      "step": 10615
    },
    {
      "entropy": 4.86848521232605,
      "epoch": 3.0025445292620865,
      "grad_norm": 1.375,
      "learning_rate": 0.0004184489277268833,
      "loss": 4.473,
      "mean_token_accuracy": 0.26172467321157455,
      "num_tokens": 8466363.0,
      "step": 10620
    },
    {
      "entropy": 4.89790358543396,
      "epoch": 3.0039581566299125,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041836966804284964,
      "loss": 4.4613,
      "mean_token_accuracy": 0.2631221368908882,
      "num_tokens": 8470425.0,
      "step": 10625
    },
    {
      "entropy": 4.924962186813355,
      "epoch": 3.005371783997738,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041829037839544186,
      "loss": 4.5023,
      "mean_token_accuracy": 0.24678574949502946,
      "num_tokens": 8474472.0,
      "step": 10630
    },
    {
      "entropy": 4.904690551757812,
      "epoch": 3.006785411365564,
      "grad_norm": 1.375,
      "learning_rate": 0.0004182110588012312,
      "loss": 4.4133,
      "mean_token_accuracy": 0.25526578575372694,
      "num_tokens": 8478078.0,
      "step": 10635
    },
    {
      "entropy": 4.816775941848755,
      "epoch": 3.0081990387333897,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041813170927679466,
      "loss": 4.3155,
      "mean_token_accuracy": 0.2779365375638008,
      "num_tokens": 8481979.0,
      "step": 10640
    },
    {
      "entropy": 4.86826639175415,
      "epoch": 3.0096126661012157,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041805232983871605,
      "loss": 4.4781,
      "mean_token_accuracy": 0.2639901876449585,
      "num_tokens": 8486125.0,
      "step": 10645
    },
    {
      "entropy": 4.872261810302734,
      "epoch": 3.0110262934690417,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004179729205035852,
      "loss": 4.4499,
      "mean_token_accuracy": 0.26033587753772736,
      "num_tokens": 8489837.0,
      "step": 10650
    },
    {
      "entropy": 4.9024085521698,
      "epoch": 3.0124399208368673,
      "grad_norm": 1.515625,
      "learning_rate": 0.000417893481287998,
      "loss": 4.4367,
      "mean_token_accuracy": 0.26533375680446625,
      "num_tokens": 8493931.0,
      "step": 10655
    },
    {
      "entropy": 4.9146946430206295,
      "epoch": 3.0138535482046933,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041781401220855683,
      "loss": 4.523,
      "mean_token_accuracy": 0.2566113412380219,
      "num_tokens": 8497862.0,
      "step": 10660
    },
    {
      "entropy": 4.813630390167236,
      "epoch": 3.015267175572519,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004177345132818703,
      "loss": 4.4172,
      "mean_token_accuracy": 0.2699525788426399,
      "num_tokens": 8501768.0,
      "step": 10665
    },
    {
      "entropy": 4.789814281463623,
      "epoch": 3.016680802940345,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004176549845245531,
      "loss": 4.4219,
      "mean_token_accuracy": 0.2667327269911766,
      "num_tokens": 8505738.0,
      "step": 10670
    },
    {
      "entropy": 4.8919168472290036,
      "epoch": 3.018094430308171,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004175754259532263,
      "loss": 4.471,
      "mean_token_accuracy": 0.2624878793954849,
      "num_tokens": 8509677.0,
      "step": 10675
    },
    {
      "entropy": 4.719310569763183,
      "epoch": 3.0195080576759965,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004174958375845171,
      "loss": 4.3477,
      "mean_token_accuracy": 0.28210090547800065,
      "num_tokens": 8513819.0,
      "step": 10680
    },
    {
      "entropy": 4.823121833801269,
      "epoch": 3.0209216850438225,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004174162194350591,
      "loss": 4.4614,
      "mean_token_accuracy": 0.2565693691372871,
      "num_tokens": 8517987.0,
      "step": 10685
    },
    {
      "entropy": 4.867356014251709,
      "epoch": 3.022335312411648,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041733657152149184,
      "loss": 4.4894,
      "mean_token_accuracy": 0.2610200747847557,
      "num_tokens": 8521650.0,
      "step": 10690
    },
    {
      "entropy": 4.768692541122436,
      "epoch": 3.023748939779474,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00041725689386046136,
      "loss": 4.3734,
      "mean_token_accuracy": 0.2736115664243698,
      "num_tokens": 8525393.0,
      "step": 10695
    },
    {
      "entropy": 4.842730236053467,
      "epoch": 3.0251625671473,
      "grad_norm": 1.453125,
      "learning_rate": 0.00041717718646861975,
      "loss": 4.479,
      "mean_token_accuracy": 0.2729305371642113,
      "num_tokens": 8529251.0,
      "step": 10700
    },
    {
      "entropy": 4.838714456558227,
      "epoch": 3.0265761945151257,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004170974493626253,
      "loss": 4.4258,
      "mean_token_accuracy": 0.2641012579202652,
      "num_tokens": 8533319.0,
      "step": 10705
    },
    {
      "entropy": 4.849391126632691,
      "epoch": 3.0279898218829517,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004170176825591426,
      "loss": 4.3417,
      "mean_token_accuracy": 0.2734531000256538,
      "num_tokens": 8537210.0,
      "step": 10710
    },
    {
      "entropy": 4.793328809738159,
      "epoch": 3.0294034492507773,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004169378860748426,
      "loss": 4.4681,
      "mean_token_accuracy": 0.2658111363649368,
      "num_tokens": 8541201.0,
      "step": 10715
    },
    {
      "entropy": 4.866433906555176,
      "epoch": 3.0308170766186033,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004168580599264021,
      "loss": 4.4861,
      "mean_token_accuracy": 0.25492009371519087,
      "num_tokens": 8545191.0,
      "step": 10720
    },
    {
      "entropy": 4.830425119400024,
      "epoch": 3.0322307039864294,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004167782041305042,
      "loss": 4.4542,
      "mean_token_accuracy": 0.2647597104310989,
      "num_tokens": 8549269.0,
      "step": 10725
    },
    {
      "entropy": 4.807202100753784,
      "epoch": 3.033644331354255,
      "grad_norm": 1.34375,
      "learning_rate": 0.00041669831870383845,
      "loss": 4.4236,
      "mean_token_accuracy": 0.26285615712404253,
      "num_tokens": 8553252.0,
      "step": 10730
    },
    {
      "entropy": 4.768797540664673,
      "epoch": 3.035057958722081,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004166184036631003,
      "loss": 4.3985,
      "mean_token_accuracy": 0.2617382690310478,
      "num_tokens": 8557120.0,
      "step": 10735
    },
    {
      "entropy": 4.772066402435303,
      "epoch": 3.0364715860899065,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004165384590249915,
      "loss": 4.3486,
      "mean_token_accuracy": 0.27808377891778946,
      "num_tokens": 8561166.0,
      "step": 10740
    },
    {
      "entropy": 4.82372817993164,
      "epoch": 3.0378852134577325,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041645848480622005,
      "loss": 4.4035,
      "mean_token_accuracy": 0.27203281968832016,
      "num_tokens": 8565052.0,
      "step": 10745
    },
    {
      "entropy": 4.83982481956482,
      "epoch": 3.0392988408255586,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004163784810235001,
      "loss": 4.5135,
      "mean_token_accuracy": 0.26158757507801056,
      "num_tokens": 8568960.0,
      "step": 10750
    },
    {
      "entropy": 4.816205549240112,
      "epoch": 3.040712468193384,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004162984476935517,
      "loss": 4.4219,
      "mean_token_accuracy": 0.26379294991493224,
      "num_tokens": 8572961.0,
      "step": 10755
    },
    {
      "entropy": 4.827414846420288,
      "epoch": 3.04212609556121,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041621838483310163,
      "loss": 4.4137,
      "mean_token_accuracy": 0.27308743000030516,
      "num_tokens": 8576942.0,
      "step": 10760
    },
    {
      "entropy": 4.809186315536499,
      "epoch": 3.0435397229290357,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004161382924588824,
      "loss": 4.458,
      "mean_token_accuracy": 0.2576703876256943,
      "num_tokens": 8580944.0,
      "step": 10765
    },
    {
      "entropy": 4.814213514328003,
      "epoch": 3.0449533502968618,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004160581705876328,
      "loss": 4.4595,
      "mean_token_accuracy": 0.2590032324194908,
      "num_tokens": 8584685.0,
      "step": 10770
    },
    {
      "entropy": 4.789428472518921,
      "epoch": 3.046366977664688,
      "grad_norm": 1.375,
      "learning_rate": 0.00041597801923609775,
      "loss": 4.4632,
      "mean_token_accuracy": 0.26336066275835035,
      "num_tokens": 8588733.0,
      "step": 10775
    },
    {
      "entropy": 4.827841281890869,
      "epoch": 3.0477806050325134,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004158978384210285,
      "loss": 4.3892,
      "mean_token_accuracy": 0.26502277702093124,
      "num_tokens": 8592662.0,
      "step": 10780
    },
    {
      "entropy": 4.830812644958496,
      "epoch": 3.0491942324003394,
      "grad_norm": 1.46875,
      "learning_rate": 0.00041581762815918235,
      "loss": 4.4503,
      "mean_token_accuracy": 0.2612713620066643,
      "num_tokens": 8596458.0,
      "step": 10785
    },
    {
      "entropy": 4.861101579666138,
      "epoch": 3.050607859768165,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00041573738846732266,
      "loss": 4.4645,
      "mean_token_accuracy": 0.26364328116178515,
      "num_tokens": 8600413.0,
      "step": 10790
    },
    {
      "entropy": 4.853553247451782,
      "epoch": 3.052021487135991,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004156571193622191,
      "loss": 4.4688,
      "mean_token_accuracy": 0.2731964275240898,
      "num_tokens": 8604384.0,
      "step": 10795
    },
    {
      "entropy": 4.84637303352356,
      "epoch": 3.053435114503817,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004155768208606474,
      "loss": 4.5462,
      "mean_token_accuracy": 0.23996562212705613,
      "num_tokens": 8608295.0,
      "step": 10800
    },
    {
      "entropy": 4.714881134033203,
      "epoch": 3.0548487418716426,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00041549649297938936,
      "loss": 4.37,
      "mean_token_accuracy": 0.2670752614736557,
      "num_tokens": 8612290.0,
      "step": 10805
    },
    {
      "entropy": 4.830947589874268,
      "epoch": 3.0562623692394686,
      "grad_norm": 1.5,
      "learning_rate": 0.00041541613573523316,
      "loss": 4.4353,
      "mean_token_accuracy": 0.270080603659153,
      "num_tokens": 8615939.0,
      "step": 10810
    },
    {
      "entropy": 4.903247928619384,
      "epoch": 3.057675996607294,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004153357491449729,
      "loss": 4.5674,
      "mean_token_accuracy": 0.25211200267076495,
      "num_tokens": 8619770.0,
      "step": 10815
    },
    {
      "entropy": 4.7475316524505615,
      "epoch": 3.05908962397512,
      "grad_norm": 1.40625,
      "learning_rate": 0.00041525533322540886,
      "loss": 4.3414,
      "mean_token_accuracy": 0.26595164835453033,
      "num_tokens": 8623820.0,
      "step": 10820
    },
    {
      "entropy": 4.845161485671997,
      "epoch": 3.060503251342946,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004151748879933475,
      "loss": 4.4692,
      "mean_token_accuracy": 0.25296606719493864,
      "num_tokens": 8627888.0,
      "step": 10825
    },
    {
      "entropy": 4.7444374561309814,
      "epoch": 3.061916878710772,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004150944134656014,
      "loss": 4.3805,
      "mean_token_accuracy": 0.26735483705997465,
      "num_tokens": 8631821.0,
      "step": 10830
    },
    {
      "entropy": 4.839508152008056,
      "epoch": 3.063330506078598,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00041501390965898913,
      "loss": 4.4948,
      "mean_token_accuracy": 0.25142394602298734,
      "num_tokens": 8635583.0,
      "step": 10835
    },
    {
      "entropy": 4.864922189712525,
      "epoch": 3.0647441334464234,
      "grad_norm": 1.5,
      "learning_rate": 0.0004149333765903356,
      "loss": 4.425,
      "mean_token_accuracy": 0.25930521041154864,
      "num_tokens": 8639455.0,
      "step": 10840
    },
    {
      "entropy": 4.849027538299561,
      "epoch": 3.0661577608142494,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004148528142764717,
      "loss": 4.4762,
      "mean_token_accuracy": 0.26472352892160417,
      "num_tokens": 8643356.0,
      "step": 10845
    },
    {
      "entropy": 4.818200254440308,
      "epoch": 3.067571388182075,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041477222273423437,
      "loss": 4.5585,
      "mean_token_accuracy": 0.2575565278530121,
      "num_tokens": 8647567.0,
      "step": 10850
    },
    {
      "entropy": 4.854152584075928,
      "epoch": 3.068985015549901,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004146916019804669,
      "loss": 4.5385,
      "mean_token_accuracy": 0.2564412966370583,
      "num_tokens": 8651426.0,
      "step": 10855
    },
    {
      "entropy": 4.783076667785645,
      "epoch": 3.070398642917727,
      "grad_norm": 1.375,
      "learning_rate": 0.0004146109520320184,
      "loss": 4.4158,
      "mean_token_accuracy": 0.26901860982179643,
      "num_tokens": 8655263.0,
      "step": 10860
    },
    {
      "entropy": 4.790256834030151,
      "epoch": 3.0718122702855526,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004145302729057443,
      "loss": 4.3586,
      "mean_token_accuracy": 0.2677400544285774,
      "num_tokens": 8659230.0,
      "step": 10865
    },
    {
      "entropy": 4.830503511428833,
      "epoch": 3.0732258976533786,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000414449564618506,
      "loss": 4.4254,
      "mean_token_accuracy": 0.26979370415210724,
      "num_tokens": 8663283.0,
      "step": 10870
    },
    {
      "entropy": 4.8951231956481935,
      "epoch": 3.0746395250212046,
      "grad_norm": 1.375,
      "learning_rate": 0.00041436882718717095,
      "loss": 4.5435,
      "mean_token_accuracy": 0.2578342393040657,
      "num_tokens": 8667656.0,
      "step": 10875
    },
    {
      "entropy": 4.885486173629761,
      "epoch": 3.07605315238903,
      "grad_norm": 1.375,
      "learning_rate": 0.00041428806062861277,
      "loss": 4.45,
      "mean_token_accuracy": 0.26596787571907043,
      "num_tokens": 8671491.0,
      "step": 10880
    },
    {
      "entropy": 4.9039905071258545,
      "epoch": 3.0774667797568562,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00041420726495971144,
      "loss": 4.4694,
      "mean_token_accuracy": 0.2594125896692276,
      "num_tokens": 8675528.0,
      "step": 10885
    },
    {
      "entropy": 4.798290491104126,
      "epoch": 3.078880407124682,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004141264401973525,
      "loss": 4.4266,
      "mean_token_accuracy": 0.26673284024000166,
      "num_tokens": 8679635.0,
      "step": 10890
    },
    {
      "entropy": 4.799066972732544,
      "epoch": 3.080294034492508,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004140455863584279,
      "loss": 4.4315,
      "mean_token_accuracy": 0.26057217121124265,
      "num_tokens": 8683951.0,
      "step": 10895
    },
    {
      "entropy": 4.802571249008179,
      "epoch": 3.0817076618603334,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004139647034598355,
      "loss": 4.3638,
      "mean_token_accuracy": 0.27314565777778627,
      "num_tokens": 8687817.0,
      "step": 10900
    },
    {
      "entropy": 4.808282518386841,
      "epoch": 3.0831212892281594,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004138837915184795,
      "loss": 4.4068,
      "mean_token_accuracy": 0.2664977476000786,
      "num_tokens": 8691748.0,
      "step": 10905
    },
    {
      "entropy": 4.847906017303467,
      "epoch": 3.0845349165959854,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041380285055126983,
      "loss": 4.4907,
      "mean_token_accuracy": 0.25744392573833463,
      "num_tokens": 8695755.0,
      "step": 10910
    },
    {
      "entropy": 4.959339570999146,
      "epoch": 3.085948543963811,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004137218805751227,
      "loss": 4.5281,
      "mean_token_accuracy": 0.26538713574409484,
      "num_tokens": 8699758.0,
      "step": 10915
    },
    {
      "entropy": 4.900239038467407,
      "epoch": 3.087362171331637,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004136408816069605,
      "loss": 4.5347,
      "mean_token_accuracy": 0.24634915888309478,
      "num_tokens": 8703469.0,
      "step": 10920
    },
    {
      "entropy": 4.702831506729126,
      "epoch": 3.0887757986994626,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004135598536637113,
      "loss": 4.331,
      "mean_token_accuracy": 0.2691888988018036,
      "num_tokens": 8707283.0,
      "step": 10925
    },
    {
      "entropy": 4.73499641418457,
      "epoch": 3.0901894260672886,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004134787967623095,
      "loss": 4.4614,
      "mean_token_accuracy": 0.27275838553905485,
      "num_tokens": 8711244.0,
      "step": 10930
    },
    {
      "entropy": 4.8522261619567875,
      "epoch": 3.0916030534351147,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041339771091969553,
      "loss": 4.5294,
      "mean_token_accuracy": 0.2592154234647751,
      "num_tokens": 8715230.0,
      "step": 10935
    },
    {
      "entropy": 4.861284923553467,
      "epoch": 3.0930166808029402,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004133165961528158,
      "loss": 4.4776,
      "mean_token_accuracy": 0.27152823358774186,
      "num_tokens": 8719071.0,
      "step": 10940
    },
    {
      "entropy": 4.813327980041504,
      "epoch": 3.0944303081707663,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041323545247862284,
      "loss": 4.4489,
      "mean_token_accuracy": 0.26622218191623687,
      "num_tokens": 8723068.0,
      "step": 10945
    },
    {
      "entropy": 4.832866477966308,
      "epoch": 3.095843935538592,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000413154279914075,
      "loss": 4.4525,
      "mean_token_accuracy": 0.26080614179372785,
      "num_tokens": 8726994.0,
      "step": 10950
    },
    {
      "entropy": 4.840227699279785,
      "epoch": 3.097257562906418,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041307307847613707,
      "loss": 4.489,
      "mean_token_accuracy": 0.25498823523521424,
      "num_tokens": 8730924.0,
      "step": 10955
    },
    {
      "entropy": 4.723715782165527,
      "epoch": 3.098671190274244,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041299184818177956,
      "loss": 4.3277,
      "mean_token_accuracy": 0.27298938781023024,
      "num_tokens": 8734936.0,
      "step": 10960
    },
    {
      "entropy": 4.771328258514404,
      "epoch": 3.1000848176420694,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004129105890479791,
      "loss": 4.3639,
      "mean_token_accuracy": 0.27125951945781707,
      "num_tokens": 8739046.0,
      "step": 10965
    },
    {
      "entropy": 4.828376245498657,
      "epoch": 3.1014984450098955,
      "grad_norm": 1.375,
      "learning_rate": 0.00041282930109171826,
      "loss": 4.5283,
      "mean_token_accuracy": 0.2564481168985367,
      "num_tokens": 8743183.0,
      "step": 10970
    },
    {
      "entropy": 4.845306015014648,
      "epoch": 3.102912072377721,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004127479843299859,
      "loss": 4.4614,
      "mean_token_accuracy": 0.254409496486187,
      "num_tokens": 8746988.0,
      "step": 10975
    },
    {
      "entropy": 4.8301167488098145,
      "epoch": 3.104325699745547,
      "grad_norm": 1.5,
      "learning_rate": 0.00041266663877977653,
      "loss": 4.3552,
      "mean_token_accuracy": 0.2831428796052933,
      "num_tokens": 8750567.0,
      "step": 10980
    },
    {
      "entropy": 4.760574626922607,
      "epoch": 3.105739327113373,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004125852644580911,
      "loss": 4.3464,
      "mean_token_accuracy": 0.2755372792482376,
      "num_tokens": 8754617.0,
      "step": 10985
    },
    {
      "entropy": 4.694981575012207,
      "epoch": 3.1071529544811987,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041250386138193615,
      "loss": 4.3114,
      "mean_token_accuracy": 0.28698567897081373,
      "num_tokens": 8758463.0,
      "step": 10990
    },
    {
      "entropy": 4.900340557098389,
      "epoch": 3.1085665818490247,
      "grad_norm": 1.421875,
      "learning_rate": 0.00041242242956832445,
      "loss": 4.5213,
      "mean_token_accuracy": 0.2583063006401062,
      "num_tokens": 8762707.0,
      "step": 10995
    },
    {
      "entropy": 4.824777364730835,
      "epoch": 3.1099802092168503,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004123409690342747,
      "loss": 4.4382,
      "mean_token_accuracy": 0.26917397528886794,
      "num_tokens": 8766618.0,
      "step": 11000
    },
    {
      "epoch": 3.1099802092168503,
      "eval_entropy": 4.600402104271633,
      "eval_loss": 5.028213977813721,
      "eval_mean_token_accuracy": 0.24619505676413833,
      "eval_num_tokens": 8766618.0,
      "eval_runtime": 4.9035,
      "eval_samples_per_second": 2301.2,
      "eval_steps_per_second": 287.752,
      "step": 11000
    },
    {
      "entropy": 4.788196277618408,
      "epoch": 3.1113938365846763,
      "grad_norm": 1.4375,
      "learning_rate": 0.00041225947979681176,
      "loss": 4.4079,
      "mean_token_accuracy": 0.25539918690919877,
      "num_tokens": 8770312.0,
      "step": 11005
    },
    {
      "entropy": 4.7783262729644775,
      "epoch": 3.1128074639525023,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004121779618729664,
      "loss": 4.3709,
      "mean_token_accuracy": 0.2623876318335533,
      "num_tokens": 8774518.0,
      "step": 11010
    },
    {
      "entropy": 4.781240463256836,
      "epoch": 3.114221091320328,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00041209641527977523,
      "loss": 4.4271,
      "mean_token_accuracy": 0.27311070561408995,
      "num_tokens": 8778465.0,
      "step": 11015
    },
    {
      "entropy": 4.825845003128052,
      "epoch": 3.115634718688154,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00041201484003428095,
      "loss": 4.4539,
      "mean_token_accuracy": 0.25929405689239504,
      "num_tokens": 8782508.0,
      "step": 11020
    },
    {
      "entropy": 4.825310468673706,
      "epoch": 3.1170483460559795,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004119332361535325,
      "loss": 4.4746,
      "mean_token_accuracy": 0.25726732760667803,
      "num_tokens": 8786416.0,
      "step": 11025
    },
    {
      "entropy": 4.807851028442383,
      "epoch": 3.1184619734238055,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004118516036545844,
      "loss": 4.5034,
      "mean_token_accuracy": 0.2617938965559006,
      "num_tokens": 8790189.0,
      "step": 11030
    },
    {
      "entropy": 4.827736234664917,
      "epoch": 3.1198756007916315,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041176994255449743,
      "loss": 4.4257,
      "mean_token_accuracy": 0.26076364815235137,
      "num_tokens": 8794234.0,
      "step": 11035
    },
    {
      "entropy": 4.863289642333984,
      "epoch": 3.121289228159457,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004116882528703383,
      "loss": 4.4489,
      "mean_token_accuracy": 0.2683545142412186,
      "num_tokens": 8797949.0,
      "step": 11040
    },
    {
      "entropy": 4.8323548316955565,
      "epoch": 3.122702855527283,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00041160653461917945,
      "loss": 4.4063,
      "mean_token_accuracy": 0.2720375880599022,
      "num_tokens": 8801856.0,
      "step": 11045
    },
    {
      "entropy": 4.7011596202850345,
      "epoch": 3.1241164828951087,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041152478781809963,
      "loss": 4.3361,
      "mean_token_accuracy": 0.28659252226352694,
      "num_tokens": 8805677.0,
      "step": 11050
    },
    {
      "entropy": 4.775273609161377,
      "epoch": 3.1255301102629347,
      "grad_norm": 1.40625,
      "learning_rate": 0.00041144301248418355,
      "loss": 4.4581,
      "mean_token_accuracy": 0.26345016807317734,
      "num_tokens": 8809777.0,
      "step": 11055
    },
    {
      "entropy": 4.758031272888184,
      "epoch": 3.1269437376307607,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004113612086345214,
      "loss": 4.3738,
      "mean_token_accuracy": 0.2683605864644051,
      "num_tokens": 8813635.0,
      "step": 11060
    },
    {
      "entropy": 4.851541471481323,
      "epoch": 3.1283573649985863,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00041127937628621,
      "loss": 4.4915,
      "mean_token_accuracy": 0.2564735636115074,
      "num_tokens": 8817529.0,
      "step": 11065
    },
    {
      "entropy": 4.878368091583252,
      "epoch": 3.1297709923664123,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004111975154563517,
      "loss": 4.4755,
      "mean_token_accuracy": 0.2653796821832657,
      "num_tokens": 8821366.0,
      "step": 11070
    },
    {
      "entropy": 4.742597150802612,
      "epoch": 3.131184619734238,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041111562616205486,
      "loss": 4.3771,
      "mean_token_accuracy": 0.274046191573143,
      "num_tokens": 8825246.0,
      "step": 11075
    },
    {
      "entropy": 4.777177524566651,
      "epoch": 3.132598247102064,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004110337084204338,
      "loss": 4.4294,
      "mean_token_accuracy": 0.26414402574300766,
      "num_tokens": 8829182.0,
      "step": 11080
    },
    {
      "entropy": 4.7075354099273685,
      "epoch": 3.13401187446989,
      "grad_norm": 1.375,
      "learning_rate": 0.0004109517622486089,
      "loss": 4.4251,
      "mean_token_accuracy": 0.27150127440690996,
      "num_tokens": 8833322.0,
      "step": 11085
    },
    {
      "entropy": 4.850328493118286,
      "epoch": 3.1354255018377155,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00041086978766370646,
      "loss": 4.4484,
      "mean_token_accuracy": 0.26836225390434265,
      "num_tokens": 8837277.0,
      "step": 11090
    },
    {
      "entropy": 4.83239107131958,
      "epoch": 3.1368391292055415,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004107877846828586,
      "loss": 4.4142,
      "mean_token_accuracy": 0.2712548181414604,
      "num_tokens": 8840917.0,
      "step": 11095
    },
    {
      "entropy": 4.780044269561768,
      "epoch": 3.138252756573367,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004107057533232034,
      "loss": 4.3493,
      "mean_token_accuracy": 0.2751594722270966,
      "num_tokens": 8844967.0,
      "step": 11100
    },
    {
      "entropy": 4.946954917907715,
      "epoch": 3.139666383941193,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00041062369360188485,
      "loss": 4.5619,
      "mean_token_accuracy": 0.2425859570503235,
      "num_tokens": 8848843.0,
      "step": 11105
    },
    {
      "entropy": 4.844552373886108,
      "epoch": 3.141080011309019,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004105416055360531,
      "loss": 4.5304,
      "mean_token_accuracy": 0.2455057755112648,
      "num_tokens": 8852893.0,
      "step": 11110
    },
    {
      "entropy": 4.801080656051636,
      "epoch": 3.1424936386768447,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00041045948914286397,
      "loss": 4.5386,
      "mean_token_accuracy": 0.2584892362356186,
      "num_tokens": 8856842.0,
      "step": 11115
    },
    {
      "entropy": 4.822932481765747,
      "epoch": 3.1439072660446707,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004103773444394792,
      "loss": 4.484,
      "mean_token_accuracy": 0.2597409293055534,
      "num_tokens": 8860585.0,
      "step": 11120
    },
    {
      "entropy": 4.7909753799438475,
      "epoch": 3.1453208934124963,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004102951714430666,
      "loss": 4.3796,
      "mean_token_accuracy": 0.272781902551651,
      "num_tokens": 8864515.0,
      "step": 11125
    },
    {
      "entropy": 4.857646226882935,
      "epoch": 3.1467345207803223,
      "grad_norm": 1.375,
      "learning_rate": 0.00041021297017079984,
      "loss": 4.5074,
      "mean_token_accuracy": 0.2645867273211479,
      "num_tokens": 8868376.0,
      "step": 11130
    },
    {
      "entropy": 4.834972238540649,
      "epoch": 3.148148148148148,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041013074063985845,
      "loss": 4.4667,
      "mean_token_accuracy": 0.2729821428656578,
      "num_tokens": 8872051.0,
      "step": 11135
    },
    {
      "entropy": 4.833029842376709,
      "epoch": 3.149561775515974,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00041004848286742783,
      "loss": 4.4997,
      "mean_token_accuracy": 0.260781791806221,
      "num_tokens": 8876260.0,
      "step": 11140
    },
    {
      "entropy": 4.767558574676514,
      "epoch": 3.1509754028838,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004099661968706995,
      "loss": 4.3655,
      "mean_token_accuracy": 0.26316431611776353,
      "num_tokens": 8880587.0,
      "step": 11145
    },
    {
      "entropy": 4.772040939331054,
      "epoch": 3.1523890302516255,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004098838826668705,
      "loss": 4.4693,
      "mean_token_accuracy": 0.263696426153183,
      "num_tokens": 8884640.0,
      "step": 11150
    },
    {
      "entropy": 4.8616025924682615,
      "epoch": 3.1538026576194516,
      "grad_norm": 1.578125,
      "learning_rate": 0.00040980154027314426,
      "loss": 4.5466,
      "mean_token_accuracy": 0.24504658281803132,
      "num_tokens": 8888381.0,
      "step": 11155
    },
    {
      "entropy": 4.846489000320434,
      "epoch": 3.1552162849872776,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00040971916970672964,
      "loss": 4.3591,
      "mean_token_accuracy": 0.2766589030623436,
      "num_tokens": 8892104.0,
      "step": 11160
    },
    {
      "entropy": 4.801296424865723,
      "epoch": 3.156629912355103,
      "grad_norm": 1.5,
      "learning_rate": 0.0004096367709848415,
      "loss": 4.3972,
      "mean_token_accuracy": 0.2722443714737892,
      "num_tokens": 8896017.0,
      "step": 11165
    },
    {
      "entropy": 4.828802585601807,
      "epoch": 3.158043539722929,
      "grad_norm": 1.4375,
      "learning_rate": 0.000409554344124701,
      "loss": 4.4784,
      "mean_token_accuracy": 0.26443993896245954,
      "num_tokens": 8899778.0,
      "step": 11170
    },
    {
      "entropy": 4.752134704589844,
      "epoch": 3.1594571670907547,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004094718891435346,
      "loss": 4.4876,
      "mean_token_accuracy": 0.25959065556526184,
      "num_tokens": 8903794.0,
      "step": 11175
    },
    {
      "entropy": 4.842313814163208,
      "epoch": 3.1608707944585808,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040938940605857475,
      "loss": 4.5254,
      "mean_token_accuracy": 0.2595626458525658,
      "num_tokens": 8907503.0,
      "step": 11180
    },
    {
      "entropy": 4.813429260253907,
      "epoch": 3.1622844218264063,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00040930689488706025,
      "loss": 4.4277,
      "mean_token_accuracy": 0.2599091663956642,
      "num_tokens": 8911388.0,
      "step": 11185
    },
    {
      "entropy": 4.81595549583435,
      "epoch": 3.1636980491942324,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040922435564623525,
      "loss": 4.3833,
      "mean_token_accuracy": 0.2725223541259766,
      "num_tokens": 8915093.0,
      "step": 11190
    },
    {
      "entropy": 4.815892744064331,
      "epoch": 3.1651116765620584,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004091417883533498,
      "loss": 4.4716,
      "mean_token_accuracy": 0.26378020346164704,
      "num_tokens": 8919568.0,
      "step": 11195
    },
    {
      "entropy": 4.895820760726929,
      "epoch": 3.166525303929884,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040905919302566015,
      "loss": 4.5725,
      "mean_token_accuracy": 0.24602432548999786,
      "num_tokens": 8924059.0,
      "step": 11200
    },
    {
      "entropy": 4.715346145629883,
      "epoch": 3.16793893129771,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004089765696804281,
      "loss": 4.3809,
      "mean_token_accuracy": 0.26678473204374314,
      "num_tokens": 8927937.0,
      "step": 11205
    },
    {
      "entropy": 4.793514919281006,
      "epoch": 3.1693525586655356,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040889391833492163,
      "loss": 4.3852,
      "mean_token_accuracy": 0.26785253137350085,
      "num_tokens": 8932226.0,
      "step": 11210
    },
    {
      "entropy": 4.776254940032959,
      "epoch": 3.1707661860333616,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004088112390064141,
      "loss": 4.4718,
      "mean_token_accuracy": 0.26632533222436905,
      "num_tokens": 8936086.0,
      "step": 11215
    },
    {
      "entropy": 4.883353042602539,
      "epoch": 3.1721798134011876,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004087285317121851,
      "loss": 4.5831,
      "mean_token_accuracy": 0.25726936012506485,
      "num_tokens": 8940338.0,
      "step": 11220
    },
    {
      "entropy": 4.710450124740601,
      "epoch": 3.173593440769013,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040864579646951996,
      "loss": 4.3509,
      "mean_token_accuracy": 0.27504410296678544,
      "num_tokens": 8944590.0,
      "step": 11225
    },
    {
      "entropy": 4.852598571777344,
      "epoch": 3.175007068136839,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00040856303329570975,
      "loss": 4.4437,
      "mean_token_accuracy": 0.2671940878033638,
      "num_tokens": 8948321.0,
      "step": 11230
    },
    {
      "entropy": 4.784152841567993,
      "epoch": 3.1764206955046648,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004084802422080515,
      "loss": 4.4251,
      "mean_token_accuracy": 0.28641738891601565,
      "num_tokens": 8952344.0,
      "step": 11235
    },
    {
      "entropy": 4.846502494812012,
      "epoch": 3.177834322872491,
      "grad_norm": 1.53125,
      "learning_rate": 0.00040839742322384807,
      "loss": 4.5246,
      "mean_token_accuracy": 0.26386879980564115,
      "num_tokens": 8956353.0,
      "step": 11240
    },
    {
      "entropy": 4.820135068893433,
      "epoch": 3.179247950240317,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004083145763604081,
      "loss": 4.5225,
      "mean_token_accuracy": 0.2618135303258896,
      "num_tokens": 8960643.0,
      "step": 11245
    },
    {
      "entropy": 4.83622260093689,
      "epoch": 3.1806615776081424,
      "grad_norm": 1.421875,
      "learning_rate": 0.000408231701635046,
      "loss": 4.4433,
      "mean_token_accuracy": 0.27006596624851226,
      "num_tokens": 8964661.0,
      "step": 11250
    },
    {
      "entropy": 4.872301483154297,
      "epoch": 3.1820752049759684,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004081487990650822,
      "loss": 4.4882,
      "mean_token_accuracy": 0.27355396151542666,
      "num_tokens": 8968631.0,
      "step": 11255
    },
    {
      "entropy": 4.827031517028809,
      "epoch": 3.183488832343794,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004080658686678428,
      "loss": 4.3979,
      "mean_token_accuracy": 0.2642911374568939,
      "num_tokens": 8972955.0,
      "step": 11260
    },
    {
      "entropy": 4.880700302124024,
      "epoch": 3.18490245971162,
      "grad_norm": 1.375,
      "learning_rate": 0.0004079829104606596,
      "loss": 4.5115,
      "mean_token_accuracy": 0.2559134602546692,
      "num_tokens": 8977393.0,
      "step": 11265
    },
    {
      "entropy": 4.771676063537598,
      "epoch": 3.186316087079446,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004078999244608705,
      "loss": 4.3253,
      "mean_token_accuracy": 0.27176550179719927,
      "num_tokens": 8981206.0,
      "step": 11270
    },
    {
      "entropy": 4.813017797470093,
      "epoch": 3.1877297144472716,
      "grad_norm": 1.515625,
      "learning_rate": 0.00040781691068581905,
      "loss": 4.473,
      "mean_token_accuracy": 0.25924657583236693,
      "num_tokens": 8985567.0,
      "step": 11275
    },
    {
      "entropy": 4.838366222381592,
      "epoch": 3.1891433418150976,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004077338691528545,
      "loss": 4.5249,
      "mean_token_accuracy": 0.2525318428874016,
      "num_tokens": 8989599.0,
      "step": 11280
    },
    {
      "entropy": 4.871576261520386,
      "epoch": 3.190556969182923,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004076507998793322,
      "loss": 4.3939,
      "mean_token_accuracy": 0.26319691389799116,
      "num_tokens": 8993662.0,
      "step": 11285
    },
    {
      "entropy": 4.793976068496704,
      "epoch": 3.191970596550749,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040756770288261296,
      "loss": 4.4795,
      "mean_token_accuracy": 0.268438883125782,
      "num_tokens": 8997598.0,
      "step": 11290
    },
    {
      "entropy": 4.770085716247559,
      "epoch": 3.1933842239185752,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00040748457818006367,
      "loss": 4.4488,
      "mean_token_accuracy": 0.2604923829436302,
      "num_tokens": 9001709.0,
      "step": 11295
    },
    {
      "entropy": 4.854135036468506,
      "epoch": 3.194797851286401,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004074014257890567,
      "loss": 4.4881,
      "mean_token_accuracy": 0.256848518550396,
      "num_tokens": 9005861.0,
      "step": 11300
    },
    {
      "entropy": 4.709980010986328,
      "epoch": 3.196211478654227,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004073182457269705,
      "loss": 4.3665,
      "mean_token_accuracy": 0.2764938071370125,
      "num_tokens": 9009897.0,
      "step": 11305
    },
    {
      "entropy": 4.825043439865112,
      "epoch": 3.1976251060220524,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040723503801118923,
      "loss": 4.4285,
      "mean_token_accuracy": 0.262653774023056,
      "num_tokens": 9014021.0,
      "step": 11310
    },
    {
      "entropy": 4.843183946609497,
      "epoch": 3.1990387333898784,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040715180265910267,
      "loss": 4.4555,
      "mean_token_accuracy": 0.2719668626785278,
      "num_tokens": 9018062.0,
      "step": 11315
    },
    {
      "entropy": 4.81154146194458,
      "epoch": 3.2004523607577045,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004070685396881066,
      "loss": 4.5742,
      "mean_token_accuracy": 0.24840006232261658,
      "num_tokens": 9022276.0,
      "step": 11320
    },
    {
      "entropy": 4.810736083984375,
      "epoch": 3.20186598812553,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004069852491156024,
      "loss": 4.4203,
      "mean_token_accuracy": 0.27539160549640657,
      "num_tokens": 9026479.0,
      "step": 11325
    },
    {
      "entropy": 4.964304113388062,
      "epoch": 3.203279615493356,
      "grad_norm": 1.375,
      "learning_rate": 0.0004069019309589973,
      "loss": 4.6217,
      "mean_token_accuracy": 0.2427675098180771,
      "num_tokens": 9030828.0,
      "step": 11330
    },
    {
      "entropy": 4.792222118377685,
      "epoch": 3.2046932428611816,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040681858523570433,
      "loss": 4.3375,
      "mean_token_accuracy": 0.27864596247673035,
      "num_tokens": 9034761.0,
      "step": 11335
    },
    {
      "entropy": 4.768745374679566,
      "epoch": 3.2061068702290076,
      "grad_norm": 1.359375,
      "learning_rate": 0.00040673521196314216,
      "loss": 4.3742,
      "mean_token_accuracy": 0.2712128132581711,
      "num_tokens": 9038778.0,
      "step": 11340
    },
    {
      "entropy": 4.821781253814697,
      "epoch": 3.2075204975968337,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040665181115873524,
      "loss": 4.4426,
      "mean_token_accuracy": 0.26132562160491946,
      "num_tokens": 9042922.0,
      "step": 11345
    },
    {
      "entropy": 4.883051538467408,
      "epoch": 3.2089341249646592,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00040656838283991396,
      "loss": 4.5558,
      "mean_token_accuracy": 0.24718158096075057,
      "num_tokens": 9047323.0,
      "step": 11350
    },
    {
      "entropy": 4.927135705947876,
      "epoch": 3.2103477523324853,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040648492702411424,
      "loss": 4.5428,
      "mean_token_accuracy": 0.26283884346485137,
      "num_tokens": 9051144.0,
      "step": 11355
    },
    {
      "entropy": 4.833128929138184,
      "epoch": 3.211761379700311,
      "grad_norm": 1.390625,
      "learning_rate": 0.000406401443728778,
      "loss": 4.5123,
      "mean_token_accuracy": 0.26705087870359423,
      "num_tokens": 9054919.0,
      "step": 11360
    },
    {
      "entropy": 4.872909355163574,
      "epoch": 3.213175007068137,
      "grad_norm": 1.453125,
      "learning_rate": 0.00040631793297135247,
      "loss": 4.4335,
      "mean_token_accuracy": 0.2689783826470375,
      "num_tokens": 9058804.0,
      "step": 11365
    },
    {
      "entropy": 4.716111278533935,
      "epoch": 3.214588634435963,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040623439476929093,
      "loss": 4.3336,
      "mean_token_accuracy": 0.281607511639595,
      "num_tokens": 9062893.0,
      "step": 11370
    },
    {
      "entropy": 4.826288175582886,
      "epoch": 3.2160022618037885,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040615082914005255,
      "loss": 4.5538,
      "mean_token_accuracy": 0.2543916776776314,
      "num_tokens": 9066861.0,
      "step": 11375
    },
    {
      "entropy": 4.740278434753418,
      "epoch": 3.2174158891716145,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00040606723610110175,
      "loss": 4.4178,
      "mean_token_accuracy": 0.2648239731788635,
      "num_tokens": 9070783.0,
      "step": 11380
    },
    {
      "entropy": 4.824033308029175,
      "epoch": 3.21882951653944,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040598361566990916,
      "loss": 4.4441,
      "mean_token_accuracy": 0.26514638066291807,
      "num_tokens": 9074778.0,
      "step": 11385
    },
    {
      "entropy": 4.767303037643432,
      "epoch": 3.220243143907266,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004058999678639509,
      "loss": 4.4237,
      "mean_token_accuracy": 0.26508931517601014,
      "num_tokens": 9078438.0,
      "step": 11390
    },
    {
      "entropy": 4.831383228302002,
      "epoch": 3.221656771275092,
      "grad_norm": 1.453125,
      "learning_rate": 0.00040581629270070876,
      "loss": 4.4207,
      "mean_token_accuracy": 0.2637964949011803,
      "num_tokens": 9082403.0,
      "step": 11395
    },
    {
      "entropy": 4.806244897842407,
      "epoch": 3.2230703986429177,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004057325901976704,
      "loss": 4.4486,
      "mean_token_accuracy": 0.2660215809941292,
      "num_tokens": 9086342.0,
      "step": 11400
    },
    {
      "entropy": 4.8769659996032715,
      "epoch": 3.2244840260107437,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004056488603723292,
      "loss": 4.5183,
      "mean_token_accuracy": 0.25663975477218626,
      "num_tokens": 9090508.0,
      "step": 11405
    },
    {
      "entropy": 4.75980772972107,
      "epoch": 3.2258976533785693,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000405565103242184,
      "loss": 4.37,
      "mean_token_accuracy": 0.2604318395256996,
      "num_tokens": 9094688.0,
      "step": 11410
    },
    {
      "entropy": 4.698468732833862,
      "epoch": 3.2273112807463953,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00040548131882473963,
      "loss": 4.4342,
      "mean_token_accuracy": 0.26802407056093214,
      "num_tokens": 9098885.0,
      "step": 11415
    },
    {
      "entropy": 4.788854312896729,
      "epoch": 3.228724908114221,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040539750713750657,
      "loss": 4.4279,
      "mean_token_accuracy": 0.26886927932500837,
      "num_tokens": 9103286.0,
      "step": 11420
    },
    {
      "entropy": 4.792214250564575,
      "epoch": 3.230138535482047,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040531366819800076,
      "loss": 4.4272,
      "mean_token_accuracy": 0.27288908809423446,
      "num_tokens": 9107406.0,
      "step": 11425
    },
    {
      "entropy": 4.745152282714844,
      "epoch": 3.231552162849873,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040522980202374414,
      "loss": 4.3724,
      "mean_token_accuracy": 0.2562165930867195,
      "num_tokens": 9111630.0,
      "step": 11430
    },
    {
      "entropy": 4.765726900100708,
      "epoch": 3.2329657902176985,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004051459086322643,
      "loss": 4.3705,
      "mean_token_accuracy": 0.2713934049010277,
      "num_tokens": 9115784.0,
      "step": 11435
    },
    {
      "entropy": 4.808067703247071,
      "epoch": 3.2343794175855245,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00040506198804109427,
      "loss": 4.4529,
      "mean_token_accuracy": 0.2565127432346344,
      "num_tokens": 9119715.0,
      "step": 11440
    },
    {
      "entropy": 4.815934562683106,
      "epoch": 3.2357930449533505,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004049780402677731,
      "loss": 4.327,
      "mean_token_accuracy": 0.2777225911617279,
      "num_tokens": 9123574.0,
      "step": 11445
    },
    {
      "entropy": 4.780207586288452,
      "epoch": 3.237206672321176,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004048940653298452,
      "loss": 4.3961,
      "mean_token_accuracy": 0.27538775354623796,
      "num_tokens": 9127329.0,
      "step": 11450
    },
    {
      "entropy": 4.76591649055481,
      "epoch": 3.238620299689002,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000404810063244861,
      "loss": 4.4286,
      "mean_token_accuracy": 0.2639049455523491,
      "num_tokens": 9130974.0,
      "step": 11455
    },
    {
      "entropy": 4.73817081451416,
      "epoch": 3.2400339270568277,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004047260340303762,
      "loss": 4.4362,
      "mean_token_accuracy": 0.26540700495243075,
      "num_tokens": 9135042.0,
      "step": 11460
    },
    {
      "entropy": 4.770730686187744,
      "epoch": 3.2414475544246537,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004046419777039525,
      "loss": 4.4597,
      "mean_token_accuracy": 0.2765744522213936,
      "num_tokens": 9138921.0,
      "step": 11465
    },
    {
      "entropy": 4.7798467636108395,
      "epoch": 3.2428611817924793,
      "grad_norm": 1.359375,
      "learning_rate": 0.00040455789428315707,
      "loss": 4.3664,
      "mean_token_accuracy": 0.26251440942287446,
      "num_tokens": 9142994.0,
      "step": 11470
    },
    {
      "entropy": 4.825553369522095,
      "epoch": 3.2442748091603053,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040447378378556307,
      "loss": 4.5908,
      "mean_token_accuracy": 0.24250887781381608,
      "num_tokens": 9146911.0,
      "step": 11475
    },
    {
      "entropy": 4.8096753597259525,
      "epoch": 3.2456884365281313,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040438964622874874,
      "loss": 4.3679,
      "mean_token_accuracy": 0.2739678740501404,
      "num_tokens": 9151026.0,
      "step": 11480
    },
    {
      "entropy": 4.782994174957276,
      "epoch": 3.247102063895957,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004043054816302986,
      "loss": 4.4365,
      "mean_token_accuracy": 0.26689216047525405,
      "num_tokens": 9154999.0,
      "step": 11485
    },
    {
      "entropy": 4.721678161621094,
      "epoch": 3.248515691263783,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00040422129000780235,
      "loss": 4.4564,
      "mean_token_accuracy": 0.2662389069795609,
      "num_tokens": 9159099.0,
      "step": 11490
    },
    {
      "entropy": 4.723671102523804,
      "epoch": 3.249929318631609,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004041370713788555,
      "loss": 4.3728,
      "mean_token_accuracy": 0.26815446764230727,
      "num_tokens": 9162752.0,
      "step": 11495
    },
    {
      "entropy": 4.818335342407226,
      "epoch": 3.2513429459994345,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004040528257610593,
      "loss": 4.487,
      "mean_token_accuracy": 0.2669894427061081,
      "num_tokens": 9166537.0,
      "step": 11500
    },
    {
      "entropy": 4.89172248840332,
      "epoch": 3.2527565733672605,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004039685531720207,
      "loss": 4.5178,
      "mean_token_accuracy": 0.25724323391914367,
      "num_tokens": 9170684.0,
      "step": 11505
    },
    {
      "entropy": 4.870863580703736,
      "epoch": 3.254170200735086,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004038842536293518,
      "loss": 4.4949,
      "mean_token_accuracy": 0.25265308916568757,
      "num_tokens": 9174642.0,
      "step": 11510
    },
    {
      "entropy": 4.902136516571045,
      "epoch": 3.255583828102912,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004037999271506711,
      "loss": 4.534,
      "mean_token_accuracy": 0.2591665357351303,
      "num_tokens": 9178653.0,
      "step": 11515
    },
    {
      "entropy": 4.852220678329468,
      "epoch": 3.2569974554707377,
      "grad_norm": 1.484375,
      "learning_rate": 0.00040371557375360194,
      "loss": 4.4833,
      "mean_token_accuracy": 0.26088032126426697,
      "num_tokens": 9182509.0,
      "step": 11520
    },
    {
      "entropy": 4.765623712539673,
      "epoch": 3.2584110828385637,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004036311934557739,
      "loss": 4.3546,
      "mean_token_accuracy": 0.2725220814347267,
      "num_tokens": 9186605.0,
      "step": 11525
    },
    {
      "entropy": 4.7598659038543705,
      "epoch": 3.2598247102063898,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004035467862748219,
      "loss": 4.4174,
      "mean_token_accuracy": 0.26481594890356064,
      "num_tokens": 9190437.0,
      "step": 11530
    },
    {
      "entropy": 4.748468446731567,
      "epoch": 3.2612383375742153,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004034623522283863,
      "loss": 4.3496,
      "mean_token_accuracy": 0.2706329122185707,
      "num_tokens": 9194757.0,
      "step": 11535
    },
    {
      "entropy": 4.856335020065307,
      "epoch": 3.2626519649420413,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004033778913341136,
      "loss": 4.507,
      "mean_token_accuracy": 0.26659811437129977,
      "num_tokens": 9199165.0,
      "step": 11540
    },
    {
      "entropy": 4.777457761764526,
      "epoch": 3.264065592309867,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00040329340360965556,
      "loss": 4.4186,
      "mean_token_accuracy": 0.27169869244098666,
      "num_tokens": 9203283.0,
      "step": 11545
    },
    {
      "entropy": 4.785129404067993,
      "epoch": 3.265479219677693,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004032088890726694,
      "loss": 4.4433,
      "mean_token_accuracy": 0.2611043304204941,
      "num_tokens": 9207038.0,
      "step": 11550
    },
    {
      "entropy": 4.79152398109436,
      "epoch": 3.266892847045519,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040312434774081817,
      "loss": 4.4429,
      "mean_token_accuracy": 0.26585907191038133,
      "num_tokens": 9211035.0,
      "step": 11555
    },
    {
      "entropy": 4.80406379699707,
      "epoch": 3.2683064744133445,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004030397796317707,
      "loss": 4.46,
      "mean_token_accuracy": 0.2597294956445694,
      "num_tokens": 9215226.0,
      "step": 11560
    },
    {
      "entropy": 4.881890916824341,
      "epoch": 3.2697201017811706,
      "grad_norm": 1.40625,
      "learning_rate": 0.000402955184763201,
      "loss": 4.4657,
      "mean_token_accuracy": 0.25572032630443575,
      "num_tokens": 9219461.0,
      "step": 11565
    },
    {
      "entropy": 4.834243249893189,
      "epoch": 3.271133729148996,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004028705631527889,
      "loss": 4.4734,
      "mean_token_accuracy": 0.2661674484610558,
      "num_tokens": 9223501.0,
      "step": 11570
    },
    {
      "entropy": 4.778223896026612,
      "epoch": 3.272547356516822,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004027859148182199,
      "loss": 4.4722,
      "mean_token_accuracy": 0.27183846980333326,
      "num_tokens": 9227649.0,
      "step": 11575
    },
    {
      "entropy": 4.835183238983154,
      "epoch": 3.273960983884648,
      "grad_norm": 1.484375,
      "learning_rate": 0.00040270123977718486,
      "loss": 4.5288,
      "mean_token_accuracy": 0.25362751483917234,
      "num_tokens": 9231512.0,
      "step": 11580
    },
    {
      "entropy": 4.884126329421997,
      "epoch": 3.2753746112524738,
      "grad_norm": 1.5,
      "learning_rate": 0.0004026165380473804,
      "loss": 4.5351,
      "mean_token_accuracy": 0.26040648967027663,
      "num_tokens": 9235413.0,
      "step": 11585
    },
    {
      "entropy": 4.876446104049682,
      "epoch": 3.2767882386202998,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004025318096465087,
      "loss": 4.4218,
      "mean_token_accuracy": 0.262123042345047,
      "num_tokens": 9239256.0,
      "step": 11590
    },
    {
      "entropy": 4.8121270656585695,
      "epoch": 3.2782018659881254,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004024470545922773,
      "loss": 4.4233,
      "mean_token_accuracy": 0.2697847053408623,
      "num_tokens": 9243193.0,
      "step": 11595
    },
    {
      "entropy": 4.799270153045654,
      "epoch": 3.2796154933559514,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004023622729023997,
      "loss": 4.5189,
      "mean_token_accuracy": 0.25766860842704775,
      "num_tokens": 9247151.0,
      "step": 11600
    },
    {
      "entropy": 4.759886455535889,
      "epoch": 3.2810291207237774,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00040227746459459455,
      "loss": 4.4481,
      "mean_token_accuracy": 0.2682533234357834,
      "num_tokens": 9250959.0,
      "step": 11605
    },
    {
      "entropy": 4.7803261280059814,
      "epoch": 3.282442748091603,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00040219262968658646,
      "loss": 4.433,
      "mean_token_accuracy": 0.26671700179576874,
      "num_tokens": 9254998.0,
      "step": 11610
    },
    {
      "entropy": 4.850271797180175,
      "epoch": 3.283856375459429,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040210776819610527,
      "loss": 4.4329,
      "mean_token_accuracy": 0.2678584411740303,
      "num_tokens": 9259023.0,
      "step": 11615
    },
    {
      "entropy": 5.021196985244751,
      "epoch": 3.2852700028272546,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004020228801408866,
      "loss": 4.6685,
      "mean_token_accuracy": 0.24355312436819077,
      "num_tokens": 9263230.0,
      "step": 11620
    },
    {
      "entropy": 4.774869298934936,
      "epoch": 3.2866836301950806,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040193796553867137,
      "loss": 4.4327,
      "mean_token_accuracy": 0.26816623359918595,
      "num_tokens": 9267296.0,
      "step": 11625
    },
    {
      "entropy": 4.8111371994018555,
      "epoch": 3.2880972575629066,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00040185302440720627,
      "loss": 4.4689,
      "mean_token_accuracy": 0.2636589899659157,
      "num_tokens": 9271535.0,
      "step": 11630
    },
    {
      "entropy": 4.926999711990357,
      "epoch": 3.289510884930732,
      "grad_norm": 1.375,
      "learning_rate": 0.0004017680567642435,
      "loss": 4.5456,
      "mean_token_accuracy": 0.25008263885974885,
      "num_tokens": 9275543.0,
      "step": 11635
    },
    {
      "entropy": 4.745850944519043,
      "epoch": 3.290924512298558,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004016830626275409,
      "loss": 4.3616,
      "mean_token_accuracy": 0.272255602478981,
      "num_tokens": 9279438.0,
      "step": 11640
    },
    {
      "entropy": 4.756447839736938,
      "epoch": 3.292338139666384,
      "grad_norm": 1.390625,
      "learning_rate": 0.00040159804201486156,
      "loss": 4.4324,
      "mean_token_accuracy": 0.26966148614883423,
      "num_tokens": 9283615.0,
      "step": 11645
    },
    {
      "entropy": 4.864556360244751,
      "epoch": 3.29375176703421,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004015129949439743,
      "loss": 4.5179,
      "mean_token_accuracy": 0.2486346811056137,
      "num_tokens": 9287593.0,
      "step": 11650
    },
    {
      "entropy": 4.841293430328369,
      "epoch": 3.2951653944020354,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00040142792143265337,
      "loss": 4.4685,
      "mean_token_accuracy": 0.26187690794467927,
      "num_tokens": 9291398.0,
      "step": 11655
    },
    {
      "entropy": 4.859081506729126,
      "epoch": 3.2965790217698614,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004013428214986787,
      "loss": 4.4786,
      "mean_token_accuracy": 0.2634968876838684,
      "num_tokens": 9295340.0,
      "step": 11660
    },
    {
      "entropy": 4.855468797683716,
      "epoch": 3.2979926491376874,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040125769515983564,
      "loss": 4.4328,
      "mean_token_accuracy": 0.2608964815735817,
      "num_tokens": 9299462.0,
      "step": 11665
    },
    {
      "entropy": 4.8139759540557865,
      "epoch": 3.299406276505513,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00040117254243391505,
      "loss": 4.4917,
      "mean_token_accuracy": 0.25454739928245546,
      "num_tokens": 9303330.0,
      "step": 11670
    },
    {
      "entropy": 4.86029634475708,
      "epoch": 3.300819903873339,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004010873633387133,
      "loss": 4.571,
      "mean_token_accuracy": 0.26291003823280334,
      "num_tokens": 9307247.0,
      "step": 11675
    },
    {
      "entropy": 4.848725318908691,
      "epoch": 3.302233531241165,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00040100215789203226,
      "loss": 4.4925,
      "mean_token_accuracy": 0.2600388079881668,
      "num_tokens": 9311340.0,
      "step": 11680
    },
    {
      "entropy": 4.83942723274231,
      "epoch": 3.3036471586089906,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004009169261116794,
      "loss": 4.4166,
      "mean_token_accuracy": 0.2839584320783615,
      "num_tokens": 9315451.0,
      "step": 11685
    },
    {
      "entropy": 4.906036806106568,
      "epoch": 3.3050607859768166,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004008316680154678,
      "loss": 4.5209,
      "mean_token_accuracy": 0.2553615361452103,
      "num_tokens": 9319657.0,
      "step": 11690
    },
    {
      "entropy": 4.918138313293457,
      "epoch": 3.306474413344642,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00040074638362121547,
      "loss": 4.4041,
      "mean_token_accuracy": 0.2638064756989479,
      "num_tokens": 9323702.0,
      "step": 11695
    },
    {
      "entropy": 4.827712059020996,
      "epoch": 3.3078880407124682,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004006610729467467,
      "loss": 4.4629,
      "mean_token_accuracy": 0.26854889243841173,
      "num_tokens": 9327701.0,
      "step": 11700
    },
    {
      "entropy": 4.794729280471802,
      "epoch": 3.309301668080294,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004005757360098907,
      "loss": 4.4936,
      "mean_token_accuracy": 0.26157944053411486,
      "num_tokens": 9331328.0,
      "step": 11705
    },
    {
      "entropy": 4.756185007095337,
      "epoch": 3.31071529544812,
      "grad_norm": 1.375,
      "learning_rate": 0.00040049037282848237,
      "loss": 4.4396,
      "mean_token_accuracy": 0.26310967803001406,
      "num_tokens": 9335187.0,
      "step": 11710
    },
    {
      "entropy": 4.845719718933106,
      "epoch": 3.312128922815946,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00040040498342036216,
      "loss": 4.5268,
      "mean_token_accuracy": 0.2610419914126396,
      "num_tokens": 9339167.0,
      "step": 11715
    },
    {
      "entropy": 4.814881372451782,
      "epoch": 3.3135425501837714,
      "grad_norm": 1.453125,
      "learning_rate": 0.000400319567803376,
      "loss": 4.4484,
      "mean_token_accuracy": 0.26064361482858656,
      "num_tokens": 9343172.0,
      "step": 11720
    },
    {
      "entropy": 4.782169198989868,
      "epoch": 3.3149561775515974,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040023412599537496,
      "loss": 4.4932,
      "mean_token_accuracy": 0.2694997534155846,
      "num_tokens": 9347296.0,
      "step": 11725
    },
    {
      "entropy": 4.763385009765625,
      "epoch": 3.3163698049194235,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004001486580142161,
      "loss": 4.417,
      "mean_token_accuracy": 0.27314908653497694,
      "num_tokens": 9350919.0,
      "step": 11730
    },
    {
      "entropy": 4.814848566055298,
      "epoch": 3.317783432287249,
      "grad_norm": 1.40625,
      "learning_rate": 0.00040006316387776166,
      "loss": 4.4061,
      "mean_token_accuracy": 0.26101131588220594,
      "num_tokens": 9354748.0,
      "step": 11735
    },
    {
      "entropy": 4.671590375900268,
      "epoch": 3.319197059655075,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003999776436038792,
      "loss": 4.3467,
      "mean_token_accuracy": 0.27772393375635146,
      "num_tokens": 9358794.0,
      "step": 11740
    },
    {
      "entropy": 4.7437032699584964,
      "epoch": 3.3206106870229006,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039989209721044217,
      "loss": 4.3441,
      "mean_token_accuracy": 0.27173826545476915,
      "num_tokens": 9362882.0,
      "step": 11745
    },
    {
      "entropy": 4.842611074447632,
      "epoch": 3.3220243143907267,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003998065247153291,
      "loss": 4.5442,
      "mean_token_accuracy": 0.26162564009428024,
      "num_tokens": 9367004.0,
      "step": 11750
    },
    {
      "entropy": 4.815105819702149,
      "epoch": 3.3234379417585522,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003997209261364241,
      "loss": 4.5379,
      "mean_token_accuracy": 0.2503095865249634,
      "num_tokens": 9371259.0,
      "step": 11755
    },
    {
      "entropy": 4.81246337890625,
      "epoch": 3.3248515691263782,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039963530149161685,
      "loss": 4.4765,
      "mean_token_accuracy": 0.26586395502090454,
      "num_tokens": 9375296.0,
      "step": 11760
    },
    {
      "entropy": 4.7608545303344725,
      "epoch": 3.3262651964942043,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003995496507988023,
      "loss": 4.4155,
      "mean_token_accuracy": 0.26774235963821413,
      "num_tokens": 9379505.0,
      "step": 11765
    },
    {
      "entropy": 4.764514636993408,
      "epoch": 3.32767882386203,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003994639740758809,
      "loss": 4.489,
      "mean_token_accuracy": 0.25749047100543976,
      "num_tokens": 9383450.0,
      "step": 11770
    },
    {
      "entropy": 4.897027397155762,
      "epoch": 3.329092451229856,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00039937827134075866,
      "loss": 4.5773,
      "mean_token_accuracy": 0.24832695573568345,
      "num_tokens": 9387632.0,
      "step": 11775
    },
    {
      "entropy": 4.817900133132935,
      "epoch": 3.330506078597682,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003992925426113468,
      "loss": 4.4198,
      "mean_token_accuracy": 0.2604187771677971,
      "num_tokens": 9391768.0,
      "step": 11780
    },
    {
      "entropy": 4.814379119873047,
      "epoch": 3.3319197059655075,
      "grad_norm": 1.484375,
      "learning_rate": 0.000399206787905562,
      "loss": 4.4336,
      "mean_token_accuracy": 0.25692488104104994,
      "num_tokens": 9395606.0,
      "step": 11785
    },
    {
      "entropy": 4.777588176727295,
      "epoch": 3.3333333333333335,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003991210072413268,
      "loss": 4.5755,
      "mean_token_accuracy": 0.2600775882601738,
      "num_tokens": 9399624.0,
      "step": 11790
    },
    {
      "entropy": 4.860274887084961,
      "epoch": 3.334746960701159,
      "grad_norm": 1.375,
      "learning_rate": 0.0003990352006365685,
      "loss": 4.5507,
      "mean_token_accuracy": 0.254029369354248,
      "num_tokens": 9403537.0,
      "step": 11795
    },
    {
      "entropy": 4.820262336730957,
      "epoch": 3.336160588068985,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003989493681092204,
      "loss": 4.4496,
      "mean_token_accuracy": 0.27350431978702544,
      "num_tokens": 9407496.0,
      "step": 11800
    },
    {
      "entropy": 4.882473182678223,
      "epoch": 3.3375742154368107,
      "grad_norm": 1.5,
      "learning_rate": 0.00039886350967722074,
      "loss": 4.5263,
      "mean_token_accuracy": 0.25446031987667084,
      "num_tokens": 9411397.0,
      "step": 11805
    },
    {
      "entropy": 4.682826280593872,
      "epoch": 3.3389878428046367,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003987776253585135,
      "loss": 4.3675,
      "mean_token_accuracy": 0.2813922196626663,
      "num_tokens": 9415404.0,
      "step": 11810
    },
    {
      "entropy": 4.774057865142822,
      "epoch": 3.3404014701724627,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00039869171517104804,
      "loss": 4.4142,
      "mean_token_accuracy": 0.2640749245882034,
      "num_tokens": 9419509.0,
      "step": 11815
    },
    {
      "entropy": 4.896359634399414,
      "epoch": 3.3418150975402883,
      "grad_norm": 1.5,
      "learning_rate": 0.000398605779132779,
      "loss": 4.5118,
      "mean_token_accuracy": 0.2581461250782013,
      "num_tokens": 9423516.0,
      "step": 11820
    },
    {
      "entropy": 4.839256954193115,
      "epoch": 3.3432287249081143,
      "grad_norm": 1.625,
      "learning_rate": 0.0003985198172616664,
      "loss": 4.5174,
      "mean_token_accuracy": 0.25860830396413803,
      "num_tokens": 9427337.0,
      "step": 11825
    },
    {
      "entropy": 4.825803565979004,
      "epoch": 3.3446423522759403,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039843382957567585,
      "loss": 4.4028,
      "mean_token_accuracy": 0.2652476578950882,
      "num_tokens": 9431188.0,
      "step": 11830
    },
    {
      "entropy": 4.754788732528686,
      "epoch": 3.346055979643766,
      "grad_norm": 1.5,
      "learning_rate": 0.0003983478160927782,
      "loss": 4.4203,
      "mean_token_accuracy": 0.2773317933082581,
      "num_tokens": 9434877.0,
      "step": 11835
    },
    {
      "entropy": 4.801159334182739,
      "epoch": 3.347469607011592,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039826177683094984,
      "loss": 4.435,
      "mean_token_accuracy": 0.2582353547215462,
      "num_tokens": 9438937.0,
      "step": 11840
    },
    {
      "entropy": 4.725914716720581,
      "epoch": 3.3488832343794175,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003981757118081722,
      "loss": 4.4719,
      "mean_token_accuracy": 0.2665604516863823,
      "num_tokens": 9443167.0,
      "step": 11845
    },
    {
      "entropy": 4.814371395111084,
      "epoch": 3.3502968617472435,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003980896210424326,
      "loss": 4.5131,
      "mean_token_accuracy": 0.2532465010881424,
      "num_tokens": 9446946.0,
      "step": 11850
    },
    {
      "entropy": 4.782845544815063,
      "epoch": 3.351710489115069,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003980035045517233,
      "loss": 4.4131,
      "mean_token_accuracy": 0.27295090556144713,
      "num_tokens": 9450886.0,
      "step": 11855
    },
    {
      "entropy": 4.8095615863800045,
      "epoch": 3.353124116482895,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003979173623540423,
      "loss": 4.4985,
      "mean_token_accuracy": 0.2658781290054321,
      "num_tokens": 9454896.0,
      "step": 11860
    },
    {
      "entropy": 4.844632244110107,
      "epoch": 3.354537743850721,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003978311944673925,
      "loss": 4.5317,
      "mean_token_accuracy": 0.2603448539972305,
      "num_tokens": 9458730.0,
      "step": 11865
    },
    {
      "entropy": 4.909362077713013,
      "epoch": 3.3559513712185467,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039774500090978274,
      "loss": 4.5164,
      "mean_token_accuracy": 0.25695322155952455,
      "num_tokens": 9462909.0,
      "step": 11870
    },
    {
      "entropy": 4.903599262237549,
      "epoch": 3.3573649985863727,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039765878169922673,
      "loss": 4.5771,
      "mean_token_accuracy": 0.25979782789945605,
      "num_tokens": 9467145.0,
      "step": 11875
    },
    {
      "entropy": 4.770998001098633,
      "epoch": 3.3587786259541983,
      "grad_norm": 1.5,
      "learning_rate": 0.00039757253685374387,
      "loss": 4.3839,
      "mean_token_accuracy": 0.26965268403291703,
      "num_tokens": 9471095.0,
      "step": 11880
    },
    {
      "entropy": 4.816092014312744,
      "epoch": 3.3601922533220243,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039748626639135885,
      "loss": 4.3634,
      "mean_token_accuracy": 0.27514705508947374,
      "num_tokens": 9474857.0,
      "step": 11885
    },
    {
      "entropy": 4.873953819274902,
      "epoch": 3.3616058806898503,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00039739997033010153,
      "loss": 4.4661,
      "mean_token_accuracy": 0.2597108379006386,
      "num_tokens": 9478676.0,
      "step": 11890
    },
    {
      "entropy": 4.828252458572388,
      "epoch": 3.363019508057676,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003973136486880073,
      "loss": 4.3989,
      "mean_token_accuracy": 0.28253997266292574,
      "num_tokens": 9482774.0,
      "step": 11895
    },
    {
      "entropy": 4.828372573852539,
      "epoch": 3.364433135425502,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003972273014831169,
      "loss": 4.4661,
      "mean_token_accuracy": 0.2546294629573822,
      "num_tokens": 9486640.0,
      "step": 11900
    },
    {
      "entropy": 4.748572206497192,
      "epoch": 3.3658467627933275,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039714092873347623,
      "loss": 4.5506,
      "mean_token_accuracy": 0.2547990188002586,
      "num_tokens": 9490425.0,
      "step": 11905
    },
    {
      "entropy": 4.846953344345093,
      "epoch": 3.3672603901611535,
      "grad_norm": 1.5,
      "learning_rate": 0.00039705453045713686,
      "loss": 4.5945,
      "mean_token_accuracy": 0.2519740447402,
      "num_tokens": 9494514.0,
      "step": 11910
    },
    {
      "entropy": 4.856832504272461,
      "epoch": 3.3686740175289795,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003969681066721553,
      "loss": 4.5085,
      "mean_token_accuracy": 0.2681466445326805,
      "num_tokens": 9498485.0,
      "step": 11915
    },
    {
      "entropy": 4.894257640838623,
      "epoch": 3.370087644896805,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003968816573965938,
      "loss": 4.4771,
      "mean_token_accuracy": 0.2712566822767258,
      "num_tokens": 9502540.0,
      "step": 11920
    },
    {
      "entropy": 4.824291563034057,
      "epoch": 3.371501272264631,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039679518264851944,
      "loss": 4.4693,
      "mean_token_accuracy": 0.2671370908617973,
      "num_tokens": 9506645.0,
      "step": 11925
    },
    {
      "entropy": 4.836324834823609,
      "epoch": 3.3729148996324567,
      "grad_norm": 1.453125,
      "learning_rate": 0.00039670868244600515,
      "loss": 4.4203,
      "mean_token_accuracy": 0.27058052122592924,
      "num_tokens": 9510894.0,
      "step": 11930
    },
    {
      "entropy": 4.816177892684936,
      "epoch": 3.3743285270002827,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003966221568071288,
      "loss": 4.535,
      "mean_token_accuracy": 0.25915530174970625,
      "num_tokens": 9514990.0,
      "step": 11935
    },
    {
      "entropy": 4.82040810585022,
      "epoch": 3.3757421543681088,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00039653560574997383,
      "loss": 4.4785,
      "mean_token_accuracy": 0.25040440559387206,
      "num_tokens": 9518815.0,
      "step": 11940
    },
    {
      "entropy": 4.8559120178222654,
      "epoch": 3.3771557817359343,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003964490292926288,
      "loss": 4.4627,
      "mean_token_accuracy": 0.2628237545490265,
      "num_tokens": 9522736.0,
      "step": 11945
    },
    {
      "entropy": 4.820240163803101,
      "epoch": 3.3785694091037604,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003963624274531877,
      "loss": 4.4899,
      "mean_token_accuracy": 0.26386402994394303,
      "num_tokens": 9526995.0,
      "step": 11950
    },
    {
      "entropy": 4.748870420455932,
      "epoch": 3.379983036471586,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003962758002497497,
      "loss": 4.4243,
      "mean_token_accuracy": 0.26155407875776293,
      "num_tokens": 9531271.0,
      "step": 11955
    },
    {
      "entropy": 4.786216974258423,
      "epoch": 3.381396663839412,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039618914770041936,
      "loss": 4.4573,
      "mean_token_accuracy": 0.27092886716127396,
      "num_tokens": 9535337.0,
      "step": 11960
    },
    {
      "entropy": 4.761487531661987,
      "epoch": 3.382810291207238,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003961024698233067,
      "loss": 4.3918,
      "mean_token_accuracy": 0.27120794653892516,
      "num_tokens": 9539526.0,
      "step": 11965
    },
    {
      "entropy": 4.725912570953369,
      "epoch": 3.3842239185750635,
      "grad_norm": 1.375,
      "learning_rate": 0.0003960157666365265,
      "loss": 4.4223,
      "mean_token_accuracy": 0.26291935294866564,
      "num_tokens": 9543563.0,
      "step": 11970
    },
    {
      "entropy": 4.735417413711548,
      "epoch": 3.3856375459428896,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003959290381581996,
      "loss": 4.4343,
      "mean_token_accuracy": 0.25969979614019395,
      "num_tokens": 9547389.0,
      "step": 11975
    },
    {
      "entropy": 4.894664525985718,
      "epoch": 3.387051173310715,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003958422844064514,
      "loss": 4.5424,
      "mean_token_accuracy": 0.2614024430513382,
      "num_tokens": 9551721.0,
      "step": 11980
    },
    {
      "entropy": 4.861048936843872,
      "epoch": 3.388464800678541,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039575550539941315,
      "loss": 4.5652,
      "mean_token_accuracy": 0.2639038637280464,
      "num_tokens": 9555610.0,
      "step": 11985
    },
    {
      "entropy": 4.916353750228882,
      "epoch": 3.3898784280463667,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003956687011552209,
      "loss": 4.5197,
      "mean_token_accuracy": 0.25161915123462675,
      "num_tokens": 9559618.0,
      "step": 11990
    },
    {
      "entropy": 4.8018158912658695,
      "epoch": 3.3912920554141928,
      "grad_norm": 1.390625,
      "learning_rate": 0.00039558187169201636,
      "loss": 4.4962,
      "mean_token_accuracy": 0.262605819106102,
      "num_tokens": 9563717.0,
      "step": 11995
    },
    {
      "entropy": 4.761739444732666,
      "epoch": 3.392705682782019,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00039549501702794616,
      "loss": 4.3856,
      "mean_token_accuracy": 0.27905588299036027,
      "num_tokens": 9567550.0,
      "step": 12000
    },
    {
      "epoch": 3.392705682782019,
      "eval_entropy": 4.5829856615384195,
      "eval_loss": 5.0075883865356445,
      "eval_mean_token_accuracy": 0.2476578903464372,
      "eval_num_tokens": 9567550.0,
      "eval_runtime": 4.8743,
      "eval_samples_per_second": 2315.018,
      "eval_steps_per_second": 289.48,
      "step": 12000
    },
    {
      "entropy": 4.809433364868164,
      "epoch": 3.3941193101498444,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003954081371811627,
      "loss": 4.5127,
      "mean_token_accuracy": 0.26408721655607226,
      "num_tokens": 9571316.0,
      "step": 12005
    },
    {
      "entropy": 4.734455919265747,
      "epoch": 3.3955329375176704,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003953212321698231,
      "loss": 4.4174,
      "mean_token_accuracy": 0.2713135316967964,
      "num_tokens": 9575231.0,
      "step": 12010
    },
    {
      "entropy": 4.746622848510742,
      "epoch": 3.3969465648854964,
      "grad_norm": 1.484375,
      "learning_rate": 0.00039523430201209,
      "loss": 4.4114,
      "mean_token_accuracy": 0.268108007311821,
      "num_tokens": 9579180.0,
      "step": 12015
    },
    {
      "entropy": 4.8090392589569095,
      "epoch": 3.398360192253322,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039514734672613125,
      "loss": 4.4895,
      "mean_token_accuracy": 0.26079668998718264,
      "num_tokens": 9583171.0,
      "step": 12020
    },
    {
      "entropy": 4.709369087219239,
      "epoch": 3.399773819621148,
      "grad_norm": 1.578125,
      "learning_rate": 0.00039506036633012,
      "loss": 4.3712,
      "mean_token_accuracy": 0.2721718207001686,
      "num_tokens": 9586780.0,
      "step": 12025
    },
    {
      "entropy": 4.8107287883758545,
      "epoch": 3.4011874469889736,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003949733608422347,
      "loss": 4.4657,
      "mean_token_accuracy": 0.26723330169916154,
      "num_tokens": 9591225.0,
      "step": 12030
    },
    {
      "entropy": 4.8450737476348875,
      "epoch": 3.4026010743567996,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00039488633028065883,
      "loss": 4.5071,
      "mean_token_accuracy": 0.2646340489387512,
      "num_tokens": 9595179.0,
      "step": 12035
    },
    {
      "entropy": 4.810352230072022,
      "epoch": 3.404014701724625,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00039479927466358125,
      "loss": 4.5478,
      "mean_token_accuracy": 0.2632232949137688,
      "num_tokens": 9599325.0,
      "step": 12040
    },
    {
      "entropy": 4.8312903881073,
      "epoch": 3.405428329092451,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003947121940091962,
      "loss": 4.5245,
      "mean_token_accuracy": 0.2561719983816147,
      "num_tokens": 9603760.0,
      "step": 12045
    },
    {
      "entropy": 4.784038019180298,
      "epoch": 3.406841956460277,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039462508833570275,
      "loss": 4.4623,
      "mean_token_accuracy": 0.2717742547392845,
      "num_tokens": 9607826.0,
      "step": 12050
    },
    {
      "entropy": 4.807305717468262,
      "epoch": 3.408255583828103,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039453795766130565,
      "loss": 4.5535,
      "mean_token_accuracy": 0.24842023998498916,
      "num_tokens": 9611859.0,
      "step": 12055
    },
    {
      "entropy": 4.887192440032959,
      "epoch": 3.409669211195929,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003944508020042145,
      "loss": 4.5831,
      "mean_token_accuracy": 0.24703286588191986,
      "num_tokens": 9615938.0,
      "step": 12060
    },
    {
      "entropy": 4.93574104309082,
      "epoch": 3.411082838563755,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003943636213826445,
      "loss": 4.5221,
      "mean_token_accuracy": 0.2632774382829666,
      "num_tokens": 9619722.0,
      "step": 12065
    },
    {
      "entropy": 4.781073808670044,
      "epoch": 3.4124964659315804,
      "grad_norm": 1.5625,
      "learning_rate": 0.00039427641581481573,
      "loss": 4.4907,
      "mean_token_accuracy": 0.26215250194072726,
      "num_tokens": 9623623.0,
      "step": 12070
    },
    {
      "entropy": 4.894560670852661,
      "epoch": 3.4139100932994064,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003941891853189536,
      "loss": 4.6062,
      "mean_token_accuracy": 0.2549786314368248,
      "num_tokens": 9627515.0,
      "step": 12075
    },
    {
      "entropy": 4.837924528121948,
      "epoch": 3.415323720667232,
      "grad_norm": 1.359375,
      "learning_rate": 0.00039410192991328884,
      "loss": 4.4384,
      "mean_token_accuracy": 0.26864873617887497,
      "num_tokens": 9631513.0,
      "step": 12080
    },
    {
      "entropy": 4.846583843231201,
      "epoch": 3.416737348035058,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003940146496160573,
      "loss": 4.4862,
      "mean_token_accuracy": 0.25882298797369,
      "num_tokens": 9635425.0,
      "step": 12085
    },
    {
      "entropy": 4.777957201004028,
      "epoch": 3.4181509754028836,
      "grad_norm": 1.421875,
      "learning_rate": 0.00039392734444549976,
      "loss": 4.4176,
      "mean_token_accuracy": 0.26225718557834626,
      "num_tokens": 9639144.0,
      "step": 12090
    },
    {
      "entropy": 4.830684232711792,
      "epoch": 3.4195646027707096,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003938400144198628,
      "loss": 4.4989,
      "mean_token_accuracy": 0.26675003170967104,
      "num_tokens": 9643492.0,
      "step": 12095
    },
    {
      "entropy": 4.744479370117188,
      "epoch": 3.4209782301385356,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003937526595573976,
      "loss": 4.3043,
      "mean_token_accuracy": 0.28800593316555023,
      "num_tokens": 9647510.0,
      "step": 12100
    },
    {
      "entropy": 4.847334957122802,
      "epoch": 3.422391857506361,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039366527987636104,
      "loss": 4.5623,
      "mean_token_accuracy": 0.2663918539881706,
      "num_tokens": 9651586.0,
      "step": 12105
    },
    {
      "entropy": 4.837295770645142,
      "epoch": 3.4238054848741872,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039357787539501466,
      "loss": 4.6002,
      "mean_token_accuracy": 0.2546368479728699,
      "num_tokens": 9655567.0,
      "step": 12110
    },
    {
      "entropy": 4.948073005676269,
      "epoch": 3.4252191122420133,
      "grad_norm": 1.515625,
      "learning_rate": 0.00039349044613162565,
      "loss": 4.6338,
      "mean_token_accuracy": 0.2445577397942543,
      "num_tokens": 9659271.0,
      "step": 12115
    },
    {
      "entropy": 4.8627911567687985,
      "epoch": 3.426632739609839,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00039340299210446607,
      "loss": 4.4908,
      "mean_token_accuracy": 0.26340765655040743,
      "num_tokens": 9663145.0,
      "step": 12120
    },
    {
      "entropy": 4.913440132141114,
      "epoch": 3.428046366977665,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039331551333181335,
      "loss": 4.5358,
      "mean_token_accuracy": 0.2540408968925476,
      "num_tokens": 9666806.0,
      "step": 12125
    },
    {
      "entropy": 4.811867666244507,
      "epoch": 3.4294599943454904,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00039322800983195,
      "loss": 4.3925,
      "mean_token_accuracy": 0.26842064410448074,
      "num_tokens": 9670619.0,
      "step": 12130
    },
    {
      "entropy": 4.7341776371002195,
      "epoch": 3.4308736217133164,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00039314048162316376,
      "loss": 4.4108,
      "mean_token_accuracy": 0.2695192500948906,
      "num_tokens": 9674600.0,
      "step": 12135
    },
    {
      "entropy": 4.78953595161438,
      "epoch": 3.432287249081142,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00039305292872374734,
      "loss": 4.463,
      "mean_token_accuracy": 0.25920013785362245,
      "num_tokens": 9678535.0,
      "step": 12140
    },
    {
      "entropy": 4.834987592697144,
      "epoch": 3.433700876448968,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00039296535115199887,
      "loss": 4.4124,
      "mean_token_accuracy": 0.27807012647390367,
      "num_tokens": 9682428.0,
      "step": 12145
    },
    {
      "entropy": 4.893732357025146,
      "epoch": 3.435114503816794,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003928777489262214,
      "loss": 4.6135,
      "mean_token_accuracy": 0.24926285296678544,
      "num_tokens": 9686497.0,
      "step": 12150
    },
    {
      "entropy": 4.818486547470092,
      "epoch": 3.4365281311846196,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003927901220647234,
      "loss": 4.4861,
      "mean_token_accuracy": 0.2663331374526024,
      "num_tokens": 9690468.0,
      "step": 12155
    },
    {
      "entropy": 4.822831773757935,
      "epoch": 3.4379417585524457,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003927024705858183,
      "loss": 4.3808,
      "mean_token_accuracy": 0.2725465580821037,
      "num_tokens": 9694041.0,
      "step": 12160
    },
    {
      "entropy": 4.738700485229492,
      "epoch": 3.4393553859202712,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039261479450782466,
      "loss": 4.3211,
      "mean_token_accuracy": 0.28103157132864,
      "num_tokens": 9698094.0,
      "step": 12165
    },
    {
      "entropy": 4.877137660980225,
      "epoch": 3.4407690132880973,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003925270938490663,
      "loss": 4.4615,
      "mean_token_accuracy": 0.2761657044291496,
      "num_tokens": 9701868.0,
      "step": 12170
    },
    {
      "entropy": 4.711083602905274,
      "epoch": 3.4421826406559233,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003924393686278721,
      "loss": 4.413,
      "mean_token_accuracy": 0.27119372338056563,
      "num_tokens": 9705912.0,
      "step": 12175
    },
    {
      "entropy": 4.778292894363403,
      "epoch": 3.443596268023749,
      "grad_norm": 1.40625,
      "learning_rate": 0.000392351618862576,
      "loss": 4.4188,
      "mean_token_accuracy": 0.2622037649154663,
      "num_tokens": 9709893.0,
      "step": 12180
    },
    {
      "entropy": 4.840904951095581,
      "epoch": 3.445009895391575,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003922638445715174,
      "loss": 4.4958,
      "mean_token_accuracy": 0.2588240757584572,
      "num_tokens": 9713786.0,
      "step": 12185
    },
    {
      "entropy": 4.769951295852661,
      "epoch": 3.4464235227594004,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003921760457730403,
      "loss": 4.419,
      "mean_token_accuracy": 0.2637442111968994,
      "num_tokens": 9717914.0,
      "step": 12190
    },
    {
      "entropy": 4.838496160507202,
      "epoch": 3.4478371501272265,
      "grad_norm": 1.375,
      "learning_rate": 0.0003920882224854942,
      "loss": 4.5211,
      "mean_token_accuracy": 0.25414435416460035,
      "num_tokens": 9721901.0,
      "step": 12195
    },
    {
      "entropy": 4.838874387741089,
      "epoch": 3.4492507774950525,
      "grad_norm": 1.375,
      "learning_rate": 0.00039200037472723377,
      "loss": 4.5101,
      "mean_token_accuracy": 0.2597037896513939,
      "num_tokens": 9726371.0,
      "step": 12200
    },
    {
      "entropy": 4.808909273147583,
      "epoch": 3.450664404862878,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003919125025166185,
      "loss": 4.4496,
      "mean_token_accuracy": 0.25829310715198517,
      "num_tokens": 9730501.0,
      "step": 12205
    },
    {
      "entropy": 4.782706880569458,
      "epoch": 3.452078032230704,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003918246058720131,
      "loss": 4.372,
      "mean_token_accuracy": 0.2658412307500839,
      "num_tokens": 9734500.0,
      "step": 12210
    },
    {
      "entropy": 4.797542381286621,
      "epoch": 3.4534916595985297,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003917366848117875,
      "loss": 4.5051,
      "mean_token_accuracy": 0.2590572193264961,
      "num_tokens": 9738258.0,
      "step": 12215
    },
    {
      "entropy": 4.943078804016113,
      "epoch": 3.4549052869663557,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003916487393543167,
      "loss": 4.5915,
      "mean_token_accuracy": 0.2503367155790329,
      "num_tokens": 9742373.0,
      "step": 12220
    },
    {
      "entropy": 4.807430744171143,
      "epoch": 3.4563189143341817,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00039156076951798064,
      "loss": 4.4731,
      "mean_token_accuracy": 0.2666779413819313,
      "num_tokens": 9746296.0,
      "step": 12225
    },
    {
      "entropy": 4.815003871917725,
      "epoch": 3.4577325417020073,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003914727753211647,
      "loss": 4.5459,
      "mean_token_accuracy": 0.2616420701146126,
      "num_tokens": 9750545.0,
      "step": 12230
    },
    {
      "entropy": 4.813952875137329,
      "epoch": 3.4591461690698333,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003913847567822588,
      "loss": 4.4778,
      "mean_token_accuracy": 0.2564672440290451,
      "num_tokens": 9754504.0,
      "step": 12235
    },
    {
      "entropy": 4.814035511016845,
      "epoch": 3.460559796437659,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003912967139196585,
      "loss": 4.5102,
      "mean_token_accuracy": 0.2529383048415184,
      "num_tokens": 9758878.0,
      "step": 12240
    },
    {
      "entropy": 4.7921422004699705,
      "epoch": 3.461973423805485,
      "grad_norm": 1.46875,
      "learning_rate": 0.00039120864675176416,
      "loss": 4.3999,
      "mean_token_accuracy": 0.26679577976465224,
      "num_tokens": 9762818.0,
      "step": 12245
    },
    {
      "entropy": 4.799564361572266,
      "epoch": 3.463387051173311,
      "grad_norm": 1.484375,
      "learning_rate": 0.00039112055529698126,
      "loss": 4.3764,
      "mean_token_accuracy": 0.2665384814143181,
      "num_tokens": 9766789.0,
      "step": 12250
    },
    {
      "entropy": 4.941089868545532,
      "epoch": 3.4648006785411365,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003910324395737203,
      "loss": 4.5927,
      "mean_token_accuracy": 0.25415632128715515,
      "num_tokens": 9770721.0,
      "step": 12255
    },
    {
      "entropy": 4.778097438812256,
      "epoch": 3.4662143059089625,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00039094429960039707,
      "loss": 4.3685,
      "mean_token_accuracy": 0.2770001068711281,
      "num_tokens": 9774568.0,
      "step": 12260
    },
    {
      "entropy": 4.762670373916626,
      "epoch": 3.467627933276788,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003908561353954322,
      "loss": 4.3597,
      "mean_token_accuracy": 0.26723783612251284,
      "num_tokens": 9778694.0,
      "step": 12265
    },
    {
      "entropy": 4.8171038150787355,
      "epoch": 3.469041560644614,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003907679469772514,
      "loss": 4.4931,
      "mean_token_accuracy": 0.250614732503891,
      "num_tokens": 9782501.0,
      "step": 12270
    },
    {
      "entropy": 4.775955104827881,
      "epoch": 3.4704551880124397,
      "grad_norm": 1.46875,
      "learning_rate": 0.00039067973436428553,
      "loss": 4.4305,
      "mean_token_accuracy": 0.26261936873197556,
      "num_tokens": 9786410.0,
      "step": 12275
    },
    {
      "entropy": 4.82211856842041,
      "epoch": 3.4718688153802657,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003905914975749707,
      "loss": 4.6125,
      "mean_token_accuracy": 0.2513530597090721,
      "num_tokens": 9790575.0,
      "step": 12280
    },
    {
      "entropy": 4.879493141174317,
      "epoch": 3.4732824427480917,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003905032366277475,
      "loss": 4.448,
      "mean_token_accuracy": 0.26213698387145995,
      "num_tokens": 9794440.0,
      "step": 12285
    },
    {
      "entropy": 4.998200416564941,
      "epoch": 3.4746960701159173,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003904149515410622,
      "loss": 4.6223,
      "mean_token_accuracy": 0.2518760398030281,
      "num_tokens": 9798287.0,
      "step": 12290
    },
    {
      "entropy": 4.865606355667114,
      "epoch": 3.4761096974837433,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003903266423333657,
      "loss": 4.454,
      "mean_token_accuracy": 0.26835622936487197,
      "num_tokens": 9802177.0,
      "step": 12295
    },
    {
      "entropy": 4.825556755065918,
      "epoch": 3.4775233248515693,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003902383090231141,
      "loss": 4.4731,
      "mean_token_accuracy": 0.2598146662116051,
      "num_tokens": 9805906.0,
      "step": 12300
    },
    {
      "entropy": 4.825527954101562,
      "epoch": 3.478936952219395,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003901499516287685,
      "loss": 4.543,
      "mean_token_accuracy": 0.25282507836818696,
      "num_tokens": 9809599.0,
      "step": 12305
    },
    {
      "entropy": 4.782500696182251,
      "epoch": 3.480350579587221,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003900615701687952,
      "loss": 4.3601,
      "mean_token_accuracy": 0.27591018974781034,
      "num_tokens": 9813439.0,
      "step": 12310
    },
    {
      "entropy": 4.71904354095459,
      "epoch": 3.4817642069550465,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003899731646616651,
      "loss": 4.4293,
      "mean_token_accuracy": 0.27002314329147337,
      "num_tokens": 9817386.0,
      "step": 12315
    },
    {
      "entropy": 4.865688943862915,
      "epoch": 3.4831778343228725,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003898847351258547,
      "loss": 4.5445,
      "mean_token_accuracy": 0.24979934394359588,
      "num_tokens": 9821589.0,
      "step": 12320
    },
    {
      "entropy": 4.85878267288208,
      "epoch": 3.484591461690698,
      "grad_norm": 1.375,
      "learning_rate": 0.0003897962815798449,
      "loss": 4.4605,
      "mean_token_accuracy": 0.26568566262722015,
      "num_tokens": 9825671.0,
      "step": 12325
    },
    {
      "entropy": 4.849913454055786,
      "epoch": 3.486005089058524,
      "grad_norm": 1.375,
      "learning_rate": 0.0003897078040421223,
      "loss": 4.5226,
      "mean_token_accuracy": 0.2562889292836189,
      "num_tokens": 9830123.0,
      "step": 12330
    },
    {
      "entropy": 4.756100606918335,
      "epoch": 3.48741871642635,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003896193025311779,
      "loss": 4.4111,
      "mean_token_accuracy": 0.2708584576845169,
      "num_tokens": 9834281.0,
      "step": 12335
    },
    {
      "entropy": 4.820545101165772,
      "epoch": 3.4888323437941757,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003895307770655081,
      "loss": 4.4826,
      "mean_token_accuracy": 0.25502536743879317,
      "num_tokens": 9837707.0,
      "step": 12340
    },
    {
      "entropy": 4.79272084236145,
      "epoch": 3.4902459711620017,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003894422276636141,
      "loss": 4.4146,
      "mean_token_accuracy": 0.27068017721176146,
      "num_tokens": 9841837.0,
      "step": 12345
    },
    {
      "entropy": 4.77443265914917,
      "epoch": 3.4916595985298278,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003893536543440022,
      "loss": 4.3648,
      "mean_token_accuracy": 0.26918865740299225,
      "num_tokens": 9846199.0,
      "step": 12350
    },
    {
      "entropy": 4.769264554977417,
      "epoch": 3.4930732258976533,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003892650571251838,
      "loss": 4.4085,
      "mean_token_accuracy": 0.27405866086483,
      "num_tokens": 9850477.0,
      "step": 12355
    },
    {
      "entropy": 4.715578937530518,
      "epoch": 3.4944868532654794,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003891764360256749,
      "loss": 4.4337,
      "mean_token_accuracy": 0.2758219987154007,
      "num_tokens": 9854639.0,
      "step": 12360
    },
    {
      "entropy": 4.864862108230591,
      "epoch": 3.495900480633305,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00038908779106399695,
      "loss": 4.6387,
      "mean_token_accuracy": 0.25746575742959976,
      "num_tokens": 9858265.0,
      "step": 12365
    },
    {
      "entropy": 4.798770093917847,
      "epoch": 3.497314108001131,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003889991222586762,
      "loss": 4.3784,
      "mean_token_accuracy": 0.2782547816634178,
      "num_tokens": 9862176.0,
      "step": 12370
    },
    {
      "entropy": 4.8223724365234375,
      "epoch": 3.4987277353689565,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00038891042962824386,
      "loss": 4.4683,
      "mean_token_accuracy": 0.2566756993532181,
      "num_tokens": 9866058.0,
      "step": 12375
    },
    {
      "entropy": 4.829181051254272,
      "epoch": 3.5001413627367826,
      "grad_norm": 1.484375,
      "learning_rate": 0.00038882171319123607,
      "loss": 4.4891,
      "mean_token_accuracy": 0.26697990745306016,
      "num_tokens": 9870000.0,
      "step": 12380
    },
    {
      "entropy": 4.791319751739502,
      "epoch": 3.5015549901046086,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003887329729661941,
      "loss": 4.437,
      "mean_token_accuracy": 0.27366954237222674,
      "num_tokens": 9873797.0,
      "step": 12385
    },
    {
      "entropy": 4.766758918762207,
      "epoch": 3.502968617472434,
      "grad_norm": 1.375,
      "learning_rate": 0.00038864420897166414,
      "loss": 4.3346,
      "mean_token_accuracy": 0.2727569743990898,
      "num_tokens": 9877808.0,
      "step": 12390
    },
    {
      "entropy": 4.818148136138916,
      "epoch": 3.50438224484026,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00038855542122619716,
      "loss": 4.4504,
      "mean_token_accuracy": 0.267840701341629,
      "num_tokens": 9881640.0,
      "step": 12395
    },
    {
      "entropy": 4.8322748184204105,
      "epoch": 3.505795872208086,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003884666097483495,
      "loss": 4.4664,
      "mean_token_accuracy": 0.26494406461715697,
      "num_tokens": 9885834.0,
      "step": 12400
    },
    {
      "entropy": 4.844805335998535,
      "epoch": 3.5072094995759118,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038837777455668184,
      "loss": 4.5046,
      "mean_token_accuracy": 0.2553685501217842,
      "num_tokens": 9890240.0,
      "step": 12405
    },
    {
      "entropy": 4.810683298110962,
      "epoch": 3.508623126943738,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003882889156697606,
      "loss": 4.4326,
      "mean_token_accuracy": 0.267569874227047,
      "num_tokens": 9893862.0,
      "step": 12410
    },
    {
      "entropy": 4.773639678955078,
      "epoch": 3.5100367543115634,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038820003310615644,
      "loss": 4.3975,
      "mean_token_accuracy": 0.2644489675760269,
      "num_tokens": 9897552.0,
      "step": 12415
    },
    {
      "entropy": 4.877600908279419,
      "epoch": 3.5114503816793894,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003881111268844454,
      "loss": 4.5714,
      "mean_token_accuracy": 0.24915522038936616,
      "num_tokens": 9901634.0,
      "step": 12420
    },
    {
      "entropy": 4.836810064315796,
      "epoch": 3.512864009047215,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003880221970232083,
      "loss": 4.4994,
      "mean_token_accuracy": 0.24990716576576233,
      "num_tokens": 9905741.0,
      "step": 12425
    },
    {
      "entropy": 4.721572542190552,
      "epoch": 3.514277636415041,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003879332435410309,
      "loss": 4.4049,
      "mean_token_accuracy": 0.26077873855829237,
      "num_tokens": 9909563.0,
      "step": 12430
    },
    {
      "entropy": 4.818407249450684,
      "epoch": 3.515691263782867,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003878442664565039,
      "loss": 4.4629,
      "mean_token_accuracy": 0.26481315195560456,
      "num_tokens": 9913750.0,
      "step": 12435
    },
    {
      "entropy": 4.866672611236572,
      "epoch": 3.5171048911506926,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00038775526578822323,
      "loss": 4.4239,
      "mean_token_accuracy": 0.26588657349348066,
      "num_tokens": 9917653.0,
      "step": 12440
    },
    {
      "entropy": 4.836006927490234,
      "epoch": 3.5185185185185186,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038766624155478914,
      "loss": 4.5154,
      "mean_token_accuracy": 0.26373470276594163,
      "num_tokens": 9921579.0,
      "step": 12445
    },
    {
      "entropy": 4.870648574829102,
      "epoch": 3.5199321458863446,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003875771937748073,
      "loss": 4.5532,
      "mean_token_accuracy": 0.25533377528190615,
      "num_tokens": 9925545.0,
      "step": 12450
    },
    {
      "entropy": 4.770530033111572,
      "epoch": 3.52134577325417,
      "grad_norm": 1.453125,
      "learning_rate": 0.00038748812246688804,
      "loss": 4.4597,
      "mean_token_accuracy": 0.26458604335784913,
      "num_tokens": 9929682.0,
      "step": 12455
    },
    {
      "entropy": 4.764917087554932,
      "epoch": 3.5227594006219958,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003873990276496468,
      "loss": 4.5,
      "mean_token_accuracy": 0.26304744333028796,
      "num_tokens": 9933488.0,
      "step": 12460
    },
    {
      "entropy": 4.871738004684448,
      "epoch": 3.524173027989822,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003873099093417039,
      "loss": 4.5446,
      "mean_token_accuracy": 0.25937757045030596,
      "num_tokens": 9937470.0,
      "step": 12465
    },
    {
      "entropy": 4.757137632369995,
      "epoch": 3.525586655357648,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00038722076756168443,
      "loss": 4.3999,
      "mean_token_accuracy": 0.2756254136562347,
      "num_tokens": 9941139.0,
      "step": 12470
    },
    {
      "entropy": 4.816866159439087,
      "epoch": 3.5270002827254734,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003871316023282185,
      "loss": 4.4951,
      "mean_token_accuracy": 0.26482260078191755,
      "num_tokens": 9944992.0,
      "step": 12475
    },
    {
      "entropy": 4.833708381652832,
      "epoch": 3.5284139100932994,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003870424136599411,
      "loss": 4.5541,
      "mean_token_accuracy": 0.2506968483328819,
      "num_tokens": 9949132.0,
      "step": 12480
    },
    {
      "entropy": 4.876145935058593,
      "epoch": 3.5298275374611254,
      "grad_norm": 1.515625,
      "learning_rate": 0.000386953201575492,
      "loss": 4.5266,
      "mean_token_accuracy": 0.24696842432022095,
      "num_tokens": 9952928.0,
      "step": 12485
    },
    {
      "entropy": 4.893138456344604,
      "epoch": 3.531241164828951,
      "grad_norm": 1.5,
      "learning_rate": 0.00038686396609351615,
      "loss": 4.5109,
      "mean_token_accuracy": 0.2544416680932045,
      "num_tokens": 9957000.0,
      "step": 12490
    },
    {
      "entropy": 4.873467969894409,
      "epoch": 3.532654792196777,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003867747072326631,
      "loss": 4.5191,
      "mean_token_accuracy": 0.2657310292124748,
      "num_tokens": 9960738.0,
      "step": 12495
    },
    {
      "entropy": 4.801416063308716,
      "epoch": 3.534068419564603,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003866854250115875,
      "loss": 4.4137,
      "mean_token_accuracy": 0.2710634216666222,
      "num_tokens": 9964756.0,
      "step": 12500
    },
    {
      "entropy": 4.787413120269775,
      "epoch": 3.5354820469324286,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003865961194489487,
      "loss": 4.3672,
      "mean_token_accuracy": 0.2700983315706253,
      "num_tokens": 9968835.0,
      "step": 12505
    },
    {
      "entropy": 4.785168552398682,
      "epoch": 3.536895674300254,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038650679056341096,
      "loss": 4.5122,
      "mean_token_accuracy": 0.24986608922481537,
      "num_tokens": 9972721.0,
      "step": 12510
    },
    {
      "entropy": 4.886654281616211,
      "epoch": 3.53830930166808,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038641743837364355,
      "loss": 4.5425,
      "mean_token_accuracy": 0.24942718744277953,
      "num_tokens": 9976758.0,
      "step": 12515
    },
    {
      "entropy": 4.808265972137451,
      "epoch": 3.5397229290359062,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003863280628983205,
      "loss": 4.4588,
      "mean_token_accuracy": 0.27493842244148253,
      "num_tokens": 9980506.0,
      "step": 12520
    },
    {
      "entropy": 4.751906633377075,
      "epoch": 3.541136556403732,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003862386641561208,
      "loss": 4.4489,
      "mean_token_accuracy": 0.27061349153518677,
      "num_tokens": 9984338.0,
      "step": 12525
    },
    {
      "entropy": 4.806198787689209,
      "epoch": 3.542550183771558,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038614924216572803,
      "loss": 4.487,
      "mean_token_accuracy": 0.2688231378793716,
      "num_tokens": 9988196.0,
      "step": 12530
    },
    {
      "entropy": 4.765150642395019,
      "epoch": 3.543963811139384,
      "grad_norm": 1.453125,
      "learning_rate": 0.00038605979694583117,
      "loss": 4.3892,
      "mean_token_accuracy": 0.27998588532209395,
      "num_tokens": 9992045.0,
      "step": 12535
    },
    {
      "entropy": 4.831715726852417,
      "epoch": 3.5453774385072094,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003859703285151234,
      "loss": 4.533,
      "mean_token_accuracy": 0.2543764129281044,
      "num_tokens": 9996026.0,
      "step": 12540
    },
    {
      "entropy": 4.80385913848877,
      "epoch": 3.5467910658750355,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003858808368923032,
      "loss": 4.5179,
      "mean_token_accuracy": 0.2589538708329201,
      "num_tokens": 9999778.0,
      "step": 12545
    },
    {
      "entropy": 4.779917526245117,
      "epoch": 3.548204693242861,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003857913220960737,
      "loss": 4.4281,
      "mean_token_accuracy": 0.27421519011259077,
      "num_tokens": 10003978.0,
      "step": 12550
    },
    {
      "entropy": 4.849533176422119,
      "epoch": 3.549618320610687,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00038570178414514314,
      "loss": 4.4673,
      "mean_token_accuracy": 0.2677691727876663,
      "num_tokens": 10007870.0,
      "step": 12555
    },
    {
      "entropy": 4.848208045959472,
      "epoch": 3.5510319479785126,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038561222305822415,
      "loss": 4.5803,
      "mean_token_accuracy": 0.24652958959341048,
      "num_tokens": 10011865.0,
      "step": 12560
    },
    {
      "entropy": 4.715923595428467,
      "epoch": 3.5524455753463386,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003855226388540345,
      "loss": 4.3306,
      "mean_token_accuracy": 0.2734778195619583,
      "num_tokens": 10015840.0,
      "step": 12565
    },
    {
      "entropy": 4.787542676925659,
      "epoch": 3.5538592027141647,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003854330315512969,
      "loss": 4.4991,
      "mean_token_accuracy": 0.26717508137226104,
      "num_tokens": 10020110.0,
      "step": 12570
    },
    {
      "entropy": 4.825817584991455,
      "epoch": 3.5552728300819902,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003853434011687386,
      "loss": 4.4874,
      "mean_token_accuracy": 0.2645777851343155,
      "num_tokens": 10024246.0,
      "step": 12575
    },
    {
      "entropy": 4.889746379852295,
      "epoch": 3.5566864574498163,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003852537477250917,
      "loss": 4.5168,
      "mean_token_accuracy": 0.26420356184244154,
      "num_tokens": 10028499.0,
      "step": 12580
    },
    {
      "entropy": 4.74712176322937,
      "epoch": 3.5581000848176423,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00038516407123909336,
      "loss": 4.3695,
      "mean_token_accuracy": 0.27932612895965575,
      "num_tokens": 10032423.0,
      "step": 12585
    },
    {
      "entropy": 4.754502534866333,
      "epoch": 3.559513712185468,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00038507437172948536,
      "loss": 4.5059,
      "mean_token_accuracy": 0.26335087418556213,
      "num_tokens": 10036201.0,
      "step": 12590
    },
    {
      "entropy": 4.878707075119019,
      "epoch": 3.560927339553294,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003849846492150144,
      "loss": 4.6218,
      "mean_token_accuracy": 0.2515658140182495,
      "num_tokens": 10040123.0,
      "step": 12595
    },
    {
      "entropy": 4.8496785163879395,
      "epoch": 3.5623409669211195,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003848949037144318,
      "loss": 4.5525,
      "mean_token_accuracy": 0.2617558628320694,
      "num_tokens": 10044074.0,
      "step": 12600
    },
    {
      "entropy": 4.906689786911011,
      "epoch": 3.5637545942889455,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038480513524649396,
      "loss": 4.562,
      "mean_token_accuracy": 0.25195698589086535,
      "num_tokens": 10048114.0,
      "step": 12605
    },
    {
      "entropy": 4.946790456771851,
      "epoch": 3.565168221656771,
      "grad_norm": 1.5,
      "learning_rate": 0.00038471534382996184,
      "loss": 4.6244,
      "mean_token_accuracy": 0.24909215718507766,
      "num_tokens": 10052145.0,
      "step": 12610
    },
    {
      "entropy": 4.79185266494751,
      "epoch": 3.566581849024597,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003846255294836013,
      "loss": 4.403,
      "mean_token_accuracy": 0.2564375713467598,
      "num_tokens": 10056033.0,
      "step": 12615
    },
    {
      "entropy": 4.782229423522949,
      "epoch": 3.567995476392423,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00038453569222618304,
      "loss": 4.3814,
      "mean_token_accuracy": 0.26911907345056535,
      "num_tokens": 10059757.0,
      "step": 12620
    },
    {
      "entropy": 4.815708446502685,
      "epoch": 3.5694091037602487,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003844458320764824,
      "loss": 4.4162,
      "mean_token_accuracy": 0.2756891369819641,
      "num_tokens": 10063995.0,
      "step": 12625
    },
    {
      "entropy": 4.828704452514648,
      "epoch": 3.5708227311280747,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038435594905327966,
      "loss": 4.4645,
      "mean_token_accuracy": 0.2560040667653084,
      "num_tokens": 10068014.0,
      "step": 12630
    },
    {
      "entropy": 4.9667884349823,
      "epoch": 3.5722363584959007,
      "grad_norm": 1.4375,
      "learning_rate": 0.00038426604317535985,
      "loss": 4.5698,
      "mean_token_accuracy": 0.25114998668432237,
      "num_tokens": 10072186.0,
      "step": 12635
    },
    {
      "entropy": 4.906714677810669,
      "epoch": 3.5736499858637263,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003841761144615126,
      "loss": 4.5437,
      "mean_token_accuracy": 0.26084943413734435,
      "num_tokens": 10076152.0,
      "step": 12640
    },
    {
      "entropy": 4.851852941513061,
      "epoch": 3.5750636132315523,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003840861629305327,
      "loss": 4.482,
      "mean_token_accuracy": 0.2658794760704041,
      "num_tokens": 10079844.0,
      "step": 12645
    },
    {
      "entropy": 4.7933738231658936,
      "epoch": 3.576477240599378,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00038399618860121917,
      "loss": 4.4445,
      "mean_token_accuracy": 0.27572684437036515,
      "num_tokens": 10083674.0,
      "step": 12650
    },
    {
      "entropy": 4.769465637207031,
      "epoch": 3.577890867967204,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003839061914923763,
      "loss": 4.5248,
      "mean_token_accuracy": 0.2513433247804642,
      "num_tokens": 10087641.0,
      "step": 12655
    },
    {
      "entropy": 4.877322578430176,
      "epoch": 3.5793044953350295,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003838161716228129,
      "loss": 4.4749,
      "mean_token_accuracy": 0.25849741101264956,
      "num_tokens": 10091507.0,
      "step": 12660
    },
    {
      "entropy": 4.9037196159362795,
      "epoch": 3.5807181227028555,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003837261290113425,
      "loss": 4.4883,
      "mean_token_accuracy": 0.2511134460568428,
      "num_tokens": 10095269.0,
      "step": 12665
    },
    {
      "entropy": 4.786405467987061,
      "epoch": 3.5821317500706815,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038363606367678343,
      "loss": 4.4105,
      "mean_token_accuracy": 0.2741294473409653,
      "num_tokens": 10099098.0,
      "step": 12670
    },
    {
      "entropy": 4.882800817489624,
      "epoch": 3.583545377438507,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00038354597563795893,
      "loss": 4.5605,
      "mean_token_accuracy": 0.24771392941474915,
      "num_tokens": 10103040.0,
      "step": 12675
    },
    {
      "entropy": 4.821696043014526,
      "epoch": 3.584959004806333,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003834558649136967,
      "loss": 4.438,
      "mean_token_accuracy": 0.2730036735534668,
      "num_tokens": 10106862.0,
      "step": 12680
    },
    {
      "entropy": 4.781372547149658,
      "epoch": 3.586372632174159,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003833657315228294,
      "loss": 4.303,
      "mean_token_accuracy": 0.27247946709394455,
      "num_tokens": 10110980.0,
      "step": 12685
    },
    {
      "entropy": 4.808914756774902,
      "epoch": 3.5877862595419847,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003832755754841943,
      "loss": 4.5875,
      "mean_token_accuracy": 0.25917208641767503,
      "num_tokens": 10114913.0,
      "step": 12690
    },
    {
      "entropy": 4.859126472473145,
      "epoch": 3.5891998869098107,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003831853968166335,
      "loss": 4.5442,
      "mean_token_accuracy": 0.25623818635940554,
      "num_tokens": 10119046.0,
      "step": 12695
    },
    {
      "entropy": 4.844850206375122,
      "epoch": 3.5906135142776363,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00038309519553899376,
      "loss": 4.5748,
      "mean_token_accuracy": 0.25189149081707,
      "num_tokens": 10123055.0,
      "step": 12700
    },
    {
      "entropy": 4.88198299407959,
      "epoch": 3.5920271416454623,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038300497167012657,
      "loss": 4.5429,
      "mean_token_accuracy": 0.24896415323019028,
      "num_tokens": 10127006.0,
      "step": 12705
    },
    {
      "entropy": 4.807261943817139,
      "epoch": 3.593440769013288,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00038291472522888816,
      "loss": 4.5246,
      "mean_token_accuracy": 0.2590080603957176,
      "num_tokens": 10130601.0,
      "step": 12710
    },
    {
      "entropy": 4.767070722579956,
      "epoch": 3.594854396381114,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038282445623413953,
      "loss": 4.5332,
      "mean_token_accuracy": 0.2640402168035507,
      "num_tokens": 10134657.0,
      "step": 12715
    },
    {
      "entropy": 4.799829816818237,
      "epoch": 3.59626802374894,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003827341647047463,
      "loss": 4.4417,
      "mean_token_accuracy": 0.270112743973732,
      "num_tokens": 10138147.0,
      "step": 12720
    },
    {
      "entropy": 4.911062002182007,
      "epoch": 3.5976816511167655,
      "grad_norm": 1.5625,
      "learning_rate": 0.00038264385065957897,
      "loss": 4.6567,
      "mean_token_accuracy": 0.25625698566436766,
      "num_tokens": 10142101.0,
      "step": 12725
    },
    {
      "entropy": 4.804557514190674,
      "epoch": 3.5990952784845915,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00038255351411751234,
      "loss": 4.3729,
      "mean_token_accuracy": 0.27259483337402346,
      "num_tokens": 10146204.0,
      "step": 12730
    },
    {
      "entropy": 4.812171220779419,
      "epoch": 3.6005089058524176,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003824631550974264,
      "loss": 4.4758,
      "mean_token_accuracy": 0.25866478979587554,
      "num_tokens": 10149996.0,
      "step": 12735
    },
    {
      "entropy": 4.869154977798462,
      "epoch": 3.601922533220243,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003823727736182056,
      "loss": 4.5696,
      "mean_token_accuracy": 0.26058095693588257,
      "num_tokens": 10154021.0,
      "step": 12740
    },
    {
      "entropy": 4.819116687774658,
      "epoch": 3.6033361605880687,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003822823696987391,
      "loss": 4.4562,
      "mean_token_accuracy": 0.257068307697773,
      "num_tokens": 10158070.0,
      "step": 12745
    },
    {
      "entropy": 4.812728214263916,
      "epoch": 3.6047497879558947,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003821919433579207,
      "loss": 4.4508,
      "mean_token_accuracy": 0.2685788542032242,
      "num_tokens": 10162122.0,
      "step": 12750
    },
    {
      "entropy": 4.8818957805633545,
      "epoch": 3.6061634153237208,
      "grad_norm": 1.46875,
      "learning_rate": 0.000382101494614649,
      "loss": 4.523,
      "mean_token_accuracy": 0.2628467470407486,
      "num_tokens": 10165892.0,
      "step": 12755
    },
    {
      "entropy": 4.718151378631592,
      "epoch": 3.6075770426915463,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003820110234878272,
      "loss": 4.3163,
      "mean_token_accuracy": 0.29305427670478823,
      "num_tokens": 10169693.0,
      "step": 12760
    },
    {
      "entropy": 4.696264266967773,
      "epoch": 3.6089906700593724,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00038192052999636337,
      "loss": 4.3762,
      "mean_token_accuracy": 0.27136590927839277,
      "num_tokens": 10173612.0,
      "step": 12765
    },
    {
      "entropy": 4.753298711776734,
      "epoch": 3.6104042974271984,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003818300141591698,
      "loss": 4.3588,
      "mean_token_accuracy": 0.26879282146692274,
      "num_tokens": 10177824.0,
      "step": 12770
    },
    {
      "entropy": 4.811779546737671,
      "epoch": 3.611817924795024,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003817394759951639,
      "loss": 4.4899,
      "mean_token_accuracy": 0.26481370627880096,
      "num_tokens": 10181712.0,
      "step": 12775
    },
    {
      "entropy": 4.823728084564209,
      "epoch": 3.61323155216285,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038164891552326766,
      "loss": 4.4622,
      "mean_token_accuracy": 0.266204397380352,
      "num_tokens": 10185942.0,
      "step": 12780
    },
    {
      "entropy": 4.867056512832642,
      "epoch": 3.614645179530676,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003815583327624075,
      "loss": 4.5345,
      "mean_token_accuracy": 0.2599469184875488,
      "num_tokens": 10190148.0,
      "step": 12785
    },
    {
      "entropy": 4.746792936325074,
      "epoch": 3.6160588068985016,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003814677277315147,
      "loss": 4.4197,
      "mean_token_accuracy": 0.2667789712548256,
      "num_tokens": 10194283.0,
      "step": 12790
    },
    {
      "entropy": 4.9190582752227785,
      "epoch": 3.617472434266327,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003813771004495253,
      "loss": 4.6008,
      "mean_token_accuracy": 0.24690867811441422,
      "num_tokens": 10198202.0,
      "step": 12795
    },
    {
      "entropy": 4.816651821136475,
      "epoch": 3.618886061634153,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003812864509353797,
      "loss": 4.4351,
      "mean_token_accuracy": 0.2721071720123291,
      "num_tokens": 10201995.0,
      "step": 12800
    },
    {
      "entropy": 4.797214794158935,
      "epoch": 3.620299689001979,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00038119577920802295,
      "loss": 4.4804,
      "mean_token_accuracy": 0.2694116622209549,
      "num_tokens": 10205938.0,
      "step": 12805
    },
    {
      "entropy": 4.748946809768677,
      "epoch": 3.6217133163698048,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003811050852864051,
      "loss": 4.4368,
      "mean_token_accuracy": 0.27859034687280654,
      "num_tokens": 10210226.0,
      "step": 12810
    },
    {
      "entropy": 4.852495765686035,
      "epoch": 3.623126943737631,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00038101436918948074,
      "loss": 4.4957,
      "mean_token_accuracy": 0.25956837832927704,
      "num_tokens": 10213986.0,
      "step": 12815
    },
    {
      "entropy": 4.848103046417236,
      "epoch": 3.624540571105457,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003809236309362087,
      "loss": 4.4425,
      "mean_token_accuracy": 0.2782969817519188,
      "num_tokens": 10218053.0,
      "step": 12820
    },
    {
      "entropy": 4.864340448379517,
      "epoch": 3.6259541984732824,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038083287054555264,
      "loss": 4.5113,
      "mean_token_accuracy": 0.261515685915947,
      "num_tokens": 10221997.0,
      "step": 12825
    },
    {
      "entropy": 4.876324415206909,
      "epoch": 3.6273678258411084,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003807420880364813,
      "loss": 4.5481,
      "mean_token_accuracy": 0.252836012840271,
      "num_tokens": 10226270.0,
      "step": 12830
    },
    {
      "entropy": 4.884369993209839,
      "epoch": 3.628781453208934,
      "grad_norm": 1.53125,
      "learning_rate": 0.00038065128342796726,
      "loss": 4.5411,
      "mean_token_accuracy": 0.25572770088911057,
      "num_tokens": 10230243.0,
      "step": 12835
    },
    {
      "entropy": 4.810578107833862,
      "epoch": 3.63019508057676,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003805604567389885,
      "loss": 4.4582,
      "mean_token_accuracy": 0.26828575432300567,
      "num_tokens": 10234388.0,
      "step": 12840
    },
    {
      "entropy": 4.835265684127807,
      "epoch": 3.6316087079445856,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038046960798852696,
      "loss": 4.455,
      "mean_token_accuracy": 0.2632140710949898,
      "num_tokens": 10238411.0,
      "step": 12845
    },
    {
      "entropy": 4.803600835800171,
      "epoch": 3.6330223353124116,
      "grad_norm": 1.484375,
      "learning_rate": 0.00038037873719556957,
      "loss": 4.4611,
      "mean_token_accuracy": 0.2635692223906517,
      "num_tokens": 10242272.0,
      "step": 12850
    },
    {
      "entropy": 4.818977975845337,
      "epoch": 3.6344359626802376,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038028784437910774,
      "loss": 4.4656,
      "mean_token_accuracy": 0.2650697395205498,
      "num_tokens": 10246165.0,
      "step": 12855
    },
    {
      "entropy": 4.762552070617676,
      "epoch": 3.635849590048063,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003801969295581375,
      "loss": 4.483,
      "mean_token_accuracy": 0.26323607563972473,
      "num_tokens": 10249909.0,
      "step": 12860
    },
    {
      "entropy": 4.826672744750977,
      "epoch": 3.637263217415889,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003801059927516595,
      "loss": 4.4779,
      "mean_token_accuracy": 0.25735963732004163,
      "num_tokens": 10254296.0,
      "step": 12865
    },
    {
      "entropy": 4.804860019683838,
      "epoch": 3.6386768447837152,
      "grad_norm": 1.53125,
      "learning_rate": 0.00038001503397867896,
      "loss": 4.4925,
      "mean_token_accuracy": 0.2701292961835861,
      "num_tokens": 10258394.0,
      "step": 12870
    },
    {
      "entropy": 4.791331100463867,
      "epoch": 3.640090472151541,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003799240532582058,
      "loss": 4.3915,
      "mean_token_accuracy": 0.2766176640987396,
      "num_tokens": 10262460.0,
      "step": 12875
    },
    {
      "entropy": 4.870790195465088,
      "epoch": 3.641504099519367,
      "grad_norm": 1.40625,
      "learning_rate": 0.00037983305060925425,
      "loss": 4.5896,
      "mean_token_accuracy": 0.25266426652669904,
      "num_tokens": 10266707.0,
      "step": 12880
    },
    {
      "entropy": 4.886058330535889,
      "epoch": 3.6429177268871924,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003797420260508434,
      "loss": 4.489,
      "mean_token_accuracy": 0.2517557621002197,
      "num_tokens": 10271110.0,
      "step": 12885
    },
    {
      "entropy": 4.784596300125122,
      "epoch": 3.6443313542550184,
      "grad_norm": 1.5625,
      "learning_rate": 0.00037965097960199684,
      "loss": 4.4224,
      "mean_token_accuracy": 0.26481255441904067,
      "num_tokens": 10275255.0,
      "step": 12890
    },
    {
      "entropy": 4.786478805541992,
      "epoch": 3.645744981622844,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003795599112817427,
      "loss": 4.5284,
      "mean_token_accuracy": 0.26404415965080263,
      "num_tokens": 10279372.0,
      "step": 12895
    },
    {
      "entropy": 4.766873788833618,
      "epoch": 3.64715860899067,
      "grad_norm": 1.375,
      "learning_rate": 0.00037946882110911364,
      "loss": 4.4494,
      "mean_token_accuracy": 0.25855106860399246,
      "num_tokens": 10283277.0,
      "step": 12900
    },
    {
      "entropy": 4.91284294128418,
      "epoch": 3.648572236358496,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000379377709103147,
      "loss": 4.6085,
      "mean_token_accuracy": 0.24989253729581834,
      "num_tokens": 10287323.0,
      "step": 12905
    },
    {
      "entropy": 4.775114059448242,
      "epoch": 3.6499858637263216,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037928657528288464,
      "loss": 4.3722,
      "mean_token_accuracy": 0.27544273138046266,
      "num_tokens": 10291191.0,
      "step": 12910
    },
    {
      "entropy": 4.754809951782226,
      "epoch": 3.6513994910941476,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003791954196673729,
      "loss": 4.4616,
      "mean_token_accuracy": 0.2691652774810791,
      "num_tokens": 10295111.0,
      "step": 12915
    },
    {
      "entropy": 4.866592216491699,
      "epoch": 3.6528131184619737,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037910424227566276,
      "loss": 4.5207,
      "mean_token_accuracy": 0.2601174205541611,
      "num_tokens": 10299418.0,
      "step": 12920
    },
    {
      "entropy": 4.815042209625244,
      "epoch": 3.6542267458297992,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003790130431268097,
      "loss": 4.4651,
      "mean_token_accuracy": 0.2678791180253029,
      "num_tokens": 10303556.0,
      "step": 12925
    },
    {
      "entropy": 4.8009082794189455,
      "epoch": 3.6556403731976252,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00037892182223987395,
      "loss": 4.5319,
      "mean_token_accuracy": 0.2714900359511375,
      "num_tokens": 10307373.0,
      "step": 12930
    },
    {
      "entropy": 4.767893266677857,
      "epoch": 3.657054000565451,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037883057963391994,
      "loss": 4.4147,
      "mean_token_accuracy": 0.26868176758289336,
      "num_tokens": 10311604.0,
      "step": 12935
    },
    {
      "entropy": 4.803636503219605,
      "epoch": 3.658467627933277,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00037873931532801685,
      "loss": 4.4953,
      "mean_token_accuracy": 0.26362473964691163,
      "num_tokens": 10315628.0,
      "step": 12940
    },
    {
      "entropy": 4.822842168807983,
      "epoch": 3.6598812553011024,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003786480293412383,
      "loss": 4.5685,
      "mean_token_accuracy": 0.24897135198116302,
      "num_tokens": 10319970.0,
      "step": 12945
    },
    {
      "entropy": 4.749465990066528,
      "epoch": 3.6612948826689284,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003785567216926626,
      "loss": 4.3694,
      "mean_token_accuracy": 0.27143624126911164,
      "num_tokens": 10323696.0,
      "step": 12950
    },
    {
      "entropy": 4.893176174163818,
      "epoch": 3.6627085100367545,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037846539240137244,
      "loss": 4.511,
      "mean_token_accuracy": 0.25924976766109464,
      "num_tokens": 10327502.0,
      "step": 12955
    },
    {
      "entropy": 4.801934337615966,
      "epoch": 3.66412213740458,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003783740414864551,
      "loss": 4.4648,
      "mean_token_accuracy": 0.27220958322286604,
      "num_tokens": 10331520.0,
      "step": 12960
    },
    {
      "entropy": 4.792726039886475,
      "epoch": 3.665535764772406,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037828266896700237,
      "loss": 4.4517,
      "mean_token_accuracy": 0.2702966034412384,
      "num_tokens": 10335504.0,
      "step": 12965
    },
    {
      "entropy": 4.714724540710449,
      "epoch": 3.666949392140232,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037819127486211045,
      "loss": 4.3858,
      "mean_token_accuracy": 0.2695778712630272,
      "num_tokens": 10339955.0,
      "step": 12970
    },
    {
      "entropy": 4.821270704269409,
      "epoch": 3.6683630195080577,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037809985919088013,
      "loss": 4.4888,
      "mean_token_accuracy": 0.2662080407142639,
      "num_tokens": 10344047.0,
      "step": 12975
    },
    {
      "entropy": 4.868264484405517,
      "epoch": 3.6697766468758837,
      "grad_norm": 1.375,
      "learning_rate": 0.0003780084219724169,
      "loss": 4.5593,
      "mean_token_accuracy": 0.2620030641555786,
      "num_tokens": 10348020.0,
      "step": 12980
    },
    {
      "entropy": 4.8076740264892575,
      "epoch": 3.6711902742437093,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037791696322583043,
      "loss": 4.3701,
      "mean_token_accuracy": 0.27459056824445727,
      "num_tokens": 10351910.0,
      "step": 12985
    },
    {
      "entropy": 4.772724723815918,
      "epoch": 3.6726039016115353,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003778254829702351,
      "loss": 4.4626,
      "mean_token_accuracy": 0.27614119797945025,
      "num_tokens": 10355937.0,
      "step": 12990
    },
    {
      "entropy": 4.7489715099334715,
      "epoch": 3.674017528979361,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003777339812247496,
      "loss": 4.4789,
      "mean_token_accuracy": 0.25907639861106874,
      "num_tokens": 10359744.0,
      "step": 12995
    },
    {
      "entropy": 4.775900983810425,
      "epoch": 3.675431156347187,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00037764245800849737,
      "loss": 4.4408,
      "mean_token_accuracy": 0.2692386120557785,
      "num_tokens": 10363628.0,
      "step": 13000
    },
    {
      "epoch": 3.675431156347187,
      "eval_entropy": 4.607662939839461,
      "eval_loss": 4.982306480407715,
      "eval_mean_token_accuracy": 0.2494852201238243,
      "eval_num_tokens": 10363628.0,
      "eval_runtime": 4.6621,
      "eval_samples_per_second": 2420.373,
      "eval_steps_per_second": 302.654,
      "step": 13000
    },
    {
      "entropy": 4.916377449035645,
      "epoch": 3.676844783715013,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003775509133406061,
      "loss": 4.5545,
      "mean_token_accuracy": 0.2488421082496643,
      "num_tokens": 10367258.0,
      "step": 13005
    },
    {
      "entropy": 4.909597110748291,
      "epoch": 3.6782584110828385,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003774593472402081,
      "loss": 4.5558,
      "mean_token_accuracy": 0.2632711723446846,
      "num_tokens": 10371383.0,
      "step": 13010
    },
    {
      "entropy": 4.846770429611206,
      "epoch": 3.6796720384506645,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003773677597264402,
      "loss": 4.5049,
      "mean_token_accuracy": 0.25370497554540633,
      "num_tokens": 10375198.0,
      "step": 13015
    },
    {
      "entropy": 4.863863945007324,
      "epoch": 3.6810856658184905,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003772761508184435,
      "loss": 4.547,
      "mean_token_accuracy": 0.2479250431060791,
      "num_tokens": 10379265.0,
      "step": 13020
    },
    {
      "entropy": 4.790130186080932,
      "epoch": 3.682499293186316,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037718452053536375,
      "loss": 4.4409,
      "mean_token_accuracy": 0.26262121498584745,
      "num_tokens": 10383449.0,
      "step": 13025
    },
    {
      "entropy": 4.730867338180542,
      "epoch": 3.6839129205541417,
      "grad_norm": 1.421875,
      "learning_rate": 0.000377092868896351,
      "loss": 4.3732,
      "mean_token_accuracy": 0.2794343039393425,
      "num_tokens": 10387282.0,
      "step": 13030
    },
    {
      "entropy": 4.772504997253418,
      "epoch": 3.6853265479219677,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037700119592056,
      "loss": 4.4574,
      "mean_token_accuracy": 0.2632947191596031,
      "num_tokens": 10391351.0,
      "step": 13035
    },
    {
      "entropy": 4.7649681091308596,
      "epoch": 3.6867401752897937,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00037690950162714995,
      "loss": 4.3626,
      "mean_token_accuracy": 0.2760659784078598,
      "num_tokens": 10395225.0,
      "step": 13040
    },
    {
      "entropy": 4.8231809616088865,
      "epoch": 3.6881538026576193,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003768177860352841,
      "loss": 4.5133,
      "mean_token_accuracy": 0.263174931704998,
      "num_tokens": 10399340.0,
      "step": 13045
    },
    {
      "entropy": 4.849117422103882,
      "epoch": 3.6895674300254453,
      "grad_norm": 1.375,
      "learning_rate": 0.00037672604916413064,
      "loss": 4.5195,
      "mean_token_accuracy": 0.26305426806211474,
      "num_tokens": 10403608.0,
      "step": 13050
    },
    {
      "entropy": 4.7903910160064695,
      "epoch": 3.6909810573932713,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037663429103286197,
      "loss": 4.4311,
      "mean_token_accuracy": 0.2608529657125473,
      "num_tokens": 10407812.0,
      "step": 13055
    },
    {
      "entropy": 4.819154882431031,
      "epoch": 3.692394684761097,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000376542511660655,
      "loss": 4.5304,
      "mean_token_accuracy": 0.2611896276473999,
      "num_tokens": 10411836.0,
      "step": 13060
    },
    {
      "entropy": 4.775899744033813,
      "epoch": 3.693808312128923,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037645071106669095,
      "loss": 4.5164,
      "mean_token_accuracy": 0.2528954714536667,
      "num_tokens": 10416230.0,
      "step": 13065
    },
    {
      "entropy": 4.892530250549316,
      "epoch": 3.695221939496749,
      "grad_norm": 1.578125,
      "learning_rate": 0.00037635888927015565,
      "loss": 4.5178,
      "mean_token_accuracy": 0.25797474235296247,
      "num_tokens": 10420396.0,
      "step": 13070
    },
    {
      "entropy": 4.886746025085449,
      "epoch": 3.6966355668645745,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037626704629023923,
      "loss": 4.5003,
      "mean_token_accuracy": 0.25889341682195666,
      "num_tokens": 10424371.0,
      "step": 13075
    },
    {
      "entropy": 4.761044836044311,
      "epoch": 3.6980491942324,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037617518214613647,
      "loss": 4.3232,
      "mean_token_accuracy": 0.28277445733547213,
      "num_tokens": 10428144.0,
      "step": 13080
    },
    {
      "entropy": 4.760215473175049,
      "epoch": 3.699462821600226,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003760832968570462,
      "loss": 4.4236,
      "mean_token_accuracy": 0.2530431717634201,
      "num_tokens": 10432251.0,
      "step": 13085
    },
    {
      "entropy": 4.806635618209839,
      "epoch": 3.700876448968052,
      "grad_norm": 1.359375,
      "learning_rate": 0.00037599139044217203,
      "loss": 4.4461,
      "mean_token_accuracy": 0.2678716629743576,
      "num_tokens": 10436244.0,
      "step": 13090
    },
    {
      "entropy": 4.860167503356934,
      "epoch": 3.7022900763358777,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003758994629207217,
      "loss": 4.5749,
      "mean_token_accuracy": 0.2553035616874695,
      "num_tokens": 10440465.0,
      "step": 13095
    },
    {
      "entropy": 4.979308938980102,
      "epoch": 3.7037037037037037,
      "grad_norm": 1.65625,
      "learning_rate": 0.00037580751431190765,
      "loss": 4.5695,
      "mean_token_accuracy": 0.26320854127407073,
      "num_tokens": 10444374.0,
      "step": 13100
    },
    {
      "entropy": 4.836717367172241,
      "epoch": 3.7051173310715297,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003757155446349464,
      "loss": 4.5057,
      "mean_token_accuracy": 0.2582481414079666,
      "num_tokens": 10448498.0,
      "step": 13105
    },
    {
      "entropy": 4.772403812408447,
      "epoch": 3.7065309584393553,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003756235539090591,
      "loss": 4.4743,
      "mean_token_accuracy": 0.26338121443986895,
      "num_tokens": 10452328.0,
      "step": 13110
    },
    {
      "entropy": 4.706276559829712,
      "epoch": 3.7079445858071813,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00037553154215347133,
      "loss": 4.4517,
      "mean_token_accuracy": 0.26304790526628496,
      "num_tokens": 10456037.0,
      "step": 13115
    },
    {
      "entropy": 4.749392747879028,
      "epoch": 3.709358213175007,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037543950938741293,
      "loss": 4.4216,
      "mean_token_accuracy": 0.27195377349853517,
      "num_tokens": 10459951.0,
      "step": 13120
    },
    {
      "entropy": 4.873599720001221,
      "epoch": 3.710771840542833,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037534745563011816,
      "loss": 4.6078,
      "mean_token_accuracy": 0.24816474318504333,
      "num_tokens": 10464023.0,
      "step": 13125
    },
    {
      "entropy": 4.801414346694946,
      "epoch": 3.7121854679106585,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037525538090082565,
      "loss": 4.4286,
      "mean_token_accuracy": 0.27520918250083926,
      "num_tokens": 10468161.0,
      "step": 13130
    },
    {
      "entropy": 4.868777704238892,
      "epoch": 3.7135990952784845,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003751632852187785,
      "loss": 4.4717,
      "mean_token_accuracy": 0.2589162826538086,
      "num_tokens": 10472141.0,
      "step": 13135
    },
    {
      "entropy": 4.811661720275879,
      "epoch": 3.7150127226463106,
      "grad_norm": 1.375,
      "learning_rate": 0.0003750711686032242,
      "loss": 4.4833,
      "mean_token_accuracy": 0.269028240442276,
      "num_tokens": 10476166.0,
      "step": 13140
    },
    {
      "entropy": 4.847944641113282,
      "epoch": 3.716426350014136,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037497903107341446,
      "loss": 4.5881,
      "mean_token_accuracy": 0.25887214839458467,
      "num_tokens": 10480320.0,
      "step": 13145
    },
    {
      "entropy": 4.789576435089112,
      "epoch": 3.717839977381962,
      "grad_norm": 1.421875,
      "learning_rate": 0.00037488687264860545,
      "loss": 4.4781,
      "mean_token_accuracy": 0.25902772396802903,
      "num_tokens": 10484398.0,
      "step": 13150
    },
    {
      "entropy": 4.776571989059448,
      "epoch": 3.719253604749788,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037479469334805775,
      "loss": 4.4579,
      "mean_token_accuracy": 0.28022659868001937,
      "num_tokens": 10488240.0,
      "step": 13155
    },
    {
      "entropy": 4.818247842788696,
      "epoch": 3.7206672321176137,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003747024931910362,
      "loss": 4.5483,
      "mean_token_accuracy": 0.26612986624240875,
      "num_tokens": 10492456.0,
      "step": 13160
    },
    {
      "entropy": 4.791763496398926,
      "epoch": 3.7220808594854398,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003746102721968102,
      "loss": 4.4285,
      "mean_token_accuracy": 0.2710942804813385,
      "num_tokens": 10496442.0,
      "step": 13165
    },
    {
      "entropy": 4.815055179595947,
      "epoch": 3.7234944868532653,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003745180303846532,
      "loss": 4.4297,
      "mean_token_accuracy": 0.25891700983047483,
      "num_tokens": 10500399.0,
      "step": 13170
    },
    {
      "entropy": 4.753982830047607,
      "epoch": 3.7249081142210914,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003744257677738433,
      "loss": 4.4585,
      "mean_token_accuracy": 0.26740929037332534,
      "num_tokens": 10504267.0,
      "step": 13175
    },
    {
      "entropy": 4.83813738822937,
      "epoch": 3.726321741588917,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037433348438366276,
      "loss": 4.4221,
      "mean_token_accuracy": 0.2664206504821777,
      "num_tokens": 10508100.0,
      "step": 13180
    },
    {
      "entropy": 4.78205304145813,
      "epoch": 3.727735368956743,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003742411802333983,
      "loss": 4.4302,
      "mean_token_accuracy": 0.26098109036684036,
      "num_tokens": 10512337.0,
      "step": 13185
    },
    {
      "entropy": 4.748507070541382,
      "epoch": 3.729148996324569,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003741488553423408,
      "loss": 4.4328,
      "mean_token_accuracy": 0.26959338337183,
      "num_tokens": 10516478.0,
      "step": 13190
    },
    {
      "entropy": 4.83568058013916,
      "epoch": 3.7305626236923946,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037405650972978567,
      "loss": 4.5923,
      "mean_token_accuracy": 0.2490632340312004,
      "num_tokens": 10520569.0,
      "step": 13195
    },
    {
      "entropy": 4.825828790664673,
      "epoch": 3.7319762510602206,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003739641434150324,
      "loss": 4.422,
      "mean_token_accuracy": 0.2647728964686394,
      "num_tokens": 10524778.0,
      "step": 13200
    },
    {
      "entropy": 4.75985894203186,
      "epoch": 3.7333898784280466,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037387175641738534,
      "loss": 4.4844,
      "mean_token_accuracy": 0.2656961664557457,
      "num_tokens": 10528949.0,
      "step": 13205
    },
    {
      "entropy": 4.776730155944824,
      "epoch": 3.734803505795872,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00037377934875615257,
      "loss": 4.4177,
      "mean_token_accuracy": 0.27775571644306185,
      "num_tokens": 10532950.0,
      "step": 13210
    },
    {
      "entropy": 4.849742841720581,
      "epoch": 3.736217133163698,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037368692045064666,
      "loss": 4.4722,
      "mean_token_accuracy": 0.2619649648666382,
      "num_tokens": 10537140.0,
      "step": 13215
    },
    {
      "entropy": 4.8097615242004395,
      "epoch": 3.7376307605315238,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00037359447152018465,
      "loss": 4.4266,
      "mean_token_accuracy": 0.2650702193379402,
      "num_tokens": 10541195.0,
      "step": 13220
    },
    {
      "entropy": 4.780143308639526,
      "epoch": 3.73904438789935,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037350200198408786,
      "loss": 4.4162,
      "mean_token_accuracy": 0.27627013623714447,
      "num_tokens": 10545289.0,
      "step": 13225
    },
    {
      "entropy": 4.770954227447509,
      "epoch": 3.7404580152671754,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003734095118616816,
      "loss": 4.4886,
      "mean_token_accuracy": 0.2526908114552498,
      "num_tokens": 10549324.0,
      "step": 13230
    },
    {
      "entropy": 4.7370130062103275,
      "epoch": 3.7418716426350014,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003733170011722961,
      "loss": 4.3907,
      "mean_token_accuracy": 0.27065504193305967,
      "num_tokens": 10553166.0,
      "step": 13235
    },
    {
      "entropy": 4.925796175003052,
      "epoch": 3.7432852700028274,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003732244699352653,
      "loss": 4.5773,
      "mean_token_accuracy": 0.25396625846624377,
      "num_tokens": 10556834.0,
      "step": 13240
    },
    {
      "entropy": 4.835301351547241,
      "epoch": 3.744698897370653,
      "grad_norm": 1.484375,
      "learning_rate": 0.00037313191816992764,
      "loss": 4.4832,
      "mean_token_accuracy": 0.2650833860039711,
      "num_tokens": 10561079.0,
      "step": 13245
    },
    {
      "entropy": 4.738225793838501,
      "epoch": 3.746112524738479,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037303934589562587,
      "loss": 4.3425,
      "mean_token_accuracy": 0.2780235350131989,
      "num_tokens": 10564936.0,
      "step": 13250
    },
    {
      "entropy": 4.7338048934936525,
      "epoch": 3.747526152106305,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003729467531317071,
      "loss": 4.4445,
      "mean_token_accuracy": 0.2598521426320076,
      "num_tokens": 10568990.0,
      "step": 13255
    },
    {
      "entropy": 4.730132246017456,
      "epoch": 3.7489397794741306,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037285413989752265,
      "loss": 4.418,
      "mean_token_accuracy": 0.2603334650397301,
      "num_tokens": 10573046.0,
      "step": 13260
    },
    {
      "entropy": 4.972252368927002,
      "epoch": 3.7503534068419566,
      "grad_norm": 1.453125,
      "learning_rate": 0.00037276150621242803,
      "loss": 4.5743,
      "mean_token_accuracy": 0.2413726791739464,
      "num_tokens": 10577131.0,
      "step": 13265
    },
    {
      "entropy": 4.883005571365357,
      "epoch": 3.751767034209782,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003726688520957832,
      "loss": 4.4687,
      "mean_token_accuracy": 0.26484550088644027,
      "num_tokens": 10581153.0,
      "step": 13270
    },
    {
      "entropy": 4.807442855834961,
      "epoch": 3.753180661577608,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037257617756695225,
      "loss": 4.4646,
      "mean_token_accuracy": 0.2522194504737854,
      "num_tokens": 10585080.0,
      "step": 13275
    },
    {
      "entropy": 4.77384672164917,
      "epoch": 3.754594288945434,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037248348264530345,
      "loss": 4.4393,
      "mean_token_accuracy": 0.2629065543413162,
      "num_tokens": 10589087.0,
      "step": 13280
    },
    {
      "entropy": 4.913646602630616,
      "epoch": 3.75600791631326,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003723907673502097,
      "loss": 4.5967,
      "mean_token_accuracy": 0.25010081231594083,
      "num_tokens": 10593077.0,
      "step": 13285
    },
    {
      "entropy": 4.783433294296264,
      "epoch": 3.757421543681086,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037229803170104784,
      "loss": 4.4773,
      "mean_token_accuracy": 0.26458549648523333,
      "num_tokens": 10596942.0,
      "step": 13290
    },
    {
      "entropy": 4.7905192375183105,
      "epoch": 3.7588351710489114,
      "grad_norm": 1.40625,
      "learning_rate": 0.000372205275717199,
      "loss": 4.4625,
      "mean_token_accuracy": 0.2576505944132805,
      "num_tokens": 10601020.0,
      "step": 13295
    },
    {
      "entropy": 4.833533048629761,
      "epoch": 3.7602487984167374,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037211249941804865,
      "loss": 4.5123,
      "mean_token_accuracy": 0.2567380264401436,
      "num_tokens": 10604959.0,
      "step": 13300
    },
    {
      "entropy": 4.811011552810669,
      "epoch": 3.7616624257845634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003720197028229864,
      "loss": 4.4171,
      "mean_token_accuracy": 0.2623270615935326,
      "num_tokens": 10609109.0,
      "step": 13305
    },
    {
      "entropy": 4.822523498535157,
      "epoch": 3.763076053152389,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003719268859514062,
      "loss": 4.5231,
      "mean_token_accuracy": 0.2655691891908646,
      "num_tokens": 10612979.0,
      "step": 13310
    },
    {
      "entropy": 4.8206805229187015,
      "epoch": 3.7644896805202146,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037183404882270627,
      "loss": 4.5403,
      "mean_token_accuracy": 0.2611152395606041,
      "num_tokens": 10617277.0,
      "step": 13315
    },
    {
      "entropy": 4.811039924621582,
      "epoch": 3.7659033078880406,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037174119145628895,
      "loss": 4.4506,
      "mean_token_accuracy": 0.2709058836102486,
      "num_tokens": 10621446.0,
      "step": 13320
    },
    {
      "entropy": 4.805000114440918,
      "epoch": 3.7673169352558666,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003716483138715608,
      "loss": 4.4487,
      "mean_token_accuracy": 0.2642622575163841,
      "num_tokens": 10625273.0,
      "step": 13325
    },
    {
      "entropy": 4.764337396621704,
      "epoch": 3.768730562623692,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003715554160879327,
      "loss": 4.4518,
      "mean_token_accuracy": 0.264188189804554,
      "num_tokens": 10629092.0,
      "step": 13330
    },
    {
      "entropy": 4.840927267074585,
      "epoch": 3.7701441899915182,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003714624981248198,
      "loss": 4.4848,
      "mean_token_accuracy": 0.2785185396671295,
      "num_tokens": 10633044.0,
      "step": 13335
    },
    {
      "entropy": 4.792979049682617,
      "epoch": 3.7715578173593443,
      "grad_norm": 1.453125,
      "learning_rate": 0.00037136956000164117,
      "loss": 4.4852,
      "mean_token_accuracy": 0.26328506171703336,
      "num_tokens": 10636993.0,
      "step": 13340
    },
    {
      "entropy": 4.770647239685059,
      "epoch": 3.77297144472717,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00037127660173782057,
      "loss": 4.3734,
      "mean_token_accuracy": 0.2813486084342003,
      "num_tokens": 10640927.0,
      "step": 13345
    },
    {
      "entropy": 4.794718933105469,
      "epoch": 3.774385072094996,
      "grad_norm": 1.5,
      "learning_rate": 0.00037118362335278543,
      "loss": 4.5363,
      "mean_token_accuracy": 0.261420501768589,
      "num_tokens": 10644651.0,
      "step": 13350
    },
    {
      "entropy": 4.7523353576660154,
      "epoch": 3.775798699462822,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003710906248659679,
      "loss": 4.3706,
      "mean_token_accuracy": 0.27413483113050463,
      "num_tokens": 10648731.0,
      "step": 13355
    },
    {
      "entropy": 4.8700157642364506,
      "epoch": 3.7772123268306474,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003709976062968039,
      "loss": 4.5588,
      "mean_token_accuracy": 0.259662900865078,
      "num_tokens": 10652585.0,
      "step": 13360
    },
    {
      "entropy": 4.831765079498291,
      "epoch": 3.778625954198473,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037090456766473393,
      "loss": 4.4627,
      "mean_token_accuracy": 0.2656571865081787,
      "num_tokens": 10656533.0,
      "step": 13365
    },
    {
      "entropy": 4.849581480026245,
      "epoch": 3.780039581566299,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037081150898920226,
      "loss": 4.5331,
      "mean_token_accuracy": 0.25857638418674467,
      "num_tokens": 10660551.0,
      "step": 13370
    },
    {
      "entropy": 4.825933742523193,
      "epoch": 3.781453208934125,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037071843028965766,
      "loss": 4.448,
      "mean_token_accuracy": 0.2606165617704391,
      "num_tokens": 10664489.0,
      "step": 13375
    },
    {
      "entropy": 4.81016263961792,
      "epoch": 3.7828668363019506,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037062533158555315,
      "loss": 4.3783,
      "mean_token_accuracy": 0.2702557995915413,
      "num_tokens": 10668391.0,
      "step": 13380
    },
    {
      "entropy": 4.787446451187134,
      "epoch": 3.7842804636697767,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003705322128963456,
      "loss": 4.5473,
      "mean_token_accuracy": 0.2618073493242264,
      "num_tokens": 10672440.0,
      "step": 13385
    },
    {
      "entropy": 4.8509564876556395,
      "epoch": 3.7856940910376027,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003704390742414964,
      "loss": 4.538,
      "mean_token_accuracy": 0.25591142773628234,
      "num_tokens": 10676366.0,
      "step": 13390
    },
    {
      "entropy": 4.768226051330567,
      "epoch": 3.7871077184054283,
      "grad_norm": 1.375,
      "learning_rate": 0.0003703459156404708,
      "loss": 4.4719,
      "mean_token_accuracy": 0.263679051399231,
      "num_tokens": 10680204.0,
      "step": 13395
    },
    {
      "entropy": 4.745037031173706,
      "epoch": 3.7885213457732543,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037025273711273837,
      "loss": 4.3832,
      "mean_token_accuracy": 0.27488472908735273,
      "num_tokens": 10684359.0,
      "step": 13400
    },
    {
      "entropy": 4.868150949478149,
      "epoch": 3.78993497314108,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037015953867777297,
      "loss": 4.6438,
      "mean_token_accuracy": 0.24635444730520248,
      "num_tokens": 10688551.0,
      "step": 13405
    },
    {
      "entropy": 4.79884295463562,
      "epoch": 3.791348600508906,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003700663203550524,
      "loss": 4.4047,
      "mean_token_accuracy": 0.27278355360031126,
      "num_tokens": 10692231.0,
      "step": 13410
    },
    {
      "entropy": 4.77059850692749,
      "epoch": 3.7927622278767315,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003699730821640587,
      "loss": 4.4164,
      "mean_token_accuracy": 0.27154659777879714,
      "num_tokens": 10696156.0,
      "step": 13415
    },
    {
      "entropy": 4.934561347961425,
      "epoch": 3.7941758552445575,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003698798241242783,
      "loss": 4.7371,
      "mean_token_accuracy": 0.2409916564822197,
      "num_tokens": 10700016.0,
      "step": 13420
    },
    {
      "entropy": 4.885385942459107,
      "epoch": 3.7955894826123835,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036978654625520127,
      "loss": 4.5842,
      "mean_token_accuracy": 0.25300799757242204,
      "num_tokens": 10704089.0,
      "step": 13425
    },
    {
      "entropy": 4.842309617996216,
      "epoch": 3.797003109980209,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036969324857632227,
      "loss": 4.4671,
      "mean_token_accuracy": 0.26495357006788256,
      "num_tokens": 10708008.0,
      "step": 13430
    },
    {
      "entropy": 4.806987524032593,
      "epoch": 3.798416737348035,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036959993110713976,
      "loss": 4.4336,
      "mean_token_accuracy": 0.26618589013814925,
      "num_tokens": 10711874.0,
      "step": 13435
    },
    {
      "entropy": 4.800780391693115,
      "epoch": 3.799830364715861,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036950659386715675,
      "loss": 4.4736,
      "mean_token_accuracy": 0.26006456911563874,
      "num_tokens": 10716297.0,
      "step": 13440
    },
    {
      "entropy": 4.833214378356933,
      "epoch": 3.8012439920836867,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036941323687587994,
      "loss": 4.4947,
      "mean_token_accuracy": 0.2547933325171471,
      "num_tokens": 10720605.0,
      "step": 13445
    },
    {
      "entropy": 4.83890061378479,
      "epoch": 3.8026576194515127,
      "grad_norm": 1.421875,
      "learning_rate": 0.00036931986015282056,
      "loss": 4.5078,
      "mean_token_accuracy": 0.26999998837709427,
      "num_tokens": 10724498.0,
      "step": 13450
    },
    {
      "entropy": 4.846611881256104,
      "epoch": 3.8040712468193383,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036922646371749355,
      "loss": 4.4974,
      "mean_token_accuracy": 0.2727766275405884,
      "num_tokens": 10728396.0,
      "step": 13455
    },
    {
      "entropy": 4.730855989456177,
      "epoch": 3.8054848741871643,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003691330475894184,
      "loss": 4.4402,
      "mean_token_accuracy": 0.26623533964157103,
      "num_tokens": 10732257.0,
      "step": 13460
    },
    {
      "entropy": 4.8069620609283445,
      "epoch": 3.80689850155499,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036903961178811836,
      "loss": 4.4392,
      "mean_token_accuracy": 0.26014135777950287,
      "num_tokens": 10736090.0,
      "step": 13465
    },
    {
      "entropy": 4.759947204589844,
      "epoch": 3.808312128922816,
      "grad_norm": 1.53125,
      "learning_rate": 0.000368946156333121,
      "loss": 4.4019,
      "mean_token_accuracy": 0.28341004550457,
      "num_tokens": 10739878.0,
      "step": 13470
    },
    {
      "entropy": 4.761424016952515,
      "epoch": 3.809725756290642,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036885268124395786,
      "loss": 4.4987,
      "mean_token_accuracy": 0.26073292046785357,
      "num_tokens": 10744053.0,
      "step": 13475
    },
    {
      "entropy": 4.768369245529175,
      "epoch": 3.8111393836584675,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003687591865401648,
      "loss": 4.3941,
      "mean_token_accuracy": 0.27157397717237475,
      "num_tokens": 10747987.0,
      "step": 13480
    },
    {
      "entropy": 4.787438583374024,
      "epoch": 3.8125530110262935,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003686656722412814,
      "loss": 4.472,
      "mean_token_accuracy": 0.2634075105190277,
      "num_tokens": 10752019.0,
      "step": 13485
    },
    {
      "entropy": 4.755287790298462,
      "epoch": 3.8139666383941195,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003685721383668519,
      "loss": 4.4374,
      "mean_token_accuracy": 0.28043030202388763,
      "num_tokens": 10756156.0,
      "step": 13490
    },
    {
      "entropy": 4.835629081726074,
      "epoch": 3.815380265761945,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000368478584936424,
      "loss": 4.41,
      "mean_token_accuracy": 0.27059364020824433,
      "num_tokens": 10759875.0,
      "step": 13495
    },
    {
      "entropy": 4.845592927932739,
      "epoch": 3.816793893129771,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003683850119695499,
      "loss": 4.56,
      "mean_token_accuracy": 0.2620421350002289,
      "num_tokens": 10763786.0,
      "step": 13500
    },
    {
      "entropy": 4.879301261901856,
      "epoch": 3.8182075204975967,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036829141948578576,
      "loss": 4.6383,
      "mean_token_accuracy": 0.24093610942363738,
      "num_tokens": 10767856.0,
      "step": 13505
    },
    {
      "entropy": 4.801182126998901,
      "epoch": 3.8196211478654227,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036819780750469187,
      "loss": 4.4809,
      "mean_token_accuracy": 0.2645676225423813,
      "num_tokens": 10772041.0,
      "step": 13510
    },
    {
      "entropy": 4.736063718795776,
      "epoch": 3.8210347752332483,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003681041760458326,
      "loss": 4.3713,
      "mean_token_accuracy": 0.27141694724559784,
      "num_tokens": 10775850.0,
      "step": 13515
    },
    {
      "entropy": 4.854523420333862,
      "epoch": 3.8224484026010743,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00036801052512877623,
      "loss": 4.5027,
      "mean_token_accuracy": 0.2583886578679085,
      "num_tokens": 10779710.0,
      "step": 13520
    },
    {
      "entropy": 4.840526580810547,
      "epoch": 3.8238620299689003,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003679168547730952,
      "loss": 4.5718,
      "mean_token_accuracy": 0.2531481921672821,
      "num_tokens": 10783627.0,
      "step": 13525
    },
    {
      "entropy": 4.928232669830322,
      "epoch": 3.825275657336726,
      "grad_norm": 1.46875,
      "learning_rate": 0.00036782316499836613,
      "loss": 4.5891,
      "mean_token_accuracy": 0.24622486680746078,
      "num_tokens": 10787675.0,
      "step": 13530
    },
    {
      "entropy": 4.811538505554199,
      "epoch": 3.826689284704552,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003677294558241696,
      "loss": 4.4524,
      "mean_token_accuracy": 0.2675464853644371,
      "num_tokens": 10791709.0,
      "step": 13535
    },
    {
      "entropy": 4.818176746368408,
      "epoch": 3.828102912072378,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036763572727009023,
      "loss": 4.4943,
      "mean_token_accuracy": 0.2675613135099411,
      "num_tokens": 10795848.0,
      "step": 13540
    },
    {
      "entropy": 4.791712093353271,
      "epoch": 3.8295165394402035,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003675419793557168,
      "loss": 4.4664,
      "mean_token_accuracy": 0.25773088485002515,
      "num_tokens": 10799662.0,
      "step": 13545
    },
    {
      "entropy": 4.805335235595703,
      "epoch": 3.8309301668080296,
      "grad_norm": 1.375,
      "learning_rate": 0.0003674482121006418,
      "loss": 4.4883,
      "mean_token_accuracy": 0.26792600452899934,
      "num_tokens": 10803968.0,
      "step": 13550
    },
    {
      "entropy": 4.803144216537476,
      "epoch": 3.832343794175855,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003673544255244622,
      "loss": 4.4762,
      "mean_token_accuracy": 0.2599541202187538,
      "num_tokens": 10807810.0,
      "step": 13555
    },
    {
      "entropy": 4.906888246536255,
      "epoch": 3.833757421543681,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036726061964677877,
      "loss": 4.622,
      "mean_token_accuracy": 0.2474090963602066,
      "num_tokens": 10811624.0,
      "step": 13560
    },
    {
      "entropy": 4.932873201370239,
      "epoch": 3.8351710489115067,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003671667944871963,
      "loss": 4.5661,
      "mean_token_accuracy": 0.2568101719021797,
      "num_tokens": 10815806.0,
      "step": 13565
    },
    {
      "entropy": 4.782061004638672,
      "epoch": 3.8365846762793328,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00036707295006532375,
      "loss": 4.3503,
      "mean_token_accuracy": 0.2701748639345169,
      "num_tokens": 10819942.0,
      "step": 13570
    },
    {
      "entropy": 4.708682727813721,
      "epoch": 3.8379983036471588,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003669790864007741,
      "loss": 4.3434,
      "mean_token_accuracy": 0.26618189215660093,
      "num_tokens": 10823975.0,
      "step": 13575
    },
    {
      "entropy": 4.756056547164917,
      "epoch": 3.8394119310149843,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003668852035131641,
      "loss": 4.3747,
      "mean_token_accuracy": 0.27381795197725295,
      "num_tokens": 10827729.0,
      "step": 13580
    },
    {
      "entropy": 4.8537229061126705,
      "epoch": 3.8408255583828104,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00036679130142211474,
      "loss": 4.5485,
      "mean_token_accuracy": 0.2569513007998466,
      "num_tokens": 10831710.0,
      "step": 13585
    },
    {
      "entropy": 4.748006772994995,
      "epoch": 3.8422391857506364,
      "grad_norm": 1.4375,
      "learning_rate": 0.000366697380147251,
      "loss": 4.3897,
      "mean_token_accuracy": 0.27698706537485124,
      "num_tokens": 10835626.0,
      "step": 13590
    },
    {
      "entropy": 4.827952909469604,
      "epoch": 3.843652813118462,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003666034397082019,
      "loss": 4.4369,
      "mean_token_accuracy": 0.2630734607577324,
      "num_tokens": 10839444.0,
      "step": 13595
    },
    {
      "entropy": 4.736324310302734,
      "epoch": 3.8450664404862875,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036650948012460034,
      "loss": 4.4516,
      "mean_token_accuracy": 0.2774525091052055,
      "num_tokens": 10843356.0,
      "step": 13600
    },
    {
      "entropy": 4.8972807884216305,
      "epoch": 3.8464800678541136,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00036641550141608333,
      "loss": 4.5564,
      "mean_token_accuracy": 0.24789429754018782,
      "num_tokens": 10847441.0,
      "step": 13605
    },
    {
      "entropy": 4.788968229293824,
      "epoch": 3.8478936952219396,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036632150360229183,
      "loss": 4.4748,
      "mean_token_accuracy": 0.2651695415377617,
      "num_tokens": 10851335.0,
      "step": 13610
    },
    {
      "entropy": 4.74798264503479,
      "epoch": 3.849307322589765,
      "grad_norm": 1.5,
      "learning_rate": 0.00036622748670287077,
      "loss": 4.4193,
      "mean_token_accuracy": 0.2696067735552788,
      "num_tokens": 10855358.0,
      "step": 13615
    },
    {
      "entropy": 4.828816890716553,
      "epoch": 3.850720949957591,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003661334507374691,
      "loss": 4.6102,
      "mean_token_accuracy": 0.25489101409912107,
      "num_tokens": 10859396.0,
      "step": 13620
    },
    {
      "entropy": 4.837277793884278,
      "epoch": 3.852134577325417,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036603939572573974,
      "loss": 4.5226,
      "mean_token_accuracy": 0.2559801012277603,
      "num_tokens": 10863471.0,
      "step": 13625
    },
    {
      "entropy": 4.798639631271362,
      "epoch": 3.8535482046932428,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003659453216873397,
      "loss": 4.3498,
      "mean_token_accuracy": 0.270724880695343,
      "num_tokens": 10867147.0,
      "step": 13630
    },
    {
      "entropy": 4.86679220199585,
      "epoch": 3.854961832061069,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036585122864192987,
      "loss": 4.5442,
      "mean_token_accuracy": 0.26415084451437,
      "num_tokens": 10871423.0,
      "step": 13635
    },
    {
      "entropy": 4.738937664031982,
      "epoch": 3.856375459428895,
      "grad_norm": 1.34375,
      "learning_rate": 0.000365757116609175,
      "loss": 4.3527,
      "mean_token_accuracy": 0.27562596648931503,
      "num_tokens": 10875526.0,
      "step": 13640
    },
    {
      "entropy": 4.7936749935150145,
      "epoch": 3.8577890867967204,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003656629856087439,
      "loss": 4.4817,
      "mean_token_accuracy": 0.2594189286231995,
      "num_tokens": 10879598.0,
      "step": 13645
    },
    {
      "entropy": 4.880722236633301,
      "epoch": 3.859202714164546,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003655688356603096,
      "loss": 4.4828,
      "mean_token_accuracy": 0.25045367479324343,
      "num_tokens": 10883536.0,
      "step": 13650
    },
    {
      "entropy": 4.866079425811767,
      "epoch": 3.860616341532372,
      "grad_norm": 1.46875,
      "learning_rate": 0.00036547466678354864,
      "loss": 4.4602,
      "mean_token_accuracy": 0.250050762295723,
      "num_tokens": 10887367.0,
      "step": 13655
    },
    {
      "entropy": 4.8325450897216795,
      "epoch": 3.862029968900198,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00036538047899814177,
      "loss": 4.5118,
      "mean_token_accuracy": 0.25915208011865615,
      "num_tokens": 10891398.0,
      "step": 13660
    },
    {
      "entropy": 4.8113929271698,
      "epoch": 3.8634435962680236,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036528627232377376,
      "loss": 4.5471,
      "mean_token_accuracy": 0.2600417956709862,
      "num_tokens": 10895602.0,
      "step": 13665
    },
    {
      "entropy": 4.820925617218018,
      "epoch": 3.8648572236358496,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003651920467801331,
      "loss": 4.4442,
      "mean_token_accuracy": 0.2677819117903709,
      "num_tokens": 10899647.0,
      "step": 13670
    },
    {
      "entropy": 4.833693885803223,
      "epoch": 3.8662708510036756,
      "grad_norm": 1.484375,
      "learning_rate": 0.00036509780238691244,
      "loss": 4.5296,
      "mean_token_accuracy": 0.25681918263435366,
      "num_tokens": 10903697.0,
      "step": 13675
    },
    {
      "entropy": 4.871029424667358,
      "epoch": 3.867684478371501,
      "grad_norm": 1.5,
      "learning_rate": 0.00036500353916380814,
      "loss": 4.5116,
      "mean_token_accuracy": 0.2642779976129532,
      "num_tokens": 10907797.0,
      "step": 13680
    },
    {
      "entropy": 4.821073913574219,
      "epoch": 3.869098105739327,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003649092571305208,
      "loss": 4.3657,
      "mean_token_accuracy": 0.27732581049203875,
      "num_tokens": 10911461.0,
      "step": 13685
    },
    {
      "entropy": 4.694756555557251,
      "epoch": 3.870511733107153,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00036481495630675463,
      "loss": 4.365,
      "mean_token_accuracy": 0.27655056267976763,
      "num_tokens": 10915160.0,
      "step": 13690
    },
    {
      "entropy": 4.792196416854859,
      "epoch": 3.871925360474979,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003647206367122181,
      "loss": 4.4495,
      "mean_token_accuracy": 0.27043226957321165,
      "num_tokens": 10919317.0,
      "step": 13695
    },
    {
      "entropy": 4.805445146560669,
      "epoch": 3.8733389878428044,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00036462629836662324,
      "loss": 4.3971,
      "mean_token_accuracy": 0.262938466668129,
      "num_tokens": 10923042.0,
      "step": 13700
    },
    {
      "entropy": 4.885599851608276,
      "epoch": 3.8747526152106304,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003645319412896863,
      "loss": 4.5399,
      "mean_token_accuracy": 0.25402490198612215,
      "num_tokens": 10927169.0,
      "step": 13705
    },
    {
      "entropy": 4.786007690429687,
      "epoch": 3.8761662425784564,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003644375655011272,
      "loss": 4.3822,
      "mean_token_accuracy": 0.27436971217393874,
      "num_tokens": 10931332.0,
      "step": 13710
    },
    {
      "entropy": 4.864028692245483,
      "epoch": 3.877579869946282,
      "grad_norm": 1.5,
      "learning_rate": 0.00036434317102067,
      "loss": 4.5798,
      "mean_token_accuracy": 0.2554190158843994,
      "num_tokens": 10935179.0,
      "step": 13715
    },
    {
      "entropy": 4.898044013977051,
      "epoch": 3.878993497314108,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003642487578680426,
      "loss": 4.6227,
      "mean_token_accuracy": 0.25033741891384126,
      "num_tokens": 10939205.0,
      "step": 13720
    },
    {
      "entropy": 4.8285990238189695,
      "epoch": 3.880407124681934,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003641543260629766,
      "loss": 4.4792,
      "mean_token_accuracy": 0.26051670163869856,
      "num_tokens": 10943355.0,
      "step": 13725
    },
    {
      "entropy": 4.800533390045166,
      "epoch": 3.8818207520497596,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003640598756252079,
      "loss": 4.4353,
      "mean_token_accuracy": 0.2625692754983902,
      "num_tokens": 10947522.0,
      "step": 13730
    },
    {
      "entropy": 4.810432720184326,
      "epoch": 3.8832343794175856,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036396540657447586,
      "loss": 4.4539,
      "mean_token_accuracy": 0.2724036112427711,
      "num_tokens": 10951639.0,
      "step": 13735
    },
    {
      "entropy": 4.791786003112793,
      "epoch": 3.8846480067854112,
      "grad_norm": 1.375,
      "learning_rate": 0.0003638709189305239,
      "loss": 4.5028,
      "mean_token_accuracy": 0.27102805078029635,
      "num_tokens": 10955874.0,
      "step": 13740
    },
    {
      "entropy": 4.8527891635894775,
      "epoch": 3.8860616341532372,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003637764127130996,
      "loss": 4.4623,
      "mean_token_accuracy": 0.2685776650905609,
      "num_tokens": 10959397.0,
      "step": 13745
    },
    {
      "entropy": 4.879961729049683,
      "epoch": 3.887475261521063,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003636818879419539,
      "loss": 4.4892,
      "mean_token_accuracy": 0.2593619778752327,
      "num_tokens": 10963260.0,
      "step": 13750
    },
    {
      "entropy": 4.841781711578369,
      "epoch": 3.888888888888889,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00036358734463684203,
      "loss": 4.4835,
      "mean_token_accuracy": 0.2633384957909584,
      "num_tokens": 10967343.0,
      "step": 13755
    },
    {
      "entropy": 4.790372371673584,
      "epoch": 3.890302516256715,
      "grad_norm": 1.296875,
      "learning_rate": 0.000363492782817523,
      "loss": 4.3795,
      "mean_token_accuracy": 0.27607927173376084,
      "num_tokens": 10971560.0,
      "step": 13760
    },
    {
      "entropy": 4.859266710281372,
      "epoch": 3.8917161436245404,
      "grad_norm": 1.390625,
      "learning_rate": 0.00036339820250375955,
      "loss": 4.5866,
      "mean_token_accuracy": 0.25819905549287797,
      "num_tokens": 10975785.0,
      "step": 13765
    },
    {
      "entropy": 4.839828681945801,
      "epoch": 3.8931297709923665,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003633036037153184,
      "loss": 4.43,
      "mean_token_accuracy": 0.26197829395532607,
      "num_tokens": 10980009.0,
      "step": 13770
    },
    {
      "entropy": 4.859458541870117,
      "epoch": 3.8945433983601925,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00036320898647197015,
      "loss": 4.4854,
      "mean_token_accuracy": 0.2547484651207924,
      "num_tokens": 10983661.0,
      "step": 13775
    },
    {
      "entropy": 4.804533290863037,
      "epoch": 3.895957025728018,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003631143507934892,
      "loss": 4.608,
      "mean_token_accuracy": 0.24772416651248932,
      "num_tokens": 10987580.0,
      "step": 13780
    },
    {
      "entropy": 4.7840879440307615,
      "epoch": 3.897370653095844,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00036301969669965377,
      "loss": 4.4526,
      "mean_token_accuracy": 0.2668111428618431,
      "num_tokens": 10991294.0,
      "step": 13785
    },
    {
      "entropy": 4.7662702083587645,
      "epoch": 3.8987842804636696,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003629250242102461,
      "loss": 4.4026,
      "mean_token_accuracy": 0.26514937430620195,
      "num_tokens": 10995084.0,
      "step": 13790
    },
    {
      "entropy": 4.829332256317139,
      "epoch": 3.9001979078314957,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000362830333345052,
      "loss": 4.5066,
      "mean_token_accuracy": 0.26717615127563477,
      "num_tokens": 10999137.0,
      "step": 13795
    },
    {
      "entropy": 4.8667449951171875,
      "epoch": 3.9016115351993212,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003627356241238614,
      "loss": 4.6613,
      "mean_token_accuracy": 0.24295440018177034,
      "num_tokens": 11003111.0,
      "step": 13800
    },
    {
      "entropy": 4.920505523681641,
      "epoch": 3.9030251625671473,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036264089656646794,
      "loss": 4.5522,
      "mean_token_accuracy": 0.2564935952425003,
      "num_tokens": 11007150.0,
      "step": 13805
    },
    {
      "entropy": 4.878567743301391,
      "epoch": 3.9044387899349733,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000362546150692669,
      "loss": 4.5156,
      "mean_token_accuracy": 0.27196300178766253,
      "num_tokens": 11011001.0,
      "step": 13810
    },
    {
      "entropy": 4.804516363143921,
      "epoch": 3.905852417302799,
      "grad_norm": 1.578125,
      "learning_rate": 0.00036245138652226587,
      "loss": 4.3789,
      "mean_token_accuracy": 0.26936547458171844,
      "num_tokens": 11015087.0,
      "step": 13815
    },
    {
      "entropy": 4.805682706832886,
      "epoch": 3.907266044670625,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036235660407506393,
      "loss": 4.4158,
      "mean_token_accuracy": 0.2760919615626335,
      "num_tokens": 11018983.0,
      "step": 13820
    },
    {
      "entropy": 4.7615966796875,
      "epoch": 3.908679672038451,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003622618033708718,
      "loss": 4.3868,
      "mean_token_accuracy": 0.26692229956388475,
      "num_tokens": 11022680.0,
      "step": 13825
    },
    {
      "entropy": 4.709288835525513,
      "epoch": 3.9100932994062765,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036216698442950235,
      "loss": 4.4139,
      "mean_token_accuracy": 0.26490636318922045,
      "num_tokens": 11026485.0,
      "step": 13830
    },
    {
      "entropy": 4.791393184661866,
      "epoch": 3.9115069267741025,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036207214727077225,
      "loss": 4.5272,
      "mean_token_accuracy": 0.2621482312679291,
      "num_tokens": 11030481.0,
      "step": 13835
    },
    {
      "entropy": 4.918068170547485,
      "epoch": 3.912920554141928,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003619772919145017,
      "loss": 4.6239,
      "mean_token_accuracy": 0.255429182946682,
      "num_tokens": 11034362.0,
      "step": 13840
    },
    {
      "entropy": 4.825625896453857,
      "epoch": 3.914334181509754,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003618824183805152,
      "loss": 4.4753,
      "mean_token_accuracy": 0.26802953630685805,
      "num_tokens": 11038481.0,
      "step": 13845
    },
    {
      "entropy": 4.893768692016602,
      "epoch": 3.9157478088775797,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00036178752668864037,
      "loss": 4.5147,
      "mean_token_accuracy": 0.2603351652622223,
      "num_tokens": 11042380.0,
      "step": 13850
    },
    {
      "entropy": 4.845927906036377,
      "epoch": 3.9171614362454057,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003616926168587092,
      "loss": 4.4244,
      "mean_token_accuracy": 0.26562248915433884,
      "num_tokens": 11046427.0,
      "step": 13855
    },
    {
      "entropy": 4.866923999786377,
      "epoch": 3.9185750636132317,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003615976889105571,
      "loss": 4.4735,
      "mean_token_accuracy": 0.2799953565001488,
      "num_tokens": 11050420.0,
      "step": 13860
    },
    {
      "entropy": 4.865629625320435,
      "epoch": 3.9199886909810573,
      "grad_norm": 1.453125,
      "learning_rate": 0.00036150274286402364,
      "loss": 4.551,
      "mean_token_accuracy": 0.2640145689249039,
      "num_tokens": 11054155.0,
      "step": 13865
    },
    {
      "entropy": 4.805466079711914,
      "epoch": 3.9214023183488833,
      "grad_norm": 1.515625,
      "learning_rate": 0.00036140777873895175,
      "loss": 4.4315,
      "mean_token_accuracy": 0.27698596715927126,
      "num_tokens": 11057994.0,
      "step": 13870
    },
    {
      "entropy": 4.7713275909423825,
      "epoch": 3.9228159457167093,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003613127965551884,
      "loss": 4.44,
      "mean_token_accuracy": 0.2728674918413162,
      "num_tokens": 11062093.0,
      "step": 13875
    },
    {
      "entropy": 4.7780656814575195,
      "epoch": 3.924229573084535,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036121779633258445,
      "loss": 4.5392,
      "mean_token_accuracy": 0.26126407235860827,
      "num_tokens": 11066000.0,
      "step": 13880
    },
    {
      "entropy": 4.7724151611328125,
      "epoch": 3.9256432004523605,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003611227780909942,
      "loss": 4.4025,
      "mean_token_accuracy": 0.2666705295443535,
      "num_tokens": 11069585.0,
      "step": 13885
    },
    {
      "entropy": 4.808309507369995,
      "epoch": 3.9270568278201865,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003610277418502758,
      "loss": 4.4385,
      "mean_token_accuracy": 0.26364396810531615,
      "num_tokens": 11073645.0,
      "step": 13890
    },
    {
      "entropy": 4.819502687454223,
      "epoch": 3.9284704551880125,
      "grad_norm": 1.375,
      "learning_rate": 0.0003609326876302914,
      "loss": 4.4618,
      "mean_token_accuracy": 0.2683993712067604,
      "num_tokens": 11077831.0,
      "step": 13895
    },
    {
      "entropy": 4.812364292144776,
      "epoch": 3.929884082555838,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00036083761545090663,
      "loss": 4.5326,
      "mean_token_accuracy": 0.24616265296936035,
      "num_tokens": 11081745.0,
      "step": 13900
    },
    {
      "entropy": 4.885935592651367,
      "epoch": 3.931297709923664,
      "grad_norm": 1.5,
      "learning_rate": 0.00036074252533199107,
      "loss": 4.5613,
      "mean_token_accuracy": 0.26064191609621046,
      "num_tokens": 11085719.0,
      "step": 13905
    },
    {
      "entropy": 4.907442665100097,
      "epoch": 3.93271133729149,
      "grad_norm": 1.5,
      "learning_rate": 0.00036064741729341794,
      "loss": 4.4885,
      "mean_token_accuracy": 0.2603288695216179,
      "num_tokens": 11089657.0,
      "step": 13910
    },
    {
      "entropy": 4.851953029632568,
      "epoch": 3.9341249646593157,
      "grad_norm": 1.4375,
      "learning_rate": 0.00036055229135506425,
      "loss": 4.4029,
      "mean_token_accuracy": 0.2713371947407722,
      "num_tokens": 11093535.0,
      "step": 13915
    },
    {
      "entropy": 4.868658494949341,
      "epoch": 3.9355385920271417,
      "grad_norm": 1.40625,
      "learning_rate": 0.00036045714753681063,
      "loss": 4.5108,
      "mean_token_accuracy": 0.2610654100775719,
      "num_tokens": 11097720.0,
      "step": 13920
    },
    {
      "entropy": 4.775626087188721,
      "epoch": 3.9369522193949678,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003603619858585417,
      "loss": 4.5409,
      "mean_token_accuracy": 0.262907849252224,
      "num_tokens": 11101852.0,
      "step": 13925
    },
    {
      "entropy": 4.775410413742065,
      "epoch": 3.9383658467627933,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003602668063401455,
      "loss": 4.366,
      "mean_token_accuracy": 0.28150447458028793,
      "num_tokens": 11105769.0,
      "step": 13930
    },
    {
      "entropy": 4.842332029342652,
      "epoch": 3.939779474130619,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003601716090015141,
      "loss": 4.5276,
      "mean_token_accuracy": 0.26394058763980865,
      "num_tokens": 11109650.0,
      "step": 13935
    },
    {
      "entropy": 4.847542810440063,
      "epoch": 3.941193101498445,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003600763938625431,
      "loss": 4.4751,
      "mean_token_accuracy": 0.26261531859636306,
      "num_tokens": 11113672.0,
      "step": 13940
    },
    {
      "entropy": 4.738144493103027,
      "epoch": 3.942606728866271,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00035998116094313186,
      "loss": 4.4394,
      "mean_token_accuracy": 0.2708406805992126,
      "num_tokens": 11117589.0,
      "step": 13945
    },
    {
      "entropy": 4.807592391967773,
      "epoch": 3.9440203562340965,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00035988591026318357,
      "loss": 4.4235,
      "mean_token_accuracy": 0.27019968926906585,
      "num_tokens": 11121311.0,
      "step": 13950
    },
    {
      "entropy": 4.845996570587158,
      "epoch": 3.9454339836019225,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003597906418426049,
      "loss": 4.4806,
      "mean_token_accuracy": 0.26426031440496445,
      "num_tokens": 11125087.0,
      "step": 13955
    },
    {
      "entropy": 4.800911235809326,
      "epoch": 3.9468476109697486,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003596953557013064,
      "loss": 4.3939,
      "mean_token_accuracy": 0.26962524503469465,
      "num_tokens": 11129020.0,
      "step": 13960
    },
    {
      "entropy": 4.8140209197998045,
      "epoch": 3.948261238337574,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00035960005185920233,
      "loss": 4.4381,
      "mean_token_accuracy": 0.2574431225657463,
      "num_tokens": 11133082.0,
      "step": 13965
    },
    {
      "entropy": 4.7580437660217285,
      "epoch": 3.9496748657054,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003595047303362105,
      "loss": 4.46,
      "mean_token_accuracy": 0.25610816180706025,
      "num_tokens": 11136863.0,
      "step": 13970
    },
    {
      "entropy": 4.859652662277222,
      "epoch": 3.9510884930732257,
      "grad_norm": 1.46875,
      "learning_rate": 0.00035940939115225266,
      "loss": 4.494,
      "mean_token_accuracy": 0.2655857652425766,
      "num_tokens": 11140824.0,
      "step": 13975
    },
    {
      "entropy": 4.767376184463501,
      "epoch": 3.9525021204410518,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00035931403432725405,
      "loss": 4.422,
      "mean_token_accuracy": 0.2671900987625122,
      "num_tokens": 11144749.0,
      "step": 13980
    },
    {
      "entropy": 4.762121629714966,
      "epoch": 3.9539157478088773,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003592186598811436,
      "loss": 4.5015,
      "mean_token_accuracy": 0.2660601630806923,
      "num_tokens": 11148755.0,
      "step": 13985
    },
    {
      "entropy": 4.7226996421813965,
      "epoch": 3.9553293751767034,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003591232678338541,
      "loss": 4.3491,
      "mean_token_accuracy": 0.2689395651221275,
      "num_tokens": 11152540.0,
      "step": 13990
    },
    {
      "entropy": 4.862719488143921,
      "epoch": 3.9567430025445294,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00035902785820532174,
      "loss": 4.5109,
      "mean_token_accuracy": 0.2685216322541237,
      "num_tokens": 11156358.0,
      "step": 13995
    },
    {
      "entropy": 4.841523027420044,
      "epoch": 3.958156629912355,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00035893243101548666,
      "loss": 4.4819,
      "mean_token_accuracy": 0.2607298955321312,
      "num_tokens": 11160395.0,
      "step": 14000
    },
    {
      "epoch": 3.958156629912355,
      "eval_entropy": 4.590682546776462,
      "eval_loss": 4.9608869552612305,
      "eval_mean_token_accuracy": 0.2509288435642133,
      "eval_num_tokens": 11160395.0,
      "eval_runtime": 4.7142,
      "eval_samples_per_second": 2393.613,
      "eval_steps_per_second": 299.308,
      "step": 14000
    }
  ],
  "logging_steps": 5,
  "max_steps": 35360,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2008001764270080.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}